1 line
486 KiB
JSON
1 line
486 KiB
JSON
{"last_node_id": 51, "last_link_id": 48, "nodes": [{"id": 4, "type": "LoraLoaderModelOnly", "pos": [4740, -240], "size": {"0": 315, "1": 82}, "flags": {}, "order": 23, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": 1}], "outputs": [{"name": "MODEL", "type": "MODEL", "links": [2], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "LoraLoaderModelOnly"}, "widgets_values": ["ip-adapter-faceid-plusv2_sd15_lora.safetensors", 1]}, {"id": 9, "type": "Anything Everywhere", "pos": [4800, -60], "size": {"0": 240, "1": 60}, "flags": {}, "order": 17, "mode": 0, "inputs": [{"name": "IPADAPTER", "type": "*", "link": 6, "color_on": ""}], "title": "IPAadapter Everywhere", "properties": {"group_restricted": false, "color_restricted": false, "Node name for S&R": "Anything Everywhere"}, "widgets_values": ["{'image_proj': {'proj.0.weight': tensor([[ 0.0247, -0.0113, -0.0276, ..., -0.0316, 0.0359, -0.0495],\n [ 0.0196, 0.0128, -0.0555, ..., 0.0481, -0.0486, -0.0173],\n [-0.0402, 0.0024, -0.0332, ..., -0.0943, 0.0126, 0.0450],\n ...,\n [-0.1112, -0.0075, 0.0141, ..., 0.0582, 0.0253, 0.0004],\n [ 0.0481, -0.0420, 0.0203, ..., 0.1025, -0.0098, -0.0918],\n [-0.0129, -0.0557, 0.0528, ..., -0.0301, -0.0118, -0.0273]], dtype=torch.float16), 'proj.0.bias': tensor([ 0.0098, -0.0213, 0.0003, ..., 0.0134, -0.0158, -0.0216], dtype=torch.float16), 'proj.2.weight': tensor([[ 0.0236, 0.0237, -0.0525, ..., 0.0217, 0.0224, 0.0496],\n [ 0.0004, -0.0344, -0.0250, ..., -0.0364, -0.0165, -0.0207],\n [ 0.0125, 0.0262, 0.0002, ..., -0.0206, 0.0230, 0.0089],\n ...,\n [ 0.0092, -0.0508, -0.0059, ..., -0.0140, 0.0518, -0.0222],\n [ 0.0687, -0.0341, -0.0788, ..., -0.0443, -0.0566, -0.0299],\n [-0.0472, -0.0559, -0.0098, ..., -0.0134, 0.0651, -0.0255]], dtype=torch.float16), 'proj.2.bias': tensor([-0.0150, 0.0126, -0.0102, ..., -0.0150, 0.0083, 0.0208], dtype=torch.float16), 'norm.weight': tensor([0.6450, 0.5938, 0.5840, 0.6499, 0.6270, 0.6772, 0.6016, 0.5483, 0.6309, 0.6528, 0.6343, 0.6240, 0.6294, 0.6543, 0.6514, 0.6289, 0.6191, 0.6479, 0.5977, 0.5635, 0.6758, 0.6431, 0.6353, 0.6274, 0.6206, 0.6523, 0.6318, 0.5938, 0.6533, 0.6318, 0.5913, 0.5874, 0.5898, 0.6240, 0.6313, 0.6963, 0.6919, 0.6519, 0.6045,\n 0.6396, 0.6255, 0.6655, 0.6201, 0.6035, 0.6045, 0.6748, 0.6162, 0.5972, 0.5972, 0.6621, 0.6167, 0.6099, 0.6440, 0.5977, 0.6499, 0.6377, 0.5728, 0.6733, 0.6641, 0.6318, 0.6372, 0.6162, 0.6719, 0.7002, 0.6284, 0.6650, 0.6323, 0.6382, 0.6064, 0.5757, 0.6343, 0.6597, 0.6577, 0.6968, 0.6533, 0.6772, 0.6201, 0.5864,\n 0.6465, 0.6021, 0.6401, 0.5786, 0.6597, 0.6514, 0.6416, 0.6777, 0.6714, 0.6211, 0.6685, 0.5898, 0.6484, 0.5908, 0.5928, 0.6196, 0.6553, 0.6035, 0.6416, 0.5732, 0.6865, 0.6670, 0.6338, 0.6094, 0.6123, 0.6846, 0.6533, 0.6313, 0.6436, 0.6348, 0.5962, 0.6372, 0.5864, 0.6362, 0.6704, 0.6177, 0.6304, 0.6553, 0.6016,\n 0.6538, 0.6294, 0.5972, 0.6255, 0.5913, 0.6235, 0.6060, 0.6274, 0.6377, 0.5723, 0.6313, 0.6582, 0.6187, 0.6069, 0.5952, 0.6787, 0.6562, 0.6572, 0.6377, 0.6440, 0.6787, 0.6211, 0.6191, 0.6123, 0.6230, 0.6450, 0.6123, 0.6260, 0.6465, 0.6191, 0.6416, 0.6675, 0.6538, 0.6763, 0.6548, 0.5933, 0.6118, 0.5825, 0.7354,\n 0.6597, 0.6421, 0.6479, 0.6318, 0.7026, 0.5835, 0.6416, 0.4976, 0.5898, 0.5796, 0.5767, 0.5928, 0.6465, 0.6699, 0.6426, 0.6411, 0.6250, 0.6167, 0.6113, 0.6172, 0.6055, 0.6255, 0.6147, 0.5996, 0.6514, 0.6631, 0.6265, 0.5942, 0.6694, 0.6323, 0.6440, 0.5884, 0.5908, 0.6328, 0.6289, 0.6284, 0.6245, 0.6147, 0.6045,\n 0.5718, 0.5801, 0.5923, 0.5581, 0.6782, 0.5859, 0.6519, 0.6274, 0.6606, 0.6465, 0.6562, 0.6182, 0.6602, 0.6621, 0.6870, 0.5957, 0.2013, 0.6235, 0.6294, 0.2688, 0.6567, 0.6680, 0.6211, 0.6494, 0.6426, 0.6079, 0.5938, 0.6396, 0.6660, 0.6562, 0.6348, 0.6968, 0.5859, 0.6372, 0.5840, 0.6069, 0.6284, 0.6504, 0.6660,\n 0.6187, 0.5776, 0.6348, 0.6680, 0.7046, 0.6221, 0.6450, 0.5078, 0.6250, 0.6318, 0.6567, 0.6494, 0.6157, 0.6260, 0.6318, 0.6455, 0.5732, 0.6675, 0.6030, 0.6191, 0.4446, 0.6265, 0.6348, 0.6543, 0.5151, 0.6426, 0.6055, 0.0934, 0.6284, 0.6123, 0.6523, 0.4749, 0.6353, 0.6558, 0.6167, 0.6333, 0.5962, 0.6143, 0.6372,\n 0.6343, 0.6318, 0.6743, 0.6167, 0.6367, 0.6221, 0.6104, 0.6313, 0.5684, 0.6641, 0.6440, 0.5435, 0.6206, 0.6514, 0.6260, 0.5957, 0.6245, 0.6279, 0.6040, 0.6025, 0.6528, 0.6230, 0.6309, 0.6318, 0.6504, 0.6191, 0.6523, 0.6240, 0.6611, 0.6016, 0.2769, 0.6182, 0.6421, 0.6567, 0.6172, 0.6543, 0.5649, 0.5918, 0.2294,\n 0.5454, 0.6299, 0.6855, 0.6250, 0.6250, 0.6670, 0.6587, 0.6177, 0.6299, 0.6260, 0.6040, 0.6108, 0.7383, 0.5952, 0.6699, 0.5708, 0.6460, 0.5986, 0.2505, 0.6401, 0.6455, 0.6196, 0.6309, 0.6338, 0.6196, 0.5679, 0.6323, 0.6196, 0.6001, 0.6348, 0.6050, 0.6421, 0.6436, 0.5903, 0.6284, 0.6387, 0.6655, 0.6221, 0.6680,\n 0.6719, 0.6260, 0.6460, 0.6416, 0.6504, 0.6196, 0.6138, 0.6665, 0.6689, 0.6235, 0.6338, 0.6636, 0.6108, 0.6470, 0.4399, 0.6675, 0.6152, 0.6016, 0.6274, 0.5029, 0.6235, 0.5825, 0.6201, 0.6519, 0.5840, 0.6250, 0.5303, 0.6084, 0.6006, 0.6152, 0.6079, 0.6265, 0.6704, 0.6196, 0.6665, 0.6538, 0.5850, 0.6079, 0.6479,\n 0.6353, 0.6299, 0.6099, 0.5791, 0.6494, 0.5884, 0.6577, 0.6465, 0.6714, 0.6621, 0.5635, 0.6167, 0.6128, 0.6675, 0.5928, 0.6206, 0.6602, 0.5859, 0.6660, 0.6567, 0.6436, 0.5259, 0.6030, 0.6392, 0.2898, 0.4231, 0.6108, 0.6729, 0.6392, 0.6367, 0.6562, 0.6299, 0.6123, 0.6509, 0.6182, 0.6216, 0.6890, 0.6460, 0.6401,\n 0.6665, 0.5884, 0.6621, 0.6284, 0.6577, 0.6821, 0.6499, 0.6470, 0.6221, 0.6846, 0.6479, 0.6147, 0.6274, 0.6011, 0.7974, 0.6226, 0.6333, 0.6157, 0.6626, 0.6274, 0.6079, 0.6006, 0.6934, 0.6558, 0.6060, 0.6172, 0.6460, 0.5278, 0.6255, 0.6353, 0.6260, 0.7095, 0.6118, 0.6724, 0.6382, 0.6309, 0.6201, 0.6157, 0.6240,\n 0.6367, 0.6973, 0.6489, 0.5947, 0.6475, 0.6196, 0.5664, 0.6704, 0.6245, 0.6489, 0.6289, 0.6816, 0.6187, 0.6255, 0.6201, 0.7012, 0.6147, 0.6675, 0.6206, 0.6689, 0.6118, 0.6836, 0.5679, 0.6001, 0.6445, 0.5757, 0.6025, 0.6396, 0.6367, 0.7212, 0.5894, 0.6812, 0.6055, 0.5938, 0.6387, 0.5801, 0.6367, 0.6548, 0.6455,\n 0.5967, 0.6118, 0.6299, 0.6426, 0.6338, 0.6606, 0.6626, 0.6304, 0.6465, 0.6128, 0.6572, 0.5737, 0.6670, 0.5933, 0.5771, 0.2396, 0.6636, 0.6030, 0.6230, 0.5776, 0.6274, 0.6406, 0.5889, 0.6250, 0.0776, 0.5864, 0.6445, 0.6240, 0.6587, 0.5874, 0.5996, 0.5820, 0.6387, 0.6372, 0.6177, 0.6338, 0.5898, 0.6270, 0.5762,\n 0.6411, 0.6265, 0.5742, 0.6582, 0.6265, 0.6357, 0.5933, 0.6558, 0.6831, 0.6494, 0.6377, 0.6392, 0.6758, 0.6206, 0.6011, 0.6069, 0.6025, 0.6470, 0.6377, 0.6641, 0.6875, 0.5830, 0.5879, 0.5898, 0.6440, 0.6240, 0.6230, 0.6494, 0.6533, 0.6255, 0.6255, 0.6387, 0.6318, 0.5576, 0.6636, 0.6309, 0.6089, 0.6851, 0.6328,\n 0.5171, 0.6377, 0.6172, 0.6289, 0.6655, 0.6069, 0.6167, 0.6880, 0.6182, 0.6411, 0.6108, 0.6006, 0.6475, 0.6084, 0.6719, 0.6646, 0.5781, 0.6089, 0.5977, 0.6440, 0.5830, 0.6074, 0.5962, 0.6152, 0.6196, 0.5498, 0.5898, 0.6099, 0.5645, 0.5864, 0.6602, 0.6406, 0.5640, 0.5698, 0.5845, 0.5840, 0.6367, 0.5801, 0.6382,\n 0.6401, 0.5854, 0.6709, 0.6006, 0.6440, 0.6406, 0.6382, 0.6694, 0.6396, 0.5444, 0.5898, 0.6655, 0.6055, 0.6431, 0.6577, 0.6279, 0.5435, 0.6309, 0.6660, 0.6143, 0.6514, 0.6040, 0.5854, 0.6367, 0.5752, 0.6084, 0.6704, 0.6460, 0.6348, 0.7100, 0.6357, 0.6284, 0.6143, 0.5649, 0.5679, 0.6758, 0.6011, 0.6260, 0.5981,\n 0.6401, 0.6304, 0.6392, 0.5771, 0.6274, 0.5762, 0.6011, 0.6655, 0.6641, 0.6177, 0.5947, 0.6113, 0.5991, 0.6167, 0.6138, 0.6338, 0.6069, 0.6104, 0.5474, 0.6929, 0.6460, 0.6104, 0.6406, 0.5967, 0.5723, 0.6323, 0.6572, 0.6475, 0.6343, 0.5957, 0.6328, 0.6245, 0.6592, 0.6348, 0.6201, 0.6929, 0.6113, 0.6670, 0.5581,\n 0.6538, 0.6328, 0.6094, 0.6406, 0.6807, 0.6528, 0.6079, 0.6523, 0.6450, 0.6255, 0.6323, 0.6753, 0.5786, 0.6318, 0.5708, 0.6226, 0.6567, 0.6426, 0.6289, 0.5693, 0.6025, 0.5757, 0.6255, 0.5884, 0.6021, 0.6182, 0.5854, 0.5938, 0.6636, 0.6704, 0.6416, 0.6909, 0.6299, 0.6177, 0.6250, 0.5996, 0.6606, 0.6636, 0.5537,\n 0.6357, 0.5840, 0.6392, 0.5996, 0.6035, 0.6045, 0.6284, 0.6416, 0.6421, 0.6455, 0.6201, 0.6133, 0.6060, 0.6846, 0.6514, 0.6074, 0.6328, 0.6318, 0.6011, 0.5859, 0.6157, 0.6367, 0.6465, 0.6191, 0.2517, 0.6509, 0.6450], dtype=torch.float16), 'norm.bias': tensor([ 7.3385e-04, -2.4242e-03, -8.8959e-03, -3.4088e-02, 1.8215e-03, 2.9874e-04, 2.1896e-02, -2.7725e-02, 5.9357e-03, 1.4084e-02, -5.0049e-02, 6.2294e-03, -5.0049e-02, -5.3902e-03, 6.3324e-03, 7.6246e-04, 1.0086e-02, -4.3854e-02, -8.3847e-03, 4.2206e-02, -4.5280e-03, -5.0201e-02, 2.9053e-02, -7.8964e-03,\n 3.9581e-02, 8.9111e-03, -1.1581e-02, -1.2329e-02, -5.8563e-02, 5.0774e-03, 2.4017e-02, -2.1667e-02, 4.4617e-02, 1.3908e-02, 1.4877e-02, -2.9648e-02, 3.8086e-02, -3.5980e-02, -6.4182e-04, 5.3978e-03, -1.2161e-02, 4.5563e-02, 2.4841e-02, 7.1678e-03, -2.0355e-02, -1.8555e-02, 2.4673e-02, -3.2318e-02,\n -6.1188e-03, 4.5197e-02, -1.9455e-02, -2.1957e-02, -8.5831e-03, -1.0017e-02, -1.5915e-02, 1.0300e-02, -2.2392e-03, 1.7548e-02, -3.1006e-02, -1.7975e-02, 2.4628e-02, 5.7831e-03, -5.5389e-02, 2.2018e-02, 7.7820e-03, -1.6357e-02, 1.8829e-02, -2.0523e-02, -9.7656e-03, -2.5589e-02, 1.6830e-02, -1.6327e-02,\n -2.5005e-03, 6.9580e-03, 1.5350e-02, -2.2064e-02, 4.5410e-02, -2.4090e-03, -1.3710e-02, 1.6220e-02, -4.5288e-02, -1.3714e-03, -1.0773e-02, 1.8616e-02, 1.5060e-02, 2.9510e-02, 3.3569e-02, -1.9638e-02, 1.0300e-02, 1.1742e-02, -7.0992e-03, -2.1484e-02, -1.2825e-02, -4.3091e-02, 1.0918e-02, 2.3804e-02,\n 1.3885e-02, -9.6359e-03, -1.9684e-02, 1.6098e-02, 7.5798e-03, -8.9340e-03, -1.0956e-02, -4.6959e-03, -3.7689e-02, 3.0991e-02, -6.4163e-03, -7.4387e-03, -7.7934e-03, -4.3640e-02, -2.4445e-02, -1.8677e-02, 5.5008e-03, -1.4877e-02, -2.7481e-02, 1.1253e-02, -9.3002e-03, 2.1500e-02, -4.1962e-03, -5.9547e-03,\n 2.0569e-02, 3.2715e-02, 1.4420e-02, 6.5727e-03, 4.1718e-02, -1.8372e-02, -2.0599e-02, -7.7744e-03, -4.9469e-02, 8.6212e-03, 1.3924e-02, -2.5665e-02, 2.5528e-02, 3.1185e-03, 6.8703e-03, -7.3318e-03, 2.2316e-03, -6.1340e-02, -3.1311e-02, 6.4774e-03, -1.4595e-02, -2.3193e-02, 7.3090e-03, -3.9185e-02,\n 3.8013e-03, 1.0773e-02, 2.4033e-03, -1.2672e-02, -5.0316e-03, -3.5305e-03, -9.2850e-03, -1.7700e-02, 2.5955e-02, -3.4760e-02, -2.3529e-02, 3.9978e-02, -1.2146e-02, -4.6387e-03, -5.1086e-02, -6.5079e-03, -5.7983e-02, 6.9237e-04, 2.0493e-02, 7.5989e-03, -1.3359e-02, 2.0813e-02, 1.6022e-02, -1.1543e-02,\n -3.4149e-02, -3.3844e-02, 4.6539e-03, -2.7573e-02, -1.7471e-02, -4.1504e-02, -5.9204e-03, -1.0704e-02, 3.1433e-02, -1.5625e-02, -2.9282e-02, 3.2593e-02, -9.8953e-03, -2.8061e-02, 2.6947e-02, -2.1271e-02, -3.7537e-02, -2.0462e-02, -6.3599e-02, 9.4238e-02, -1.2108e-02, 1.2245e-02, 5.6152e-03, 5.0926e-03,\n -1.2253e-02, -2.1454e-02, -4.8447e-03, -1.8454e-03, -2.9984e-02, 2.0866e-03, -1.1673e-02, 1.9455e-02, -2.7405e-02, 1.9165e-02, -3.7994e-02, 3.1769e-02, -3.7811e-02, -1.0361e-02, 1.5083e-02, 2.5158e-03, 1.5640e-02, -1.0681e-02, -9.3994e-03, 7.9041e-02, -6.2294e-03, 3.0045e-02, -1.2164e-01, -4.7188e-03,\n 1.2375e-02, -2.1515e-02, -2.7069e-02, 2.0569e-02, -5.1842e-03, -7.8354e-03, -2.5597e-03, -2.9572e-02, -1.5289e-02, 1.2329e-02, -2.1133e-02, -3.7079e-02, 3.8879e-02, -1.5533e-02, -1.2764e-02, -3.1464e-02, -1.8784e-02, 1.4290e-02, 7.2021e-03, -8.1863e-03, -6.3171e-03, 7.4501e-03, 1.1292e-02, -2.5375e-02,\n -3.5553e-02, 1.9872e-04, -6.1836e-03, -3.6640e-03, -7.1564e-03, -1.7868e-02, 1.3649e-02, -5.1971e-02, 9.7122e-03, -4.0955e-02, -4.9896e-03, -3.3417e-03, -4.5280e-03, -3.1174e-02, -2.6154e-02, 2.1957e-02, 1.1230e-02, 1.4000e-02, -6.0511e-04, 1.3855e-02, 7.3385e-04, -2.5732e-01, -7.5226e-03, 1.1887e-02,\n -1.0786e-03, -4.8523e-02, 2.2945e-03, -2.7969e-02, 7.5607e-03, -1.8188e-02, 5.2490e-03, -1.8448e-02, 8.1711e-03, -1.6907e-02, -3.5156e-02, -1.9226e-02, -1.5762e-02, 3.0537e-03, -1.4481e-02, -1.8768e-02, 9.5701e-04, 5.5054e-02, -3.4058e-02, -9.7733e-03, -3.2867e-02, 6.1493e-03, -1.9592e-02, 3.8544e-02,\n 8.4839e-03, 8.8501e-03, -2.1164e-02, 2.6413e-02, 8.3694e-03, -3.2444e-03, -2.1759e-02, 6.3820e-03, -1.6031e-03, -4.0222e-02, -2.0126e-02, -7.3700e-03, -1.1063e-02, 3.9864e-03, -7.4883e-03, 1.0822e-01, -2.0714e-03, 9.2316e-03, -5.5084e-03, -7.0953e-03, -9.7656e-03, 4.4403e-02, 2.7878e-02, 5.0323e-02,\n 6.3038e-04, 2.0874e-02, -2.9755e-02, -1.5625e-02, 2.0157e-02, -2.0630e-02, 1.8311e-02, 1.2169e-02, -2.1820e-02, 1.1368e-02, -2.6584e-05, 1.9501e-02, 6.8909e-02, 1.8036e-02, -4.0131e-02, 2.8477e-03, -3.5797e-02, -3.6041e-02, -1.1896e-01, 3.1799e-02, -2.9465e-02, 2.1210e-02, -1.0014e-03, 4.8752e-03,\n -4.0314e-02, 1.0048e-02, -4.9820e-03, -2.3178e-02, 7.9575e-03, 2.7725e-02, 1.8997e-02, -3.4698e-02, -1.0155e-02, -5.1498e-03, 1.9699e-02, 9.2850e-03, -9.5291e-03, 2.3605e-02, 3.5583e-02, -1.1841e-02, -5.4321e-03, 4.6173e-02, -4.6814e-02, -4.8737e-02, 8.8577e-03, -1.1879e-02, 1.8616e-02, -3.2387e-03,\n 5.8517e-03, -1.0628e-02, 2.2934e-02, -3.3783e-02, -1.1063e-02, -7.6771e-04, 3.1395e-03, -9.4604e-03, -1.5621e-03, -2.8564e-02, -2.7084e-02, -2.7771e-02, 4.0558e-02, 5.3177e-03, -9.7580e-03, -1.2779e-02, 1.3222e-02, -2.5879e-02, -2.1393e-02, -1.0841e-02, -2.4460e-02, 5.4855e-03, 3.1021e-02, 7.5226e-03,\n 2.7344e-02, -1.3664e-02, 2.7786e-02, -2.8488e-02, -1.1238e-02, -8.0948e-03, 4.0283e-02, 2.9602e-02, 3.5858e-03, -1.1909e-02, 2.9926e-03, -3.3684e-03, 3.9398e-02, 7.1487e-03, -4.9622e-02, -9.7809e-03, -3.4576e-02, -2.8896e-03, -3.5706e-02, -1.0376e-02, 5.1804e-03, -2.9831e-02, 2.9251e-02, 1.1002e-02,\n -4.8943e-03, 2.4933e-02, 4.1504e-02, -1.4671e-02, 8.6288e-03, 7.3120e-02, 1.4819e-01, 3.6530e-02, 9.8267e-03, -2.5192e-02, -1.5343e-02, -6.0364e-02, 3.4668e-02, -1.2978e-02, -2.7145e-02, -1.7883e-02, -1.0090e-03, -1.4267e-02, 1.8265e-02, -3.3386e-02, 8.2626e-03, -2.0580e-03, 2.2675e-02, -1.8463e-02,\n 1.8539e-02, 4.1626e-02, -1.8250e-02, 1.2962e-02, -5.8289e-02, 4.1962e-02, -2.9922e-02, -8.3008e-03, -5.3192e-02, 2.1545e-02, -3.7811e-02, 4.4250e-03, -8.6823e-03, -1.7868e-02, 3.6740e-04, 5.9395e-03, -1.2596e-02, -2.8057e-03, -3.2318e-02, -3.0502e-02, -1.8555e-02, 2.8488e-02, 2.1820e-03, -1.8311e-02,\n 1.2466e-02, 2.7664e-02, -5.3329e-03, -3.1128e-02, 2.1988e-02, 3.1204e-02, -9.0790e-03, 2.8015e-02, -2.1912e-02, -2.4796e-02, 3.9444e-03, -2.0248e-02, 6.0005e-03, 1.5404e-02, -2.0905e-02, -2.4658e-02, 2.0187e-02, 1.4191e-02, 2.3193e-02, -2.2263e-02, 2.0981e-02, 3.3903e-04, -3.7659e-02, 4.1626e-02,\n -6.0272e-03, -2.3575e-02, -3.3245e-03, -3.3661e-02, 5.0995e-02, -2.5725e-04, 7.4425e-03, -4.0970e-03, -5.4932e-02, -6.6795e-03, 1.6891e-02, -1.3321e-02, -9.2621e-03, -7.5455e-03, 5.2414e-03, -3.6285e-02, -5.4504e-02, -1.6266e-02, -2.3544e-02, -4.8920e-02, 6.7673e-03, 5.8380e-02, -9.1248e-03, -1.9012e-02,\n -4.8676e-02, -3.2463e-03, -9.9564e-03, 1.3474e-02, -2.0254e-04, 2.7603e-02, 9.9106e-03, -9.3307e-03, 6.0883e-03, 7.7095e-03, -1.1253e-03, -1.7761e-02, -1.6571e-02, 1.5821e-03, -2.8133e-03, 7.7171e-03, 1.3000e-02, -2.6260e-02, -1.4258e-01, -7.1793e-03, -3.4607e-02, -1.1093e-02, -4.7684e-03, 1.0956e-02,\n 1.7059e-02, 2.0050e-02, 3.4332e-02, 3.2886e-01, 7.0229e-03, -8.6517e-03, -4.2267e-02, -3.2257e-02, -7.5569e-03, 6.4697e-03, -1.1383e-02, -2.4841e-02, -2.8946e-02, 9.3536e-03, 1.7509e-03, 5.5542e-03, -1.5099e-02, 1.4771e-02, -1.2299e-02, 1.7792e-02, 2.9816e-02, -1.4671e-02, -1.9806e-02, -3.3813e-02,\n -1.1932e-02, 1.9958e-02, -2.8458e-02, 2.5574e-02, -1.1948e-02, -7.5102e-04, -3.9062e-02, -2.7985e-02, 3.3817e-03, 1.5602e-03, -1.0666e-02, 8.1863e-03, 6.8512e-03, 2.4155e-02, -2.6382e-02, 1.6846e-02, 1.1505e-02, -1.8005e-02, 5.2185e-03, 1.3466e-02, -1.7061e-03, -6.0616e-03, -1.8356e-02, 1.3367e-02,\n -2.6932e-03, -4.7264e-03, 8.7585e-03, 1.1894e-02, 1.0559e-02, 2.1683e-02, 3.4618e-03, 2.9812e-03, 2.3651e-03, 1.6281e-02, -8.3160e-03, -1.3702e-02, 2.2736e-02, 1.0460e-02, -3.1830e-02, 2.3849e-02, -6.5899e-04, 3.4637e-02, -6.6566e-03, -3.2654e-02, 3.1624e-03, -2.2354e-02, -2.9602e-02, -1.0910e-02,\n -3.2471e-02, -3.4454e-02, -2.5742e-02, -5.1613e-03, -2.2324e-02, -2.4750e-02, 4.7699e-02, 1.6693e-02, 9.3613e-03, -7.7553e-03, -3.7170e-02, 6.1073e-03, -3.0380e-02, -7.7209e-03, -5.2002e-02, 1.4849e-03, 6.4850e-03, -9.0408e-03, 1.8265e-02, 9.5139e-03, -8.7509e-03, -1.8539e-02, -5.6419e-03, 1.2314e-02,\n -3.8177e-02, -6.0349e-03, -3.3142e-02, 2.9282e-02, -4.5685e-02, 5.4245e-03, -2.5436e-02, -2.1790e-02, 2.5192e-02, -2.7885e-03, -8.7280e-03, -1.4191e-02, 3.0304e-02, -2.6184e-02, -2.3556e-03, 7.1068e-03, -2.9736e-03, -1.9379e-02, -1.0773e-02, -1.9653e-02, 7.0610e-03, 1.5839e-02, -8.4991e-03, -2.0859e-02,\n 1.2875e-03, 1.4458e-02, 2.4673e-02, -2.2125e-02, -1.6602e-02, -2.4109e-02, 1.1292e-03, 1.4633e-02, -2.1942e-02, -7.3891e-03, -1.6281e-02, 3.6144e-03, -3.6621e-02, 3.8357e-03, -1.7380e-02, 2.7466e-02, 1.8051e-02, -2.9175e-02, 3.2959e-02, 3.0106e-02, 2.3804e-02, -1.1436e-02, -1.3000e-02, -2.7204e-04,\n 4.1504e-03, -4.8920e-02, 2.2934e-02, 9.8343e-03, -3.1490e-03, -5.5695e-03, -1.2268e-02, -6.2332e-03, 7.8087e-03, -1.8311e-02, 2.8854e-02, -8.7967e-03, -3.5034e-02, 2.0111e-02, -5.8861e-03, -2.3026e-02, -1.8021e-02, 1.8358e-04, -5.2643e-02, 1.1597e-02, 1.9470e-02, -9.7961e-03, -4.1565e-02, -4.1016e-02,\n -1.1177e-02, -2.2583e-02, 1.4122e-02, 4.1382e-02, -8.2855e-03, 2.0340e-02, 1.1587e-03, -4.7913e-02, -2.3865e-02, 4.8904e-03, 3.3112e-02, -1.1200e-02, 2.5833e-02, -1.5808e-02, 8.7433e-03, 1.4465e-02, 3.4180e-03, 3.2013e-02, -8.7357e-03, -2.0027e-03, -8.3208e-05, -2.2469e-03, 1.0681e-02, -5.6580e-02,\n -3.1799e-02, -1.9333e-02, 5.9723e-02, 5.9586e-03, 1.6907e-02, 2.6489e-02, 2.5543e-02, -2.5223e-02, -7.6408e-03, 1.6052e-02, 1.5312e-02, -4.4937e-03, -2.0981e-02, -1.2219e-04, -7.6332e-03, -4.3060e-02, 1.1391e-02, 9.7809e-03, -6.5231e-03, 2.4887e-02, -2.5620e-02, -1.1032e-02, -5.0323e-02, -3.0075e-02,\n -2.1957e-02, -2.4246e-02, -5.3864e-03, -1.1627e-02, 4.8370e-03, -3.0766e-03, 2.4399e-02, -6.2714e-03, 2.6947e-02, -1.6006e-02, 1.4328e-02, 1.3824e-02, 7.4348e-03, 2.6947e-02, 3.3283e-04, -2.0111e-02, -1.0384e-02, -2.4567e-02, 1.9882e-02, 9.3002e-03, 2.2064e-02, 4.5044e-02, -2.0538e-02, -9.6664e-03],\n dtype=torch.float16), 'perceiver_resampler.proj_in.weight': tensor([[ 0.0016, 0.0183, 0.0078, ..., -0.0225, 0.0149, 0.0149],\n [-0.0461, -0.0285, -0.0263, ..., -0.0616, -0.0199, 0.0468],\n [-0.0015, -0.0100, -0.0414, ..., 0.0117, -0.0006, 0.0171],\n ...,\n [ 0.0208, -0.0064, -0.0189, ..., 0.0435, 0.0561, 0.0537],\n [-0.0318, -0.0083, -0.0089, ..., -0.0413, -0.0336, 0.0226],\n [-0.0257, 0.0165, 0.0391, ..., 0.0706, 0.0294, 0.0717]], dtype=torch.float16), 'perceiver_resampler.proj_in.bias': tensor([-9.3126e-04, -1.6769e-02, 2.9846e-02, -1.1375e-02, -2.8400e-03, -2.2842e-02, 6.0539e-03, 3.9330e-03, -2.5040e-02, -2.3560e-02, 2.7752e-03, 1.8524e-02, 8.2626e-03, 5.2567e-03, 7.2441e-03, -1.7120e-02, 1.3626e-02, -1.7792e-02, 5.0735e-03, -1.7334e-02, 2.7252e-02, -2.6417e-03, 2.9205e-02, -1.2772e-02,\n 2.5375e-02, 2.1561e-02, -2.1973e-02, -2.3174e-04, -1.9026e-03, -8.2169e-03, 9.3918e-03, 2.2526e-03, -9.7961e-03, 7.8506e-03, 1.2886e-02, -4.5624e-03, 9.9564e-03, -4.2267e-03, 9.6359e-03, -8.0032e-03, 1.2192e-02, 7.0724e-03, 1.5488e-02, 4.6539e-03, 2.0660e-02, -2.3392e-02, -4.5242e-03, 7.4043e-03,\n -7.7171e-03, 4.8332e-03, 1.7471e-02, 1.9638e-02, 1.3969e-02, 2.9373e-03, 7.8201e-03, -1.4572e-03, 1.2505e-02, 3.0689e-03, 1.7639e-02, -1.0214e-03, -6.8436e-03, -3.1464e-02, 1.6479e-02, -9.1858e-03, -1.0529e-02, -1.4313e-02, -8.9188e-03, 8.4543e-04, -6.6795e-03, 3.4637e-02, 2.5360e-02, -3.9177e-03,\n 5.1079e-03, 1.4534e-02, -2.6505e-02, -2.0084e-03, -9.6817e-03, -6.4583e-03, -2.6047e-02, -9.5444e-03, -1.7578e-02, -1.2558e-02, 1.8415e-03, -4.5166e-03, -1.6037e-02, -4.6654e-03, -7.2823e-03, 1.9867e-02, 1.4748e-02, 2.2903e-02, 5.2681e-03, -7.0267e-03, 9.5749e-03, 6.4049e-03, 1.9470e-02, 4.0779e-03,\n -2.1324e-03, -1.7075e-02, -1.0216e-02, 1.9089e-02, 1.8311e-02, 1.0162e-02, -3.2410e-02, 1.3412e-02, -1.1093e-02, 5.4970e-03, -2.3773e-02, 7.4844e-03, 1.2253e-02, 6.9923e-03, -1.6068e-02, -1.3733e-02, -1.6606e-04, 6.9733e-03, 1.9257e-02, -3.0670e-02, 2.4353e-02, 1.3916e-02, -1.3908e-02, -5.2605e-03,\n 2.9129e-02, 1.6342e-02, 2.4750e-02, -5.5389e-03, 9.7351e-03, -1.7685e-02, -2.4597e-02, -3.8700e-03, -3.0136e-02, 1.2367e-02, -1.2184e-02, -1.7075e-02, 2.2552e-02, 4.2915e-03, 2.0390e-03, 2.0798e-02, 1.3161e-03, -1.2887e-04, -2.3483e-02, -5.9814e-03, -3.4271e-02, -2.1072e-02, 2.7943e-03, 2.2447e-04,\n 1.5574e-03, -1.1444e-02, 1.7441e-02, 5.9967e-03, -1.8204e-02, -4.9057e-03, 1.0155e-02, 3.5767e-02, 8.5297e-03, 8.7662e-03, -6.6185e-03, -1.1147e-02, 2.0325e-02, 5.2147e-03, 1.2840e-02, -2.1347e-02, 3.5645e-02, -8.9798e-03, -7.1449e-03, -1.4706e-03, 6.1188e-03, -1.5556e-02, -3.4447e-03, -1.0834e-02,\n -5.4207e-03, 2.2690e-02, -6.5994e-03, -1.2619e-02, -1.6815e-02, -7.6904e-03, 1.7044e-02, -8.6260e-04, -5.8327e-03, -2.4979e-02, 1.1002e-02, 2.6428e-02, 4.7607e-03, 1.4809e-02, 2.2095e-02, 1.6846e-02, -1.8494e-02, -1.8358e-03, 9.0088e-02, 2.1088e-02, -8.3389e-03, 1.3100e-02, -1.8524e-02, 9.1476e-03,\n -1.2802e-02, -1.9684e-03, -5.7640e-03, -1.5900e-02, 1.0567e-02, -2.0660e-02, -2.1423e-02, -1.7380e-02, 2.4368e-02, 1.0345e-02, -6.0310e-03, -1.3138e-02, -1.3824e-02, 4.8218e-03, 1.1459e-02, 1.5778e-02, 2.3315e-02, -1.0948e-02, 1.1353e-02, 3.0029e-02, -6.3362e-03, -1.4053e-02, -1.0704e-02, 8.3237e-03,\n -9.0485e-03, 1.5488e-02, 1.7822e-02, -1.7197e-02, 5.6190e-03, -5.5008e-03, -1.6006e-02, -1.6373e-02, 2.3975e-03, -1.0498e-02, 9.0485e-03, -1.9577e-02, 1.5457e-02, -7.7515e-03, -1.4420e-02, -1.4130e-02, 1.8387e-02, -2.5223e-02, -3.2787e-03, 1.5358e-02, 1.4221e-02, 9.0561e-03, -1.4257e-03, -1.7929e-02,\n 1.0208e-02, -9.2316e-03, -5.3644e-04, 1.7120e-02, -2.2430e-02, 1.7853e-02, 9.1629e-03, 7.6904e-03, 1.9058e-02, 1.3016e-02, 1.9470e-02, -2.8122e-02, 1.2192e-02, 1.5495e-02, -4.5967e-03, 8.6365e-03, 6.5079e-03, 2.1095e-03, -1.2581e-02, 1.5795e-05, 4.3583e-04, -2.5116e-02, 1.2474e-02, -6.7520e-03,\n 1.2360e-02, 2.4979e-02, -1.5244e-02, -6.1913e-03, -1.4008e-02, 5.3835e-04, -8.3847e-03, 1.4870e-02, -1.4603e-02, 6.5613e-03, -3.3417e-03, -1.3214e-02, -4.9057e-03, -1.8860e-02, -7.4577e-04, 1.2665e-02, 1.2177e-02, 1.9638e-02, -2.3590e-02, -2.1103e-02, -1.5106e-02, 1.2817e-02, 1.7685e-02, 2.1423e-02,\n 2.8976e-02, -2.8839e-03, 2.1820e-02, -7.7820e-03, -1.9592e-02, -1.7914e-02, 9.0103e-03, -7.8869e-04, 8.6451e-04, -1.5808e-02, -2.4261e-02, -1.6800e-02, 2.0279e-02, 2.9343e-02, -3.7518e-03, -2.0020e-02, -3.6774e-03, -7.4501e-03, 7.0076e-03, 2.1210e-02, -1.0361e-02, -3.1643e-03, -1.8127e-02, 1.4679e-02,\n 1.2321e-02, 1.4580e-02, -4.0855e-03, 2.3254e-02, -7.2670e-03, 1.7136e-02, 1.8112e-02, 1.2541e-03, 2.1011e-02, -2.9316e-03, -1.3321e-02, 1.8892e-03, 6.6681e-03, 1.9180e-02, 4.7913e-03, 8.2932e-03, -1.0063e-02, -1.2321e-02, -2.6978e-02, -1.6449e-02, 7.2174e-03, 3.7933e-02, 1.6006e-02, 4.0169e-03,\n 7.6332e-03, -2.3788e-02, -1.1803e-02, -1.4702e-02, 6.3858e-03, -2.0504e-03, -1.5884e-02, -9.9335e-03, 8.9035e-03, 1.8280e-02, -3.0346e-03, 8.5831e-03, 6.8521e-04, -1.5961e-02, 1.2146e-02, -2.3651e-02, 1.4893e-02, 2.2316e-04, 1.8173e-02, -1.9989e-03, -6.3286e-03, 1.7181e-02, 7.2556e-03, 2.5818e-02,\n -6.9313e-03, -8.3771e-03, 1.1246e-02, 6.3896e-03, -1.7746e-02, 1.4938e-02, -1.4336e-02, -1.9852e-02, 3.5229e-03, -6.7978e-03, -7.9193e-03, 1.1398e-02, 8.1491e-04, 2.6535e-02, -1.2650e-02, 3.8414e-03, -2.3056e-02, 1.3763e-02, -1.2856e-02, -2.2995e-02, 8.4686e-03, 2.1088e-02, 1.6815e-02, 2.4643e-02,\n 2.0020e-02, 6.3820e-03, 1.2445e-03, -2.0218e-02, 2.5345e-02, -3.2215e-03, 1.9699e-02, 2.9541e-02, -7.3814e-03, 1.1482e-02, -9.8877e-03, 5.8250e-03, -2.2263e-02, 2.8320e-02, -5.5122e-03, 5.8632e-03, -1.6785e-02, -1.0452e-03, -1.9394e-02, -5.2452e-03, -1.1559e-03, -2.5139e-03, -1.2741e-02, 1.8219e-02,\n 1.4900e-02, -1.1169e-02, 3.3600e-02, 5.6267e-03, -1.8158e-02, 8.4839e-03, -6.0349e-03, 6.1569e-03, -1.2267e-04, 3.3665e-03, 2.7863e-02, -9.3002e-03, -1.0109e-02, 6.0501e-03, -1.3876e-03, -2.2717e-03, 2.3056e-02, -1.4397e-02, -2.3804e-02, -2.8046e-02, 2.2415e-02, 1.5640e-02, -3.1525e-02, -8.3237e-03,\n 5.5199e-03, 1.7654e-02, 1.0178e-02, 8.0414e-03, 1.8204e-02, 7.7972e-03, -1.3332e-03, -8.9035e-03, -4.4060e-03, -3.6850e-03, -9.6560e-04, 1.9569e-03, 1.8814e-02, -1.3145e-02, -1.1497e-02, 1.8494e-02, -1.8234e-02, 1.1730e-03, -9.7198e-03, -2.0340e-02, 2.5177e-02, 2.5520e-03, -2.1744e-02, -2.2675e-02,\n 2.0485e-03, -1.4885e-02, -1.2341e-03, 8.0948e-03, 9.3842e-03, -2.8946e-02, 9.5673e-03, -8.7814e-03, 1.0597e-02, -1.0063e-02, -4.6768e-03, 1.5774e-03, 2.3689e-03, 9.9182e-03, 2.4719e-02, 1.0231e-02, -2.4094e-02, 1.7731e-02, 1.8295e-02, -1.0727e-02, -2.7370e-03, 2.7893e-02, -1.1169e-02, 2.6962e-02,\n -9.6970e-03, -1.6312e-02, -1.0612e-02, -2.0355e-02, -2.6749e-02, -1.5114e-02, -1.6983e-02, -2.5940e-02, 4.8981e-03, 1.9684e-02, 1.4320e-02, -5.7068e-03, 1.9241e-02, 1.4519e-02, 5.4550e-03, -2.2446e-02, 3.9330e-03, 2.6093e-02, 1.6296e-02, 3.8853e-03, 1.2375e-02, 8.3923e-03, 2.0432e-02, 1.4048e-03,\n -1.2611e-02, -1.0742e-02, 1.1742e-02, 1.4519e-02, 8.8654e-03, 8.4229e-03, 1.6174e-02, 1.6251e-02, 1.3573e-02, 1.2550e-02, -2.8934e-03, -3.4599e-03, -1.7487e-02, 1.6876e-02, -3.8605e-03, 1.5022e-02, 2.2812e-03, -1.3313e-02, 8.0414e-03, -2.4673e-02, 1.9821e-02, -1.9150e-02, -1.3046e-02, 2.1839e-03,\n 1.6861e-02, -5.7755e-03, 1.2512e-02, 1.6541e-02, -1.5572e-02, -1.9394e-02, -1.5764e-03, 1.8661e-02, -3.2471e-02, -7.2632e-03, 9.3307e-03, 6.5689e-03, -1.1513e-02, -7.7486e-04, -7.5417e-03, -2.3102e-02, -1.4359e-02, -1.9775e-02, 3.6669e-04, -3.0613e-04, 8.9417e-03, -1.8326e-02, -2.5436e-02, -1.6937e-02,\n 1.0048e-02, 2.2106e-03, 9.4604e-03, 2.0172e-02, 9.0714e-03, -1.8982e-02, -2.1515e-02, 1.7868e-02, -1.3954e-02, 1.7433e-03, 5.6915e-03, 1.1314e-02, 7.0763e-03, -1.0880e-02, 1.6270e-03, -1.1719e-02, -1.1925e-02, 8.9188e-03, 1.1078e-02, -2.2156e-02, 2.9587e-02, -2.0523e-02, 1.1505e-02, -1.0300e-02,\n -3.9368e-03, -1.0674e-02, 2.8973e-03, 1.4648e-02, 1.2383e-02, -1.3992e-02, -1.9287e-02, 5.1460e-03, 4.7836e-03, 3.7155e-03, 5.5313e-03, 2.8877e-03, 1.0271e-03, -1.0178e-02, -1.6541e-02, -1.3596e-02, 4.0131e-02, 1.3573e-02, -2.4902e-02, 6.2218e-03, 1.0628e-02, -9.8953e-03, -1.3313e-02, 1.1816e-03,\n -1.8585e-02, -1.1986e-02, 2.1576e-02, 2.1149e-02, 1.7456e-02, -3.8834e-03, 1.5602e-02, -3.7231e-03, -1.1505e-02, 7.6752e-03, -4.0092e-03, 1.7975e-02, 5.8784e-03, 9.9640e-03, 2.4918e-02, 1.1803e-02, 1.9073e-02, -7.0000e-03, 1.5991e-02, 3.3661e-02, -3.3627e-03, 1.7014e-02, 1.3039e-02, -2.7227e-04,\n 2.4986e-03, -2.1160e-05, 1.9363e-02, -1.4107e-02, 1.8005e-02, 1.1871e-02, 5.2261e-03, -7.2289e-03, 1.3323e-03, 6.2294e-03, -3.0640e-02, -1.0384e-02, -2.0584e-02, -5.0850e-03, 2.2678e-03, -1.3390e-02, 2.5436e-02, -1.3054e-02, 2.1072e-02, -2.3270e-02, -4.1656e-03, 1.9760e-02, -2.2842e-02, 2.5162e-02,\n 9.0361e-04, -7.5188e-03, 2.5436e-02, 2.3994e-03, 7.1678e-03, -3.4142e-03, -3.1235e-02, 3.2806e-02, -7.0906e-04, 7.9575e-03, -2.2415e-02, 5.6038e-03, -2.8900e-02, 8.7357e-03, -1.3008e-02, 7.3204e-03, -2.1591e-02, 4.2267e-03, -1.2894e-03, 6.3896e-03, 5.2681e-03, 7.2060e-03, -1.6617e-02, 8.5449e-03,\n -1.3748e-02, 1.0424e-03, 2.4338e-02, 2.3697e-02, -6.9962e-03, 3.0804e-03, 1.2512e-02, 7.4158e-03, 8.9111e-03, 8.1863e-03, -1.7487e-02, 4.1246e-04, 1.3596e-02, -1.8250e-02, 1.0887e-02, 7.0152e-03, -1.1063e-02, -3.2440e-02, -1.9852e-02, 2.7328e-02, 3.8242e-03, 2.0004e-02, -1.1635e-02, -2.5425e-03,\n 7.6790e-03, 1.0521e-02, 8.7509e-03, -1.1871e-02, 1.7715e-02, 2.3232e-03, -3.0792e-02, 1.9119e-02, -2.8419e-04, 2.7054e-02, -2.7046e-03, -1.8417e-02, -2.1255e-02, -1.6907e-02, -9.1324e-03, 1.9043e-02, -9.8724e-03, 1.4557e-02, -2.6875e-03, -2.2751e-02, -3.0880e-03, -1.3191e-02, -1.2146e-02, -4.0550e-03,\n 2.9495e-02, 2.3346e-02, -1.3573e-02, -6.8665e-03, 8.8882e-03, -6.8703e-03, 1.2001e-02, -5.6791e-04, -1.1665e-02, 6.6605e-03, 2.1530e-02, -2.2217e-02, 1.8829e-02, 2.5513e-02, -7.0457e-03, -4.3564e-03, -1.0023e-03, 1.1177e-02, 1.5900e-02, 4.5204e-03, 7.2174e-03, 2.6306e-02, -8.9417e-03, -1.1604e-02,\n 5.5008e-03, -3.0041e-03, -1.0239e-02, 5.9967e-03, 1.2917e-02, 4.7379e-03, -3.6983e-03, 7.5188e-03, 6.6910e-03, 1.6098e-02, -8.7738e-05, 3.0384e-03, -9.3536e-03, 3.8269e-02, 1.3710e-02, 1.6891e-02, 1.9974e-02, -2.9812e-03, -2.7435e-02, 1.3260e-02, 3.5362e-03, -4.7379e-03, 5.1498e-03, 2.6131e-03],\n dtype=torch.float16), 'perceiver_resampler.proj_out.weight': tensor([[ 0.0334, 0.0221, -0.0682, ..., -0.0543, -0.0063, 0.0121],\n [ 0.0471, 0.0061, 0.0254, ..., -0.0036, -0.0243, 0.0395],\n [-0.0163, 0.0471, -0.0521, ..., -0.0257, 0.0210, -0.0237],\n ...,\n [ 0.0466, 0.0340, -0.1034, ..., -0.0661, 0.0071, -0.0772],\n [ 0.0465, 0.0299, 0.0292, ..., -0.0231, 0.0225, -0.0161],\n [-0.0034, 0.0245, -0.0088, ..., 0.0025, -0.0512, 0.0449]], dtype=torch.float16), 'perceiver_resampler.proj_out.bias': tensor([ 1.6602e-02, 1.6876e-02, -1.7395e-02, 2.7481e-02, -3.5004e-02, -2.9831e-02, 4.2458e-03, 3.8116e-02, 1.6968e-02, -6.2485e-03, 4.7516e-02, -4.5166e-02, -2.2858e-02, -1.7670e-02, 2.2797e-02, 2.5848e-02, 3.4607e-02, 4.9622e-02, -1.3748e-02, 2.5806e-01, 1.6403e-02, -2.6321e-02, 1.5854e-02, 2.1820e-02,\n 1.8814e-02, 2.8934e-03, 2.1973e-02, -1.7900e-03, -2.1225e-02, -5.2338e-02, -1.3466e-02, 2.2400e-02, -1.1909e-02, 3.5522e-02, -8.4000e-03, -8.3160e-03, -9.4223e-03, 8.5144e-03, 4.2786e-02, 1.5495e-02, 2.4658e-02, 2.2217e-02, 8.4991e-03, 2.9312e-02, -3.6346e-02, -6.7505e-02, -5.3883e-04, -2.7466e-02,\n -1.1612e-02, -5.2757e-03, -7.4463e-02, 2.9526e-02, 7.3280e-03, -1.2331e-03, 1.2871e-02, 1.2032e-02, 2.3849e-02, -5.7770e-02, 1.5587e-02, -4.1595e-02, 2.1866e-02, 4.0924e-02, 3.7994e-03, 2.4292e-02, 3.2578e-03, 4.2236e-02, 1.2369e-03, 6.0959e-03, -4.2267e-03, 5.3162e-02, 1.6281e-02, 1.8478e-02,\n 3.9154e-02, 3.0838e-02, -5.2246e-02, -2.9099e-02, 1.0933e-02, -3.2806e-02, -1.8082e-02, -1.4481e-02, 2.7176e-02, 2.2217e-02, 5.2307e-02, 7.2449e-02, 1.2497e-02, 4.9858e-03, -3.5675e-02, -1.7654e-02, 5.7487e-03, -2.8015e-02, -6.8436e-03, -1.4732e-02, 3.0079e-03, 2.9785e-02, 2.4509e-03, 4.1138e-02,\n 4.8462e-02, -7.9529e-02, 7.8049e-03, -1.7471e-02, 2.6749e-02, 5.0659e-02, -1.3138e-02, -1.1833e-02, -1.3039e-02, -4.0558e-02, 2.0248e-02, -4.4227e-05, -1.1475e-02, -9.4482e-02, -1.0109e-03, -4.8523e-02, -1.2985e-02, 1.8463e-02, 4.6570e-02, 4.0192e-02, -5.0163e-03, -1.0368e-02, -3.3112e-02, 2.8748e-02,\n -5.3436e-02, -1.5503e-02, 1.4420e-02, 4.8752e-03, 3.7201e-02, 8.2397e-03, -3.0212e-02, 3.0731e-02, 1.0231e-02, -8.2703e-03, 3.4607e-02, 3.3356e-02, 1.6281e-02, 2.3209e-02, -3.0956e-03, -7.3471e-03, 9.3994e-03, -2.6810e-02, -3.4790e-02, -1.6678e-02, 3.2410e-02, -1.0277e-02, 2.2919e-02, 4.2725e-02,\n 1.7776e-02, -3.1372e-02, -7.5607e-03, -4.6295e-02, -1.9485e-02, 3.1952e-02, 2.4399e-02, 2.3865e-02, 3.0380e-02, 1.9714e-02, -5.8807e-02, -1.9623e-02, -3.3386e-02, 4.9622e-02, -1.1490e-02, -2.3087e-02, 3.2227e-02, 3.2837e-02, -8.1711e-03, -2.7740e-02, -3.7567e-02, 8.0228e-05, -1.4511e-02, -1.3039e-02,\n -7.6752e-03, 3.2593e-02, 7.8964e-03, -5.2826e-02, -6.1646e-02, 8.0872e-02, -6.3599e-02, -2.0584e-02, -1.9150e-02, 3.1757e-03, -5.4970e-03, -1.4778e-02, 1.6815e-02, -2.0264e-02, 3.2349e-02, 4.1626e-02, -9.7656e-03, 2.7657e-03, -2.2552e-02, -3.6560e-02, -1.2466e-02, 3.3081e-02, 1.2684e-03, 3.7781e-02,\n -2.0264e-02, -1.1047e-02, -1.8864e-03, -6.3972e-03, 1.7654e-02, -1.8356e-02, 4.7119e-02, -1.4259e-02, 1.1091e-03, -2.6474e-02, -2.1851e-02, 3.2471e-02, 7.2021e-03, -1.5202e-03, -3.2104e-02, 5.5084e-02, -1.2199e-02, -3.4943e-02, -4.9683e-02, -6.8298e-02, -4.6173e-02, 1.6586e-02, 1.0889e-01, -3.1586e-02,\n -6.0425e-02, 3.7598e-02, 1.9119e-02, 5.1331e-02, 4.5319e-02, 5.5733e-03, -6.6895e-02, 1.1894e-02, 1.4755e-02, -4.9133e-03, 7.8430e-02, 5.3894e-02, -1.2436e-03, -7.4120e-03, -5.1498e-04, 1.3039e-02, -2.2125e-02, 3.2074e-02, -4.5013e-02, -6.9389e-03, -2.0447e-02, 1.7365e-02, 5.9624e-03, 2.1076e-03,\n -5.6725e-03, 2.1179e-02, 1.6312e-02, 3.0533e-02, -1.9272e-02, 5.0621e-03, -1.1719e-02, -1.4839e-03, -4.0817e-03, -5.0774e-03, 3.6499e-02, 9.8610e-04, 3.2867e-02, 4.7363e-02, -1.4183e-02, 1.1711e-02, 2.5604e-02, -4.8494e-04, -9.0576e-02, -7.3967e-03, -3.4485e-02, -1.2744e-01, -6.0272e-02, -2.4979e-02,\n -2.4750e-02, 6.4659e-03, 4.9377e-02, -8.0633e-04, -1.6739e-02, 1.6022e-02, 9.6588e-03, 1.4214e-02, -1.6312e-02, -1.8417e-02, 1.0941e-02, -1.4473e-02, -1.6708e-02, -1.7334e-02, -6.7749e-02, -2.0493e-02, 1.5305e-02, -1.4038e-03, -1.3535e-02, 3.8452e-02, 3.1372e-02, 4.4495e-02, -1.8387e-02, -1.9638e-02,\n 2.6150e-03, 2.2232e-02, 1.8799e-02, 3.6041e-02, -2.7267e-02, 2.3071e-02, 1.8906e-02, -1.7517e-02, 4.0466e-02, 2.8763e-02, 1.4664e-02, 2.9434e-02, -1.8661e-02, 4.2786e-02, -2.1988e-02, -9.9060e-02, -3.1555e-02, -4.1138e-02, -4.0314e-02, -8.7357e-03, -2.2278e-02, -2.2675e-02, 1.9503e-03, -1.4496e-02,\n 3.8025e-02, -9.2392e-03, -2.9663e-02, 3.3478e-02, -2.3865e-02, 1.7136e-02, -1.7824e-03, -1.4389e-02, -3.0121e-02, 3.2187e-04, 5.8746e-03, 3.2745e-02, 3.4692e-01, -1.6357e-02, 5.0507e-03, 1.9165e-02, 2.3972e-02, -1.6785e-02, 8.5510e-02, -9.6512e-03, 2.6875e-03, -1.5774e-03, 5.9784e-02, 1.1139e-02,\n -2.5314e-02, -3.8757e-02, -5.2612e-02, 4.6692e-03, 4.1840e-02, 6.3972e-03, -6.4087e-02, -8.6594e-03, 3.0136e-02, 3.5675e-02, -1.7960e-02, 1.5419e-02, -2.5284e-02, -4.1473e-02, -3.4363e-02, -1.4984e-02, 1.8188e-02, 2.7740e-02, 8.7509e-03, 3.7018e-02, 3.5706e-03, -1.0300e-02, -1.1719e-02, -1.9426e-03,\n -5.7495e-02, 2.5864e-03, 8.5878e-04, 2.0721e-02, -2.5024e-02, -1.9089e-02, 3.7903e-02, -1.5778e-02, -9.4604e-03, 4.3907e-03, -3.2471e-02, 1.7822e-02, -2.6215e-02, -2.8091e-02, -5.4779e-02, -4.5654e-02, -4.2114e-02, 1.6479e-02, 1.4923e-02, 9.0408e-03, -2.8915e-02, 4.1565e-02, 1.8127e-02, 1.5015e-02,\n -4.3549e-02, -4.1565e-02, -5.1605e-02, 6.6185e-03, -2.5711e-02, 4.0222e-02, 4.9164e-02, -2.5314e-02, 1.0071e-02, 1.4122e-02, 1.7227e-02, -3.3966e-02, 3.4027e-02, -6.1073e-03, -4.0436e-02, 1.3344e-02, -2.3782e-04, -2.1774e-02, 3.2368e-03, -1.5945e-02, -7.4615e-03, 1.2505e-02, 2.6962e-02, -2.5681e-02,\n 3.8269e-02, 8.2474e-03, -6.0181e-02, 1.2192e-02, 7.6782e-02, -3.2640e-04, -4.4518e-03, -2.5000e-01, 1.2188e-03, -1.3323e-03, 2.1118e-02, 8.2703e-03, 2.6794e-02, -3.5767e-02, -2.8229e-02, -4.9095e-03, 4.7073e-03, -2.7756e-02, -1.0635e-02, -2.7817e-02, -4.2664e-02, 2.0477e-02, 4.3060e-02, -5.0583e-03,\n -1.2367e-02, -9.5139e-03, -7.8506e-03, 2.9556e-02, -3.8330e-02, 1.0231e-02, 1.5535e-03, -2.1255e-02, 1.3145e-02, 3.1281e-02, -3.6659e-03, 2.7294e-03, -2.4246e-02, 4.0398e-03, -3.6407e-02, -1.5686e-02, -5.1231e-03, 3.5801e-03, 3.0937e-03, 1.2817e-02, -1.9455e-02, -1.8539e-02, -2.2919e-02, 3.0975e-02,\n -1.0046e-01, 1.4091e-02, 1.3275e-02, 3.2806e-02, 3.4180e-02, 4.4037e-02, 1.0284e-02, -9.0561e-03, -4.0222e-02, 5.6671e-02, -4.6326e-02, -1.5793e-02, 8.2855e-03, -5.7411e-03, 2.8412e-02, 1.1284e-02, -2.8564e-02, 2.1553e-03, -9.3699e-04, -1.5991e-02, -3.6865e-02, 2.7351e-03, -4.5837e-02, -4.6844e-03,\n -7.7705e-03, -2.3880e-02, -1.9703e-03, 4.6997e-02, 3.1158e-02, -1.1749e-02, -4.5074e-02, -4.9286e-03, 4.0375e-02, 3.0777e-02, 1.2108e-02, -1.0536e-02, -8.6212e-03, -4.9347e-02, 2.6230e-02, 1.4610e-02, -6.5384e-03, -1.9928e-02, 1.0536e-02, -3.1021e-02, -1.7033e-03, -3.1686e-04, -3.5004e-02, 6.5125e-02,\n -3.2501e-02, -1.2161e-02, -1.6769e-02, 5.0934e-02, -2.8610e-02, 5.5603e-02, 3.3447e-02, 2.0142e-02, -3.4790e-02, -4.2694e-02, -2.1393e-02, -1.0216e-02, 5.7922e-02, -7.2021e-03, -4.6722e-02, -9.8343e-03, 3.6621e-02, -4.3610e-02, 7.5500e-02, 2.2202e-02, -6.4575e-02, 4.1847e-03, -1.4481e-02, -1.4351e-02,\n -2.7740e-02, -5.1765e-03, -3.4149e-02, 2.7197e-01, -8.4457e-03, -2.4399e-02, -8.5545e-04, -9.3842e-03, -5.2612e-02, -3.3783e-02, 6.6589e-02, 3.9886e-02, 2.3621e-02, 4.1473e-02, -1.0117e-02, -1.0201e-02, -2.7039e-02, 1.1292e-02, 8.9340e-03, -2.8091e-02, 2.6352e-02, 9.3555e-04, -1.3748e-02, -4.7668e-02,\n -1.6769e-02, -2.9278e-03, 4.9782e-04, 6.7902e-03, 1.1475e-02, -2.5665e-02, 2.3529e-02, -4.8706e-02, -1.5335e-03, -2.5101e-03, -1.6546e-03, 7.6408e-03, 3.1311e-02, 6.2981e-03, 4.3144e-03, -1.1511e-03, 2.8305e-02, -3.9215e-02, -3.6835e-02, 5.7144e-03, -1.4694e-02, 1.8585e-02, 6.9397e-02, 2.3758e-02,\n 1.0002e-02, 5.7640e-03, -2.7512e-02, -2.9282e-02, -2.5757e-02, -2.7313e-02, 2.6306e-02, 7.2388e-02, 7.7744e-03, -7.6172e-02, 6.5002e-03, -1.2466e-02, 1.0246e-02, -1.3924e-02, -8.5907e-03, -1.1368e-02, 5.2299e-03, 3.9764e-02, 2.9922e-02, -2.6947e-02, -7.7553e-03, -3.3417e-02, 3.6346e-02, 4.5128e-03,\n -4.2480e-02, -2.9465e-02, 2.6505e-02, 6.1989e-03, -4.2664e-02, 2.0733e-03, 4.5471e-02, -7.4097e-02, 2.7496e-02, 2.6382e-02, 6.1310e-02, -2.7409e-03, 3.2990e-02, 2.8519e-02, 6.3562e-04, 6.6795e-03, 2.0920e-02, 7.7705e-03, -7.4890e-02, 3.4241e-02, -5.1537e-03, 1.1223e-02, 1.2810e-02, 2.0008e-03,\n 7.4120e-03, 4.1260e-02, -1.5572e-02, -1.7731e-02, 2.7618e-02, 7.6599e-03, 3.8116e-02, -2.3849e-02, -6.6711e-02, 9.7351e-03, -4.4006e-02, -4.9477e-03, -5.5664e-02, 3.8338e-03, -2.7481e-02, 1.4465e-02, -1.3237e-02, -4.5349e-02, -2.3937e-03, -2.5406e-02, 1.9730e-02, -3.5645e-02, 1.4481e-02, -8.5907e-03,\n -9.3384e-03, 3.6287e-04, 3.3569e-02, -7.2670e-04, 8.9493e-03, -4.0619e-02, -9.9277e-04, 3.4103e-03, -1.0345e-02, -4.0955e-02, -6.2378e-02, -3.2024e-03, 2.9434e-02, 6.6528e-02, 4.0283e-02, -2.4033e-02, -1.3550e-02, 8.0872e-03, -1.5808e-02, -1.4191e-02, 4.2633e-02, -8.2932e-03, -3.3474e-03, 1.0330e-02,\n 1.3268e-02, -1.8356e-02, 1.1169e-02, -8.2703e-03, 2.7435e-02, 6.2164e-02, -3.2368e-03, -1.3375e-04, -3.5156e-02, -3.2275e-01, -1.0834e-02, -6.8626e-03, 3.7720e-02, 1.8415e-03, -2.4155e-02, 5.3558e-02, -2.8839e-02, -2.1622e-02, 1.4030e-02, -1.0292e-02, -2.4307e-02, -7.1564e-03, -1.5745e-03, 2.1362e-02,\n -3.3417e-02, 1.8044e-03, 2.3819e-02, -5.4216e-04, 1.3866e-03, -4.1718e-02, -1.5343e-02, 2.2598e-02, -4.7150e-02, 4.2755e-02, -1.2146e-02, -4.1466e-03, 3.8116e-02, -1.8478e-02, 4.1870e-02, 4.7340e-03, 2.7100e-02, -1.6586e-02, 6.0425e-02, -2.2476e-02, -2.0096e-02, -1.9226e-02, 2.9240e-03, 3.5828e-02,\n 1.6266e-02, 9.9335e-03, 3.4454e-02, -1.4030e-02, 3.7098e-03, -4.1885e-03, 2.4231e-02, -1.0406e-02, -1.2787e-02, -2.9221e-02, 1.0498e-02, 8.0566e-03, 4.0253e-02, -4.5990e-02, 3.4729e-02, -2.5635e-02, -9.7046e-03, -4.5868e-02, -9.2840e-04, 2.5988e-04, -2.6672e-02, 8.4915e-03, 2.6169e-02, 2.6306e-02,\n 2.6169e-02, 5.2118e-04, 1.4114e-02, 4.0375e-02, 5.1346e-03, 3.1586e-02, -1.3268e-02, 4.1618e-03, -3.6163e-02, -2.5162e-02, -3.9307e-02, 4.0894e-03, -1.4565e-02, 7.6981e-03, 2.7405e-02, 3.3752e-02, -3.4607e-02, 2.8183e-02, 5.7953e-02, -2.7023e-02, -1.6312e-02, 8.7433e-03, 3.6743e-02, 5.8105e-02],\n dtype=torch.float16), 'perceiver_resampler.norm_out.weight': tensor([0.5762, 0.5093, 0.5498, 0.5068, 0.5679, 0.4724, 0.4976, 0.5518, 0.5532, 0.5005, 0.5342, 0.5498, 0.4966, 0.5181, 0.5347, 0.5400, 0.5483, 0.5400, 0.5229, 0.7134, 0.4944, 0.5054, 0.5278, 0.5010, 0.5205, 0.5107, 0.4937, 0.5493, 0.5649, 0.4995, 0.5259, 0.5435, 0.5122, 0.5337, 0.5337, 0.4846, 0.5112, 0.5210, 0.5361,\n 0.5073, 0.4844, 0.4885, 0.5503, 0.5557, 0.5088, 0.5259, 0.5269, 0.4795, 0.5718, 0.5142, 0.5342, 0.5376, 0.5059, 0.5640, 0.4683, 0.5166, 0.5137, 0.5347, 0.5171, 0.5376, 0.5186, 0.5269, 0.5063, 0.5366, 0.5322, 0.5142, 0.5283, 0.5552, 0.5576, 0.6133, 0.5049, 0.5186, 0.5410, 0.5034, 0.5112, 0.4795, 0.5278, 0.5405,\n 0.5161, 0.5674, 0.5630, 0.5376, 0.5303, 0.4871, 0.5298, 0.5420, 0.5029, 0.5317, 0.5229, 0.5527, 0.5034, 0.5103, 0.5176, 0.5522, 0.5225, 0.5391, 0.5303, 0.5337, 0.5044, 0.5161, 0.5312, 0.5898, 0.5513, 0.5332, 0.4946, 0.5190, 0.5063, 0.4819, 0.5669, 0.5195, 0.5371, 0.5518, 0.4807, 0.4976, 0.5415, 0.5308, 0.5571,\n 0.4700, 0.5347, 0.4695, 0.5195, 0.5151, 0.5259, 0.5439, 0.5200, 0.5220, 0.5332, 0.4819, 0.5288, 0.5054, 0.5234, 0.5400, 0.5059, 0.4832, 0.4829, 0.5288, 0.5200, 0.5112, 0.5347, 0.5298, 0.5410, 0.5278, 0.5005, 0.5332, 0.4756, 0.5288, 0.5132, 0.5293, 0.5493, 0.4768, 0.5029, 0.5225, 0.5308, 0.4709, 0.5234, 0.5713,\n 0.5098, 0.5420, 0.5205, 0.5137, 0.4927, 0.5298, 0.5161, 0.6079, 0.5410, 0.5522, 0.5127, 0.5249, 0.5093, 0.4827, 0.5225, 0.4836, 0.5425, 0.5527, 0.5601, 0.5513, 0.5537, 0.5166, 0.5190, 0.4714, 0.5239, 0.5493, 0.5068, 0.5073, 0.4631, 0.5342, 0.5298, 0.5591, 0.5322, 0.5103, 0.5581, 0.5317, 0.5439, 0.5063, 0.4868,\n 0.5542, 0.5347, 0.5371, 0.5005, 0.5015, 0.5317, 0.5332, 0.5479, 0.5288, 0.5288, 0.5020, 0.5132, 0.5166, 0.5762, 0.4932, 0.5093, 0.5176, 0.5601, 0.5312, 0.4226, 0.5508, 0.5430, 0.5596, 0.5044, 0.5366, 0.5522, 0.5439, 0.5786, 0.4543, 0.5215, 0.4546, 0.5439, 0.5732, 0.5430, 0.5332, 0.4998, 0.5146, 0.5146, 0.5356,\n 0.5522, 0.5000, 0.5151, 0.5166, 0.5078, 0.5063, 0.4998, 0.6182, 0.5049, 0.4883, 0.5117, 0.5117, 0.5352, 0.5493, 0.4919, 0.4822, 0.5513, 0.4875, 0.5312, 0.5044, 0.5659, 0.4919, 0.5474, 0.4805, 0.5488, 0.4905, 0.5942, 0.0469, 0.4915, 0.5239, 0.5229, 0.5884, 0.5151, 0.5342, 0.5288, 0.5142, 0.5264, 0.5127, 0.4751,\n 0.5088, 0.4836, 0.4961, 0.5239, 0.5215, 0.5381, 0.5488, 0.5122, 0.5225, 0.5166, 0.4763, 0.5571, 0.5415, 0.5049, 0.5537, 0.5425, 0.5181, 0.5195, 0.5103, 0.5625, 0.5200, 0.5142, 0.5137, 0.5635, 0.4985, 0.5479, 0.5029, 0.5566, 0.5200, 0.5103, 0.5020, 0.4729, 0.4949, 0.4719, 0.5220, 0.5132, 0.6226, 0.5020, 0.5552,\n 0.5303, 0.4802, 0.4932, 0.5562, 0.4817, 0.5288, 0.5513, 0.5332, 0.5322, 0.5454, 0.5190, 0.5259, 0.7764, 0.5205, 0.5020, 0.5347, 0.5459, 0.5020, 0.5088, 0.5371, 0.5576, 0.5171, 0.5063, 0.4626, 0.5605, 0.5420, 0.4946, 0.4846, 0.5244, 0.5254, 0.5532, 0.5356, 0.5425, 0.5366, 0.4500, 0.5117, 0.5371, 0.5405, 0.5571,\n 0.5063, 0.4871, 0.5444, 0.5371, 0.5078, 0.5400, 0.5215, 0.4675, 0.4902, 0.5405, 0.5254, 0.4988, 0.5386, 0.4927, 0.5430, 0.5410, 0.5405, 0.5547, 0.5132, 0.5522, 0.5244, 0.5269, 0.5098, 0.5420, 0.5278, 0.5127, 0.5615, 0.4761, 0.4951, 0.5659, 0.5830, 0.4912, 0.4634, 0.5039, 0.4829, 0.5278, 0.5103, 0.5117, 0.5371,\n 0.5293, 0.5420, 0.5347, 0.5220, 0.5273, 0.5244, 0.5156, 0.5078, 0.5117, 0.5078, 0.5698, 0.5239, 0.5679, 0.5327, 0.5669, 0.5430, 0.5352, 0.5649, 0.5278, 0.5015, 0.4688, 0.5381, 0.5503, 0.5898, 0.5576, 0.6841, 0.5024, 0.5396, 0.5415, 0.5425, 0.5190, 0.5503, 0.5054, 0.5557, 0.5415, 0.5176, 0.5010, 0.4727, 0.5132,\n 0.4973, 0.4875, 0.5356, 0.4988, 0.5068, 0.5049, 0.4724, 0.5400, 0.5469, 0.5322, 0.5430, 0.5063, 0.5132, 0.5171, 0.5347, 0.5303, 0.5415, 0.5239, 0.5146, 0.4917, 0.5454, 0.4890, 0.5503, 0.4890, 0.5332, 0.5488, 0.4983, 0.5181, 0.5513, 0.5210, 0.5181, 0.4761, 0.4749, 0.5195, 0.4910, 0.5371, 0.5132, 0.5493, 0.5103,\n 0.5410, 0.5352, 0.5176, 0.5039, 0.5391, 0.5337, 0.5601, 0.5400, 0.5195, 0.5200, 0.5479, 0.5107, 0.5215, 0.5308, 0.4653, 0.5190, 0.4980, 0.5146, 0.4990, 0.5449, 0.4548, 0.4954, 0.5371, 0.5298, 0.4995, 0.5264, 0.4700, 0.5103, 0.5093, 0.5283, 0.5132, 0.5049, 0.5444, 0.5225, 0.5244, 0.5161, 0.5420, 0.5239, 0.5293,\n 0.5439, 0.5234, 0.5371, 0.5225, 0.5366, 0.5293, 0.5239, 0.5327, 0.5552, 0.5342, 0.5112, 0.5732, 0.5239, 0.5547, 0.5054, 0.4338, 0.5171, 0.5312, 0.5781, 0.5176, 0.5098, 0.4775, 0.4995, 0.5508, 0.0340, 0.5269, 0.5234, 0.4939, 0.5073, 0.5244, 0.5107, 0.5747, 0.5327, 0.5190, 0.5229, 0.4749, 0.5259, 0.5347, 0.5269,\n 0.5645, 0.5269, 0.5474, 0.5293, 0.4719, 0.5127, 0.5288, 0.5117, 0.5112, 0.4993, 0.5361, 0.5254, 0.5210, 0.5586, 0.5029, 0.5361, 0.5342, 0.5264, 0.5464, 0.5488, 0.4331, 0.5562, 0.5493, 0.5415, 0.5107, 0.4841, 0.5552, 0.4956, 0.5571, 0.5361, 0.5024, 0.5503, 0.4924, 0.5215, 0.5161, 0.5605, 0.5405, 0.5435, 0.4976,\n 0.5459, 0.5146, 0.5474, 0.5376, 0.5566, 0.5312, 0.4692, 0.5044, 0.5000, 0.5151, 0.5400, 0.5200, 0.4729, 0.5444, 0.5112, 0.5029, 0.5298, 0.5352, 0.5566, 0.5479, 0.4824, 0.5342, 0.5083, 0.5029, 0.4814, 0.5928, 0.5142, 0.4648, 0.5181, 0.4937, 0.5127, 0.5669, 0.5693, 0.5630, 0.5547, 0.5508, 0.5083, 0.5205, 0.5103,\n 0.5200, 0.5542, 0.5269, 0.5449, 0.5127, 0.5098, 0.5244, 0.4758, 0.4963, 0.5967, 0.5093, 0.5029, 0.5405, 0.4744, 0.4482, 0.5000, 0.5562, 0.4773, 0.5132, 0.4973, 0.5508, 0.5493, 0.5142, 0.5488, 0.5322, 0.5137, 0.4961, 0.4958, 0.4619, 0.4951, 0.5210, 0.5269, 0.5200, 0.5278, 0.5664, 0.5303, 0.5635, 0.5391, 0.5674,\n 0.5205, 0.5425, 0.5156, 0.5654, 0.5029, 0.5571, 0.5376, 0.4910, 0.5410, 0.5527, 0.5225, 0.5195, 0.6001, 0.5073, 0.5391, 0.4849, 0.5142, 0.5659, 0.6533, 0.5176, 0.5200, 0.5010, 0.5010, 0.5703, 0.5254, 0.5742, 0.5386, 0.5083, 0.4907, 0.5903, 0.5190, 0.5151, 0.4927, 0.5293, 0.4746, 0.5283, 0.4863, 0.5254, 0.5176,\n 0.4866, 0.5088, 0.5200, 0.5269, 0.5361, 0.4744, 0.4998, 0.5420, 0.5371, 0.5356, 0.5034, 0.5132, 0.5527, 0.4871, 0.5259, 0.5308, 0.5049, 0.5146, 0.5234, 0.5430, 0.5122, 0.5527, 0.5410, 0.5127, 0.6147, 0.5439, 0.5420, 0.5244, 0.4905, 0.5015, 0.5308, 0.4717, 0.5093, 0.4983, 0.5474, 0.5640, 0.5073, 0.5005, 0.5366,\n 0.5220, 0.5244, 0.4812, 0.5229, 0.5435, 0.5581, 0.5435, 0.5146, 0.4707, 0.5195, 0.5225, 0.5488, 0.4954, 0.4482, 0.4900, 0.4985, 0.5312, 0.5425, 0.5200, 0.5674, 0.5693, 0.5171, 0.5171, 0.5166, 0.5674, 0.5293, 0.5400], dtype=torch.float16), 'perceiver_resampler.norm_out.bias': tensor([ 1.7738e-03, 1.0252e-03, -2.2507e-04, 5.8327e-03, 5.9433e-03, 6.6936e-05, -2.8825e-04, -2.5578e-03, 2.6169e-03, 8.1558e-03, 7.6561e-03, 1.8501e-03, -5.1613e-03, -7.7286e-03, -3.0308e-03, 3.2215e-03, 6.9389e-03, 5.3520e-03, 3.3112e-03, 5.8258e-02, -7.6675e-03, 1.9205e-04, 4.4861e-03, 8.6927e-04,\n 5.3549e-04, -7.7057e-03, -2.6321e-03, 4.3640e-03, 5.1737e-04, -4.5853e-03, -3.8033e-03, -7.7581e-04, 6.6414e-03, 1.8196e-03, -5.0926e-03, -9.8953e-03, 7.0343e-03, 5.9433e-03, 7.7782e-03, -5.0449e-04, -6.6528e-03, 2.0962e-03, 1.7052e-03, 5.4741e-04, 5.7755e-03, -2.3174e-03, 5.2910e-03, -7.0095e-04,\n 1.0315e-02, 5.8289e-03, -3.2635e-03, 2.7008e-03, 4.2419e-03, -2.9125e-03, 4.4250e-03, -2.3670e-03, 6.4850e-03, -9.0790e-03, 1.3565e-02, -7.2784e-03, 7.5493e-03, 2.2812e-03, -4.0817e-03, 2.6379e-03, 2.1057e-03, 9.5520e-03, -5.8670e-03, -4.4556e-03, -9.6035e-04, 9.3002e-03, -6.4468e-04, -2.9945e-03,\n 1.6037e-02, 3.6125e-03, -2.2011e-03, -6.6185e-03, 6.9771e-03, -6.7043e-04, -1.5917e-03, 2.9297e-03, -1.6105e-04, 6.5079e-03, 2.6093e-03, 8.7585e-03, 9.2459e-04, -1.4915e-03, -5.0392e-03, 4.0398e-03, -9.4509e-04, 9.2316e-03, -1.8253e-03, 2.5883e-03, 5.7030e-04, 4.7417e-03, 4.3907e-03, 7.5035e-03,\n 2.7027e-03, -3.1624e-03, 9.9468e-04, 4.8943e-03, 2.8706e-03, 1.3895e-03, 2.8992e-03, 6.3515e-03, 2.1515e-03, -1.0624e-03, 4.0169e-03, 7.5912e-03, 3.9597e-03, -7.7553e-03, 2.6855e-03, -3.3436e-03, 2.2926e-03, 2.1992e-03, 1.4572e-03, 4.7379e-03, 4.1428e-03, 6.7139e-03, -5.6839e-03, -7.4196e-03,\n -1.0118e-03, 3.0785e-03, 2.9011e-03, -1.9522e-03, -6.5269e-03, -2.4490e-03, -3.6068e-03, 7.9803e-03, 2.1992e-03, 4.4746e-03, 5.2757e-03, -2.9411e-03, 4.0703e-03, 4.1084e-03, 8.9121e-04, 3.8643e-03, 9.2010e-03, -7.8430e-03, -4.5929e-03, -1.7424e-03, -3.7575e-03, -6.6643e-03, -2.0409e-03, 5.9280e-03,\n -1.4248e-03, 3.6716e-03, 8.9111e-03, -2.8019e-03, 4.0016e-03, 4.2305e-03, 7.2174e-03, 5.1498e-03, 4.2381e-03, -7.3433e-04, -7.5264e-03, 9.7580e-03, -4.1428e-03, 1.1429e-02, -2.5883e-03, -8.2636e-04, 7.9012e-04, 6.7177e-03, 3.0556e-03, -1.1528e-02, 6.2485e-03, 8.1329e-03, 5.5313e-04, 1.6222e-03,\n -3.5620e-04, 5.6076e-03, -5.2528e-03, -1.9627e-03, -8.4763e-03, 8.7204e-03, -2.8763e-03, -3.1071e-03, 4.4785e-03, -2.8477e-03, -5.2261e-03, 3.9482e-03, -2.3007e-04, -3.4752e-03, 1.7071e-03, 4.5052e-03, -4.5547e-03, -6.6805e-04, -2.1935e-03, -5.6076e-03, 3.3588e-03, 3.5686e-03, 4.2152e-03, 2.1381e-03,\n -1.9913e-03, -8.3923e-04, 3.6678e-03, -6.4659e-03, 1.4143e-03, -3.8910e-03, 9.8190e-03, 3.5501e-04, -2.3651e-03, -2.5539e-03, -4.8804e-04, 6.8130e-03, -7.5626e-04, -2.0695e-03, -5.8899e-03, -2.9206e-04, 1.8806e-03, -1.9646e-03, 4.8141e-03, -5.0140e-02, 4.0030e-04, 8.9493e-03, 8.5693e-02, -2.2202e-03,\n -1.8864e-03, 2.0885e-03, 5.6190e-03, 2.4605e-03, 4.5242e-03, 3.0670e-03, -7.1373e-03, -2.8248e-03, 5.8670e-03, 6.7291e-03, 1.2770e-03, 6.3782e-03, -1.6346e-03, 3.7098e-03, 1.7891e-03, 8.2932e-03, -6.6109e-03, 1.0037e-04, 4.4560e-04, -9.2087e-03, 1.0133e-05, 1.2306e-02, 3.0174e-03, -2.0635e-04,\n 3.6507e-03, -9.1553e-04, 3.6106e-03, 7.7400e-03, -2.5120e-03, 6.8207e-03, 2.2049e-03, -3.8395e-03, 1.3628e-03, 7.6294e-04, 6.2332e-03, 2.2202e-03, 8.9111e-03, -1.4105e-03, 1.6800e-02, -1.6947e-03, 5.2567e-03, -1.1463e-03, 4.5013e-03, 2.7275e-03, 8.1062e-04, 2.0215e-01, -1.0643e-02, -8.1863e-03,\n 4.9973e-03, 8.7814e-03, 8.1253e-03, -5.6305e-03, -4.2801e-03, 8.9836e-04, -5.4626e-03, 1.2770e-03, -1.0777e-03, -1.2941e-03, 5.8031e-04, 3.9978e-03, -4.3564e-03, -1.5793e-03, -3.4256e-03, -1.4400e-03, 7.7667e-03, 3.2353e-04, 7.7133e-03, -1.7729e-03, 7.6103e-03, 9.8038e-03, 2.7657e-03, 3.6411e-03,\n -1.3900e-04, 1.6724e-02, 7.1526e-03, 7.1869e-03, -1.1482e-03, 3.7098e-04, -3.0041e-03, 8.0185e-03, 9.3842e-03, 6.2141e-03, -2.7065e-03, 7.9880e-03, 3.2806e-03, 3.8776e-03, 2.6798e-04, -6.6406e-02, -1.3666e-03, -2.6073e-03, 3.2444e-03, 5.6219e-04, -1.5421e-03, -1.3298e-02, 2.0814e-04, -4.6143e-02,\n 3.2635e-03, 9.0714e-03, -4.2953e-03, -1.6842e-03, -2.0847e-03, 7.9203e-04, 3.7251e-03, 3.8414e-03, -6.9618e-03, -3.8481e-04, -3.1590e-04, 2.9793e-03, 3.3447e-02, -1.8177e-03, 3.7804e-03, 1.7500e-03, 1.7090e-03, -1.7691e-03, 5.2643e-02, 1.9798e-03, -2.4509e-03, 1.7366e-03, -1.7099e-03, 2.8000e-03,\n -3.5596e-04, 5.8479e-03, 8.2207e-04, 3.0937e-03, 3.9444e-03, 7.2174e-03, -3.4752e-03, 2.2640e-03, 9.8877e-03, -3.4332e-03, -3.0193e-03, -5.5313e-04, -8.1100e-03, -1.2827e-03, -1.9531e-03, -9.9564e-04, 6.0921e-03, -3.8195e-04, 1.6327e-03, -1.5414e-04, 1.0635e-02, 4.6086e-04, -1.9836e-03, -3.3207e-03,\n -5.0049e-03, 8.1940e-03, 2.5196e-03, -1.6441e-03, -2.6646e-03, 7.8011e-03, 5.4131e-03, 1.9169e-03, -8.1730e-04, 4.1103e-04, 6.4926e-03, 1.6670e-03, 3.4866e-03, -1.3504e-03, -6.2895e-04, -1.2741e-03, -7.8659e-03, 3.4828e-03, -1.8034e-03, -3.2959e-03, -2.2793e-03, 3.8109e-03, 9.5444e-03, 4.7755e-04,\n -1.0328e-03, 3.6201e-03, 2.0599e-03, -1.3113e-03, -1.0330e-02, 8.6594e-03, -2.1515e-03, 2.3880e-03, 1.3695e-03, -4.5090e-03, 3.7651e-03, -1.0449e-04, 4.3373e-03, 7.0238e-04, -4.4212e-03, -5.6124e-04, 8.1406e-03, -1.4651e-04, -5.3978e-03, 2.4204e-03, 8.0795e-03, -3.4733e-03, 7.2784e-03, -3.3331e-04,\n -1.7872e-03, 8.1253e-03, -9.9850e-04, 1.5312e-02, 3.3207e-03, 1.0658e-02, -7.8613e-02, -2.5146e-02, 6.0692e-03, 1.4954e-03, -1.7090e-03, 5.9166e-03, 8.4915e-03, -6.5842e-03, 3.7346e-03, -4.1542e-03, 5.0621e-03, 2.7084e-03, 1.6117e-03, -6.8588e-03, 2.1782e-03, 2.1095e-03, 6.5536e-03, 2.9125e-03,\n -1.6727e-03, -1.6718e-03, 4.9114e-04, 6.5088e-04, -1.6279e-03, -5.8174e-03, 5.1575e-03, 4.6082e-03, -4.8103e-03, 3.6221e-03, 1.8377e-03, -3.0537e-03, -4.5815e-03, -5.2261e-03, -2.0962e-03, -6.3248e-03, 2.3842e-03, 6.6490e-03, -4.2558e-04, -9.0561e-03, -3.0060e-03, 3.8815e-04, 7.5531e-03, 1.9150e-03,\n -6.7177e-03, 7.8812e-03, 6.9847e-03, 7.0305e-03, 6.5422e-03, 6.9199e-03, -2.8114e-03, 7.5645e-03, 6.0034e-04, 8.3237e-03, -7.8857e-05, -3.6955e-04, 4.8409e-03, 1.6232e-03, 5.1613e-03, 5.4626e-03, 2.7924e-03, 4.9515e-03, 1.2283e-02, 4.7493e-03, -2.4490e-03, 5.8472e-05, -8.1921e-04, -3.3550e-03,\n 2.1973e-03, -3.7212e-03, 3.9787e-03, -5.0430e-03, 5.0278e-03, 8.4305e-03, -4.4746e-03, 1.4820e-03, 5.6124e-04, 7.0610e-03, -3.7613e-03, -4.5300e-05, 9.7656e-04, -4.3373e-03, 3.3474e-03, 4.0741e-03, -3.9749e-03, -1.3227e-03, 6.0463e-03, -4.6921e-03, 4.9210e-03, 3.1805e-04, 5.9013e-03, 1.3933e-03,\n -5.5428e-03, 1.2124e-04, 2.4242e-03, 1.2560e-03, 9.1743e-04, 8.6899e-03, 8.6746e-03, -1.5564e-03, -2.2316e-03, 4.9934e-03, -2.1420e-03, 2.6836e-03, 5.0201e-03, -3.5191e-03, -5.7182e-03, 3.5954e-04, 4.2076e-03, -3.5973e-03, 9.8145e-02, -3.4466e-03, -8.9493e-03, 4.3755e-03, 1.0672e-03, 1.4238e-03,\n -1.7977e-03, -6.6452e-03, 7.3814e-03, -1.8164e-01, 3.9368e-03, 2.8706e-03, -2.2869e-03, -1.4830e-03, -1.3714e-03, 2.4929e-03, 5.6076e-03, 4.0207e-03, 4.2534e-04, 5.4359e-03, -8.0185e-03, 4.2877e-03, -3.0861e-03, -6.3324e-04, 6.5117e-03, -3.3035e-03, 3.0251e-03, -3.3703e-03, 1.8768e-03, -1.6298e-03,\n 1.8883e-03, -2.3580e-04, 1.2941e-03, 5.0507e-03, 3.7346e-03, -5.7335e-03, -1.7986e-03, -3.9406e-03, -3.4142e-03, 8.1635e-03, -1.3840e-04, 1.9112e-03, 6.8712e-04, 1.8370e-04, 1.7872e-03, -6.8321e-03, 8.0109e-04, 7.0648e-03, 2.5482e-03, -3.3913e-03, -3.4008e-03, 5.2357e-04, 4.7493e-03, 2.1954e-03,\n 3.3989e-03, 2.2106e-03, -4.3983e-03, -6.2943e-04, -2.2812e-03, 1.1375e-02, 1.8320e-03, 2.7847e-03, 1.4221e-02, -4.0016e-03, 2.1553e-03, -1.9741e-03, 1.1551e-02, 3.4981e-03, 3.2158e-03, 1.7414e-03, 6.1455e-03, -1.4400e-03, 4.1008e-03, -2.6226e-03, 1.9894e-03, 3.6640e-03, 6.8426e-04, -1.8902e-03,\n -3.3016e-03, 1.0329e-04, 1.1909e-02, 1.7271e-03, -3.0365e-03, -6.7749e-03, 1.1078e-02, -2.7847e-03, 1.9760e-03, 1.3494e-03, 2.0340e-02, 6.9504e-03, -2.4581e-04, 3.3112e-03, -1.5564e-03, -7.9489e-04, 2.8114e-03, 1.0710e-03, 8.8120e-03, 3.0346e-03, 1.9026e-03, 6.6042e-04, -8.4305e-04, 5.0011e-03,\n 9.3794e-04, -2.4581e-04, -2.7466e-03, 7.2336e-04, -7.8559e-05, 3.7251e-03, 1.7624e-03, 1.3762e-03, 5.9776e-03, -7.4081e-03, -4.8294e-03, -1.6165e-04, -6.7558e-03, -5.3368e-03, -6.3515e-03, -2.5806e-03, 7.6447e-03, -8.0109e-03, -2.5725e-04, -1.8196e-03, 7.5493e-03, -3.8052e-03, 2.0905e-03, -2.5711e-03,\n -1.4694e-02, -1.6022e-03, 3.2883e-03, 7.6115e-05, 2.2087e-03, -3.4924e-03, 5.1117e-03, 4.6806e-03, -2.9697e-03, -6.3477e-03, -2.7122e-03, -3.0613e-03, -2.2030e-03, 1.5945e-02, 1.1337e-02, -1.7786e-04, 4.0970e-03, 2.2869e-03, -4.3225e-04, 3.1929e-03, -1.7424e-03, -8.1177e-03, 2.5043e-03, -3.1033e-03,\n 9.0027e-03, -3.0565e-04, 5.7945e-03, -1.4572e-03, -2.2678e-03, 3.5992e-03, -9.3222e-04, 1.2703e-03, -6.1607e-03, -4.9438e-02, -4.4022e-03, 2.5959e-03, -1.0691e-03, 4.2267e-03, 1.3912e-04, -1.9627e-03, -3.5419e-03, -5.3177e-03, -1.9178e-03, 2.5902e-03, -9.3155e-03, 8.4915e-03, 5.4407e-04, 2.1496e-03,\n -6.1989e-03, 1.5583e-03, 2.3708e-03, 2.4567e-03, 4.0588e-03, -3.6478e-04, -6.6948e-04, -2.0046e-03, -4.6539e-03, 4.7684e-03, 6.3419e-04, -1.4248e-03, 3.3951e-03, 6.4316e-03, 1.1835e-03, 5.9557e-04, 7.7858e-03, 4.3297e-03, 4.8218e-03, 2.7733e-03, 7.2632e-03, 9.0027e-04, 2.1954e-03, 5.8353e-05,\n 3.9062e-03, 3.2673e-03, 5.8670e-03, 3.6926e-03, -1.7061e-03, 5.9624e-03, 9.9106e-03, -3.2845e-03, 3.7041e-03, 9.0790e-04, 9.7122e-03, 6.4430e-03, 5.0879e-04, 3.8223e-03, 8.5983e-03, -5.5428e-03, -3.9160e-05, -2.7752e-03, -4.4594e-03, 4.8208e-04, 1.0406e-02, 3.5400e-03, -4.7827e-04, 9.0933e-04,\n -2.3060e-03, -1.6289e-03, 7.2899e-03, 7.9117e-03, -4.0855e-03, 2.2583e-03, 2.6340e-03, -2.9469e-03, -4.1885e-03, -9.1028e-04, -8.7814e-03, 7.0114e-03, 4.4785e-03, 2.2430e-03, 7.0267e-03, -9.5463e-04, -4.0245e-03, 9.0637e-03, 5.0468e-03, -2.8019e-03, 2.1973e-03, -3.2349e-02, 2.1350e-04, 5.4131e-03],\n dtype=torch.float16), 'perceiver_resampler.layers.0.0.norm1.weight': tensor([0.7563, 0.7856, 0.7852, 0.7539, 0.7373, 0.7803, 0.7656, 0.7964, 0.7974, 0.7524, 0.7632, 0.7485, 0.7695, 0.8027, 0.8057, 0.7700, 0.7744, 0.6948, 0.7754, 0.7773, 0.8145, 0.7349, 0.7666, 0.6685, 0.7520, 0.7949, 0.7534, 0.8105, 0.6270, 0.7349, 0.7563, 0.7993, 0.7769, 0.7563, 0.7930, 0.7632, 0.6504, 0.7568, 0.7573,\n 0.7495, 0.7383, 0.7974, 0.7720, 0.8081, 0.7812, 0.7715, 0.7061, 0.7705, 0.7515, 0.7432, 0.7295, 0.7329, 0.7617, 0.7759, 0.7412, 0.8188, 0.7812, 0.6357, 0.7847, 0.7349, 0.7324, 0.7832, 0.7393, 0.7485, 0.7554, 0.7866, 0.8042, 0.7007, 0.7817, 0.8457, 0.7705, 0.7290, 0.7925, 0.7539, 0.8066, 0.7690, 0.7251, 0.7241,\n 0.7373, 0.7764, 0.7646, 0.7866, 0.7520, 0.7002, 0.7754, 0.7559, 0.7651, 0.7573, 0.7939, 0.7202, 0.7461, 0.7578, 0.7729, 0.7402, 0.7988, 0.6782, 0.7803, 0.7266, 0.7715, 0.7524, 0.7441, 0.7651, 0.7881, 0.7720, 0.7725, 0.7505, 0.7563, 0.7827, 0.7656, 0.7173, 0.7803, 0.7690, 0.7651, 0.7446, 0.7505, 0.7700, 0.7231,\n 0.8003, 0.7559, 0.7236, 0.7490, 0.7471, 0.7715, 0.8677, 0.7046, 0.6704, 0.7500, 0.7744, 0.6274, 0.7563, 0.7104, 0.7646, 0.7393, 0.7583, 0.7422, 0.7905, 0.7832, 0.7295, 0.7666, 0.7822, 0.7471, 0.7446, 0.7119, 0.7471, 0.7739, 0.8257, 0.7612, 0.7246, 0.7764, 0.7070, 0.7461, 0.7129, 0.7725, 0.7295, 0.7827, 0.7412,\n 0.7222, 0.7275, 0.7461, 0.7427, 0.7788, 0.7368, 0.7930, 0.7153, 0.7168, 0.7769, 0.7349, 0.7285, 0.7866, 0.7368, 0.7534, 0.7622, 0.7534, 0.7900, 0.7476, 0.7441, 0.7637, 0.7183, 0.7319, 0.7622, 0.7612, 0.8071, 0.8018, 0.8125, 0.7925, 0.7388, 0.6338, 0.7773, 0.7021, 0.7412, 0.7788, 0.6318, 0.7729, 0.7920, 0.7935,\n 0.7476, 0.7207, 0.7305, 0.7515, 0.8291, 0.7290, 0.7559, 0.7134, 0.7217, 0.7778, 0.7344, 0.7393, 0.8013, 0.7095, 0.7075, 0.7485, 0.7056, 0.6880, 0.7856, 0.7026, 0.6841, 0.7700, 0.7349, 0.7705, 0.7617, 0.7241, 0.7642, 0.8096, 0.8013, 0.7739, 0.6396, 0.7812, 0.7808, 0.7573, 0.7505, 0.7373, 0.7612, 0.8145, 0.7109,\n 0.7505, 0.8071, 0.8125, 0.7627, 0.7271, 0.6826, 0.7700, 0.7593, 0.7637, 0.7236, 0.7002, 0.7344, 0.7305, 0.7588, 0.7646, 0.7852, 0.7817, 0.7476, 0.7051, 0.7031, 0.7583, 0.6982, 0.7588, 0.7603, 0.6523, 0.7368, 0.7456, 0.6987, 0.7778, 0.7393, 0.7754, 0.6821, 0.7412, 0.8120, 0.7456, 0.7437, 0.8086, 0.7134, 0.8550,\n 0.7500, 0.7793, 0.7715, 0.7769, 0.7188, 0.7012, 0.7354, 0.7036, 0.7754, 0.7661, 0.7310, 0.7197, 0.6997, 0.7246, 0.7671, 0.7627, 0.7607, 0.7456, 0.6992, 0.7095, 0.6465, 0.7720, 0.7339, 0.7554, 0.7461, 0.7427, 0.7534, 0.7466, 0.7163, 0.6699, 0.7544, 0.7681, 0.6450, 0.7021, 0.7954, 0.7163, 0.7261, 0.7334, 0.7554,\n 0.7446, 0.7671, 0.7310, 0.7720, 0.7368, 0.7690, 0.7031, 0.8047, 0.7407, 0.7280, 0.7261, 0.7129, 0.7607, 0.7739, 0.7993, 0.7964, 0.7759, 0.7446, 0.7266, 0.7266, 0.8208, 0.7705, 0.7212, 0.8037, 0.7861, 0.7505, 0.7651, 0.7529, 0.7729, 0.7241, 0.7017, 0.7070, 0.7744, 0.7373, 0.7515, 0.7666, 0.7793, 0.7671, 0.7539,\n 0.7988, 0.7319, 0.7476, 0.7573, 0.7832, 0.8066, 0.7388, 0.7480, 0.8223, 0.7817, 0.7070, 0.7720, 0.7314, 0.7734, 0.7007, 0.7598, 0.7290, 0.7515, 0.7700, 0.7891, 0.7588, 0.6909, 0.7534, 0.7700, 0.7622, 0.7637, 0.7759, 0.6709, 0.7593, 0.6997, 0.7261, 0.7529, 0.7725, 0.7168, 0.7529, 0.7524, 0.7622, 0.7354, 0.7432,\n 0.7607, 0.7974, 0.7588, 0.7080, 0.7373, 0.7466, 0.7461, 0.7441, 0.7690, 0.7607, 0.8057, 0.7114, 0.7363, 0.7632, 0.7705, 0.7642, 0.7910, 0.7549, 0.7275, 0.7192, 0.7549, 0.7812, 0.7261, 0.7734, 0.6694, 0.7944, 0.7432, 0.6553, 0.7529, 0.7798, 0.7383, 0.7798, 0.7886, 0.7603, 0.7466, 0.7720, 0.7192, 0.8071, 0.7124,\n 0.7031, 0.7881, 0.7310, 0.7578, 0.7515, 0.7695, 0.7998, 0.7773, 0.7158, 0.7822, 0.7842, 0.7827, 0.7998, 0.7793, 0.7607, 0.7583, 0.7949, 0.7808, 0.7456, 0.6870, 0.7349, 0.7148, 0.7178, 0.7593, 0.7012, 0.7378, 0.7422, 0.7339, 0.7466, 0.7622, 0.7915, 0.7588, 0.7520, 0.7368, 0.7324, 0.7485, 0.7959, 0.7773, 0.7573,\n 0.7915, 0.7393, 0.7856, 0.7246, 0.7256, 0.6699, 0.8018, 0.7593, 0.7612, 0.7603, 0.7671, 0.7808, 0.7290, 0.7319, 0.7466, 0.7612, 0.7334, 0.6938, 0.7607, 0.6499, 0.7559, 0.6758, 0.7632, 0.6177, 0.7197, 0.7334, 0.7925, 0.8003, 0.7651, 0.7539, 0.7446, 0.7539, 0.7397, 0.7534, 0.7363, 0.7163, 0.7451, 0.7725, 0.7520,\n 0.7520, 0.7715, 0.7095, 0.7192, 0.7573, 0.7705, 0.7202, 0.7539, 0.7329, 0.7534, 0.7256, 0.7622, 0.7007, 0.7769, 0.7920, 0.7827, 0.8140, 0.7861, 0.8027, 0.6460, 0.7827, 0.7432, 0.6055, 0.7744, 0.6968, 0.7349, 0.7290, 0.8301, 0.7910, 0.7466, 0.7417, 0.7686, 0.7080, 0.7036, 0.8135, 0.7251, 0.7305, 0.7710, 0.7974,\n 0.7793, 0.7300, 0.7749, 0.7603, 0.7539, 0.7568, 0.7417, 0.7661, 0.8228, 0.7446, 0.7900, 0.7319, 0.7578, 0.7207, 0.6963, 0.7974, 0.7349, 0.7480, 0.7529, 0.7983, 0.7695, 0.7603, 0.7578, 0.7456, 0.7705, 0.7969, 0.7837, 0.6582, 0.7305, 0.7642, 0.7227, 0.7271, 0.7661, 0.7329, 0.7349, 0.7437, 0.7686, 0.8135, 0.8076,\n 0.7129, 0.7158, 0.6709, 0.7563, 0.7573, 0.7202, 0.7524, 0.7515, 0.8042, 0.7549, 0.7485, 0.7847, 0.7563, 0.7451, 0.7319, 0.7192, 0.7563, 0.6709, 0.7358, 0.7515, 0.7134, 0.7500, 0.7764, 0.7666, 0.6963, 0.7544, 0.7681, 0.7432, 0.7427, 0.7808, 0.7397, 0.7021, 0.7373, 0.8042, 0.7871, 0.7251, 0.7891, 0.7104, 0.7842,\n 0.7656, 0.6919, 0.7637, 0.7158, 0.7183, 0.7783, 0.7476, 0.7803, 0.7090, 0.8149, 0.7544, 0.7036, 0.7505, 0.7739, 0.7778, 0.7778, 0.7681, 0.7725, 0.7832, 0.7607, 0.7227, 0.7456, 0.7896, 0.7012, 0.7661, 0.7754, 0.6929, 0.7153, 0.7856, 0.7896, 0.7988, 0.7856, 0.7554, 0.7451, 0.7583, 0.7930, 0.6880, 0.7319, 0.7358,\n 0.6895, 0.7378, 0.7803, 0.7920, 0.7598, 0.7363, 0.7358, 0.6797, 0.7676, 0.7695, 0.7251, 0.7549, 0.8086, 0.7734, 0.7637, 0.7983, 0.7544, 0.7578, 0.7437, 0.7300, 0.7808, 0.7310, 0.6802, 0.7559, 0.7598, 0.7485, 0.7285, 0.7451, 0.7407, 0.7812, 0.7241, 0.7520, 0.8140, 0.7725, 0.7241, 0.7612, 0.7402, 0.6919, 0.7339,\n 0.7886, 0.7368, 0.7642, 0.7368, 0.7671, 0.7705, 0.8076, 0.6904, 0.6836, 0.7075, 0.7593, 0.7368, 0.7437, 0.7905, 0.7563, 0.7754, 0.7573, 0.6025, 0.7612, 0.7847, 0.8037, 0.7847, 0.7646, 0.7534, 0.7104, 0.6797, 0.7534, 0.7944, 0.7104, 0.7783, 0.7334, 0.7236, 0.7622, 0.7632, 0.7168, 0.7568, 0.7295, 0.7524, 0.7319,\n 0.6826, 0.7876, 0.7437, 0.7505, 0.7925, 0.7085, 0.7827, 0.7549, 0.7261, 0.7646, 0.8154, 0.7915, 0.7622, 0.7515, 0.7046, 0.7896, 0.6992, 0.7617, 0.7168, 0.7847, 0.6978, 0.7256, 0.7515, 0.7690, 0.7080, 0.7271, 0.6885], dtype=torch.float16), 'perceiver_resampler.layers.0.0.norm1.bias': tensor([-9.3765e-03, 1.7910e-03, -2.6836e-03, -2.1500e-02, 1.5587e-02, -3.0319e-02, -6.2513e-04, 3.2501e-03, 1.3443e-02, -1.9897e-02, 2.0645e-02, -1.4809e-02, -5.4169e-03, -3.5973e-03, -2.0401e-02, 1.2024e-02, 1.3924e-02, -3.4180e-02, -1.9226e-02, -1.9791e-02, 3.7567e-02, -2.5604e-02, -5.1331e-02, -1.1932e-02,\n -1.0796e-02, -8.8406e-04, -4.4312e-02, -1.5793e-02, 4.6173e-02, -3.6888e-03, 2.7496e-02, -4.6539e-02, 1.5038e-02, 1.6441e-03, -2.9259e-03, 6.6528e-03, 3.2562e-02, 4.0169e-03, -7.8154e-04, 1.6487e-04, 2.4261e-02, -4.1260e-02, 7.6485e-03, 2.4994e-02, -1.2703e-02, -1.3107e-02, -2.0294e-02, -2.6443e-02,\n -6.4697e-03, 2.7054e-02, 1.4601e-03, 5.5206e-02, 1.2787e-02, 3.0869e-02, 2.3346e-02, 3.0090e-02, 2.4750e-02, 2.2476e-02, 2.0981e-02, -2.7359e-02, 4.0344e-02, -1.9302e-03, -1.3382e-02, -4.5547e-03, -1.3456e-03, -3.3661e-02, -1.0506e-02, -3.1891e-02, -9.3994e-03, -6.9275e-03, -9.8114e-03, -4.3915e-02,\n 4.1840e-02, -2.0050e-02, 6.3400e-03, -6.0242e-02, 4.0710e-02, -1.1795e-02, -3.2318e-02, -3.5583e-02, -6.9313e-03, -2.5101e-02, -1.3924e-02, 3.2593e-02, -7.3929e-03, 5.7888e-04, 7.6561e-03, 1.0994e-02, 1.6800e-02, -1.3924e-02, 1.6281e-02, 2.5131e-02, 1.3771e-02, 1.3985e-02, 9.3765e-03, -3.7811e-02,\n 5.9242e-03, 5.7373e-03, 3.2139e-03, 1.0811e-02, 2.8839e-02, 1.5144e-02, 4.6959e-03, 3.0167e-02, 1.6632e-02, -2.9354e-03, -3.1490e-03, -1.1551e-02, -3.4424e-02, -1.5472e-02, 1.1563e-04, -1.6785e-02, -9.9945e-03, 1.6571e-02, 1.6617e-02, -4.7516e-02, 1.9913e-02, -3.9577e-04, -3.4821e-02, 1.0941e-02,\n 7.1182e-03, -2.0889e-02, -1.1703e-02, 4.0321e-03, -1.1009e-02, -1.1299e-02, 2.0279e-02, 3.1738e-02, 1.0291e-01, 3.8239e-02, 2.8934e-03, 4.3144e-03, 3.5919e-02, -2.2552e-02, 3.2616e-04, 1.2024e-02, 3.1799e-02, 2.3010e-02, -1.8707e-02, 1.8692e-02, -1.8982e-02, -4.8584e-02, 4.5738e-03, -1.9333e-02,\n 4.0863e-02, -1.3580e-02, -5.8708e-03, -1.4603e-02, 6.1531e-03, 3.6041e-02, -1.4626e-02, -3.1433e-02, -1.8723e-02, 5.3711e-02, -1.3817e-02, 2.5482e-02, -1.2581e-02, -1.9638e-02, 2.9129e-02, -1.4816e-02, -3.7498e-03, 3.6041e-02, -9.4681e-03, 3.7415e-02, 1.0361e-02, 3.0701e-02, 2.6398e-03, 6.5536e-03,\n -1.1387e-03, -1.9424e-02, -3.8605e-02, 6.7627e-02, -1.4811e-03, 3.7170e-02, 1.1757e-02, -1.6394e-03, 3.3447e-02, 1.0361e-02, 4.0344e-02, 2.3468e-02, -1.2444e-02, 1.5221e-02, 2.5665e-02, 6.1646e-03, 1.0376e-02, -2.0737e-02, -1.2012e-01, -5.9166e-03, 3.5065e-02, 2.3468e-02, -9.8495e-03, 1.4185e-01,\n 8.7051e-03, -1.6739e-02, 5.6686e-03, 4.6173e-02, 1.6281e-02, -3.4256e-03, 3.0014e-02, 1.5656e-02, 3.1677e-02, -2.9251e-02, -2.7939e-02, 4.3144e-03, 2.3834e-02, -4.9164e-02, -4.3976e-02, -1.8341e-02, 2.5009e-02, -1.1276e-02, 2.0477e-02, 1.0406e-02, 3.7170e-02, -6.9580e-03, -8.9722e-03, 2.8793e-02,\n -6.1684e-03, 1.9703e-03, 4.7058e-02, 7.8506e-03, 2.3666e-02, 1.6342e-02, -7.7868e-04, 8.7204e-03, -1.0521e-02, -5.8655e-02, -3.0533e-02, -1.1124e-02, -8.7891e-03, 3.8147e-02, -2.4658e-02, 3.6049e-03, -4.4441e-04, 2.6871e-02, 2.1915e-03, -1.1337e-02, -2.0508e-02, -1.1299e-02, 2.0035e-02, 9.7351e-03,\n 5.1666e-02, -1.6876e-02, -1.9653e-02, 2.1317e-02, -4.3060e-02, -1.9745e-02, -2.5101e-02, -1.4900e-02, -1.8600e-02, -3.6888e-03, 4.0680e-02, -1.8188e-02, -5.9631e-02, 8.2550e-03, -8.2169e-03, 6.0669e-02, 5.2147e-03, -1.4107e-02, -5.6244e-02, -2.0813e-02, -5.3520e-03, -1.9104e-02, 5.5733e-03, -1.8066e-02,\n 1.3924e-02, 3.8513e-02, -1.4748e-02, -1.7899e-02, -5.4901e-02, -1.4984e-02, 1.2703e-02, 4.8096e-02, -1.3100e-02, -4.8218e-03, -1.8387e-02, -1.1620e-02, 5.6953e-03, 2.4445e-02, -1.7273e-02, 6.7854e-04, 8.1863e-03, 1.2718e-02, -3.8422e-02, 6.9824e-02, 1.6312e-02, 9.1095e-03, -3.0518e-02, 3.6530e-02,\n 1.6006e-02, -1.4763e-02, -1.5083e-02, 2.9053e-02, -3.4847e-03, 2.4628e-02, 9.8801e-03, 1.6266e-02, -5.4741e-03, 2.6199e-02, -4.0344e-02, 1.6830e-02, -7.7171e-03, 2.6627e-02, -7.6477e-02, -2.6596e-02, 1.5717e-03, 3.4943e-02, 4.4647e-02, -6.1531e-03, -8.1024e-03, 1.1147e-02, -3.3588e-03, 1.4133e-03,\n 1.9974e-02, 3.0533e-02, -3.2684e-02, 2.1881e-02, -6.6338e-03, 1.4145e-02, 5.7922e-02, 9.0837e-05, -4.1122e-03, -2.0905e-02, -3.0441e-03, 3.3379e-03, -1.1063e-02, 6.0997e-03, -2.1622e-02, 2.1534e-03, -4.8981e-03, -4.1016e-02, -4.6814e-02, -3.3417e-02, 1.3550e-02, 2.7710e-02, -2.5208e-02, -2.6642e-02,\n -3.5286e-03, -3.0533e-02, -7.5054e-04, -8.3084e-03, 3.4760e-02, -4.0985e-02, -2.4033e-02, 1.5564e-02, 1.0353e-02, -2.6810e-02, -7.4577e-03, -2.7863e-02, -2.1286e-02, -8.0109e-03, 1.7838e-02, 2.2614e-02, 5.0720e-02, -1.8295e-02, 1.1314e-02, -2.0218e-02, -1.8524e-02, 7.2289e-03, -3.7140e-02, -9.8267e-03,\n -2.2675e-02, -1.6815e-02, -1.2947e-02, 5.7648e-02, -4.2603e-02, -3.4515e-02, -7.1945e-03, -2.8931e-02, 1.2360e-02, -1.4877e-02, 3.1403e-02, -3.5839e-03, -9.0332e-03, 3.0670e-02, -1.3496e-02, -9.6283e-03, 1.8341e-02, 2.1381e-03, -1.5419e-02, -2.3098e-03, 5.0323e-02, 2.3346e-02, -1.0674e-02, -9.6970e-03,\n -4.4250e-03, -2.1912e-02, 5.3024e-03, -3.1281e-02, 3.5896e-03, 2.0615e-02, 2.0981e-02, 8.7357e-03, 1.3283e-02, 1.8082e-02, -9.9106e-03, -1.7120e-02, 4.8103e-03, 1.7685e-02, 5.6744e-04, -3.9185e-02, -1.0513e-02, -2.9160e-02, 2.9861e-02, -1.1017e-02, -1.5526e-02, -2.9755e-02, -2.1774e-02, -6.4564e-04,\n -3.7117e-03, -7.2365e-03, -1.3756e-02, 4.0405e-02, -1.0330e-02, -1.3176e-02, 3.2745e-02, 4.9515e-03, -1.0895e-02, 2.6672e-02, 2.4826e-02, -2.1530e-02, 5.6124e-04, 5.9471e-03, -1.6449e-02, 6.7673e-03, 6.8932e-03, 5.1956e-03, -1.7258e-02, 6.9847e-03, 2.3315e-02, 7.0190e-02, 1.7075e-02, -7.2975e-03,\n -1.4847e-02, 2.8229e-02, -1.6510e-02, -2.1072e-02, 9.8419e-03, -1.8372e-02, 2.1042e-02, -4.4830e-02, -2.0485e-03, -3.1281e-03, -1.3351e-02, -4.8103e-03, 2.0905e-02, 1.9608e-02, -2.7103e-03, -2.1164e-02, 5.2551e-02, -6.1188e-02, -4.3106e-03, -4.6265e-02, -2.0798e-02, 1.1917e-02, 1.2321e-02, 3.8086e-02,\n 2.6749e-02, -8.4381e-03, 3.8971e-02, -5.1689e-03, -2.7832e-02, 3.8757e-02, -3.9276e-02, -1.3113e-03, -1.8112e-02, 3.6774e-03, 3.2501e-03, 2.2491e-02, -1.6342e-02, 4.8492e-02, -1.7487e-02, 4.4525e-02, -1.9241e-02, -2.2202e-02, -1.2369e-03, -5.0659e-02, 7.7171e-03, -9.8343e-03, -2.0370e-02, 3.2043e-02,\n 2.2934e-02, -2.8946e-02, 3.5744e-03, 1.3458e-02, 3.0838e-02, -1.2497e-02, -1.0109e-02, -4.2633e-02, -2.1687e-03, -2.1774e-02, -3.4393e-02, 7.2876e-02, 2.2202e-02, -1.0933e-02, 5.3894e-02, -2.3361e-02, -1.4763e-02, -3.4451e-04, -1.9821e-02, 1.1078e-02, 6.2256e-03, -7.3853e-03, -1.4442e-02, -5.3528e-02,\n 2.7802e-02, -1.7532e-02, -8.4915e-03, 1.9627e-03, -1.9073e-02, -4.1351e-03, 1.0727e-02, 2.2583e-03, -8.2550e-03, 7.5302e-03, -1.2684e-03, -2.8152e-02, -5.1422e-03, 2.6207e-03, -1.0658e-02, -2.8381e-02, -6.4125e-03, 3.2227e-02, 2.7557e-02, 1.1543e-02, 8.3983e-05, 2.6108e-02, -4.9255e-02, -1.1978e-02,\n 1.1568e-03, -8.9661e-02, -4.6539e-03, 3.3970e-03, 3.7262e-02, -1.0658e-02, -6.9962e-03, 7.9498e-03, -2.6062e-02, -5.9814e-03, 3.5065e-02, -1.8585e-02, -4.9316e-02, -1.7334e-02, -4.8309e-02, 8.9417e-03, -8.1787e-03, -2.2110e-02, 2.8778e-02, 1.4992e-02, -2.8046e-02, -2.7084e-02, 2.5192e-02, 5.2681e-03,\n 1.8326e-02, 6.7711e-03, 9.4032e-04, -4.8676e-02, -5.6114e-03, 4.8157e-02, -8.7814e-03, -1.4046e-02, -2.4063e-02, -4.2847e-02, -1.3161e-02, -1.0223e-02, 8.2321e-03, -1.6083e-02, 1.8173e-02, -1.0254e-02, 7.4234e-03, -2.0767e-02, -9.9182e-03, 2.0275e-03, -1.7075e-02, 3.4027e-02, -1.4633e-02, 1.3161e-02,\n 5.7106e-03, -4.5776e-02, -4.1290e-02, -1.9104e-02, 1.2238e-02, 2.8397e-02, -2.9434e-02, 1.3771e-03, 2.8564e-02, 1.4221e-02, 1.8768e-03, -5.6458e-04, -1.2909e-02, -3.8300e-02, -2.8656e-02, -1.0376e-02, 1.9592e-02, 2.2980e-02, 3.0537e-03, 1.7120e-02, 1.0086e-02, 1.1292e-02, -1.7929e-02, -9.4757e-03,\n -1.0185e-02, 2.8717e-02, 1.1549e-03, -6.5565e-05, 5.6610e-03, -1.7883e-02, 3.2558e-03, -5.1994e-03, 5.8556e-03, 1.0008e-04, -2.7298e-02, -3.4637e-02, -1.3952e-03, 2.2526e-03, -1.2871e-02, 9.6130e-03, 3.5339e-02, 4.0100e-02, 3.9520e-02, 1.9409e-02, -2.7710e-02, 1.6708e-02, -2.3987e-02, -2.7527e-02,\n -1.7487e-02, 4.6448e-02, 1.0925e-02, -2.6978e-02, 5.3062e-03, 1.9135e-02, 8.9569e-03, -2.1194e-02, 3.2867e-02, 1.3008e-02, 1.9424e-02, -2.4094e-02, 4.9973e-03, 1.3313e-02, -5.2681e-03, -2.5131e-02, 2.2217e-02, -1.8555e-02, 2.1240e-02, -2.5116e-02, -3.8109e-03, -9.1248e-03, -3.3112e-02, -2.3331e-02,\n 2.0279e-02, 8.9645e-03, -2.4567e-02, -2.4963e-02, -9.9564e-03, -1.2083e-03, 2.1606e-02, 4.0283e-03, -5.3978e-03, 6.3038e-04, -2.7756e-02, -2.2339e-02, 1.8219e-02, 5.6976e-02, 1.9379e-02, 3.1143e-02, 1.4038e-02, -4.6425e-03, 4.3297e-03, 1.1383e-02, -1.4137e-02, 7.7286e-03, -3.8177e-02, -5.5511e-02,\n -3.1910e-03, -1.7462e-03, 7.0267e-03, -3.1021e-02, -8.7051e-03, -1.4633e-02, -6.2256e-03, 3.6049e-03, -2.3972e-02, -1.0025e-02, -2.7084e-02, -2.2202e-02, -4.7421e-04, -3.4389e-03, 5.0392e-03, -5.2376e-03, 1.0328e-03, 6.2406e-05, -1.7731e-02, 1.3153e-02, 1.4610e-02, 2.7786e-02, -2.6917e-02, 1.4496e-02,\n -1.0132e-02, 3.5278e-02, 2.4658e-02, 1.7654e-02, 2.4002e-02, 4.4067e-02, -2.1393e-02, 2.5055e-02, 2.5787e-02, -3.2867e-02, -2.9633e-02, -2.7039e-02, -2.0828e-03, -2.5116e-02, -3.5522e-02, -2.2308e-02, -1.1225e-03, 6.6772e-02, 7.5722e-03, -1.9681e-04, -5.5756e-02, 2.9312e-02, 2.4734e-02, 1.3989e-01,\n 2.3186e-04, -9.4843e-04, -3.1525e-02, -1.7670e-02, -2.7145e-02, 5.6038e-03, -1.6144e-02, -5.2521e-02, 1.5312e-02, 8.2016e-03, -4.2786e-02, -2.7542e-02, -1.0658e-02, 4.1924e-03, -3.0243e-02, 1.3411e-05, 2.7435e-02, -2.1896e-02, -6.8779e-03, 1.3466e-02, 1.7166e-02, 2.2018e-02, 1.1703e-02, -1.3580e-02,\n -6.9656e-03, -7.3357e-03, -1.1749e-02, -3.9673e-03, 1.7960e-02, 3.6255e-02, 5.0079e-02, 2.3468e-02, 2.3514e-02, 1.6159e-02, -2.1149e-02, 3.0193e-03, -3.9697e-04, -4.2450e-02, -1.0574e-02, 5.9235e-02, 2.1439e-03, 2.8809e-02, -1.5762e-02, -1.6617e-02, 1.9974e-02, 1.6434e-02, -3.9581e-02, -8.5999e-02],\n dtype=torch.float16), 'perceiver_resampler.layers.0.0.norm2.weight': tensor([0.6997, 0.7153, 0.7490, 0.7002, 0.6899, 0.7007, 0.7324, 0.6167, 0.7505, 0.7310, 0.7046, 0.7329, 0.7168, 0.7188, 0.7427, 0.7095, 0.6890, 0.6943, 0.7227, 0.5552, 0.6831, 0.7803, 0.6890, 0.6992, 0.6973, 0.7271, 0.6875, 0.7075, 0.7529, 0.7085, 0.7471, 0.6899, 0.7217, 0.7412, 0.7310, 0.6660, 0.7598, 0.6724, 0.7710,\n 0.6934, 0.6802, 0.7344, 0.7017, 0.7603, 0.6606, 0.6875, 0.7305, 0.7114, 0.7012, 0.7080, 0.7295, 0.7441, 0.7178, 0.7295, 0.7026, 0.7178, 0.6914, 0.7334, 0.7017, 0.7100, 0.7026, 0.7217, 0.7520, 0.7368, 0.7065, 0.6924, 0.6948, 0.6343, 0.7617, 0.7378, 0.7397, 0.7197, 0.7554, 0.6836, 0.7466, 0.7178, 0.7432, 0.7622,\n 0.6987, 0.7212, 0.7427, 0.7100, 0.7012, 0.7104, 0.7393, 0.7290, 0.7559, 0.7427, 0.6963, 0.7300, 0.6797, 0.7100, 0.7476, 0.7764, 0.7319, 0.7500, 0.7021, 0.7676, 0.7031, 0.6934, 0.7407, 0.7290, 0.7495, 0.6982, 0.7612, 0.7188, 0.6934, 0.7407, 0.7383, 0.7056, 0.7246, 0.7271, 0.7056, 0.7334, 0.7671, 0.7759, 0.7744,\n 0.7158, 0.6919, 0.6968, 0.7344, 0.7573, 0.7490, 0.8120, 0.6802, 0.7563, 0.7217, 0.6948, 0.7314, 0.7085, 0.6797, 0.6641, 0.7002, 0.6919, 0.7129, 0.7495, 0.7007, 0.7661, 0.7129, 0.7212, 0.7373, 0.7437, 0.6631, 0.7026, 0.7134, 0.7158, 0.6519, 0.7285, 0.7354, 0.7280, 0.7261, 0.6729, 0.7661, 0.7095, 0.7026, 0.7354,\n 0.6943, 0.6963, 0.7300, 0.7065, 0.7407, 0.6782, 0.6982, 0.6172, 0.7334, 0.7202, 0.7163, 0.7231, 0.7344, 0.6675, 0.6992, 0.7539, 0.7280, 0.7842, 0.7173, 0.7598, 0.6943, 0.6870, 0.7495, 0.6763, 0.7095, 0.6831, 0.7358, 0.7144, 0.6865, 0.7446, 0.6880, 0.7324, 0.7632, 0.7363, 0.7061, 0.6807, 0.6978, 0.7271, 0.7241,\n 0.7422, 0.7202, 0.6724, 0.6992, 0.6943, 0.7407, 0.6592, 0.7412, 0.7803, 0.7173, 0.7568, 0.6680, 0.7349, 0.7266, 0.7017, 0.7031, 0.6030, 0.6421, 0.6963, 0.5283, 0.7383, 0.7046, 0.7397, 0.7310, 0.7344, 0.7305, 0.7646, 0.6753, 0.7637, 0.6895, 0.7437, 0.7012, 0.7759, 0.6782, 0.7212, 0.7202, 0.6797, 0.6538, 0.6953,\n 0.7378, 0.6455, 0.7041, 0.6519, 0.6826, 0.7222, 0.7383, 0.6196, 0.7266, 0.7378, 0.7065, 0.7124, 0.7612, 0.7583, 0.6929, 0.7266, 0.6875, 0.7617, 0.6421, 0.6846, 0.4973, 0.7285, 0.7031, 0.6665, 0.5063, 0.6968, 0.7324, 0.5679, 0.6675, 0.7344, 0.6914, 0.4683, 0.7104, 0.7012, 0.7451, 0.7339, 0.7466, 0.6895, 0.6860,\n 0.7271, 0.7412, 0.6992, 0.7266, 0.7334, 0.7129, 0.7524, 0.6978, 0.6470, 0.7041, 0.6904, 0.6323, 0.7451, 0.7627, 0.7285, 0.6890, 0.6831, 0.6499, 0.7422, 0.7070, 0.6733, 0.7168, 0.7432, 0.6875, 0.7808, 0.7256, 0.6621, 0.7051, 0.6968, 0.7324, 0.5537, 0.7544, 0.6729, 0.7139, 0.6782, 0.7163, 0.6528, 0.7275, 0.5161,\n 0.7246, 0.6968, 0.6860, 0.7617, 0.7222, 0.7129, 0.7002, 0.7695, 0.7334, 0.7007, 0.7031, 0.6699, 0.6909, 0.7749, 0.7607, 0.7524, 0.6909, 0.6787, 0.4934, 0.7837, 0.7202, 0.7344, 0.7749, 0.7437, 0.7427, 0.6890, 0.7026, 0.7334, 0.7329, 0.7168, 0.6968, 0.6860, 0.7080, 0.7407, 0.7368, 0.7109, 0.7021, 0.7437, 0.6948,\n 0.7476, 0.7231, 0.7617, 0.7095, 0.6782, 0.7124, 0.7500, 0.7412, 0.6772, 0.7065, 0.7031, 0.7231, 0.7344, 0.7568, 0.5762, 0.7061, 0.7715, 0.7129, 0.7100, 0.6191, 0.6831, 0.7183, 0.7437, 0.7017, 0.7222, 0.6904, 0.6226, 0.6650, 0.6948, 0.7339, 0.7349, 0.6758, 0.7207, 0.7437, 0.7275, 0.6914, 0.7007, 0.7012, 0.7075,\n 0.7212, 0.7373, 0.7310, 0.6890, 0.6846, 0.7261, 0.7617, 0.7622, 0.6992, 0.7070, 0.6597, 0.6562, 0.7363, 0.6973, 0.7119, 0.7124, 0.7070, 0.7300, 0.7300, 0.7446, 0.7168, 0.6670, 0.7476, 0.7676, 0.5122, 0.7764, 0.7222, 0.6650, 0.7266, 0.7280, 0.7593, 0.7700, 0.7402, 0.7412, 0.7334, 0.7114, 0.7725, 0.7363, 0.6763,\n 0.7363, 0.7505, 0.6846, 0.6899, 0.7349, 0.6875, 0.6987, 0.7402, 0.7446, 0.7114, 0.6963, 0.7305, 0.7178, 0.7700, 0.7119, 0.7261, 0.7534, 0.6733, 0.6465, 0.7227, 0.7246, 0.7676, 0.6807, 0.7153, 0.8096, 0.7700, 0.6855, 0.6240, 0.7085, 0.7300, 0.7588, 0.7070, 0.7114, 0.7344, 0.7300, 0.7451, 0.7769, 0.7358, 0.7627,\n 0.6689, 0.7417, 0.7183, 0.7412, 0.7285, 0.7754, 0.7344, 0.7305, 0.6807, 0.7197, 0.7690, 0.6685, 0.6997, 0.7227, 0.7227, 0.7383, 0.7646, 0.6768, 0.7710, 0.6699, 0.7310, 0.7178, 0.7563, 0.7524, 0.7148, 0.7354, 0.7339, 0.7119, 0.7314, 0.6978, 0.6899, 0.6689, 0.7017, 0.7822, 0.7700, 0.7080, 0.7017, 0.7222, 0.7080,\n 0.6626, 0.6748, 0.6929, 0.7217, 0.7588, 0.7236, 0.6729, 0.7109, 0.7202, 0.7002, 0.6899, 0.7446, 0.7329, 0.7783, 0.7271, 0.5601, 0.7095, 0.7412, 0.7070, 0.7480, 0.6831, 0.7183, 0.6929, 0.6899, 0.5977, 0.7598, 0.6802, 0.7017, 0.6851, 0.7129, 0.7324, 0.7310, 0.6875, 0.6865, 0.7207, 0.7178, 0.7568, 0.7334, 0.7549,\n 0.7212, 0.7021, 0.7451, 0.6807, 0.7583, 0.7114, 0.8145, 0.6782, 0.7095, 0.6997, 0.7744, 0.7168, 0.7124, 0.7305, 0.7227, 0.7192, 0.7661, 0.7524, 0.7593, 0.7441, 0.6689, 0.7056, 0.7158, 0.5757, 0.7051, 0.7202, 0.7676, 0.7144, 0.6792, 0.7070, 0.7100, 0.7280, 0.7002, 0.7373, 0.7812, 0.6904, 0.7554, 0.6821, 0.6694,\n 0.7529, 0.7305, 0.7573, 0.7378, 0.7251, 0.7051, 0.7119, 0.7134, 0.7358, 0.7056, 0.8018, 0.7358, 0.7290, 0.7646, 0.6934, 0.6826, 0.7212, 0.6719, 0.7217, 0.7407, 0.7485, 0.6470, 0.7344, 0.7407, 0.7734, 0.6011, 0.7427, 0.6841, 0.6821, 0.7720, 0.6973, 0.7339, 0.7017, 0.6821, 0.7666, 0.7207, 0.6953, 0.7007, 0.7144,\n 0.7227, 0.7769, 0.6992, 0.7598, 0.7339, 0.7080, 0.7217, 0.7397, 0.7388, 0.5771, 0.7188, 0.7417, 0.7114, 0.6836, 0.6475, 0.6909, 0.6641, 0.7031, 0.7036, 0.7539, 0.6572, 0.7227, 0.6572, 0.6846, 0.6846, 0.7354, 0.7266, 0.7373, 0.7188, 0.7031, 0.7349, 0.7153, 0.7529, 0.7427, 0.6094, 0.6914, 0.7192, 0.6460, 0.7695,\n 0.7280, 0.6792, 0.7427, 0.7925, 0.7847, 0.7944, 0.7383, 0.7158, 0.7290, 0.6743, 0.7407, 0.6846, 0.7505, 0.7388, 0.7124, 0.7373, 0.6797, 0.7158, 0.6577, 0.6909, 0.7593, 0.7603, 0.7412, 0.7690, 0.7637, 0.6880, 0.6772, 0.7539, 0.6606, 0.7300, 0.7363, 0.6992, 0.7817, 0.7466, 0.7461, 0.6919, 0.7451, 0.6821, 0.6636,\n 0.6885, 0.7524, 0.6509, 0.7017, 0.7539, 0.7119, 0.6934, 0.6968, 0.7231, 0.6865, 0.7285, 0.6597, 0.7661, 0.7373, 0.7139, 0.6958, 0.7017, 0.7842, 0.7012, 0.7676, 0.7207, 0.7061, 0.7480, 0.7705, 0.7207, 0.6006, 0.6704, 0.7256, 0.6826, 0.6787, 0.6797, 0.7085, 0.7080, 0.7324, 0.8032, 0.7349, 0.6646, 0.7065, 0.6499,\n 0.7168, 0.7637, 0.7046, 0.7744, 0.6909, 0.7456, 0.7427, 0.7241, 0.7476, 0.7271, 0.7100, 0.7207, 0.6948, 0.6982, 0.6636, 0.7173, 0.7153, 0.6807, 0.7012, 0.7363, 0.6909, 0.7144, 0.7241, 0.7251, 0.5645, 0.7090, 0.7051], dtype=torch.float16), 'perceiver_resampler.layers.0.0.norm2.bias': tensor([ 6.2744e-02, 1.2494e-01, 1.9562e-02, 9.8572e-02, 5.7404e-02, -8.5220e-03, -2.5757e-02, -1.7310e-01, 2.6226e-03, 8.9874e-03, -4.2603e-02, 2.9724e-02, -7.3303e-02, -7.7095e-03, -2.7817e-02, -4.1901e-02, -2.4231e-02, -5.2032e-03, -1.0309e-01, 5.0278e-03, -8.2642e-02, -6.0822e-02, 2.2186e-02, -7.5378e-02,\n 4.3152e-02, 2.1622e-02, -6.7688e-02, 1.1200e-01, 7.7881e-02, -7.8064e-02, -5.8319e-02, -3.0640e-02, 2.1790e-02, -7.4219e-02, -1.0139e-02, 2.5543e-02, -9.1064e-02, 4.8676e-03, 4.8309e-02, 9.8724e-03, 5.6982e-04, -1.6144e-02, -4.3610e-02, 2.0561e-03, -6.0150e-02, 3.1052e-02, 3.4515e-02, -9.3201e-02,\n -4.4281e-02, 1.5610e-02, 1.7792e-02, 2.6047e-02, -6.0150e-02, 7.4425e-03, -1.9485e-02, -3.5950e-02, -4.1351e-02, -4.9133e-02, 1.0388e-01, -1.2291e-02, -9.2697e-03, -6.8176e-02, 4.2763e-03, 2.7191e-02, 8.7036e-02, -7.3181e-02, -4.9500e-02, 6.2622e-02, 3.0975e-02, -3.5156e-02, -3.2410e-02, -7.6660e-02,\n 4.1870e-02, -1.1389e-01, -9.9087e-04, 3.8574e-02, 1.7395e-02, 6.2134e-02, -9.4070e-03, 2.9099e-02, 2.7740e-02, 4.6875e-02, -1.7715e-02, 2.9938e-02, -5.3497e-02, -3.5217e-02, 3.8391e-02, 6.9702e-02, -3.6987e-02, 6.8542e-02, -2.9099e-02, 5.5176e-02, -2.9846e-02, -4.7424e-02, 3.5431e-02, 1.1177e-02,\n 3.9795e-02, -3.6469e-02, -4.2419e-03, -5.0568e-02, 5.9662e-02, -5.8899e-02, -9.8267e-02, 2.9236e-02, -1.0361e-02, -8.5876e-02, -8.3801e-02, -2.3880e-02, -5.6885e-02, -2.1434e-04, 2.2430e-02, 7.0618e-02, 9.3811e-02, 7.2754e-02, -9.2896e-02, -1.6851e-03, -5.1086e-02, 3.1021e-02, 5.2673e-02, 3.1250e-02,\n 9.1858e-03, -8.7158e-02, 1.5480e-02, 5.4131e-03, 3.2562e-02, -8.6365e-02, 3.2196e-02, 3.5706e-02, -3.2253e-03, -1.2527e-02, -3.7750e-02, -4.8920e-02, 1.0382e-01, -6.6467e-02, 7.0007e-02, -2.2182e-03, 4.2114e-03, -5.0049e-02, -1.8024e-03, -2.0676e-02, 1.4793e-02, -8.7646e-02, -2.3956e-02, -7.8918e-02,\n 5.4260e-02, 2.6871e-02, 4.2938e-02, 8.0109e-03, -7.4951e-02, 2.4216e-02, 3.5187e-02, -6.4209e-02, -1.3367e-02, 1.0406e-01, 1.2573e-02, 8.2214e-02, 6.7688e-02, -4.0985e-02, -2.1572e-03, -6.5125e-02, 7.1106e-02, 2.9354e-03, 2.0340e-02, 1.1902e-02, -2.8443e-04, 4.2389e-02, 1.4259e-02, -2.0126e-02,\n -2.5467e-02, -7.9102e-02, -6.0539e-03, 3.7231e-02, 5.2216e-02, -3.3356e-02, 2.3209e-02, -4.5349e-02, 2.5787e-02, -7.2937e-02, 1.8478e-02, -3.0121e-02, -9.5581e-02, -7.8308e-02, -5.6061e-02, 2.3621e-02, 7.3624e-03, 2.9678e-02, 3.3131e-03, 1.8042e-01, 3.6835e-02, -2.7023e-02, -9.4421e-02, -4.2305e-03,\n -1.8112e-02, 3.5126e-02, 1.0547e-01, 2.5574e-02, 9.6130e-02, -7.1899e-02, -1.1432e-01, 5.2582e-02, 1.2006e-01, -8.0688e-02, 2.4509e-03, 2.3926e-02, -2.9800e-02, -1.8326e-02, -2.6436e-03, -3.6652e-02, 7.7629e-03, 1.3831e-01, -3.9703e-02, -1.0413e-01, 3.8239e-02, 4.3823e-02, 2.3682e-01, -4.2755e-02,\n -4.0619e-02, 5.7617e-02, 3.3203e-02, 3.0823e-03, 2.2980e-02, 9.4849e-02, -4.4617e-02, 6.3744e-03, 6.7566e-02, 1.2543e-02, -8.4473e-02, 3.1643e-03, -6.2195e-02, -2.3758e-02, 9.1431e-02, -7.2449e-02, 4.5319e-02, 3.5248e-02, 2.0081e-02, 7.8735e-02, -4.6997e-02, -2.9556e-02, 3.8147e-02, -1.6068e-02,\n -3.7201e-02, -1.1621e-01, -9.5215e-02, -4.0817e-03, 1.6968e-02, -5.1208e-02, 4.4037e-02, -1.6298e-03, -7.3303e-02, -6.7253e-03, -2.5421e-02, -5.0201e-02, -9.4482e-02, -4.3884e-02, 1.0626e-01, 1.0242e-01, 6.8115e-02, 2.6550e-02, 1.2524e-01, -4.8798e-02, 1.0535e-01, 1.8750e-01, 8.2764e-02, 9.8999e-02,\n -4.8248e-02, 1.2463e-01, 3.5309e-02, -3.6774e-02, -2.0844e-02, 1.0468e-01, -2.2171e-02, -9.2834e-02, 9.8145e-02, 3.8452e-03, -4.6753e-02, 4.7882e-02, -4.6509e-02, 1.3329e-02, -2.9266e-02, 2.4094e-02, 4.6143e-02, -4.2633e-02, 4.5013e-03, 4.9835e-02, -3.6812e-03, 3.7628e-02, 3.8208e-02, -8.4991e-03,\n -5.0568e-02, -1.3565e-02, -1.9135e-02, -5.8746e-02, -2.5604e-02, -3.3478e-02, 8.1970e-02, 3.1052e-02, -3.8269e-02, 3.5645e-02, -5.2124e-02, -1.3870e-02, 1.5778e-02, 3.0289e-03, -6.5727e-03, -2.2705e-01, -4.7546e-02, 6.8054e-02, 6.6528e-02, -3.0807e-02, -5.7449e-03, 1.7654e-02, -6.5918e-02, -8.5693e-02,\n 7.7332e-02, 2.3468e-02, -1.0327e-01, -8.7524e-02, 4.9164e-02, 2.3911e-02, 6.1249e-02, -1.5343e-02, 5.5969e-02, 1.4336e-02, -1.0635e-02, 2.1835e-02, 5.7098e-02, -2.2980e-02, 1.6647e-02, -3.4428e-03, -3.3512e-03, 1.1078e-01, 1.4612e-01, -7.7438e-03, 7.7087e-02, -9.4666e-02, 3.9215e-02, 9.1064e-02,\n 3.6102e-02, 7.1220e-03, 2.2873e-02, -1.9043e-02, 6.2042e-02, -6.9153e-02, 1.0681e-01, 9.3018e-02, -8.0261e-02, -3.9764e-02, -4.9286e-02, -6.1310e-02, -4.9408e-02, -1.2962e-02, 9.5444e-03, 5.5511e-02, 1.0699e-01, -1.1206e-01, 3.6011e-02, -1.7349e-02, -7.9529e-02, -1.1528e-02, 2.1866e-02, 8.0383e-02,\n 9.0179e-03, 1.2970e-02, -5.7434e-02, 2.8168e-02, -1.6739e-02, 2.7954e-02, -6.9946e-02, -1.0773e-01, -6.7322e-02, 1.0364e-01, 7.5317e-02, -3.9612e-02, -3.5229e-03, 4.6875e-02, 2.1851e-02, 3.6133e-02, -9.2163e-02, -1.1792e-01, 7.2449e-02, -6.5918e-02, -2.5864e-03, 1.0712e-01, -3.0426e-02, -9.2590e-02,\n -3.9703e-02, -6.4148e-02, 1.8631e-02, 8.7214e-04, -5.1003e-03, 2.7023e-02, 2.4963e-02, 8.7952e-02, 2.1915e-03, -3.5828e-02, -8.7158e-02, 9.0637e-02, 2.4261e-03, -1.2650e-02, -9.2850e-03, -2.6283e-03, 2.8122e-02, -2.9465e-02, -2.8824e-02, 1.5121e-02, -3.4943e-02, 6.6895e-02, -3.1433e-02, 5.9326e-02,\n 6.3416e-02, 1.3294e-03, -5.1910e-02, 7.8491e-02, 3.9642e-02, -2.6794e-02, -1.4148e-01, -1.1505e-01, -7.4280e-02, 2.6520e-02, -7.0923e-02, -1.9379e-02, 1.9989e-03, 8.2886e-02, -3.3234e-02, 5.5046e-03, 1.5979e-01, -2.4994e-02, 7.2021e-02, 2.8091e-02, 7.4997e-03, -3.7506e-02, 4.3213e-02, -3.2196e-02,\n -7.1594e-02, 5.8411e-02, -1.8417e-02, -8.7051e-03, 3.8269e-02, 4.3564e-03, 4.8462e-02, 3.2272e-03, -1.9562e-02, -6.8787e-02, 1.7593e-02, -2.4338e-03, 6.0181e-02, -1.3535e-02, 5.9875e-02, -8.4351e-02, 7.1533e-02, -2.0897e-04, 4.0802e-02, 4.2053e-02, -9.3307e-03, 2.4323e-02, -6.7215e-03, 1.7960e-02,\n 1.3086e-01, 3.3020e-02, 2.9663e-02, -7.2517e-03, -5.9784e-02, 7.8857e-02, 2.4658e-02, -1.1023e-01, 1.8066e-02, 4.7699e-02, 5.5542e-02, -3.3142e-02, 2.4748e-04, 5.2277e-02, 2.3178e-02, -4.0131e-02, -5.3558e-02, 7.0076e-03, 4.8615e-02, 9.1553e-02, -7.2693e-02, -1.2848e-02, 9.7561e-04, 6.2065e-03,\n -3.8910e-02, 9.9548e-02, -2.0721e-02, 1.1810e-01, 4.0436e-02, -5.1544e-02, -4.2664e-02, -5.8365e-03, -9.0088e-02, -8.9264e-03, 2.5940e-02, -6.7635e-03, -2.0660e-02, 7.7271e-02, 4.9171e-03, -5.8014e-02, -5.0385e-02, 3.1586e-03, 8.5526e-03, 2.6459e-02, 3.0334e-02, 5.4749e-02, -5.6572e-03, -1.9287e-02,\n -4.7882e-02, -3.0441e-02, -8.1055e-02, -8.4778e-02, 1.4124e-01, -4.3793e-02, 4.9896e-02, -3.7231e-02, 3.7170e-02, -1.5326e-03, -6.1737e-02, -8.1253e-03, -1.1663e-03, 1.0675e-01, -1.6724e-02, -6.4453e-02, -2.6550e-02, 5.9326e-02, 2.0081e-01, 6.8237e-02, -8.6914e-02, 6.4758e-02, 1.2976e-01, -9.8572e-02,\n -3.1464e-02, -9.1324e-03, -4.7188e-03, -1.7676e-01, -5.8411e-02, 1.4618e-02, -1.6418e-02, -7.3120e-02, 4.4159e-02, -3.7872e-02, -1.4137e-02, -1.2672e-02, 6.6185e-03, -4.5929e-02, 5.8228e-02, 5.5115e-02, 5.3162e-02, 3.4637e-02, -8.6670e-03, 2.5726e-02, -1.3626e-02, -4.2358e-02, 3.3020e-02, 5.8563e-02,\n 3.5919e-02, -7.1907e-03, -6.7444e-02, 4.7028e-02, 8.6641e-04, 1.0266e-01, -1.0522e-01, 1.0262e-02, 2.0981e-02, -7.7019e-03, -2.3834e-02, -1.1757e-02, -6.7673e-03, -2.2797e-02, 3.3112e-02, -1.4453e-01, 1.9547e-02, 7.7637e-02, 6.8115e-02, -1.9073e-02, 2.7145e-02, 6.1493e-02, -3.1525e-02, -6.0883e-02,\n 3.8605e-02, -4.4189e-02, -5.7739e-02, 5.4871e-02, 6.0760e-02, -1.3153e-02, 1.2927e-01, -2.1698e-02, -6.7177e-03, 1.3672e-01, -1.6602e-02, 8.3252e-02, -9.0515e-02, -1.8402e-02, -4.8828e-02, 2.5436e-02, 2.9526e-02, -6.3965e-02, 2.0126e-02, 8.9294e-02, 2.6825e-02, -3.0624e-02, -2.9572e-02, -4.0649e-02,\n 8.9417e-02, -5.0781e-02, 5.2124e-02, -8.0444e-02, 2.4445e-02, 1.8784e-02, 4.8447e-03, -1.4679e-02, -3.6865e-02, 6.0913e-02, 9.6863e-02, 2.6306e-02, -3.5706e-02, -7.0679e-02, 5.3375e-02, 3.7720e-02, -6.1615e-02, -1.7960e-02, 3.3630e-02, -2.0020e-02, -2.4204e-03, -5.1117e-02, -4.3335e-02, -6.6223e-03,\n -1.7212e-02, -1.8860e-02, -2.5681e-02, 2.3155e-03, 3.8086e-02, -2.2781e-02, -6.3782e-02, 4.6570e-02, -2.2659e-02, -1.2500e-01, 5.6824e-02, -2.6047e-02, 6.3049e-02, 1.3501e-01, 3.8116e-02, 3.9856e-02, 6.8970e-03, 1.8295e-02, -5.3772e-02, 4.7150e-02, -3.5034e-02, 7.6065e-03, -1.5601e-01, -2.1545e-02,\n 2.7313e-02, 3.0060e-02, -4.3884e-02, -5.8838e-02, 4.4128e-02, -3.0182e-02, -5.2643e-04, 5.4626e-02, 1.1353e-02, 8.2458e-02, 8.9355e-02, 2.5055e-02, 4.0314e-02, 1.3313e-02, 1.9531e-02, 9.0820e-02, -2.8854e-02, 2.7161e-02, 1.0971e-02, 3.8727e-02, 2.5345e-02, 5.1422e-02, -5.5542e-02, 5.1727e-02,\n -4.6448e-02, -2.9449e-02, 2.3941e-02, -1.3916e-02, 1.0956e-02, -2.0035e-02, 4.7241e-02, 3.8116e-02, -2.3880e-02, -1.0992e-01, -4.1077e-02, 1.0788e-02, -2.1774e-02, 3.7872e-02, 5.8685e-02, -2.0645e-02, 8.6182e-02, 8.8623e-02, 1.8143e-02, 1.2756e-02, -8.2581e-02, 2.8397e-02, -6.7505e-02, -9.5367e-03,\n -8.3496e-02, 1.3451e-02, 4.2542e-02, -6.1493e-02, 2.3022e-03, 1.4258e-01, 8.5815e-02, -7.6355e-02, 5.5145e-02, -6.0028e-02, 2.8137e-02, -3.6102e-02, 9.4238e-02, 7.5378e-02, -5.2277e-02, 1.2413e-02, -5.0842e-02, 4.2175e-02, 8.6060e-02, -3.0106e-02, 4.0405e-02, 9.2896e-02, 6.5979e-02, -9.1934e-03,\n 3.6072e-02, -1.0002e-02, -4.6722e-02, -4.0359e-03, -5.1605e-02, 2.9388e-02, -5.4260e-02, -3.3569e-02, -5.0110e-02, 2.9327e-02, 1.4870e-02, 5.2460e-02, 2.5406e-02, 3.8605e-03, 8.8425e-03, 1.2955e-02, 2.8259e-02, 1.2390e-02, -8.3862e-02, -4.8370e-02, -6.0455e-02, 7.6843e-02, -5.5206e-02, -4.8553e-02,\n 1.8860e-02, 3.6255e-02, 3.9856e-02, -1.7624e-02, -4.7951e-03, -7.7942e-02, -2.7145e-02, 5.1558e-05, 1.5015e-02, 2.7954e-02, 1.2451e-02, 2.8305e-02, -2.5345e-02, 7.3303e-02, -6.9946e-02, 4.4098e-02, -9.4528e-03, 1.2152e-01, -4.3274e-02, 3.1647e-02, -3.0350e-02, -9.4116e-02, 7.0801e-02, -2.4948e-02],\n dtype=torch.float16), 'perceiver_resampler.layers.0.0.to_q.weight': tensor([[-0.0090, -0.0633, -0.0323, ..., -0.0058, 0.0754, -0.0359],\n [ 0.0576, 0.0070, 0.0030, ..., -0.0203, 0.0003, 0.0314],\n [-0.0240, 0.0208, -0.0360, ..., -0.0010, -0.0404, 0.0696],\n ...,\n [ 0.0079, 0.0317, -0.0323, ..., 0.0110, -0.0570, -0.0291],\n [ 0.0416, 0.0310, -0.0346, ..., 0.0112, 0.0074, -0.0094],\n [ 0.0048, -0.0309, -0.0564, ..., -0.0239, 0.0063, 0.0325]], dtype=torch.float16), 'perceiver_resampler.layers.0.0.to_kv.weight': tensor([[-0.0353, 0.0108, 0.0043, ..., 0.0024, 0.0720, -0.0053],\n [ 0.0175, -0.0551, -0.0290, ..., 0.0240, 0.0048, -0.0214],\n [-0.0016, -0.0078, -0.0381, ..., -0.0343, -0.0310, 0.0211],\n ...,\n [-0.0146, -0.0365, 0.0404, ..., -0.0135, -0.0226, 0.0115],\n [-0.0023, -0.0291, -0.0297, ..., -0.0549, 0.0235, 0.0103],\n [ 0.0178, 0.0337, 0.0328, ..., -0.0328, -0.0355, 0.0657]], dtype=torch.float16), 'perceiver_resampler.layers.0.0.to_out.weight': tensor([[ 3.1769e-02, 3.4637e-02, 1.4778e-02, ..., -8.3847e-03, 4.4365e-03, -3.6743e-02],\n [ 1.6632e-03, 1.1009e-02, 9.5010e-05, ..., 2.7039e-02, -3.8177e-02, -4.6021e-02],\n [-7.1899e-02, -2.6978e-02, -3.1036e-02, ..., 2.8931e-02, 5.8937e-03, 2.2568e-02],\n ...,\n [ 6.5979e-02, 3.1830e-02, 3.9734e-02, ..., 6.0974e-02, 2.3819e-02, 2.4277e-02],\n [ 1.1757e-02, -1.7792e-02, -1.1391e-02, ..., -2.7420e-02, -8.9233e-02, 1.5574e-03],\n [-2.1072e-02, 3.4332e-02, -1.1040e-02, ..., -2.1652e-02, 7.9803e-03, -3.6530e-02]], dtype=torch.float16), 'perceiver_resampler.layers.0.1.0.weight': tensor([0.8110, 0.7798, 0.8525, 0.8252, 0.7432, 0.8491, 0.8359, 0.7935, 0.8652, 0.8887, 0.8428, 0.7935, 0.8750, 0.7651, 0.8945, 0.8140, 0.8662, 0.8711, 0.7847, 0.7578, 0.8643, 0.8003, 0.7910, 0.8076, 0.7627, 0.9404, 0.7866, 0.8784, 0.8638, 0.8564, 0.8135, 0.8325, 0.8164, 0.8125, 0.8169, 0.8311, 0.8311, 0.8447, 0.8145,\n 0.8398, 0.8271, 0.7944, 0.7842, 0.9062, 0.7900, 0.7837, 0.8237, 0.8564, 0.8032, 0.9136, 0.8770, 0.8447, 0.8242, 0.8516, 0.7910, 0.8735, 0.7593, 0.7461, 0.9155, 0.8384, 0.7485, 0.8428, 0.9170, 0.8921, 0.8789, 0.8447, 0.8452, 0.8013, 0.8774, 0.8813, 0.7539, 0.8340, 0.8823, 0.7646, 0.8076, 0.8340, 0.8652, 0.7363,\n 0.8242, 0.7954, 0.8281, 0.7354, 0.8672, 0.8936, 0.8813, 0.8438, 0.8711, 0.8398, 0.8809, 0.7998, 0.8799, 0.7861, 0.8950, 0.9248, 0.8315, 0.8687, 0.7920, 0.8940, 0.7783, 0.7949, 0.8408, 0.8569, 0.7876, 0.8169, 0.8872, 0.8823, 0.8257, 0.8247, 0.8853, 0.7544, 0.8716, 0.9253, 0.7031, 0.8105, 0.8135, 0.8765, 0.8760,\n 0.7979, 0.8950, 0.7798, 0.8711, 0.9233, 0.8491, 0.9185, 0.7729, 0.8286, 0.7729, 0.7793, 0.8452, 0.8921, 0.8364, 0.7935, 0.8198, 0.8203, 0.8696, 0.8169, 0.8315, 0.9077, 0.9326, 0.8232, 0.8237, 0.8418, 0.8564, 0.7827, 0.8408, 0.8716, 0.7676, 0.8931, 0.8530, 0.8047, 0.9023, 0.7910, 0.8525, 0.8179, 0.8613, 0.8252,\n 0.8545, 0.8501, 0.7676, 0.8379, 0.9238, 0.7847, 0.8667, 0.7100, 0.7969, 0.8877, 0.8604, 0.8091, 0.8589, 0.8521, 0.7939, 0.8569, 0.8433, 0.8506, 0.8999, 0.8042, 0.8096, 0.8350, 0.8633, 0.8550, 0.8804, 0.9185, 0.7866, 0.7842, 0.7896, 0.8970, 0.8730, 0.8428, 0.8613, 0.8379, 0.8325, 0.8110, 0.8457, 0.8887, 0.8496,\n 0.8936, 0.8330, 0.8574, 0.8638, 0.7925, 0.8120, 0.8721, 0.8340, 0.8540, 0.7231, 0.8677, 0.9141, 0.8940, 0.9248, 0.8091, 0.8281, 0.6743, 0.7900, 0.9121, 0.5332, 0.9189, 0.7832, 0.8506, 0.8462, 0.8267, 0.8228, 0.8486, 0.9019, 0.8750, 0.8618, 0.8633, 0.8203, 0.8315, 0.8501, 0.8247, 0.7983, 0.8335, 0.8110, 0.8711,\n 0.8838, 0.8071, 0.8218, 0.8574, 0.8657, 0.8735, 0.7881, 0.6631, 0.8281, 0.8228, 0.8774, 0.8345, 0.8770, 0.8135, 0.8203, 0.8770, 0.7114, 0.8315, 0.7524, 0.9028, 0.6904, 0.8945, 0.8975, 0.7593, 0.7935, 0.8579, 0.8374, 0.5820, 0.7861, 0.8418, 0.8169, 0.7358, 0.8262, 0.7886, 0.8022, 0.8325, 0.7739, 0.7832, 0.7959,\n 0.8994, 0.8384, 0.8770, 0.8501, 0.8262, 0.8516, 0.8799, 0.8081, 0.6362, 0.8135, 0.8496, 0.7930, 0.8188, 0.7676, 0.9058, 0.8398, 0.8081, 0.8408, 0.7817, 0.8574, 0.8574, 0.8193, 0.8213, 0.7812, 0.9468, 0.8516, 0.8379, 0.8169, 0.7998, 0.8042, 0.5557, 0.8164, 0.6392, 0.8149, 0.8042, 0.8086, 0.6216, 0.8457, 0.5503,\n 0.7935, 0.8364, 0.8745, 0.8037, 0.8271, 0.7988, 0.8496, 0.8906, 0.9165, 0.8315, 0.7866, 0.8301, 0.9478, 0.9077, 0.8687, 0.8228, 0.8130, 0.8413, 0.6938, 0.8755, 0.9072, 0.8228, 0.8730, 0.7773, 0.8335, 0.8477, 0.8594, 0.8672, 0.8662, 0.8545, 0.8369, 0.8447, 0.8433, 0.8281, 0.9326, 0.8389, 0.8203, 0.7651, 0.9697,\n 0.7954, 0.7720, 0.8584, 0.8643, 0.8682, 0.8135, 0.8252, 0.8022, 0.8188, 0.8042, 0.8340, 0.8394, 0.8203, 0.8755, 0.7847, 0.8037, 0.8706, 0.8433, 0.8018, 0.6963, 0.8696, 0.8853, 0.8486, 0.8857, 0.8228, 0.7832, 0.7881, 0.7363, 0.8096, 0.7847, 0.8525, 0.8521, 0.8398, 0.8813, 0.8882, 0.8862, 0.8218, 0.8379, 0.8975,\n 0.8838, 0.8057, 0.8687, 0.8286, 0.7959, 0.8066, 0.9375, 0.9307, 0.8027, 0.8896, 0.7290, 0.7500, 0.8345, 0.9204, 0.9072, 0.8511, 0.8066, 0.8672, 0.9121, 0.8403, 0.8325, 0.7109, 0.8311, 0.6689, 0.5186, 0.7886, 0.8462, 0.8252, 0.8652, 0.8779, 0.8608, 0.9160, 0.7886, 0.8423, 0.7549, 0.9019, 0.8237, 0.8628, 0.7754,\n 0.8335, 0.8149, 0.8711, 0.7729, 0.8604, 0.8359, 0.8125, 0.7559, 0.9214, 0.7358, 0.8804, 0.8652, 0.8242, 0.7969, 0.8833, 0.8623, 0.8613, 0.7969, 0.8379, 0.8774, 0.7930, 0.7993, 0.8511, 0.8965, 0.8237, 0.8062, 0.7954, 0.7534, 0.7661, 0.8276, 0.8335, 0.8359, 0.8516, 0.8555, 0.8154, 0.8364, 0.8003, 0.8027, 0.9033,\n 0.8511, 0.8945, 0.9248, 0.7837, 0.8423, 0.8599, 0.7236, 0.8306, 0.8623, 0.9707, 0.8579, 0.7974, 0.8101, 0.7690, 0.8369, 0.8335, 0.7832, 0.6875, 0.8442, 0.8262, 0.8037, 0.8608, 0.8970, 0.7944, 0.7681, 0.8442, 0.8149, 0.8413, 0.8926, 0.8037, 0.8315, 0.8452, 0.7949, 0.8291, 0.8496, 0.7383, 0.8325, 0.8203, 0.8276,\n 0.8179, 0.6626, 0.8350, 0.8345, 0.8872, 0.7905, 0.8843, 0.8384, 0.9111, 0.8750, 0.8579, 0.8281, 0.9233, 0.7949, 0.8555, 0.5952, 0.7695, 0.6938, 0.8105, 0.7734, 0.7930, 0.8003, 0.8091, 0.8491, 0.4612, 0.9106, 0.8022, 0.8320, 0.7954, 0.7974, 0.7808, 0.9136, 0.8218, 0.8418, 0.8071, 0.8105, 0.8022, 0.7163, 0.8179,\n 0.8813, 0.8364, 0.8115, 0.8716, 0.8071, 0.8955, 0.9033, 0.8652, 0.8867, 0.8760, 0.8818, 0.8511, 0.8149, 0.8174, 0.8374, 0.8115, 0.8374, 0.8711, 0.9004, 0.9121, 0.7520, 0.7993, 0.8203, 0.7700, 0.9155, 0.8999, 0.8364, 0.8755, 0.7944, 0.7739, 0.7939, 0.8506, 0.8652, 0.7949, 0.8584, 0.9146, 0.8789, 0.8691, 0.8452,\n 0.6919, 0.8379, 0.8022, 0.8511, 0.8340, 0.8022, 0.9160, 0.8071, 0.8862, 0.9097, 0.8096, 0.8340, 0.8770, 0.9014, 0.8467, 0.8228, 0.8350, 0.8350, 0.8013, 0.8345, 0.8066, 0.7817, 0.8379, 0.8872, 0.8628, 0.7861, 0.8516, 0.8242, 0.8062, 0.8276, 0.8101, 0.8120, 0.8193, 0.7847, 0.8276, 0.8892, 0.8438, 0.8369, 0.8618,\n 0.8682, 0.8823, 0.8516, 0.8711, 0.7271, 0.7505, 0.8364, 0.8765, 0.8696, 0.8081, 0.8188, 0.8027, 0.7876, 0.8301, 0.7036, 0.7529, 0.8682, 0.8770, 0.8652, 0.8667, 0.9019, 0.8462, 0.8125, 0.8115, 0.8125, 0.8843, 0.7954, 0.8394, 0.8418, 0.7881, 0.8892, 0.8340, 0.8438, 0.8555, 0.8208, 0.8452, 0.8213, 0.8481, 0.8140,\n 0.7632, 0.8237, 0.8750, 0.8477, 0.8755, 0.7559, 0.8218, 0.9019, 0.9146, 0.7832, 0.8506, 0.8799, 0.8682, 0.8843, 0.8828, 0.7437, 0.8506, 0.7915, 0.8530, 0.9189, 0.8550, 0.7949, 0.8789, 0.8179, 0.8281, 0.7905, 0.7427, 0.8760, 0.8506, 0.8218, 0.8003, 0.8765, 0.8979, 0.8174, 0.8638, 0.8413, 0.6631, 0.8174, 0.7778,\n 0.8398, 0.8926, 0.7832, 0.9419, 0.8672, 0.8589, 0.8867, 0.8535, 0.7939, 0.8696, 0.8076, 0.8101, 0.8379, 0.8555, 0.8682, 0.8564, 0.7988, 0.7974, 0.8203, 0.7700, 0.8325, 0.8091, 0.8535, 0.8237, 0.8408, 0.8301, 0.7710, 0.8154, 0.8027, 0.7866, 0.8052, 0.7231, 0.7729, 0.9116, 0.8472, 0.8618, 0.8276, 0.8354, 0.8354,\n 0.7871, 0.8232, 0.8345, 0.8179, 0.8403, 0.8511, 0.8574, 0.7793, 0.8618, 0.9312, 0.8198, 0.7700, 0.8350, 0.9038, 0.7554, 0.7935, 0.8291, 0.8838, 0.8467, 0.8667, 0.8271, 0.8633, 0.8374, 0.8750, 0.5581, 0.8184, 0.8115], dtype=torch.float16), 'perceiver_resampler.layers.0.1.0.bias': tensor([ 1.1194e-01, 5.2551e-02, 2.6608e-03, 7.2083e-02, -4.6158e-03, -5.8685e-02, 1.2878e-01, 2.8168e-02, -2.2369e-02, 7.5150e-03, -1.9116e-01, 1.4966e-01, -1.2695e-01, 2.6581e-02, -8.2275e-02, -4.5441e-02, 4.8126e-02, -1.9031e-01, -3.0273e-02, 1.1407e-01, 1.2000e-01, -8.9539e-02, 1.1572e-01, -2.1008e-01,\n 1.7371e-01, -1.3696e-01, -1.0016e-01, 1.3696e-01, 1.5841e-03, -1.0181e-01, 1.1493e-01, 9.3323e-02, 1.7847e-01, -4.6814e-02, -9.8190e-03, -6.5918e-02, -1.2903e-01, -1.7136e-02, 1.8237e-01, 6.5918e-02, 6.9092e-02, 9.3567e-02, 4.3823e-02, -3.2990e-02, -1.7139e-01, 1.2489e-02, 9.6464e-04, -1.8286e-01,\n 6.3293e-02, 2.6665e-03, 5.4718e-02, -1.2195e-01, -1.0565e-01, -2.3819e-02, -1.7798e-01, -7.5073e-02, 3.0502e-02, -7.2693e-02, 7.1960e-02, -1.3037e-01, 5.9631e-02, -3.2227e-02, -1.0967e-03, 6.3286e-03, 1.5967e-01, -4.5624e-02, -4.5258e-02, -1.6406e-01, -8.8867e-02, -1.5344e-01, 3.9520e-02, -1.7059e-02,\n 1.1145e-01, 4.0253e-02, 1.5961e-02, -6.1920e-02, 1.6467e-01, 1.3135e-01, -6.3904e-02, 5.7831e-02, -1.2408e-01, 9.2224e-02, -1.5710e-01, -4.8706e-02, 2.1057e-02, -1.8509e-02, 5.2032e-03, 4.1382e-02, 9.2834e-02, 2.0312e-01, -1.5930e-01, 5.4230e-02, -1.5511e-02, 3.0426e-02, 1.1926e-01, 4.4189e-02,\n -1.5282e-02, -7.1526e-03, 2.1149e-02, -5.4657e-02, -2.7512e-02, -1.1810e-01, -1.6342e-02, 1.2573e-01, -1.6272e-01, -3.1342e-02, 1.5869e-02, 1.0834e-01, -6.1798e-03, -5.7037e-02, 6.0974e-02, -1.1002e-02, 8.4412e-02, 2.4582e-02, -5.6152e-02, -6.1615e-02, -5.7190e-02, 7.6477e-02, 8.9966e-02, -4.8141e-03,\n 1.7502e-02, 1.3477e-01, -3.6346e-02, 7.8659e-03, 1.1749e-01, -9.7107e-02, -8.4457e-03, -1.9562e-02, -9.5093e-02, 1.3940e-01, 7.3792e-02, -1.2952e-01, 3.0197e-02, -1.6571e-02, 1.6602e-01, -6.7078e-02, 1.5808e-01, -2.2766e-01, 3.3894e-03, -4.4220e-02, 3.1067e-02, 1.4282e-02, -8.0566e-02, -1.1865e-01,\n 1.8213e-01, 1.4062e-01, -2.4796e-02, -9.3323e-02, -1.3892e-01, 5.4169e-02, 9.2224e-02, -7.4707e-02, 1.4978e-01, 1.0669e-01, 5.5817e-02, 2.0154e-01, -1.9409e-01, -4.4952e-02, -2.0007e-01, -9.3018e-02, -4.6936e-02, 4.5929e-02, 1.0541e-01, 2.2766e-01, -4.5776e-02, 9.9243e-02, 1.1469e-01, 2.6886e-02,\n -8.0078e-02, -2.0471e-01, 6.1646e-02, -3.4363e-02, 5.7068e-02, 3.6438e-02, 6.3477e-02, 3.6530e-02, 8.6060e-02, -1.3440e-01, -2.9053e-02, 1.2427e-01, -5.1697e-02, -8.0566e-02, -9.1492e-02, 1.9073e-02, -1.0614e-01, 1.1676e-01, -2.6392e-01, 2.5806e-01, -7.3059e-02, -1.3489e-01, 6.3416e-02, -4.7326e-04,\n -1.1279e-01, 4.5380e-02, 1.0284e-01, 1.2610e-01, -1.0797e-01, 1.4307e-01, -2.0496e-01, 9.0637e-02, 3.0289e-02, -8.8928e-02, -9.8633e-02, 1.9775e-02, -4.1443e-02, -1.0590e-02, 1.4404e-01, 8.4595e-02, -7.4585e-02, 1.0651e-01, -1.0455e-01, -1.7044e-02, 1.2764e-02, 2.2351e-01, -1.9641e-01, -9.6130e-02,\n -5.8807e-02, 1.6980e-01, 1.5572e-02, -1.1481e-01, 1.1768e-01, 5.4077e-02, 5.6274e-02, -6.2622e-02, 3.0060e-02, -3.8300e-02, -1.3770e-01, -4.3427e-02, 1.7383e-01, -4.3030e-02, 1.0510e-01, -1.6492e-01, -9.6069e-02, -1.2769e-01, 9.0149e-02, -4.9164e-02, 9.4604e-02, 1.5930e-01, -2.0032e-01, -1.8640e-01,\n -1.1053e-01, -1.2009e-02, -1.6174e-02, -1.0229e-01, 3.7048e-02, 3.7598e-02, -1.6880e-03, -5.1147e-02, 1.1658e-01, -1.6321e-01, -4.8737e-02, -9.5947e-02, -1.4270e-01, -9.9609e-02, -1.2769e-01, 1.2622e-01, 5.1636e-02, -1.1322e-02, -7.2754e-02, -8.6548e-02, 1.8213e-01, -1.7957e-01, -5.5176e-02, 1.1032e-02,\n 6.4209e-02, -1.6504e-01, -3.9635e-03, -6.5674e-02, -1.2524e-01, -7.0129e-02, -3.6743e-02, -1.6602e-01, -7.6485e-03, 8.8013e-02, -1.6260e-01, 5.6793e-02, -7.5989e-02, -5.1727e-02, -5.0049e-02, -4.8340e-02, -3.5034e-02, 2.0532e-01, -1.6895e-01, 9.0759e-02, 9.2407e-02, -1.1511e-01, 1.2695e-01, 1.4551e-01,\n -1.2939e-01, -7.2388e-02, 5.8807e-02, 1.5015e-02, 4.9042e-02, 9.8450e-02, -4.3915e-02, -1.0938e-01, 7.8491e-02, -1.7920e-01, -7.3181e-02, 6.2790e-03, 3.2864e-03, 6.1005e-02, 6.1859e-02, 1.6736e-01, -5.2582e-02, -7.9285e-02, -1.1415e-03, -1.7017e-01, -3.2593e-02, 1.6150e-01, 8.5938e-02, 1.8518e-01,\n 8.4473e-02, -4.9774e-02, -7.4829e-02, -1.5308e-01, 8.1177e-02, 9.6985e-02, 2.0288e-01, -1.5173e-01, -1.5784e-01, 1.6772e-01, 6.0974e-02, 1.4502e-01, 1.9690e-01, -1.0352e-01, -1.6272e-01, 1.0449e-01, 4.5074e-02, -5.7709e-02, -1.3525e-01, 2.3636e-02, -7.0618e-02, -1.8030e-01, -1.5244e-02, -6.4758e-02,\n -2.3026e-02, 8.6853e-02, -2.2469e-03, 1.2537e-01, -1.0858e-01, 4.0039e-02, 2.2290e-01, -2.3712e-02, -1.2073e-01, -1.6284e-01, -3.9612e-02, -1.0168e-01, -5.1910e-02, 9.5093e-02, 1.2622e-01, 5.5023e-02, 8.2092e-02, -3.8086e-02, -1.4478e-01, 5.4138e-02, -8.0139e-02, 5.0507e-02, -1.9363e-02, 7.1228e-02,\n 1.2732e-01, 4.8218e-03, 8.1116e-02, -1.1273e-01, 1.1908e-01, -2.3209e-02, 1.8036e-02, -8.8928e-02, 1.9272e-02, 3.0304e-02, 7.7820e-02, 2.0966e-02, 1.2274e-01, 1.3440e-01, -3.8177e-02, -5.6427e-02, 6.4880e-02, 5.7526e-02, 1.2903e-01, -2.1622e-02, 1.1658e-02, 1.1940e-02, 7.1960e-02, -3.0731e-02,\n 5.3986e-02, 8.0414e-03, 1.4294e-01, -3.6377e-02, -5.8167e-02, 1.1572e-01, 3.9276e-02, 1.2238e-01, 1.3086e-01, -5.6519e-02, -4.9042e-02, -5.7640e-03, 6.3293e-02, -9.2392e-03, -1.3879e-01, -1.5417e-01, -1.9067e-01, 3.4485e-02, -7.9712e-02, -3.1616e-02, 3.2166e-02, 4.3976e-02, 1.0017e-02, 1.6223e-01,\n 1.5961e-02, 1.6406e-01, 1.4685e-01, 3.3447e-02, 6.5857e-02, 1.1969e-01, 2.2180e-01, 1.2756e-01, 6.6406e-02, -1.1987e-01, -5.5420e-02, -3.2898e-02, -3.0899e-02, 4.5837e-02, -1.2756e-01, 3.9307e-02, 2.6108e-02, -1.0077e-01, 3.2684e-02, -1.2964e-01, 8.0750e-02, -3.2837e-02, 1.0663e-01, 8.6609e-02,\n -5.8746e-02, 1.6479e-01, 1.6736e-01, -2.6688e-02, 5.4474e-03, 4.9472e-05, -2.0340e-02, -1.7578e-02, -1.6309e-01, 6.2195e-02, -1.1951e-01, -4.4861e-02, -1.9531e-02, -4.3457e-02, 1.1467e-02, -1.0333e-01, -4.1931e-02, 1.0760e-01, -1.0284e-01, 9.3750e-02, -2.6871e-02, 9.3994e-02, -7.4036e-02, -1.4697e-01,\n 3.0075e-02, 2.5928e-01, -4.1840e-02, -2.7512e-02, 1.2122e-01, 1.5173e-01, 1.4270e-01, -1.1487e-01, 3.5767e-02, 1.8066e-02, 1.1652e-01, -1.9028e-02, -1.9745e-02, 8.3923e-03, 1.7798e-01, -1.1084e-01, 1.9180e-02, 1.2671e-01, 2.6993e-02, -3.1586e-02, -8.5205e-02, -1.1993e-01, -6.5369e-02, -2.9190e-02,\n 2.8931e-02, 7.3242e-02, 3.4302e-02, -5.1880e-02, 6.6650e-02, -2.0523e-02, 1.0040e-02, -1.4885e-02, -1.2830e-01, 1.4746e-01, 8.4351e-02, -2.9640e-03, 7.3242e-03, 9.3323e-02, 4.6204e-02, -1.1169e-01, -2.0691e-01, -6.2744e-02, -1.8835e-01, -6.1096e-02, 5.2429e-02, -1.9638e-02, 3.0651e-03, 7.2876e-02,\n 4.6295e-02, -1.4148e-01, 6.8970e-02, 1.0675e-01, -4.9530e-02, -1.9150e-02, 4.8431e-02, -1.0645e-01, -9.7290e-02, -3.7354e-02, -1.2366e-01, 3.9001e-02, -1.1102e-01, 8.0795e-03, 1.8042e-01, 4.4586e-02, -1.0986e-01, 6.4331e-02, -1.1273e-01, 1.2976e-01, 9.2346e-02, -6.6528e-02, 1.7664e-01, 1.1804e-01,\n 8.5693e-02, 1.1029e-01, 3.7476e-02, 1.6882e-01, 1.7212e-02, -5.7465e-02, -8.4106e-02, -5.5237e-02, 1.3039e-02, 1.1887e-02, -4.9866e-02, -1.3660e-01, -1.3733e-01, -1.8997e-02, -2.9373e-02, -1.2427e-01, -1.6467e-01, -1.2201e-01, -1.1151e-01, 2.7863e-02, -3.8269e-02, -1.7004e-01, -3.7476e-02, 7.5928e-02,\n -5.0262e-02, 2.2003e-02, -1.9214e-01, 1.2787e-02, -1.1841e-01, 1.8372e-01, -1.3757e-01, -3.6377e-02, 3.6713e-02, 1.3757e-01, -1.0754e-01, 1.8799e-02, -4.2053e-02, 7.0007e-02, -1.9424e-02, 5.0995e-02, -3.5954e-03, -1.2830e-01, 1.1688e-01, 5.9052e-02, 9.4727e-02, -7.7454e-02, 1.1041e-01, -2.8687e-02,\n -2.6123e-02, -7.6965e-02, -9.1919e-02, 4.7646e-03, 7.9651e-02, 9.3262e-02, 7.3486e-02, -5.7617e-02, 1.3008e-02, 4.3854e-02, -1.2323e-01, 1.5656e-02, -1.7065e-01, 7.2083e-02, -1.5686e-01, 1.3391e-01, 8.3160e-03, 1.1780e-02, -3.8605e-02, -1.0620e-01, 3.3417e-02, -4.7112e-03, -1.6858e-01, 8.9172e-02,\n -1.2140e-01, 1.1920e-01, 1.1267e-01, -6.1920e-02, -3.5828e-02, -1.4984e-02, 1.6711e-01, -1.0090e-03, -4.0222e-02, 3.9368e-02, -1.3452e-01, -5.5481e-02, -1.7542e-01, 9.6802e-02, -6.7139e-02, 1.2079e-01, -3.5950e-02, 2.8152e-02, 1.9031e-01, 3.8971e-02, 8.4473e-02, -5.3650e-02, -4.7516e-02, -1.4809e-02,\n -1.4429e-01, 4.6043e-03, 2.2400e-02, 5.8319e-02, -1.6150e-01, 1.2622e-01, -1.1346e-01, 5.7037e-02, -1.7407e-01, 1.6882e-01, 1.2000e-01, 5.3120e-04, 9.1980e-02, 1.0185e-02, -9.6436e-02, 1.0345e-01, 5.7983e-02, -2.5818e-02, -1.3075e-03, -1.6617e-02, 1.6638e-01, -4.3365e-02, -6.7383e-02, -6.2225e-02,\n 9.5703e-02, 1.1426e-01, 1.8129e-03, -5.4398e-03, 9.4299e-03, -2.6199e-02, -2.5940e-02, 1.0773e-02, -1.0034e-01, 2.4521e-02, -1.5405e-01, 1.0120e-01, 5.1727e-02, 9.4360e-02, -1.4490e-01, 1.2085e-01, 9.4238e-02, 1.8372e-02, 1.1940e-02, 3.9948e-02, 2.9358e-02, -2.9846e-02, -6.3477e-02, -3.4119e-02,\n 4.5349e-02, -7.0190e-02, 1.0956e-01, 8.5815e-02, -5.0354e-02, -4.3701e-02, -2.0355e-02, -7.2266e-02, -4.2084e-02, -1.7554e-01, 4.9225e-02, 1.3379e-01, -9.6741e-02, 7.0801e-02, -5.4979e-04, -9.5886e-02, 1.0071e-01, 5.0018e-02, -8.2458e-02, 1.0687e-01, -7.0007e-02, -8.4839e-02, -9.4910e-02, -9.5032e-02,\n 2.4933e-02, -1.1316e-01, 8.8684e-02, 5.7182e-03, -1.3245e-01, -2.0813e-02, 8.6853e-02, -1.3635e-01, -7.3853e-03, 9.4910e-02, 1.1499e-01, 1.3550e-01, 1.0468e-01, 9.3262e-02, -3.1143e-02, 9.3445e-02, -1.0962e-01, 1.3293e-01, 1.7920e-01, 5.0476e-02, 7.6233e-02, -1.3046e-02, 4.8798e-02, -1.8372e-01,\n -5.5298e-02, -5.7983e-02, 1.2360e-01, 1.2781e-01, -5.4810e-02, 1.5063e-01, 1.2878e-01, -2.0984e-01, 7.5623e-02, -7.2083e-02, -6.4575e-02, 3.2867e-02, -1.7407e-01, -2.5330e-02, -9.9121e-02, -9.0027e-02, 4.2847e-02, 6.8176e-02, 1.3458e-02, 1.0553e-01, -1.0413e-01, -6.8726e-02, 2.4780e-02, -1.1853e-01,\n -4.6539e-03, -6.1096e-02, 4.0955e-02, -8.0322e-02, 1.1285e-01, -5.1392e-02, -3.7262e-02, 1.1017e-01, -8.3496e-02, -5.1147e-02, -4.0283e-03, 9.4482e-02, -8.6243e-02, 1.9287e-01, -2.0264e-01, -1.4294e-01, -1.5137e-01, -7.9712e-02, 3.7628e-02, 1.5356e-01, 9.1309e-02, -4.8676e-02, -4.9316e-02, 1.2878e-01],\n dtype=torch.float16), 'perceiver_resampler.layers.0.1.1.weight': tensor([[ 0.0111, 0.0131, 0.0102, ..., -0.0450, 0.0174, -0.0191],\n [-0.0102, -0.0194, 0.0225, ..., -0.0038, 0.0156, -0.0237],\n [-0.0165, -0.0603, -0.0249, ..., 0.0233, -0.0103, 0.0068],\n ...,\n [-0.0605, -0.0161, 0.0474, ..., -0.0003, 0.0023, 0.0216],\n [-0.0233, -0.0164, 0.0380, ..., 0.0239, -0.0186, -0.0095],\n [ 0.0146, -0.0038, 0.0016, ..., 0.0161, 0.0284, -0.0195]], dtype=torch.float16), 'perceiver_resampler.layers.0.1.3.weight': tensor([[ 0.0175, -0.0067, -0.0249, ..., -0.0043, -0.0408, 0.0269],\n [-0.0065, 0.0123, -0.0020, ..., -0.0112, -0.0229, -0.0201],\n [-0.0096, 0.0199, -0.0064, ..., 0.0189, -0.0211, -0.0100],\n ...,\n [-0.0486, 0.0464, 0.0287, ..., 0.0411, -0.0128, -0.0560],\n [-0.0336, 0.0217, -0.0131, ..., -0.0031, 0.0258, 0.0439],\n [-0.0204, -0.0156, -0.0273, ..., -0.0037, -0.0097, -0.0296]], dtype=torch.float16), 'perceiver_resampler.layers.1.0.norm1.weight': tensor([0.7344, 0.7256, 0.7466, 0.7925, 0.7656, 0.7939, 0.7515, 0.7773, 0.7979, 0.7534, 0.7363, 0.7285, 0.7246, 0.7964, 0.8198, 0.7783, 0.7358, 0.6494, 0.7285, 0.7471, 0.7754, 0.7354, 0.7700, 0.6919, 0.6997, 0.7661, 0.7607, 0.7676, 0.6670, 0.7681, 0.7461, 0.8281, 0.7852, 0.7817, 0.7617, 0.6885, 0.6162, 0.7505, 0.7246,\n 0.7207, 0.7729, 0.8003, 0.7725, 0.7451, 0.8086, 0.7500, 0.7222, 0.7749, 0.6733, 0.7188, 0.6821, 0.7681, 0.7314, 0.7607, 0.7480, 0.7710, 0.7422, 0.6284, 0.7578, 0.7617, 0.7217, 0.7676, 0.6792, 0.7427, 0.7764, 0.6978, 0.8071, 0.7651, 0.7178, 0.8242, 0.7354, 0.7656, 0.8022, 0.8081, 0.7524, 0.7837, 0.7139, 0.7222,\n 0.7002, 0.7617, 0.7485, 0.7510, 0.7510, 0.6836, 0.7749, 0.7310, 0.6895, 0.7476, 0.7930, 0.7583, 0.7461, 0.7705, 0.7285, 0.7461, 0.8008, 0.7437, 0.7686, 0.7080, 0.7808, 0.7158, 0.7046, 0.7207, 0.7314, 0.7158, 0.7310, 0.7402, 0.7793, 0.7739, 0.7925, 0.7524, 0.7778, 0.7788, 0.8135, 0.7812, 0.7524, 0.7998, 0.7603,\n 0.8188, 0.6470, 0.7451, 0.8057, 0.7056, 0.7646, 0.8442, 0.6987, 0.7251, 0.7397, 0.7988, 0.6172, 0.7295, 0.6978, 0.7471, 0.7559, 0.7705, 0.6841, 0.8105, 0.7266, 0.7563, 0.7959, 0.7432, 0.7310, 0.7300, 0.7036, 0.8066, 0.7344, 0.8130, 0.7754, 0.6948, 0.7930, 0.7324, 0.7227, 0.6831, 0.7373, 0.7446, 0.7437, 0.7217,\n 0.7861, 0.7500, 0.7520, 0.7559, 0.7295, 0.7559, 0.7207, 0.7241, 0.6831, 0.7939, 0.7666, 0.7227, 0.7998, 0.7271, 0.7480, 0.7212, 0.7881, 0.7451, 0.7910, 0.7529, 0.7520, 0.6870, 0.7188, 0.7041, 0.7324, 0.7524, 0.7905, 0.8052, 0.7607, 0.7705, 0.5698, 0.7568, 0.6802, 0.7266, 0.8120, 0.6152, 0.7661, 0.8076, 0.7490,\n 0.7695, 0.7378, 0.6938, 0.7583, 0.7710, 0.6963, 0.7500, 0.7134, 0.6929, 0.7314, 0.7314, 0.7070, 0.7944, 0.7070, 0.7856, 0.6973, 0.7432, 0.6660, 0.7832, 0.7095, 0.7559, 0.7905, 0.7402, 0.7690, 0.7432, 0.7183, 0.7642, 0.7305, 0.7617, 0.7505, 0.6851, 0.7900, 0.7935, 0.7178, 0.7920, 0.7739, 0.7305, 0.7559, 0.7036,\n 0.7329, 0.7168, 0.7827, 0.7681, 0.6743, 0.7700, 0.7061, 0.7876, 0.7324, 0.6943, 0.7212, 0.7388, 0.7739, 0.7827, 0.7744, 0.7607, 0.7441, 0.7329, 0.6792, 0.7002, 0.7246, 0.7197, 0.7397, 0.6528, 0.6953, 0.7217, 0.6992, 0.7344, 0.7705, 0.7871, 0.7466, 0.7251, 0.7222, 0.7637, 0.7700, 0.7778, 0.7612, 0.7578, 0.8438,\n 0.6978, 0.8047, 0.7383, 0.7822, 0.7471, 0.7280, 0.6953, 0.7290, 0.7646, 0.7368, 0.6401, 0.7485, 0.6733, 0.7358, 0.8179, 0.7544, 0.7739, 0.7275, 0.7075, 0.7275, 0.6289, 0.7866, 0.7510, 0.7290, 0.6992, 0.7593, 0.6455, 0.6650, 0.6929, 0.6523, 0.7515, 0.7202, 0.6152, 0.6479, 0.7095, 0.7695, 0.6865, 0.6982, 0.6982,\n 0.8071, 0.7529, 0.7339, 0.7861, 0.7075, 0.7007, 0.7607, 0.7866, 0.7036, 0.6919, 0.6836, 0.7134, 0.7256, 0.7715, 0.7925, 0.7886, 0.7310, 0.7612, 0.7573, 0.7163, 0.7593, 0.7363, 0.7339, 0.7837, 0.7842, 0.7139, 0.7886, 0.7217, 0.7734, 0.7202, 0.7461, 0.7236, 0.7095, 0.6851, 0.7710, 0.7793, 0.7466, 0.7222, 0.7021,\n 0.7534, 0.6426, 0.7329, 0.7109, 0.7778, 0.7656, 0.7739, 0.7495, 0.7905, 0.7681, 0.7124, 0.7788, 0.7876, 0.7529, 0.7173, 0.7129, 0.7651, 0.7461, 0.7612, 0.7495, 0.7642, 0.7427, 0.7437, 0.7275, 0.7666, 0.7764, 0.7710, 0.7114, 0.7637, 0.7017, 0.7256, 0.7305, 0.7231, 0.7583, 0.7578, 0.7290, 0.6890, 0.7295, 0.6885,\n 0.7715, 0.7964, 0.7915, 0.7271, 0.7461, 0.7485, 0.7329, 0.7275, 0.7500, 0.7441, 0.7593, 0.7095, 0.7808, 0.7383, 0.7168, 0.7900, 0.6978, 0.7510, 0.7275, 0.7505, 0.7393, 0.7578, 0.7378, 0.7720, 0.6865, 0.7935, 0.7427, 0.7295, 0.7852, 0.7422, 0.7324, 0.7490, 0.8027, 0.7251, 0.7690, 0.7891, 0.8018, 0.7808, 0.7725,\n 0.6782, 0.7544, 0.6885, 0.7109, 0.7568, 0.7822, 0.7837, 0.7212, 0.7446, 0.7573, 0.8032, 0.7583, 0.7900, 0.7188, 0.8042, 0.6890, 0.7734, 0.7578, 0.7661, 0.6978, 0.7725, 0.6885, 0.6865, 0.7573, 0.7007, 0.7549, 0.7202, 0.6729, 0.7451, 0.7041, 0.8096, 0.7368, 0.7100, 0.7090, 0.7217, 0.7412, 0.7803, 0.7827, 0.7388,\n 0.7793, 0.7739, 0.6948, 0.7344, 0.8071, 0.7002, 0.7451, 0.7632, 0.7905, 0.7627, 0.7480, 0.6899, 0.7427, 0.7476, 0.7876, 0.7378, 0.7661, 0.6885, 0.7344, 0.6821, 0.7905, 0.6528, 0.7842, 0.6587, 0.6978, 0.7104, 0.7461, 0.7534, 0.7471, 0.7866, 0.7339, 0.6846, 0.7153, 0.7754, 0.7812, 0.7285, 0.8052, 0.7266, 0.7344,\n 0.7446, 0.7637, 0.7588, 0.7661, 0.7573, 0.7715, 0.7603, 0.7681, 0.7705, 0.7744, 0.7090, 0.7422, 0.7314, 0.7681, 0.7446, 0.8027, 0.8086, 0.7646, 0.7803, 0.6294, 0.7197, 0.6865, 0.6333, 0.7681, 0.6816, 0.7456, 0.7246, 0.7905, 0.7783, 0.6768, 0.7314, 0.8125, 0.7173, 0.7505, 0.7329, 0.8179, 0.7583, 0.7700, 0.7744,\n 0.7544, 0.7344, 0.7495, 0.7632, 0.7896, 0.7905, 0.7285, 0.8047, 0.7510, 0.7388, 0.7856, 0.7324, 0.7334, 0.6953, 0.7734, 0.7378, 0.7627, 0.7480, 0.7744, 0.7725, 0.7661, 0.7129, 0.7612, 0.7710, 0.7217, 0.7959, 0.7554, 0.6470, 0.7227, 0.7329, 0.7822, 0.7905, 0.7119, 0.6992, 0.6797, 0.8062, 0.7119, 0.7881, 0.7329,\n 0.7349, 0.7261, 0.7568, 0.7954, 0.7617, 0.6846, 0.6890, 0.7153, 0.7363, 0.7124, 0.7949, 0.7607, 0.7334, 0.7407, 0.6787, 0.7417, 0.7290, 0.6519, 0.7002, 0.7690, 0.7563, 0.7397, 0.7358, 0.7681, 0.6841, 0.7280, 0.7974, 0.6748, 0.7188, 0.6987, 0.7969, 0.7256, 0.7661, 0.7646, 0.7383, 0.7378, 0.7563, 0.7261, 0.7852,\n 0.7476, 0.6240, 0.7622, 0.7954, 0.7021, 0.7612, 0.7417, 0.7583, 0.7773, 0.7617, 0.7573, 0.6860, 0.7686, 0.8076, 0.7700, 0.7832, 0.7007, 0.7070, 0.7939, 0.7817, 0.7427, 0.7427, 0.8047, 0.7036, 0.7080, 0.7778, 0.7188, 0.7114, 0.7090, 0.7437, 0.7793, 0.7441, 0.7285, 0.7681, 0.7354, 0.8267, 0.6758, 0.7227, 0.7676,\n 0.6729, 0.7549, 0.8145, 0.7842, 0.7700, 0.6816, 0.7178, 0.6523, 0.7622, 0.7744, 0.7173, 0.7495, 0.7847, 0.8179, 0.7603, 0.7373, 0.7446, 0.8076, 0.7456, 0.7378, 0.7695, 0.6846, 0.7119, 0.7451, 0.7456, 0.7578, 0.7090, 0.7559, 0.6958, 0.7397, 0.7466, 0.7207, 0.7661, 0.7090, 0.8115, 0.7617, 0.7222, 0.7441, 0.7632,\n 0.7817, 0.7407, 0.7910, 0.7456, 0.7671, 0.7637, 0.7651, 0.6958, 0.6758, 0.6997, 0.7778, 0.7305, 0.8042, 0.7178, 0.7520, 0.7759, 0.7231, 0.5933, 0.7896, 0.7441, 0.7427, 0.7466, 0.7295, 0.7207, 0.7529, 0.6523, 0.7256, 0.7832, 0.7163, 0.7349, 0.7373, 0.7246, 0.6948, 0.7334, 0.7134, 0.7686, 0.7666, 0.7358, 0.6606,\n 0.6948, 0.7935, 0.6797, 0.7339, 0.7236, 0.7085, 0.7725, 0.7417, 0.7515, 0.7544, 0.7744, 0.7593, 0.7437, 0.7622, 0.6748, 0.7598, 0.6152, 0.7222, 0.7207, 0.7505, 0.6826, 0.7007, 0.7773, 0.7588, 0.7231, 0.7256, 0.6992], dtype=torch.float16), 'perceiver_resampler.layers.1.0.norm1.bias': tensor([ 1.5091e-02, 3.3783e-02, -1.0979e-02, -8.3351e-04, 2.3682e-02, -4.1412e-02, -3.3142e-02, 2.9182e-03, 2.0340e-02, -9.3155e-03, 1.6022e-02, 1.0672e-03, -3.1433e-02, 2.3890e-04, -4.5967e-03, -1.4055e-04, -2.0657e-03, -6.9214e-02, -1.3344e-02, -2.0905e-02, 9.4681e-03, 1.1158e-03, -3.2959e-03, -2.6321e-02,\n -2.4368e-02, 3.3417e-02, -2.5620e-02, 2.8839e-02, 2.6352e-02, 2.9984e-02, 1.0513e-02, -3.7518e-03, 1.7834e-03, 6.5002e-03, -1.5099e-02, 2.2095e-02, 6.1829e-02, -1.3933e-03, 3.0914e-02, -6.4941e-02, -2.5787e-02, -3.7899e-03, 2.8973e-03, 3.8422e-02, 2.0370e-03, 7.1335e-03, 5.1605e-02, 2.9984e-02,\n 3.3813e-02, 5.7587e-02, -1.1276e-02, 3.0960e-02, 1.7227e-02, -2.6306e-02, 3.6346e-02, -3.1143e-02, 2.0523e-02, 4.2610e-03, 3.0441e-02, 7.2479e-03, 2.2781e-02, -8.2064e-04, -1.4206e-02, 3.2990e-02, 7.1754e-03, -7.4654e-03, -2.4536e-02, -1.4954e-02, 4.5776e-02, -5.4932e-02, -1.3252e-02, -4.0436e-02,\n -2.4704e-02, 2.0447e-02, -1.2810e-02, 1.6296e-02, 2.8595e-02, 4.1687e-02, 3.0273e-02, -1.5144e-02, -3.1281e-02, -3.4119e-02, -5.6274e-02, 7.8659e-03, -3.4058e-02, -1.7443e-03, -4.4373e-02, -4.0283e-03, 1.1696e-02, 1.5594e-02, 2.8183e-02, -7.5073e-03, 1.5823e-02, 7.8735e-03, 1.9806e-02, -1.4687e-02,\n 3.8948e-03, -3.5583e-02, -1.2178e-03, -4.4983e-02, -2.2507e-02, 2.0798e-02, 2.1225e-02, -1.7853e-02, 2.5284e-02, 2.6798e-03, 2.6947e-02, -2.7298e-02, -3.9154e-02, -3.4088e-02, 3.4973e-02, -1.8188e-02, 1.4603e-02, 2.3804e-03, 4.3091e-02, -2.3239e-02, 2.8381e-02, -3.3752e-02, -3.7651e-03, 1.3580e-02,\n -3.9185e-02, -3.3905e-02, -6.1989e-03, -1.2955e-02, -4.2694e-02, -4.3762e-02, -1.1024e-02, 1.3374e-02, 1.0321e-01, -2.1713e-02, -8.5449e-04, -5.6793e-02, 4.2358e-02, -1.7944e-02, 7.2813e-04, 6.7101e-03, -2.4857e-02, -3.0136e-03, -1.2810e-02, 2.6360e-03, -3.6346e-02, -1.5993e-03, -4.9805e-02, 2.0523e-03,\n 1.1520e-02, -3.3264e-02, 1.9665e-03, 2.2522e-02, 6.5956e-03, 5.6244e-02, -8.5449e-03, -2.6367e-02, -3.1799e-02, 1.6556e-02, -9.2773e-03, -9.6283e-03, -1.1997e-03, 2.5284e-02, 4.5593e-02, -3.0746e-02, -1.7252e-03, -7.8487e-04, 4.1626e-02, 2.1240e-02, 1.7838e-02, -6.3095e-03, -4.4617e-02, 8.8425e-03,\n -2.6123e-02, 1.1284e-02, -4.0100e-02, 3.1647e-02, -7.4654e-03, 4.6173e-02, 1.6937e-02, 2.2705e-02, 1.0429e-02, 4.9072e-02, 1.9653e-02, -1.8940e-03, -1.5266e-02, 1.1162e-02, 6.2866e-03, 5.1758e-02, -1.0462e-03, -3.3417e-02, -6.0699e-02, 2.3880e-02, 6.1401e-02, -1.7881e-03, 1.9653e-02, -8.2703e-03,\n -5.2223e-03, -3.6850e-03, 1.8997e-02, -1.0033e-02, -1.0681e-02, 2.3315e-02, -9.3536e-03, 4.2572e-02, 3.4973e-02, -1.2329e-02, -2.5330e-02, 1.0208e-02, -5.0087e-03, -3.5828e-02, -6.7749e-02, -1.2405e-02, -1.5884e-02, -2.6840e-02, 7.0801e-02, -3.3661e-02, 2.9251e-02, 1.7181e-02, 4.6753e-02, 4.3373e-03,\n -3.7201e-02, 2.0416e-02, 1.4893e-02, 1.9913e-03, -1.0948e-03, 1.8661e-02, 3.6316e-02, 1.1948e-02, -2.8717e-02, -6.3965e-02, -5.4245e-03, -1.9455e-02, -5.6885e-02, -1.7853e-02, -1.7996e-03, -1.4244e-02, 2.9159e-04, -4.5624e-02, -1.1307e-02, 2.9938e-02, -6.5460e-03, -7.9803e-03, -4.2908e-02, 5.6305e-02,\n -3.5156e-02, -4.1924e-03, -2.7817e-02, -1.1734e-02, 7.6942e-03, 1.8806e-03, -3.0869e-02, 7.9407e-02, -1.7349e-02, -2.4094e-02, 3.4607e-02, 1.7349e-02, -3.7079e-02, -4.6173e-02, -5.4321e-03, 9.9487e-03, 3.6072e-02, -2.1561e-02, -5.8533e-02, -3.9490e-02, -3.7842e-02, 4.7180e-02, 1.8066e-02, -4.8431e-02,\n 4.2236e-02, 5.9998e-02, 4.1290e-02, 4.1962e-02, 1.7639e-02, 1.9119e-02, -9.9792e-03, -4.3121e-02, 1.2741e-02, 7.1754e-03, 2.3056e-02, 2.9846e-02, -7.5188e-03, -2.0695e-03, 9.4452e-03, -4.0161e-02, 3.9886e-02, -1.4465e-02, -1.3199e-02, 2.5070e-02, 9.3994e-03, -2.2949e-02, -1.2779e-02, 6.6566e-03,\n 1.3420e-02, -4.1733e-03, -7.1899e-02, 1.4381e-02, 1.9913e-02, 7.3059e-02, -1.2827e-03, 2.3861e-03, 2.0370e-02, 3.6240e-03, -2.1362e-02, -1.2016e-02, -3.9856e-02, -2.8793e-02, -1.8906e-02, -1.2207e-02, -2.2720e-02, 3.0029e-02, 6.8237e-02, -3.7628e-02, 3.4149e-02, -1.0384e-02, 3.6743e-02, -8.2932e-03,\n 2.4490e-02, 4.2381e-03, 1.7792e-02, -3.1494e-02, -8.9417e-03, 3.1403e-02, 1.4740e-02, -1.1110e-03, -4.4212e-03, 5.2404e-04, 3.4973e-02, -1.5335e-02, -2.1378e-02, 2.3880e-02, 5.2063e-02, -3.5248e-03, 3.8788e-02, -2.2705e-02, -8.6975e-03, -5.1117e-02, -2.7130e-02, -3.0167e-02, 3.8757e-03, 5.0831e-04,\n -4.8157e-02, -4.4647e-02, -4.8184e-04, 1.2560e-03, -6.2132e-04, 3.1071e-03, 9.9716e-03, -3.1815e-03, -1.6413e-03, 3.6499e-02, 2.3560e-02, -4.0375e-02, -3.8242e-03, 5.2673e-02, -4.4617e-02, 9.7580e-03, 5.3284e-02, 5.3497e-02, -2.9205e-02, 1.3657e-02, -3.6285e-02, 1.5221e-03, 7.6538e-02, -2.6779e-03,\n -2.8732e-02, 1.4229e-02, 2.3880e-02, 2.9160e-02, -2.7817e-02, -3.3997e-02, 3.5217e-02, 2.7710e-02, 5.1880e-03, -1.5625e-02, 3.8818e-02, -4.2191e-03, -4.3457e-02, 1.6510e-02, 4.8828e-02, 6.6910e-03, -6.0364e-02, -4.1556e-04, -1.4664e-02, -6.0730e-02, 1.3344e-02, -4.5242e-03, -1.3336e-02, -3.7994e-02,\n -3.6716e-03, -6.3553e-03, 1.6251e-02, 1.6541e-02, 4.9072e-02, -3.6987e-02, 1.5221e-02, -5.4359e-03, 5.8098e-03, 3.1616e-02, 1.3123e-02, -1.2184e-02, -1.1803e-02, -2.2903e-02, 4.0314e-02, 3.8300e-03, 3.9825e-02, -3.1952e-02, -2.2354e-03, -4.3549e-02, 2.7527e-02, 2.1915e-03, 2.4780e-02, 9.0179e-03,\n 6.5041e-03, -1.8875e-02, 1.1501e-03, 1.2573e-02, 3.8055e-02, -1.5762e-02, 2.1454e-02, 4.8889e-02, -1.3725e-02, -2.3193e-02, -7.8011e-04, -2.2507e-02, 9.0714e-03, -4.0588e-02, -2.1271e-02, 3.5439e-03, -3.9856e-02, 8.7509e-03, -1.5091e-02, 6.0730e-02, -4.3427e-02, 3.3112e-02, 1.6052e-02, 4.5349e-02,\n 3.5810e-04, 2.2392e-03, 1.5383e-03, -2.7145e-02, -1.8723e-02, 1.6830e-02, -1.9272e-02, 2.4033e-03, 8.5907e-03, 6.8703e-03, 1.3451e-02, -2.1271e-02, -1.3527e-02, 2.3117e-02, 3.1555e-02, 3.6041e-02, 1.1398e-02, 1.2642e-02, -1.7303e-02, -3.0411e-02, -3.9520e-02, 2.4979e-02, -2.0046e-03, 6.5491e-02,\n -1.7395e-02, 6.6071e-03, 3.6255e-02, -2.1515e-03, -3.7292e-02, 3.7933e-02, 5.4665e-03, -6.4735e-03, 1.0612e-02, 1.5778e-02, -1.3428e-02, 2.4277e-02, -8.3923e-03, -6.8550e-03, -1.0590e-01, 1.8127e-02, 2.6230e-02, 5.2460e-02, 3.9948e-02, 2.1553e-03, -9.2697e-03, -1.7395e-02, 1.6754e-02, -1.5572e-02,\n 2.0370e-03, 2.4933e-02, -1.5900e-02, 4.1321e-02, 9.3536e-03, 1.2535e-02, -2.8961e-02, -3.7842e-02, -2.9053e-02, -6.3721e-02, 1.5450e-03, 2.5894e-02, 3.1052e-02, -5.2460e-02, -5.1003e-03, -1.4259e-02, -3.7140e-02, 1.9791e-02, -6.6185e-03, -1.0757e-02, 1.3947e-02, -4.1779e-02, 1.1940e-02, -2.7710e-02,\n 7.0000e-03, 3.5522e-02, 7.9041e-03, 1.9569e-03, -3.7109e-02, -1.6418e-02, 7.0152e-03, -2.1896e-02, 3.6774e-02, 6.4087e-02, -3.6835e-02, -2.3987e-02, -9.3689e-03, 2.0309e-02, -1.4359e-02, 6.3610e-04, -8.0032e-03, 4.9622e-02, -5.8594e-02, -3.1338e-03, -4.6349e-03, -1.7120e-02, -2.4548e-03, 3.0422e-03,\n 1.1269e-02, 2.3895e-02, -1.5450e-03, -2.3239e-02, -1.6556e-02, 3.0472e-02, -2.8519e-02, 1.9974e-02, 4.8561e-03, -2.5055e-02, 1.4961e-02, -2.5558e-02, -7.4463e-03, -2.1420e-03, -1.3725e-02, -2.1095e-03, -1.9958e-02, -7.0496e-03, -2.2545e-03, 8.7547e-04, 2.2781e-02, 3.7201e-02, 1.6464e-02, -1.3466e-02,\n -3.0548e-02, 3.8087e-05, -5.0903e-02, 3.0212e-02, -9.6207e-03, 1.2573e-02, 4.0009e-02, 5.8502e-02, -7.0679e-02, -4.7821e-02, 6.5193e-03, 6.6589e-02, -3.2166e-02, 1.9531e-02, -5.8784e-03, -1.5190e-02, -7.2861e-03, -2.5314e-02, -7.6294e-03, -4.9377e-02, 1.4755e-02, 1.8143e-02, -8.7433e-03, -3.1952e-02,\n -2.4185e-02, 5.3284e-02, 8.5068e-03, -1.9913e-02, 8.3618e-03, -1.2428e-02, -5.9906e-02, -9.0027e-04, 1.8902e-03, 3.7598e-02, -3.4332e-02, -2.9037e-02, -1.8936e-02, -1.6571e-02, 3.3997e-02, -5.3253e-02, 7.6599e-03, 1.8265e-02, 3.4607e-02, -1.7670e-02, -2.8290e-02, 5.2109e-03, 1.6327e-02, -9.3155e-03,\n -1.0620e-02, -8.8120e-03, -2.7130e-02, -3.2288e-02, -7.9727e-03, -1.0971e-02, 3.1250e-02, 1.2856e-02, 2.7695e-03, -1.3153e-02, 2.6798e-03, 2.5787e-02, -3.2104e-02, 9.8190e-03, -2.4292e-02, -3.7262e-02, 1.9287e-02, -7.9956e-03, -3.6194e-02, -4.6616e-03, -1.0651e-02, 3.1677e-02, 2.6138e-02, 6.1264e-03,\n 1.9104e-02, 8.6731e-02, -3.3630e-02, 9.8190e-03, 2.1896e-02, -1.5656e-02, -2.7695e-02, -5.0926e-03, -1.7746e-02, 1.1375e-02, 4.7638e-02, -5.9570e-02, -4.0924e-02, 1.2840e-02, 1.7426e-02, -1.2344e-02, 1.8906e-02, -1.4664e-02, -3.1338e-03, 3.6793e-03, 1.9073e-02, -2.9022e-02, 2.0950e-02, -6.9946e-02,\n 2.0096e-02, 4.9683e-02, -2.0294e-02, 4.4220e-02, -2.3651e-02, -1.5808e-02, -4.7569e-03, -3.5736e-02, -2.3483e-02, 2.2980e-02, 1.3489e-02, -1.5945e-02, 2.8687e-02, 3.8300e-02, 1.5839e-02, 1.9989e-02, -3.7415e-02, -7.4234e-03, 1.7639e-02, 1.5182e-02, -1.9012e-02, -2.2934e-02, -5.2460e-02, 2.4323e-02,\n -3.6755e-03, 1.6769e-02, -2.6581e-02, 1.4008e-02, 6.4354e-03, 1.6998e-02, -6.9214e-02, -2.3708e-03, -1.4534e-02, -7.3090e-03, -5.2155e-02, -2.7390e-02, 1.5915e-02, 2.7191e-02, -9.3994e-03, 1.8677e-02, 7.4219e-02, 1.6403e-02, 1.0460e-02, 2.5726e-02, -1.4572e-02, 4.5166e-03, 1.9577e-02, -2.3468e-02,\n 1.9196e-02, 1.0521e-02, 2.1255e-02, -6.6772e-02, 3.5839e-03, -1.6174e-02, 3.7323e-02, 4.6173e-02, -1.5343e-02, 2.3880e-02, -2.3346e-03, 8.7357e-03, 8.9340e-03, -2.5757e-02, 1.4862e-02, 2.3365e-03, 6.9336e-02, 1.0551e-02, -8.9035e-03, -4.5090e-03, -2.8549e-02, 2.5406e-02, 1.2520e-02, 1.1877e-01,\n -4.9072e-02, 5.0392e-03, -3.4454e-02, 1.2589e-02, -4.2877e-02, 4.1687e-02, 4.8340e-02, -7.4097e-02, -1.8036e-02, -1.3618e-02, -3.6804e-02, 4.2633e-02, 3.1143e-02, 1.0277e-02, -2.6733e-02, 3.8849e-02, -6.5689e-03, 4.9324e-03, -4.9362e-03, -6.4270e-02, 6.0997e-03, 1.3649e-02, 3.7262e-02, -4.6600e-02,\n -2.8549e-02, -2.9297e-02, 5.0262e-02, -1.2093e-03, 2.6749e-02, -2.6413e-02, -1.8127e-02, -3.5309e-02, 1.2421e-02, 1.3924e-02, 9.5062e-03, -1.6403e-02, 4.3549e-02, 9.2850e-03, 4.0092e-03, -2.4223e-03, -4.6783e-02, 8.2764e-02, 3.2196e-02, -3.0014e-02, 8.8196e-03, 2.0111e-02, 2.1576e-02, -2.0493e-02],\n dtype=torch.float16), 'perceiver_resampler.layers.1.0.norm2.weight': tensor([0.5908, 0.6177, 0.6460, 0.5747, 0.5874, 0.6499, 0.5996, 0.5723, 0.6030, 0.5938, 0.6069, 0.5176, 0.6099, 0.5522, 0.6294, 0.5674, 0.5986, 0.6182, 0.5771, 0.5454, 0.5718, 0.5737, 0.5825, 0.5361, 0.5938, 0.6152, 0.6367, 0.6499, 0.6025, 0.6099, 0.6377, 0.6045, 0.6626, 0.5708, 0.5786, 0.6465, 0.5430, 0.5977, 0.5684,\n 0.5742, 0.6084, 0.5884, 0.5908, 0.6074, 0.5806, 0.5063, 0.6167, 0.5977, 0.5693, 0.6353, 0.6172, 0.5693, 0.6455, 0.5962, 0.5103, 0.6401, 0.5674, 0.5254, 0.6221, 0.5493, 0.5703, 0.6265, 0.6187, 0.6284, 0.6108, 0.6001, 0.5918, 0.5269, 0.6885, 0.5693, 0.6055, 0.6196, 0.6396, 0.6099, 0.6294, 0.6587, 0.6108, 0.5322,\n 0.5615, 0.6216, 0.6387, 0.5117, 0.5840, 0.6343, 0.6016, 0.6035, 0.6079, 0.5273, 0.6016, 0.4609, 0.5928, 0.6045, 0.6162, 0.6602, 0.6191, 0.6318, 0.5513, 0.5801, 0.5786, 0.5679, 0.6226, 0.5718, 0.6094, 0.5493, 0.5737, 0.6104, 0.5781, 0.5986, 0.5977, 0.5776, 0.5835, 0.5933, 0.5757, 0.5918, 0.6138, 0.6636, 0.5845,\n 0.5889, 0.6294, 0.5664, 0.6118, 0.6274, 0.6265, 0.6084, 0.6270, 0.5840, 0.6240, 0.5752, 0.5947, 0.6353, 0.6157, 0.5190, 0.6431, 0.5977, 0.6235, 0.5854, 0.5869, 0.5923, 0.5981, 0.6064, 0.6387, 0.6284, 0.6187, 0.5117, 0.5967, 0.5913, 0.5957, 0.6255, 0.5815, 0.6318, 0.6631, 0.5532, 0.5586, 0.5879, 0.6548, 0.5781,\n 0.5474, 0.6333, 0.5601, 0.6016, 0.6406, 0.5557, 0.6021, 0.5181, 0.6489, 0.6411, 0.6138, 0.5864, 0.6367, 0.6060, 0.5654, 0.6372, 0.5952, 0.6636, 0.5474, 0.5376, 0.5806, 0.5586, 0.6113, 0.6084, 0.6016, 0.6313, 0.5669, 0.5151, 0.5728, 0.6162, 0.5815, 0.5454, 0.6152, 0.6206, 0.6191, 0.5869, 0.5591, 0.6499, 0.6201,\n 0.6177, 0.5620, 0.6016, 0.6416, 0.5981, 0.5728, 0.6099, 0.6177, 0.6343, 0.5322, 0.6260, 0.6035, 0.6079, 0.6943, 0.5342, 0.5737, 0.6631, 0.5757, 0.6099, 0.4790, 0.5835, 0.5898, 0.5825, 0.6436, 0.5518, 0.5479, 0.6191, 0.5938, 0.6045, 0.5981, 0.6494, 0.5601, 0.5894, 0.5483, 0.6421, 0.6274, 0.5356, 0.6035, 0.5942,\n 0.6460, 0.6006, 0.6396, 0.6108, 0.5693, 0.6064, 0.5479, 0.5264, 0.6196, 0.5903, 0.6226, 0.6011, 0.5752, 0.5762, 0.6436, 0.5327, 0.5605, 0.5693, 0.5669, 0.6343, 0.5591, 0.6719, 0.6572, 0.6064, 0.6025, 0.6514, 0.6050, 0.5195, 0.5728, 0.5786, 0.6274, 0.5889, 0.6382, 0.5918, 0.5981, 0.6060, 0.6387, 0.5640, 0.5474,\n 0.6274, 0.5356, 0.6440, 0.6074, 0.6265, 0.5444, 0.6221, 0.5654, 0.4685, 0.5376, 0.6040, 0.5981, 0.6074, 0.5322, 0.6572, 0.5630, 0.6343, 0.6084, 0.6045, 0.6504, 0.6455, 0.6343, 0.5742, 0.5649, 0.6167, 0.6382, 0.6001, 0.6655, 0.5479, 0.6719, 0.4529, 0.6021, 0.5371, 0.6304, 0.5527, 0.6235, 0.4666, 0.6050, 0.5713,\n 0.6025, 0.6426, 0.6392, 0.5557, 0.6113, 0.6323, 0.5830, 0.6133, 0.6162, 0.5850, 0.5366, 0.5854, 0.6826, 0.5806, 0.6123, 0.5996, 0.6274, 0.5796, 0.5879, 0.6094, 0.6450, 0.6504, 0.6138, 0.6128, 0.6040, 0.6226, 0.6030, 0.5962, 0.6470, 0.5850, 0.6440, 0.6104, 0.6235, 0.5874, 0.6265, 0.5742, 0.6567, 0.5278, 0.6440,\n 0.5649, 0.5908, 0.6519, 0.5332, 0.5879, 0.6040, 0.5391, 0.6143, 0.6143, 0.6416, 0.6060, 0.6504, 0.5894, 0.6260, 0.5781, 0.5908, 0.6265, 0.6597, 0.6162, 0.5532, 0.5728, 0.6401, 0.6108, 0.6040, 0.5923, 0.6104, 0.6416, 0.5723, 0.5728, 0.5840, 0.6504, 0.5752, 0.5928, 0.6304, 0.5889, 0.6182, 0.5796, 0.6108, 0.6064,\n 0.6309, 0.5190, 0.5718, 0.6094, 0.5605, 0.6304, 0.6553, 0.6011, 0.5903, 0.6479, 0.6152, 0.6045, 0.5757, 0.6382, 0.6230, 0.6455, 0.6353, 0.6450, 0.6401, 0.5757, 0.6006, 0.5864, 0.6465, 0.4712, 0.5405, 0.5840, 0.5552, 0.5522, 0.6460, 0.6680, 0.6250, 0.6333, 0.6069, 0.5513, 0.6650, 0.5420, 0.5771, 0.5864, 0.5508,\n 0.6245, 0.6182, 0.6016, 0.5449, 0.6416, 0.6104, 0.5908, 0.5806, 0.5752, 0.5840, 0.6069, 0.5645, 0.6318, 0.5864, 0.5952, 0.6406, 0.6387, 0.5981, 0.6143, 0.6309, 0.5483, 0.5547, 0.6060, 0.5752, 0.5713, 0.6006, 0.5283, 0.6987, 0.5884, 0.6304, 0.5088, 0.5972, 0.6255, 0.6245, 0.6284, 0.5830, 0.5854, 0.5537, 0.6338,\n 0.6064, 0.6328, 0.6201, 0.5815, 0.5542, 0.6089, 0.5513, 0.5991, 0.5112, 0.6382, 0.6045, 0.6284, 0.6021, 0.5957, 0.6382, 0.6460, 0.6011, 0.5498, 0.5957, 0.6162, 0.5601, 0.5835, 0.6060, 0.5850, 0.5645, 0.5762, 0.6138, 0.5957, 0.5693, 0.6401, 0.5557, 0.6099, 0.5527, 0.5781, 0.6021, 0.6187, 0.5815, 0.5640, 0.5854,\n 0.6406, 0.6050, 0.6392, 0.6069, 0.5776, 0.5820, 0.6152, 0.5527, 0.6401, 0.5859, 0.6309, 0.6396, 0.6436, 0.5986, 0.6577, 0.5381, 0.5469, 0.5010, 0.6016, 0.5190, 0.5337, 0.5435, 0.6465, 0.6172, 0.4670, 0.6694, 0.5972, 0.6016, 0.6196, 0.6294, 0.5024, 0.6035, 0.5532, 0.6240, 0.6099, 0.6191, 0.5850, 0.5249, 0.6108,\n 0.5928, 0.6831, 0.6187, 0.5737, 0.5889, 0.5581, 0.6338, 0.6182, 0.5796, 0.6128, 0.5518, 0.6084, 0.6064, 0.6533, 0.6572, 0.5488, 0.5801, 0.5889, 0.6196, 0.6245, 0.5610, 0.6226, 0.6167, 0.5698, 0.5918, 0.6299, 0.6191, 0.6128, 0.5742, 0.5195, 0.6030, 0.6201, 0.6323, 0.5889, 0.5791, 0.6216, 0.6372, 0.6484, 0.6201,\n 0.5508, 0.5850, 0.5928, 0.5581, 0.6108, 0.5679, 0.6157, 0.6138, 0.6475, 0.6494, 0.6235, 0.5928, 0.6323, 0.5430, 0.6050, 0.6284, 0.6167, 0.6011, 0.5786, 0.6484, 0.5728, 0.5996, 0.5586, 0.5576, 0.5942, 0.5908, 0.6157, 0.5757, 0.5801, 0.6113, 0.6318, 0.6060, 0.6323, 0.6733, 0.6099, 0.5986, 0.6216, 0.6118, 0.6104,\n 0.5830, 0.6318, 0.6597, 0.5830, 0.4954, 0.5532, 0.6001, 0.6592, 0.6025, 0.5469, 0.6250, 0.5889, 0.5933, 0.5864, 0.6011, 0.5986, 0.6187, 0.6216, 0.6743, 0.6401, 0.6001, 0.6294, 0.6655, 0.5996, 0.6567, 0.6177, 0.5610, 0.6133, 0.5962, 0.6162, 0.5938, 0.6226, 0.6162, 0.5977, 0.6343, 0.6074, 0.6621, 0.6060, 0.6328,\n 0.5972, 0.6475, 0.6309, 0.5864, 0.6035, 0.5884, 0.5728, 0.6235, 0.5923, 0.5454, 0.6401, 0.5874, 0.5776, 0.5439, 0.5742, 0.5737, 0.6074, 0.6206, 0.6523, 0.6489, 0.6050, 0.6318, 0.5815, 0.5547, 0.5913, 0.4934, 0.6411, 0.6450, 0.6470, 0.6206, 0.5435, 0.6187, 0.6050, 0.6177, 0.6284, 0.5430, 0.5947, 0.5425, 0.5845,\n 0.6191, 0.6279, 0.6357, 0.6460, 0.6040, 0.6162, 0.6240, 0.6284, 0.6152, 0.5825, 0.6162, 0.6226, 0.5840, 0.6206, 0.5767, 0.6094, 0.5527, 0.5122, 0.5806, 0.6011, 0.6230, 0.5918, 0.6069, 0.6250, 0.5757, 0.5249, 0.5977, 0.6431, 0.6206, 0.5645, 0.5205, 0.6216, 0.5859, 0.6372, 0.6055, 0.6392, 0.6309, 0.5942, 0.5947,\n 0.5210, 0.6079, 0.6504, 0.6001, 0.5449, 0.6143, 0.6431, 0.6094, 0.6421, 0.6060, 0.6270, 0.5962, 0.5859, 0.6270, 0.5518, 0.5322, 0.5562, 0.6211, 0.5830, 0.5859, 0.6201, 0.5918, 0.6152, 0.5415, 0.5317, 0.5562, 0.5508], dtype=torch.float16), 'perceiver_resampler.layers.1.0.norm2.bias': tensor([-0.0422, -0.0167, -0.0359, -0.0708, -0.0438, 0.0478, -0.0759, 0.0074, 0.0266, -0.0459, 0.0218, -0.1760, 0.0204, -0.1188, 0.0970, 0.0414, 0.0106, 0.1142, -0.0253, -0.0555, -0.1027, -0.0375, -0.0173, 0.1320, -0.0970, 0.0576, 0.0412, -0.0168, 0.0079, 0.0325, 0.0428, 0.0651, -0.0253, 0.0529,\n 0.1335, 0.0374, 0.0939, 0.0639, -0.0533, -0.0845, -0.0287, 0.0424, -0.0056, 0.0232, 0.0552, -0.0948, 0.0210, 0.0862, -0.0332, 0.0126, -0.0236, 0.1144, 0.0448, 0.0812, 0.0663, 0.0518, -0.0273, 0.0736, -0.0336, 0.0998, -0.0862, 0.0037, 0.0477, -0.0246, -0.0574, 0.0231, 0.0713, 0.1678,\n 0.0264, 0.0524, -0.0893, -0.0196, -0.0039, -0.0904, -0.0125, 0.0058, -0.0667, -0.1813, 0.1344, 0.0079, 0.0754, -0.0016, 0.0873, 0.0232, 0.0374, 0.0778, -0.0254, -0.1335, -0.1011, -0.1963, 0.1009, -0.0516, 0.0574, -0.0111, -0.0602, -0.0972, -0.0204, -0.0787, -0.0861, 0.0684, 0.0792, 0.0696,\n 0.0440, -0.1500, 0.0804, 0.0319, 0.0154, -0.0573, 0.0412, 0.0828, -0.0777, 0.0306, -0.0435, -0.0982, 0.0458, -0.0212, 0.1274, -0.0645, 0.0112, -0.0410, -0.0043, -0.0651, -0.0393, -0.0272, 0.0064, 0.1152, 0.0372, -0.0430, 0.0373, -0.0675, -0.0231, 0.1663, 0.0260, 0.0634, -0.0406, -0.0096,\n -0.0861, 0.0854, -0.0650, 0.0363, 0.0335, -0.0608, 0.0927, 0.1477, -0.0973, -0.1138, 0.0164, -0.0213, 0.0584, -0.0108, -0.0027, 0.0274, -0.1136, -0.0412, -0.0360, -0.1257, 0.1193, 0.0317, 0.0259, -0.0431, 0.0784, -0.0605, -0.0192, -0.1431, 0.0008, -0.0496, -0.0554, -0.0175, 0.0298, 0.0513,\n -0.0049, 0.0228, 0.0188, -0.0407, -0.1527, -0.1804, -0.0587, 0.0684, 0.0451, -0.1313, 0.0625, 0.0944, 0.1362, 0.0255, 0.0538, -0.0502, 0.0441, -0.0838, 0.0035, 0.0226, -0.0592, 0.0959, 0.0911, -0.0246, -0.0485, -0.0095, 0.1257, -0.0474, 0.0569, -0.0723, 0.0795, 0.0336, 0.0118, 0.0469,\n 0.0864, 0.0354, -0.0740, -0.0368, 0.0900, -0.0114, 0.1133, -0.0504, -0.0845, -0.0905, 0.1709, 0.1294, -0.0023, -0.0823, -0.0147, 0.1689, -0.0298, -0.0236, -0.0649, 0.1024, -0.0388, 0.0832, 0.1749, 0.0886, -0.1090, -0.0123, -0.0384, 0.1616, 0.0637, 0.0793, -0.0234, 0.0748, -0.0375, -0.0911,\n 0.1681, 0.0522, -0.0219, 0.0311, 0.0762, 0.0618, -0.0900, -0.0031, 0.0866, 0.0253, -0.0482, 0.1206, 0.0670, 0.0523, 0.1237, 0.0452, 0.1128, -0.0218, -0.0652, 0.0862, 0.0941, 0.0626, -0.0555, 0.1833, 0.0820, -0.0311, -0.0122, -0.0023, 0.0241, -0.0101, 0.0530, 0.0683, 0.0641, 0.1445,\n -0.0450, -0.0556, 0.1252, 0.0002, 0.0997, 0.0598, 0.0778, -0.0029, 0.0091, -0.1753, 0.1677, -0.1025, -0.0493, 0.0518, -0.0248, 0.0118, 0.0882, -0.0006, -0.0519, 0.0879, -0.0606, -0.0302, -0.0081, 0.1068, -0.1080, 0.1058, 0.0376, -0.0179, 0.0472, -0.0306, -0.0381, -0.2400, 0.0506, -0.0253,\n 0.0583, 0.1548, 0.0107, -0.1840, 0.0433, -0.0738, -0.0501, 0.0305, 0.0641, 0.1223, -0.0207, -0.0353, -0.1076, 0.0880, 0.1021, -0.1320, -0.0635, -0.1255, -0.0836, 0.0237, 0.0568, -0.0291, -0.0243, 0.0587, 0.1267, -0.0694, 0.0768, 0.0572, 0.0285, 0.0116, -0.0135, -0.0317, 0.0618, -0.0498,\n 0.0677, -0.0609, -0.0273, 0.0516, 0.0734, 0.1155, -0.0204, 0.0884, 0.0218, 0.0137, -0.0535, -0.1541, -0.0726, -0.0359, 0.1177, -0.0630, 0.0312, -0.1079, 0.0864, -0.0480, -0.0008, 0.0575, -0.0102, 0.1298, -0.0621, 0.0539, 0.0808, 0.0068, 0.0217, -0.0205, -0.0743, 0.0136, -0.0079, -0.0645,\n 0.0093, 0.0665, -0.0922, -0.0842, -0.0632, -0.0117, -0.1025, 0.0124, -0.0509, 0.0696, 0.0120, 0.0114, -0.0706, 0.0743, 0.0555, -0.0864, 0.0252, -0.1602, -0.1493, 0.0231, 0.0878, 0.0204, -0.0268, -0.0165, 0.0065, 0.0092, 0.0770, 0.0321, 0.0889, 0.0691, -0.0804, -0.0260, -0.0670, 0.0286,\n -0.0282, -0.0808, -0.0243, -0.0652, 0.0377, -0.1335, -0.1489, -0.0574, -0.1694, 0.1489, 0.0313, 0.0435, 0.0670, -0.0618, -0.0237, -0.0315, 0.0315, 0.1653, -0.0605, 0.1102, -0.1138, 0.0008, -0.0693, -0.0157, 0.1267, -0.0556, -0.0324, 0.1287, -0.0179, -0.0963, 0.0692, 0.0515, 0.1333, -0.0132,\n 0.0309, 0.0499, 0.0407, -0.0566, -0.1146, 0.0260, 0.0301, -0.1327, -0.0086, -0.0535, 0.0527, -0.0418, 0.0861, 0.1615, 0.0177, -0.0269, 0.0397, 0.1309, -0.0580, -0.0767, 0.0054, 0.0413, -0.1010, -0.0109, -0.0273, -0.0174, -0.0226, 0.0246, -0.0771, -0.0777, -0.1035, -0.1169, -0.0927, -0.0490,\n 0.1351, 0.0714, 0.0285, -0.0246, 0.0259, 0.0652, -0.0753, 0.0145, 0.0594, -0.0609, -0.0650, -0.0012, 0.0622, -0.0348, 0.0406, 0.1013, -0.0060, -0.0991, -0.0345, 0.0881, 0.0829, 0.0064, 0.1420, 0.1197, -0.1036, 0.0963, -0.0864, 0.0393, -0.0846, 0.0854, -0.1097, -0.0075, 0.0323, 0.0690,\n -0.0949, 0.1094, -0.0049, 0.0323, 0.1312, -0.0353, 0.0268, -0.0104, -0.0345, -0.0775, 0.0953, -0.0368, 0.1841, -0.1456, -0.0779, 0.0302, -0.1434, -0.1182, -0.0891, -0.0064, -0.0512, -0.2045, 0.0157, 0.0381, 0.0222, -0.0280, -0.0202, 0.0016, 0.0269, 0.1172, 0.0374, -0.0142, 0.0490, 0.0820,\n 0.1036, 0.0836, 0.0194, -0.0151, 0.0694, 0.1232, 0.0267, -0.0996, -0.0236, -0.0091, 0.0542, -0.1229, 0.1497, -0.0740, 0.0739, -0.0200, -0.0140, -0.1888, -0.0306, 0.0323, 0.0206, -0.0506, 0.0524, -0.0120, -0.0147, 0.1119, 0.0387, -0.0274, -0.0666, 0.0935, -0.1166, -0.1230, 0.0860, 0.0733,\n 0.0259, -0.0437, -0.0582, -0.0419, -0.0160, 0.0272, -0.0454, -0.0060, 0.0715, -0.0080, 0.1288, -0.0359, 0.0708, -0.0872, 0.0421, -0.0753, 0.0153, -0.0247, -0.0706, 0.0189, 0.0675, -0.0569, 0.0273, -0.0291, -0.0825, 0.0879, 0.0509, 0.1039, -0.0620, 0.0101, 0.0300, -0.0126, 0.0710, -0.0128,\n 0.0790, -0.0520, 0.0192, 0.0138, -0.0031, 0.0422, -0.0308, -0.0106, -0.0891, 0.0054, 0.0109, 0.0277, 0.0848, -0.0658, 0.0413, -0.1002, 0.1125, -0.1002, -0.0058, -0.0196, 0.1053, -0.1157, -0.0034, -0.0172, -0.0726, 0.0714, -0.0599, -0.0723, 0.0174, -0.0188, -0.0238, -0.0324, -0.0967, 0.0618,\n -0.0077, 0.0351, 0.0022, -0.0509, 0.0901, -0.0173, 0.0678, 0.0091, 0.0213, 0.0478, 0.0765, -0.0840, 0.0707, -0.0779, 0.0330, -0.0519, 0.0565, -0.0996, -0.0916, -0.0224, -0.0607, -0.0108, 0.0494, 0.0912, 0.0703, 0.0819, 0.0074, 0.0436, -0.0496, -0.0715, 0.0892, 0.0930, -0.0488, 0.0884,\n 0.0082, 0.0490, -0.0198, -0.0955, 0.0161, -0.0193, -0.1538, 0.0087, -0.1232, -0.0554, -0.0117, 0.0045, 0.0385, 0.0140, 0.0251, 0.0668, -0.0685, 0.0309, -0.0846, -0.0737, 0.1155, -0.0193, 0.0024, -0.0365, 0.0793, -0.0397, -0.0560, -0.0968, -0.0042, -0.0123, -0.0089, -0.0122, 0.0259, -0.0424,\n -0.0869, -0.0332, -0.0951, 0.0850, -0.1057, 0.1836, -0.0425, 0.0056, -0.0497, -0.0700, 0.0636, 0.0318, -0.1088, 0.1511, -0.0744, 0.1226, 0.1201, -0.0871, 0.1676, 0.0215, 0.1089, 0.0331, -0.0454, -0.0643, -0.0316, -0.0961, 0.0442, 0.1526, 0.0234, -0.0181, 0.0765, 0.0737, -0.0352, 0.0850,\n -0.0601, -0.0025, 0.0974, -0.0586, 0.1166, 0.0298, -0.0582, -0.0380, 0.1611, -0.1367, 0.0591, 0.0715, 0.0789, -0.0192, -0.0129, -0.0609, -0.0895, -0.0133, 0.1616, -0.1062], dtype=torch.float16), 'perceiver_resampler.layers.1.0.to_q.weight': tensor([[ 0.0194, -0.0086, 0.0033, ..., -0.0130, -0.0089, -0.0125],\n [-0.0436, -0.0211, -0.0741, ..., 0.0263, 0.0086, -0.0072],\n [-0.0245, 0.0075, 0.0044, ..., -0.0078, 0.0191, 0.0349],\n ...,\n [-0.0406, -0.0367, 0.0094, ..., -0.0025, -0.0050, 0.0237],\n [-0.0049, -0.0241, 0.0273, ..., -0.0068, 0.0256, 0.0175],\n [ 0.0265, 0.0104, 0.0020, ..., 0.0693, -0.0046, -0.0053]], dtype=torch.float16), 'perceiver_resampler.layers.1.0.to_kv.weight': tensor([[-0.0125, 0.0191, 0.0296, ..., -0.0424, 0.0221, -0.0014],\n [ 0.0467, -0.0298, 0.0335, ..., 0.0205, 0.0122, -0.0267],\n [ 0.0193, 0.0361, 0.0268, ..., 0.0159, 0.0149, 0.0088],\n ...,\n [ 0.0079, 0.0013, -0.0250, ..., -0.0016, -0.0070, 0.0668],\n [ 0.0010, -0.0267, -0.0192, ..., -0.0006, 0.0104, -0.0113],\n [-0.0036, 0.0090, 0.0305, ..., 0.0272, 0.0240, -0.0384]], dtype=torch.float16), 'perceiver_resampler.layers.1.0.to_out.weight': tensor([[ 0.0042, -0.0216, 0.0410, ..., 0.0030, 0.0493, -0.0262],\n [-0.0586, -0.0706, 0.0138, ..., 0.0468, 0.0006, 0.0101],\n [-0.0097, 0.0205, -0.0143, ..., -0.0306, -0.0126, -0.0384],\n ...,\n [-0.0481, -0.0646, 0.0159, ..., 0.0194, 0.0019, -0.0172],\n [-0.0116, 0.0259, -0.0192, ..., 0.0010, 0.0635, -0.0098],\n [ 0.0185, -0.0068, 0.0145, ..., -0.0860, -0.0174, 0.0303]], dtype=torch.float16), 'perceiver_resampler.layers.1.1.0.weight': tensor([0.8760, 0.7798, 0.8755, 0.8711, 0.7456, 0.8687, 0.8535, 0.8169, 0.8140, 0.8179, 0.9160, 0.9043, 0.8667, 0.7856, 0.8516, 0.7642, 0.8472, 0.8130, 0.8584, 0.8281, 0.9297, 0.8110, 0.7544, 0.8174, 0.8516, 0.8540, 0.8096, 0.8584, 0.9072, 0.8838, 0.7793, 0.6880, 0.8867, 0.9355, 0.8599, 0.8223, 0.8062, 0.7861, 0.9126,\n 0.9004, 0.8335, 0.6885, 0.8535, 0.8281, 0.7686, 0.7715, 0.8252, 0.9111, 0.8501, 0.8838, 0.8848, 0.8457, 0.8677, 0.8315, 0.6826, 0.8418, 0.8516, 0.8804, 0.8745, 0.8052, 0.7651, 0.8481, 0.9180, 0.8584, 0.8599, 0.8501, 0.8325, 0.9316, 0.9292, 0.8970, 0.7148, 0.8774, 0.8677, 0.8672, 0.8545, 0.8145, 0.8394, 0.8857,\n 0.8872, 0.8877, 0.8706, 0.7534, 0.8657, 0.7959, 0.8657, 0.8584, 0.8286, 0.9180, 0.8862, 0.8682, 0.9355, 0.7651, 0.8496, 0.8838, 0.8540, 0.9229, 0.7051, 0.8667, 0.7954, 0.8374, 0.9009, 0.7695, 0.8296, 0.8755, 0.8589, 0.9028, 0.8804, 0.8418, 0.9111, 0.8726, 0.8037, 0.8584, 0.8022, 0.6777, 0.9287, 0.8940, 0.9160,\n 0.8550, 0.9048, 0.8081, 0.8506, 0.9009, 0.8745, 0.8945, 0.8438, 0.8467, 0.7339, 0.8130, 0.7739, 0.9351, 0.8682, 0.7744, 0.8193, 0.9116, 0.8779, 0.8506, 0.9136, 0.8564, 0.7798, 0.8306, 0.8384, 0.8296, 0.8760, 0.7476, 0.9053, 0.8799, 0.7578, 0.8179, 0.8652, 0.8228, 0.8496, 0.8271, 0.9248, 0.8589, 0.7988, 0.8867,\n 0.8970, 0.8071, 0.6748, 0.8574, 0.8862, 0.7651, 0.8687, 0.6870, 0.8433, 0.8652, 0.8496, 0.8574, 0.8525, 0.9019, 0.8037, 0.8599, 0.8818, 0.8945, 0.8340, 0.9165, 0.8086, 0.8496, 0.8423, 0.9380, 0.7705, 0.8477, 0.8110, 0.8066, 0.7788, 0.8701, 0.8242, 0.7568, 0.8491, 0.8706, 0.8901, 0.8936, 0.8896, 0.9023, 0.8042,\n 0.8975, 0.8442, 0.8569, 0.8755, 0.8198, 0.6479, 0.8818, 0.8135, 0.8955, 0.8237, 0.8882, 0.9136, 0.7363, 0.9048, 0.8730, 0.8906, 0.7773, 0.8452, 0.9019, 0.6787, 0.8623, 0.8057, 0.7480, 0.8911, 0.9321, 0.8525, 0.8838, 0.8833, 0.8667, 0.8389, 0.8315, 0.9678, 0.8921, 0.8018, 0.7900, 0.8618, 0.9238, 0.8442, 0.8794,\n 0.8867, 0.8291, 0.8638, 0.9160, 0.9429, 0.7739, 0.7666, 0.7188, 0.7910, 0.8164, 0.8838, 0.8472, 0.8521, 0.7944, 0.8428, 0.9126, 0.7393, 0.7832, 0.8364, 0.8374, 0.8916, 0.9038, 0.9043, 0.8086, 0.8335, 0.7905, 0.9414, 0.8130, 0.8730, 0.8335, 0.8540, 0.6982, 0.8696, 0.8071, 0.7949, 0.8320, 0.7559, 0.9053, 0.8418,\n 0.8774, 0.9214, 0.8442, 0.9229, 0.7832, 0.8330, 0.9458, 0.8467, 0.7842, 0.9424, 0.9395, 0.8916, 0.8408, 0.6694, 0.8872, 0.8896, 0.8687, 0.8926, 0.8667, 0.8535, 0.8950, 0.8594, 0.9087, 0.8042, 0.8970, 0.8359, 0.8369, 0.8638, 0.8853, 0.8506, 0.7681, 0.8340, 0.5962, 0.8530, 0.8452, 0.8564, 0.7900, 0.7993, 0.7812,\n 0.8491, 0.8110, 0.9351, 0.8198, 0.8379, 0.8491, 0.8638, 0.8486, 0.9023, 0.8652, 0.7827, 0.9219, 1.0215, 0.7505, 0.8970, 0.8853, 0.8521, 0.8711, 0.9722, 0.8257, 0.9028, 0.8838, 0.7700, 0.7769, 0.8438, 0.8604, 0.8784, 0.8667, 0.8755, 0.8398, 0.8438, 0.9189, 0.8555, 0.8696, 0.8501, 0.8799, 0.8613, 0.6714, 0.8765,\n 0.7710, 0.9316, 0.8584, 0.9287, 0.8706, 0.8013, 0.7920, 0.8584, 0.8511, 0.9033, 0.8169, 0.8408, 0.8633, 0.8691, 0.8550, 0.8638, 0.8428, 0.8604, 0.8730, 0.8267, 0.8311, 0.8921, 0.9219, 0.8228, 0.8047, 0.8184, 0.9370, 0.7671, 0.8457, 0.8857, 0.8882, 0.8672, 0.8735, 0.9199, 0.8594, 0.8901, 0.8423, 0.8374, 0.8843,\n 0.8770, 0.8672, 0.9380, 0.8364, 0.8716, 0.8618, 0.8604, 0.9082, 0.7529, 0.8862, 0.8667, 0.7866, 0.8081, 0.9448, 0.8931, 0.9004, 0.8062, 0.8262, 0.8379, 0.9229, 0.8779, 0.8120, 0.8677, 0.7876, 0.7471, 0.8418, 0.9136, 0.8872, 0.8975, 0.9106, 0.8511, 0.8979, 0.8066, 0.8198, 0.8208, 0.9160, 0.7183, 0.9102, 0.8623,\n 0.8784, 0.7988, 0.8862, 0.8174, 0.8481, 0.8643, 0.8506, 0.7749, 0.9180, 0.7012, 0.8911, 0.9282, 0.8184, 0.7017, 0.8252, 0.9121, 0.8833, 0.8652, 0.8574, 0.8188, 0.8921, 0.8887, 0.8193, 0.8071, 0.8296, 0.8716, 0.9141, 0.8135, 0.7925, 0.8486, 0.7788, 0.8511, 0.8335, 0.7739, 0.8325, 0.7891, 0.8115, 0.7681, 0.8984,\n 0.8804, 0.8701, 0.9189, 0.7935, 0.9023, 0.8374, 0.7251, 0.8198, 0.8428, 0.9531, 0.8491, 0.7676, 0.7910, 0.7876, 0.8721, 0.8511, 0.7642, 0.6831, 0.8145, 0.8154, 0.8735, 0.8901, 0.8486, 0.8105, 0.7310, 0.8115, 0.7710, 0.8784, 0.9175, 0.8354, 0.8989, 0.7861, 0.8120, 0.8457, 0.8657, 0.7695, 0.7700, 0.9194, 0.8555,\n 0.8979, 0.7871, 0.8071, 0.9126, 0.8872, 0.7700, 0.8335, 0.8896, 0.9028, 0.9180, 0.8501, 0.8506, 0.8237, 0.8960, 0.8936, 0.7915, 0.8647, 0.6040, 0.8481, 0.8823, 0.8486, 0.8359, 0.8379, 0.8418, 0.7578, 0.8848, 0.8872, 0.8115, 0.8374, 0.8203, 0.7642, 0.8989, 0.9160, 0.9038, 0.8574, 0.8423, 0.8267, 0.7256, 0.8755,\n 0.8271, 0.8579, 0.8496, 0.9365, 0.8120, 0.8862, 0.8604, 0.9004, 0.7275, 0.8857, 0.9272, 0.9248, 0.8350, 0.8755, 0.8223, 0.9263, 0.8408, 0.8379, 0.8701, 0.8706, 0.6143, 0.8413, 0.8926, 0.9219, 0.8857, 0.8965, 0.8540, 0.9023, 0.8765, 0.7183, 0.8315, 0.8555, 0.8584, 0.8335, 0.8530, 0.9302, 0.8530, 0.8584, 0.8657,\n 0.7002, 0.8613, 0.8145, 0.8882, 0.8442, 0.8677, 0.8711, 0.7720, 0.8438, 0.8848, 0.8477, 0.9097, 0.8662, 0.8032, 0.9297, 0.8730, 0.8115, 0.8896, 0.7109, 0.8286, 0.8062, 0.8311, 0.8442, 0.7905, 0.8022, 0.8828, 0.8066, 0.8823, 0.7939, 0.8315, 0.8540, 0.7886, 0.8462, 0.8125, 0.8931, 0.8550, 0.8306, 0.8730, 0.8613,\n 0.8501, 0.8906, 0.8682, 0.8521, 0.7437, 0.8218, 0.8638, 0.8428, 0.8677, 0.8818, 0.8994, 0.8691, 0.8638, 0.8721, 0.6694, 0.7793, 0.8394, 0.8984, 0.8716, 0.8462, 0.9043, 0.8428, 0.8672, 0.8525, 0.8154, 0.9155, 0.6406, 0.8555, 0.8418, 0.7656, 0.8511, 0.8301, 0.8784, 0.8535, 0.8984, 0.9233, 0.8867, 0.8760, 0.8794,\n 0.8730, 0.8569, 0.8936, 0.7886, 0.8442, 0.8379, 0.8213, 0.8657, 0.7910, 0.7603, 0.8604, 0.8823, 0.8872, 0.8066, 0.7759, 0.7656, 0.8911, 0.8364, 0.8652, 0.8721, 0.9033, 0.8164, 0.9253, 0.8442, 0.8369, 0.9233, 0.8203, 0.8535, 0.8315, 0.7866, 0.7109, 0.8613, 0.8828, 0.8745, 0.8618, 0.8926, 0.6162, 0.8618, 0.7417,\n 0.7939, 0.8691, 0.8613, 0.8857, 0.8931, 0.9014, 0.8623, 0.8716, 0.8696, 0.8101, 0.8809, 0.8091, 0.9209, 0.8877, 0.9365, 0.8530, 0.7441, 0.8857, 0.9150, 0.8022, 0.8271, 0.8203, 0.8516, 0.8213, 0.8501, 0.9351, 0.8862, 0.8682, 0.8823, 0.9004, 0.8394, 0.8599, 0.8506, 0.9087, 0.8652, 0.8052, 0.8691, 0.8921, 0.8735,\n 0.7656, 0.8521, 0.8677, 0.8438, 0.8506, 0.7754, 0.8540, 0.8032, 0.8760, 0.8921, 0.8438, 0.7422, 0.8716, 0.8149, 0.8711, 0.7949, 0.9287, 0.9243, 0.8916, 0.8232, 0.8652, 0.8213, 0.8735, 0.8711, 0.5288, 0.7681, 0.8135], dtype=torch.float16), 'perceiver_resampler.layers.1.1.0.bias': tensor([ 1.0559e-01, -4.7089e-02, -2.0493e-02, 1.2286e-01, 2.4353e-02, -7.2205e-02, 4.4289e-03, 1.8936e-02, -5.7587e-02, -1.4465e-02, -1.7004e-01, 1.8970e-01, -8.7214e-04, 3.1616e-02, -8.8074e-02, -8.7280e-02, 6.5979e-02, -1.2927e-01, -1.6003e-01, 3.8391e-02, 1.0504e-01, 2.0275e-03, 1.3867e-01, -1.4404e-01,\n 1.4050e-01, -8.5815e-02, -1.1322e-02, 9.1064e-02, -2.7466e-02, -1.0223e-01, 1.0468e-01, 8.8577e-03, 1.5735e-01, -1.5649e-01, -1.6772e-01, -1.2329e-01, -9.4299e-02, -5.5756e-02, 1.2744e-01, 1.7822e-01, 4.1260e-02, 1.5114e-02, 4.8279e-02, 1.1612e-02, -6.8970e-02, 1.2549e-01, 6.5613e-02, -1.1700e-01,\n 1.0760e-01, 4.2877e-02, -4.9072e-02, -1.2622e-01, -7.6111e-02, -3.2257e-02, -1.7493e-01, -5.9570e-02, 4.0527e-02, -1.7786e-01, 1.7426e-02, -1.2201e-01, 5.0446e-02, -2.3758e-02, -1.4404e-01, 3.2768e-03, 6.9824e-02, -3.9185e-02, -2.0477e-02, -1.8945e-01, -3.2196e-02, -1.8042e-01, 1.2769e-01, -4.2572e-03,\n 6.5918e-02, 8.3740e-02, -3.2410e-02, -4.1008e-03, 1.1523e-01, 1.4954e-01, -1.2878e-01, -5.3436e-02, -1.1523e-01, 1.1969e-01, -1.6492e-01, 2.9388e-02, -6.7322e-02, -1.2152e-01, 2.0752e-02, 1.0895e-01, 1.3574e-01, 1.4514e-01, -1.8665e-01, -6.1951e-02, -1.2671e-01, 3.1372e-02, 7.2388e-02, 1.1652e-01,\n -1.2291e-02, 1.4478e-01, 1.1157e-01, -1.6003e-01, -6.0158e-03, -8.6670e-02, -5.5275e-03, 1.4038e-01, -9.4971e-02, -7.3547e-02, -7.9712e-02, 1.0022e-01, -1.3025e-01, -1.4587e-01, 4.8340e-02, 3.2104e-02, 6.0892e-04, 4.6501e-03, -9.1064e-02, -3.0884e-02, -1.7786e-01, 5.3040e-02, 6.3171e-02, 2.7176e-02,\n 1.0779e-01, 1.4783e-01, 4.1168e-02, 8.4167e-02, 9.3140e-02, -4.8615e-02, -1.1848e-02, 6.4148e-02, 1.2000e-01, 1.3733e-01, 3.8391e-02, -5.3101e-02, -3.1052e-02, -1.4905e-01, 5.6122e-02, -5.1819e-02, 1.5979e-01, -1.1804e-01, 3.2898e-02, -4.9469e-02, 1.7609e-02, -1.2402e-01, -8.8379e-02, -6.1310e-02,\n 1.3452e-01, 6.3904e-02, 3.2013e-02, -5.5389e-02, -1.3489e-01, 6.2012e-02, -5.1537e-03, -1.3123e-01, 1.5430e-01, 9.0210e-02, 8.0505e-02, 1.6785e-01, -1.7358e-01, -7.5760e-03, -1.0876e-01, 1.1682e-01, 5.5420e-02, 1.2878e-01, -5.7892e-02, 1.9006e-01, -3.7323e-02, 2.1561e-02, 8.9722e-02, 2.7145e-02,\n -9.1492e-02, -1.3171e-01, -5.7907e-03, -6.7627e-02, -9.9609e-02, 8.2474e-03, 1.1420e-01, 1.9873e-01, 6.5857e-02, -9.2407e-02, -8.1909e-02, 1.6797e-01, 9.4116e-02, -4.5288e-02, -1.4954e-01, -4.1016e-02, -8.2092e-02, 5.8197e-02, -1.2213e-01, 1.4380e-01, 8.0872e-02, -8.1848e-02, 5.7526e-02, -3.0792e-02,\n -1.2817e-01, 1.0272e-01, 1.3184e-01, 6.1798e-02, -9.7412e-02, -1.5259e-02, -2.2186e-02, 3.5248e-02, -4.6112e-02, -1.0504e-01, -7.3486e-02, -1.6028e-01, -5.7770e-02, -8.8989e-02, 1.0352e-01, 7.6111e-02, -9.2712e-02, 9.3750e-02, -1.5198e-01, 2.9572e-02, 1.2366e-01, 1.2683e-01, -1.7041e-01, -9.4543e-02,\n 1.2688e-02, 3.3438e-05, 8.7280e-02, -1.7310e-01, 5.3406e-02, 1.1395e-01, 5.3955e-02, 2.5330e-02, 4.0039e-02, -4.7485e-02, -1.9019e-01, -4.7150e-02, 1.0486e-01, -4.8706e-02, 3.1235e-02, -2.0422e-01, -1.2463e-01, -4.3335e-02, 8.9111e-02, -7.7881e-02, 5.4199e-02, 1.6089e-01, -1.9458e-01, -1.2891e-01,\n -4.0955e-02, 3.5400e-02, -7.8491e-02, -4.9438e-02, 1.2323e-01, -5.6549e-02, -1.4160e-01, -6.9153e-02, 1.9775e-02, -1.6321e-01, 1.0510e-01, -1.3885e-02, -1.4868e-01, -9.8389e-02, -1.4185e-01, 1.1890e-01, 9.1980e-02, -5.3345e-02, -5.7312e-02, -9.4910e-03, 1.7078e-01, -1.1011e-01, -1.0986e-01, -6.4270e-02,\n 6.7078e-02, -1.1407e-01, -9.0881e-02, -1.1177e-02, -2.5330e-02, -5.6427e-02, -1.3367e-01, -1.5137e-01, 4.8370e-02, 5.6915e-02, -1.8164e-01, 6.6040e-02, -1.4758e-01, -2.9037e-02, -7.4829e-02, 1.3135e-01, -4.2648e-03, 1.5588e-01, -1.8274e-01, 1.4868e-01, 1.8872e-01, -1.8021e-02, 2.2980e-02, 2.9907e-02,\n -1.5662e-01, -4.7180e-02, 1.3049e-01, -5.5267e-02, -2.2186e-02, 5.8289e-02, 1.2636e-03, -6.8359e-02, -2.1271e-02, -1.3037e-01, -9.6924e-02, 1.4038e-01, 1.7029e-02, 1.3831e-01, 2.1057e-02, 1.5356e-01, -4.1046e-02, -7.3853e-02, 1.0400e-01, -1.3428e-01, -8.2153e-02, 1.5845e-01, 6.8909e-02, 1.6174e-01,\n 9.4299e-02, -9.3079e-02, -1.2549e-01, -6.8726e-02, 2.2919e-02, 8.2947e-02, 1.3977e-01, -7.8918e-02, -9.5276e-02, 1.6028e-01, 8.3008e-03, 1.7053e-01, 1.7883e-01, -3.9642e-02, -1.2598e-01, 8.0994e-02, -3.7506e-02, -2.5818e-02, -1.5479e-01, 6.8542e-02, -4.8431e-02, -2.7420e-02, -3.7842e-02, -1.3879e-01,\n 1.1969e-01, 3.0624e-02, -7.0190e-02, 9.4666e-02, -1.3501e-01, 1.0211e-01, 1.4551e-01, -1.5161e-01, -1.8005e-02, -1.2439e-01, 3.9444e-03, -1.3965e-01, -3.3813e-02, 5.5847e-02, 3.1708e-02, 3.5248e-02, 1.3782e-01, 8.3862e-02, -1.5845e-01, 1.0986e-01, -2.5253e-02, 1.1444e-02, -4.7424e-02, -1.8110e-03,\n -1.0321e-01, 6.6101e-02, 7.8049e-03, -1.0522e-01, 1.1871e-01, -4.0710e-02, -8.2336e-02, -6.3843e-02, 1.4908e-02, 2.5955e-02, 1.0596e-01, -3.9703e-02, 6.7932e-02, 1.6699e-01, 3.7476e-02, -9.1309e-02, 5.8441e-02, 1.3440e-01, 4.4067e-02, -1.1459e-02, 1.5405e-01, 5.1025e-02, 5.5267e-02, -1.3293e-01,\n -2.8244e-02, -3.2444e-03, 1.2250e-01, -4.2023e-02, -1.0944e-01, 1.1108e-01, 3.6883e-04, 1.6443e-01, 1.6968e-01, -2.5604e-02, -1.0107e-01, -5.2094e-02, 2.6154e-02, -3.2410e-02, -8.5327e-02, -1.1957e-01, -1.3831e-01, -1.1963e-02, -9.7351e-02, -1.2634e-01, 7.5500e-02, 9.9548e-02, 8.8379e-02, 5.4596e-02,\n 3.2410e-02, 1.2201e-01, -3.2013e-02, 9.5337e-02, -1.1084e-01, 1.8726e-01, 1.8188e-01, 3.6926e-02, 9.6924e-02, -1.2598e-01, -1.3147e-01, -1.4343e-01, -1.1981e-01, 9.8877e-02, 7.4341e-02, 2.5681e-02, 1.3074e-01, -1.6138e-01, -3.0457e-02, -1.4539e-01, 1.2085e-01, -4.0894e-02, 8.9417e-02, 3.8361e-02,\n -1.1688e-01, 7.5867e-02, 1.4441e-01, -1.0797e-01, -3.5431e-02, 1.2549e-01, -8.4900e-02, -3.6896e-02, -1.4880e-01, 1.1810e-02, -5.7953e-02, -9.5749e-03, -1.0986e-02, 4.6448e-02, 6.0791e-02, -1.0187e-01, -8.8623e-02, 1.1694e-01, 8.3847e-03, 6.8420e-02, 1.4372e-03, 7.5195e-02, -7.2693e-02, -1.6455e-01,\n -1.6312e-02, 9.8450e-02, 1.2527e-02, -1.2079e-01, 4.7760e-02, 3.8422e-02, 7.6233e-02, -5.5603e-02, 3.4210e-02, -6.3293e-02, 4.5441e-02, 6.9153e-02, 2.6226e-03, 6.4735e-03, 1.4917e-01, 7.7057e-04, 1.2231e-01, -3.6133e-02, 1.7670e-02, 7.1899e-02, -1.2866e-01, -8.8623e-02, -8.3801e-02, -3.5370e-02,\n -9.0027e-02, -2.1225e-02, 1.1926e-01, -1.9531e-02, 9.0088e-02, 2.3804e-02, 8.4534e-02, -6.8550e-03, -9.0576e-02, 4.0466e-02, 4.4922e-02, -7.9041e-02, 1.9577e-02, 1.3831e-01, 1.1633e-01, -1.2323e-01, -1.9238e-01, -1.2219e-01, -1.4172e-01, -8.6853e-02, 4.4312e-02, -3.0777e-02, 6.8176e-02, -4.1962e-03,\n 4.2496e-03, -1.7749e-01, 1.2634e-01, 1.7358e-01, -1.1206e-01, -6.7932e-02, 1.0852e-01, -1.0785e-01, -1.4172e-01, -1.2001e-02, -1.3257e-01, 4.5685e-02, -6.0547e-02, -3.1311e-02, -1.3199e-02, 5.7220e-02, -1.2469e-01, 3.6163e-02, -1.3525e-01, 1.0730e-01, 9.8022e-02, -4.0375e-02, 1.3403e-01, 1.6162e-01,\n 1.2262e-01, 1.2256e-01, 6.2683e-02, 1.6577e-01, -9.2102e-02, 3.8116e-02, -1.1517e-01, 2.4200e-02, -1.3359e-02, 1.4551e-01, -9.7595e-02, -1.7273e-01, -1.1322e-01, 1.5063e-01, -4.6051e-02, -9.4299e-02, -1.4771e-01, -1.6382e-01, -5.5817e-02, 2.1545e-02, -9.1736e-02, -1.5381e-01, -6.9153e-02, 1.3037e-01,\n 6.7383e-02, 2.8610e-02, -1.1938e-01, 1.0876e-01, -1.5332e-01, 1.6223e-01, -7.9834e-02, -9.3994e-02, 2.2934e-02, 1.9543e-01, -7.0129e-02, -1.3718e-02, 7.9041e-02, 7.1678e-03, 7.6660e-02, 1.3306e-01, 6.5735e-02, -1.6321e-01, -7.8369e-02, 6.0577e-02, 6.2012e-02, -1.3037e-01, 1.3611e-01, 2.9251e-02,\n -4.8065e-02, -1.0913e-01, -7.4036e-02, 8.1543e-02, 7.6172e-02, 1.5369e-01, 6.5247e-02, -9.0820e-02, 1.2866e-01, 4.1901e-02, -1.8219e-02, 3.0182e-02, -1.2024e-01, 1.2164e-01, -1.3855e-01, 1.0321e-01, -6.9275e-02, 1.0400e-01, -2.5620e-02, 6.5552e-02, 1.8469e-01, 7.2289e-03, -2.3865e-02, 1.4709e-01,\n -1.3342e-01, 4.2938e-02, 1.4893e-01, 2.0462e-02, -4.7455e-02, -4.0710e-02, 1.2842e-01, -6.9214e-02, -2.1149e-02, 1.3220e-01, -1.2585e-01, -5.3444e-03, -1.4307e-01, 7.2815e-02, -8.8745e-02, 8.1177e-02, -4.0192e-02, -1.1511e-01, 1.0535e-01, 2.6520e-02, 8.5373e-03, -6.5918e-02, -1.1700e-01, -9.2468e-02,\n -5.1819e-02, 1.4001e-01, -1.8768e-02, 8.0933e-02, -1.7163e-01, 1.1975e-01, -9.4238e-02, 9.3201e-02, -5.3894e-02, 1.4355e-01, -1.2299e-01, 5.1575e-02, 7.2510e-02, -1.9318e-02, -3.4485e-02, 1.4221e-01, -1.7395e-03, 8.1543e-02, -5.4352e-02, 1.0010e-01, 1.3440e-01, -7.3547e-02, 1.4000e-02, -1.0980e-01,\n -2.1545e-02, 1.2360e-01, -1.6327e-02, 1.4534e-02, -3.8071e-03, 1.1314e-02, -1.7929e-02, -4.7729e-02, -7.2876e-02, 1.0132e-02, -9.4299e-02, 1.6602e-01, -2.7954e-02, 9.0515e-02, -1.3013e-01, 1.2634e-01, 1.4404e-01, 1.0193e-01, 9.5398e-02, -1.7761e-02, 2.3453e-02, -5.8777e-02, -9.1736e-02, 1.1414e-02,\n 1.1041e-01, -2.4323e-02, 1.1554e-01, 3.9978e-02, -1.6251e-02, 2.6352e-02, -7.9407e-02, -6.7688e-02, -1.7944e-02, -1.4624e-01, -3.4637e-03, 1.4929e-01, -5.1544e-02, 2.0615e-02, 1.1981e-01, -6.0638e-02, 1.4771e-01, 8.8867e-02, 7.1167e-02, -4.7646e-03, 3.1815e-03, 3.4210e-02, 4.3030e-02, -8.2336e-02,\n -2.1927e-02, -7.6172e-02, 1.2469e-01, 1.1920e-01, -9.9854e-02, -1.7258e-02, 4.7943e-02, 7.2594e-03, -1.0449e-01, 6.9519e-02, 1.9067e-01, 1.3818e-01, 3.1464e-02, 2.4734e-02, -9.4421e-02, -1.8509e-02, -6.9885e-02, 8.0994e-02, 1.3293e-01, 1.0474e-01, 1.7456e-01, -7.0557e-02, 6.8604e-02, -1.2115e-01,\n 8.5205e-02, 5.5351e-03, -4.2847e-02, 1.0864e-01, -9.4727e-02, 1.6663e-02, 1.1627e-01, -1.9617e-01, 1.3647e-01, -1.2476e-01, -1.1420e-01, 1.3452e-01, -1.1597e-01, 8.6121e-02, -1.1377e-01, -4.5868e-02, 8.8440e-02, 6.1188e-02, 7.4768e-02, 8.6914e-02, -9.0149e-02, -8.2626e-03, -3.9001e-02, -8.7830e-02,\n -8.8440e-02, -9.8572e-02, 2.5085e-02, -1.0272e-01, 9.7580e-03, -3.5065e-02, -1.5393e-01, 1.0278e-01, -5.3802e-02, -7.0251e-02, 8.0139e-02, 1.6101e-01, -8.3862e-02, 1.2140e-01, -1.2140e-01, -7.7332e-02, -8.1177e-02, 1.2769e-01, 5.3070e-02, 9.3201e-02, 7.6172e-02, 1.0162e-01, -1.0803e-01, 9.5337e-02],\n dtype=torch.float16), 'perceiver_resampler.layers.1.1.1.weight': tensor([[ 0.0072, -0.0214, 0.0046, ..., 0.0155, 0.0296, 0.0099],\n [ 0.0120, -0.0180, 0.0327, ..., 0.0120, -0.0205, 0.0174],\n [-0.0111, -0.0159, 0.0089, ..., -0.0241, 0.0120, -0.0293],\n ...,\n [ 0.0061, 0.0196, 0.0041, ..., 0.0142, 0.0300, -0.0172],\n [-0.0403, -0.0425, 0.0425, ..., -0.0350, 0.0617, -0.0200],\n [-0.0137, 0.0262, 0.0500, ..., 0.0085, 0.0399, 0.0032]], dtype=torch.float16), 'perceiver_resampler.layers.1.1.3.weight': tensor([[-0.0103, -0.0249, -0.0071, ..., -0.0133, -0.0111, -0.0232],\n [ 0.0072, -0.0017, -0.0133, ..., 0.0093, -0.0363, -0.0386],\n [ 0.0221, 0.0071, -0.0087, ..., -0.0022, 0.0006, -0.0188],\n ...,\n [-0.0251, 0.0392, 0.0216, ..., 0.0779, -0.0274, 0.0182],\n [-0.0049, 0.0042, 0.0534, ..., 0.0202, -0.0014, 0.0227],\n [ 0.0158, -0.0044, -0.0285, ..., 0.0107, 0.0215, -0.0209]], dtype=torch.float16), 'perceiver_resampler.layers.2.0.norm1.weight': tensor([0.7285, 0.7319, 0.7065, 0.7954, 0.7510, 0.7715, 0.7700, 0.7017, 0.7676, 0.7583, 0.6919, 0.7441, 0.7559, 0.8003, 0.7466, 0.6855, 0.7866, 0.6943, 0.7979, 0.7935, 0.8413, 0.8008, 0.7861, 0.7422, 0.7495, 0.7231, 0.7515, 0.6919, 0.6904, 0.7393, 0.7773, 0.8262, 0.7715, 0.7632, 0.7461, 0.7397, 0.6250, 0.7554, 0.7266,\n 0.7612, 0.7671, 0.7817, 0.8130, 0.7480, 0.7456, 0.7935, 0.7168, 0.7471, 0.7388, 0.6616, 0.7124, 0.7827, 0.7402, 0.7119, 0.7432, 0.7856, 0.7363, 0.6899, 0.7471, 0.7930, 0.7354, 0.7891, 0.7549, 0.7544, 0.7915, 0.7539, 0.7896, 0.8105, 0.7666, 0.7148, 0.7852, 0.7920, 0.7886, 0.7998, 0.7393, 0.8198, 0.7188, 0.7236,\n 0.8013, 0.6968, 0.7876, 0.7900, 0.7339, 0.6943, 0.7793, 0.7422, 0.7778, 0.7656, 0.8047, 0.7178, 0.7739, 0.7119, 0.7441, 0.6772, 0.7837, 0.7388, 0.7769, 0.6890, 0.7739, 0.7817, 0.7310, 0.7729, 0.7109, 0.7695, 0.7061, 0.7373, 0.7266, 0.7856, 0.7319, 0.7891, 0.7456, 0.7935, 0.8174, 0.7554, 0.7241, 0.7676, 0.7695,\n 0.8433, 0.7144, 0.7397, 0.7905, 0.7192, 0.7524, 0.8335, 0.7241, 0.7217, 0.7739, 0.7563, 0.6167, 0.7505, 0.7056, 0.6768, 0.7559, 0.7817, 0.7383, 0.7759, 0.7861, 0.7090, 0.7808, 0.7373, 0.6934, 0.7065, 0.6807, 0.8311, 0.7563, 0.7827, 0.7881, 0.7202, 0.7954, 0.7637, 0.7349, 0.7271, 0.7583, 0.7227, 0.7759, 0.7188,\n 0.7026, 0.6997, 0.8057, 0.7148, 0.6597, 0.7554, 0.7725, 0.7627, 0.7202, 0.7778, 0.7710, 0.7485, 0.7373, 0.7598, 0.7681, 0.7231, 0.7607, 0.7402, 0.7456, 0.7207, 0.7490, 0.7300, 0.6270, 0.7119, 0.7749, 0.6606, 0.8027, 0.7832, 0.8232, 0.7344, 0.5908, 0.7783, 0.7100, 0.7334, 0.7866, 0.6162, 0.7656, 0.7749, 0.7373,\n 0.7368, 0.6919, 0.6616, 0.7339, 0.8154, 0.7573, 0.7910, 0.7373, 0.8037, 0.7158, 0.7686, 0.6196, 0.7793, 0.7808, 0.7622, 0.7036, 0.7339, 0.6543, 0.7764, 0.8096, 0.6963, 0.7583, 0.7778, 0.7852, 0.7271, 0.7471, 0.7593, 0.7461, 0.7334, 0.7627, 0.6777, 0.7808, 0.7715, 0.6802, 0.7544, 0.7368, 0.7017, 0.8120, 0.7163,\n 0.7368, 0.6938, 0.8018, 0.7637, 0.6929, 0.7144, 0.6938, 0.7681, 0.7002, 0.7725, 0.7188, 0.7466, 0.7559, 0.7734, 0.7559, 0.7925, 0.7490, 0.7104, 0.7026, 0.6919, 0.6802, 0.7715, 0.7832, 0.7632, 0.7427, 0.7485, 0.7241, 0.7622, 0.7754, 0.7349, 0.7314, 0.7241, 0.7173, 0.7769, 0.7412, 0.7578, 0.7749, 0.7070, 0.8594,\n 0.7222, 0.7676, 0.7764, 0.7568, 0.7070, 0.7617, 0.6729, 0.7896, 0.7686, 0.7666, 0.6729, 0.7832, 0.6909, 0.7427, 0.7793, 0.7529, 0.7656, 0.7280, 0.6753, 0.7090, 0.6338, 0.7656, 0.7139, 0.7041, 0.7783, 0.7461, 0.6797, 0.7671, 0.6924, 0.6221, 0.7905, 0.7510, 0.6318, 0.7241, 0.7529, 0.7310, 0.7251, 0.7451, 0.7070,\n 0.7173, 0.7637, 0.7578, 0.7637, 0.7300, 0.7710, 0.7534, 0.7793, 0.7007, 0.7495, 0.7012, 0.6719, 0.7974, 0.7612, 0.7417, 0.7251, 0.7471, 0.8081, 0.7285, 0.7212, 0.7339, 0.7480, 0.7222, 0.7661, 0.7729, 0.7368, 0.7969, 0.7866, 0.7671, 0.7725, 0.7285, 0.7129, 0.7544, 0.7466, 0.7593, 0.7861, 0.8354, 0.7256, 0.7471,\n 0.7148, 0.6938, 0.7441, 0.7651, 0.8027, 0.7603, 0.7441, 0.7832, 0.7378, 0.7559, 0.7651, 0.7310, 0.8306, 0.7041, 0.7002, 0.7441, 0.7764, 0.7578, 0.7798, 0.7432, 0.8052, 0.7715, 0.7656, 0.7505, 0.7876, 0.7959, 0.7344, 0.8105, 0.7910, 0.7319, 0.7573, 0.7559, 0.7515, 0.7310, 0.7275, 0.7305, 0.6982, 0.7319, 0.7378,\n 0.7515, 0.7749, 0.7983, 0.6792, 0.7559, 0.7397, 0.7642, 0.7021, 0.7939, 0.7656, 0.8208, 0.7417, 0.7827, 0.7876, 0.7461, 0.7944, 0.6978, 0.7666, 0.7515, 0.7339, 0.7744, 0.7490, 0.7241, 0.7661, 0.6426, 0.7544, 0.7539, 0.7554, 0.7759, 0.7241, 0.6763, 0.7651, 0.8623, 0.7617, 0.7568, 0.7793, 0.7622, 0.7183, 0.7144,\n 0.6807, 0.8271, 0.7261, 0.7822, 0.7812, 0.7607, 0.7627, 0.7554, 0.7485, 0.7451, 0.7930, 0.7642, 0.7417, 0.7251, 0.7031, 0.7158, 0.7749, 0.8066, 0.8003, 0.7368, 0.7686, 0.7051, 0.7397, 0.7593, 0.7710, 0.7637, 0.7163, 0.6890, 0.7422, 0.7134, 0.7969, 0.6187, 0.7451, 0.6772, 0.6934, 0.6943, 0.7554, 0.7607, 0.7422,\n 0.7446, 0.7524, 0.6680, 0.7705, 0.7734, 0.7271, 0.7632, 0.7402, 0.7646, 0.7886, 0.7681, 0.6846, 0.7104, 0.7402, 0.8037, 0.7148, 0.7974, 0.6792, 0.7480, 0.6875, 0.7622, 0.7046, 0.6924, 0.7129, 0.7515, 0.7310, 0.7632, 0.7539, 0.7388, 0.7881, 0.7651, 0.7402, 0.7373, 0.7773, 0.7544, 0.7310, 0.7847, 0.7183, 0.7969,\n 0.7393, 0.7710, 0.7505, 0.8018, 0.7466, 0.7402, 0.7632, 0.7588, 0.8037, 0.8066, 0.7031, 0.7305, 0.7671, 0.7461, 0.7549, 0.7769, 0.7944, 0.8193, 0.7437, 0.6787, 0.7539, 0.7378, 0.6729, 0.7017, 0.6631, 0.7837, 0.7310, 0.8252, 0.7476, 0.6836, 0.7041, 0.7642, 0.7490, 0.7695, 0.7578, 0.7949, 0.7695, 0.6938, 0.8159,\n 0.7285, 0.7900, 0.7729, 0.7798, 0.7222, 0.7993, 0.7334, 0.8193, 0.8032, 0.7021, 0.8301, 0.7329, 0.8145, 0.7246, 0.7549, 0.7246, 0.6938, 0.7637, 0.8135, 0.7344, 0.7637, 0.7852, 0.7793, 0.6860, 0.7290, 0.7690, 0.7734, 0.6919, 0.6978, 0.7471, 0.7686, 0.7480, 0.6831, 0.7349, 0.7080, 0.7808, 0.7549, 0.8149, 0.7397,\n 0.7207, 0.7808, 0.7446, 0.7549, 0.7476, 0.6807, 0.7212, 0.7285, 0.7998, 0.7119, 0.8062, 0.8467, 0.7188, 0.7563, 0.6748, 0.7090, 0.6953, 0.7041, 0.6797, 0.7197, 0.7329, 0.7021, 0.7168, 0.8047, 0.6733, 0.7520, 0.7769, 0.6982, 0.7173, 0.6982, 0.7764, 0.7417, 0.7886, 0.7427, 0.7788, 0.7256, 0.7949, 0.7295, 0.7939,\n 0.7822, 0.6953, 0.7183, 0.7271, 0.7886, 0.7676, 0.7334, 0.7969, 0.7661, 0.7744, 0.7241, 0.7617, 0.7656, 0.8232, 0.7866, 0.7495, 0.7065, 0.7324, 0.7573, 0.7817, 0.7202, 0.8052, 0.8286, 0.7686, 0.7466, 0.7681, 0.7251, 0.7432, 0.7153, 0.6826, 0.7593, 0.7080, 0.6924, 0.7993, 0.7568, 0.8579, 0.7480, 0.7998, 0.8052,\n 0.6577, 0.7153, 0.8052, 0.8369, 0.7729, 0.7441, 0.7290, 0.6777, 0.7422, 0.7593, 0.7163, 0.7837, 0.7480, 0.7349, 0.7700, 0.7871, 0.7153, 0.7764, 0.7466, 0.7393, 0.8018, 0.7544, 0.6440, 0.7520, 0.7524, 0.7397, 0.6860, 0.7617, 0.7134, 0.7598, 0.6821, 0.7739, 0.7588, 0.6978, 0.7578, 0.7773, 0.7764, 0.7271, 0.7935,\n 0.7847, 0.6904, 0.7266, 0.7412, 0.7451, 0.7490, 0.8120, 0.7456, 0.6812, 0.7363, 0.7368, 0.7627, 0.7725, 0.7549, 0.7944, 0.7627, 0.7305, 0.6260, 0.8086, 0.7808, 0.7471, 0.7393, 0.7681, 0.7690, 0.7817, 0.6724, 0.7495, 0.7651, 0.7397, 0.7773, 0.7710, 0.6709, 0.7344, 0.7324, 0.7368, 0.7319, 0.7368, 0.7095, 0.7354,\n 0.7358, 0.8184, 0.6841, 0.7139, 0.7295, 0.7134, 0.7705, 0.7769, 0.7803, 0.7559, 0.7529, 0.8052, 0.7876, 0.7891, 0.6924, 0.7646, 0.5928, 0.7290, 0.7256, 0.7583, 0.6699, 0.7822, 0.7754, 0.7651, 0.7686, 0.7007, 0.7319], dtype=torch.float16), 'perceiver_resampler.layers.2.0.norm1.bias': tensor([ 8.3466e-03, 8.0566e-03, 6.3416e-02, 8.8882e-03, -1.9592e-02, 3.3234e-02, 2.8320e-02, -1.4206e-02, -1.5930e-02, -2.9953e-02, 2.6062e-02, 8.7814e-03, -1.1963e-02, 3.6255e-02, -6.9733e-03, -2.6367e-02, -7.4387e-03, 1.1406e-02, -9.4070e-03, 2.6291e-02, 7.1907e-03, 3.7262e-02, 3.0396e-02, -2.4353e-02,\n 2.1118e-02, 1.0345e-02, 5.3368e-03, 5.7251e-02, -2.4582e-02, 3.4370e-03, 3.1311e-02, 4.6295e-02, 3.6163e-02, 6.4507e-03, -4.5609e-04, 1.6846e-02, 8.4305e-03, -1.0826e-02, 1.2947e-02, -7.8125e-03, -2.9694e-02, 1.3969e-02, 2.2675e-02, -3.4760e-02, 3.2349e-02, -3.1403e-02, -3.0579e-02, 8.2016e-03,\n 1.8295e-02, 2.9846e-02, 3.2745e-02, -2.6398e-02, -1.7536e-04, 1.0719e-02, -4.0207e-03, -8.0200e-02, -2.2293e-02, -3.8086e-02, -3.1525e-02, 3.1052e-02, 3.4943e-02, -3.0716e-02, -1.4404e-02, -1.7975e-02, -1.6861e-02, -1.0323e-02, 7.9956e-03, 3.8513e-02, -4.0863e-02, 7.1838e-02, -9.9335e-03, 1.6388e-02,\n 1.0223e-02, -1.3285e-03, 3.5896e-03, 2.4765e-02, 8.1787e-03, -2.4033e-02, -3.2410e-02, 5.4260e-02, -9.3002e-03, 2.6443e-02, 2.4246e-02, -3.0365e-02, -6.4011e-03, 1.6205e-02, 1.0048e-02, 1.9012e-02, 6.5994e-03, -7.8440e-04, -3.6896e-02, -1.1848e-02, -2.7283e-02, -3.9001e-02, -1.7792e-02, 3.0884e-02,\n -1.5945e-02, -5.0659e-03, -3.3447e-02, 2.1164e-02, -2.6825e-02, 1.2787e-02, 5.7411e-03, -1.0513e-02, -9.4681e-03, -4.9782e-03, 2.7924e-02, 2.4841e-02, -1.3306e-02, -2.5772e-02, -2.8076e-02, 2.1652e-02, -1.6220e-02, 1.4969e-02, 4.6959e-03, 2.2705e-02, -1.5106e-02, 2.2385e-02, 8.9645e-03, -1.2484e-03,\n -3.9368e-02, 1.7288e-02, 2.0691e-02, 1.5312e-02, -2.5439e-04, -2.1652e-02, -1.5961e-02, -4.8981e-02, 2.6207e-03, -1.7975e-02, -1.8555e-02, 5.7037e-02, 2.9160e-02, 2.2995e-02, 3.3813e-02, -3.0319e-02, 4.6844e-03, 4.3640e-02, -3.9787e-03, -3.0716e-02, -2.2446e-02, -2.9030e-03, 4.2725e-03, -3.1433e-02,\n -3.2318e-02, 1.2497e-02, -2.0809e-03, 2.0493e-02, -3.2501e-02, 3.5715e-04, -3.0121e-02, 4.6326e-02, 1.4221e-02, -6.4697e-03, 2.6611e-02, -3.4180e-02, 1.7502e-02, 3.2410e-02, -3.4912e-02, 6.6406e-02, 3.3417e-02, -2.3575e-02, 2.7634e-02, -5.3711e-02, 5.5145e-02, -2.8778e-02, -8.0795e-03, -1.0101e-02,\n 2.9022e-02, 1.9180e-02, -6.1035e-03, -5.2551e-02, 1.8753e-02, -1.5915e-02, 5.6686e-03, -8.8654e-03, -4.8859e-02, -1.7822e-02, 4.7882e-02, 3.3875e-02, -2.4048e-02, 1.2276e-02, 5.1346e-03, 2.5085e-02, -1.6861e-02, 7.4425e-03, -2.3108e-01, -4.5593e-02, -5.9906e-02, 1.1986e-02, -1.0309e-03, 2.5528e-02,\n -2.5192e-02, -3.5034e-02, -3.1082e-02, -3.1662e-04, -4.8256e-03, -1.4526e-02, -4.8828e-02, -2.1076e-03, -2.6657e-02, 2.5681e-02, -1.3710e-02, 7.5722e-03, -2.4872e-03, -2.7847e-03, 1.1463e-03, 3.5095e-02, 3.9177e-03, -5.9242e-03, -7.3730e-02, 5.3291e-03, -1.0358e-01, -4.3762e-02, 1.7624e-02, 3.4424e-02,\n -1.3603e-02, -3.6011e-02, -3.2074e-02, 1.9867e-02, -1.1787e-02, 1.0864e-02, 2.4216e-02, 2.6672e-02, -1.1055e-02, -2.7756e-02, 2.5421e-02, 8.8806e-03, 2.2446e-02, 1.8387e-02, 2.6836e-03, 3.1281e-02, 1.2314e-02, 5.3740e-04, 9.2010e-03, -7.6599e-03, -1.2121e-03, 8.5068e-03, 2.6413e-02, 2.0370e-02,\n 1.4908e-02, 4.6921e-03, 9.4910e-03, 2.4399e-02, -2.5406e-02, -1.7075e-02, -9.1267e-04, -3.0575e-03, 6.2752e-03, 2.2186e-02, 1.3329e-02, -5.4932e-02, 1.6220e-02, -1.1635e-04, 1.3794e-02, -2.4681e-03, -4.0924e-02, 3.5591e-03, -2.1591e-02, 1.0590e-02, 4.2572e-02, -9.5978e-03, -1.2413e-02, -1.4175e-02,\n 1.8341e-02, 6.0730e-03, -3.7611e-05, 3.3569e-02, 1.3145e-02, -1.4305e-02, 2.4414e-02, 2.1378e-02, 4.2847e-02, 5.4138e-02, 1.7105e-02, 9.6970e-03, -3.6682e-02, -2.1469e-02, 5.8594e-02, 2.7359e-02, 6.7978e-03, 3.8483e-02, -2.9755e-02, -1.4008e-02, -9.9716e-03, -2.4155e-02, 2.4963e-02, -2.9648e-02,\n 4.8370e-03, 3.6049e-03, 4.6021e-02, 2.3499e-03, 1.1368e-02, 2.1515e-02, 6.7787e-03, 2.2552e-02, 2.5314e-02, 1.4420e-02, 2.2049e-02, 9.5596e-03, 3.7598e-02, -7.4844e-03, -2.5574e-02, 5.0316e-03, 1.5427e-02, -3.8605e-02, -1.1978e-02, -2.3102e-02, 2.8961e-02, -5.4657e-02, -3.2623e-02, -8.5526e-03,\n -1.6617e-02, 1.2276e-02, 3.3386e-02, 3.7384e-02, 1.9211e-02, -1.3924e-02, -2.4033e-02, -1.3916e-02, -1.3405e-02, 4.3144e-03, 5.4321e-03, -2.5818e-02, -1.8112e-02, -1.4544e-03, 1.6541e-02, 1.0361e-02, -1.4748e-02, 1.5511e-02, -1.1978e-02, 4.6173e-02, -2.4475e-02, 1.6861e-02, 3.9093e-02, 1.0841e-02,\n 2.4811e-02, -2.7145e-02, 8.6441e-03, 2.9236e-02, -4.0588e-02, 3.8879e-02, 3.7689e-02, -4.1382e-02, -1.8044e-03, -3.8757e-03, 1.9043e-02, 1.1871e-02, -1.4029e-03, 4.8866e-03, 4.1504e-03, 3.3936e-02, -5.0888e-03, 5.2399e-02, 1.8753e-02, 3.4668e-02, 3.9368e-02, -5.5176e-02, -2.8629e-03, 2.6550e-02,\n -1.9669e-02, -1.3290e-02, 9.1982e-04, -2.5040e-02, 4.2206e-02, 4.8187e-02, 5.4810e-02, -5.5790e-04, 1.2825e-02, 2.1622e-02, -2.5345e-02, 1.9302e-02, 1.1169e-02, 1.1452e-02, -3.8483e-02, 1.9318e-02, -9.1705e-03, 8.1711e-03, 2.0538e-02, 4.1138e-02, -3.8605e-02, -6.8207e-03, 5.0497e-04, -1.8936e-02,\n -1.7288e-02, 5.9357e-03, 3.4454e-02, 1.9287e-02, 1.3237e-02, 3.8361e-02, -1.4381e-03, -2.3499e-02, 3.3386e-02, 9.8953e-03, -1.3931e-02, -2.1378e-02, 4.0985e-02, -1.6556e-02, 7.1068e-03, 2.6733e-02, -1.2932e-02, 3.4027e-02, -4.1161e-03, 4.6425e-03, -6.5880e-03, 2.8183e-02, -1.7471e-02, 1.3611e-02,\n 1.2560e-03, 3.5400e-02, 6.1218e-02, -4.3549e-02, 6.8207e-03, -1.2001e-02, -3.0670e-02, -6.5247e-02, 5.3192e-02, -6.2294e-03, 4.8004e-02, 4.1077e-02, -2.1133e-02, 4.8566e-04, 2.5742e-02, 1.6220e-02, 4.8462e-02, -2.1515e-02, -2.7420e-02, -2.1378e-02, 2.0950e-02, -6.1584e-02, 3.5896e-03, 7.5226e-03,\n 2.0523e-02, -2.3270e-02, 2.5513e-02, 9.9411e-03, -2.4002e-02, -7.7019e-03, 2.1347e-02, 3.8055e-02, -5.3596e-03, 3.6865e-02, 3.7659e-02, -4.1748e-02, -3.0167e-02, -4.7058e-02, 1.7380e-02, 1.4168e-02, -3.2043e-02, 3.6682e-02, -4.0863e-02, 2.7054e-02, 2.2537e-02, 1.2482e-02, 2.0370e-02, -2.6169e-02,\n -1.1101e-02, -3.9787e-03, -7.1594e-02, 3.8147e-02, -3.0869e-02, 2.2293e-02, 4.7546e-02, -8.2642e-02, -1.0109e-02, -5.8556e-03, 1.4046e-02, -3.7598e-02, 1.5945e-02, -2.6901e-02, 7.1289e-02, -3.0579e-02, 2.2491e-02, -1.2840e-02, 2.3514e-02, 6.3354e-02, 3.8357e-03, 3.1464e-02, -2.8992e-03, 5.5359e-02,\n -1.7761e-02, -5.7259e-03, 1.9073e-02, -3.0994e-03, -2.6749e-02, -1.0178e-02, -3.0457e-02, 3.1067e-02, -1.8600e-02, -2.4052e-03, -5.5008e-03, 1.5249e-03, 6.6528e-03, 2.7496e-02, -1.1757e-02, 1.0155e-02, 6.5002e-03, -2.5223e-02, -2.0172e-02, -3.9459e-02, 4.4525e-02, 4.4922e-02, 7.2441e-03, 3.6469e-02,\n -1.5411e-02, 5.7068e-03, -1.7563e-02, 6.3782e-03, 4.7150e-02, 6.1798e-03, -3.9734e-02, 1.8251e-04, -9.7733e-03, 1.2550e-02, 3.4058e-02, -1.3176e-02, -2.1133e-02, -2.4277e-02, 1.7532e-02, -5.9853e-03, -1.2009e-02, -3.9246e-02, -9.3994e-03, -1.8600e-02, -4.0009e-02, -8.7662e-03, 3.5461e-02, 2.5436e-02,\n -5.9570e-02, 2.4857e-02, 4.4800e-02, -7.0751e-05, -1.3306e-02, -1.4915e-02, 7.1831e-03, 3.0960e-02, -2.2186e-02, -1.0506e-02, 4.7112e-03, 5.0995e-02, 4.2023e-02, 9.8572e-03, 2.6962e-02, 1.5732e-02, -1.0881e-03, -3.5629e-03, 1.1177e-02, -2.9335e-03, 1.5388e-02, 1.7792e-02, -2.4109e-02, -4.4212e-03,\n -7.4005e-03, -2.3178e-02, 2.8076e-02, -3.5706e-03, 3.7140e-02, 1.4664e-02, -4.0131e-02, -1.2390e-02, 2.8610e-02, 3.4332e-02, -1.4603e-02, -1.9241e-02, 2.1011e-02, 4.4975e-03, -3.9398e-02, -1.1864e-02, 5.8136e-03, -1.4305e-02, 1.9501e-02, 2.5925e-02, 3.4912e-02, -1.1206e-01, 1.5478e-03, 5.5939e-02,\n -9.2392e-03, 9.3842e-03, -4.8737e-02, 9.7885e-03, -9.1782e-03, -2.5692e-03, 2.4689e-02, -5.1003e-03, -2.5574e-02, -2.6718e-02, 2.3575e-03, -2.1759e-02, -1.1162e-02, 4.5837e-02, 4.4220e-02, 2.7191e-02, 2.4887e-02, -3.2379e-02, -2.4323e-02, 1.0574e-02, 3.2578e-03, 1.5137e-02, 2.7008e-03, 1.5793e-02,\n -1.1902e-02, 5.1025e-02, -4.6417e-02, -5.4810e-02, 2.8687e-02, -3.6201e-03, -2.2186e-02, -1.2970e-02, 2.8687e-02, -1.5213e-02, -1.1963e-02, 1.3168e-02, 1.5869e-02, -2.6031e-02, 3.2196e-02, 1.2794e-02, -1.5099e-02, -2.0020e-02, 2.6978e-02, 5.0049e-03, -6.6948e-03, 3.3894e-03, 9.3079e-03, 3.7384e-02,\n -3.6087e-03, -4.7546e-02, 6.0760e-02, -8.7662e-03, -9.2621e-03, 1.8253e-03, -1.9287e-02, -1.1765e-02, -2.4780e-02, -3.7041e-03, 1.0574e-02, 1.6037e-02, 9.5901e-03, 2.0966e-02, 3.4729e-02, 1.0170e-02, 6.4888e-03, 4.7379e-03, 1.7639e-02, -3.1342e-02, -5.7861e-02, 4.2633e-02, 5.6953e-03, 5.6000e-02,\n -3.5278e-02, -3.7872e-02, 1.3344e-02, -1.9806e-02, -2.8351e-02, 2.4548e-03, 3.8483e-02, 5.3741e-02, -7.0763e-03, -4.6806e-03, -2.9755e-02, 1.4786e-02, -4.5898e-02, -1.5747e-02, -1.5869e-02, -7.9651e-03, 1.0681e-02, 2.1805e-02, 2.1027e-02, 7.4043e-03, -1.7136e-02, 5.3558e-02, 4.5593e-02, 2.1896e-02,\n -8.8263e-04, 5.8105e-02, 1.0834e-02, 3.6469e-02, 1.8082e-02, -2.0218e-02, 4.2450e-02, -4.8828e-02, 3.5858e-02, 3.6133e-02, -9.8419e-03, 1.6113e-02, 3.9856e-02, -1.7120e-02, 1.5306e-03, -1.8951e-02, -6.3232e-02, 9.4681e-03, 1.2489e-02, 4.3678e-03, -6.7863e-03, -2.6779e-02, 3.6133e-02, 1.8951e-02,\n -4.6515e-04, 1.0712e-02, -1.1665e-02, -2.7130e-02, 4.1382e-02, -3.0884e-02, -2.9449e-02, -6.1493e-03, -2.6001e-02, 4.8920e-02, -8.5449e-03, 6.5193e-03, 1.2077e-02, -3.1700e-03, -1.1024e-02, -5.4131e-03, 7.0953e-03, -1.9760e-02, 2.3079e-03, 2.1347e-02, -6.3896e-03, 1.1932e-02, -6.4087e-03, 8.8135e-02,\n 1.7929e-02, 1.9638e-02, 2.9312e-02, 2.7878e-02, 1.9165e-02, 1.7929e-02, 1.3214e-02, 8.0017e-02, 1.6403e-02, -9.6664e-03, 2.5101e-02, 1.0231e-02, -9.0637e-03, 3.9856e-02, 6.1920e-02, -3.2291e-03, -2.8305e-02, -3.1281e-02, -5.0697e-03, 4.0588e-02, 3.0701e-02, -4.0131e-03, -3.7048e-02, -1.7426e-02,\n 2.3315e-02, 1.0300e-02, -3.2898e-02, -1.6060e-03, -3.7201e-02, -3.6285e-02, -7.4921e-03, -1.6861e-03, 2.0554e-02, -4.0436e-02, 2.9556e-02, 6.8115e-02, -3.4546e-02, -1.3757e-01, 2.2675e-02, 2.1393e-02, 3.5370e-02, -4.9255e-02, -5.2063e-02, 1.6281e-02, -4.5654e-02, -5.0011e-03, 4.6692e-02, 5.9652e-04],\n dtype=torch.float16), 'perceiver_resampler.layers.2.0.norm2.weight': tensor([0.5522, 0.6113, 0.6685, 0.5625, 0.5713, 0.6206, 0.6377, 0.6030, 0.5776, 0.5967, 0.6030, 0.6113, 0.6392, 0.5322, 0.5591, 0.5679, 0.6284, 0.5757, 0.6606, 0.6426, 0.5874, 0.6064, 0.5732, 0.5269, 0.5347, 0.6138, 0.6084, 0.6265, 0.6382, 0.6406, 0.6030, 0.5513, 0.6343, 0.6030, 0.5767, 0.6494, 0.5674, 0.6504, 0.6431,\n 0.6284, 0.6289, 0.5659, 0.6436, 0.6055, 0.5986, 0.5581, 0.5928, 0.5508, 0.6079, 0.5908, 0.6289, 0.5947, 0.5962, 0.6846, 0.5635, 0.6763, 0.6289, 0.6045, 0.6196, 0.5811, 0.5942, 0.6172, 0.5908, 0.6323, 0.6152, 0.5513, 0.6172, 0.6016, 0.6709, 0.6191, 0.5674, 0.6099, 0.6201, 0.6392, 0.6318, 0.6172, 0.6328, 0.5151,\n 0.5396, 0.6201, 0.6177, 0.5713, 0.5942, 0.5977, 0.6182, 0.6118, 0.6479, 0.6001, 0.5591, 0.5034, 0.6118, 0.5815, 0.6182, 0.6162, 0.5977, 0.6357, 0.5791, 0.6489, 0.6235, 0.6094, 0.6357, 0.5820, 0.6440, 0.5459, 0.6069, 0.6270, 0.5820, 0.6260, 0.5693, 0.6182, 0.5942, 0.5693, 0.6753, 0.5845, 0.6353, 0.6577, 0.5757,\n 0.6392, 0.6187, 0.6118, 0.5801, 0.6025, 0.6382, 0.6245, 0.6060, 0.6240, 0.6226, 0.6250, 0.5688, 0.6421, 0.6094, 0.5547, 0.6040, 0.6309, 0.5742, 0.6665, 0.6074, 0.5942, 0.6094, 0.6030, 0.6631, 0.6143, 0.6265, 0.5586, 0.6377, 0.6079, 0.6514, 0.5947, 0.6230, 0.6064, 0.5908, 0.5952, 0.6113, 0.5786, 0.5674, 0.5684,\n 0.6113, 0.6152, 0.5552, 0.6714, 0.6523, 0.5728, 0.6753, 0.5396, 0.6025, 0.6138, 0.6055, 0.5986, 0.6094, 0.6187, 0.6440, 0.6235, 0.5737, 0.6089, 0.5713, 0.5078, 0.6216, 0.5864, 0.6572, 0.5659, 0.5942, 0.6035, 0.5215, 0.5591, 0.6401, 0.5957, 0.5845, 0.5747, 0.6025, 0.6240, 0.5913, 0.6147, 0.6035, 0.6802, 0.6113,\n 0.6426, 0.6001, 0.6118, 0.6504, 0.6099, 0.5791, 0.6646, 0.6406, 0.6309, 0.6006, 0.6123, 0.6260, 0.5752, 0.6392, 0.6060, 0.5859, 0.6304, 0.6104, 0.6128, 0.5981, 0.6030, 0.5645, 0.5923, 0.6118, 0.5259, 0.6216, 0.6279, 0.6040, 0.5952, 0.5542, 0.5957, 0.5625, 0.6133, 0.5972, 0.6621, 0.6147, 0.5864, 0.5327, 0.5947,\n 0.6426, 0.5742, 0.6709, 0.5962, 0.6309, 0.5596, 0.6460, 0.5728, 0.6011, 0.5400, 0.6318, 0.5747, 0.6436, 0.6221, 0.6143, 0.5938, 0.6079, 0.5361, 0.6343, 0.6362, 0.5508, 0.6167, 0.6489, 0.5786, 0.6187, 0.5923, 0.6445, 0.6045, 0.6440, 0.6655, 0.6533, 0.5674, 0.6367, 0.6079, 0.6553, 0.5947, 0.6172, 0.5767, 0.5752,\n 0.6655, 0.5674, 0.6333, 0.6064, 0.5796, 0.6162, 0.6592, 0.6182, 0.4844, 0.5859, 0.5972, 0.6460, 0.6191, 0.5430, 0.6299, 0.5796, 0.6357, 0.5679, 0.5972, 0.6294, 0.5981, 0.6772, 0.5693, 0.6040, 0.6279, 0.6157, 0.5898, 0.6621, 0.5942, 0.5918, 0.5640, 0.6689, 0.5659, 0.6572, 0.6021, 0.6250, 0.5654, 0.6118, 0.6167,\n 0.5908, 0.6704, 0.5669, 0.5508, 0.6064, 0.6567, 0.5815, 0.5659, 0.6802, 0.5488, 0.5986, 0.5903, 0.6899, 0.5293, 0.6084, 0.5791, 0.6172, 0.6328, 0.6108, 0.5522, 0.5830, 0.6470, 0.6260, 0.5938, 0.6128, 0.5825, 0.5542, 0.6162, 0.5981, 0.5928, 0.6714, 0.6279, 0.5918, 0.5645, 0.6274, 0.6025, 0.6509, 0.5654, 0.6572,\n 0.5425, 0.5762, 0.6289, 0.5649, 0.6602, 0.6318, 0.5825, 0.5791, 0.5913, 0.6494, 0.6230, 0.5781, 0.6016, 0.5967, 0.5977, 0.5908, 0.6265, 0.6201, 0.6201, 0.6362, 0.6079, 0.6685, 0.5996, 0.6221, 0.6113, 0.6050, 0.6289, 0.5737, 0.6025, 0.6025, 0.6089, 0.5845, 0.6357, 0.6025, 0.6064, 0.5962, 0.5776, 0.5781, 0.5762,\n 0.5952, 0.5601, 0.6528, 0.6270, 0.5771, 0.6484, 0.5820, 0.5962, 0.6069, 0.6265, 0.6191, 0.5317, 0.5894, 0.6323, 0.6074, 0.6118, 0.6001, 0.6426, 0.6123, 0.5718, 0.6279, 0.5767, 0.5957, 0.5327, 0.5298, 0.6455, 0.5488, 0.6250, 0.6289, 0.6177, 0.6694, 0.6060, 0.6211, 0.5532, 0.6919, 0.5601, 0.5723, 0.6475, 0.6030,\n 0.6362, 0.5986, 0.5913, 0.5601, 0.5854, 0.6582, 0.6138, 0.5991, 0.5933, 0.6562, 0.6582, 0.5181, 0.5952, 0.5474, 0.5884, 0.5781, 0.6230, 0.5786, 0.6602, 0.5830, 0.6050, 0.6060, 0.5659, 0.5615, 0.6519, 0.5767, 0.5132, 0.6460, 0.5728, 0.6084, 0.5654, 0.6440, 0.6284, 0.5801, 0.5688, 0.6069, 0.6060, 0.5586, 0.6523,\n 0.6182, 0.6382, 0.6196, 0.6011, 0.6479, 0.6323, 0.6040, 0.5811, 0.6006, 0.6470, 0.5791, 0.6548, 0.6030, 0.6270, 0.6665, 0.6411, 0.5645, 0.5801, 0.5669, 0.5498, 0.5767, 0.5840, 0.5752, 0.6343, 0.5674, 0.5679, 0.5547, 0.5664, 0.6763, 0.6333, 0.5981, 0.5757, 0.5522, 0.5996, 0.5835, 0.6104, 0.6167, 0.5659, 0.5576,\n 0.6479, 0.6802, 0.6431, 0.6289, 0.6465, 0.5776, 0.6187, 0.5400, 0.6406, 0.6152, 0.5713, 0.6211, 0.6191, 0.5688, 0.6719, 0.6250, 0.6060, 0.6191, 0.6367, 0.5142, 0.5562, 0.6211, 0.6040, 0.6323, 0.5269, 0.6255, 0.6108, 0.5967, 0.6147, 0.6250, 0.5781, 0.6167, 0.5518, 0.6162, 0.6123, 0.6548, 0.6074, 0.5239, 0.6519,\n 0.6050, 0.6011, 0.6436, 0.6221, 0.6162, 0.5967, 0.6211, 0.6646, 0.5825, 0.6167, 0.6147, 0.5879, 0.5752, 0.6367, 0.5542, 0.5811, 0.6479, 0.6250, 0.6191, 0.6328, 0.5894, 0.6211, 0.6318, 0.6060, 0.5972, 0.5508, 0.6216, 0.6216, 0.5938, 0.5400, 0.6211, 0.5278, 0.6240, 0.6846, 0.5771, 0.6289, 0.6040, 0.6401, 0.5942,\n 0.5454, 0.6309, 0.6152, 0.6157, 0.6445, 0.5918, 0.5967, 0.5444, 0.5747, 0.6187, 0.6099, 0.5781, 0.6294, 0.5825, 0.6641, 0.6289, 0.6372, 0.6089, 0.5601, 0.6445, 0.5991, 0.5913, 0.5879, 0.5957, 0.6626, 0.5820, 0.6509, 0.6030, 0.5732, 0.5703, 0.6021, 0.5435, 0.5791, 0.6172, 0.5996, 0.6504, 0.5566, 0.5884, 0.5869,\n 0.5757, 0.6348, 0.5630, 0.6230, 0.5742, 0.6606, 0.6157, 0.6289, 0.6030, 0.5962, 0.6772, 0.6050, 0.6025, 0.6113, 0.5312, 0.5981, 0.6440, 0.6074, 0.6841, 0.6201, 0.6499, 0.5830, 0.6543, 0.6348, 0.6616, 0.6777, 0.5317, 0.6079, 0.5693, 0.6167, 0.6660, 0.6240, 0.6372, 0.6021, 0.6416, 0.5981, 0.6240, 0.6289, 0.6367,\n 0.6465, 0.6245, 0.6577, 0.5854, 0.6416, 0.6187, 0.5962, 0.6338, 0.5718, 0.6055, 0.5923, 0.6396, 0.6138, 0.5933, 0.5850, 0.6426, 0.6118, 0.5986, 0.6694, 0.6284, 0.5815, 0.5366, 0.5791, 0.6177, 0.6172, 0.6050, 0.6616, 0.5146, 0.6201, 0.6162, 0.5674, 0.5840, 0.6338, 0.6182, 0.6294, 0.6143, 0.5874, 0.6025, 0.6099,\n 0.6128, 0.5962, 0.6270, 0.6401, 0.6191, 0.5903, 0.5576, 0.6025, 0.6421, 0.6401, 0.5586, 0.6152, 0.6089, 0.6265, 0.6050, 0.6172, 0.5508, 0.5332, 0.6338, 0.6577, 0.5859, 0.5835, 0.6060, 0.6104, 0.5659, 0.6489, 0.5747, 0.6187, 0.6323, 0.5898, 0.5796, 0.6279, 0.5913, 0.6650, 0.6333, 0.5967, 0.6016, 0.5884, 0.6245,\n 0.5630, 0.6069, 0.6699, 0.6104, 0.5693, 0.6191, 0.6040, 0.5957, 0.6089, 0.6118, 0.6045, 0.6040, 0.6265, 0.5845, 0.6182, 0.5371, 0.5996, 0.6201, 0.6211, 0.6216, 0.6138, 0.5845, 0.6294, 0.6211, 0.5967, 0.5693, 0.6152], dtype=torch.float16), 'perceiver_resampler.layers.2.0.norm2.bias': tensor([-4.0009e-02, -1.4091e-02, -1.5465e-02, -1.0492e-01, -4.6173e-02, 2.5192e-02, -2.0096e-02, 2.9190e-02, 4.5197e-02, -5.4962e-02, 4.9927e-02, -9.3567e-02, -1.2611e-02, -1.1743e-01, 9.5032e-02, 2.4185e-02, -9.3918e-03, 3.6041e-02, -2.3026e-02, -4.6265e-02, -8.4473e-02, -5.4054e-03, 3.7022e-03, 9.9548e-02,\n -8.8623e-02, -1.6632e-02, 6.1310e-02, -4.4220e-02, 2.3163e-02, 2.8152e-02, 4.8981e-02, 1.0291e-01, -4.3304e-02, 5.7739e-02, 8.1665e-02, 1.2901e-02, 5.5939e-02, 3.3966e-02, 4.9362e-03, -1.4450e-02, 6.5575e-03, 8.2886e-02, -3.3356e-02, -6.8703e-03, -1.2238e-02, -3.4576e-02, -3.2997e-03, 8.2886e-02,\n 6.7101e-03, 5.2071e-03, 1.3702e-02, 4.2847e-02, 3.2898e-02, 3.3783e-02, 5.0293e-02, -1.3885e-02, 8.6517e-03, 6.3965e-02, -5.1483e-02, 9.0637e-02, -1.2311e-01, 2.1439e-02, 2.2736e-02, -2.2598e-02, -2.7664e-02, 4.1931e-02, 6.1218e-02, 1.1847e-01, 1.4191e-02, 5.4291e-02, -6.6895e-02, -4.6509e-02,\n 1.0033e-02, -7.5073e-02, -3.8361e-02, -6.3286e-03, -5.5664e-02, -1.3501e-01, 1.6394e-01, 1.9226e-02, 3.0060e-02, 1.9093e-03, 4.4739e-02, -2.7161e-02, 1.0696e-02, 9.6863e-02, -3.1433e-03, -8.4534e-02, -1.0907e-01, -1.8005e-01, 8.7891e-02, -6.8054e-02, -1.2093e-02, -3.2684e-02, -2.6337e-02, -6.0181e-02,\n -5.8411e-02, -6.1989e-04, -2.6047e-02, 5.0812e-02, 6.3416e-02, -6.5918e-03, 3.6346e-02, -1.4612e-01, 4.7546e-02, -1.7471e-02, 1.4229e-02, -4.2572e-02, 9.3140e-02, 4.6051e-02, -6.8359e-02, 3.5065e-02, -4.1077e-02, -1.0913e-01, 4.1321e-02, 2.6688e-02, 9.3018e-02, -5.7281e-02, -1.9516e-02, -2.9358e-02,\n 5.3215e-03, -5.4688e-02, 1.3458e-02, 4.2305e-03, -3.7567e-02, 5.6488e-02, -2.0874e-02, -5.3284e-02, 6.4453e-02, -5.2460e-02, -1.6617e-02, 1.3428e-01, -9.0179e-03, 3.4729e-02, -4.9561e-02, 9.8705e-05, -8.1238e-02, 9.1309e-02, -6.3110e-02, 6.1920e-02, -2.6302e-03, 5.1727e-03, 2.7832e-02, 1.8762e-01,\n -5.7861e-02, -7.8552e-02, 8.0948e-03, -2.7695e-02, 4.9133e-02, 1.3817e-02, 4.7913e-03, 5.5695e-02, -4.8981e-02, -7.0862e-02, -3.7079e-02, -9.1980e-02, 8.7769e-02, 1.6083e-02, -2.6062e-02, -4.8676e-02, 3.6194e-02, -6.2805e-02, -3.2745e-02, -5.6885e-02, 5.1193e-03, -7.2815e-02, -2.5375e-02, -4.0466e-02,\n 1.2909e-02, 2.1057e-02, 1.1627e-02, 3.3936e-02, -6.8855e-04, -7.3576e-04, -8.0750e-02, -1.3745e-01, -3.1769e-02, 7.6843e-02, 2.7618e-02, -8.2275e-02, 7.0068e-02, 6.0913e-02, 1.2476e-01, 4.7821e-02, 5.1331e-02, -3.7018e-02, -5.4741e-03, -7.9712e-02, -1.4091e-02, 3.3478e-02, -3.6896e-02, 6.5674e-02,\n 7.8613e-02, -1.4153e-02, -5.0323e-02, -1.4572e-02, 1.2537e-01, -1.6083e-02, 4.5380e-02, -6.0333e-02, 1.0809e-01, 1.5778e-02, -2.4826e-02, 4.3335e-02, 9.4543e-02, 1.4374e-02, -4.3427e-02, -2.1790e-02, 7.3730e-02, 2.7115e-02, 5.3192e-02, -3.9917e-02, -5.6824e-02, -3.8483e-02, 1.2311e-01, 4.3579e-02,\n -9.5749e-03, -6.5979e-02, 6.9656e-03, 1.2610e-01, 3.5801e-03, -4.9530e-02, -3.5492e-02, 1.8127e-02, -4.1077e-02, 5.4779e-02, 1.0040e-01, 5.8868e-02, -4.6295e-02, -1.6994e-03, -4.1229e-02, 9.9121e-02, 7.7393e-02, 5.4504e-02, -2.3560e-02, 5.2795e-02, -1.6769e-02, -8.2581e-02, 7.8430e-02, 4.4281e-02,\n -3.0243e-02, 7.2327e-02, 5.1270e-02, 1.3354e-01, -5.6915e-02, 1.6769e-02, 2.3712e-02, 7.3486e-02, -3.5858e-02, 1.0126e-01, 5.3650e-02, 8.4412e-02, 7.0496e-02, 4.7791e-02, 9.5642e-02, -2.5604e-02, -1.7059e-02, 8.9417e-02, 8.2397e-02, 1.4374e-02, -3.9948e-02, 1.5881e-01, 5.9967e-03, -4.0100e-02,\n -2.1317e-02, -5.3986e-02, 2.9907e-02, 2.8076e-02, -2.8519e-02, 2.7008e-02, 4.2389e-02, 1.0779e-01, -7.2083e-02, -3.1372e-02, 1.1206e-01, 2.7599e-03, 7.5684e-02, 2.3529e-02, 5.1819e-02, -1.0315e-02, 2.7847e-02, -1.4819e-01, 9.4604e-02, -8.4656e-02, -4.5807e-02, 1.1971e-02, -7.5012e-02, 1.3474e-02,\n 5.0934e-02, 1.6037e-02, -5.5756e-02, 4.2053e-02, -3.9429e-02, -5.0018e-02, -2.7176e-02, 5.8350e-02, -7.8308e-02, 5.6946e-02, -2.7985e-02, -5.8708e-03, 3.6041e-02, -1.9363e-02, -2.5009e-02, -1.8005e-01, 2.1286e-02, -4.5990e-02, -4.1885e-03, 9.5703e-02, 3.5114e-03, -9.4238e-02, 3.4760e-02, -8.7585e-02,\n -2.3331e-02, -3.4485e-02, 7.8735e-02, 1.6980e-01, -2.4277e-02, -5.1117e-02, -1.4050e-01, 9.4299e-02, 2.3270e-02, -8.3679e-02, -5.0934e-02, -8.1848e-02, -8.4961e-02, -2.7649e-02, 5.4779e-02, -4.9530e-02, -2.5711e-02, 4.2938e-02, 1.2469e-01, -1.2384e-01, 9.8450e-02, 4.2908e-02, -2.8515e-03, -2.2369e-02,\n -1.9897e-02, 4.0817e-03, 7.9773e-02, -3.8452e-02, 5.5267e-02, -4.9377e-02, -5.2338e-02, 3.2959e-02, 5.3741e-02, 8.4961e-02, -2.7299e-04, 2.9922e-02, 1.0080e-03, 2.3376e-02, -8.3084e-03, -1.7468e-01, -6.1890e-02, -4.4823e-03, 8.9294e-02, -1.2939e-02, 4.0253e-02, -4.5837e-02, 7.0618e-02, -3.0884e-02,\n -2.3758e-02, 2.7908e-02, 1.5991e-02, 6.4148e-02, -2.4689e-02, 6.4148e-02, 5.3009e-02, 2.3972e-02, 1.0658e-02, -4.6997e-02, -5.4779e-02, 3.9703e-02, -4.2206e-02, -5.6305e-02, -3.8727e-02, 6.0059e-02, -9.3079e-02, -3.9246e-02, -2.2217e-02, -1.8082e-02, -5.0262e-02, -2.3560e-02, -7.6408e-03, 3.6041e-02,\n 9.8133e-04, 1.6190e-02, -3.9246e-02, 4.3549e-02, 6.1432e-02, -3.4363e-02, 1.5774e-03, -1.0040e-01, -8.9233e-02, 7.5264e-03, 1.0699e-01, 1.7578e-02, -5.6427e-02, 3.7109e-02, 2.1591e-03, 3.7506e-02, 4.7943e-02, 2.6413e-02, 5.1392e-02, 8.3069e-02, -1.4984e-02, 1.2856e-02, -2.8534e-02, 2.2217e-02,\n 2.5330e-02, -6.4880e-02, -4.0802e-02, -6.3477e-02, 1.1345e-02, -7.5317e-02, -1.4441e-01, -5.4810e-02, -1.5698e-01, 1.1902e-01, 4.2084e-02, 3.2562e-02, 1.2680e-02, -1.9363e-02, -2.7054e-02, -6.9153e-02, 1.5701e-02, 1.2219e-01, -3.1891e-02, 3.1891e-02, -6.2073e-02, 2.8885e-02, -1.1432e-01, -1.3390e-03,\n 5.4321e-02, -7.8613e-02, -4.8714e-03, 9.2041e-02, -1.5671e-02, -7.1899e-02, 6.3354e-02, 1.6861e-02, 1.5344e-01, -5.2429e-02, -3.3054e-03, 9.1309e-02, 2.0706e-02, -8.8120e-03, -8.8806e-02, -1.5808e-02, 7.8583e-03, -4.9835e-02, 7.0610e-03, -5.3284e-02, 4.2480e-02, 8.7662e-03, 4.4342e-02, 1.5674e-01,\n -6.5384e-03, 1.2497e-02, -8.9798e-03, 9.3689e-02, -3.0746e-02, -3.9612e-02, 3.1769e-02, 3.1799e-02, -3.1204e-02, -5.6488e-02, 3.2623e-02, -1.8890e-02, -5.3589e-02, 1.8295e-02, -5.7739e-02, -1.0730e-01, -5.9052e-02, -7.2144e-02, -4.7150e-02, -1.9745e-02, 8.4961e-02, 7.7393e-02, 5.2490e-02, -4.1962e-02,\n 2.7740e-02, 1.7975e-02, -6.2012e-02, 2.9068e-03, 3.2135e-02, -7.0068e-02, -3.5278e-02, 3.9482e-03, 2.3712e-02, -2.3712e-02, 1.8585e-02, 6.2866e-02, 3.0930e-02, -7.2205e-02, -5.4779e-02, 7.0923e-02, 4.4617e-02, 5.8228e-02, 1.0901e-01, 8.3008e-02, -8.4045e-02, 4.8187e-02, -1.0254e-01, 8.7708e-02,\n -3.8788e-02, 1.0352e-01, -6.7322e-02, -1.3031e-02, 9.4757e-03, 7.2876e-02, -6.0669e-02, 6.0883e-02, 2.5444e-03, 5.1331e-02, 9.2773e-02, -1.4824e-02, 3.7445e-02, 4.5776e-02, -2.6245e-02, -6.6406e-02, 8.1482e-02, 1.7380e-02, 1.4526e-01, -8.6121e-02, -1.1223e-02, -2.9266e-02, -1.8091e-01, -5.9082e-02,\n -4.5837e-02, -4.1931e-02, -5.1086e-02, -1.8616e-01, 5.5603e-02, 3.0319e-02, -1.7197e-02, -1.0788e-02, -2.5726e-02, 7.9422e-03, -5.0354e-03, 9.2102e-02, 5.4199e-02, -1.9379e-02, 3.0746e-02, 2.6443e-02, 6.5125e-02, 7.9346e-02, 1.8158e-02, 3.8879e-02, 5.0934e-02, 7.4524e-02, -5.8823e-03, -4.7638e-02,\n 1.5998e-04, 4.5276e-04, -1.3878e-02, -1.2476e-01, 7.2571e-02, -2.8397e-02, 6.8848e-02, 3.0396e-02, 2.4124e-02, -1.3965e-01, -1.0544e-02, -1.4900e-02, -2.6978e-02, -2.6489e-02, 6.1768e-02, 2.7313e-03, 1.3222e-02, 7.4036e-02, -1.5869e-02, -4.8950e-02, -4.2633e-02, 7.1045e-02, -1.1578e-01, -1.6772e-01,\n 7.4219e-02, 7.4158e-02, 1.8723e-02, 1.3561e-03, -3.6407e-02, -2.8839e-02, 3.3844e-02, 6.3820e-03, -5.6854e-02, -4.1779e-02, 9.7961e-03, -1.8951e-02, 1.0358e-01, -3.9825e-02, 5.6152e-02, -7.0251e-02, 5.7617e-02, -6.8787e-02, 3.4424e-02, -3.0701e-02, -4.7729e-02, -1.5259e-02, 5.5939e-02, -3.3600e-02,\n 2.1698e-02, -2.0676e-02, -3.9581e-02, 1.1713e-01, 6.9031e-02, 8.3130e-02, -7.3242e-02, 1.4572e-02, 6.2141e-03, 3.9276e-02, 3.7994e-02, -4.4861e-02, 2.3972e-02, -8.0795e-03, -2.2522e-02, -1.5198e-02, 1.2589e-02, 6.5002e-02, -4.5959e-02, -1.6693e-02, -2.7725e-02, -1.3618e-02, 8.1787e-02, 6.5660e-04,\n 7.2632e-02, -5.5237e-02, -3.1891e-02, -6.6040e-02, 5.3314e-02, -8.5022e-02, 9.4681e-03, -2.1759e-02, 4.6234e-02, -5.7526e-02, 2.2614e-02, 4.4098e-03, -8.9905e-02, 7.7454e-02, -1.3843e-01, -5.2948e-02, -1.5511e-02, -2.2308e-02, -7.7286e-03, -1.5587e-02, -5.8044e-02, 2.3071e-02, 1.2604e-02, 3.4363e-02,\n 6.4850e-03, -4.2786e-02, 8.8135e-02, 6.2370e-03, 5.6915e-02, -7.4730e-03, 1.1147e-02, 3.0716e-02, -1.5732e-02, -3.7537e-02, 3.8177e-02, -8.8928e-02, 3.4760e-02, 2.4445e-02, 2.4063e-02, -6.9336e-02, -8.1726e-02, -2.8809e-02, -5.0903e-02, 3.7613e-03, 3.2257e-02, 7.2388e-02, 4.9774e-02, 4.8492e-02,\n -2.1896e-03, 4.3854e-02, -3.6560e-02, -6.8542e-02, 9.4666e-02, 8.2703e-02, -5.6152e-02, 3.6530e-02, -9.8114e-03, -4.6997e-03, -1.7960e-02, -7.7881e-02, 6.0547e-02, -2.9007e-02, -1.3025e-01, 4.1901e-02, -6.7444e-02, -4.4312e-02, 8.3008e-03, 1.3138e-02, 2.6825e-02, 6.8298e-02, 2.3193e-02, 5.3741e-02,\n 1.7929e-02, 3.2623e-02, -1.1615e-01, -6.4819e-02, 8.7463e-02, -6.0303e-02, 2.4673e-02, -4.2145e-02, 9.2102e-02, 2.9205e-02, -6.1981e-02, -6.1798e-02, 2.9602e-02, -1.1047e-02, -2.0889e-02, -5.0385e-02, 4.0894e-02, -2.1835e-02, -4.3976e-02, -1.3374e-02, -5.8777e-02, 4.0466e-02, -1.4429e-01, 1.8774e-01,\n -1.9516e-02, 1.0910e-02, -1.7853e-02, -3.3325e-02, 8.0505e-02, 4.6539e-02, -9.1980e-02, 7.8796e-02, -6.9092e-02, 8.0750e-02, 6.2103e-02, -7.8796e-02, 1.1877e-01, 1.3885e-02, 2.3315e-02, 1.8234e-02, 1.1086e-02, -1.2581e-02, -4.3732e-02, -8.5754e-02, 3.1143e-02, 1.5039e-01, 1.1772e-02, 5.8533e-02,\n 7.2266e-02, 4.5471e-02, -2.3819e-02, 6.7017e-02, -7.1106e-02, -1.5459e-03, 5.6091e-02, -4.5593e-02, 1.1017e-01, 3.4271e-02, -3.2928e-02, -3.9459e-02, 1.1926e-01, -1.0052e-01, 4.5319e-02, 4.9042e-02, 8.0322e-02, 1.7288e-02, -1.9119e-02, -5.5084e-02, -2.8000e-02, 3.9551e-02, 1.5991e-01, -5.4382e-02],\n dtype=torch.float16), 'perceiver_resampler.layers.2.0.to_q.weight': tensor([[ 0.0217, 0.0137, -0.0125, ..., 0.0424, -0.0128, 0.0031],\n [ 0.0041, 0.0040, 0.0129, ..., -0.0088, -0.0063, 0.0078],\n [-0.0030, -0.0035, -0.0001, ..., -0.0232, 0.0200, -0.0351],\n ...,\n [-0.0147, -0.0140, -0.0330, ..., 0.0146, 0.0033, 0.0205],\n [-0.0128, 0.0254, 0.0010, ..., -0.0455, -0.0190, 0.0170],\n [ 0.0066, 0.0154, 0.0076, ..., -0.0394, 0.0031, 0.0238]], dtype=torch.float16), 'perceiver_resampler.layers.2.0.to_kv.weight': tensor([[ 0.0113, -0.0146, -0.0035, ..., 0.0418, -0.0186, -0.0181],\n [ 0.0133, -0.0096, -0.0021, ..., -0.0115, 0.0052, -0.0010],\n [ 0.0184, -0.0028, -0.0626, ..., -0.0403, 0.0356, 0.0384],\n ...,\n [-0.0335, -0.0046, 0.0169, ..., -0.0597, 0.0431, 0.0125],\n [-0.0151, 0.0163, 0.0404, ..., -0.0164, -0.0241, -0.0293],\n [-0.0302, 0.0595, -0.0186, ..., -0.0146, -0.0128, 0.0302]], dtype=torch.float16), 'perceiver_resampler.layers.2.0.to_out.weight': tensor([[ 0.0096, -0.0101, -0.0124, ..., 0.0016, 0.0067, -0.0339],\n [ 0.0197, 0.0172, -0.0244, ..., 0.0006, 0.0232, -0.0168],\n [-0.0040, -0.0419, -0.0266, ..., -0.0014, 0.0219, 0.0247],\n ...,\n [-0.0225, 0.0254, -0.0050, ..., 0.0626, -0.0133, 0.0148],\n [-0.0476, -0.0088, 0.0788, ..., 0.0168, -0.0053, -0.0501],\n [-0.0184, 0.0003, 0.0027, ..., 0.0190, -0.0260, -0.0013]], dtype=torch.float16), 'perceiver_resampler.layers.2.1.0.weight': tensor([0.7998, 0.8193, 0.8350, 0.8613, 0.7798, 0.8765, 0.8604, 0.8110, 0.8203, 0.8213, 0.9028, 0.9116, 0.8828, 0.7832, 0.8530, 0.7871, 0.8657, 0.7676, 0.7935, 0.8423, 0.8418, 0.7886, 0.7734, 0.8169, 0.8340, 0.9189, 0.7710, 0.8989, 0.8228, 0.8696, 0.7720, 0.7402, 0.8872, 0.9268, 0.8013, 0.8135, 0.7573, 0.8325, 0.8599,\n 0.8574, 0.8271, 0.7280, 0.8643, 0.8032, 0.8086, 0.7759, 0.8286, 0.9141, 0.8066, 0.8735, 0.8999, 0.8184, 0.8320, 0.8354, 0.7148, 0.8569, 0.8408, 0.8491, 0.8384, 0.8164, 0.8252, 0.8950, 0.8867, 0.8066, 0.8760, 0.8159, 0.7861, 0.8794, 0.8423, 0.9092, 0.7661, 0.8579, 0.8638, 0.8564, 0.8257, 0.7817, 0.8467, 0.8843,\n 0.8813, 0.8535, 0.8589, 0.7456, 0.8428, 0.7842, 0.8315, 0.8843, 0.7852, 0.8843, 0.9082, 0.8672, 0.9360, 0.7632, 0.8267, 0.8750, 0.8433, 0.9072, 0.7349, 0.8545, 0.8052, 0.7676, 0.8398, 0.7422, 0.8428, 0.9028, 0.7832, 0.8604, 0.8486, 0.9053, 0.9014, 0.8311, 0.7808, 0.8198, 0.8066, 0.7520, 0.8955, 0.8662, 0.8369,\n 0.8027, 0.9019, 0.8530, 0.8037, 0.8750, 0.8828, 0.8403, 0.7993, 0.8452, 0.7617, 0.8359, 0.8022, 0.9004, 0.8472, 0.8208, 0.8188, 0.8853, 0.8657, 0.7998, 0.8989, 0.8223, 0.7563, 0.8491, 0.8525, 0.7910, 0.8320, 0.7808, 0.9131, 0.8730, 0.7739, 0.7915, 0.8135, 0.8184, 0.8237, 0.8438, 0.9404, 0.8242, 0.8062, 0.8760,\n 0.8628, 0.8447, 0.7646, 0.8975, 0.8750, 0.7817, 0.8091, 0.7612, 0.8516, 0.8345, 0.8594, 0.8130, 0.8398, 0.8389, 0.8506, 0.8267, 0.8467, 0.8516, 0.8252, 0.9023, 0.7788, 0.8091, 0.8550, 0.9502, 0.7939, 0.8794, 0.8042, 0.8291, 0.7559, 0.9097, 0.7900, 0.7559, 0.8066, 0.8823, 0.8735, 0.8486, 0.8936, 0.8862, 0.8184,\n 0.8745, 0.8618, 0.8345, 0.8960, 0.8042, 0.7202, 0.7969, 0.8052, 0.8784, 0.8394, 0.8218, 0.9175, 0.7554, 0.8330, 0.8198, 0.8486, 0.7412, 0.8276, 0.8877, 0.7007, 0.8970, 0.8188, 0.7695, 0.8345, 0.8906, 0.7832, 0.8735, 0.9185, 0.8193, 0.8008, 0.8433, 0.9375, 0.8584, 0.7969, 0.7876, 0.8486, 0.8950, 0.8174, 0.8657,\n 0.8330, 0.7783, 0.7964, 0.8892, 0.9453, 0.7725, 0.7432, 0.8101, 0.7998, 0.7822, 0.8706, 0.8916, 0.8535, 0.7910, 0.8560, 0.9058, 0.7686, 0.7666, 0.8584, 0.8711, 0.8882, 0.8442, 0.8726, 0.8062, 0.8408, 0.8281, 0.9272, 0.8599, 0.8745, 0.8330, 0.8506, 0.6899, 0.8682, 0.7744, 0.7876, 0.8477, 0.8540, 0.8638, 0.8198,\n 0.8560, 0.9233, 0.8457, 0.9282, 0.8027, 0.8735, 0.8926, 0.8633, 0.8018, 0.9277, 0.9131, 0.8428, 0.8086, 0.6968, 0.8647, 0.8667, 0.8926, 0.8511, 0.8730, 0.8506, 0.8540, 0.7915, 0.9106, 0.8008, 0.8350, 0.8164, 0.8125, 0.8711, 0.7954, 0.8120, 0.7612, 0.8755, 0.7070, 0.7998, 0.8315, 0.8130, 0.8110, 0.8521, 0.7554,\n 0.8105, 0.7998, 0.9185, 0.8394, 0.8438, 0.8223, 0.8447, 0.8828, 0.9048, 0.8296, 0.8037, 0.8906, 0.9722, 0.7661, 0.8491, 0.8594, 0.8384, 0.8403, 0.9814, 0.8071, 0.8940, 0.8979, 0.8071, 0.8013, 0.8521, 0.8696, 0.8652, 0.8677, 0.8950, 0.8638, 0.8984, 0.8169, 0.8970, 0.8853, 0.8062, 0.7979, 0.8867, 0.7158, 0.8521,\n 0.7856, 0.9165, 0.8511, 0.8604, 0.9023, 0.8164, 0.8198, 0.8472, 0.8569, 0.8516, 0.8110, 0.7759, 0.8643, 0.8750, 0.7700, 0.8564, 0.8320, 0.8208, 0.8511, 0.8384, 0.8804, 0.8174, 0.8804, 0.7949, 0.8433, 0.8726, 0.9058, 0.7598, 0.9062, 0.8623, 0.8276, 0.8657, 0.8784, 0.8569, 0.9165, 0.8574, 0.8022, 0.8535, 0.8179,\n 0.8525, 0.8530, 0.8579, 0.8511, 0.8789, 0.8896, 0.8105, 0.8418, 0.7539, 0.8403, 0.8848, 0.8022, 0.8032, 0.8901, 0.8506, 0.8394, 0.8213, 0.8389, 0.8696, 0.8662, 0.8530, 0.7993, 0.8638, 0.7275, 0.7510, 0.8232, 0.8721, 0.8896, 0.8647, 0.8691, 0.9263, 0.8911, 0.8052, 0.8110, 0.7681, 0.8892, 0.7368, 0.9141, 0.8467,\n 0.8477, 0.7368, 0.8433, 0.7803, 0.8403, 0.8315, 0.8481, 0.7856, 0.8882, 0.7222, 0.8105, 0.9214, 0.8145, 0.7183, 0.8516, 0.8418, 0.8652, 0.8643, 0.8423, 0.7983, 0.8154, 0.8691, 0.8828, 0.8096, 0.7812, 0.8384, 0.8628, 0.8179, 0.7544, 0.8066, 0.8135, 0.8706, 0.8354, 0.7822, 0.8130, 0.7998, 0.7422, 0.7969, 0.8467,\n 0.8833, 0.8188, 0.8530, 0.7808, 0.8584, 0.8291, 0.7632, 0.7856, 0.8364, 0.8989, 0.8486, 0.7300, 0.8257, 0.8091, 0.8481, 0.8193, 0.7671, 0.7212, 0.8037, 0.8379, 0.7920, 0.8418, 0.8867, 0.8394, 0.7397, 0.7832, 0.7861, 0.8843, 0.9243, 0.7925, 0.9189, 0.7764, 0.7710, 0.8403, 0.8452, 0.7441, 0.7593, 0.8726, 0.8335,\n 0.8555, 0.8379, 0.8276, 0.8750, 0.8589, 0.7656, 0.8276, 0.8848, 0.8726, 0.9043, 0.7993, 0.8306, 0.8164, 0.8335, 0.8730, 0.7129, 0.8628, 0.6641, 0.8296, 0.9272, 0.7822, 0.8149, 0.9082, 0.8647, 0.7495, 0.8545, 0.8228, 0.7856, 0.8296, 0.8281, 0.7705, 0.8545, 0.8818, 0.8853, 0.8374, 0.8457, 0.8364, 0.7314, 0.8730,\n 0.8262, 0.8315, 0.9155, 0.9404, 0.8320, 0.8682, 0.8989, 0.8438, 0.7627, 0.9263, 0.9038, 0.8926, 0.8486, 0.8237, 0.8096, 1.0010, 0.8218, 0.8018, 0.8447, 0.8691, 0.7070, 0.8252, 0.8198, 0.9038, 0.8584, 0.8789, 0.8530, 0.8550, 0.8545, 0.7886, 0.9014, 0.8306, 0.8335, 0.8164, 0.9048, 0.8984, 0.8032, 0.8789, 0.8154,\n 0.6665, 0.8237, 0.8164, 0.8809, 0.8589, 0.8555, 0.8896, 0.7817, 0.8481, 0.8384, 0.7456, 0.8813, 0.9048, 0.8022, 0.8906, 0.8169, 0.8169, 0.8281, 0.7974, 0.8164, 0.8359, 0.8154, 0.7793, 0.7979, 0.7954, 0.8535, 0.8442, 0.8369, 0.7852, 0.8667, 0.8105, 0.8345, 0.8428, 0.8140, 0.8359, 0.8896, 0.8032, 0.8799, 0.8159,\n 0.9038, 0.8550, 0.8833, 0.8896, 0.7725, 0.8774, 0.8403, 0.8369, 0.8594, 0.8228, 0.8237, 0.7837, 0.8911, 0.9077, 0.7393, 0.7490, 0.8428, 0.8921, 0.8335, 0.8813, 0.9434, 0.8042, 0.8652, 0.8701, 0.8667, 0.8999, 0.6943, 0.8311, 0.8101, 0.8159, 0.8799, 0.9058, 0.8481, 0.8677, 0.8472, 0.8838, 0.9092, 0.8545, 0.8379,\n 0.8730, 0.8457, 0.9136, 0.7803, 0.7197, 0.8013, 0.7915, 0.8799, 0.8145, 0.7485, 0.7891, 0.9180, 0.8921, 0.7817, 0.7993, 0.7910, 0.8584, 0.8423, 0.7983, 0.8188, 0.8945, 0.7651, 0.8408, 0.8477, 0.8613, 0.8584, 0.8325, 0.8306, 0.8940, 0.7798, 0.7397, 0.8374, 0.8716, 0.8701, 0.8452, 0.8818, 0.6792, 0.8232, 0.7656,\n 0.7432, 0.8486, 0.8472, 0.8652, 0.7983, 0.8247, 0.8374, 0.8438, 0.8408, 0.8457, 0.8208, 0.7983, 0.8770, 0.9009, 0.8755, 0.8154, 0.7876, 0.8833, 0.8628, 0.8159, 0.8447, 0.8188, 0.8408, 0.8252, 0.8105, 0.8857, 0.8320, 0.8169, 0.8970, 0.9233, 0.8198, 0.8828, 0.8306, 0.8506, 0.8315, 0.7861, 0.7979, 0.9111, 0.8354,\n 0.8003, 0.8647, 0.8535, 0.8042, 0.8174, 0.7827, 0.8506, 0.7622, 0.8408, 0.8789, 0.8330, 0.8008, 0.8491, 0.8223, 0.8105, 0.7837, 0.9111, 0.8901, 0.8896, 0.8394, 0.8491, 0.8320, 0.8496, 0.8730, 0.6167, 0.7734, 0.7979], dtype=torch.float16), 'perceiver_resampler.layers.2.1.0.bias': tensor([ 1.2012e-01, 1.9562e-02, -5.5634e-02, 1.2067e-01, 9.6741e-03, 9.8572e-03, 2.9938e-02, -2.3544e-02, -1.3100e-02, 4.9561e-02, -1.3879e-01, 2.0410e-01, -1.1383e-02, 5.9540e-02, -6.7139e-02, -7.8125e-02, 4.3304e-02, -7.5928e-02, -9.3079e-02, -2.7313e-02, 1.0565e-01, 3.7956e-03, 8.9111e-02, -1.3074e-01,\n 1.0797e-01, -1.0852e-01, -3.1494e-02, 1.2164e-01, -2.8687e-02, -1.2830e-01, 2.0035e-02, 3.5187e-02, 1.4331e-01, -1.8176e-01, -1.5356e-01, -1.0938e-01, 4.5280e-03, -8.3191e-02, 9.2224e-02, 1.5967e-01, -2.1408e-02, 3.3234e-02, 4.4037e-02, -6.4270e-02, -9.2773e-02, 9.7717e-02, 5.6641e-02, -8.9294e-02,\n -1.6693e-02, 6.9397e-02, -4.1351e-02, -6.8481e-02, -6.8176e-02, -4.1779e-02, -8.9722e-02, -7.7637e-02, 4.1733e-03, -1.8091e-01, 8.6853e-02, -5.8655e-02, 6.6223e-02, -8.0750e-02, -8.6975e-02, 1.7319e-02, 1.2335e-01, -7.1411e-02, 2.4300e-03, -1.3184e-01, -2.5005e-03, -1.2451e-01, 1.0925e-01, -1.4229e-02,\n -2.8809e-02, 7.2754e-02, -6.3049e-02, 3.5095e-04, 7.7332e-02, 1.0486e-01, -1.0138e-01, -7.8430e-02, -5.1453e-02, 8.1482e-02, -1.4172e-01, 1.7654e-02, -6.5430e-02, -1.7126e-01, 2.0966e-02, 1.2903e-01, 1.8701e-01, 1.6162e-01, -2.0496e-01, 2.7817e-02, -1.1877e-01, 6.2408e-02, 7.8613e-02, 8.0872e-02,\n -1.7761e-02, 1.2598e-01, 1.1035e-01, -5.4871e-02, -6.3538e-02, -8.1055e-02, -1.0022e-01, 1.3049e-01, -5.9753e-02, -6.7505e-02, -1.0162e-01, 7.4402e-02, -1.2341e-01, -1.2488e-01, 5.1880e-02, -1.9012e-02, 4.9469e-02, 4.0192e-02, -1.4819e-01, -5.8517e-03, -1.3232e-01, 5.0476e-02, -7.4463e-03, 6.1737e-02,\n 6.8176e-02, 1.1505e-01, 1.1475e-01, 1.0303e-01, 6.9153e-02, -8.9294e-02, -5.8899e-03, 7.4707e-02, 7.3669e-02, 2.0752e-01, 2.4475e-02, -8.4229e-02, 1.8295e-02, -1.1719e-01, 5.0720e-02, -5.0385e-02, 1.4819e-01, -9.7839e-02, -1.4503e-02, -7.3547e-02, -2.7283e-02, -4.5288e-02, -6.8237e-02, -8.1055e-02,\n 1.5088e-01, 3.4790e-02, -1.7151e-02, 5.3864e-02, -9.2041e-02, 2.9388e-02, -2.0111e-02, -1.4331e-01, 1.8579e-01, 9.1187e-02, 7.1655e-02, 1.6956e-01, -1.3916e-01, 1.9455e-02, -4.4495e-02, 1.0980e-01, 9.8083e-02, 1.2347e-01, -5.9479e-02, 1.1859e-01, -6.7673e-03, 5.8258e-02, 1.1847e-01, 7.5439e-02,\n -9.1431e-02, -1.1523e-01, 4.5074e-02, -8.1604e-02, -5.4932e-02, -3.2867e-02, 1.3037e-01, 1.8433e-01, 5.8022e-03, -4.3060e-02, -6.9702e-02, 2.3645e-01, 6.7261e-02, -4.9622e-02, -1.0760e-01, -1.0187e-01, 1.7197e-02, 1.2659e-01, -8.1116e-02, 1.1432e-01, 1.4130e-02, -7.1838e-02, 2.8183e-02, -5.8411e-02,\n -1.0376e-01, 1.7236e-01, 1.1652e-01, 1.1543e-02, -1.5076e-01, 6.1401e-02, -5.3528e-02, 3.9642e-02, 1.2489e-02, -9.2896e-02, -2.1400e-03, -1.2537e-01, -9.3689e-02, -2.1759e-02, 1.1969e-01, 8.5571e-02, 1.6357e-02, 5.7648e-02, -1.0962e-01, 1.2650e-02, 7.5989e-02, 1.1963e-01, -9.4543e-02, -1.1450e-01,\n -6.3477e-03, -1.1299e-02, 6.8542e-02, -1.5674e-01, 5.7678e-02, 1.1096e-01, 9.6436e-02, -9.2163e-03, 8.2016e-03, -8.1604e-02, -1.7908e-01, -4.5563e-02, 7.9773e-02, -4.4067e-02, 3.7079e-02, -2.0471e-01, -1.1157e-01, -9.3628e-02, 8.2458e-02, -4.5441e-02, 1.4709e-02, 1.9775e-01, -1.8884e-01, -1.0590e-01,\n 7.0496e-02, 1.8738e-02, -4.1138e-02, -1.1978e-02, 1.1725e-01, -6.9771e-03, -1.3428e-01, -1.8646e-02, 1.1572e-01, -1.4966e-01, 7.9346e-02, 7.4348e-03, -9.5825e-02, -1.3037e-01, -1.6260e-01, 7.5928e-02, 8.6548e-02, -2.6642e-02, -6.8481e-02, -7.7332e-02, 1.8225e-01, -7.7332e-02, -1.8262e-01, -2.7695e-02,\n 7.7881e-02, -7.1777e-02, -3.3478e-02, -2.7725e-02, -5.1605e-02, -4.5746e-02, -9.1370e-02, -1.4136e-01, 5.9448e-02, 4.8950e-02, -1.5784e-01, 1.6403e-02, -1.6553e-01, -8.4656e-02, -8.1604e-02, 1.3635e-01, -1.8509e-02, 1.5674e-01, -2.1082e-01, 1.4307e-01, 1.4490e-01, -3.9825e-02, -6.7673e-03, 5.4047e-02,\n -1.3843e-01, -2.9861e-02, 1.2695e-01, -1.1212e-01, 4.3060e-02, 7.8735e-02, 3.3691e-02, -1.3574e-01, 4.0009e-02, -2.7939e-02, -5.1994e-03, 9.0210e-02, 6.7932e-02, 1.3684e-01, -3.5767e-02, 1.0870e-01, -1.0175e-01, 3.2635e-03, 1.5335e-02, -1.0028e-01, -9.2224e-02, 1.4209e-01, -7.5134e-02, 1.2537e-01,\n 4.2847e-02, -9.8999e-02, -1.2030e-01, -6.3904e-02, 3.2623e-02, 4.3427e-02, 9.8511e-02, -7.8857e-02, -1.1023e-01, 1.1584e-01, 4.9316e-02, 1.6394e-01, 1.7078e-01, 2.5314e-02, -5.6610e-02, 1.1292e-01, -9.8389e-02, -4.7791e-02, -1.6809e-01, 4.6997e-02, -1.0944e-01, 1.5354e-04, 1.9836e-02, -7.2388e-02,\n 1.3672e-01, 3.4241e-02, -1.0016e-01, 1.2329e-01, -1.0516e-01, 8.3313e-02, 1.8286e-01, -7.8125e-02, -3.6926e-02, -1.3354e-01, 2.8275e-02, -1.2646e-01, -3.4271e-02, 2.4292e-02, 1.1528e-02, -1.6203e-03, 2.0154e-01, 7.3608e-02, -9.9426e-02, 1.2781e-01, -3.9490e-02, 6.7566e-02, -1.2189e-01, 1.6388e-02,\n -1.1304e-01, 7.1533e-02, 2.6245e-03, -1.3171e-01, 7.3486e-02, 5.9235e-02, -1.0028e-01, -7.2083e-02, 3.9185e-02, 6.0883e-02, 1.3916e-01, -4.0863e-02, 2.2018e-02, 1.6577e-01, 3.9043e-03, -4.4128e-02, 6.3782e-02, 1.2891e-01, 2.4323e-02, 1.9547e-02, 1.5515e-01, -4.2633e-02, 2.2720e-02, -1.0876e-01,\n 2.2415e-02, -1.4124e-01, 4.8096e-02, -5.8823e-03, -8.7402e-02, 9.7290e-02, 1.0376e-02, 1.9177e-01, 1.4783e-01, -2.5101e-02, -6.4392e-02, -1.0858e-01, -1.3027e-03, -2.5177e-02, -2.4948e-03, -1.4917e-01, -8.5144e-02, -9.6008e-02, -1.0205e-01, -1.8433e-01, 9.7107e-02, 8.4961e-02, 7.6355e-02, 9.7122e-03,\n 3.8910e-02, 1.2939e-01, -4.6265e-02, 1.1902e-01, -1.4893e-01, 1.4648e-01, 1.5198e-01, 1.2001e-02, 1.1115e-01, -1.2213e-01, -1.2384e-01, -1.3269e-01, -1.1768e-01, 9.9670e-02, 8.6121e-02, 9.0393e-02, 1.1279e-01, -1.5759e-01, 3.9005e-03, -1.7944e-01, 1.2988e-01, 1.2360e-02, 1.0178e-02, 1.2093e-02,\n -1.0126e-01, 8.9111e-02, 1.6833e-01, -6.8909e-02, -6.2866e-02, 1.4319e-01, -3.6682e-02, -5.6152e-02, -1.5247e-01, -4.9973e-04, -6.2012e-02, -1.7868e-02, -7.3767e-04, 8.0200e-02, 7.8308e-02, -2.8366e-02, -6.0791e-02, 1.3721e-01, -3.9330e-03, 4.3427e-02, 4.0558e-02, 5.8632e-03, -1.3831e-01, -1.3464e-01,\n -1.7151e-02, -1.5396e-02, -2.4872e-02, -1.2500e-01, 8.2458e-02, 4.2694e-02, 7.0007e-02, -3.1250e-02, 3.6285e-02, 2.6520e-02, -2.0187e-02, 8.4167e-02, -2.6794e-02, 1.0425e-01, 1.0822e-01, 2.5314e-02, 7.4158e-02, 4.6844e-03, 1.5926e-03, 7.0007e-02, -7.8186e-02, -1.1377e-01, -7.4097e-02, 6.7825e-03,\n -1.0040e-01, -2.0874e-02, 6.5674e-02, -8.2092e-03, 6.8481e-02, -6.7291e-03, 3.7048e-02, -4.9629e-03, -7.8918e-02, -1.2199e-02, -1.0602e-01, -8.6609e-02, 3.6682e-02, 3.3051e-02, 4.7150e-02, -1.4893e-01, -1.7261e-01, -8.8562e-02, -1.6956e-01, -3.6743e-02, 7.1335e-03, -9.6130e-02, 1.4380e-01, -3.4119e-02,\n 3.6285e-02, -1.5442e-01, 1.0785e-01, 9.1553e-02, -8.5876e-02, -5.3467e-02, 1.0449e-01, -9.4238e-02, -1.0529e-01, -9.1858e-02, -1.3940e-01, 1.1523e-01, -8.4656e-02, -3.5736e-02, -6.2561e-02, 3.2562e-02, -1.1755e-01, 6.8298e-02, -5.9418e-02, 1.1151e-01, 2.1057e-02, -5.9929e-03, 1.5100e-01, 1.0480e-01,\n 1.1383e-02, 1.1951e-01, 4.3213e-02, 1.1169e-01, -2.4277e-02, 1.5030e-02, -7.0557e-02, -5.2765e-02, -2.9602e-02, 6.6650e-02, -4.5502e-02, -1.5125e-01, -1.3696e-01, 1.3440e-01, -3.4149e-02, -1.2744e-01, -8.3130e-02, -1.1133e-01, -4.1351e-02, -5.7617e-02, -1.3049e-01, -1.3806e-01, -1.0986e-01, 9.3811e-02,\n -5.1155e-03, 2.7832e-02, -6.6406e-02, 1.4758e-01, -1.3806e-01, 1.4941e-01, -6.0913e-02, -6.7688e-02, -2.1912e-02, 2.2729e-01, -4.7760e-02, -7.0862e-02, 4.4098e-02, 4.3915e-02, 2.4826e-02, 5.4626e-02, 6.7566e-02, -1.7810e-01, -3.0991e-02, 9.6924e-02, 6.7749e-02, -1.4062e-01, 1.4514e-01, 9.4177e-02,\n -1.5625e-01, -8.2092e-02, -2.6749e-02, 1.1276e-02, 7.2754e-02, 1.6626e-01, -1.4748e-02, -9.9670e-02, 1.0645e-01, 3.1872e-03, 9.5673e-03, 4.0802e-02, -1.2006e-01, 7.8918e-02, -1.4709e-01, 1.0022e-01, -6.2134e-02, 4.6478e-02, 3.0106e-02, 5.4596e-02, 1.2878e-01, 3.5278e-02, -4.8462e-02, 1.3464e-01,\n -1.0266e-01, 4.3701e-02, 1.6504e-01, 1.5007e-02, -9.2102e-02, -6.7825e-03, 2.2049e-02, 1.2146e-02, -1.4221e-02, 3.5339e-02, -7.0129e-02, 8.2169e-03, -1.1755e-01, 9.2468e-02, -7.8979e-02, 7.3425e-02, 3.6255e-02, -8.0261e-02, 9.0027e-02, -2.5101e-03, 5.6976e-02, -4.8737e-02, -1.1914e-01, -5.1361e-02,\n -2.9694e-02, 1.2018e-01, 2.6566e-02, 7.7393e-02, -9.8816e-02, 1.5222e-01, -7.9468e-02, 8.1055e-02, -5.8044e-02, 1.3147e-01, -5.9143e-02, 5.1361e-02, 1.1755e-01, -1.4148e-01, 1.2955e-02, 5.9967e-03, -4.5105e-02, 1.1353e-01, -6.6528e-02, 9.0759e-02, 1.6919e-01, -7.5500e-02, -2.0233e-02, -1.0187e-01,\n -9.3445e-02, 1.4453e-01, 2.9240e-03, 4.0192e-02, -5.4565e-02, 4.5044e-02, 6.0987e-04, 6.3782e-02, -4.3221e-03, -1.7715e-02, -5.1056e-02, 1.3855e-01, -1.3928e-01, 3.6259e-03, -9.1919e-02, 1.3623e-01, 2.5589e-02, 1.3171e-01, 1.0272e-01, -3.2379e-02, -5.2460e-02, -7.0068e-02, -9.2163e-02, -4.3945e-02,\n -1.8372e-02, 4.9530e-02, 1.4136e-01, 9.0881e-02, 3.3234e-02, -2.2339e-02, -6.5979e-02, -7.0129e-02, -8.4381e-03, -1.0962e-01, 2.2385e-02, 1.2903e-01, 3.0655e-02, -3.1494e-02, 1.1945e-01, -5.3192e-02, 1.3782e-01, -9.3842e-03, 5.3009e-02, 3.2928e-02, -6.2141e-03, 3.0579e-02, 7.3669e-02, -4.6112e-02,\n -1.5465e-02, -5.8807e-02, 1.1029e-01, 3.7323e-02, -7.4219e-02, -1.5930e-02, 2.7267e-02, 4.2145e-02, -1.0089e-01, 4.2328e-02, 1.1761e-01, 1.0474e-01, -7.5912e-03, 2.9022e-02, -8.5815e-02, 5.2429e-02, -3.0640e-02, 1.3586e-01, 4.1626e-02, 7.1106e-02, 1.0663e-01, -2.4750e-02, 6.7383e-02, -6.0303e-02,\n 7.3303e-02, 6.2347e-02, -1.3329e-02, 9.3079e-03, -7.0862e-02, -5.3070e-02, 9.4421e-02, -1.9617e-01, 5.8411e-02, -1.0797e-01, -7.4890e-02, 1.6467e-01, -9.9609e-02, 1.3416e-01, -1.4514e-01, -9.2468e-03, 1.1407e-01, 4.8126e-02, 4.4434e-02, 1.0449e-01, -6.2927e-02, -3.0807e-02, -5.8258e-02, -3.7567e-02,\n -4.1656e-02, 4.2908e-02, 5.2551e-02, -9.3384e-02, -3.5309e-02, -8.3160e-04, -7.7332e-02, 3.1311e-02, -8.0627e-02, -1.1456e-01, 4.8676e-02, 1.3464e-01, -7.3364e-02, 1.0980e-01, -2.0410e-01, -4.5074e-02, -9.8633e-02, 1.2744e-01, -2.8564e-02, 1.1572e-01, 7.7026e-02, 9.2087e-03, -9.6741e-02, 1.5656e-02],\n dtype=torch.float16), 'perceiver_resampler.layers.2.1.1.weight': tensor([[ 1.9928e-02, 2.5558e-02, -3.7384e-02, ..., -3.4821e-02, -1.8555e-02, 4.2763e-03],\n [ 3.6335e-03, -2.1729e-02, 4.2305e-03, ..., -1.3031e-02, 1.1833e-02, -2.2369e-02],\n [ 2.8839e-02, -1.6298e-03, 4.6959e-03, ..., -2.1912e-02, 1.8738e-02, 1.8326e-02],\n ...,\n [-4.5357e-03, 9.6817e-03, -4.2999e-02, ..., -2.9007e-02, 2.4796e-02, 2.2293e-02],\n [-2.8172e-03, -3.2898e-02, 5.1928e-04, ..., -1.6907e-02, 1.6422e-03, 1.5610e-02],\n [ 3.5309e-02, -5.1856e-06, -2.0126e-02, ..., -8.4763e-03, 4.6539e-02, -3.2410e-02]], dtype=torch.float16), 'perceiver_resampler.layers.2.1.3.weight': tensor([[-0.0481, 0.0066, 0.0099, ..., -0.0383, -0.0097, -0.0163],\n [ 0.0493, -0.0042, 0.0010, ..., 0.0362, -0.0059, -0.0194],\n [ 0.0165, -0.0070, -0.0018, ..., -0.0112, 0.0005, -0.0239],\n ...,\n [ 0.0082, -0.0343, -0.0194, ..., 0.0151, -0.0472, -0.0011],\n [-0.0166, -0.0325, 0.0084, ..., 0.0244, -0.0204, 0.0396],\n [ 0.0023, 0.0115, -0.0123, ..., 0.0142, 0.0110, -0.0283]], dtype=torch.float16), 'perceiver_resampler.layers.3.0.norm1.weight': tensor([0.7251, 0.7427, 0.6470, 0.7793, 0.7197, 0.7930, 0.7305, 0.7314, 0.7876, 0.7451, 0.7222, 0.6836, 0.7402, 0.7573, 0.8003, 0.7446, 0.7837, 0.6792, 0.7495, 0.7368, 0.7510, 0.7729, 0.7251, 0.6948, 0.6919, 0.7676, 0.7319, 0.6909, 0.6880, 0.7397, 0.7705, 0.7275, 0.7344, 0.7256, 0.7363, 0.6655, 0.6372, 0.7119, 0.6831,\n 0.7710, 0.8105, 0.7993, 0.7715, 0.7300, 0.7466, 0.7344, 0.6631, 0.7319, 0.7183, 0.6904, 0.7090, 0.7217, 0.7324, 0.7227, 0.6968, 0.7598, 0.7256, 0.6553, 0.7490, 0.7827, 0.7427, 0.7817, 0.7266, 0.7632, 0.7866, 0.6875, 0.7729, 0.7490, 0.7417, 0.7705, 0.7715, 0.7256, 0.7383, 0.7637, 0.6992, 0.7588, 0.6714, 0.7329,\n 0.7368, 0.7368, 0.7603, 0.7803, 0.7827, 0.6948, 0.7944, 0.7661, 0.7041, 0.7139, 0.8008, 0.7158, 0.7510, 0.7153, 0.6768, 0.7427, 0.7373, 0.7075, 0.7837, 0.7363, 0.7451, 0.7598, 0.6826, 0.7266, 0.7524, 0.7368, 0.7915, 0.7241, 0.7490, 0.7393, 0.7471, 0.7686, 0.7227, 0.7388, 0.7554, 0.7451, 0.7158, 0.7354, 0.6968,\n 0.7905, 0.6523, 0.7183, 0.7656, 0.7212, 0.7397, 0.8955, 0.7354, 0.7119, 0.7178, 0.7495, 0.6055, 0.7334, 0.6914, 0.6904, 0.7275, 0.7324, 0.7134, 0.7690, 0.7573, 0.7178, 0.7861, 0.6948, 0.7217, 0.7202, 0.7373, 0.7017, 0.6816, 0.7788, 0.7900, 0.7114, 0.8071, 0.7461, 0.7173, 0.7485, 0.7627, 0.7329, 0.7148, 0.7446,\n 0.7095, 0.6543, 0.7456, 0.7266, 0.7007, 0.7056, 0.7095, 0.7485, 0.7031, 0.7773, 0.7422, 0.6855, 0.7935, 0.7520, 0.7388, 0.7153, 0.7261, 0.7490, 0.7622, 0.6660, 0.7236, 0.7026, 0.6763, 0.7153, 0.7446, 0.7168, 0.7441, 0.7852, 0.8008, 0.7744, 0.5327, 0.7510, 0.6758, 0.6836, 0.8003, 0.6113, 0.7271, 0.7363, 0.7275,\n 0.7163, 0.7119, 0.7148, 0.7632, 0.8369, 0.6685, 0.7290, 0.6704, 0.6890, 0.7324, 0.7637, 0.6899, 0.7314, 0.7261, 0.7549, 0.6558, 0.7402, 0.6357, 0.7949, 0.6851, 0.6958, 0.7739, 0.7754, 0.7715, 0.7822, 0.7163, 0.7461, 0.7085, 0.7456, 0.7612, 0.6460, 0.7959, 0.7729, 0.6943, 0.7158, 0.7476, 0.7441, 0.7261, 0.6919,\n 0.7368, 0.7114, 0.7583, 0.7891, 0.6758, 0.7466, 0.7144, 0.7212, 0.6963, 0.6821, 0.6855, 0.7061, 0.7554, 0.7866, 0.7295, 0.7983, 0.7993, 0.7031, 0.6807, 0.7021, 0.6860, 0.7021, 0.7866, 0.7334, 0.6943, 0.7334, 0.7344, 0.7456, 0.7642, 0.7114, 0.7246, 0.7100, 0.7017, 0.7422, 0.7681, 0.7290, 0.7329, 0.7212, 0.8057,\n 0.7202, 0.7280, 0.7446, 0.8047, 0.6978, 0.7549, 0.6729, 0.7583, 0.7061, 0.7300, 0.6353, 0.7324, 0.6323, 0.7603, 0.7832, 0.7725, 0.7524, 0.7275, 0.6514, 0.7358, 0.5942, 0.7100, 0.7266, 0.7769, 0.7349, 0.7529, 0.6562, 0.7349, 0.6851, 0.6167, 0.7866, 0.7202, 0.6206, 0.7432, 0.7476, 0.7544, 0.7046, 0.7241, 0.7231,\n 0.7422, 0.7271, 0.7017, 0.7925, 0.7192, 0.7153, 0.7256, 0.8340, 0.7275, 0.7339, 0.6978, 0.6738, 0.7485, 0.7017, 0.7661, 0.7207, 0.7461, 0.7007, 0.6089, 0.6777, 0.7134, 0.7290, 0.6914, 0.7231, 0.7900, 0.7275, 0.7925, 0.6855, 0.7476, 0.7358, 0.7144, 0.7173, 0.7339, 0.7603, 0.7197, 0.7666, 0.7510, 0.7256, 0.6699,\n 0.6968, 0.7104, 0.7334, 0.7134, 0.7773, 0.7544, 0.7324, 0.7656, 0.6772, 0.7314, 0.7363, 0.7344, 0.8047, 0.7036, 0.6821, 0.7148, 0.7690, 0.7529, 0.7256, 0.7397, 0.7925, 0.7573, 0.7725, 0.7676, 0.7573, 0.7783, 0.7349, 0.6992, 0.7510, 0.7056, 0.6904, 0.7026, 0.7095, 0.7402, 0.7261, 0.7417, 0.7124, 0.7026, 0.7109,\n 0.7510, 0.7700, 0.7607, 0.7197, 0.7031, 0.7354, 0.7046, 0.6772, 0.7495, 0.7573, 0.7988, 0.7163, 0.7612, 0.7310, 0.7495, 0.7949, 0.6914, 0.7544, 0.7021, 0.7671, 0.7095, 0.7231, 0.6919, 0.7256, 0.6479, 0.7671, 0.7017, 0.7432, 0.7246, 0.7544, 0.6826, 0.7539, 0.7695, 0.8120, 0.7612, 0.7139, 0.7158, 0.7158, 0.7236,\n 0.7329, 0.7383, 0.6641, 0.7720, 0.7793, 0.7427, 0.7090, 0.7397, 0.7612, 0.7510, 0.7515, 0.7212, 0.7344, 0.6997, 0.7695, 0.7070, 0.7734, 0.7949, 0.7593, 0.7222, 0.7544, 0.6543, 0.6689, 0.7437, 0.6895, 0.7607, 0.7256, 0.6699, 0.7271, 0.7446, 0.8203, 0.7134, 0.7646, 0.7285, 0.6328, 0.7109, 0.7158, 0.7666, 0.7148,\n 0.7339, 0.7759, 0.6963, 0.7485, 0.7637, 0.6733, 0.7495, 0.7314, 0.7192, 0.7114, 0.7646, 0.7104, 0.7412, 0.7202, 0.7817, 0.7129, 0.7690, 0.6875, 0.7886, 0.6802, 0.7212, 0.6851, 0.7427, 0.6289, 0.7314, 0.6694, 0.7388, 0.7583, 0.7607, 0.7266, 0.6963, 0.7324, 0.7461, 0.7407, 0.6826, 0.6611, 0.7461, 0.6792, 0.7231,\n 0.7329, 0.6655, 0.7495, 0.7104, 0.7622, 0.6787, 0.7798, 0.7329, 0.7930, 0.7490, 0.7373, 0.7666, 0.7358, 0.7319, 0.7529, 0.8076, 0.7798, 0.7715, 0.7295, 0.6396, 0.7104, 0.6548, 0.5938, 0.7451, 0.6538, 0.7393, 0.6772, 0.7622, 0.7515, 0.6748, 0.7441, 0.7739, 0.6992, 0.7344, 0.7192, 0.7974, 0.6895, 0.7031, 0.7471,\n 0.7251, 0.7598, 0.7437, 0.7612, 0.7627, 0.7930, 0.7554, 0.7949, 0.8091, 0.6685, 0.8110, 0.6768, 0.7095, 0.7295, 0.7778, 0.6963, 0.7417, 0.7456, 0.7339, 0.6616, 0.7305, 0.7578, 0.7563, 0.7212, 0.7119, 0.8052, 0.7305, 0.6567, 0.6689, 0.7017, 0.7798, 0.6636, 0.7324, 0.7261, 0.6909, 0.7783, 0.7471, 0.8276, 0.7285,\n 0.6924, 0.7666, 0.7168, 0.7607, 0.7534, 0.6865, 0.7515, 0.6812, 0.7764, 0.6885, 0.7393, 0.7773, 0.7471, 0.7104, 0.6904, 0.7095, 0.6641, 0.6372, 0.7119, 0.7622, 0.7163, 0.6978, 0.7124, 0.7808, 0.6465, 0.7344, 0.7778, 0.6973, 0.7144, 0.7080, 0.7407, 0.6973, 0.7612, 0.7861, 0.7534, 0.7515, 0.7480, 0.7451, 0.7764,\n 0.7490, 0.6528, 0.7358, 0.7583, 0.7090, 0.7324, 0.7148, 0.7461, 0.7598, 0.7578, 0.7988, 0.7314, 0.7183, 0.7480, 0.6992, 0.6943, 0.7480, 0.7036, 0.7832, 0.7329, 0.6943, 0.7256, 0.8022, 0.7158, 0.6899, 0.7788, 0.7144, 0.6792, 0.7388, 0.6802, 0.6797, 0.7026, 0.7212, 0.7485, 0.7266, 0.7886, 0.7163, 0.7163, 0.7319,\n 0.6719, 0.7241, 0.7451, 0.8291, 0.7168, 0.7314, 0.6899, 0.6831, 0.7251, 0.7480, 0.6963, 0.7808, 0.7451, 0.7212, 0.7764, 0.7231, 0.7217, 0.7778, 0.7451, 0.7070, 0.7319, 0.7212, 0.7007, 0.7471, 0.7217, 0.7500, 0.7368, 0.7710, 0.7017, 0.7485, 0.6948, 0.7349, 0.7422, 0.6924, 0.7725, 0.7705, 0.7207, 0.6831, 0.7388,\n 0.7534, 0.6304, 0.6895, 0.7036, 0.7500, 0.7368, 0.8047, 0.7041, 0.7207, 0.6626, 0.7144, 0.7295, 0.7930, 0.7129, 0.7358, 0.7402, 0.6792, 0.6089, 0.7495, 0.7905, 0.7388, 0.7144, 0.7153, 0.7002, 0.6953, 0.6777, 0.6904, 0.7920, 0.7188, 0.7358, 0.7368, 0.6851, 0.7363, 0.7529, 0.7515, 0.6997, 0.7212, 0.7046, 0.7246,\n 0.6411, 0.7422, 0.7129, 0.7446, 0.7144, 0.6943, 0.7251, 0.7510, 0.7671, 0.7607, 0.7378, 0.7471, 0.7471, 0.7183, 0.7129, 0.7681, 0.6294, 0.7192, 0.7144, 0.7378, 0.6855, 0.7319, 0.7397, 0.7466, 0.6816, 0.6851, 0.6797], dtype=torch.float16), 'perceiver_resampler.layers.3.0.norm1.bias': tensor([ 1.4565e-02, -2.9388e-02, 4.8187e-02, -3.0327e-03, 3.2806e-02, -3.1738e-02, 7.3090e-03, 2.7866e-03, -2.2447e-04, -2.4681e-03, 3.1643e-03, 4.4220e-02, 1.9882e-02, -1.4313e-02, 3.0228e-02, 1.0155e-02, -1.1375e-02, 7.1259e-03, 1.4099e-02, -3.3295e-02, -2.6536e-04, -4.7684e-03, -2.0676e-02, 2.1408e-02,\n 2.6871e-02, 2.0611e-04, -1.6312e-02, 2.5085e-02, -6.0638e-02, -4.9805e-02, -1.5656e-02, -1.2146e-02, 3.2074e-02, -2.5696e-02, -2.5606e-04, -3.9253e-03, 1.0292e-02, -2.0027e-03, -4.0588e-03, 2.5757e-02, 3.0716e-02, -4.7035e-03, -1.7212e-02, -2.6855e-02, -1.6968e-02, -5.4626e-03, -4.1473e-02, -2.0569e-02,\n 1.6281e-02, 1.9989e-02, 1.9363e-02, 1.0361e-02, -4.6463e-03, -7.8344e-04, -4.9973e-03, -1.2268e-02, 4.3915e-02, 2.1729e-02, -1.3718e-02, -8.6594e-03, -7.0152e-03, 1.6127e-03, -5.0485e-05, 1.2001e-02, 2.4681e-03, 2.4857e-02, 1.2505e-02, -1.6144e-02, -3.2898e-02, 5.5908e-02, 1.8524e-02, 1.8768e-02,\n 4.5562e-04, -1.4183e-02, -3.3569e-02, -2.8473e-02, 4.3144e-03, -2.7256e-03, -2.1149e-02, 3.0823e-02, -5.3635e-03, -1.6846e-02, -5.4855e-03, -2.9907e-02, -2.2751e-02, 2.9907e-02, 4.7211e-02, -1.8661e-02, 9.1553e-03, 9.3079e-03, -2.5818e-02, -1.7328e-03, -5.2277e-02, 3.3844e-02, -3.9307e-02, 3.1219e-02,\n -3.6804e-02, -1.4084e-02, 8.4381e-03, -2.8046e-02, 3.9795e-02, -2.6417e-03, 3.1757e-03, -1.9867e-02, 6.9962e-03, -1.3008e-02, 1.2512e-02, -3.6041e-02, 3.3081e-02, -1.9211e-02, -4.6501e-03, 1.8005e-02, 1.0788e-02, -2.7061e-04, -3.3691e-02, -3.4809e-03, -1.2283e-02, -5.0049e-03, -8.1940e-03, 1.0101e-02,\n 1.7776e-02, -6.3591e-03, -4.0016e-03, -7.7553e-03, -1.2711e-02, 9.2545e-03, 1.1795e-02, 8.1558e-03, 4.3091e-02, 4.9591e-02, 2.0203e-02, 3.5919e-02, -1.7105e-02, -1.4000e-02, -2.2278e-02, 1.2527e-02, 4.1504e-02, -5.4073e-04, -1.0208e-02, -2.2491e-02, 4.2367e-04, -6.1340e-03, 2.9816e-02, -3.2532e-02,\n 1.6678e-02, -2.8244e-02, 2.1469e-02, 1.2306e-02, -1.7563e-02, 1.0635e-02, 3.6835e-02, 6.0005e-03, 1.2749e-02, 1.9760e-02, -3.2825e-03, 2.4719e-03, -5.2681e-03, -3.9154e-02, -7.3776e-03, -2.3697e-02, 1.6983e-02, 4.4525e-02, -2.9984e-02, -1.4618e-02, -4.2801e-03, 2.8107e-02, 2.1759e-02, -1.4893e-02,\n 2.6749e-02, -2.1790e-02, 1.2245e-02, 2.8290e-02, -7.4272e-03, -4.3518e-02, -3.3951e-03, -5.5237e-02, 1.7560e-04, 4.6326e-02, 4.7913e-02, 4.0222e-02, 3.0384e-03, 1.6312e-02, -1.0170e-02, -4.3091e-02, 3.5217e-02, 2.9602e-02, -2.8223e-01, 4.0527e-02, 3.2166e-02, -2.9469e-03, -1.3176e-02, 4.2999e-02,\n 3.9062e-02, 3.6224e-02, 4.5433e-03, 4.2648e-03, 5.0171e-02, 3.1067e-02, 1.4029e-03, -2.1942e-02, 7.3120e-02, -4.2206e-02, -1.0635e-02, -1.2306e-02, -3.4008e-03, 2.7420e-02, 8.8348e-03, 1.6953e-02, 8.1253e-03, 2.3743e-02, -2.2263e-02, 7.6771e-04, -8.9294e-02, 3.7384e-02, 1.9348e-02, 1.8356e-02,\n -1.2596e-02, 2.4384e-02, 1.5306e-03, -1.5282e-02, 1.0401e-04, -4.1046e-02, -3.8727e-02, 2.0035e-02, -2.6566e-02, -4.3945e-02, 1.5465e-02, 2.5528e-02, 4.0558e-02, -7.4730e-03, -3.9101e-03, -4.2969e-02, -1.8158e-02, 5.0934e-02, -1.9503e-03, -1.7441e-02, 2.6817e-03, 1.8814e-02, 1.6388e-02, -6.8970e-02,\n 2.5528e-02, 2.6672e-02, -2.0523e-02, 6.9580e-02, -2.5848e-02, 2.4521e-02, 2.2324e-02, -2.6917e-02, -1.7899e-02, 6.6895e-02, -3.8208e-02, -4.2847e-02, 4.0398e-03, 1.1818e-02, 6.0921e-03, -3.5934e-03, 1.6159e-02, -1.9007e-03, 2.6886e-02, 4.4159e-02, -4.2999e-02, -7.1411e-03, 1.9703e-03, -3.9635e-03,\n -7.3280e-03, -1.6754e-02, 2.5253e-02, 1.9012e-02, -1.2634e-02, -9.3079e-03, -4.8141e-03, -1.5020e-03, -1.0056e-02, -3.9215e-03, 2.6718e-02, -5.6732e-02, 3.0289e-02, -3.7170e-02, -3.6545e-03, 2.4765e-02, -1.3153e-02, 3.4241e-02, -5.8594e-03, -7.0419e-03, -1.5701e-02, 1.5556e-02, -1.9501e-02, 3.3997e-02,\n 3.2379e-02, -3.0685e-02, 1.4175e-02, 8.3435e-02, 1.7029e-02, 2.5162e-02, -3.2593e-02, -4.5357e-03, -1.0239e-02, -2.9404e-02, 1.3412e-02, 2.7557e-02, 2.7542e-03, -2.7420e-02, -3.3875e-02, 3.1204e-03, 1.9058e-02, -9.4681e-03, -2.6550e-02, 2.6718e-02, 9.0256e-03, 5.3940e-03, 4.3427e-02, -5.4474e-03,\n 9.1171e-03, -2.0538e-02, 1.0727e-02, -2.6779e-02, 5.0850e-03, 2.0874e-02, -1.5717e-02, 2.2934e-02, 2.5772e-02, 1.0445e-02, -8.5220e-03, -3.6682e-02, 1.8463e-02, -7.0419e-03, -4.9591e-02, -3.1586e-02, -9.0551e-04, 5.8441e-03, -3.0533e-02, 4.1992e-02, 3.7994e-02, 2.3376e-02, 5.2223e-03, -9.8324e-04,\n 3.6316e-02, -5.9090e-03, 2.2568e-02, -1.5045e-02, -1.0880e-02, -6.9031e-02, 4.0649e-02, -1.2901e-02, -2.8046e-02, 2.1118e-02, 3.5522e-02, 2.7344e-02, 5.0278e-03, 5.4565e-02, 4.5502e-02, -2.3544e-02, 2.1500e-02, -2.7130e-02, 2.1469e-02, 1.7166e-02, -1.9104e-02, 1.4519e-02, -1.5945e-02, -2.2827e-02,\n -3.5210e-03, -1.3641e-02, 4.8920e-02, 2.3819e-02, -2.1439e-02, 4.7119e-02, 8.8453e-04, -1.3275e-03, -1.8860e-02, -1.5083e-02, 2.1652e-02, -1.0391e-02, 1.2999e-03, -1.1978e-02, -1.7365e-02, 2.0706e-02, 2.7344e-02, 1.3634e-02, 3.2837e-02, 5.6610e-02, 1.5381e-02, 1.2375e-02, 1.7227e-02, 4.9652e-02,\n 8.2321e-03, 6.9885e-03, 2.9114e-02, -1.4946e-02, -3.6499e-02, -3.5019e-03, 2.2141e-02, 1.2512e-02, 2.3079e-04, -1.0605e-02, -2.1408e-02, 3.4607e-02, -3.1769e-02, 1.8768e-02, -1.6479e-02, 1.5564e-03, -5.6076e-03, -1.2741e-02, 1.1988e-03, 6.6910e-03, 5.3864e-03, -3.4618e-03, -4.3152e-02, -4.1626e-02,\n 7.9163e-02, -8.8882e-03, 1.3306e-02, 2.5284e-02, 2.6817e-03, -3.4882e-02, 4.1229e-02, -2.7405e-02, -1.8036e-02, -1.2024e-02, -2.8336e-02, 5.9776e-03, -2.5681e-02, 4.4800e-02, 1.0925e-02, 1.7731e-02, -2.1957e-02, -3.5065e-02, -3.7506e-02, -3.1921e-02, 2.4841e-02, 3.7964e-02, 6.3972e-03, -2.0233e-02,\n 1.3466e-02, -2.0752e-02, -1.9501e-02, 3.5400e-02, 5.9357e-02, 2.1118e-02, 1.7792e-02, -6.6528e-03, -5.8022e-03, -2.1877e-03, -2.5467e-02, 3.4851e-02, 1.4740e-02, 1.2024e-02, -1.9470e-02, -1.6769e-02, -1.0765e-02, 1.9522e-03, -3.2867e-02, 1.6434e-02, 6.2180e-03, 9.2773e-03, -1.4526e-02, 3.0151e-02,\n -4.8401e-02, 1.4351e-02, -4.4067e-02, -8.6136e-03, 1.0147e-02, -2.8748e-02, -1.5488e-02, -4.7546e-02, -2.9541e-02, 1.3359e-02, 2.5955e-02, 3.7079e-02, 3.6652e-02, -4.0283e-02, 6.4026e-02, -2.0370e-02, -3.4393e-02, 4.3259e-03, 1.1225e-03, 2.0199e-03, 6.5613e-03, 1.4915e-02, -2.5208e-02, 1.8005e-02,\n -2.6688e-02, -6.7444e-03, -1.5686e-02, -2.4200e-02, -1.0506e-02, 9.6989e-04, 1.4175e-02, 1.0974e-01, 7.8735e-03, 3.2654e-02, -6.5804e-03, 5.4504e-02, -5.2109e-03, -2.5955e-02, -5.1849e-02, 7.6180e-03, -2.1652e-02, 2.0477e-02, 5.5908e-02, -8.0948e-03, -4.0894e-02, -4.5891e-03, -1.3000e-02, 2.9278e-03,\n -1.0857e-02, -1.7593e-02, 8.6823e-03, -2.6123e-02, 1.7975e-02, 1.1925e-02, 2.6169e-02, 7.7605e-05, -9.3613e-03, -1.7212e-02, -1.1627e-02, 3.0090e-02, -1.3046e-02, -2.8397e-02, -1.4069e-02, 7.8087e-03, 2.1317e-02, 3.0106e-02, 2.0981e-03, -3.3779e-03, 3.7476e-02, -1.3229e-02, -1.7349e-02, -7.8583e-03,\n -2.4384e-02, 3.1952e-02, 3.1567e-03, -1.9379e-02, 2.9556e-02, -2.0004e-02, -2.3895e-02, -5.1270e-03, -3.7415e-02, 4.3793e-03, -3.4088e-02, 2.4170e-02, -2.3651e-02, -4.4403e-02, 2.1038e-03, 2.0264e-02, -1.2428e-02, 4.1840e-02, -2.1606e-02, 3.0727e-03, -1.7593e-02, 2.2736e-02, -2.3895e-02, 3.6438e-02,\n 3.5950e-02, 4.3884e-02, 2.5940e-02, 9.0942e-03, 4.0703e-03, 1.7136e-02, 4.5105e-02, 1.9287e-02, -1.9272e-02, 8.6212e-03, -1.8265e-02, -3.3325e-02, -6.8550e-03, -2.7542e-03, -2.2202e-02, 1.2108e-02, 2.1530e-02, 1.2383e-02, -1.8097e-02, 3.9024e-03, 5.1910e-02, -4.1962e-02, -1.4938e-02, -2.0020e-02,\n -3.3905e-02, -5.5115e-02, 1.7197e-02, 1.1040e-02, 2.7679e-02, -2.1038e-03, -1.1078e-02, 2.1179e-02, -4.1656e-02, 2.9343e-02, 1.0124e-02, 4.0527e-02, 3.3051e-02, -5.2124e-02, 1.4153e-02, 7.4196e-03, 3.1555e-02, -7.0190e-03, -2.2934e-02, -2.4475e-02, -5.8603e-04, -4.4037e-02, 1.6083e-02, 1.3664e-02,\n -8.8654e-03, 2.4078e-02, -4.2839e-03, 1.9180e-02, -1.1642e-02, -4.1901e-02, -1.9165e-02, -3.3661e-02, 9.5596e-03, 1.5373e-02, -1.4397e-02, -5.5817e-02, -1.6336e-03, 8.8501e-03, 1.1856e-02, 2.6031e-02, 2.8610e-02, -4.9782e-03, 1.6754e-02, -1.1391e-02, 3.6804e-02, 3.2410e-02, 8.4152e-03, -1.9241e-02,\n 2.0386e-02, 1.7490e-03, 3.4828e-03, -9.1705e-03, -1.9852e-02, 7.7963e-04, 4.4189e-02, 5.0629e-02, -3.0193e-03, 7.1430e-04, -3.1281e-02, 2.7065e-03, -4.7302e-02, 4.0817e-03, -3.7708e-03, 6.0310e-03, 6.4430e-03, 3.2043e-03, 1.0460e-02, 1.0498e-02, 2.6441e-04, 2.4857e-02, -3.3356e-02, -2.8286e-03,\n 3.7750e-02, -2.6073e-03, 1.6830e-02, -3.4943e-02, 1.3847e-02, 3.6743e-02, 1.0574e-02, 1.2863e-02, 1.7319e-02, -7.3486e-02, -1.0643e-02, 3.0212e-02, -3.7537e-03, -2.6657e-02, 2.3880e-02, -6.3801e-04, -5.4884e-04, -2.4765e-02, 1.4771e-02, 3.0060e-02, 1.4359e-02, 3.4546e-02, -3.5744e-03, -6.0043e-03,\n -4.0588e-02, -5.3215e-03, 3.4515e-02, -2.0142e-03, -1.2161e-02, -1.2886e-02, 2.6077e-02, -5.6038e-03, -4.4823e-03, -2.2537e-02, -5.9662e-02, 1.6861e-02, -5.0201e-03, -1.5373e-02, -2.6646e-03, -7.9117e-03, -2.9297e-03, -1.6968e-02, -2.1927e-02, -2.2766e-02, 5.6553e-04, 1.8677e-02, -1.5327e-02, 1.0910e-02,\n 6.2828e-03, -1.3657e-03, -8.7357e-03, 2.2202e-02, 2.8015e-02, 1.2039e-02, 3.3975e-04, -5.9624e-03, -1.0559e-02, -4.4975e-03, -2.2411e-03, -3.5095e-02, -9.8877e-03, -6.0120e-03, 4.9713e-02, 1.8280e-02, -9.1019e-03, 5.8823e-03, 2.3060e-03, -2.0599e-02, -1.1871e-02, 7.3547e-03, -4.1443e-02, 8.7036e-02,\n 2.1286e-02, 1.3847e-02, 8.4152e-03, 3.9856e-02, 1.1696e-02, -6.2073e-02, -3.7018e-02, 8.0017e-02, 1.1147e-02, 1.0773e-02, 1.5144e-02, 1.4153e-02, -2.1408e-02, 5.1636e-02, -3.2043e-04, -2.0248e-02, 2.2751e-02, -4.1161e-03, 3.9429e-02, 1.2733e-02, -4.3274e-02, 2.7176e-02, -4.8409e-03, 4.6539e-03,\n 4.6310e-03, 3.0899e-03, -5.2338e-03, -2.1973e-03, -2.6138e-02, -1.9241e-02, 4.3259e-03, 6.5842e-03, -1.2398e-02, -3.4912e-02, 1.8356e-02, -8.1635e-03, -5.5847e-02, -8.9600e-02, 2.1637e-02, 2.7481e-02, -1.2550e-02, -3.6285e-02, -1.8219e-02, 2.9144e-02, 2.7939e-02, -2.0264e-02, -1.4214e-02, -6.1111e-03],\n dtype=torch.float16), 'perceiver_resampler.layers.3.0.norm2.weight': tensor([0.6196, 0.6577, 0.6226, 0.6465, 0.5889, 0.6069, 0.6382, 0.6548, 0.6030, 0.5825, 0.6323, 0.6255, 0.6265, 0.6006, 0.5581, 0.5952, 0.6309, 0.6274, 0.7100, 0.5986, 0.6621, 0.6143, 0.5747, 0.5854, 0.5771, 0.6104, 0.5698, 0.6406, 0.6709, 0.6172, 0.5957, 0.6260, 0.6460, 0.6074, 0.5552, 0.6445, 0.6133, 0.6289, 0.6714,\n 0.6274, 0.6162, 0.5791, 0.6191, 0.5674, 0.6191, 0.5786, 0.6709, 0.6528, 0.6304, 0.5933, 0.6523, 0.6313, 0.5972, 0.7134, 0.6685, 0.7637, 0.6626, 0.6157, 0.6401, 0.6489, 0.6113, 0.6587, 0.5938, 0.6147, 0.6543, 0.6763, 0.5913, 0.6279, 0.6577, 0.6797, 0.6538, 0.5972, 0.6660, 0.6479, 0.6494, 0.7139, 0.6143, 0.5420,\n 0.4788, 0.6572, 0.6719, 0.6118, 0.5977, 0.6104, 0.6099, 0.6694, 0.6504, 0.6123, 0.5874, 0.6060, 0.6294, 0.5972, 0.6060, 0.6372, 0.6289, 0.6147, 0.6172, 0.6724, 0.5850, 0.6030, 0.6616, 0.6294, 0.6875, 0.5737, 0.6313, 0.6304, 0.6367, 0.6455, 0.5718, 0.6958, 0.6123, 0.5781, 0.6841, 0.6362, 0.6519, 0.6108, 0.5620,\n 0.7295, 0.6709, 0.6792, 0.6182, 0.6123, 0.6104, 0.6445, 0.5898, 0.6611, 0.5488, 0.6831, 0.6025, 0.6270, 0.6699, 0.6016, 0.6714, 0.6167, 0.6177, 0.6675, 0.6147, 0.6523, 0.5952, 0.6113, 0.6528, 0.5923, 0.6914, 0.6104, 0.6479, 0.6440, 0.6470, 0.6724, 0.6626, 0.6421, 0.6353, 0.6602, 0.5903, 0.6138, 0.5972, 0.6396,\n 0.6875, 0.5776, 0.5098, 0.7402, 0.6768, 0.5884, 0.6777, 0.5620, 0.6519, 0.6650, 0.6587, 0.5581, 0.6299, 0.6489, 0.6152, 0.6206, 0.6089, 0.6362, 0.6182, 0.5757, 0.6704, 0.5527, 0.5898, 0.5923, 0.5879, 0.6064, 0.5713, 0.6060, 0.6118, 0.6499, 0.5903, 0.6016, 0.6128, 0.6567, 0.5581, 0.6514, 0.6421, 0.6572, 0.6035,\n 0.6885, 0.6147, 0.6450, 0.6440, 0.6396, 0.6738, 0.6797, 0.6509, 0.6558, 0.6582, 0.6187, 0.6338, 0.5879, 0.6807, 0.6396, 0.6094, 0.6992, 0.6665, 0.6470, 0.6885, 0.6377, 0.5444, 0.6191, 0.6797, 0.5591, 0.6499, 0.6538, 0.6113, 0.6084, 0.5903, 0.6328, 0.5942, 0.6528, 0.6055, 0.6836, 0.6479, 0.6309, 0.6182, 0.5938,\n 0.6631, 0.5723, 0.6362, 0.6650, 0.6704, 0.6333, 0.6611, 0.5747, 0.6460, 0.6646, 0.6201, 0.6377, 0.6211, 0.6260, 0.6313, 0.6084, 0.6338, 0.6313, 0.6182, 0.6543, 0.6597, 0.6641, 0.6714, 0.5820, 0.5967, 0.6655, 0.6816, 0.6055, 0.6909, 0.7026, 0.6641, 0.6284, 0.6260, 0.6387, 0.6226, 0.6060, 0.6548, 0.5918, 0.6157,\n 0.6382, 0.6079, 0.6699, 0.6392, 0.6118, 0.6240, 0.6748, 0.6450, 0.5396, 0.6353, 0.6240, 0.6489, 0.6748, 0.6030, 0.6558, 0.6538, 0.6152, 0.6694, 0.6802, 0.6333, 0.6611, 0.6660, 0.6084, 0.5654, 0.6792, 0.6670, 0.5771, 0.6899, 0.6196, 0.6094, 0.6348, 0.6768, 0.6362, 0.6895, 0.6509, 0.6250, 0.6494, 0.6538, 0.6304,\n 0.6357, 0.6313, 0.5859, 0.5122, 0.6235, 0.6177, 0.6162, 0.6074, 0.6743, 0.5591, 0.6143, 0.6182, 0.7417, 0.5474, 0.6641, 0.6030, 0.6245, 0.6333, 0.6929, 0.5566, 0.6694, 0.7158, 0.6011, 0.6260, 0.5835, 0.6577, 0.6069, 0.6724, 0.6719, 0.6479, 0.6367, 0.6421, 0.6719, 0.6348, 0.6602, 0.5728, 0.6880, 0.5391, 0.6484,\n 0.5684, 0.5610, 0.7056, 0.5684, 0.6440, 0.6094, 0.6187, 0.6938, 0.6440, 0.6514, 0.6519, 0.6475, 0.5747, 0.6865, 0.5894, 0.5845, 0.5972, 0.6714, 0.6030, 0.6396, 0.6294, 0.6587, 0.6577, 0.6064, 0.5972, 0.6030, 0.6792, 0.6313, 0.6812, 0.6401, 0.6226, 0.6353, 0.6558, 0.6763, 0.6436, 0.6450, 0.5952, 0.5479, 0.5693,\n 0.6162, 0.5698, 0.6133, 0.6533, 0.6401, 0.6465, 0.6357, 0.6396, 0.5801, 0.7017, 0.6836, 0.6255, 0.5781, 0.6411, 0.6235, 0.6875, 0.6104, 0.5933, 0.6401, 0.6382, 0.6807, 0.6465, 0.6167, 0.5674, 0.5996, 0.6221, 0.6001, 0.6567, 0.6572, 0.6821, 0.6836, 0.6553, 0.6299, 0.5840, 0.7451, 0.5967, 0.6055, 0.6792, 0.6475,\n 0.6401, 0.5444, 0.5908, 0.6050, 0.6084, 0.6851, 0.6274, 0.5962, 0.6309, 0.6616, 0.6499, 0.6050, 0.6245, 0.5850, 0.6426, 0.6021, 0.6968, 0.5996, 0.6763, 0.5703, 0.5928, 0.5981, 0.6431, 0.5962, 0.6831, 0.5967, 0.5684, 0.7559, 0.5811, 0.6475, 0.6069, 0.6426, 0.6265, 0.6133, 0.5747, 0.6709, 0.5610, 0.5547, 0.6890,\n 0.6274, 0.6143, 0.6562, 0.5488, 0.6772, 0.6460, 0.6162, 0.6533, 0.5977, 0.6880, 0.6641, 0.6631, 0.6523, 0.6572, 0.6895, 0.6729, 0.6050, 0.6167, 0.6338, 0.6328, 0.6152, 0.6626, 0.6392, 0.6875, 0.5732, 0.5947, 0.6055, 0.5732, 0.6885, 0.6006, 0.6074, 0.5884, 0.5513, 0.6250, 0.5967, 0.6201, 0.6113, 0.6377, 0.6333,\n 0.6826, 0.7349, 0.6748, 0.6328, 0.6133, 0.6782, 0.6221, 0.5952, 0.6680, 0.5952, 0.6357, 0.6699, 0.5742, 0.5977, 0.6382, 0.7051, 0.6533, 0.6528, 0.6660, 0.5332, 0.5698, 0.6191, 0.6431, 0.6304, 0.6411, 0.6499, 0.6792, 0.6221, 0.5820, 0.6353, 0.6646, 0.6470, 0.6196, 0.6489, 0.6348, 0.6128, 0.5928, 0.6172, 0.6953,\n 0.6558, 0.6250, 0.6680, 0.7056, 0.6533, 0.6104, 0.6523, 0.7104, 0.5957, 0.6455, 0.6509, 0.6582, 0.6255, 0.6479, 0.6548, 0.6890, 0.6504, 0.6392, 0.6226, 0.6235, 0.6699, 0.5874, 0.6665, 0.6738, 0.6191, 0.5952, 0.6133, 0.6172, 0.5996, 0.5679, 0.6377, 0.5942, 0.6187, 0.6235, 0.5649, 0.6660, 0.6284, 0.6416, 0.6216,\n 0.5693, 0.6372, 0.5757, 0.6660, 0.7036, 0.6069, 0.6768, 0.5469, 0.6484, 0.6528, 0.6157, 0.6841, 0.6035, 0.6406, 0.6348, 0.6094, 0.6538, 0.6802, 0.6094, 0.6484, 0.5942, 0.6138, 0.6196, 0.6396, 0.6763, 0.6465, 0.6460, 0.6304, 0.6118, 0.6025, 0.6606, 0.6553, 0.6143, 0.6499, 0.6431, 0.6592, 0.5874, 0.6108, 0.6152,\n 0.6733, 0.5703, 0.5967, 0.6768, 0.5977, 0.6855, 0.6729, 0.6831, 0.6313, 0.6294, 0.6890, 0.6118, 0.6411, 0.6797, 0.6313, 0.6880, 0.6626, 0.6655, 0.6802, 0.6265, 0.6909, 0.6084, 0.6201, 0.6909, 0.6904, 0.6416, 0.5791, 0.5894, 0.5840, 0.5981, 0.6553, 0.6611, 0.6523, 0.6465, 0.6279, 0.6440, 0.6450, 0.6367, 0.6406,\n 0.6211, 0.5928, 0.6807, 0.5854, 0.6045, 0.6567, 0.6021, 0.6152, 0.6279, 0.5767, 0.5854, 0.6514, 0.6055, 0.6362, 0.5532, 0.6982, 0.6338, 0.6396, 0.7065, 0.6289, 0.5981, 0.6396, 0.6055, 0.6245, 0.6777, 0.6309, 0.6694, 0.5728, 0.6387, 0.6294, 0.5532, 0.5923, 0.6582, 0.6450, 0.6631, 0.6660, 0.6377, 0.6714, 0.6221,\n 0.6328, 0.6523, 0.6328, 0.6157, 0.6353, 0.5918, 0.5796, 0.6016, 0.6924, 0.6426, 0.6343, 0.6377, 0.6177, 0.6621, 0.6699, 0.6396, 0.6230, 0.5610, 0.6323, 0.6738, 0.6240, 0.6562, 0.6270, 0.6792, 0.5972, 0.6655, 0.6387, 0.6343, 0.7080, 0.6323, 0.6128, 0.6851, 0.6372, 0.6108, 0.6606, 0.6216, 0.6206, 0.6743, 0.6401,\n 0.6279, 0.6313, 0.6987, 0.6143, 0.5996, 0.5874, 0.6606, 0.5645, 0.6621, 0.6807, 0.5923, 0.6460, 0.6143, 0.6226, 0.6558, 0.5693, 0.6333, 0.6831, 0.6069, 0.6372, 0.5898, 0.6201, 0.6562, 0.6382, 0.6367, 0.6709, 0.6587], dtype=torch.float16), 'perceiver_resampler.layers.3.0.norm2.bias': tensor([-1.8967e-02, -4.8828e-02, 2.4929e-03, -2.4643e-02, -9.1980e-02, 6.2332e-03, 1.9424e-02, -4.8637e-03, 6.5552e-02, -1.2482e-01, 1.4153e-02, -8.7097e-02, -2.4796e-02, -1.6394e-01, 1.4282e-01, -2.5055e-02, 1.8860e-02, -1.0384e-02, -2.4368e-02, -8.7967e-03, -5.2826e-02, 6.0394e-02, 3.7201e-02, 1.8890e-02,\n -2.9358e-02, 6.6471e-04, 1.0266e-01, -8.5571e-02, 1.7731e-02, 2.7847e-03, 1.1218e-01, 1.9141e-01, -2.7710e-02, 6.0883e-02, -5.8508e-04, -5.2246e-02, 5.9601e-02, 1.5472e-02, -2.7924e-02, -5.0323e-02, -6.9389e-03, 1.4685e-01, -4.8981e-03, -4.1199e-02, -8.1299e-02, 3.3051e-02, -9.0561e-03, 1.1299e-02,\n 5.5733e-03, -5.4717e-05, 2.0981e-02, -1.1490e-02, 1.8097e-02, 2.4719e-02, -6.4575e-02, -1.4023e-02, 3.9490e-02, 6.1859e-02, -5.8746e-02, 9.5215e-02, -1.4124e-01, 7.1983e-03, 1.1986e-02, -1.0078e-02, -6.2317e-02, 6.1859e-02, 6.2988e-02, 9.2957e-02, -1.7441e-02, 7.9346e-02, -9.1600e-04, -5.5542e-02,\n 5.2261e-03, -4.6204e-02, -7.0679e-02, 1.0307e-02, -8.5144e-02, -7.5256e-02, 2.1301e-01, 1.2207e-02, -1.7426e-02, 5.6335e-02, 2.2335e-03, -4.1962e-02, 6.0806e-03, 9.3994e-02, -6.0181e-02, -1.3318e-01, -9.2834e-02, -6.0760e-02, 1.5182e-02, -1.0498e-01, -4.8889e-02, -2.5421e-02, -7.5195e-02, -8.2336e-02,\n -8.7463e-02, 1.0246e-02, -2.2598e-02, 7.0496e-02, 2.7893e-02, -8.7830e-02, 4.0375e-02, -1.5625e-01, 3.8261e-03, -5.3635e-03, 7.5928e-02, -2.4719e-02, 1.2744e-01, 2.0950e-02, -9.2651e-02, 3.0308e-03, -2.1286e-02, -1.8201e-01, 5.4260e-02, 7.1411e-02, 5.0537e-02, -2.2095e-02, -3.2257e-02, 7.8630e-04,\n -1.5656e-02, -8.1177e-03, -5.4962e-02, 1.7385e-03, 3.5919e-02, 7.7271e-02, -9.8633e-02, -3.0945e-02, 3.5614e-02, -5.2765e-02, -7.1411e-02, 1.6687e-01, -6.7932e-02, 2.0187e-02, -5.5908e-02, 1.9150e-03, -5.9784e-02, 9.7534e-02, -9.4910e-02, 1.0797e-01, 4.2969e-02, 1.7426e-02, -1.5793e-02, 1.9153e-01,\n -5.7098e-02, 3.0308e-03, 1.7197e-02, -2.0569e-02, 3.5675e-02, 4.7668e-02, 4.8096e-02, 6.9031e-02, -4.3030e-02, -6.3416e-02, 2.5223e-02, -4.9652e-02, 2.0218e-02, 5.7098e-02, -1.0480e-01, -3.2959e-02, -1.7176e-03, 1.5457e-02, -2.2827e-02, 1.2924e-02, 3.4454e-02, -3.5858e-02, -1.0551e-02, -8.2031e-02,\n -1.3603e-02, -3.3569e-03, 4.9042e-02, 3.2837e-02, -3.6133e-02, -5.2214e-05, -1.4626e-02, -1.4246e-01, -5.7220e-02, 9.3628e-02, 4.7546e-02, -7.2693e-02, 1.2952e-01, -7.2861e-03, 3.2196e-02, 4.5319e-02, 4.0039e-02, -1.6327e-02, -5.9387e-02, -2.2491e-02, -4.9194e-02, 2.9678e-02, -8.0505e-02, 6.7749e-02,\n 3.5431e-02, -1.0086e-02, 2.9434e-02, -1.7099e-03, 1.0608e-01, -1.8829e-02, 4.9866e-02, -5.2032e-03, 1.3538e-01, -2.6840e-02, -7.6599e-02, 2.9770e-02, 7.4646e-02, 2.9358e-02, -2.8809e-02, 8.6121e-02, 2.3270e-02, 1.7105e-02, -5.7755e-03, -1.2474e-02, -9.0515e-02, -2.3956e-02, -1.3672e-02, 4.2633e-02,\n 4.1290e-02, -1.2128e-01, -9.5444e-03, 7.8064e-02, -3.9062e-02, -4.7546e-02, -2.3560e-02, 8.4534e-02, -8.7769e-02, 6.5857e-02, 1.1542e-01, 3.9124e-02, 2.8839e-02, -6.8054e-02, 9.1858e-03, 3.0411e-02, -4.0436e-03, 4.7089e-02, -4.0161e-02, 9.8206e-02, 1.0811e-02, -3.2074e-02, 7.4890e-02, -5.8929e-02,\n -4.6539e-02, 9.9182e-02, 6.0913e-02, 1.1298e-01, -2.4048e-02, 1.2833e-02, -1.6129e-02, 1.3074e-01, -2.0355e-02, 1.2323e-01, 7.0740e-02, 1.3110e-01, 1.3580e-02, 2.7451e-02, 2.4902e-02, -5.3101e-02, -2.2774e-03, 1.2372e-01, 9.4727e-02, 2.6718e-02, 9.7427e-03, 1.0162e-01, 2.8275e-02, -3.8940e-02,\n -1.4305e-02, -1.3464e-01, 6.1340e-02, 8.5449e-02, -8.8928e-02, -1.4435e-02, -5.8517e-03, 2.6871e-02, -1.1176e-01, -2.5192e-02, 8.1604e-02, 3.2837e-02, 6.9275e-02, -7.8918e-02, 9.1003e-02, -3.9581e-02, 3.3356e-02, -8.1909e-02, 1.1285e-01, -8.8318e-02, -5.7373e-03, -1.6037e-02, -1.1255e-01, 1.6922e-02,\n -2.0050e-02, 4.1351e-02, -6.2317e-02, 2.4521e-02, -3.4943e-02, 1.2741e-03, 5.7251e-02, 2.3697e-02, -1.0095e-01, 3.3020e-02, -1.6739e-02, 1.1070e-02, 1.5991e-02, 2.6581e-02, -6.3660e-02, -7.6111e-02, 5.7800e-02, -8.3862e-02, -2.0554e-02, 9.4666e-02, 1.7105e-02, 8.5144e-03, 6.5247e-02, 3.2730e-03,\n -4.6295e-02, -9.0881e-02, 7.8430e-02, 2.0068e-01, -1.6495e-02, -8.7769e-02, -1.5466e-01, 7.7148e-02, 1.5144e-02, 6.3992e-04, -9.6985e-02, -5.6061e-02, -3.4882e-02, -8.3984e-02, 2.4246e-02, 1.5289e-02, -5.2856e-02, -4.9667e-03, 1.2244e-01, -1.1395e-01, 1.5640e-02, 3.3051e-02, -5.6946e-02, -7.7209e-02,\n 4.3945e-02, 1.6647e-02, 4.4098e-02, -4.4922e-02, 7.6408e-03, -1.5717e-02, -5.5420e-02, -2.4223e-03, 6.8237e-02, 1.4885e-02, -3.9154e-02, 5.4817e-03, 4.4312e-02, 1.0999e-01, 2.8717e-02, -2.0081e-01, -7.4646e-02, 1.7838e-02, 1.4717e-02, 8.4305e-03, 1.3135e-01, -1.4319e-01, 4.6173e-02, -2.1698e-02,\n 1.0330e-02, 2.0874e-02, 6.8604e-02, -4.0627e-03, -1.4267e-02, 1.0114e-01, 3.1719e-03, -1.7609e-02, 4.1351e-02, -3.7384e-02, -7.5562e-02, 1.0490e-02, 1.9705e-04, -7.7477e-03, 2.1378e-02, -3.1311e-02, -1.1713e-01, -2.3834e-02, 4.6326e-02, 5.7869e-03, -3.5522e-02, 1.6041e-03, 6.4240e-03, 5.9174e-02,\n 2.8629e-03, 2.1500e-02, -1.3817e-02, 1.2543e-02, 1.7059e-02, -3.1681e-03, 3.1319e-03, -4.4739e-02, -2.2217e-02, 3.5095e-02, 1.4966e-01, -2.8122e-02, -8.6182e-02, 5.7556e-02, -6.8970e-02, 3.8818e-02, 2.6550e-02, 4.5395e-03, 1.3232e-01, 5.8807e-02, -8.7891e-03, 3.4046e-03, 2.3392e-02, 1.0596e-01,\n -8.9417e-03, -1.0536e-02, -1.8585e-02, -9.4727e-02, -2.4139e-02, -2.1240e-02, -8.6731e-02, -3.5645e-02, -1.7493e-01, 1.1084e-01, 5.6458e-02, 8.8043e-03, 3.6041e-02, -7.2021e-02, 3.0396e-02, -3.8544e-02, 3.7964e-02, 1.4427e-02, -1.2372e-01, 4.2908e-02, -8.6487e-02, -9.5367e-03, -1.4392e-01, 2.4994e-02,\n 7.6218e-03, -1.2781e-01, 2.1500e-02, 1.4050e-01, -7.1106e-02, -6.8726e-02, 9.1553e-03, -1.7990e-02, 1.3721e-01, -5.6152e-02, -5.0507e-02, 8.4595e-02, 9.6512e-03, -2.2568e-02, -1.3098e-01, -4.5090e-03, -8.5205e-02, 2.5635e-03, 7.6342e-04, 3.5992e-03, 1.0168e-01, 2.9724e-02, -4.1428e-03, 1.2244e-01,\n -4.4403e-02, 6.2988e-02, -4.8706e-02, 2.6627e-02, -9.7580e-03, -4.3854e-02, 1.0547e-01, 5.3253e-02, -4.5197e-02, -4.8615e-02, 8.0139e-02, -4.2694e-02, -1.1856e-02, -2.2781e-02, 2.9926e-03, -1.8054e-01, -2.8076e-02, -1.0352e-01, -6.0120e-03, 1.9608e-03, 1.5967e-01, 4.9469e-02, 3.8269e-02, -6.4697e-02,\n -1.9562e-02, 6.1890e-02, -6.4758e-02, -4.3976e-02, 5.1575e-02, -1.4062e-01, -8.8623e-02, 3.2288e-02, 9.0313e-04, 4.3335e-03, 3.2227e-02, -2.2598e-02, 8.4473e-02, -1.1877e-01, -9.7839e-02, 3.6255e-02, 4.0771e-02, 6.2469e-02, 7.6599e-02, 1.2976e-01, -1.2805e-01, 4.6112e-02, -1.1151e-01, 1.1530e-01,\n 2.3468e-02, 5.5511e-02, -6.6345e-02, 3.2745e-02, 2.9739e-02, 6.3843e-02, -6.1493e-02, 7.5073e-02, -1.0394e-01, 1.8631e-02, 4.7974e-02, 2.6951e-03, 5.1994e-03, 2.9037e-02, -2.4246e-02, -8.9905e-02, 7.7576e-02, -1.8967e-02, 1.0919e-01, -6.9885e-02, 4.9805e-02, -1.2863e-02, -1.1359e-01, 5.3314e-02,\n 5.0316e-03, -1.9257e-02, -5.8075e-02, -1.2268e-01, 2.4887e-02, 3.3112e-02, -5.1758e-02, -3.0869e-02, -5.4688e-02, 9.6252e-02, 1.0872e-03, 2.3819e-02, 5.5511e-02, 2.5845e-03, 1.7029e-02, -2.6978e-02, -3.3783e-02, 2.8519e-02, -3.3295e-02, 8.9722e-03, 9.9609e-02, 4.4434e-02, -1.9470e-02, -7.8735e-02,\n 2.9373e-02, 2.7390e-02, -6.6772e-02, -1.1426e-01, 4.7485e-02, -2.4094e-02, 5.1697e-02, 3.7018e-02, -3.1677e-02, -4.2145e-02, -5.0018e-02, 6.8283e-03, -4.2603e-02, 1.9516e-02, 1.0486e-01, 2.7710e-02, 3.2654e-02, 8.8501e-02, 2.5725e-04, -8.5144e-03, 6.3416e-02, 8.6899e-03, -1.3562e-01, -1.8823e-01,\n 8.3130e-02, 6.0059e-02, 5.2277e-02, -3.9520e-02, 6.6948e-03, -1.8234e-02, -1.4511e-02, 3.9734e-02, 4.4060e-03, -7.9102e-02, 5.9296e-02, 3.0258e-02, 7.7759e-02, -2.3636e-02, -8.2626e-03, -1.0101e-01, -2.1790e-02, -4.8706e-02, 3.3630e-02, -8.9783e-02, -4.7028e-02, -3.8239e-02, 5.1819e-02, -2.1942e-02,\n 2.4002e-02, -1.8921e-02, -1.0841e-02, 1.3354e-01, 2.7893e-02, 1.3989e-01, -9.7473e-02, -6.6895e-02, -5.0049e-02, 4.5929e-02, -2.1622e-02, -8.8257e-02, -3.1204e-03, 3.2715e-02, -1.7014e-02, 4.7989e-03, 8.4167e-02, 4.5380e-02, -5.8990e-02, -1.7624e-02, -3.9642e-02, -7.1960e-02, 4.2999e-02, -1.9501e-02,\n 7.4951e-02, -2.9037e-02, -1.8906e-02, -3.6499e-02, -8.6899e-03, -9.8450e-02, -4.2953e-03, -3.2227e-02, 4.1565e-02, -2.6398e-02, 2.6321e-04, 4.1229e-02, -9.0027e-02, 2.6230e-02, -1.7554e-01, -1.9894e-03, -7.7705e-03, -3.5980e-02, 1.3161e-02, -1.0948e-02, -5.5725e-02, -1.6508e-03, 4.1534e-02, 3.9215e-02,\n -2.3308e-03, -3.2593e-02, 1.1804e-01, 3.2440e-02, 5.4169e-02, -8.0261e-02, 1.8188e-02, 2.5879e-02, -2.8290e-02, -2.9541e-02, 3.5645e-02, -7.6782e-02, 1.2634e-02, -2.9068e-02, -1.5015e-02, -4.8309e-02, -5.2612e-02, -3.6194e-02, 3.5980e-02, 4.6631e-02, 2.7313e-02, 1.0077e-01, 7.2510e-02, 1.0718e-01,\n 6.7078e-02, 9.6313e-02, -1.5778e-02, -4.7760e-02, 1.0040e-01, 1.3928e-01, -6.6101e-02, 4.5532e-02, -1.2680e-02, -1.8829e-02, 8.4152e-03, -4.0192e-02, 1.1816e-01, 5.2223e-03, -1.4880e-01, 7.2815e-02, -3.9154e-02, -6.9763e-02, 4.2236e-02, 3.6888e-03, 3.2227e-02, 9.2407e-02, 7.0618e-02, 4.7424e-02,\n 2.7237e-02, -1.5404e-02, -4.3610e-02, 2.0218e-03, 6.4516e-04, -9.9915e-02, 8.2031e-02, -2.0828e-02, 1.0736e-01, 6.2500e-02, 1.9424e-02, -9.6817e-03, 3.0319e-02, 2.6733e-02, -1.8829e-02, -8.8501e-02, 4.4708e-02, 2.6688e-02, -2.8763e-02, -2.5177e-02, -1.8478e-02, -1.4313e-02, -2.1716e-01, 1.4868e-01,\n -4.3304e-02, 1.9073e-03, -5.4016e-03, -8.4534e-03, 6.9153e-02, 4.3854e-02, -1.0266e-01, 7.9956e-02, -3.2990e-02, 5.9052e-03, 4.9561e-02, -1.0675e-01, 1.5820e-01, 3.0609e-02, -8.4000e-03, 4.5380e-02, -5.0430e-03, 6.2195e-02, -1.4061e-02, -2.2522e-02, -1.5778e-02, 2.1240e-01, 8.9417e-03, 2.6215e-02,\n 1.1658e-01, 8.1299e-02, -7.4341e-02, 5.9967e-02, -7.6416e-02, 4.0955e-02, 3.9597e-03, -4.1962e-02, -1.7105e-02, 1.8250e-02, -1.9028e-02, 2.8580e-02, 1.7102e-01, -6.9397e-02, 2.9892e-02, 4.0863e-02, 8.0200e-02, -4.6722e-02, -5.8105e-02, -2.6245e-02, -7.4402e-02, 9.3323e-02, 1.5100e-01, -7.6965e-02],\n dtype=torch.float16), 'perceiver_resampler.layers.3.0.to_q.weight': tensor([[ 0.0386, -0.0179, -0.0174, ..., 0.0194, -0.0677, 0.0275],\n [-0.0135, -0.0568, -0.0020, ..., -0.0047, 0.0032, -0.0115],\n [ 0.0179, -0.0244, -0.0112, ..., -0.0906, 0.0062, 0.0127],\n ...,\n [-0.0074, 0.0180, -0.0195, ..., -0.0172, -0.0135, 0.0225],\n [-0.0409, 0.0365, 0.0287, ..., -0.0124, 0.0260, 0.0072],\n [-0.0222, 0.0417, 0.0291, ..., 0.0258, 0.0225, -0.0034]], dtype=torch.float16), 'perceiver_resampler.layers.3.0.to_kv.weight': tensor([[ 0.0409, -0.0547, -0.0110, ..., -0.0069, 0.0216, -0.0055],\n [-0.0392, 0.0143, 0.0070, ..., -0.0282, -0.0337, 0.0356],\n [ 0.0481, 0.0353, 0.0017, ..., -0.0263, 0.0644, 0.0185],\n ...,\n [-0.0333, -0.0234, -0.0199, ..., 0.0106, 0.0590, 0.0106],\n [-0.0434, 0.0033, 0.0183, ..., -0.0162, -0.0135, 0.0409],\n [-0.0199, 0.0087, -0.0107, ..., -0.0157, 0.0091, 0.0186]], dtype=torch.float16), 'perceiver_resampler.layers.3.0.to_out.weight': tensor([[-0.0442, 0.0138, -0.0303, ..., 0.0286, 0.0154, 0.0099],\n [-0.0356, 0.0436, -0.0408, ..., -0.0613, -0.0045, -0.0146],\n [-0.0327, -0.0137, -0.0378, ..., -0.0595, 0.0145, -0.0474],\n ...,\n [ 0.0557, 0.0381, 0.0774, ..., 0.0527, 0.0424, 0.0166],\n [-0.0141, 0.0557, -0.0219, ..., -0.0053, -0.0038, -0.0423],\n [ 0.0601, -0.0295, -0.0213, ..., 0.0404, 0.0591, 0.0320]], dtype=torch.float16), 'perceiver_resampler.layers.3.1.0.weight': tensor([0.7939, 0.8066, 0.7764, 0.8760, 0.7505, 0.8076, 0.8276, 0.7676, 0.7490, 0.7666, 0.8066, 0.9062, 0.7769, 0.7627, 0.7949, 0.7734, 0.8281, 0.7832, 0.7686, 0.7842, 0.8447, 0.7720, 0.7622, 0.8037, 0.8032, 0.8325, 0.7046, 0.8257, 0.7290, 0.7837, 0.7441, 0.7129, 0.8174, 0.8452, 0.8096, 0.8208, 0.7349, 0.8369, 0.8481,\n 0.8662, 0.8325, 0.7354, 0.8208, 0.7314, 0.7881, 0.7817, 0.7935, 0.9043, 0.7676, 0.7964, 0.7954, 0.7739, 0.8086, 0.8433, 0.6953, 0.8843, 0.7881, 0.8706, 0.7896, 0.7979, 0.7964, 0.8198, 0.8584, 0.7476, 0.8584, 0.8032, 0.7793, 0.8750, 0.7705, 0.9009, 0.7729, 0.8135, 0.8438, 0.8022, 0.7559, 0.7788, 0.8467, 0.8989,\n 0.8740, 0.8301, 0.8008, 0.7295, 0.8350, 0.7964, 0.8052, 0.9111, 0.7412, 0.8755, 0.9072, 0.8945, 0.9316, 0.7090, 0.8198, 0.7954, 0.8081, 0.8569, 0.7065, 0.7842, 0.7310, 0.7119, 0.8320, 0.7026, 0.7876, 0.8325, 0.7632, 0.8174, 0.8672, 0.8457, 0.9023, 0.8335, 0.7383, 0.7603, 0.7832, 0.7642, 0.8345, 0.8184, 0.8638,\n 0.8174, 0.8120, 0.7832, 0.7900, 0.8267, 0.8521, 0.8296, 0.7793, 0.8159, 0.7412, 0.8086, 0.7607, 0.8774, 0.7847, 0.7534, 0.7588, 0.8579, 0.7949, 0.7715, 0.8984, 0.7642, 0.6855, 0.7964, 0.7637, 0.7847, 0.8301, 0.7881, 0.8618, 0.8281, 0.7998, 0.8120, 0.7852, 0.7661, 0.8174, 0.8535, 0.9102, 0.7627, 0.7759, 0.8872,\n 0.8560, 0.7827, 0.7153, 0.8784, 0.8521, 0.7524, 0.7993, 0.7627, 0.7891, 0.8257, 0.8955, 0.7725, 0.7832, 0.8140, 0.7979, 0.8101, 0.8149, 0.8281, 0.8071, 0.9209, 0.7788, 0.7754, 0.7793, 0.9077, 0.7480, 0.8115, 0.8301, 0.8438, 0.7412, 0.8589, 0.7720, 0.7144, 0.8188, 0.8457, 0.8379, 0.8262, 0.8193, 0.8042, 0.8071,\n 0.8130, 0.8022, 0.8013, 0.8345, 0.7959, 0.6973, 0.7422, 0.7949, 0.9048, 0.8325, 0.7915, 0.8774, 0.7363, 0.8140, 0.7876, 0.8154, 0.7661, 0.7524, 0.8828, 0.7168, 0.8818, 0.7471, 0.7129, 0.8345, 0.8550, 0.7271, 0.8550, 0.8853, 0.7632, 0.7661, 0.8130, 0.9614, 0.8457, 0.7734, 0.7832, 0.8267, 0.8477, 0.8091, 0.8237,\n 0.8550, 0.7788, 0.7881, 0.8608, 0.9141, 0.7358, 0.7622, 0.7759, 0.7769, 0.7456, 0.8750, 0.8730, 0.8174, 0.8086, 0.7920, 0.8672, 0.6958, 0.7300, 0.7949, 0.8271, 0.8687, 0.8242, 0.8369, 0.7817, 0.7407, 0.8135, 0.8770, 0.8638, 0.8550, 0.7988, 0.7979, 0.7334, 0.8110, 0.7563, 0.7480, 0.8286, 0.7998, 0.8940, 0.8447,\n 0.7896, 0.8711, 0.8237, 0.9590, 0.7896, 0.8335, 0.8608, 0.8320, 0.8013, 0.9058, 0.9106, 0.8213, 0.7646, 0.6792, 0.7812, 0.8276, 0.8320, 0.7866, 0.8779, 0.7925, 0.8389, 0.8062, 0.8491, 0.7246, 0.7803, 0.7866, 0.7969, 0.8389, 0.8037, 0.7949, 0.7930, 0.8408, 0.7056, 0.7949, 0.8042, 0.8018, 0.8193, 0.8232, 0.7681,\n 0.7446, 0.7832, 0.9106, 0.8008, 0.8442, 0.7715, 0.7803, 0.8042, 0.8472, 0.7788, 0.8003, 0.8960, 0.9336, 0.7319, 0.8242, 0.8091, 0.7769, 0.8311, 1.0156, 0.7456, 0.8066, 0.8184, 0.7607, 0.7827, 0.8364, 0.7886, 0.8071, 0.8584, 0.7920, 0.7939, 0.8154, 0.8105, 0.8208, 0.8481, 0.7900, 0.7998, 0.8213, 0.6724, 0.8579,\n 0.7427, 0.8965, 0.8223, 0.8018, 0.7988, 0.7808, 0.7852, 0.8735, 0.8335, 0.8179, 0.7896, 0.7705, 0.8271, 0.8301, 0.6948, 0.8213, 0.7754, 0.7939, 0.7842, 0.8535, 0.8223, 0.7925, 0.8779, 0.7759, 0.8120, 0.8452, 0.8799, 0.7617, 0.8457, 0.8843, 0.8101, 0.7944, 0.8242, 0.7842, 0.8398, 0.8335, 0.7402, 0.8394, 0.8242,\n 0.8042, 0.7847, 0.8091, 0.7812, 0.9048, 0.8169, 0.7769, 0.8086, 0.6714, 0.7998, 0.8447, 0.7935, 0.7930, 0.8066, 0.8599, 0.8130, 0.7983, 0.8018, 0.8110, 0.7812, 0.8081, 0.7871, 0.8315, 0.7734, 0.8115, 0.7061, 0.8408, 0.9087, 0.8257, 0.8096, 0.8574, 0.8423, 0.7832, 0.8091, 0.7949, 0.8652, 0.6909, 0.8472, 0.7729,\n 0.8169, 0.6934, 0.8086, 0.7812, 0.7886, 0.7666, 0.8994, 0.7490, 0.8350, 0.7495, 0.8008, 0.9414, 0.7593, 0.7305, 0.7822, 0.7808, 0.8140, 0.8613, 0.8110, 0.7144, 0.8257, 0.8223, 0.8066, 0.7349, 0.7734, 0.7979, 0.8135, 0.8052, 0.6875, 0.8008, 0.7808, 0.8198, 0.7495, 0.7603, 0.7524, 0.8008, 0.7290, 0.8115, 0.8101,\n 0.8286, 0.7886, 0.8423, 0.7266, 0.8086, 0.7598, 0.7622, 0.7729, 0.8018, 0.8550, 0.8589, 0.7188, 0.7803, 0.8135, 0.8784, 0.8110, 0.7529, 0.6890, 0.7725, 0.7578, 0.8350, 0.7456, 0.8887, 0.8149, 0.7173, 0.7144, 0.7720, 0.8169, 0.8721, 0.7939, 0.9194, 0.7891, 0.7524, 0.8608, 0.7915, 0.7485, 0.7563, 0.8999, 0.8579,\n 0.8340, 0.7598, 0.7964, 0.8828, 0.8335, 0.7261, 0.8159, 0.8438, 0.8008, 0.8350, 0.8018, 0.8486, 0.8350, 0.7852, 0.8540, 0.7803, 0.7710, 0.6899, 0.8599, 0.8970, 0.7617, 0.7920, 0.8315, 0.7949, 0.8218, 0.8364, 0.8066, 0.7490, 0.7861, 0.8042, 0.7935, 0.8149, 0.8018, 0.8174, 0.7900, 0.7910, 0.7729, 0.6982, 0.8320,\n 0.7837, 0.7910, 0.9180, 0.9229, 0.8364, 0.8423, 0.8291, 0.8076, 0.7500, 0.9785, 0.9004, 0.8301, 0.8242, 0.8032, 0.7769, 0.9438, 0.8218, 0.8271, 0.7983, 0.8223, 0.7104, 0.8027, 0.7720, 0.9150, 0.8491, 0.8354, 0.7944, 0.8618, 0.8110, 0.7695, 0.9023, 0.8193, 0.7754, 0.7681, 0.8228, 0.8145, 0.7744, 0.7900, 0.7744,\n 0.6826, 0.7847, 0.7627, 0.8550, 0.8345, 0.7925, 0.8354, 0.7583, 0.8267, 0.8311, 0.7051, 0.7993, 0.8242, 0.7915, 0.9707, 0.7183, 0.8057, 0.8252, 0.7378, 0.8022, 0.7964, 0.7505, 0.7710, 0.7627, 0.7773, 0.7988, 0.8066, 0.8335, 0.7529, 0.7925, 0.8179, 0.7935, 0.8032, 0.7544, 0.8032, 0.8716, 0.7837, 0.8452, 0.8262,\n 0.8359, 0.8335, 0.8135, 0.8276, 0.7690, 0.9077, 0.8252, 0.8135, 0.8462, 0.8276, 0.8198, 0.7876, 0.8467, 0.9263, 0.7148, 0.7339, 0.8135, 0.7827, 0.8213, 0.8428, 0.9761, 0.7744, 0.7842, 0.8350, 0.8232, 0.8716, 0.6172, 0.7998, 0.7964, 0.8120, 0.7773, 0.8770, 0.8013, 0.8193, 0.7896, 0.8413, 0.8066, 0.7749, 0.8022,\n 0.8755, 0.8125, 0.8447, 0.7676, 0.6719, 0.8105, 0.7412, 0.8184, 0.7754, 0.7358, 0.7607, 0.8350, 0.8257, 0.7603, 0.7915, 0.7681, 0.8604, 0.8145, 0.7349, 0.7876, 0.8604, 0.7139, 0.8408, 0.8823, 0.7759, 0.8765, 0.7842, 0.7886, 0.8643, 0.7529, 0.6855, 0.7812, 0.7832, 0.8057, 0.8081, 0.8115, 0.7383, 0.7979, 0.7832,\n 0.7681, 0.7925, 0.8418, 0.7974, 0.7905, 0.8359, 0.8125, 0.7886, 0.8018, 0.8467, 0.7876, 0.8340, 0.7979, 0.8516, 0.8320, 0.8247, 0.7363, 0.7954, 0.8159, 0.8101, 0.8027, 0.7881, 0.8169, 0.7954, 0.7866, 0.8633, 0.8110, 0.8354, 0.8706, 0.9233, 0.7993, 0.8662, 0.8218, 0.7900, 0.8125, 0.7515, 0.7275, 0.8135, 0.7422,\n 0.7905, 0.7998, 0.8218, 0.8022, 0.7769, 0.7515, 0.8462, 0.7627, 0.7817, 0.8320, 0.8018, 0.7568, 0.8164, 0.7871, 0.8149, 0.7832, 0.8706, 0.8613, 0.7964, 0.8052, 0.7808, 0.7979, 0.8257, 0.8286, 0.6670, 0.7617, 0.7734], dtype=torch.float16), 'perceiver_resampler.layers.3.1.0.bias': tensor([ 2.0935e-02, -1.8616e-02, -2.5085e-02, 1.6907e-01, 3.4161e-03, 5.3192e-02, 2.9678e-02, 3.2898e-02, 4.4495e-02, -5.2734e-02, -1.4209e-01, 2.0239e-01, -7.6637e-03, 1.5053e-02, -7.4310e-03, -1.1926e-01, 8.3313e-02, -9.4116e-02, -9.4971e-02, -9.0881e-02, 8.5388e-02, -5.5313e-03, 9.6191e-02, -1.4160e-01,\n 1.0907e-01, -1.0693e-01, 7.7393e-02, 1.1157e-01, -4.5013e-02, -1.3806e-01, 1.8997e-02, 4.0466e-02, 1.3660e-01, -1.7102e-01, -1.3928e-01, -1.2732e-01, 1.0895e-01, -1.0358e-01, 1.1572e-01, 1.4954e-01, -2.7054e-02, 4.4464e-02, 7.2266e-02, -6.6833e-02, -5.8075e-02, 1.2720e-01, 4.7150e-03, -1.2866e-01,\n -7.5928e-02, 4.6936e-02, -2.0981e-02, -1.9241e-02, -2.1530e-02, -8.0688e-02, -8.8867e-02, -1.2445e-01, 6.6833e-02, -1.6846e-01, 4.9194e-02, 3.0914e-02, 7.1106e-02, -1.0779e-01, -9.3689e-02, -3.5004e-02, 1.4478e-01, -9.0088e-02, 9.0027e-03, -1.0767e-01, -5.7861e-02, -1.2939e-01, 6.4636e-02, 4.7516e-02,\n -5.6427e-02, 9.2346e-02, -6.5552e-02, 1.5747e-02, 7.3669e-02, 1.6150e-01, -8.4290e-02, -1.3062e-01, -6.9580e-02, 1.1584e-01, -1.3257e-01, 1.1787e-02, -1.1115e-01, -1.6699e-01, -3.3905e-02, 1.5125e-01, 2.0837e-01, 1.7627e-01, -2.3047e-01, -7.3975e-02, -1.2976e-01, 1.0796e-02, 1.6174e-02, 1.4795e-01,\n -8.9050e-02, 1.2122e-01, 1.1438e-01, -3.6407e-02, -6.9702e-02, -1.2415e-01, -2.4399e-02, 8.3618e-02, -1.2213e-01, -9.8755e-02, -1.2915e-01, 9.6558e-02, -1.6235e-01, -1.2207e-01, -8.0566e-03, 5.2338e-03, 1.4282e-02, 3.9581e-02, -1.2036e-01, 7.0923e-02, -1.1670e-01, 6.5369e-02, -3.5400e-02, 1.0901e-01,\n 9.6191e-02, 1.1627e-01, 9.7961e-02, 1.5332e-01, 5.6610e-02, -1.0767e-01, -4.3884e-02, 1.3818e-01, 1.1505e-01, 1.9946e-01, -7.7881e-02, -2.5085e-02, -6.7139e-03, -1.9446e-01, 8.2626e-03, 3.9215e-02, 1.7578e-01, -5.6824e-02, -5.2948e-02, 1.6754e-02, -1.1024e-02, -6.2683e-02, -1.1237e-01, -4.0649e-02,\n 1.5002e-01, 5.0201e-02, -3.6285e-02, 8.5388e-02, 1.7166e-02, 2.3941e-02, -6.7017e-02, -1.8762e-01, 1.8359e-01, 1.5222e-01, 5.8960e-02, 1.6162e-01, -1.9202e-01, 1.0216e-02, -7.1350e-02, 1.7224e-01, 1.3977e-01, 1.1047e-01, -1.2238e-02, 1.2494e-01, -1.6891e-02, 9.9854e-02, 1.5466e-01, -2.8687e-02,\n -5.5054e-02, -1.6040e-01, 4.9896e-02, -8.8684e-02, -1.0870e-01, -2.1801e-03, 1.2445e-01, 2.0544e-01, 5.6671e-02, 5.7173e-04, -1.4473e-02, 2.3071e-01, 8.1238e-02, -4.4830e-02, -1.5051e-01, -1.5149e-01, 3.5492e-02, 1.5503e-01, -1.1310e-01, 1.0675e-01, -6.1676e-02, -4.9530e-02, 6.2370e-03, -2.7267e-02,\n -8.3130e-02, 1.4185e-01, 1.2598e-01, -8.0139e-02, -9.5337e-02, 8.4045e-02, -2.8503e-02, -2.6993e-02, 5.4596e-02, -8.7524e-02, -5.3596e-03, -1.8542e-01, -9.9426e-02, 4.8920e-02, 1.0791e-01, 1.2012e-01, -4.8401e-02, -6.0120e-02, -8.6243e-02, 8.3237e-03, -4.0833e-02, 1.2708e-01, -7.8674e-02, -1.1542e-01,\n 5.4993e-02, -7.3425e-02, 2.3712e-02, -1.5454e-01, 6.3660e-02, 1.1157e-01, 1.0431e-01, -1.7822e-02, -2.1610e-03, -9.1370e-02, -2.1936e-01, -6.3599e-02, 7.3975e-02, -7.2510e-02, -5.3375e-02, -1.9922e-01, -1.2097e-01, -1.1768e-01, 8.0566e-02, 2.7218e-03, 3.5973e-03, 1.8958e-01, -1.9543e-01, -8.5571e-02,\n 2.3010e-02, 4.7241e-02, -1.1127e-01, 2.4307e-04, 1.5698e-01, -1.0242e-01, -1.2177e-01, -2.2568e-02, 1.0254e-01, -1.1499e-01, 7.5562e-02, 5.2032e-02, -1.0242e-01, -9.9365e-02, -1.7700e-01, 4.1962e-02, 1.4307e-01, -2.7008e-02, 7.2632e-02, -5.8472e-02, 1.9055e-01, 6.1035e-03, -2.2070e-01, -2.7206e-02,\n 1.1438e-01, -1.1401e-01, -6.9275e-02, -4.4983e-02, -4.9591e-02, -9.6130e-02, -1.0352e-01, -2.0801e-01, 1.1029e-01, 9.3384e-02, -1.8469e-01, 9.5703e-02, -2.4048e-01, -9.9426e-02, -4.7974e-02, 2.0398e-01, 2.2919e-02, 1.5430e-01, -1.9434e-01, 1.7371e-01, 1.8091e-01, -4.6204e-02, -5.7343e-02, 2.9724e-02,\n -1.4917e-01, -7.0251e-02, 6.3110e-02, -1.7322e-01, 9.3811e-02, 1.0046e-01, 1.0822e-01, -1.3245e-01, -5.8293e-05, -8.0811e-02, -1.6800e-02, 9.6863e-02, 6.6650e-02, 1.3196e-01, -3.9307e-02, 6.7444e-02, -3.0380e-02, -2.0897e-04, 1.8930e-03, -1.7471e-02, 2.6369e-04, 1.3464e-01, -1.0950e-01, 8.6853e-02,\n -5.0598e-02, -1.0236e-01, -1.8018e-01, -6.1432e-02, 1.2292e-01, -9.4070e-03, 1.0175e-01, -7.6965e-02, -1.5881e-01, 1.1230e-01, 2.2903e-02, 2.1265e-01, 1.7969e-01, -4.0283e-02, -8.5266e-02, 7.8796e-02, -1.1072e-01, -8.5144e-02, -1.7957e-01, 2.8488e-02, -8.6121e-02, -1.0040e-02, -5.1758e-02, -4.4525e-02,\n 1.4075e-01, -3.6278e-03, -1.0614e-01, 1.0364e-01, 6.0699e-02, 1.2646e-01, 1.2500e-01, 1.5839e-02, -2.5513e-02, -1.4026e-01, -2.3079e-03, -1.4929e-01, 4.2801e-03, 5.2643e-02, -6.3416e-02, -2.2858e-02, 1.8665e-01, 1.0962e-01, -8.4229e-02, 1.3086e-01, -1.0475e-02, -3.4695e-03, -8.2520e-02, 2.0660e-02,\n -1.4062e-01, 2.9922e-02, 1.4381e-02, -1.1554e-01, 2.8732e-02, 4.5868e-02, -1.3367e-01, -3.4363e-02, -4.5837e-02, 5.5084e-02, 1.7041e-01, -4.8309e-02, -2.6150e-03, 2.2986e-01, 8.4290e-02, -4.5837e-02, 7.6904e-02, 1.2164e-01, 1.8112e-02, 9.0454e-02, 1.6467e-01, -3.1082e-02, 2.0493e-02, -1.1719e-01,\n -3.3417e-03, -9.9243e-02, 1.1243e-01, -5.4504e-02, -1.1371e-01, 9.1858e-02, 5.6915e-02, 1.4026e-01, 1.7102e-01, 1.2070e-02, -1.5552e-01, -1.1414e-01, -7.3242e-02, -1.1406e-02, -5.3833e-02, -1.4246e-01, -6.2561e-02, -1.2756e-01, -8.8806e-02, -1.8091e-01, 1.5967e-01, 1.0651e-01, 9.8145e-02, 2.5833e-02,\n 1.4626e-02, 7.1716e-02, -1.3843e-01, 1.8884e-01, -1.2463e-01, 1.6907e-01, 1.6504e-01, -1.3977e-02, 1.1987e-01, -1.5112e-01, -1.0071e-01, -1.7090e-01, -1.3269e-01, 7.2937e-02, 1.1676e-01, 8.2520e-02, 1.3367e-01, -1.5894e-01, -5.7556e-02, -1.8408e-01, 1.0461e-01, 5.3589e-02, -8.3435e-02, -4.4464e-02,\n -1.1511e-01, -4.0222e-02, 1.8225e-01, -1.2549e-01, -3.7720e-02, 1.2939e-01, -9.4788e-02, -9.8450e-02, -1.5857e-01, -6.0852e-02, -8.9844e-02, 1.0691e-03, 2.1866e-02, 6.2103e-02, 1.1145e-01, 1.1711e-02, -8.4351e-02, 1.3782e-01, 3.6926e-03, 4.7729e-02, 5.4871e-02, 3.8055e-02, -1.7981e-01, -1.8091e-01,\n -5.6122e-02, -5.8990e-02, 7.8247e-02, -9.8755e-02, 1.1060e-01, -6.7810e-02, 1.0156e-01, 2.3972e-02, 1.4453e-01, -4.6448e-02, -1.5747e-02, 8.4839e-02, 1.1237e-01, 8.7402e-02, 1.1078e-01, 6.0577e-03, 1.0986e-01, -3.5492e-02, 1.9730e-02, 2.0477e-02, -3.4729e-02, -1.0748e-01, -4.5746e-02, -1.2497e-02,\n -9.7046e-02, -7.5439e-02, 6.8909e-02, -2.4597e-02, 1.2158e-01, -2.8900e-02, -1.6663e-02, -1.2772e-02, -9.9792e-02, 2.8122e-02, -2.3779e-01, -7.8308e-02, 7.6599e-02, -2.9633e-02, 8.2932e-03, -1.6089e-01, -1.3525e-01, -1.1182e-01, -2.1289e-01, -2.1042e-02, -2.2964e-02, -1.5564e-01, 1.3049e-01, -4.9072e-02,\n 6.5727e-03, -1.8115e-01, 1.6101e-01, 1.0663e-01, -8.0627e-02, -4.9057e-03, 1.7920e-01, -1.3525e-01, -8.6731e-02, -1.3452e-01, -1.4783e-01, 1.8848e-01, -9.5032e-02, -6.4270e-02, -1.2561e-01, 5.9143e-02, -1.1578e-01, 1.0681e-01, 6.9824e-02, 4.5868e-02, 7.1655e-02, -1.4319e-01, 1.8567e-01, 1.1438e-01,\n 3.5763e-03, 1.2866e-01, 7.7454e-02, 8.0750e-02, -1.0626e-01, 7.5836e-03, -7.4768e-02, 5.2124e-02, -6.2073e-02, 1.1243e-01, -3.6194e-02, -1.3025e-01, -4.8706e-02, 1.5649e-01, 3.8696e-02, -8.8928e-02, -4.7455e-02, -1.1664e-01, -5.3894e-02, -2.6978e-02, -1.6626e-01, -1.8298e-01, -9.5886e-02, 1.3232e-01,\n 5.4016e-02, 6.8298e-02, -1.1755e-01, 2.1643e-01, -1.8286e-01, 1.6626e-01, -1.1670e-01, -3.0098e-03, -7.0129e-02, 2.1484e-01, 3.3630e-02, -5.5176e-02, 5.2673e-02, 4.8401e-02, 6.7688e-02, 5.5603e-02, -5.8784e-03, -1.9482e-01, -8.8989e-02, 8.0750e-02, 7.0068e-02, -1.0779e-01, 1.1658e-01, 6.7871e-02,\n -1.7798e-01, -7.4402e-02, 2.4166e-03, 5.0468e-03, 8.8013e-02, 1.8311e-01, 6.3232e-02, -1.3321e-02, 1.2671e-01, -2.7359e-02, -9.9426e-02, 8.7158e-02, -1.2421e-01, 6.5125e-02, -1.3904e-01, 6.7993e-02, -9.9487e-02, 7.6355e-02, 1.0046e-01, 8.9111e-03, 9.2224e-02, 1.0742e-01, -1.4931e-02, 2.1277e-01,\n -1.1157e-01, 2.4811e-02, 1.5845e-01, 6.5308e-02, -9.2712e-02, -2.1271e-02, -2.0782e-02, -9.5276e-02, -2.6062e-02, 3.7720e-02, 1.1663e-03, -2.9495e-02, -1.2537e-01, 8.7341e-02, -7.4829e-02, 1.1810e-01, 2.9556e-02, -1.2225e-01, 5.0903e-02, -1.6647e-02, 7.6660e-02, -1.1646e-01, -1.6724e-01, -9.6680e-02,\n 1.4236e-02, 1.3000e-01, -5.6114e-03, 2.8198e-02, -1.1169e-01, 1.4282e-01, -3.8666e-02, 6.4148e-02, -4.7028e-02, 1.5308e-01, -1.5723e-01, 9.4055e-02, 9.2651e-02, -2.4292e-01, 1.0292e-02, 9.3613e-03, -8.8272e-03, 1.1206e-01, -7.3814e-03, 1.0876e-01, 2.3450e-01, -9.9304e-02, 1.7746e-02, -9.2468e-02,\n -1.6309e-01, 1.3513e-01, 2.4994e-02, 2.5543e-02, -6.9763e-02, 6.3477e-02, -1.9026e-03, 1.4893e-01, -5.9570e-02, -3.5187e-02, 1.1505e-02, 1.3135e-01, -8.9600e-02, 1.2520e-02, -9.3079e-02, 1.3562e-01, 1.1377e-01, 1.2854e-01, 1.1096e-01, -4.0436e-02, 2.0660e-02, -3.6163e-02, -7.6660e-02, -4.8431e-02,\n 6.9580e-02, 5.9143e-02, 1.4734e-01, 1.2537e-01, 1.0596e-01, -7.2327e-02, -5.0415e-02, -3.4302e-02, -2.1027e-02, -1.1511e-01, -3.7872e-02, 1.7004e-01, 2.8900e-02, -7.6355e-02, 1.0864e-01, 2.2537e-02, 1.9946e-01, -6.2469e-02, 1.1273e-01, -8.3237e-03, 3.3081e-02, 5.1117e-02, 8.7341e-02, 6.5689e-03,\n 7.0129e-02, -2.4612e-02, 7.6416e-02, 4.3121e-02, -7.3486e-02, 2.0111e-02, 6.5369e-02, 3.7842e-02, -1.1719e-01, -6.1378e-03, 1.2250e-01, 1.2720e-01, -6.7520e-03, -2.9556e-02, -4.4952e-02, 6.5247e-02, -2.9251e-02, 1.2756e-01, 7.9102e-02, 7.5134e-02, 1.0388e-01, -7.3792e-02, 1.5251e-02, 1.6541e-02,\n 1.3037e-01, 1.6541e-01, 3.5278e-02, -1.0889e-01, -9.5978e-03, -1.1548e-01, -7.1678e-03, -1.9336e-01, 1.0693e-01, -1.7310e-01, -1.4001e-01, 2.0764e-01, -9.2346e-02, 2.0251e-01, -1.0614e-01, 6.3538e-02, 9.1431e-02, 9.4910e-02, 9.1309e-02, 1.1444e-01, -3.6438e-02, -5.8716e-02, -6.0059e-02, -4.2084e-02,\n -3.2959e-03, 8.0872e-02, 3.9139e-03, -1.1554e-01, -6.6528e-02, 8.5266e-02, -8.5449e-02, 1.5533e-02, -7.8857e-02, -8.8440e-02, 8.4106e-02, 1.7456e-01, -5.4688e-02, 1.1810e-01, -1.7346e-01, -1.2195e-01, -8.6426e-02, 9.3994e-02, -6.5063e-02, 4.5135e-02, 1.0419e-01, 6.5804e-03, -1.2108e-02, 2.4094e-02],\n dtype=torch.float16), 'perceiver_resampler.layers.3.1.1.weight': tensor([[-0.0322, -0.0437, -0.0197, ..., -0.0043, 0.0517, -0.0419],\n [ 0.0536, 0.0302, -0.0020, ..., 0.0126, 0.0499, -0.0424],\n [-0.0124, -0.0222, 0.0005, ..., 0.0091, -0.0066, -0.0098],\n ...,\n [-0.0009, 0.0446, 0.0613, ..., -0.0051, 0.0278, 0.0296],\n [ 0.0334, 0.0059, -0.0445, ..., 0.0316, 0.0006, -0.0243],\n [ 0.0090, 0.0136, 0.0447, ..., 0.0454, -0.0065, 0.0046]], dtype=torch.float16), 'perceiver_resampler.layers.3.1.3.weight': tensor([[-0.0043, -0.0175, 0.0053, ..., 0.0230, -0.0069, 0.0076],\n [-0.0277, -0.0464, 0.0064, ..., -0.0082, -0.0087, 0.0188],\n [ 0.0004, -0.0076, -0.0089, ..., 0.0115, 0.0164, 0.0320],\n ...,\n [ 0.0355, 0.0283, -0.0493, ..., 0.0159, -0.0205, 0.0029],\n [ 0.0559, -0.0486, -0.0224, ..., 0.0228, 0.0062, -0.0037],\n [-0.0226, 0.0041, 0.0044, ..., -0.0124, -0.0036, 0.0327]], dtype=torch.float16)}, 'ip_adapter': {'0.to_q_lora.down.weight': tensor([[-0.0083, -0.0340, -0.0008, ..., -0.0457, 0.0269, 0.0078],\n [-0.0006, 0.0085, -0.0652, ..., 0.0388, -0.0289, 0.0378],\n [ 0.0127, -0.0153, -0.0151, ..., -0.0001, 0.0084, 0.0513],\n ...,\n [-0.0022, 0.0599, -0.0169, ..., 0.0137, -0.0083, -0.0359],\n [ 0.0024, -0.0182, 0.0214, ..., 0.0252, -0.0084, -0.0355],\n [ 0.0003, 0.0338, 0.0332, ..., -0.0763, -0.0269, 0.0018]], dtype=torch.float16), '0.to_q_lora.up.weight': tensor([[ 0.0347, 0.0541, 0.0391, ..., 0.0062, 0.0078, 0.0328],\n [ 0.0497, 0.0592, 0.0225, ..., 0.0158, -0.1052, 0.0530],\n [ 0.0105, 0.0356, 0.0318, ..., 0.0231, -0.0152, -0.0042],\n ...,\n [-0.0160, 0.0340, 0.0046, ..., -0.0273, 0.0198, -0.0226],\n [ 0.0020, 0.0014, 0.0014, ..., -0.0037, 0.0035, 0.0080],\n [ 0.0063, 0.0360, -0.0784, ..., 0.0332, -0.0166, 0.0634]], dtype=torch.float16), '0.to_k_lora.down.weight': tensor([[-0.0357, 0.0169, 0.0457, ..., 0.0352, -0.0088, 0.0009],\n [ 0.0182, 0.0141, 0.0091, ..., -0.0181, -0.0045, 0.0092],\n [-0.0131, -0.0286, 0.0040, ..., 0.0561, -0.0492, 0.0029],\n ...,\n [ 0.0090, -0.0320, -0.0104, ..., -0.0110, 0.0381, 0.0427],\n [ 0.0077, -0.0180, 0.0090, ..., 0.0033, 0.0544, 0.0151],\n [ 0.0155, -0.0710, -0.0267, ..., -0.0367, -0.0011, -0.0070]], dtype=torch.float16), '0.to_k_lora.up.weight': tensor([[ 0.0473, 0.0066, 0.0140, ..., 0.0343, -0.0673, 0.0008],\n [ 0.0100, 0.0220, 0.0316, ..., -0.0334, 0.0222, -0.0088],\n [ 0.0765, -0.0349, -0.0074, ..., -0.0158, 0.0199, -0.0183],\n ...,\n [-0.0395, 0.0088, -0.0113, ..., 0.0056, 0.0349, -0.0163],\n [ 0.0003, 0.0099, -0.0176, ..., 0.0039, 0.0116, -0.0260],\n [-0.0438, -0.0121, -0.0238, ..., 0.0091, 0.0506, -0.0126]], dtype=torch.float16), '0.to_v_lora.down.weight': tensor([[-0.0158, 0.0020, 0.0022, ..., 0.0025, -0.0108, 0.0253],\n [ 0.0041, 0.0333, 0.0105, ..., -0.0426, -0.0345, 0.0067],\n [-0.0217, 0.0132, 0.0196, ..., 0.0262, -0.0340, 0.0329],\n ...,\n [ 0.0195, -0.0055, -0.0214, ..., -0.0635, -0.0191, 0.0083],\n [ 0.0293, -0.0016, -0.0177, ..., 0.0546, 0.0313, -0.0203],\n [ 0.0094, -0.0258, 0.0036, ..., -0.0187, -0.0152, 0.0160]], dtype=torch.float16), '0.to_v_lora.up.weight': tensor([[ 0.0146, -0.0312, -0.0091, ..., 0.0098, -0.0402, -0.0022],\n [-0.0155, -0.0289, 0.0009, ..., -0.0085, 0.0067, -0.0488],\n [ 0.0150, 0.0123, -0.0187, ..., -0.0140, 0.0169, 0.0057],\n ...,\n [-0.0355, 0.0093, -0.0113, ..., -0.0174, -0.0232, 0.0183],\n [ 0.0022, -0.0271, -0.0435, ..., -0.0404, 0.0113, 0.0006],\n [ 0.0194, 0.0343, 0.0051, ..., -0.0091, -0.0049, -0.0606]], dtype=torch.float16), '0.to_out_lora.down.weight': tensor([[-0.0206, -0.0316, -0.0614, ..., -0.0508, 0.0531, 0.0112],\n [-0.0060, 0.0432, -0.0080, ..., -0.0287, -0.0403, 0.0237],\n [-0.0256, 0.0532, 0.0023, ..., 0.0211, -0.0130, 0.0291],\n ...,\n [-0.0271, 0.0031, -0.0259, ..., -0.0264, -0.0214, -0.0312],\n [ 0.0397, -0.0135, -0.0009, ..., -0.0815, 0.0292, 0.0223],\n [-0.0273, -0.0252, 0.0330, ..., -0.0005, 0.0133, 0.0700]], dtype=torch.float16), '0.to_out_lora.up.weight': tensor([[-0.0191, -0.0210, 0.0496, ..., -0.0087, -0.0202, 0.0322],\n [ 0.0031, -0.0204, 0.0136, ..., -0.0051, -0.0091, -0.0129],\n [-0.0106, -0.0472, -0.0279, ..., -0.0480, -0.0178, 0.0009],\n ...,\n [ 0.0343, 0.0047, -0.0153, ..., -0.0076, -0.0089, -0.0075],\n [-0.0126, -0.0089, 0.0114, ..., 0.0207, -0.0172, -0.0101],\n [-0.0011, 0.0274, -0.0397, ..., 0.0220, -0.0067, 0.0104]], dtype=torch.float16), '1.to_q_lora.down.weight': tensor([[-0.0070, -0.0288, -0.0152, ..., -0.0521, -0.0058, 0.0133],\n [-0.0147, 0.0085, -0.0309, ..., 0.0304, 0.0230, -0.0207],\n [-0.0309, -0.0248, 0.0064, ..., 0.0159, 0.0003, -0.0388],\n ...,\n [-0.0800, 0.0439, -0.0122, ..., 0.0243, 0.0046, -0.0456],\n [ 0.0122, -0.0224, 0.0148, ..., 0.0446, 0.0591, -0.0587],\n [-0.0204, 0.0176, 0.0243, ..., 0.0427, -0.0271, -0.0294]], dtype=torch.float16), '1.to_q_lora.up.weight': tensor([[-1.7948e-03, 6.9580e-03, -3.0899e-02, ..., -3.3188e-03, 2.9968e-02, -1.3351e-02],\n [-6.0654e-03, 3.1281e-02, 4.2053e-02, ..., -2.4643e-02, -5.1361e-02, -1.6785e-02],\n [ 2.0844e-02, -3.1769e-02, 3.4542e-03, ..., 1.1314e-02, 1.8417e-02, 1.6575e-03],\n ...,\n [-3.8666e-02, 1.8997e-02, 2.2385e-02, ..., -8.1711e-03, -1.8082e-02, -1.3817e-02],\n [-4.0855e-03, -8.1003e-05, -1.6146e-03, ..., -8.6365e-03, 1.3191e-02, -1.1360e-02],\n [-1.2459e-02, 2.3285e-02, -9.0485e-03, ..., -5.6038e-03, 4.1565e-02, 7.1335e-03]], dtype=torch.float16), '1.to_k_lora.down.weight': tensor([[ 0.0417, 0.0344, -0.0069, ..., -0.0188, 0.0256, 0.0169],\n [ 0.0096, -0.0132, 0.0456, ..., 0.0283, 0.0222, -0.0252],\n [ 0.0105, 0.0699, -0.0793, ..., -0.0422, -0.0141, -0.0124],\n ...,\n [-0.0098, -0.0652, 0.0206, ..., -0.0295, -0.0274, 0.0256],\n [ 0.0350, -0.0076, -0.0437, ..., 0.0026, 0.0315, 0.0055],\n [ 0.0132, -0.0159, 0.0480, ..., -0.0344, -0.0062, -0.0196]], dtype=torch.float16), '1.to_k_lora.up.weight': tensor([[ 0.0256, 0.0703, 0.0474, ..., 0.1281, 0.0234, 0.0753],\n [-0.0015, -0.0087, 0.0188, ..., 0.0728, 0.0195, -0.0004],\n [ 0.0152, 0.0053, -0.0080, ..., 0.0136, -0.0146, -0.0195],\n ...,\n [ 0.0201, 0.0103, -0.0080, ..., 0.0068, -0.0054, 0.0206],\n [ 0.0174, 0.0311, -0.0281, ..., -0.0448, -0.0195, -0.0079],\n [ 0.0019, 0.0022, 0.0022, ..., 0.0314, 0.0500, -0.0289]], dtype=torch.float16), '1.to_v_lora.down.weight': tensor([[ 0.0144, 0.0100, 0.0160, ..., 0.0179, -0.0427, 0.0188],\n [-0.0136, -0.0156, 0.0408, ..., -0.0082, -0.0191, -0.0264],\n [-0.0063, 0.0022, -0.0161, ..., 0.0047, 0.0250, -0.0056],\n ...,\n [ 0.0094, 0.0002, 0.0211, ..., 0.0061, -0.0138, -0.0212],\n [-0.0156, 0.0028, -0.0011, ..., -0.0279, 0.0113, 0.0029],\n [-0.0118, -0.0336, -0.0221, ..., -0.0046, 0.0159, 0.0217]], dtype=torch.float16), '1.to_v_lora.up.weight': tensor([[-0.0008, 0.0039, -0.0132, ..., -0.0139, -0.0274, 0.0126],\n [ 0.0190, -0.0387, -0.0103, ..., 0.0164, -0.0022, 0.0074],\n [-0.0199, 0.0016, 0.0127, ..., -0.0343, -0.0012, -0.0025],\n ...,\n [ 0.0296, 0.0017, -0.0004, ..., 0.0312, -0.0120, 0.0071],\n [ 0.0182, -0.0072, -0.0263, ..., -0.0051, 0.0232, -0.0424],\n [ 0.0131, -0.0083, -0.0098, ..., 0.0063, -0.0061, 0.0236]], dtype=torch.float16), '1.to_out_lora.down.weight': tensor([[-0.0182, 0.0056, -0.0116, ..., 0.0112, -0.0276, 0.0047],\n [ 0.0002, -0.0403, 0.0034, ..., -0.0229, -0.0321, 0.0327],\n [ 0.0141, -0.0162, 0.0267, ..., -0.0074, 0.0293, -0.0012],\n ...,\n [-0.0055, 0.0240, 0.0310, ..., 0.0350, -0.0039, -0.0285],\n [ 0.0001, -0.0103, -0.0059, ..., 0.0163, -0.0262, 0.0144],\n [-0.0145, 0.0084, -0.0149, ..., -0.0104, -0.0035, -0.0056]], dtype=torch.float16), '1.to_out_lora.up.weight': tensor([[-2.6520e-02, -1.0338e-02, 9.0241e-05, ..., -4.9477e-03, -2.9388e-02, 2.4933e-02],\n [-1.1959e-03, 3.8948e-03, 2.8549e-02, ..., -1.3878e-02, -2.2736e-02, 3.4409e-03],\n [ 1.0582e-02, -2.6566e-02, 1.7975e-02, ..., 5.7564e-03, -2.4536e-02, 9.2163e-03],\n ...,\n [-1.2421e-02, 1.7471e-02, 3.0899e-02, ..., -5.4741e-03, 2.1484e-02, -1.9821e-02],\n [ 3.3798e-03, -1.3763e-02, 7.0801e-03, ..., -8.6546e-04, 1.1681e-02, -3.4576e-02],\n [-3.3569e-02, 3.9490e-02, 3.3569e-02, ..., -2.8656e-02, 1.9135e-02, 3.4962e-03]], dtype=torch.float16), '1.to_k_ip.weight': tensor([[-0.0012, 0.0028, 0.0191, ..., -0.0283, -0.0695, -0.1240],\n [ 0.0163, -0.0452, -0.0709, ..., -0.0866, 0.0860, 0.1132],\n [-0.0317, 0.0233, 0.0203, ..., -0.0282, 0.0212, -0.0451],\n ...,\n [-0.0419, 0.1031, -0.0316, ..., -0.0422, -0.0124, -0.0270],\n [ 0.0013, 0.0603, -0.0479, ..., 0.0206, 0.0181, -0.0554],\n [ 0.0381, -0.0086, 0.0709, ..., -0.0477, -0.0379, -0.0657]], dtype=torch.float16), '1.to_v_ip.weight': tensor([[ 0.0232, 0.0130, -0.0120, ..., 0.0298, 0.0013, 0.0156],\n [-0.0285, -0.0284, 0.0332, ..., -0.0126, 0.0185, -0.0109],\n [-0.0119, -0.0229, -0.0602, ..., -0.0703, -0.0301, 0.0040],\n ...,\n [ 0.0540, -0.0362, -0.0247, ..., -0.0556, 0.0670, 0.0755],\n [ 0.0281, -0.0139, 0.0426, ..., 0.0177, 0.0703, -0.0274],\n [-0.0337, -0.0100, 0.0585, ..., 0.0645, 0.0056, -0.0118]], dtype=torch.float16), '2.to_q_lora.down.weight': tensor([[-0.0174, -0.0211, -0.0003, ..., 0.0426, -0.0094, -0.0244],\n [-0.0084, 0.0201, 0.0252, ..., 0.0536, -0.0475, 0.0305],\n [ 0.0097, -0.0007, 0.0367, ..., 0.0075, -0.0059, 0.0041],\n ...,\n [ 0.0231, -0.0169, -0.0461, ..., -0.0161, 0.0066, -0.0165],\n [-0.0141, 0.0002, -0.0404, ..., -0.0149, -0.0359, 0.0201],\n [-0.0424, 0.0416, 0.0210, ..., 0.0548, 0.0354, -0.0080]], dtype=torch.float16), '2.to_q_lora.up.weight': tensor([[-0.0057, 0.0348, 0.0144, ..., 0.0267, -0.0127, 0.0495],\n [ 0.0210, 0.0446, -0.0231, ..., -0.0011, -0.0077, -0.0162],\n [-0.0008, 0.0108, -0.0334, ..., 0.0384, -0.0040, 0.0428],\n ...,\n [ 0.0298, 0.0129, -0.0488, ..., 0.0009, -0.0003, 0.0044],\n [ 0.0297, -0.0195, 0.0269, ..., 0.0083, -0.0025, 0.0193],\n [ 0.0119, 0.0226, -0.0002, ..., 0.0380, -0.0089, 0.0167]], dtype=torch.float16), '2.to_k_lora.down.weight': tensor([[-0.0033, 0.0243, -0.0199, ..., -0.0475, 0.0123, -0.0156],\n [ 0.0267, -0.0450, 0.0047, ..., 0.0135, -0.0463, -0.0370],\n [-0.0065, 0.0039, 0.0066, ..., -0.0402, -0.0330, -0.0217],\n ...,\n [ 0.0006, -0.0040, -0.0604, ..., 0.0118, -0.0088, 0.0158],\n [-0.0009, 0.0501, -0.0350, ..., 0.0245, 0.0119, -0.0387],\n [ 0.0240, 0.0221, -0.0020, ..., -0.0107, 0.0209, -0.0619]], dtype=torch.float16), '2.to_k_lora.up.weight': tensor([[-0.0305, 0.0177, -0.0265, ..., 0.0218, 0.0110, -0.0369],\n [ 0.0010, -0.0404, -0.0116, ..., -0.0071, -0.0090, -0.0158],\n [-0.0446, -0.0243, -0.0064, ..., 0.0089, 0.0329, -0.0040],\n ...,\n [ 0.0164, 0.0158, 0.0121, ..., 0.0367, 0.0211, -0.0284],\n [-0.0127, 0.0367, 0.0351, ..., -0.0018, 0.0021, 0.0252],\n [ 0.0182, 0.0415, -0.0013, ..., -0.0257, -0.0119, -0.0033]], dtype=torch.float16), '2.to_v_lora.down.weight': tensor([[-0.0018, 0.0040, 0.0006, ..., -0.0272, -0.0138, -0.0175],\n [ 0.0079, 0.0052, -0.0533, ..., -0.0482, 0.0227, -0.0248],\n [-0.0064, 0.0106, 0.0202, ..., 0.0638, 0.0276, -0.0096],\n ...,\n [-0.0038, 0.0358, -0.0501, ..., 0.0080, -0.0136, -0.0238],\n [ 0.0230, 0.0031, -0.0083, ..., -0.0084, 0.0024, 0.0221],\n [-0.0396, 0.0043, -0.0206, ..., -0.0004, 0.0063, -0.0084]], dtype=torch.float16), '2.to_v_lora.up.weight': tensor([[ 1.8753e-02, 2.4891e-03, 5.6000e-03, ..., -1.8036e-02, -2.0538e-02, 2.0889e-02],\n [ 2.9968e-02, -1.5930e-02, -4.0222e-02, ..., -1.1070e-02, 9.2239e-03, -2.8168e-02],\n [-6.8169e-03, -1.7609e-02, -2.1164e-02, ..., -1.1353e-02, 2.7618e-02, 4.8757e-05],\n ...,\n [ 8.4991e-03, -8.7357e-03, -3.0670e-02, ..., -3.6850e-03, -9.6436e-03, -3.6896e-02],\n [ 1.7761e-02, -2.6306e-02, -2.6367e-02, ..., -4.3335e-02, 3.4485e-02, -1.0536e-02],\n [ 4.9019e-03, -2.9449e-02, 6.7253e-03, ..., -4.5746e-02, -1.0521e-02, 2.7802e-02]], dtype=torch.float16), '2.to_out_lora.down.weight': tensor([[ 0.0115, 0.0105, 0.0166, ..., -0.0150, -0.0311, -0.0180],\n [ 0.0312, 0.0045, 0.0268, ..., 0.0102, -0.0136, -0.0016],\n [ 0.0153, 0.0175, 0.0117, ..., -0.0067, -0.0170, 0.0238],\n ...,\n [-0.0037, 0.0216, -0.0304, ..., -0.0240, 0.0575, 0.0156],\n [-0.0358, 0.0399, -0.0111, ..., -0.0056, -0.0048, 0.0094],\n [ 0.0142, 0.0300, -0.0231, ..., 0.0213, 0.0685, -0.0118]], dtype=torch.float16), '2.to_out_lora.up.weight': tensor([[ 0.0301, -0.0036, 0.0028, ..., -0.0059, -0.0033, -0.0464],\n [ 0.0429, -0.0155, -0.0455, ..., 0.0071, 0.0424, -0.0032],\n [ 0.0168, 0.0252, -0.0269, ..., -0.0187, -0.0086, 0.0267],\n ...,\n [ 0.0200, 0.0224, -0.0125, ..., 0.0048, 0.0153, -0.0181],\n [ 0.0015, -0.0131, -0.0051, ..., 0.0023, -0.0327, 0.0189],\n [-0.0291, 0.0181, -0.0228, ..., -0.0332, -0.0027, -0.0069]], dtype=torch.float16), '3.to_q_lora.down.weight': tensor([[ 0.0068, -0.0055, 0.0258, ..., 0.0379, 0.0116, 0.0214],\n [ 0.0232, 0.0224, -0.0007, ..., -0.0008, -0.0033, -0.0195],\n [ 0.0424, 0.0018, 0.0132, ..., 0.0094, -0.0170, 0.0594],\n ...,\n [ 0.0183, -0.0031, 0.0226, ..., 0.0501, -0.0050, 0.0334],\n [ 0.0083, -0.0090, 0.0260, ..., -0.0025, 0.0107, -0.0272],\n [-0.0509, 0.0195, -0.0270, ..., 0.0094, 0.0220, 0.0076]], dtype=torch.float16), '3.to_q_lora.up.weight': tensor([[ 0.0147, 0.0634, 0.0062, ..., -0.0186, -0.0321, -0.0546],\n [-0.0165, -0.0022, -0.0101, ..., -0.0122, 0.0150, -0.0581],\n [ 0.0266, 0.0053, 0.0021, ..., -0.0466, -0.0067, -0.0160],\n ...,\n [-0.0203, -0.0317, -0.0465, ..., -0.0170, -0.0418, 0.0466],\n [-0.0068, 0.0100, -0.0246, ..., -0.0451, 0.0215, -0.0232],\n [-0.0100, 0.0205, 0.0346, ..., 0.0547, -0.0162, 0.0017]], dtype=torch.float16), '3.to_k_lora.down.weight': tensor([[-0.0101, -0.0079, 0.0034, ..., 0.0231, -0.0386, -0.0171],\n [-0.0125, -0.0273, -0.0261, ..., 0.0008, 0.0853, 0.0022],\n [ 0.0545, -0.0167, 0.0290, ..., -0.0026, 0.0281, -0.0221],\n ...,\n [ 0.0110, -0.0126, 0.0300, ..., 0.0194, 0.0285, -0.0016],\n [-0.0164, -0.0064, 0.0358, ..., 0.0433, -0.0267, 0.0118],\n [ 0.0078, 0.0315, 0.0059, ..., -0.0390, 0.0518, -0.0679]], dtype=torch.float16), '3.to_k_lora.up.weight': tensor([[ 0.0269, 0.0074, 0.0151, ..., 0.0360, -0.0259, 0.0671],\n [-0.0227, 0.0538, 0.0313, ..., -0.0117, -0.0428, 0.0036],\n [ 0.0040, 0.0122, -0.0061, ..., 0.0555, 0.0037, -0.0147],\n ...,\n [-0.0122, 0.0452, 0.0253, ..., 0.0330, -0.0046, -0.0031],\n [ 0.0265, -0.0052, -0.0163, ..., -0.0081, 0.0135, -0.0305],\n [-0.0321, -0.0487, 0.0310, ..., -0.0113, 0.0231, 0.0155]], dtype=torch.float16), '3.to_v_lora.down.weight': tensor([[-1.7380e-02, -2.0264e-02, -2.0161e-03, ..., 1.7319e-02, -7.5951e-03, 1.1932e-02],\n [-1.9470e-02, -3.0220e-05, 2.1881e-02, ..., 9.9564e-04, 8.8272e-03, -1.8890e-02],\n [ 1.4359e-02, 1.1436e-02, -2.8152e-02, ..., -1.5976e-02, -1.5442e-02, 3.5217e-02],\n ...,\n [-1.3474e-02, 2.0798e-02, -6.9008e-03, ..., -6.3553e-03, -6.6605e-03, -2.9541e-02],\n [ 1.6312e-02, 4.5227e-02, 4.4952e-02, ..., -6.8779e-03, 1.9989e-02, -2.6306e-02],\n [-1.6586e-02, 1.7563e-02, -6.5575e-03, ..., 8.5144e-03, -3.2043e-02, 4.2152e-03]], dtype=torch.float16), '3.to_v_lora.up.weight': tensor([[-0.0094, -0.0146, 0.0054, ..., -0.0102, -0.0086, -0.0180],\n [-0.0025, -0.0081, 0.0218, ..., -0.0006, 0.0015, -0.0003],\n [-0.0114, 0.0135, 0.0013, ..., -0.0027, 0.0199, -0.0193],\n ...,\n [-0.0013, 0.0112, 0.0133, ..., -0.0165, 0.0131, 0.0074],\n [-0.0145, -0.0019, 0.0026, ..., 0.0140, -0.0030, 0.0039],\n [ 0.0123, -0.0024, -0.0375, ..., 0.0047, -0.0243, -0.0045]], dtype=torch.float16), '3.to_out_lora.down.weight': tensor([[ 0.0016, 0.0161, 0.0259, ..., -0.0336, 0.0171, -0.0008],\n [-0.0030, 0.0112, -0.0156, ..., 0.0106, -0.0042, 0.0066],\n [ 0.0058, 0.0216, -0.0267, ..., 0.0063, 0.0002, 0.0016],\n ...,\n [-0.0020, -0.0076, -0.0023, ..., -0.0030, -0.0275, 0.0172],\n [ 0.0104, -0.0112, -0.0083, ..., -0.0017, 0.0136, 0.0166],\n [-0.0012, 0.0058, 0.0014, ..., -0.0103, -0.0032, 0.0055]], dtype=torch.float16), '3.to_out_lora.up.weight': tensor([[-0.0107, 0.0011, 0.0058, ..., 0.0365, -0.0035, -0.0240],\n [ 0.0152, -0.0053, -0.0278, ..., 0.0177, -0.0150, 0.0126],\n [-0.0021, -0.0154, 0.0211, ..., -0.0114, -0.0231, 0.0214],\n ...,\n [-0.0098, -0.0059, 0.0157, ..., -0.0329, -0.0076, -0.0233],\n [-0.0046, -0.0070, -0.0055, ..., 0.0092, -0.0098, 0.0003],\n [-0.0126, 0.0058, 0.0268, ..., -0.0166, -0.0149, 0.0414]], dtype=torch.float16), '3.to_k_ip.weight': tensor([[ 0.0554, -0.0490, -0.0469, ..., -0.0246, 0.0235, 0.0731],\n [-0.0564, -0.0002, -0.0220, ..., -0.0625, 0.0382, 0.0593],\n [ 0.0414, 0.0420, -0.0714, ..., -0.0182, -0.0197, 0.0561],\n ...,\n [ 0.0935, -0.0190, -0.0173, ..., -0.0069, -0.0063, -0.0013],\n [-0.0223, 0.0238, 0.0003, ..., 0.0392, -0.0250, 0.0314],\n [ 0.0244, -0.0066, -0.0121, ..., -0.0193, -0.0032, -0.0020]], dtype=torch.float16), '3.to_v_ip.weight': tensor([[ 0.0084, 0.0228, 0.0120, ..., -0.0076, 0.0552, 0.0204],\n [ 0.0331, 0.0072, -0.0293, ..., 0.0164, -0.0036, 0.0323],\n [ 0.0021, -0.0209, 0.0299, ..., -0.0017, -0.0100, 0.0117],\n ...,\n [-0.0034, 0.0363, 0.0073, ..., -0.0031, 0.0087, 0.0099],\n [ 0.0081, 0.0061, -0.0373, ..., 0.0027, -0.0471, -0.0352],\n [ 0.0055, -0.0406, -0.0107, ..., 0.0125, -0.0013, 0.0202]], dtype=torch.float16), '4.to_q_lora.down.weight': tensor([[ 0.0105, 0.0398, -0.0283, ..., -0.0562, 0.0139, -0.0380],\n [ 0.0039, 0.0039, -0.0477, ..., 0.0167, 0.0087, 0.0084],\n [ 0.0384, 0.0257, 0.0023, ..., -0.0619, -0.0091, 0.0246],\n ...,\n [ 0.0641, 0.0136, 0.0216, ..., -0.0024, -0.0526, -0.0142],\n [ 0.0044, 0.0116, 0.0118, ..., -0.0126, -0.0399, -0.0183],\n [-0.0031, 0.0124, 0.0090, ..., -0.0202, 0.0093, 0.0176]], dtype=torch.float16), '4.to_q_lora.up.weight': tensor([[-0.0427, -0.0086, -0.0576, ..., -0.0122, -0.0030, -0.0087],\n [-0.0105, 0.0136, -0.0057, ..., 0.0006, 0.0406, 0.0035],\n [-0.0206, -0.0548, 0.0563, ..., -0.0173, 0.0262, 0.0219],\n ...,\n [-0.0112, -0.0324, 0.0230, ..., -0.0151, -0.0132, -0.0048],\n [ 0.0115, -0.0323, -0.0638, ..., -0.0197, -0.0440, 0.0061],\n [ 0.0280, -0.0352, -0.0148, ..., 0.0188, 0.0137, -0.0093]], dtype=torch.float16), '4.to_k_lora.down.weight': tensor([[ 0.0034, -0.0438, 0.0731, ..., 0.0487, -0.0331, -0.0143],\n [-0.0023, -0.0403, -0.0186, ..., 0.0917, -0.0001, -0.0219],\n [ 0.0044, 0.0175, 0.0078, ..., -0.0751, 0.0200, -0.0173],\n ...,\n [-0.0113, 0.0528, 0.0312, ..., -0.0072, 0.0150, 0.0059],\n [ 0.0066, 0.0191, 0.0559, ..., -0.0609, 0.0168, 0.0175],\n [-0.0663, 0.0261, 0.0133, ..., -0.0237, -0.0006, -0.0110]], dtype=torch.float16), '4.to_k_lora.up.weight': tensor([[ 3.8391e-02, -1.1627e-02, -3.0365e-02, ..., 3.0853e-02, 4.2816e-02, 1.1933e-04],\n [-6.7353e-05, 7.5645e-03, -6.0234e-03, ..., 1.4854e-02, -1.2215e-02, -1.9485e-02],\n [ 7.1777e-02, -3.3417e-03, -1.0399e-02, ..., -5.0842e-02, -8.2016e-03, 3.5553e-02],\n ...,\n [ 2.6337e-02, 2.7649e-02, -4.6082e-02, ..., -5.0232e-02, 1.6541e-02, -1.3405e-02],\n [-4.6692e-02, -2.9465e-02, 3.1872e-03, ..., 5.3192e-02, 3.7823e-03, 1.0406e-02],\n [-2.1149e-02, -2.0264e-02, -2.0233e-02, ..., 1.0330e-02, -8.1329e-03, -5.1689e-04]], dtype=torch.float16), '4.to_v_lora.down.weight': tensor([[ 0.0361, 0.0061, -0.0210, ..., 0.0321, 0.0016, -0.0147],\n [-0.0078, 0.0113, -0.0402, ..., 0.0327, -0.0173, 0.0081],\n [ 0.0425, -0.0508, 0.0083, ..., 0.0081, -0.0538, 0.0111],\n ...,\n [-0.0003, 0.0003, -0.0102, ..., -0.0026, 0.0172, -0.0059],\n [-0.0126, 0.0177, 0.0330, ..., -0.0122, 0.0171, 0.0006],\n [-0.0037, 0.0006, -0.0150, ..., -0.0338, -0.0291, 0.0539]], dtype=torch.float16), '4.to_v_lora.up.weight': tensor([[-0.0070, 0.0031, 0.0086, ..., 0.0431, -0.0090, -0.0259],\n [ 0.0533, 0.0060, -0.0487, ..., 0.0304, 0.0221, 0.0256],\n [-0.0228, 0.0220, -0.0105, ..., 0.0011, -0.0199, 0.0054],\n ...,\n [-0.0102, -0.0075, -0.0015, ..., 0.0102, 0.0085, 0.0215],\n [ 0.0540, -0.0321, -0.0074, ..., -0.0058, 0.0253, -0.0104],\n [ 0.0010, -0.0117, -0.0368, ..., 0.0146, -0.0012, -0.0162]], dtype=torch.float16), '4.to_out_lora.down.weight': tensor([[ 0.0076, 0.0028, -0.0026, ..., 0.0580, 0.0177, -0.0222],\n [-0.0253, 0.0127, -0.0196, ..., -0.0128, 0.0098, -0.0162],\n [-0.0041, 0.0379, 0.0108, ..., 0.0312, 0.0173, -0.0064],\n ...,\n [-0.0024, 0.0066, -0.0212, ..., -0.0606, 0.0436, 0.0134],\n [ 0.0039, -0.0035, -0.0016, ..., 0.0568, 0.0086, 0.0025],\n [-0.0129, 0.0325, -0.0099, ..., -0.0091, -0.0070, -0.0324]], dtype=torch.float16), '4.to_out_lora.up.weight': tensor([[ 0.0294, 0.0170, -0.0299, ..., 0.0061, 0.0091, 0.0135],\n [-0.0577, -0.0248, 0.0054, ..., 0.0083, -0.0020, -0.0200],\n [-0.0470, -0.0107, 0.0070, ..., 0.0037, 0.0012, 0.0065],\n ...,\n [ 0.0008, 0.0033, -0.0165, ..., 0.0258, 0.0102, -0.0278],\n [ 0.0104, -0.0238, 0.0215, ..., 0.0061, -0.0038, -0.0614],\n [-0.0536, 0.0231, 0.0127, ..., 0.0200, -0.0001, 0.0102]], dtype=torch.float16), '5.to_q_lora.down.weight': tensor([[ 4.7943e-02, 6.0883e-02, 3.0918e-03, ..., -5.1384e-03, 2.9190e-02, 3.9276e-02],\n [-2.6779e-02, 2.9541e-02, -1.4626e-02, ..., -3.1647e-02, -3.6469e-02, -2.9221e-02],\n [-4.5746e-02, 1.3313e-02, -4.1321e-02, ..., 2.7832e-02, 2.5034e-06, 2.0554e-02],\n ...,\n [-4.2419e-03, -3.1097e-02, -5.5573e-02, ..., 5.7037e-02, -9.1505e-04, 9.1782e-03],\n [-3.2043e-02, -7.9422e-03, -1.8646e-02, ..., -7.6294e-02, 5.1636e-02, -1.2505e-02],\n [ 5.2719e-03, 1.6785e-02, 2.8763e-02, ..., 1.9897e-02, -2.9087e-03, -4.5837e-02]], dtype=torch.float16), '5.to_q_lora.up.weight': tensor([[ 0.0345, -0.0108, -0.0026, ..., 0.0157, -0.0059, -0.0507],\n [-0.0108, -0.0363, -0.0260, ..., 0.0462, -0.0301, 0.0162],\n [ 0.0279, 0.0197, 0.0335, ..., -0.0384, 0.0323, 0.0326],\n ...,\n [ 0.0254, -0.0130, -0.0231, ..., 0.0162, -0.0030, 0.0058],\n [-0.0022, -0.0315, 0.0424, ..., 0.0184, -0.0042, -0.0316],\n [ 0.0122, 0.0097, -0.0102, ..., 0.0062, -0.0195, 0.0472]], dtype=torch.float16), '5.to_k_lora.down.weight': tensor([[ 0.0240, -0.0131, 0.0253, ..., 0.0476, 0.0243, 0.0182],\n [ 0.0136, -0.0171, 0.0273, ..., 0.0021, 0.0117, 0.0332],\n [ 0.0194, -0.0778, -0.0372, ..., -0.0285, -0.0089, 0.0057],\n ...,\n [ 0.0038, 0.0131, -0.0018, ..., -0.0703, -0.0741, -0.0569],\n [ 0.0348, 0.0501, -0.0799, ..., 0.0510, -0.0049, -0.0197],\n [ 0.0096, 0.0590, -0.0275, ..., 0.0060, 0.0512, 0.0789]], dtype=torch.float16), '5.to_k_lora.up.weight': tensor([[-0.0152, 0.0555, 0.0072, ..., -0.0529, -0.0149, 0.0307],\n [-0.0104, 0.0068, -0.0116, ..., 0.0429, -0.0529, 0.0046],\n [ 0.0086, -0.0570, -0.0335, ..., -0.0131, 0.0362, 0.0050],\n ...,\n [-0.0335, 0.0710, 0.0488, ..., 0.0358, -0.0064, 0.0468],\n [-0.0249, 0.0096, -0.0822, ..., -0.0418, 0.0072, 0.0125],\n [-0.0270, 0.0007, 0.0076, ..., -0.0119, -0.0475, -0.0070]], dtype=torch.float16), '5.to_v_lora.down.weight': tensor([[-0.0012, -0.0172, -0.0377, ..., -0.0093, -0.0106, 0.0184],\n [-0.0114, 0.0146, 0.0033, ..., 0.0116, -0.0135, 0.0180],\n [-0.0278, -0.0014, 0.0019, ..., 0.0097, 0.0057, 0.0135],\n ...,\n [ 0.0076, -0.0115, -0.0281, ..., 0.0159, -0.0139, 0.0230],\n [ 0.0083, -0.0028, 0.0688, ..., 0.0127, -0.0029, 0.0236],\n [-0.0194, -0.0289, -0.0148, ..., -0.0032, 0.0300, -0.0278]], dtype=torch.float16), '5.to_v_lora.up.weight': tensor([[-0.0033, 0.0131, -0.0105, ..., 0.0245, -0.0103, 0.0317],\n [-0.0126, 0.0231, -0.0188, ..., 0.0068, 0.0094, -0.0002],\n [ 0.0216, -0.0182, 0.0326, ..., -0.0293, -0.0055, 0.0002],\n ...,\n [ 0.0101, -0.0293, 0.0276, ..., -0.0039, -0.0505, -0.0337],\n [ 0.0118, 0.0175, -0.0299, ..., -0.0243, 0.0103, 0.0370],\n [ 0.0033, 0.0253, 0.0028, ..., 0.0023, 0.0230, 0.0126]], dtype=torch.float16), '5.to_out_lora.down.weight': tensor([[ 0.0073, -0.0613, 0.0266, ..., -0.0475, -0.0173, -0.0040],\n [-0.0190, 0.0176, 0.0392, ..., 0.0183, 0.0027, 0.0215],\n [ 0.0104, 0.0053, 0.0034, ..., 0.0496, 0.0073, -0.0121],\n ...,\n [-0.0334, -0.0040, -0.0013, ..., -0.0499, -0.0007, 0.0057],\n [-0.0167, -0.0303, -0.0003, ..., -0.0041, 0.0094, 0.0077],\n [ 0.0259, 0.0107, -0.0053, ..., 0.0500, 0.0042, -0.0252]], dtype=torch.float16), '5.to_out_lora.up.weight': tensor([[-0.0052, -0.0160, 0.1041, ..., -0.0117, 0.0414, 0.0246],\n [-0.0248, -0.0041, 0.0013, ..., 0.0014, -0.0172, -0.0110],\n [ 0.0089, -0.0160, 0.0238, ..., -0.0081, 0.0246, 0.0051],\n ...,\n [ 0.0134, 0.0113, 0.0452, ..., 0.0021, 0.0182, 0.0004],\n [ 0.0162, -0.0203, 0.0076, ..., -0.0173, 0.0186, 0.0059],\n [ 0.0026, -0.0035, -0.0169, ..., 0.0252, -0.0166, 0.0350]], dtype=torch.float16), '5.to_k_ip.weight': tensor([[-0.1010, 0.0600, -0.0645, ..., 0.0426, -0.0567, -0.0177],\n [ 0.0040, 0.0463, -0.0204, ..., -0.0144, 0.0437, 0.0743],\n [-0.0228, 0.0831, 0.0446, ..., 0.0691, -0.0226, 0.0092],\n ...,\n [-0.0911, 0.0332, -0.0202, ..., 0.0256, 0.0076, 0.0010],\n [ 0.0594, 0.0321, 0.0190, ..., 0.0555, 0.0382, -0.0960],\n [-0.0220, -0.0724, 0.0565, ..., -0.0079, -0.0367, 0.0372]], dtype=torch.float16), '5.to_v_ip.weight': tensor([[ 0.0171, -0.0116, -0.0728, ..., -0.0285, 0.0687, 0.0025],\n [-0.0047, -0.0591, -0.0107, ..., -0.0174, 0.0306, 0.0133],\n [ 0.0279, 0.0591, -0.0199, ..., -0.0209, 0.0692, 0.0411],\n ...,\n [ 0.0606, -0.0008, -0.0346, ..., 0.0156, -0.0162, 0.0830],\n [-0.0021, -0.0510, 0.0095, ..., 0.0288, 0.0034, -0.0203],\n [-0.0102, 0.0255, -0.0151, ..., -0.0202, 0.0430, 0.0077]], dtype=torch.float16), '6.to_q_lora.down.weight': tensor([[-0.0361, -0.0382, -0.0066, ..., -0.0162, 0.0253, 0.0018],\n [-0.0224, -0.0009, 0.0283, ..., 0.0145, 0.0210, -0.0066],\n [ 0.0324, -0.0303, 0.0088, ..., -0.0053, -0.0446, -0.0098],\n ...,\n [-0.0002, -0.0576, 0.0157, ..., 0.0438, -0.0255, -0.0092],\n [ 0.0523, -0.0154, -0.0017, ..., -0.0303, -0.0096, -0.0579],\n [ 0.0141, 0.0339, -0.0025, ..., -0.0059, 0.0414, 0.0100]], dtype=torch.float16), '6.to_q_lora.up.weight': tensor([[ 0.0253, -0.0468, -0.0176, ..., 0.0077, 0.0404, 0.0068],\n [-0.0079, -0.0085, 0.0027, ..., -0.0077, 0.0178, 0.0216],\n [-0.0392, 0.0521, -0.0005, ..., -0.0110, 0.0302, 0.0228],\n ...,\n [-0.0008, 0.0390, -0.0285, ..., 0.0018, 0.0231, 0.0579],\n [ 0.0071, 0.0562, 0.0099, ..., 0.0100, -0.0228, 0.0148],\n [-0.0228, 0.0211, -0.0274, ..., 0.0144, 0.0461, 0.0264]], dtype=torch.float16), '6.to_k_lora.down.weight': tensor([[ 0.0046, 0.0485, -0.0013, ..., 0.0259, 0.0204, -0.0043],\n [-0.0128, -0.0422, 0.0430, ..., -0.0221, -0.0344, -0.0021],\n [ 0.0084, 0.0072, -0.0112, ..., 0.0072, 0.0054, 0.0108],\n ...,\n [ 0.0272, -0.0259, 0.0296, ..., -0.0595, -0.0128, -0.0016],\n [-0.0193, -0.0051, 0.0188, ..., 0.0369, 0.0210, 0.0183],\n [ 0.0558, -0.0431, 0.0088, ..., 0.0387, -0.0098, -0.0148]], dtype=torch.float16), '6.to_k_lora.up.weight': tensor([[-0.0256, -0.0057, -0.0121, ..., -0.0246, -0.0019, 0.0130],\n [ 0.0122, -0.0091, 0.0103, ..., -0.0074, 0.0289, 0.0191],\n [ 0.0436, -0.0083, -0.0049, ..., -0.0124, -0.0199, 0.0330],\n ...,\n [-0.0230, -0.0058, 0.0180, ..., 0.0056, -0.0085, 0.0055],\n [ 0.0349, 0.0397, -0.0123, ..., -0.0032, -0.0194, 0.0165],\n [-0.0433, -0.0602, -0.0217, ..., 0.0042, 0.0009, -0.0423]], dtype=torch.float16), '6.to_v_lora.down.weight': tensor([[ 0.0184, 0.0329, 0.0349, ..., -0.0060, -0.0130, -0.0208],\n [ 0.0082, 0.0099, 0.0152, ..., -0.0051, -0.0182, -0.0297],\n [ 0.0397, 0.0436, 0.0119, ..., -0.0205, -0.0553, 0.0224],\n ...,\n [ 0.0261, 0.0121, -0.0103, ..., -0.0154, -0.0191, -0.0163],\n [ 0.0082, 0.0303, 0.0046, ..., 0.0169, 0.0065, -0.0333],\n [-0.0324, 0.0116, 0.0054, ..., 0.0405, 0.0147, 0.0124]], dtype=torch.float16), '6.to_v_lora.up.weight': tensor([[-0.0113, -0.0085, -0.0211, ..., -0.0283, 0.0095, 0.0039],\n [ 0.0384, -0.0138, -0.0088, ..., -0.0084, -0.0119, -0.0221],\n [ 0.0015, -0.0060, -0.0226, ..., 0.0008, -0.0289, -0.0195],\n ...,\n [-0.0077, -0.0062, -0.0094, ..., 0.0013, -0.0281, 0.0288],\n [-0.0418, 0.0192, 0.0090, ..., -0.0087, -0.0056, -0.0170],\n [ 0.0246, -0.0019, -0.0010, ..., 0.0040, 0.0063, 0.0104]], dtype=torch.float16), '6.to_out_lora.down.weight': tensor([[-0.0177, -0.0056, -0.0384, ..., -0.0417, -0.0129, -0.0425],\n [ 0.0128, -0.0029, 0.0164, ..., -0.0155, -0.0119, -0.0041],\n [ 0.0095, -0.0085, 0.0082, ..., -0.0289, -0.0218, -0.0497],\n ...,\n [ 0.0094, 0.0151, 0.0063, ..., -0.0062, -0.0273, 0.0197],\n [-0.0348, 0.0065, -0.0142, ..., -0.0304, -0.0100, -0.0006],\n [ 0.0058, 0.0285, 0.0005, ..., 0.0097, -0.0195, 0.0452]], dtype=torch.float16), '6.to_out_lora.up.weight': tensor([[ 0.0278, 0.0175, -0.0200, ..., 0.0116, -0.0248, -0.0238],\n [-0.0365, -0.0252, 0.0213, ..., 0.0178, -0.0129, 0.0168],\n [ 0.0131, -0.0475, -0.0030, ..., -0.0183, -0.0246, -0.0169],\n ...,\n [ 0.0079, -0.0204, 0.0162, ..., 0.0199, -0.0124, 0.0171],\n [ 0.0059, 0.0197, -0.0221, ..., 0.0147, -0.0252, 0.0114],\n [ 0.0136, -0.0145, 0.0228, ..., 0.0370, 0.0053, 0.0116]], dtype=torch.float16), '7.to_q_lora.down.weight': tensor([[ 0.0075, 0.0398, -0.0160, ..., -0.0135, 0.0430, -0.0416],\n [ 0.0144, 0.0113, 0.0197, ..., -0.0254, 0.0185, -0.0169],\n [-0.0286, 0.0429, 0.0154, ..., 0.0068, -0.0272, -0.0174],\n ...,\n [ 0.0343, 0.0367, 0.0019, ..., -0.0197, -0.0016, -0.0483],\n [ 0.0389, 0.0007, -0.0079, ..., 0.0710, 0.0099, -0.0301],\n [-0.0223, 0.0434, 0.0061, ..., 0.0370, -0.0215, -0.0421]], dtype=torch.float16), '7.to_q_lora.up.weight': tensor([[ 0.0186, -0.0018, 0.0287, ..., -0.0135, 0.0197, 0.0144],\n [ 0.0183, -0.0059, -0.0185, ..., -0.0249, -0.0012, 0.0053],\n [-0.0132, 0.0306, 0.0154, ..., 0.0489, 0.0027, -0.0205],\n ...,\n [ 0.0186, -0.0086, 0.0214, ..., 0.0180, 0.0081, 0.0149],\n [ 0.0064, -0.0080, 0.0195, ..., 0.0054, 0.0101, 0.0521],\n [ 0.0223, 0.0312, -0.0052, ..., 0.0135, -0.0043, -0.0526]], dtype=torch.float16), '7.to_k_lora.down.weight': tensor([[-0.0113, 0.0307, -0.0190, ..., -0.0119, -0.0172, 0.0182],\n [-0.0135, -0.0012, -0.0029, ..., 0.0207, 0.0049, -0.0017],\n [-0.0038, 0.0027, -0.0310, ..., 0.0102, 0.0078, -0.0394],\n ...,\n [-0.0275, 0.0375, -0.0031, ..., 0.0193, 0.0403, 0.0046],\n [-0.0090, 0.0555, 0.0191, ..., 0.0073, 0.0264, 0.0087],\n [-0.0121, -0.0060, -0.0021, ..., -0.0185, -0.0085, -0.0259]], dtype=torch.float16), '7.to_k_lora.up.weight': tensor([[-0.0400, 0.0374, -0.0214, ..., 0.0200, -0.0488, -0.0352],\n [-0.0005, 0.0452, -0.0028, ..., 0.0019, 0.0364, -0.0140],\n [ 0.0072, 0.0034, -0.0261, ..., -0.0203, -0.0267, -0.0623],\n ...,\n [ 0.0050, 0.0048, -0.0091, ..., 0.0046, -0.0065, -0.0062],\n [-0.0075, 0.0073, 0.0581, ..., 0.0159, 0.0524, 0.0368],\n [-0.0105, 0.0226, 0.0204, ..., 0.0124, 0.0406, 0.0302]], dtype=torch.float16), '7.to_v_lora.down.weight': tensor([[-0.0101, 0.0161, -0.0087, ..., -0.0071, -0.0140, -0.0234],\n [-0.0017, -0.0076, 0.0307, ..., 0.0137, -0.0005, -0.0364],\n [-0.0036, 0.0356, -0.0648, ..., -0.0131, -0.0122, 0.0449],\n ...,\n [ 0.0013, -0.0229, 0.0320, ..., 0.0040, 0.0127, 0.0066],\n [ 0.0171, -0.0308, 0.0219, ..., -0.0042, 0.0011, 0.0028],\n [ 0.0194, -0.0289, -0.0217, ..., -0.0082, -0.0054, -0.0077]], dtype=torch.float16), '7.to_v_lora.up.weight': tensor([[ 0.0221, -0.0041, -0.0006, ..., -0.0027, -0.0333, 0.0039],\n [-0.0041, -0.0139, 0.0134, ..., 0.0137, -0.0035, -0.0067],\n [ 0.0188, 0.0062, 0.0024, ..., -0.0162, -0.0073, 0.0269],\n ...,\n [-0.0026, -0.0195, -0.0127, ..., -0.0277, -0.0159, -0.0024],\n [ 0.0048, -0.0322, 0.0136, ..., 0.0374, -0.0160, -0.0095],\n [ 0.0012, -0.0113, -0.0043, ..., 0.0089, 0.0158, -0.0170]], dtype=torch.float16), '7.to_out_lora.down.weight': tensor([[ 0.0138, 0.0209, -0.0200, ..., -0.0389, 0.0205, 0.0095],\n [-0.0085, 0.0215, 0.0496, ..., 0.0198, -0.0081, -0.0351],\n [ 0.0050, -0.0133, 0.0067, ..., -0.0279, 0.0123, -0.0017],\n ...,\n [-0.0048, -0.0243, -0.0374, ..., -0.0145, -0.0104, 0.0073],\n [ 0.0319, -0.0472, -0.0056, ..., -0.0001, -0.0127, 0.0316],\n [-0.0111, 0.0412, 0.0139, ..., -0.0314, -0.0112, 0.0090]], dtype=torch.float16), '7.to_out_lora.up.weight': tensor([[ 0.0105, -0.0085, 0.0313, ..., 0.0239, -0.0024, 0.0036],\n [ 0.0113, -0.0045, -0.0202, ..., -0.0134, -0.0135, -0.0177],\n [-0.0116, 0.0170, -0.0128, ..., -0.0347, 0.0046, -0.0031],\n ...,\n [-0.0349, -0.0447, -0.0051, ..., 0.0238, -0.0249, -0.0113],\n [-0.0107, 0.0036, 0.0048, ..., -0.0147, 0.0011, 0.0081],\n [ 0.0217, -0.0192, 0.0044, ..., -0.0029, 0.0478, -0.0154]], dtype=torch.float16), '7.to_k_ip.weight': tensor([[ 0.0515, 0.0342, 0.0703, ..., 0.0291, -0.0582, 0.0043],\n [ 0.1266, 0.0355, 0.0117, ..., 0.0290, 0.0145, -0.1508],\n [-0.0590, 0.0266, 0.1051, ..., -0.0795, -0.0664, -0.0117],\n ...,\n [ 0.0574, 0.0366, -0.0504, ..., 0.0066, 0.0056, 0.0704],\n [-0.0444, 0.0836, 0.0374, ..., 0.0334, 0.0363, -0.0437],\n [-0.0106, -0.0690, -0.0056, ..., 0.0252, -0.0214, 0.0230]], dtype=torch.float16), '7.to_v_ip.weight': tensor([[ 0.0187, 0.0383, -0.0091, ..., -0.0076, -0.0225, 0.0644],\n [-0.0123, 0.0146, -0.0094, ..., 0.0208, -0.0614, -0.0091],\n [-0.0135, -0.0609, 0.0123, ..., 0.0442, -0.0138, -0.0240],\n ...,\n [ 0.0411, 0.0151, -0.0709, ..., -0.0092, -0.0274, -0.0396],\n [ 0.0338, -0.0015, -0.0092, ..., 0.0298, -0.0417, 0.0029],\n [ 0.0125, -0.0193, 0.0156, ..., -0.0231, 0.0053, -0.0388]], dtype=torch.float16), '8.to_q_lora.down.weight': tensor([[ 0.0078, 0.0401, -0.0351, ..., 0.0106, 0.0163, -0.0391],\n [ 0.0273, -0.0100, -0.0267, ..., -0.0098, -0.0182, -0.0235],\n [ 0.0195, -0.0326, 0.0555, ..., -0.0457, 0.0195, -0.0099],\n ...,\n [ 0.0144, 0.0043, -0.0157, ..., 0.0244, 0.0257, 0.0286],\n [-0.0103, -0.0611, 0.0697, ..., 0.0114, -0.0569, -0.0297],\n [ 0.0104, 0.0098, 0.0555, ..., -0.0222, -0.0001, -0.0179]], dtype=torch.float16), '8.to_q_lora.up.weight': tensor([[ 0.0171, 0.0006, -0.0415, ..., -0.0173, 0.0506, 0.0585],\n [-0.0113, 0.0153, 0.0156, ..., 0.0180, 0.0405, -0.0334],\n [-0.0190, 0.0028, 0.0219, ..., -0.0002, -0.0396, 0.0424],\n ...,\n [-0.0105, -0.0124, 0.0367, ..., -0.0083, 0.0325, 0.0093],\n [-0.0149, 0.0052, 0.0331, ..., 0.0205, 0.0063, -0.0392],\n [-0.0103, 0.0126, -0.0422, ..., -0.0285, -0.0240, 0.0470]], dtype=torch.float16), '8.to_k_lora.down.weight': tensor([[ 0.0600, -0.0590, 0.0303, ..., -0.0261, -0.0305, -0.0252],\n [-0.0304, 0.0493, 0.0077, ..., 0.0367, 0.0201, 0.0615],\n [ 0.0090, -0.0225, -0.0036, ..., -0.0247, 0.0218, 0.0224],\n ...,\n [ 0.0013, -0.0249, 0.0077, ..., 0.0267, 0.0219, -0.0955],\n [ 0.0202, 0.0069, 0.0025, ..., 0.0132, 0.0106, 0.0046],\n [-0.0058, -0.0062, 0.0086, ..., -0.0138, 0.0204, -0.0157]], dtype=torch.float16), '8.to_k_lora.up.weight': tensor([[ 0.0256, 0.0109, 0.0176, ..., 0.0045, -0.0167, 0.0078],\n [ 0.0003, 0.0358, 0.0153, ..., -0.0127, 0.0003, -0.0041],\n [ 0.0350, -0.0030, 0.0071, ..., -0.0077, 0.0016, -0.0435],\n ...,\n [ 0.0056, 0.0066, -0.0100, ..., 0.0108, 0.0371, 0.0367],\n [-0.0033, 0.0305, -0.0510, ..., 0.0320, -0.0031, 0.0245],\n [ 0.0031, 0.0594, -0.0258, ..., -0.0002, -0.0094, 0.0563]], dtype=torch.float16), '8.to_v_lora.down.weight': tensor([[-0.0077, 0.0507, -0.0349, ..., 0.0116, 0.0352, -0.0033],\n [-0.0173, -0.0018, 0.0050, ..., -0.0160, 0.0067, 0.0207],\n [-0.0312, -0.0134, 0.0594, ..., -0.0271, 0.0162, -0.0115],\n ...,\n [-0.0602, -0.0170, 0.0350, ..., -0.0100, -0.0255, -0.0049],\n [-0.0403, 0.0100, -0.0130, ..., -0.0089, 0.0099, 0.0028],\n [ 0.0080, 0.0238, -0.0243, ..., 0.0424, -0.0137, 0.0261]], dtype=torch.float16), '8.to_v_lora.up.weight': tensor([[-0.0104, 0.0002, 0.0381, ..., 0.0052, 0.0209, 0.0631],\n [-0.0186, -0.0386, 0.0051, ..., -0.0196, -0.0182, -0.0263],\n [-0.0233, -0.0063, -0.0226, ..., -0.0334, 0.0555, 0.0245],\n ...,\n [ 0.0264, -0.0272, 0.0119, ..., -0.0179, 0.0504, 0.0505],\n [ 0.0076, -0.0251, 0.0087, ..., -0.0356, -0.0203, -0.0413],\n [-0.0081, -0.0600, -0.0352, ..., -0.0017, -0.0129, 0.0901]], dtype=torch.float16), '8.to_out_lora.down.weight': tensor([[-0.0093, 0.0134, -0.0180, ..., 0.0182, -0.0709, 0.0184],\n [ 0.0096, -0.0325, -0.0168, ..., -0.0797, -0.0759, 0.0420],\n [-0.0179, 0.0178, -0.0125, ..., -0.0149, 0.0114, 0.0203],\n ...,\n [ 0.0195, 0.0134, 0.0075, ..., 0.0753, 0.0242, 0.1122],\n [-0.0297, -0.0102, -0.0302, ..., 0.0538, -0.0334, 0.0364],\n [-0.0222, 0.0317, -0.0016, ..., 0.0133, -0.1031, -0.0067]], dtype=torch.float16), '8.to_out_lora.up.weight': tensor([[-4.8187e-02, 5.1537e-03, -4.0512e-03, ..., 2.1667e-02, -9.8343e-03, -2.8793e-02],\n [-2.1423e-02, 6.8367e-05, 1.0614e-03, ..., -2.9297e-02, -1.1368e-02, 2.5970e-02],\n [-3.1799e-02, -1.4786e-02, -6.4354e-03, ..., 3.8422e-02, 1.3680e-02, 2.9678e-02],\n ...,\n [ 2.8137e-02, 1.0786e-03, -5.4352e-02, ..., -6.2256e-03, 1.6041e-03, 1.4145e-02],\n [-3.6163e-02, 3.1204e-02, 5.6244e-02, ..., -5.7922e-02, -1.1971e-02, 1.1574e-02],\n [-1.5366e-02, -2.4445e-02, 2.5558e-02, ..., -1.9440e-02, 1.2451e-02, 3.8361e-02]], dtype=torch.float16), '9.to_q_lora.down.weight': tensor([[ 0.0050, 0.0070, -0.0181, ..., 0.0732, 0.0468, 0.0110],\n [ 0.0382, 0.0164, -0.0196, ..., 0.0079, -0.0369, -0.0011],\n [-0.0308, -0.0161, 0.0040, ..., -0.0110, -0.0373, -0.0285],\n ...,\n [ 0.0243, 0.0205, -0.0014, ..., 0.0721, -0.0599, -0.0502],\n [-0.0240, -0.0104, 0.0203, ..., -0.0284, 0.0279, -0.0178],\n [ 0.0117, -0.0194, 0.0255, ..., 0.0579, -0.0201, -0.0467]], dtype=torch.float16), '9.to_q_lora.up.weight': tensor([[ 0.0522, 0.0432, -0.0002, ..., 0.0229, -0.0387, 0.0279],\n [-0.0063, 0.0302, 0.0039, ..., -0.0041, -0.0082, -0.0332],\n [ 0.0250, -0.0227, 0.0285, ..., -0.0132, 0.0068, -0.0117],\n ...,\n [-0.0053, -0.0427, 0.0168, ..., 0.0402, 0.0151, 0.0356],\n [-0.0068, 0.0136, 0.0004, ..., 0.0048, -0.0407, 0.0430],\n [-0.0009, 0.0245, 0.0170, ..., -0.0580, 0.0531, -0.0116]], dtype=torch.float16), '9.to_k_lora.down.weight': tensor([[ 0.0021, 0.0048, -0.0359, ..., 0.0242, 0.0195, 0.0404],\n [ 0.0251, -0.0041, 0.0561, ..., 0.0504, 0.0018, 0.0654],\n [-0.0067, 0.0276, -0.0018, ..., -0.0227, -0.0023, -0.0453],\n ...,\n [-0.0126, 0.0173, -0.0646, ..., 0.0334, 0.0135, 0.0570],\n [-0.0440, -0.0183, 0.0255, ..., 0.0135, -0.0230, 0.0204],\n [-0.0123, 0.0095, -0.0018, ..., 0.0098, -0.0019, -0.0349]], dtype=torch.float16), '9.to_k_lora.up.weight': tensor([[-0.0101, -0.0098, -0.0319, ..., -0.0339, -0.0018, -0.0022],\n [ 0.0110, -0.0086, 0.0040, ..., -0.0052, -0.0119, 0.0602],\n [-0.0165, 0.0260, -0.0460, ..., -0.0510, -0.0219, 0.0063],\n ...,\n [ 0.0488, 0.1022, 0.0120, ..., 0.0503, 0.0369, 0.0029],\n [ 0.0439, -0.0378, -0.0291, ..., -0.0495, -0.0741, -0.0548],\n [ 0.0328, -0.0263, -0.0266, ..., -0.0050, -0.0238, 0.0026]], dtype=torch.float16), '9.to_v_lora.down.weight': tensor([[ 0.0031, -0.0403, -0.0011, ..., -0.0017, 0.0031, -0.0245],\n [ 0.0309, 0.0444, -0.0094, ..., 0.0118, 0.0101, 0.0127],\n [-0.0108, -0.0515, -0.0417, ..., 0.0202, -0.0123, -0.0119],\n ...,\n [-0.0172, -0.1044, -0.0265, ..., -0.0049, -0.0040, -0.0395],\n [ 0.0163, -0.0374, -0.0793, ..., -0.0027, -0.0018, 0.0463],\n [ 0.0219, -0.0255, -0.0071, ..., 0.0018, -0.0049, -0.0485]], dtype=torch.float16), '9.to_v_lora.up.weight': tensor([[ 9.0714e-03, -1.0826e-02, 2.8931e-02, ..., 1.8829e-02, -1.8951e-02, -5.8403e-03],\n [ 8.0049e-05, -1.0590e-02, 2.6627e-02, ..., 9.4833e-03, -3.8738e-03, -5.9624e-03],\n [ 1.1848e-02, 2.1103e-02, 9.1019e-03, ..., 3.8574e-02, -1.1055e-02, 4.2816e-02],\n ...,\n [ 1.7202e-04, -2.8793e-02, 7.6027e-03, ..., 1.2001e-02, -1.0185e-02, 1.8196e-03],\n [ 3.0640e-02, -3.3600e-02, 3.9520e-02, ..., -4.0070e-02, -1.6541e-02, 9.9792e-03],\n [ 5.1392e-02, -6.0577e-02, -1.8356e-02, ..., 6.8207e-03, -3.0487e-02, -8.7814e-03]], dtype=torch.float16), '9.to_out_lora.down.weight': tensor([[ 0.0374, 0.0438, 0.0346, ..., 0.0060, -0.0210, -0.0224],\n [-0.0106, 0.0120, -0.0180, ..., -0.0158, 0.0384, 0.0092],\n [-0.0012, -0.0074, 0.0037, ..., -0.0427, -0.0552, 0.0092],\n ...,\n [ 0.0291, -0.0338, -0.0210, ..., 0.0533, 0.0507, 0.0399],\n [ 0.0106, -0.0180, -0.0209, ..., 0.0124, -0.0252, -0.0245],\n [-0.0345, 0.0447, -0.0425, ..., 0.0093, 0.0350, 0.0028]], dtype=torch.float16), '9.to_out_lora.up.weight': tensor([[ 0.0155, -0.0327, -0.0274, ..., -0.0288, -0.0324, 0.0459],\n [ 0.0321, 0.0218, 0.0341, ..., -0.0078, 0.0253, -0.0417],\n [-0.0080, 0.0346, 0.0324, ..., -0.0170, 0.0206, -0.0108],\n ...,\n [ 0.0176, 0.0497, -0.0019, ..., -0.0299, 0.0114, -0.0243],\n [-0.0006, -0.0170, 0.0348, ..., -0.0321, -0.0439, 0.0526],\n [ 0.0058, -0.0190, 0.0197, ..., 0.0598, -0.0252, 0.0372]], dtype=torch.float16), '9.to_k_ip.weight': tensor([[ 0.1166, -0.0426, 0.0198, ..., 0.0103, -0.0036, -0.0684],\n [ 0.0610, 0.0475, -0.0798, ..., -0.0407, 0.0372, -0.0996],\n [-0.0154, 0.0985, -0.0359, ..., 0.0165, -0.0688, -0.0470],\n ...,\n [-0.0156, -0.0219, -0.0538, ..., 0.0323, 0.0087, 0.0401],\n [ 0.0325, -0.0226, -0.0890, ..., -0.0191, 0.0613, -0.0225],\n [-0.0548, 0.0435, 0.0153, ..., 0.0229, -0.0003, -0.0853]], dtype=torch.float16), '9.to_v_ip.weight': tensor([[-0.0508, -0.0253, -0.0466, ..., -0.0306, -0.0143, 0.0067],\n [ 0.0099, 0.0360, -0.0769, ..., 0.0343, 0.0247, 0.0593],\n [-0.0587, 0.0188, -0.0018, ..., 0.0107, -0.0524, -0.0115],\n ...,\n [ 0.0257, 0.0060, -0.0308, ..., 0.0177, -0.0121, 0.0235],\n [-0.0238, -0.0359, 0.0052, ..., 0.0484, 0.0125, 0.0329],\n [-0.0352, -0.0556, 0.0226, ..., 0.0079, -0.0452, 0.0170]], dtype=torch.float16), '10.to_q_lora.down.weight': tensor([[-0.0667, 0.0273, 0.0053, ..., 0.0054, 0.0548, -0.0284],\n [-0.0028, -0.0282, -0.0106, ..., -0.0153, -0.0164, -0.0213],\n [ 0.0026, 0.0163, -0.0171, ..., 0.0209, 0.0267, -0.0241],\n ...,\n [ 0.0124, -0.0290, 0.0443, ..., -0.0199, 0.0256, -0.0266],\n [-0.0296, -0.0075, -0.0555, ..., -0.0011, -0.0208, -0.0522],\n [ 0.0188, -0.0123, 0.0558, ..., 0.0414, -0.0162, 0.0392]], dtype=torch.float16), '10.to_q_lora.up.weight': tensor([[-0.0092, 0.0541, -0.0293, ..., -0.0163, 0.0235, -0.0285],\n [ 0.0406, 0.0188, -0.0079, ..., -0.0329, 0.0092, -0.0267],\n [ 0.0562, -0.0408, 0.0286, ..., -0.0569, 0.0378, 0.0737],\n ...,\n [-0.0101, -0.0601, 0.0039, ..., 0.0404, 0.0468, 0.0300],\n [ 0.0127, 0.0058, 0.0002, ..., -0.0002, -0.0513, -0.0008],\n [ 0.0388, -0.0401, 0.0138, ..., -0.0194, 0.0305, -0.0263]], dtype=torch.float16), '10.to_k_lora.down.weight': tensor([[-0.0391, 0.0191, -0.0307, ..., 0.0175, -0.0093, 0.0032],\n [-0.0067, 0.0012, -0.0011, ..., -0.0346, -0.0159, -0.0195],\n [ 0.0202, -0.0036, 0.0471, ..., 0.0128, 0.0026, -0.0328],\n ...,\n [-0.0430, 0.0475, -0.0312, ..., -0.0029, 0.0135, -0.0424],\n [-0.0086, 0.0240, -0.0498, ..., 0.0008, -0.0236, 0.0298],\n [-0.0226, 0.0272, -0.0050, ..., 0.0345, 0.0140, 0.0208]], dtype=torch.float16), '10.to_k_lora.up.weight': tensor([[-0.0046, -0.0195, -0.0089, ..., -0.0294, -0.0140, 0.0173],\n [-0.0305, -0.0356, 0.0174, ..., -0.0426, -0.0102, -0.0265],\n [ 0.0167, -0.0545, 0.0298, ..., 0.0517, 0.0002, -0.0245],\n ...,\n [-0.0297, 0.0777, 0.0201, ..., 0.0658, 0.0065, -0.0185],\n [ 0.0101, 0.0276, -0.0211, ..., 0.0227, -0.0335, -0.0323],\n [ 0.0098, 0.0151, 0.0313, ..., 0.0358, 0.0072, -0.0110]], dtype=torch.float16), '10.to_v_lora.down.weight': tensor([[ 0.0341, 0.0121, 0.0060, ..., 0.0485, 0.0580, -0.0042],\n [-0.0170, 0.0276, 0.0490, ..., 0.0094, -0.0043, -0.0346],\n [ 0.0267, -0.0488, 0.0041, ..., -0.0295, -0.0032, -0.0157],\n ...,\n [ 0.0536, 0.0300, -0.0030, ..., -0.0217, -0.0139, -0.0174],\n [-0.0604, -0.0258, -0.0258, ..., -0.0021, 0.0495, -0.0215],\n [ 0.0196, -0.0550, -0.0448, ..., 0.0417, -0.0011, 0.0167]], dtype=torch.float16), '10.to_v_lora.up.weight': tensor([[-2.4734e-02, 1.3599e-03, 3.1235e-02, ..., -3.1311e-02, 1.9089e-02, -3.0502e-02],\n [ 3.8696e-02, -1.4252e-02, -2.8412e-02, ..., 1.4679e-02, 1.2833e-02, -2.1698e-02],\n [-9.3994e-03, -2.9083e-02, -2.0279e-02, ..., 1.1475e-02, 7.2021e-02, 4.5593e-02],\n ...,\n [ 2.3270e-02, -3.2928e-02, 1.5022e-02, ..., -2.7863e-02, -4.0710e-02, -2.4933e-02],\n [-1.6953e-02, -7.9727e-03, -2.3407e-02, ..., -2.7008e-03, 6.2637e-03, -1.0864e-02],\n [ 2.0813e-02, -6.6459e-05, 4.5441e-02, ..., -1.2718e-02, -9.8648e-03, 2.4506e-02]], dtype=torch.float16), '10.to_out_lora.down.weight': tensor([[ 0.0213, 0.0114, -0.0131, ..., 0.0101, -0.0220, -0.0204],\n [-0.0048, 0.0974, 0.0132, ..., 0.0248, -0.0249, 0.0206],\n [ 0.0376, 0.0118, 0.0371, ..., 0.0237, 0.0117, -0.0204],\n ...,\n [ 0.0034, -0.0003, -0.0060, ..., 0.0341, -0.0116, -0.0015],\n [-0.0058, -0.0130, -0.0268, ..., -0.0141, -0.0234, -0.0280],\n [ 0.0229, -0.0186, -0.0248, ..., 0.0057, 0.0094, -0.0281]], dtype=torch.float16), '10.to_out_lora.up.weight': tensor([[ 0.0075, 0.0170, 0.0040, ..., -0.0126, -0.0269, -0.0284],\n [ 0.0057, 0.0451, 0.0166, ..., 0.0235, -0.0044, 0.0751],\n [ 0.0096, 0.0082, 0.0496, ..., -0.0335, -0.0395, 0.0189],\n ...,\n [ 0.0205, 0.0387, -0.0478, ..., 0.0303, 0.0219, 0.0248],\n [-0.0320, 0.0009, 0.0515, ..., 0.0245, 0.0052, -0.0367],\n [ 0.0005, -0.0259, -0.0281, ..., -0.0100, 0.0412, 0.0080]], dtype=torch.float16), '11.to_q_lora.down.weight': tensor([[-0.0089, -0.0202, 0.0074, ..., 0.0144, 0.0721, 0.0079],\n [ 0.0078, 0.0192, -0.0481, ..., -0.0152, -0.0242, 0.0424],\n [-0.0183, -0.0393, -0.0089, ..., -0.0153, -0.0486, -0.0345],\n ...,\n [ 0.0172, -0.1351, -0.0259, ..., 0.0082, 0.0145, -0.0287],\n [-0.0085, 0.0043, -0.0746, ..., 0.0093, -0.0170, 0.0797],\n [-0.0217, -0.0079, 0.0020, ..., -0.0064, 0.0394, 0.0111]], dtype=torch.float16), '11.to_q_lora.up.weight': tensor([[ 4.0253e-02, 2.7695e-02, -1.0277e-02, ..., -7.1045e-02, 2.9202e-03, -4.4159e-02],\n [-5.3711e-03, 6.7200e-02, 6.8542e-02, ..., 6.1768e-02, -4.1046e-02, -7.5645e-03],\n [-4.2084e-02, 2.9404e-02, -3.4424e-02, ..., -6.8779e-03, 2.8351e-02, 4.4006e-02],\n ...,\n [-1.9547e-02, 7.3509e-03, -8.5602e-03, ..., 1.4915e-02, -6.2469e-02, 6.4659e-04],\n [ 2.4963e-02, 2.1423e-02, -3.4668e-02, ..., -4.0161e-02, -1.3588e-02, -9.0103e-03],\n [-3.3386e-02, -6.0692e-03, -4.2877e-02, ..., -4.6204e-02, -4.0352e-05, -3.0655e-02]], dtype=torch.float16), '11.to_k_lora.down.weight': tensor([[-0.0371, 0.0077, -0.0582, ..., -0.0369, 0.0733, -0.0227],\n [ 0.0008, 0.0403, 0.0263, ..., -0.0070, 0.0332, -0.0232],\n [-0.0223, 0.0417, -0.0875, ..., -0.0597, -0.0307, 0.0108],\n ...,\n [ 0.0066, -0.0567, 0.0467, ..., 0.0060, -0.0274, -0.0148],\n [-0.0320, 0.0424, 0.0070, ..., 0.0005, 0.0192, -0.0043],\n [ 0.0141, 0.0042, 0.0304, ..., 0.0179, 0.0449, 0.0239]], dtype=torch.float16), '11.to_k_lora.up.weight': tensor([[-0.0245, 0.0658, 0.0480, ..., 0.0096, -0.0028, -0.0822],\n [ 0.0638, 0.0364, 0.0193, ..., 0.0421, -0.0116, -0.0049],\n [-0.0554, 0.0369, 0.0683, ..., 0.0263, 0.0463, 0.0713],\n ...,\n [-0.0191, 0.0023, 0.0139, ..., -0.0202, -0.0296, 0.0360],\n [ 0.0038, 0.0793, -0.0286, ..., 0.0062, 0.0216, 0.0338],\n [-0.0367, 0.0384, 0.0117, ..., -0.0250, 0.0118, 0.0075]], dtype=torch.float16), '11.to_v_lora.down.weight': tensor([[-3.3051e-02, -1.0880e-02, 6.0913e-02, ..., 1.2383e-02, -2.3956e-02, 7.2174e-03],\n [ 6.2408e-03, -1.0742e-02, -2.7206e-02, ..., -7.0038e-03, 6.8970e-03, 1.5732e-02],\n [ 8.0719e-03, 6.5918e-02, 3.5934e-03, ..., 8.4152e-03, 5.1270e-03, -5.2834e-03],\n ...,\n [-1.2741e-02, 1.5190e-02, -6.5552e-02, ..., 2.7332e-03, 1.2703e-02, -5.5408e-04],\n [ 5.6038e-03, 1.2140e-01, 5.1677e-05, ..., -6.7329e-04, 3.7823e-03, 9.1400e-03],\n [ 1.8072e-03, -2.0157e-02, -3.5004e-02, ..., 2.2583e-02, 4.3396e-02, 1.9073e-02]], dtype=torch.float16), '11.to_v_lora.up.weight': tensor([[-0.0271, 0.0272, 0.0166, ..., 0.0256, -0.0136, -0.0031],\n [ 0.0012, -0.0131, -0.0061, ..., 0.0271, -0.0090, 0.0056],\n [ 0.0183, 0.0021, -0.0258, ..., 0.0138, -0.0201, -0.0461],\n ...,\n [ 0.0342, 0.0018, 0.0273, ..., -0.0340, 0.0134, 0.0505],\n [ 0.0200, -0.0432, 0.0237, ..., -0.0156, 0.0116, 0.0088],\n [-0.0131, -0.0142, 0.0173, ..., 0.0236, 0.0478, 0.0112]], dtype=torch.float16), '11.to_out_lora.down.weight': tensor([[ 0.0109, -0.0020, -0.0357, ..., 0.0365, -0.0056, 0.0101],\n [ 0.0007, -0.0231, -0.0077, ..., -0.0272, 0.0687, -0.0540],\n [ 0.0388, -0.1191, 0.0175, ..., 0.0521, -0.0249, -0.0328],\n ...,\n [-0.0105, 0.0387, 0.0095, ..., 0.0035, -0.0738, -0.0397],\n [-0.0044, -0.0146, 0.0223, ..., 0.0470, -0.0148, 0.0307],\n [ 0.0014, 0.0256, -0.0573, ..., -0.0060, 0.0045, -0.0180]], dtype=torch.float16), '11.to_out_lora.up.weight': tensor([[-0.0806, 0.0043, 0.0213, ..., 0.0372, -0.0324, -0.0126],\n [-0.0018, -0.0153, 0.0318, ..., -0.0416, 0.0046, -0.0287],\n [-0.0559, -0.0434, 0.0674, ..., 0.0641, -0.0520, 0.0408],\n ...,\n [ 0.0721, 0.0142, -0.0312, ..., -0.0629, 0.0720, -0.0242],\n [ 0.0180, -0.0137, -0.0093, ..., 0.0164, 0.0092, -0.0530],\n [ 0.0147, -0.0215, 0.0386, ..., 0.0274, -0.0085, 0.0150]], dtype=torch.float16), '11.to_k_ip.weight': tensor([[ 0.0505, 0.0121, 0.0168, ..., -0.0060, 0.0004, -0.0392],\n [ 0.0015, -0.1176, 0.0599, ..., 0.0224, 0.0618, 0.0305],\n [ 0.0250, -0.0767, -0.0028, ..., -0.0520, 0.0062, -0.0521],\n ...,\n [-0.0280, -0.0037, -0.0543, ..., -0.0393, -0.0091, 0.0507],\n [ 0.0102, -0.0606, 0.0386, ..., -0.0330, 0.0550, 0.0482],\n [-0.0583, 0.0045, 0.0329, ..., -0.0044, -0.0034, 0.0532]], dtype=torch.float16), '11.to_v_ip.weight': tensor([[-0.0102, -0.0243, -0.0319, ..., 0.0005, 0.0003, -0.0740],\n [-0.0257, 0.0177, -0.0367, ..., -0.0136, 0.0144, -0.0135],\n [ 0.0244, -0.0295, 0.0173, ..., 0.0029, -0.0162, -0.0232],\n ...,\n [-0.0200, -0.0031, -0.0412, ..., 0.0072, -0.0185, 0.0075],\n [-0.0433, 0.0233, 0.0453, ..., -0.0214, -0.0060, -0.0068],\n [ 0.0257, -0.0225, 0.0257, ..., -0.0044, -0.0449, 0.0363]], dtype=torch.float16), '12.to_q_lora.down.weight': tensor([[-0.0341, 0.0527, 0.0126, ..., 0.0004, 0.0356, -0.0332],\n [ 0.0245, -0.0286, -0.0172, ..., 0.0202, 0.0069, 0.0083],\n [-0.0546, -0.0092, -0.0186, ..., -0.0045, -0.0083, -0.0019],\n ...,\n [-0.0510, 0.0453, -0.0108, ..., 0.0342, -0.0489, 0.0528],\n [-0.0464, 0.1046, 0.0434, ..., 0.0154, 0.0049, -0.0240],\n [-0.0017, 0.0127, -0.0073, ..., -0.0295, -0.0436, 0.0076]], dtype=torch.float16), '12.to_q_lora.up.weight': tensor([[-0.0350, 0.0245, -0.0690, ..., -0.0380, -0.0225, -0.0298],\n [ 0.0662, 0.0647, -0.0087, ..., -0.0298, -0.0298, -0.0452],\n [-0.0246, 0.0627, 0.0511, ..., 0.0061, -0.0341, -0.0454],\n ...,\n [-0.0241, -0.0063, 0.0170, ..., -0.0294, -0.0305, 0.0376],\n [-0.0621, 0.0159, -0.0180, ..., -0.0529, 0.0114, 0.0710],\n [-0.0098, 0.0082, -0.0024, ..., 0.0529, 0.0368, 0.0703]], dtype=torch.float16), '12.to_k_lora.down.weight': tensor([[-0.0158, -0.0171, -0.0142, ..., -0.0372, -0.0668, 0.0366],\n [-0.0410, 0.0004, 0.0368, ..., 0.0098, -0.0272, 0.0416],\n [-0.0187, 0.0194, -0.0181, ..., 0.0593, 0.0037, -0.0611],\n ...,\n [-0.0049, -0.0221, -0.0012, ..., 0.0383, 0.0358, 0.0163],\n [ 0.0669, -0.0290, -0.0515, ..., -0.0093, 0.0155, 0.0225],\n [ 0.0021, -0.0326, -0.0243, ..., -0.0351, 0.0409, -0.0375]], dtype=torch.float16), '12.to_k_lora.up.weight': tensor([[-0.0504, -0.0387, -0.0173, ..., -0.1059, -0.0354, -0.0128],\n [ 0.0515, 0.0106, -0.0179, ..., 0.0100, -0.0018, -0.0598],\n [-0.0066, 0.0156, 0.0204, ..., -0.0030, -0.1027, 0.0086],\n ...,\n [ 0.0273, 0.0654, 0.0004, ..., -0.0403, 0.0757, 0.0491],\n [-0.0081, -0.0354, -0.0038, ..., -0.0375, -0.0192, -0.0396],\n [ 0.0174, 0.0346, 0.0307, ..., 0.0028, 0.0063, 0.0084]], dtype=torch.float16), '12.to_v_lora.down.weight': tensor([[ 0.0334, 0.0138, 0.0648, ..., -0.0035, -0.0399, -0.0011],\n [ 0.0100, -0.0192, 0.0027, ..., 0.0292, 0.0029, -0.0116],\n [-0.0008, 0.0652, 0.0154, ..., 0.0024, 0.0391, 0.0360],\n ...,\n [ 0.0617, 0.0052, 0.0290, ..., 0.0142, -0.0314, 0.0153],\n [ 0.0184, 0.0068, 0.0623, ..., 0.0234, -0.0017, -0.0363],\n [-0.0097, 0.0033, -0.0213, ..., 0.0037, -0.0023, -0.0298]], dtype=torch.float16), '12.to_v_lora.up.weight': tensor([[-1.8372e-02, -5.3802e-02, 3.0875e-05, ..., 4.0588e-02, 1.4801e-02, 2.4307e-02],\n [ 3.1189e-02, -1.7319e-02, -5.5122e-03, ..., 4.3060e-02, 3.0121e-02, 2.6978e-02],\n [-2.2461e-02, -1.8204e-02, -6.8188e-04, ..., -2.8458e-02, 5.4749e-02, 2.2766e-02],\n ...,\n [ 2.1652e-02, 2.0103e-03, -4.4670e-03, ..., -3.3081e-02, -4.0039e-02, 2.4887e-02],\n [ 5.3711e-03, 8.4534e-03, -7.3059e-02, ..., 5.7755e-03, 2.8336e-02, 3.3203e-02],\n [ 1.2085e-02, -1.4915e-02, 1.7868e-02, ..., 4.5013e-02, 6.7215e-03, 1.3664e-02]], dtype=torch.float16), '12.to_out_lora.down.weight': tensor([[-0.0231, 0.0211, -0.0112, ..., -0.0243, 0.0138, -0.0291],\n [-0.0284, 0.0630, 0.0312, ..., 0.0158, -0.0009, -0.0099],\n [ 0.0276, -0.0332, 0.0308, ..., -0.0294, 0.0024, 0.0187],\n ...,\n [-0.0270, -0.0119, 0.0459, ..., -0.0611, 0.0128, 0.0241],\n [ 0.0193, -0.0441, 0.0160, ..., -0.0201, -0.0198, -0.0571],\n [ 0.0218, -0.0093, 0.0264, ..., -0.0143, -0.0033, 0.0315]], dtype=torch.float16), '12.to_out_lora.up.weight': tensor([[ 0.0114, 0.0482, 0.0164, ..., 0.0093, 0.0217, -0.0476],\n [-0.0250, 0.0266, -0.0066, ..., -0.0228, 0.0063, -0.0100],\n [-0.0009, 0.0209, 0.0032, ..., 0.0064, 0.0228, -0.0309],\n ...,\n [ 0.0384, 0.0451, -0.0004, ..., -0.0103, -0.0064, 0.0212],\n [ 0.0078, 0.0082, -0.0096, ..., 0.0380, -0.0485, -0.0432],\n [ 0.0234, 0.0302, -0.0154, ..., 0.0166, 0.0072, 0.0078]], dtype=torch.float16), '13.to_q_lora.down.weight': tensor([[-0.0301, -0.0220, -0.0241, ..., 0.0494, 0.0004, 0.0056],\n [ 0.0082, -0.0251, 0.0327, ..., -0.0402, 0.0065, 0.0048],\n [-0.0342, -0.0270, 0.0049, ..., -0.0439, -0.0291, -0.0080],\n ...,\n [ 0.0027, -0.0167, 0.0199, ..., 0.0287, 0.0157, -0.0070],\n [-0.0389, 0.0077, 0.0732, ..., -0.0499, 0.0729, 0.0508],\n [-0.0106, -0.0531, 0.0011, ..., -0.0163, 0.0324, -0.0109]], dtype=torch.float16), '13.to_q_lora.up.weight': tensor([[ 0.0042, 0.0050, -0.0177, ..., -0.0001, 0.0206, -0.0192],\n [-0.0038, 0.0151, 0.0177, ..., 0.0353, 0.0264, 0.0087],\n [ 0.0170, 0.0219, 0.0087, ..., -0.0194, 0.0333, -0.1019],\n ...,\n [ 0.0315, -0.0179, -0.0598, ..., 0.0568, 0.0439, -0.0715],\n [ 0.0288, -0.0012, 0.0416, ..., 0.0746, -0.0129, -0.0081],\n [ 0.0854, -0.0755, -0.0102, ..., -0.0016, -0.0583, -0.0674]], dtype=torch.float16), '13.to_k_lora.down.weight': tensor([[-0.0397, -0.0507, -0.0155, ..., 0.0139, 0.0055, 0.0226],\n [ 0.0530, -0.0023, 0.0130, ..., -0.0109, 0.0191, -0.0034],\n [-0.0401, -0.0011, -0.0280, ..., -0.0194, 0.0509, -0.0182],\n ...,\n [-0.0219, 0.0454, -0.0531, ..., -0.0036, -0.0200, -0.0521],\n [-0.0222, 0.0066, -0.0264, ..., -0.0015, -0.0184, 0.0261],\n [-0.0461, 0.0240, -0.0455, ..., 0.0093, 0.0099, -0.0097]], dtype=torch.float16), '13.to_k_lora.up.weight': tensor([[ 0.0526, 0.0099, -0.0274, ..., -0.0135, -0.0547, -0.0091],\n [-0.0604, 0.0098, -0.0380, ..., -0.0334, 0.0040, 0.0460],\n [ 0.0213, -0.0119, 0.0453, ..., 0.0580, -0.1019, 0.0323],\n ...,\n [-0.0559, -0.0117, 0.1141, ..., -0.0241, -0.0244, 0.0280],\n [-0.0508, -0.0167, 0.0482, ..., -0.0247, 0.0556, -0.0354],\n [-0.0716, 0.0319, -0.0299, ..., 0.0388, -0.0275, 0.0444]], dtype=torch.float16), '13.to_v_lora.down.weight': tensor([[-0.0020, 0.0303, 0.0302, ..., 0.0117, 0.0210, -0.0597],\n [ 0.0146, -0.0077, 0.0701, ..., 0.0122, -0.0028, 0.0271],\n [ 0.0045, 0.0961, -0.0455, ..., -0.0211, -0.0096, -0.0369],\n ...,\n [-0.0225, -0.0291, 0.0807, ..., -0.0120, 0.0309, 0.0826],\n [-0.0175, -0.0432, 0.0114, ..., -0.0081, -0.0068, -0.0357],\n [-0.0449, 0.0149, -0.0798, ..., 0.0012, -0.0018, 0.0154]], dtype=torch.float16), '13.to_v_lora.up.weight': tensor([[ 0.0013, 0.0179, -0.0145, ..., -0.0062, 0.0157, -0.0104],\n [ 0.0313, 0.0224, 0.0456, ..., -0.0450, -0.0149, 0.0226],\n [-0.0291, 0.0031, -0.0562, ..., -0.0073, -0.0183, 0.0283],\n ...,\n [ 0.0231, -0.0129, -0.0398, ..., -0.0054, 0.0094, -0.0115],\n [-0.0002, 0.0107, -0.0127, ..., 0.0061, 0.0075, 0.0621],\n [-0.0387, -0.0337, -0.0117, ..., -0.0633, -0.0024, -0.0108]], dtype=torch.float16), '13.to_out_lora.down.weight': tensor([[-0.0487, -0.0304, -0.0195, ..., -0.0505, 0.0192, 0.0196],\n [-0.0218, -0.0562, 0.0355, ..., 0.0438, 0.0156, -0.0698],\n [ 0.0260, 0.0406, 0.0542, ..., -0.0324, 0.0273, -0.0698],\n ...,\n [ 0.0275, -0.0197, 0.0404, ..., 0.0065, -0.0119, 0.0421],\n [ 0.0049, -0.0275, 0.0288, ..., -0.0008, -0.0057, 0.0240],\n [-0.0408, 0.0004, -0.0163, ..., -0.0156, 0.0403, 0.0733]], dtype=torch.float16), '13.to_out_lora.up.weight': tensor([[-0.0591, 0.0239, -0.0064, ..., 0.0126, 0.1027, 0.0055],\n [ 0.0676, 0.0034, -0.0553, ..., 0.0380, -0.1110, -0.0218],\n [-0.0257, -0.0006, 0.0532, ..., -0.0023, 0.0107, 0.0216],\n ...,\n [-0.0027, -0.0616, 0.0698, ..., 0.0258, 0.0814, -0.0090],\n [ 0.0154, 0.0108, -0.0252, ..., 0.0043, -0.0039, 0.0461],\n [-0.0217, -0.0109, -0.0573, ..., -0.0714, 0.0430, 0.0427]], dtype=torch.float16), '13.to_k_ip.weight': tensor([[ 0.0292, -0.0186, -0.0551, ..., -0.0097, -0.0714, 0.0273],\n [ 0.0242, -0.0964, 0.0848, ..., -0.0204, 0.0629, 0.0103],\n [-0.0088, 0.0252, 0.0025, ..., -0.0697, -0.0425, 0.0167],\n ...,\n [ 0.0764, 0.0285, 0.0559, ..., 0.0651, -0.0242, -0.0769],\n [-0.0534, 0.0738, -0.0905, ..., -0.0613, -0.0048, -0.0382],\n [ 0.0479, 0.0413, 0.0388, ..., -0.0113, -0.0115, 0.0583]], dtype=torch.float16), '13.to_v_ip.weight': tensor([[-0.0008, 0.0128, -0.0355, ..., -0.0410, 0.0821, -0.0719],\n [-0.0246, 0.0459, -0.0729, ..., 0.0264, -0.0806, -0.0704],\n [-0.0775, -0.0560, 0.0355, ..., 0.0173, -0.0227, -0.0760],\n ...,\n [ 0.0209, -0.0128, 0.1073, ..., -0.0407, -0.0969, 0.0851],\n [-0.0860, -0.0927, 0.0196, ..., 0.0449, 0.0041, 0.0940],\n [ 0.0605, 0.0433, 0.0202, ..., 0.0393, -0.0586, 0.0390]], dtype=torch.float16), '14.to_q_lora.down.weight': tensor([[ 0.0353, -0.0200, -0.0464, ..., -0.0230, 0.0206, 0.0012],\n [ 0.0077, -0.0488, 0.0275, ..., -0.0209, -0.0087, -0.0130],\n [-0.0243, -0.0070, -0.0088, ..., 0.0479, -0.0246, -0.0205],\n ...,\n [-0.0137, 0.0349, 0.0017, ..., 0.0043, -0.0069, 0.0097],\n [-0.0042, 0.0092, 0.0326, ..., -0.0109, -0.0203, 0.0360],\n [-0.0126, 0.0159, 0.0018, ..., -0.0325, 0.0229, -0.0175]], dtype=torch.float16), '14.to_q_lora.up.weight': tensor([[-0.0446, 0.0071, -0.0195, ..., -0.0703, 0.0476, -0.0272],\n [-0.0153, -0.0086, 0.0331, ..., -0.0603, 0.0179, 0.0034],\n [-0.0103, 0.0204, -0.0239, ..., -0.0056, -0.0616, 0.0165],\n ...,\n [ 0.0251, -0.0325, 0.0140, ..., 0.0446, 0.0509, -0.0436],\n [ 0.0112, -0.0349, 0.0165, ..., -0.0225, -0.0078, 0.0188],\n [ 0.0101, 0.0241, 0.0601, ..., 0.0212, -0.0210, 0.0100]], dtype=torch.float16), '14.to_k_lora.down.weight': tensor([[ 0.0104, -0.0355, 0.0447, ..., 0.0443, -0.0352, -0.0022],\n [ 0.0030, -0.0319, -0.0045, ..., 0.0084, 0.0040, 0.0133],\n [-0.0198, 0.0307, -0.0561, ..., -0.0414, 0.0062, -0.0285],\n ...,\n [ 0.0098, 0.0171, 0.0517, ..., 0.0602, 0.0729, -0.0267],\n [-0.0231, -0.0267, 0.0170, ..., 0.0044, -0.0197, -0.0103],\n [-0.0539, 0.0226, 0.0198, ..., -0.0637, 0.0401, -0.0137]], dtype=torch.float16), '14.to_k_lora.up.weight': tensor([[-0.0291, 0.0044, 0.0215, ..., 0.0281, -0.0013, -0.0231],\n [ 0.0154, 0.0040, -0.0564, ..., 0.0728, 0.0354, -0.0116],\n [ 0.0082, 0.0049, 0.0059, ..., -0.0337, -0.0114, 0.0751],\n ...,\n [ 0.0179, 0.0511, -0.0255, ..., 0.0277, 0.0017, 0.0526],\n [ 0.0023, 0.0449, -0.0199, ..., -0.0059, -0.0197, -0.0183],\n [-0.0025, -0.0018, 0.0413, ..., 0.0577, -0.0285, 0.0127]], dtype=torch.float16), '14.to_v_lora.down.weight': tensor([[-0.0136, -0.0982, -0.0172, ..., -0.0083, -0.0112, -0.0166],\n [-0.0013, 0.0412, 0.0350, ..., 0.0262, -0.0175, 0.0377],\n [-0.0165, -0.0417, 0.0061, ..., -0.0040, 0.0105, 0.0948],\n ...,\n [-0.0195, 0.0157, -0.0024, ..., -0.0256, 0.0224, 0.0844],\n [-0.0075, 0.0026, 0.0103, ..., -0.0061, -0.0489, -0.0330],\n [ 0.0203, 0.0006, 0.0262, ..., 0.0127, 0.0107, 0.0562]], dtype=torch.float16), '14.to_v_lora.up.weight': tensor([[ 0.0022, 0.0080, -0.0093, ..., 0.0064, -0.0131, 0.0414],\n [ 0.0201, -0.0590, 0.0113, ..., 0.0223, -0.0100, -0.0370],\n [ 0.0417, 0.0620, -0.0112, ..., -0.0148, 0.0020, -0.0131],\n ...,\n [-0.0385, 0.0059, -0.0275, ..., -0.0668, -0.0378, -0.0144],\n [ 0.0084, -0.0220, 0.0615, ..., -0.0025, -0.0035, -0.0165],\n [ 0.0471, -0.0016, 0.0140, ..., 0.0020, -0.0014, 0.0164]], dtype=torch.float16), '14.to_out_lora.down.weight': tensor([[-0.0090, 0.0108, -0.0104, ..., 0.0033, -0.0243, 0.0027],\n [ 0.0075, 0.0182, 0.0062, ..., 0.0375, -0.0299, 0.0041],\n [-0.0157, 0.0216, 0.0653, ..., 0.0501, 0.0370, 0.0077],\n ...,\n [-0.0667, 0.0077, 0.0620, ..., -0.0040, -0.0309, 0.0526],\n [-0.0129, -0.0125, -0.0591, ..., 0.0247, 0.0104, -0.0113],\n [ 0.0016, -0.0218, -0.0337, ..., -0.0399, -0.0509, 0.0323]], dtype=torch.float16), '14.to_out_lora.up.weight': tensor([[-0.0222, 0.0170, -0.0173, ..., 0.0220, -0.0359, 0.0091],\n [ 0.0092, -0.0076, -0.0422, ..., 0.0146, -0.0189, -0.0010],\n [ 0.0047, 0.0456, 0.0478, ..., -0.0011, -0.0260, 0.0395],\n ...,\n [-0.0194, -0.0002, -0.0036, ..., 0.0236, -0.0112, -0.0040],\n [ 0.0441, -0.0200, -0.0311, ..., -0.0199, -0.0353, 0.0160],\n [-0.0142, 0.0516, -0.0111, ..., -0.0459, -0.0352, 0.0313]], dtype=torch.float16), '15.to_q_lora.down.weight': tensor([[ 0.0149, -0.0141, -0.0181, ..., 0.0074, -0.0097, 0.0512],\n [-0.0269, 0.0037, -0.0097, ..., 0.0121, -0.0137, 0.0284],\n [ 0.0130, -0.0412, 0.0083, ..., -0.0137, -0.0384, -0.0048],\n ...,\n [-0.0363, 0.0427, -0.0183, ..., -0.0145, 0.0137, -0.0401],\n [-0.0145, 0.0129, 0.0369, ..., -0.0041, 0.0298, -0.0056],\n [-0.0910, 0.0399, 0.0590, ..., -0.0012, 0.0123, 0.0323]], dtype=torch.float16), '15.to_q_lora.up.weight': tensor([[-0.0037, 0.0016, -0.0097, ..., 0.0081, -0.0138, -0.0470],\n [-0.0092, 0.0193, -0.0304, ..., 0.0079, 0.0029, 0.0292],\n [ 0.0194, -0.0052, -0.0474, ..., 0.0082, -0.0341, -0.0265],\n ...,\n [ 0.0544, 0.0563, 0.0058, ..., 0.0066, 0.0561, 0.0872],\n [ 0.0362, 0.0290, -0.0305, ..., -0.0414, 0.0269, 0.0063],\n [ 0.0230, 0.0494, 0.0009, ..., 0.0104, 0.0268, -0.0068]], dtype=torch.float16), '15.to_k_lora.down.weight': tensor([[ 0.0615, 0.0586, 0.0095, ..., -0.0967, 0.0282, -0.0085],\n [ 0.0442, 0.0743, 0.0116, ..., -0.0103, -0.0426, -0.0607],\n [-0.0079, 0.0379, -0.0009, ..., 0.0119, 0.0421, 0.0299],\n ...,\n [-0.0199, -0.0589, -0.0739, ..., 0.0222, 0.0168, -0.0759],\n [-0.0522, -0.0272, -0.0054, ..., 0.0223, -0.0149, 0.0065],\n [ 0.0098, -0.0017, -0.0385, ..., 0.0041, 0.0138, -0.0124]], dtype=torch.float16), '15.to_k_lora.up.weight': tensor([[ 0.0465, 0.1011, 0.0071, ..., 0.0133, 0.0168, 0.0356],\n [ 0.0638, 0.0172, -0.0102, ..., 0.0285, -0.0267, -0.0003],\n [-0.1103, -0.0178, -0.0125, ..., -0.0275, 0.0281, -0.0101],\n ...,\n [-0.0082, -0.0183, -0.0494, ..., 0.0401, -0.0091, -0.0014],\n [ 0.0257, -0.0252, 0.0210, ..., 0.0158, -0.0114, 0.0049],\n [ 0.0435, -0.0220, -0.0235, ..., 0.0569, 0.0139, 0.0385]], dtype=torch.float16), '15.to_v_lora.down.weight': tensor([[-0.0234, -0.0128, -0.0818, ..., 0.0109, -0.0196, -0.0022],\n [-0.0075, -0.0704, -0.0190, ..., 0.0299, 0.0127, 0.0161],\n [-0.0094, -0.0106, -0.0255, ..., 0.0051, 0.0193, 0.0684],\n ...,\n [ 0.0132, 0.0525, 0.0773, ..., 0.0106, 0.0166, 0.0459],\n [ 0.0182, -0.0405, -0.0318, ..., 0.0269, -0.0218, -0.0419],\n [-0.0140, 0.0293, -0.0610, ..., 0.0038, 0.0106, -0.0110]], dtype=torch.float16), '15.to_v_lora.up.weight': tensor([[ 0.0678, 0.0087, 0.0171, ..., 0.0030, 0.0126, 0.0171],\n [-0.0162, -0.0106, 0.0119, ..., 0.0387, 0.0275, -0.0547],\n [-0.0151, -0.0246, -0.0316, ..., 0.0111, -0.0210, 0.0253],\n ...,\n [-0.0151, -0.0375, 0.0238, ..., -0.0032, 0.0027, 0.0391],\n [ 0.0007, -0.0010, -0.0101, ..., -0.0756, -0.0347, -0.0195],\n [-0.0602, 0.0417, -0.0139, ..., -0.0222, 0.0149, 0.0434]], dtype=torch.float16), '15.to_out_lora.down.weight': tensor([[-0.0643, -0.0103, 0.0407, ..., 0.0047, -0.0565, 0.0340],\n [-0.0604, 0.0216, 0.0540, ..., -0.0205, 0.0105, 0.0289],\n [-0.0584, -0.0787, 0.0436, ..., 0.0967, 0.0012, -0.0351],\n ...,\n [-0.0527, -0.0422, 0.0101, ..., 0.0397, 0.0696, 0.0521],\n [ 0.0047, 0.0376, 0.0151, ..., 0.0157, 0.0536, 0.0287],\n [-0.0055, 0.0421, 0.0341, ..., -0.0780, -0.0230, -0.0489]], dtype=torch.float16), '15.to_out_lora.up.weight': tensor([[ 0.0149, -0.0532, 0.0050, ..., 0.0072, 0.0799, -0.0024],\n [ 0.0220, 0.0164, 0.0096, ..., -0.0237, -0.0062, -0.0048],\n [ 0.0171, -0.0372, -0.0103, ..., 0.0172, 0.0056, -0.0435],\n ...,\n [-0.0010, -0.0185, 0.0003, ..., -0.0551, 0.0271, 0.0505],\n [ 0.0339, -0.0384, -0.0316, ..., -0.0212, 0.0063, 0.0152],\n [-0.0088, -0.0416, 0.0412, ..., 0.0190, 0.0346, -0.0550]], dtype=torch.float16), '15.to_k_ip.weight': tensor([[-0.0010, 0.0097, -0.0138, ..., -0.0245, -0.1068, 0.0293],\n [-0.0458, -0.0192, 0.0149, ..., -0.0815, -0.0267, 0.0616],\n [ 0.0789, -0.0194, -0.0627, ..., -0.0457, -0.0019, -0.0554],\n ...,\n [-0.0963, -0.0731, -0.0138, ..., -0.0263, 0.0583, 0.0574],\n [-0.0252, -0.0328, 0.0114, ..., -0.0117, 0.0142, -0.0101],\n [ 0.0859, -0.0495, 0.0359, ..., -0.0726, 0.0145, 0.0856]], dtype=torch.float16), '15.to_v_ip.weight': tensor([[-0.0425, 0.0970, -0.0168, ..., -0.0486, -0.0696, -0.0646],\n [-0.0179, 0.0094, -0.0278, ..., 0.0320, 0.0112, -0.0942],\n [ 0.0106, 0.0805, 0.0233, ..., -0.0097, 0.0171, 0.0179],\n ...,\n [ 0.0070, 0.0515, 0.0393, ..., -0.0468, -0.0525, 0.0114],\n [ 0.0327, 0.0627, 0.0222, ..., 0.0004, 0.0065, 0.0745],\n [-0.0289, -0.0040, 0.0172, ..., -0.0042, 0.0284, 0.0127]], dtype=torch.float16), '16.to_q_lora.down.weight': tensor([[-0.0606, 0.0300, 0.0144, ..., 0.0378, -0.0570, -0.0764],\n [-0.0112, -0.0215, -0.0630, ..., 0.0554, 0.0729, 0.0040],\n [-0.0663, -0.0276, -0.0443, ..., 0.0128, -0.0180, -0.0390],\n ...,\n [ 0.0479, -0.0847, 0.0515, ..., -0.0254, -0.0390, 0.0434],\n [ 0.0091, -0.0587, 0.0182, ..., 0.0027, 0.0035, 0.0071],\n [ 0.0545, 0.0623, 0.0457, ..., -0.0200, -0.0059, -0.0002]], dtype=torch.float16), '16.to_q_lora.up.weight': tensor([[-0.0298, -0.0107, 0.0065, ..., -0.0190, 0.0087, -0.0159],\n [ 0.0179, -0.0242, 0.0334, ..., -0.1215, 0.0398, -0.0194],\n [ 0.0207, -0.0365, 0.0022, ..., -0.0634, -0.0427, 0.0215],\n ...,\n [-0.0091, 0.0067, -0.0562, ..., -0.0793, -0.0354, -0.0572],\n [-0.1216, 0.0229, -0.0224, ..., 0.0045, 0.0032, 0.0151],\n [-0.0609, -0.0464, -0.0563, ..., 0.0392, 0.0714, 0.0279]], dtype=torch.float16), '16.to_k_lora.down.weight': tensor([[ 0.0309, 0.1279, 0.0087, ..., 0.0139, -0.0275, -0.0128],\n [-0.0550, 0.0098, 0.0176, ..., -0.0502, -0.0128, -0.0528],\n [-0.0013, 0.0185, 0.0303, ..., 0.0096, -0.0111, 0.0504],\n ...,\n [ 0.0009, 0.0241, 0.0225, ..., 0.0324, -0.0364, 0.0444],\n [ 0.0102, -0.0174, 0.0046, ..., -0.0233, -0.0483, -0.0825],\n [-0.0411, -0.0297, -0.0115, ..., -0.0611, 0.0177, 0.0026]], dtype=torch.float16), '16.to_k_lora.up.weight': tensor([[-0.0334, -0.0386, 0.0292, ..., -0.0398, 0.0111, 0.0329],\n [ 0.0809, -0.0432, 0.0442, ..., -0.0022, -0.0073, 0.0154],\n [-0.0332, 0.0446, -0.0228, ..., -0.0141, -0.0066, -0.0398],\n ...,\n [ 0.0175, 0.0866, -0.0169, ..., -0.0329, -0.0442, 0.0054],\n [ 0.0286, -0.0164, 0.0601, ..., 0.0064, 0.0130, -0.0148],\n [ 0.0143, 0.0573, 0.0402, ..., 0.0510, -0.0060, -0.0479]], dtype=torch.float16), '16.to_v_lora.down.weight': tensor([[-1.9503e-03, -1.4977e-02, 9.1782e-03, ..., 4.4373e-02, 3.8574e-02, -2.6581e-02],\n [ 4.6425e-03, 1.1009e-02, -1.8322e-04, ..., 6.8588e-03, 1.3161e-02, 2.0630e-02],\n [ 4.8767e-02, -3.3783e-02, 1.7128e-03, ..., 2.7740e-02, -1.6373e-02, 5.7892e-02],\n ...,\n [-2.7054e-02, 3.9673e-02, -3.7872e-02, ..., -5.3802e-02, -3.3073e-03, -4.3121e-02],\n [-6.1035e-05, 1.1467e-02, 3.6285e-02, ..., -2.0416e-02, -4.3060e-02, 3.4821e-02],\n [-4.4785e-03, -2.5833e-02, -7.1167e-02, ..., -7.2632e-03, 3.1174e-02, 6.2866e-02]], dtype=torch.float16), '16.to_v_lora.up.weight': tensor([[ 0.0008, 0.0526, -0.0054, ..., 0.0071, -0.0051, -0.0108],\n [-0.0229, 0.0523, 0.0463, ..., 0.0180, -0.0258, -0.0456],\n [ 0.0336, 0.0657, 0.0553, ..., 0.0546, -0.0716, -0.0696],\n ...,\n [-0.0078, 0.0058, -0.0071, ..., -0.0839, -0.0595, -0.0504],\n [ 0.0473, 0.0059, -0.0372, ..., -0.0308, -0.0238, -0.0427],\n [-0.0363, 0.0224, 0.0157, ..., 0.0352, -0.0404, 0.0019]], dtype=torch.float16), '16.to_out_lora.down.weight': tensor([[ 0.0561, -0.0674, 0.0738, ..., -0.0150, -0.0092, -0.0217],\n [ 0.0184, 0.0073, -0.0225, ..., -0.0017, 0.0389, 0.0021],\n [ 0.0024, 0.0345, -0.0583, ..., -0.0161, 0.0210, -0.0381],\n ...,\n [ 0.0038, -0.0356, -0.0137, ..., 0.0177, 0.0502, 0.0156],\n [ 0.0285, 0.0536, -0.0210, ..., 0.0235, 0.0235, 0.0175],\n [-0.0092, 0.0186, 0.0527, ..., -0.0101, 0.0639, 0.0282]], dtype=torch.float16), '16.to_out_lora.up.weight': tensor([[-0.0364, -0.0274, -0.0163, ..., 0.0158, -0.0654, 0.0273],\n [-0.0037, -0.0276, -0.0427, ..., 0.0611, -0.0079, -0.0087],\n [ 0.0037, 0.0526, -0.0442, ..., -0.0558, 0.0685, 0.0107],\n ...,\n [-0.0074, -0.0163, -0.0282, ..., 0.0048, -0.0177, 0.0102],\n [-0.0420, 0.0095, 0.0341, ..., 0.0235, 0.0121, -0.0499],\n [-0.0201, 0.0069, 0.0359, ..., -0.0157, 0.0020, -0.0509]], dtype=torch.float16), '17.to_q_lora.down.weight': tensor([[-0.0027, -0.0341, 0.0181, ..., -0.0494, 0.0489, 0.0178],\n [-0.0024, -0.0235, 0.0056, ..., 0.0127, -0.0621, -0.0067],\n [-0.0119, 0.0076, 0.0588, ..., -0.0108, 0.0194, -0.0208],\n ...,\n [-0.0182, -0.1096, 0.0861, ..., -0.0716, 0.0333, -0.0254],\n [-0.0289, -0.0438, 0.0191, ..., -0.0277, -0.0002, 0.0159],\n [-0.0648, -0.0119, 0.0197, ..., -0.0087, 0.0121, 0.0111]], dtype=torch.float16), '17.to_q_lora.up.weight': tensor([[-3.6438e-02, -1.0529e-02, -2.1072e-02, ..., 1.0773e-02, 3.7689e-03, -2.7649e-02],\n [-1.2932e-02, 2.3079e-03, 2.2873e-02, ..., -2.0157e-02, -1.4938e-02, 2.0248e-02],\n [ 4.1199e-02, -4.5105e-02, -3.7689e-02, ..., 2.0538e-02, 4.7119e-02, 7.9727e-03],\n ...,\n [-2.0874e-02, 2.3849e-02, 2.3758e-02, ..., -2.4872e-02, -2.9465e-02, -3.8204e-03],\n [-6.0028e-02, 2.8610e-02, 3.5065e-02, ..., 7.2777e-05, 7.9163e-02, -5.4199e-02],\n [ 2.4170e-02, 1.8377e-03, -1.9867e-02, ..., -4.4708e-03, -8.1863e-03, -3.0212e-02]], dtype=torch.float16), '17.to_k_lora.down.weight': tensor([[-0.0348, 0.0090, 0.0178, ..., -0.0252, 0.0446, 0.0116],\n [ 0.0127, 0.0115, -0.0473, ..., 0.0154, -0.0278, 0.0552],\n [ 0.0287, 0.0135, 0.0369, ..., -0.0549, -0.0275, 0.0144],\n ...,\n [-0.0325, -0.0251, -0.0536, ..., 0.0431, -0.0431, 0.0231],\n [ 0.0119, -0.0142, -0.0042, ..., -0.0066, 0.0547, -0.0144],\n [-0.0155, 0.0254, 0.0138, ..., 0.0303, -0.0345, -0.0714]], dtype=torch.float16), '17.to_k_lora.up.weight': tensor([[ 0.0551, 0.0158, -0.0251, ..., -0.0315, -0.0255, -0.0266],\n [-0.0003, 0.0186, -0.0515, ..., 0.0127, -0.0276, 0.0640],\n [ 0.0500, -0.0807, -0.1013, ..., -0.0198, 0.0374, -0.0400],\n ...,\n [ 0.0385, -0.0314, 0.0198, ..., -0.0099, 0.0271, -0.0375],\n [-0.1029, -0.0018, -0.0497, ..., 0.0082, -0.0246, 0.0155],\n [-0.0035, 0.0588, 0.0542, ..., -0.0007, -0.0003, -0.0150]], dtype=torch.float16), '17.to_v_lora.down.weight': tensor([[-0.0297, 0.0613, -0.0246, ..., 0.0241, -0.0080, 0.0297],\n [ 0.0089, 0.1199, 0.0818, ..., 0.0181, 0.0004, 0.0351],\n [ 0.0056, 0.0219, 0.0143, ..., -0.0146, -0.0060, 0.0142],\n ...,\n [ 0.0133, -0.0068, -0.0838, ..., -0.0013, 0.0229, 0.0314],\n [ 0.0177, 0.0138, -0.0209, ..., 0.0055, -0.0191, 0.0392],\n [-0.0232, 0.0295, 0.0201, ..., 0.0005, -0.0129, 0.0233]], dtype=torch.float16), '17.to_v_lora.up.weight': tensor([[-0.0061, -0.0201, -0.0364, ..., 0.0020, -0.0072, 0.0723],\n [-0.0245, -0.0371, 0.0406, ..., -0.0031, -0.0116, 0.0322],\n [-0.0245, 0.0279, 0.0043, ..., 0.0086, -0.0049, 0.0032],\n ...,\n [-0.0238, 0.0106, -0.0097, ..., 0.0278, 0.0249, 0.0478],\n [-0.0197, 0.0185, -0.0328, ..., -0.0246, -0.0029, 0.0275],\n [-0.0168, -0.0023, -0.0194, ..., 0.0196, -0.0242, -0.0090]], dtype=torch.float16), '17.to_out_lora.down.weight': tensor([[ 0.0263, 0.0392, 0.0240, ..., -0.0407, 0.0091, 0.0620],\n [ 0.0128, -0.0227, -0.0186, ..., -0.0296, -0.0183, -0.0019],\n [-0.0241, -0.0079, -0.0304, ..., -0.0055, 0.0098, 0.0195],\n ...,\n [-0.0563, 0.0148, 0.0170, ..., 0.0407, -0.0380, -0.0096],\n [ 0.0259, -0.0386, -0.0281, ..., -0.0406, 0.0125, 0.0522],\n [-0.0261, -0.0894, 0.0255, ..., 0.0241, 0.0255, 0.0513]], dtype=torch.float16), '17.to_out_lora.up.weight': tensor([[ 0.0455, 0.0363, -0.0501, ..., -0.0354, -0.0455, 0.0013],\n [ 0.0049, -0.0349, 0.0055, ..., 0.0283, -0.0033, 0.0368],\n [ 0.0045, -0.0524, 0.0116, ..., -0.0473, 0.0484, -0.0266],\n ...,\n [-0.0633, 0.0127, -0.0132, ..., -0.0385, 0.0069, 0.0305],\n [ 0.0126, -0.0739, 0.0263, ..., -0.0137, -0.0213, 0.0284],\n [ 0.0147, -0.0224, 0.0186, ..., -0.0383, 0.0332, -0.0170]], dtype=torch.float16), '17.to_k_ip.weight': tensor([[ 0.0771, 0.1186, 0.0352, ..., -0.0376, -0.0165, 0.0631],\n [-0.0826, -0.0259, -0.0371, ..., 0.0990, 0.0905, -0.0191],\n [-0.0556, -0.0721, -0.0071, ..., -0.0703, 0.0508, 0.0816],\n ...,\n [-0.0936, 0.0450, -0.0002, ..., 0.0043, -0.0300, 0.0231],\n [-0.0686, -0.0283, -0.0126, ..., -0.0152, 0.0946, -0.0204],\n [ 0.0451, -0.0355, -0.0842, ..., 0.1437, -0.0174, 0.0063]], dtype=torch.float16), '17.to_v_ip.weight': tensor([[-0.0537, 0.0021, 0.0670, ..., -0.0063, -0.0781, 0.0275],\n [ 0.0076, 0.0620, 0.0202, ..., 0.0336, 0.0250, -0.0246],\n [ 0.0077, 0.0096, -0.0382, ..., -0.0340, 0.0540, 0.0043],\n ...,\n [ 0.0699, -0.0308, 0.0353, ..., 0.0204, 0.0626, 0.0327],\n [ 0.0129, -0.0848, -0.0764, ..., -0.0177, 0.0430, 0.0242],\n [ 0.0538, -0.0302, 0.0044, ..., 0.0227, 0.0304, 0.0425]], dtype=torch.float16), '18.to_q_lora.down.weight': tensor([[ 0.0165, 0.0320, -0.0222, ..., -0.0501, -0.0077, -0.0330],\n [-0.0350, -0.0061, -0.0066, ..., 0.0119, 0.0656, -0.0131],\n [ 0.0292, -0.0129, -0.0006, ..., -0.0462, 0.0776, -0.0431],\n ...,\n [ 0.0447, -0.0060, -0.0011, ..., -0.0276, -0.0137, 0.0011],\n [ 0.0064, -0.0583, 0.0222, ..., -0.0114, 0.0199, 0.0221],\n [-0.0257, 0.0251, -0.0698, ..., 0.0620, 0.0201, -0.0502]], dtype=torch.float16), '18.to_q_lora.up.weight': tensor([[-0.0500, -0.0051, 0.0447, ..., 0.0287, -0.0525, 0.0407],\n [-0.0451, 0.0007, 0.0230, ..., -0.0267, -0.0041, -0.0246],\n [-0.0417, -0.0891, 0.0072, ..., 0.0411, 0.0065, 0.0385],\n ...,\n [ 0.0293, 0.0470, 0.0084, ..., -0.0063, 0.0422, 0.0284],\n [ 0.1082, 0.0292, -0.0395, ..., -0.0392, -0.0213, 0.0184],\n [-0.0338, -0.0093, -0.0019, ..., -0.0507, -0.0155, 0.0288]], dtype=torch.float16), '18.to_k_lora.down.weight': tensor([[-0.0091, 0.0053, 0.0021, ..., 0.0275, -0.0809, -0.0242],\n [ 0.0350, 0.0288, -0.0267, ..., 0.0030, 0.0063, -0.0276],\n [ 0.0066, 0.0389, 0.0553, ..., -0.0137, -0.0183, -0.0281],\n ...,\n [ 0.0180, -0.0471, 0.0111, ..., 0.0228, 0.0120, -0.0074],\n [ 0.0313, 0.0121, 0.0065, ..., 0.0334, 0.0286, -0.0059],\n [-0.0139, -0.0663, -0.0292, ..., -0.0166, -0.0064, -0.0397]], dtype=torch.float16), '18.to_k_lora.up.weight': tensor([[ 0.0074, -0.0071, -0.0385, ..., -0.0308, -0.0092, -0.0035],\n [ 0.0134, -0.0006, 0.0250, ..., 0.0022, -0.0015, -0.0509],\n [-0.0139, 0.0112, 0.0143, ..., -0.0511, -0.0182, 0.0066],\n ...,\n [-0.0609, -0.0347, 0.0261, ..., -0.0084, -0.0041, -0.0286],\n [-0.0471, -0.0163, -0.0006, ..., -0.0138, -0.0019, -0.0148],\n [-0.0051, 0.0086, 0.0145, ..., 0.0057, -0.0205, -0.0096]], dtype=torch.float16), '18.to_v_lora.down.weight': tensor([[-0.0375, -0.0144, -0.0025, ..., 0.0463, -0.0027, 0.0106],\n [-0.0201, -0.0057, -0.0266, ..., 0.0146, -0.0055, -0.0227],\n [ 0.0233, -0.0189, -0.0192, ..., -0.0362, -0.0504, -0.0267],\n ...,\n [-0.0394, 0.0323, 0.0367, ..., -0.0419, 0.0633, 0.0147],\n [ 0.0112, -0.0094, 0.0092, ..., -0.0410, -0.0232, 0.0079],\n [ 0.0037, -0.0282, -0.0003, ..., -0.0051, -0.0089, -0.0090]], dtype=torch.float16), '18.to_v_lora.up.weight': tensor([[ 0.0056, 0.0112, 0.0042, ..., -0.0141, 0.0012, -0.0064],\n [ 0.0391, 0.0205, 0.0094, ..., -0.0072, 0.0143, -0.0487],\n [-0.0090, -0.0136, 0.0191, ..., 0.0505, 0.0270, 0.0013],\n ...,\n [ 0.0385, 0.0131, 0.0635, ..., 0.0374, 0.0193, -0.0158],\n [-0.0517, 0.0049, 0.0451, ..., -0.0054, -0.0003, 0.0051],\n [ 0.0104, -0.0203, 0.0105, ..., 0.0432, -0.0122, 0.0259]], dtype=torch.float16), '18.to_out_lora.down.weight': tensor([[-0.0297, -0.0198, -0.0392, ..., -0.0322, -0.0038, 0.0048],\n [-0.0275, 0.0104, 0.0162, ..., 0.0050, 0.0046, 0.0187],\n [ 0.0450, 0.0249, 0.0899, ..., -0.0137, 0.0100, 0.0009],\n ...,\n [ 0.0340, -0.0127, 0.0396, ..., 0.0038, -0.0085, -0.0306],\n [-0.0549, 0.0245, 0.1094, ..., 0.0414, -0.0320, 0.0465],\n [ 0.0075, -0.0381, 0.0478, ..., -0.0084, 0.0464, 0.0018]], dtype=torch.float16), '18.to_out_lora.up.weight': tensor([[ 0.0088, 0.0052, 0.0345, ..., 0.0077, -0.0180, 0.0244],\n [-0.0112, 0.0086, 0.0021, ..., 0.0190, -0.0043, 0.0288],\n [ 0.0160, 0.0079, -0.0187, ..., -0.0020, 0.0007, 0.0027],\n ...,\n [ 0.0082, -0.0244, 0.0239, ..., -0.0559, 0.0016, -0.0359],\n [ 0.0155, 0.0108, 0.0039, ..., 0.0072, -0.0332, -0.0106],\n [-0.0003, 0.0288, -0.0311, ..., 0.0662, -0.0032, -0.0309]], dtype=torch.float16), '19.to_q_lora.down.weight': tensor([[ 0.0268, -0.0283, -0.0389, ..., -0.0231, 0.0416, 0.0070],\n [ 0.0448, 0.0536, 0.0396, ..., 0.0121, 0.0277, 0.0417],\n [-0.0459, -0.0489, 0.0232, ..., -0.0511, 0.0443, 0.0307],\n ...,\n [-0.0180, -0.0159, -0.0085, ..., 0.0062, -0.0052, -0.0437],\n [ 0.0430, -0.0543, -0.0407, ..., -0.0189, -0.0410, -0.0007],\n [-0.0481, -0.0500, -0.0519, ..., 0.0574, -0.0223, -0.0106]], dtype=torch.float16), '19.to_q_lora.up.weight': tensor([[-0.0137, 0.0081, -0.0048, ..., 0.0236, -0.0192, 0.0103],\n [ 0.0503, -0.0176, 0.0303, ..., 0.0076, 0.0042, 0.0083],\n [ 0.0356, 0.0101, -0.0461, ..., 0.0113, -0.0243, 0.0029],\n ...,\n [ 0.0083, 0.0113, -0.0134, ..., -0.0711, -0.0332, -0.0114],\n [ 0.0037, 0.0112, 0.0096, ..., 0.0638, 0.0342, -0.0239],\n [ 0.0149, 0.0091, 0.0035, ..., -0.0084, 0.0321, -0.0119]], dtype=torch.float16), '19.to_k_lora.down.weight': tensor([[ 9.5291e-03, 1.2886e-02, -5.1270e-02, ..., -9.9487e-03, 2.7588e-02, 4.5105e-02],\n [ 4.6039e-04, -7.1645e-05, -7.3051e-03, ..., 1.3161e-02, -2.0081e-02, -9.9609e-02],\n [-1.1650e-02, -1.8555e-02, -2.8030e-02, ..., -4.3457e-02, 2.5146e-02, 2.3636e-02],\n ...,\n [-4.3488e-02, 8.9951e-03, 2.6993e-02, ..., -1.3206e-02, 3.2227e-02, -1.2154e-02],\n [-4.9683e-02, 1.0811e-02, 3.3051e-02, ..., -1.9943e-02, -6.7673e-03, -1.7792e-02],\n [-1.3077e-02, 2.1851e-02, -5.2528e-03, ..., -2.3254e-02, 7.2205e-02, 8.3466e-03]], dtype=torch.float16), '19.to_k_lora.up.weight': tensor([[ 0.0158, -0.0406, -0.0162, ..., 0.0328, 0.0142, 0.0011],\n [ 0.0540, -0.0043, 0.0242, ..., 0.0159, 0.0298, 0.0144],\n [-0.0456, 0.0037, 0.0158, ..., 0.0343, -0.0205, -0.0258],\n ...,\n [ 0.0166, -0.0157, -0.0156, ..., -0.0181, 0.0011, -0.0213],\n [ 0.0327, 0.0075, 0.0227, ..., -0.0140, 0.0047, 0.0032],\n [-0.0077, 0.0354, 0.0053, ..., 0.0299, -0.0211, 0.0114]], dtype=torch.float16), '19.to_v_lora.down.weight': tensor([[-0.0272, -0.0004, 0.0136, ..., -0.0027, -0.0102, -0.0060],\n [ 0.0010, -0.0109, -0.0025, ..., 0.0090, 0.0063, 0.0116],\n [-0.0020, 0.0025, -0.0046, ..., -0.0019, 0.0041, -0.0013],\n ...,\n [-0.0018, -0.0236, -0.0203, ..., -0.0091, 0.0160, 0.0141],\n [ 0.0046, -0.0110, 0.0408, ..., 0.0046, 0.0012, 0.0347],\n [ 0.0057, 0.0013, -0.0162, ..., -0.0017, 0.0002, -0.0041]], dtype=torch.float16), '19.to_v_lora.up.weight': tensor([[ 6.1493e-03, -2.5833e-02, -2.9469e-03, ..., 6.5002e-03, 4.6234e-03, 9.5901e-03],\n [ 2.1408e-02, 1.1429e-02, -1.9470e-02, ..., 1.2291e-02, -6.1493e-03, 1.0864e-02],\n [ 6.0081e-03, 1.4198e-04, -5.5580e-03, ..., 2.3895e-02, 2.5539e-03, 1.1475e-02],\n ...,\n [ 2.3651e-02, -6.8321e-03, 6.1214e-05, ..., -1.3847e-02, 2.5616e-03, 7.0305e-03],\n [ 3.6964e-03, 1.6479e-02, -5.2261e-03, ..., -7.2479e-03, 5.1079e-03, 4.2458e-03],\n [-1.4702e-02, -1.1223e-02, -5.8632e-03, ..., 1.7456e-02, -1.2505e-02, 4.5853e-03]], dtype=torch.float16), '19.to_out_lora.down.weight': tensor([[ 0.0640, 0.0300, -0.0583, ..., 0.1260, 0.0228, 0.0173],\n [-0.0126, 0.0289, -0.0699, ..., -0.0149, 0.0878, 0.0123],\n [-0.0497, -0.0234, 0.0207, ..., 0.0439, -0.0074, -0.0005],\n ...,\n [ 0.0077, 0.0036, -0.0132, ..., 0.0103, -0.0525, 0.0440],\n [ 0.0773, 0.0692, -0.0054, ..., -0.0184, -0.0036, 0.0268],\n [-0.0647, -0.0250, -0.0463, ..., 0.0035, -0.0314, -0.0278]], dtype=torch.float16), '19.to_out_lora.up.weight': tensor([[-0.0120, -0.0076, -0.0008, ..., -0.0625, -0.0105, -0.0615],\n [ 0.0197, -0.0563, 0.0141, ..., -0.0057, 0.0083, -0.0424],\n [-0.0954, -0.0108, 0.0098, ..., 0.0688, 0.0174, 0.0318],\n ...,\n [ 0.0588, 0.0180, 0.0482, ..., -0.0518, -0.0323, 0.0446],\n [-0.0013, -0.0541, -0.0187, ..., 0.0275, -0.0203, 0.0208],\n [ 0.0120, -0.0827, -0.0241, ..., 0.0343, -0.0419, -0.0081]], dtype=torch.float16), '19.to_k_ip.weight': tensor([[-0.0055, 0.0013, -0.0226, ..., -0.0262, -0.0167, -0.0558],\n [ 0.0063, -0.0118, 0.0537, ..., 0.0711, -0.0618, -0.0216],\n [-0.0460, 0.0076, 0.0760, ..., 0.0494, 0.1062, 0.0174],\n ...,\n [ 0.0026, 0.0277, 0.0286, ..., 0.0615, -0.0099, -0.0854],\n [-0.0845, -0.0445, 0.0639, ..., 0.0187, 0.1151, 0.0108],\n [ 0.0111, -0.0324, 0.0656, ..., -0.0764, 0.0349, 0.0068]], dtype=torch.float16), '19.to_v_ip.weight': tensor([[ 0.0171, 0.0866, 0.0286, ..., -0.0444, 0.0420, 0.0116],\n [-0.0269, 0.0972, 0.0890, ..., 0.0465, 0.0319, -0.0007],\n [-0.0007, -0.0743, -0.0192, ..., -0.0568, 0.0354, 0.0379],\n ...,\n [-0.0366, -0.0085, -0.0026, ..., 0.0182, 0.0358, 0.0576],\n [-0.0106, -0.0036, 0.0374, ..., 0.0387, 0.0444, -0.0169],\n [-0.0117, -0.0914, -0.0739, ..., 0.0913, -0.0168, -0.0383]], dtype=torch.float16), '20.to_q_lora.down.weight': tensor([[ 2.8778e-02, -1.0132e-02, 2.7256e-03, ..., -5.5695e-03, 9.1782e-03, -1.7899e-02],\n [ 4.2847e-02, 4.5319e-02, 3.7292e-02, ..., 4.2496e-03, 2.0123e-03, 3.6259e-03],\n [ 3.4851e-02, -2.3300e-02, -8.6427e-05, ..., -3.0167e-02, 4.2908e-02, 3.3875e-02],\n ...,\n [ 2.3537e-03, -2.3911e-02, -1.6212e-03, ..., -1.0042e-03, 2.8763e-02, -8.6594e-03],\n [ 6.1615e-02, -1.2856e-02, 5.2734e-02, ..., -5.0293e-02, 1.7288e-02, 4.0436e-04],\n [ 1.8082e-02, -1.5135e-03, -2.5711e-02, ..., 5.4413e-02, -3.0014e-02, -5.8228e-02]], dtype=torch.float16), '20.to_q_lora.up.weight': tensor([[ 0.0386, 0.0006, -0.0124, ..., -0.0119, -0.0216, 0.0053],\n [-0.0181, -0.0318, -0.0157, ..., 0.0616, -0.0012, -0.0157],\n [-0.0015, -0.0445, -0.0653, ..., 0.0520, 0.0511, -0.0194],\n ...,\n [ 0.0273, -0.0036, -0.0056, ..., 0.0129, 0.0238, 0.0667],\n [ 0.0071, -0.0369, 0.0093, ..., 0.0099, 0.0035, 0.0257],\n [-0.0164, -0.0170, 0.0060, ..., 0.0414, -0.0120, 0.0309]], dtype=torch.float16), '20.to_k_lora.down.weight': tensor([[-0.0299, -0.0148, 0.0225, ..., 0.0284, 0.0290, 0.0203],\n [-0.0081, -0.0678, -0.0251, ..., -0.0107, -0.0126, -0.0141],\n [-0.0137, -0.0260, 0.0186, ..., 0.0216, -0.0300, 0.0032],\n ...,\n [ 0.0095, -0.0489, -0.0208, ..., -0.0262, 0.0007, -0.0110],\n [ 0.0285, 0.0064, 0.0278, ..., 0.0352, 0.0411, -0.0064],\n [-0.0249, -0.0004, -0.0536, ..., -0.0175, -0.0109, -0.0284]], dtype=torch.float16), '20.to_k_lora.up.weight': tensor([[ 0.0497, -0.0073, -0.0139, ..., -0.0160, -0.0109, 0.0183],\n [ 0.0220, 0.0105, 0.0302, ..., -0.0009, -0.0154, -0.0065],\n [-0.0055, -0.0013, -0.0079, ..., -0.0354, -0.0333, -0.0342],\n ...,\n [ 0.0420, -0.0169, -0.0275, ..., 0.0053, 0.0001, 0.0194],\n [-0.0017, -0.0230, -0.0114, ..., -0.0049, -0.0494, -0.0104],\n [ 0.0323, 0.0089, -0.0272, ..., -0.0064, -0.0171, 0.0053]], dtype=torch.float16), '20.to_v_lora.down.weight': tensor([[-0.0176, 0.0537, 0.0138, ..., -0.0338, -0.0288, 0.0447],\n [ 0.0092, -0.0119, 0.0027, ..., -0.0031, -0.0457, 0.0071],\n [-0.0047, 0.0018, 0.0165, ..., -0.0084, -0.0005, 0.0412],\n ...,\n [-0.0123, -0.0586, -0.0096, ..., 0.0250, -0.0270, -0.0298],\n [-0.0016, 0.0466, 0.0148, ..., -0.0102, -0.0293, -0.0235],\n [-0.0391, -0.0045, -0.0232, ..., -0.0397, 0.0239, -0.0501]], dtype=torch.float16), '20.to_v_lora.up.weight': tensor([[ 0.0111, 0.0318, 0.0258, ..., -0.0092, -0.0025, 0.0115],\n [ 0.0088, 0.0237, -0.0228, ..., -0.0368, -0.0220, -0.0317],\n [-0.0195, 0.0135, -0.0232, ..., -0.0018, -0.0143, 0.0743],\n ...,\n [ 0.0178, 0.0373, 0.0053, ..., 0.0267, -0.0232, -0.0105],\n [ 0.0257, 0.0311, 0.0268, ..., 0.0016, -0.0073, 0.0059],\n [-0.0020, -0.0042, -0.0260, ..., -0.0173, -0.0127, 0.0609]], dtype=torch.float16), '20.to_out_lora.down.weight': tensor([[-0.0145, -0.0122, -0.0117, ..., 0.0003, -0.0406, -0.0147],\n [-0.0685, 0.0450, 0.0013, ..., -0.0233, -0.0230, 0.0031],\n [-0.0111, -0.0246, -0.0114, ..., 0.0358, 0.0374, -0.0216],\n ...,\n [ 0.0747, 0.0067, -0.0081, ..., -0.0188, -0.0222, 0.0572],\n [-0.0333, -0.0170, 0.0042, ..., -0.0032, 0.0205, 0.0125],\n [ 0.0304, 0.0182, -0.0246, ..., 0.0065, -0.0565, -0.0073]], dtype=torch.float16), '20.to_out_lora.up.weight': tensor([[-0.0663, 0.0056, -0.0154, ..., 0.0019, -0.0164, -0.0365],\n [-0.0107, 0.0686, -0.0168, ..., 0.0031, -0.0106, -0.0574],\n [ 0.0686, -0.0135, 0.0168, ..., -0.0501, 0.0378, 0.0080],\n ...,\n [-0.0410, -0.0262, 0.0166, ..., 0.0050, 0.0173, -0.0050],\n [-0.0242, 0.0301, -0.0123, ..., -0.0066, 0.0541, -0.0184],\n [-0.0648, 0.0150, -0.0502, ..., 0.0252, -0.0701, -0.0185]], dtype=torch.float16), '21.to_q_lora.down.weight': tensor([[-0.0439, 0.0146, 0.0767, ..., -0.0668, 0.0302, 0.0206],\n [-0.0085, 0.0315, -0.0106, ..., 0.0444, 0.0395, -0.0039],\n [-0.0513, 0.0506, 0.0039, ..., 0.0244, -0.0076, -0.0139],\n ...,\n [-0.0106, -0.0367, -0.0735, ..., 0.0239, -0.0016, 0.0057],\n [-0.0364, -0.0360, 0.0216, ..., 0.0082, 0.0300, -0.0618],\n [ 0.0236, 0.0315, -0.0200, ..., 0.0316, 0.0191, -0.0228]], dtype=torch.float16), '21.to_q_lora.up.weight': tensor([[ 0.0424, -0.0729, -0.0740, ..., -0.0030, -0.0376, 0.0340],\n [-0.0530, -0.0110, -0.0325, ..., -0.0394, 0.0071, -0.0842],\n [-0.0613, -0.0262, 0.0489, ..., -0.0019, 0.0132, 0.0174],\n ...,\n [ 0.0430, -0.0054, 0.0079, ..., 0.0148, 0.0350, -0.0333],\n [ 0.0069, -0.0677, -0.0040, ..., -0.0410, -0.0170, 0.0019],\n [ 0.0781, -0.0376, 0.0147, ..., -0.1201, 0.0170, -0.0204]], dtype=torch.float16), '21.to_k_lora.down.weight': tensor([[ 0.0233, -0.0601, 0.0334, ..., -0.0184, 0.0127, 0.0158],\n [-0.0862, -0.0192, 0.0541, ..., -0.0052, -0.0284, 0.0206],\n [ 0.0597, -0.0314, 0.0211, ..., -0.0220, 0.0286, -0.0104],\n ...,\n [ 0.0047, -0.0456, -0.0350, ..., -0.0105, 0.0148, -0.0261],\n [-0.0334, 0.0216, 0.0362, ..., 0.0390, 0.0619, -0.0050],\n [-0.0075, -0.0157, -0.0299, ..., -0.0188, -0.0279, 0.0010]], dtype=torch.float16), '21.to_k_lora.up.weight': tensor([[ 0.0236, 0.0022, -0.0069, ..., 0.0322, 0.0119, -0.0169],\n [ 0.0234, -0.0292, 0.0043, ..., 0.0087, -0.0562, 0.0833],\n [ 0.0143, -0.0501, -0.0267, ..., -0.0120, 0.0053, 0.0337],\n ...,\n [-0.0184, 0.0334, -0.0194, ..., 0.0010, 0.0269, -0.0352],\n [-0.0008, -0.1081, -0.1008, ..., 0.0448, 0.0229, -0.0711],\n [-0.0613, -0.0381, -0.0462, ..., 0.0502, 0.0212, -0.0637]], dtype=torch.float16), '21.to_v_lora.down.weight': tensor([[-0.0163, -0.0224, 0.0146, ..., -0.0117, -0.0033, -0.0117],\n [ 0.0051, -0.0159, 0.0628, ..., -0.0117, 0.0053, -0.0479],\n [-0.0201, 0.0151, -0.0152, ..., -0.0124, -0.0228, 0.0571],\n ...,\n [-0.0173, -0.0271, 0.0019, ..., 0.0067, 0.0054, -0.0100],\n [-0.0156, 0.0323, -0.0643, ..., -0.0184, -0.0208, 0.0137],\n [-0.0085, -0.0002, -0.0650, ..., 0.0081, -0.0347, -0.0426]], dtype=torch.float16), '21.to_v_lora.up.weight': tensor([[ 0.0132, 0.0352, -0.0225, ..., 0.0104, 0.0352, 0.0294],\n [-0.0162, -0.0449, 0.0403, ..., 0.0447, 0.0155, -0.0018],\n [-0.0069, -0.0079, -0.0087, ..., 0.0125, 0.0073, -0.0371],\n ...,\n [-0.0228, 0.0092, 0.0040, ..., 0.0012, 0.0098, -0.0130],\n [ 0.0047, 0.0458, -0.0275, ..., -0.0217, -0.0207, -0.0388],\n [ 0.0107, 0.0046, -0.0283, ..., 0.0365, -0.0020, 0.0271]], dtype=torch.float16), '21.to_out_lora.down.weight': tensor([[ 0.0227, -0.0292, -0.0874, ..., 0.0230, -0.0121, -0.0794],\n [ 0.0475, 0.0419, 0.0332, ..., -0.0024, 0.0344, 0.0464],\n [-0.0581, -0.0380, -0.0206, ..., 0.0022, -0.0051, 0.0440],\n ...,\n [ 0.0514, 0.0152, 0.0423, ..., 0.0552, 0.0859, 0.0150],\n [-0.0264, -0.0778, -0.0397, ..., -0.0481, -0.0269, -0.0315],\n [-0.0218, 0.0165, -0.0564, ..., 0.0099, -0.0597, -0.0186]], dtype=torch.float16), '21.to_out_lora.up.weight': tensor([[-0.0467, -0.0526, 0.0422, ..., -0.0034, 0.0023, -0.0355],\n [-0.0134, 0.0667, 0.0372, ..., -0.0605, 0.0246, -0.0005],\n [-0.0256, 0.0251, -0.0416, ..., -0.0064, 0.0043, -0.0919],\n ...,\n [ 0.0289, 0.0127, 0.0329, ..., -0.0417, 0.0487, 0.0069],\n [ 0.0099, 0.0119, 0.0449, ..., -0.0728, 0.0833, -0.0349],\n [-0.0324, -0.0123, 0.0001, ..., -0.0211, -0.0549, 0.0156]], dtype=torch.float16), '21.to_k_ip.weight': tensor([[-0.0684, -0.0425, -0.0553, ..., -0.0514, 0.0385, -0.0999],\n [ 0.0274, -0.0231, 0.0065, ..., 0.0410, 0.0800, -0.0859],\n [-0.0495, 0.0453, 0.0130, ..., 0.0681, 0.0042, -0.0649],\n ...,\n [-0.0127, 0.0597, 0.0166, ..., -0.0195, -0.0573, -0.0345],\n [ 0.0398, 0.0229, 0.0218, ..., -0.0568, 0.0109, 0.0321],\n [ 0.0310, -0.0337, -0.0583, ..., 0.0811, -0.0125, -0.0429]], dtype=torch.float16), '21.to_v_ip.weight': tensor([[-0.0067, -0.0698, -0.0584, ..., 0.0301, -0.0522, 0.0159],\n [-0.0308, -0.0358, 0.0017, ..., -0.0315, 0.0295, -0.0098],\n [-0.0570, 0.0051, 0.0353, ..., 0.0061, 0.0232, 0.0472],\n ...,\n [-0.0543, -0.0155, -0.1055, ..., 0.0177, -0.0533, 0.0091],\n [-0.0208, -0.0594, 0.0195, ..., 0.0547, -0.0046, -0.0786],\n [ 0.0377, 0.0217, -0.0314, ..., -0.0471, -0.0018, 0.0374]], dtype=torch.float16), '22.to_q_lora.down.weight': tensor([[-0.0302, -0.0073, -0.0587, ..., 0.0801, -0.0399, 0.0079],\n [-0.0337, -0.0020, 0.0513, ..., -0.0340, 0.0044, -0.0673],\n [ 0.0504, 0.0448, 0.0044, ..., -0.0287, 0.0227, -0.0222],\n ...,\n [ 0.0023, -0.0162, -0.0682, ..., 0.0423, -0.0660, -0.0380],\n [ 0.0721, -0.0322, -0.0434, ..., 0.0117, -0.0098, 0.0062],\n [ 0.0031, -0.0584, -0.0499, ..., 0.0347, 0.0262, 0.0337]], dtype=torch.float16), '22.to_q_lora.up.weight': tensor([[-0.0648, -0.0396, 0.0005, ..., -0.0176, -0.0078, 0.0109],\n [ 0.0186, -0.0312, 0.0450, ..., 0.0372, -0.0183, 0.0005],\n [-0.0183, 0.0113, -0.0184, ..., -0.0055, 0.0120, 0.0114],\n ...,\n [ 0.0172, -0.0092, 0.0479, ..., 0.0643, -0.0066, 0.0013],\n [-0.0304, -0.0093, -0.0269, ..., 0.0608, 0.0174, 0.0358],\n [-0.0462, -0.0340, 0.0104, ..., 0.0173, 0.0226, 0.0587]], dtype=torch.float16), '22.to_k_lora.down.weight': tensor([[ 0.0105, -0.0084, -0.0228, ..., -0.0434, -0.0432, -0.0140],\n [-0.0065, 0.0206, -0.0564, ..., 0.0369, -0.0175, 0.0226],\n [-0.0005, -0.0886, 0.0436, ..., -0.0010, -0.0291, 0.0509],\n ...,\n [-0.0175, 0.0645, -0.0015, ..., -0.0571, 0.0187, -0.0331],\n [-0.0089, -0.0090, 0.0279, ..., 0.0302, 0.0144, 0.0317],\n [ 0.0272, 0.0483, 0.0364, ..., 0.0740, 0.0255, 0.0306]], dtype=torch.float16), '22.to_k_lora.up.weight': tensor([[-0.0457, -0.0183, -0.0443, ..., 0.0047, -0.0363, -0.0355],\n [ 0.0109, -0.0394, 0.0096, ..., 0.0439, 0.0200, -0.0038],\n [ 0.0112, 0.0118, 0.0342, ..., 0.0009, -0.0027, -0.0215],\n ...,\n [ 0.0797, -0.0091, 0.0133, ..., 0.0646, -0.0056, 0.0656],\n [-0.0468, -0.0127, 0.0073, ..., 0.0709, -0.1467, -0.0025],\n [ 0.0139, -0.0775, 0.0044, ..., 0.0396, 0.0440, 0.0583]], dtype=torch.float16), '22.to_v_lora.down.weight': tensor([[-0.0515, 0.0102, -0.0009, ..., -0.0103, 0.0643, -0.0014],\n [ 0.0312, -0.0229, 0.0364, ..., -0.0337, -0.0337, 0.0099],\n [-0.0559, -0.0246, 0.0219, ..., 0.0185, 0.0451, 0.0252],\n ...,\n [-0.0240, -0.0098, 0.0096, ..., 0.0223, -0.0208, -0.0540],\n [ 0.0125, 0.0225, 0.0235, ..., -0.0802, -0.0521, -0.0143],\n [ 0.0102, 0.0226, -0.0604, ..., 0.0425, -0.0205, -0.0168]], dtype=torch.float16), '22.to_v_lora.up.weight': tensor([[-0.0702, 0.0069, -0.0636, ..., -0.0300, 0.0147, 0.0257],\n [-0.0506, 0.0544, -0.0325, ..., -0.0173, 0.0413, -0.0437],\n [-0.0527, -0.0171, -0.0453, ..., 0.0213, -0.0138, -0.0155],\n ...,\n [-0.0415, -0.0097, 0.0250, ..., 0.0225, 0.0163, 0.0029],\n [ 0.0395, -0.0450, 0.0163, ..., -0.0403, 0.0318, -0.0023],\n [-0.0271, 0.0137, -0.0339, ..., -0.0220, 0.0375, 0.0509]], dtype=torch.float16), '22.to_out_lora.down.weight': tensor([[-2.4490e-02, 5.2338e-02, 9.8419e-03, ..., -1.3756e-02, -7.4029e-05, 1.9196e-02],\n [-2.2141e-02, -3.1921e-02, -6.1264e-03, ..., 2.2888e-02, -4.6356e-02, 6.4011e-03],\n [ 2.4780e-02, 2.2964e-02, -2.5177e-03, ..., -1.4206e-02, 1.8921e-02, -3.1113e-02],\n ...,\n [-2.7130e-02, -4.8920e-02, -5.2948e-02, ..., -2.8030e-02, 2.2476e-02, -1.2146e-02],\n [ 6.3293e-02, 2.7939e-02, 4.8523e-03, ..., -1.6815e-02, -1.4656e-02, 2.3697e-02],\n [-2.5345e-02, -3.8643e-03, 3.7933e-02, ..., 2.3102e-02, 2.1896e-03, 3.0441e-02]], dtype=torch.float16), '22.to_out_lora.up.weight': tensor([[ 0.0247, -0.0428, -0.0638, ..., 0.0186, -0.0110, -0.0017],\n [ 0.0670, -0.0242, -0.0036, ..., 0.0283, -0.0019, -0.0389],\n [-0.0171, -0.0031, -0.0160, ..., 0.0176, -0.0207, 0.0131],\n ...,\n [ 0.0091, -0.0269, -0.0231, ..., 0.0351, -0.0522, 0.0328],\n [-0.0231, -0.0275, 0.0199, ..., 0.0004, -0.0204, -0.0179],\n [-0.0232, -0.0226, -0.0237, ..., -0.0125, 0.0013, -0.0224]], dtype=torch.float16), '23.to_q_lora.down.weight': tensor([[ 0.0274, -0.0319, -0.0488, ..., 0.0312, 0.0544, 0.0564],\n [ 0.0106, 0.0068, 0.0250, ..., 0.0195, 0.0072, -0.0485],\n [ 0.0158, 0.0152, 0.0504, ..., -0.0132, -0.0092, 0.0158],\n ...,\n [ 0.0038, 0.0177, -0.0069, ..., 0.0283, 0.0234, -0.0084],\n [ 0.0045, 0.0117, 0.0042, ..., 0.0101, 0.0408, -0.0588],\n [-0.0509, 0.0149, -0.0131, ..., -0.0603, 0.0356, 0.0847]], dtype=torch.float16), '23.to_q_lora.up.weight': tensor([[-0.0253, -0.0584, 0.0001, ..., -0.0749, 0.0051, 0.0161],\n [ 0.0077, 0.0097, 0.0142, ..., 0.0198, 0.0150, 0.0304],\n [-0.0059, -0.0476, 0.0083, ..., 0.0214, 0.0396, -0.0892],\n ...,\n [ 0.0449, -0.0706, 0.0101, ..., 0.0388, 0.0257, -0.0170],\n [ 0.0216, -0.0085, 0.0652, ..., 0.0322, -0.0450, -0.0096],\n [-0.0730, -0.0128, -0.0063, ..., -0.0252, 0.0954, -0.0418]], dtype=torch.float16), '23.to_k_lora.down.weight': tensor([[-4.3213e-02, -1.7563e-02, -8.1177e-03, ..., 3.7262e-02, 3.0231e-03, -1.1635e-02],\n [ 9.6054e-03, 2.4734e-02, -7.7248e-04, ..., -1.3710e-02, -4.4220e-02, 1.2726e-02],\n [-1.4381e-02, 4.1626e-02, 6.9702e-02, ..., -3.3295e-02, 7.2784e-03, -2.7239e-05],\n ...,\n [-1.0826e-02, -2.8748e-02, -5.0262e-02, ..., 4.2786e-02, -3.0472e-02, -3.2440e-02],\n [ 4.9438e-03, -2.5131e-02, -1.9531e-03, ..., 4.0039e-02, -4.5288e-02, 3.8177e-02],\n [ 4.1016e-02, -6.1157e-02, 2.3056e-02, ..., 4.0680e-02, -1.8950e-03, -1.9348e-02]], dtype=torch.float16), '23.to_k_lora.up.weight': tensor([[ 0.0931, -0.0067, -0.0192, ..., 0.0031, -0.0933, -0.0015],\n [-0.0462, 0.0385, -0.0240, ..., -0.0290, -0.0252, 0.0409],\n [-0.0260, -0.0284, -0.0469, ..., -0.0380, 0.0318, 0.0091],\n ...,\n [ 0.0004, -0.0564, -0.0162, ..., 0.0464, 0.0263, 0.0026],\n [-0.0535, -0.0325, -0.0317, ..., 0.0410, -0.0368, -0.0202],\n [ 0.0810, -0.0347, 0.0446, ..., 0.0164, 0.0308, 0.0023]], dtype=torch.float16), '23.to_v_lora.down.weight': tensor([[-0.0218, 0.0457, 0.0589, ..., 0.0144, -0.0057, 0.0153],\n [-0.0242, -0.0017, -0.0541, ..., 0.0015, 0.0188, -0.0023],\n [-0.0053, -0.0120, -0.0323, ..., 0.0056, 0.0094, -0.0306],\n ...,\n [-0.0003, 0.0035, -0.0093, ..., -0.0158, 0.0199, -0.0009],\n [ 0.0072, -0.0211, -0.0225, ..., 0.0154, 0.0113, -0.1106],\n [ 0.0181, 0.0139, -0.0569, ..., -0.0095, 0.0164, -0.0273]], dtype=torch.float16), '23.to_v_lora.up.weight': tensor([[-0.0133, -0.0032, -0.0088, ..., 0.0066, -0.0156, -0.0161],\n [ 0.0256, 0.0327, 0.0344, ..., 0.0140, -0.0210, -0.0066],\n [-0.0165, 0.0182, 0.0031, ..., -0.0058, -0.0406, -0.0025],\n ...,\n [-0.0105, -0.0275, -0.0076, ..., 0.0471, -0.0188, 0.0403],\n [-0.0054, -0.0327, -0.0023, ..., -0.0269, 0.0355, -0.0245],\n [-0.0024, -0.0175, 0.0230, ..., 0.0071, 0.0085, 0.0141]], dtype=torch.float16), '23.to_out_lora.down.weight': tensor([[ 0.0169, -0.0471, 0.0068, ..., 0.0515, -0.0859, 0.0424],\n [ 0.0130, -0.0324, 0.0459, ..., -0.0411, 0.0345, -0.0196],\n [ 0.0003, -0.0098, 0.0005, ..., 0.0035, -0.0078, 0.0110],\n ...,\n [ 0.0223, 0.0056, -0.0560, ..., -0.0076, 0.0515, 0.0129],\n [ 0.0101, -0.0288, -0.0410, ..., -0.0162, 0.0341, 0.0057],\n [ 0.0081, 0.0136, -0.0249, ..., -0.0075, 0.0356, 0.0313]], dtype=torch.float16), '23.to_out_lora.up.weight': tensor([[ 0.0511, -0.0341, 0.0291, ..., -0.0546, -0.0136, 0.0076],\n [ 0.0060, -0.0026, -0.0728, ..., 0.0169, -0.0322, 0.0656],\n [ 0.0492, -0.0728, 0.0128, ..., 0.0352, -0.0621, -0.0322],\n ...,\n [-0.0119, 0.0129, -0.0286, ..., 0.0101, -0.0252, 0.0136],\n [-0.0046, -0.0028, 0.0451, ..., -0.0486, -0.0030, -0.0238],\n [ 0.0201, 0.0321, 0.0142, ..., -0.0110, 0.0229, -0.0080]], dtype=torch.float16), '23.to_k_ip.weight': tensor([[ 0.0480, -0.0283, 0.0183, ..., -0.0164, -0.0476, 0.0313],\n [ 0.0269, -0.0999, -0.0815, ..., 0.0266, 0.0038, -0.0084],\n [ 0.0225, -0.0883, 0.0318, ..., 0.0546, 0.0020, 0.0441],\n ...,\n [-0.0046, -0.0180, -0.0649, ..., -0.0618, 0.0721, -0.0052],\n [ 0.0020, -0.0208, 0.0380, ..., 0.0236, 0.0255, 0.0676],\n [ 0.0797, 0.1193, -0.0069, ..., 0.0058, 0.0695, -0.0043]], dtype=torch.float16), '23.to_v_ip.weight': tensor([[-0.0299, -0.0095, 0.0407, ..., 0.0397, 0.0448, -0.0011],\n [ 0.0219, -0.0190, 0.0581, ..., -0.0061, -0.0052, 0.0304],\n [ 0.0546, -0.0491, 0.0601, ..., -0.0107, -0.0315, -0.0110],\n ...,\n [-0.0771, -0.0478, -0.0086, ..., 0.0403, -0.0024, -0.0896],\n [-0.0880, -0.0552, 0.0675, ..., -0.0144, 0.0424, -0.0227],\n [-0.0325, -0.0343, -0.0263, ..., 0.0128, 0.0417, 0.0430]], dtype=torch.float16), '24.to_q_lora.down.weight': tensor([[-0.0383, 0.0090, -0.0484, ..., 0.0420, -0.0720, -0.0046],\n [-0.0271, 0.0753, -0.0451, ..., -0.0457, -0.0341, -0.0886],\n [-0.0905, -0.0493, -0.0183, ..., 0.0312, 0.0006, 0.0464],\n ...,\n [ 0.0205, -0.0237, -0.0093, ..., -0.0051, -0.0156, 0.0566],\n [ 0.0140, -0.0181, 0.0166, ..., 0.0419, -0.0434, 0.1016],\n [-0.0792, 0.0463, 0.0007, ..., -0.0044, -0.0427, -0.0288]], dtype=torch.float16), '24.to_q_lora.up.weight': tensor([[ 0.0233, 0.0311, 0.0555, ..., 0.0083, 0.0395, -0.0236],\n [ 0.0181, 0.0198, -0.0836, ..., -0.0061, 0.0324, -0.0497],\n [ 0.0395, 0.0154, -0.0563, ..., 0.0334, -0.0211, 0.0331],\n ...,\n [-0.0297, -0.0338, -0.0171, ..., -0.0432, 0.0060, -0.0133],\n [ 0.0318, -0.0202, -0.0801, ..., -0.0270, 0.0195, -0.0685],\n [ 0.0425, -0.0334, 0.0482, ..., -0.0944, -0.0198, -0.0182]], dtype=torch.float16), '24.to_k_lora.down.weight': tensor([[-3.8269e-02, 5.0354e-02, 3.3417e-02, ..., -1.7075e-02, 3.6678e-03, -1.6312e-02],\n [-8.3847e-03, -9.7427e-03, -3.7689e-02, ..., -3.5858e-02, 2.1553e-03, 1.8814e-02],\n [-2.6531e-03, -2.6260e-02, -1.4793e-02, ..., 4.5898e-02, 1.2207e-02, -7.9651e-03],\n ...,\n [-2.2769e-05, 1.2161e-02, 5.7343e-02, ..., 6.6681e-03, -5.1483e-02, 1.6769e-02],\n [-5.4779e-02, -1.0918e-02, -5.0934e-02, ..., 4.2419e-03, -6.8542e-02, -7.1655e-02],\n [ 9.2850e-03, 2.5192e-02, 3.8940e-02, ..., -1.4931e-02, -8.0261e-02, 1.2756e-02]], dtype=torch.float16), '24.to_k_lora.up.weight': tensor([[-0.0118, -0.0701, -0.0480, ..., -0.0439, -0.0091, 0.0916],\n [-0.0460, 0.0334, 0.0033, ..., -0.0025, 0.0209, 0.0325],\n [-0.0140, -0.0216, 0.0740, ..., 0.0047, -0.0027, 0.0261],\n ...,\n [ 0.0209, 0.0337, 0.0177, ..., 0.0242, -0.0056, -0.0431],\n [ 0.0494, 0.0175, 0.0623, ..., 0.0406, -0.0438, -0.0145],\n [ 0.0587, -0.0429, 0.0770, ..., -0.0529, 0.0246, 0.0091]], dtype=torch.float16), '24.to_v_lora.down.weight': tensor([[-2.4445e-02, -3.4912e-02, 1.4015e-02, ..., -2.8793e-02, -5.6793e-02, 4.1840e-02],\n [-2.4147e-03, -7.7454e-02, 3.7109e-02, ..., 6.0120e-03, -5.5237e-03, -6.7505e-02],\n [-1.1719e-02, -3.5370e-02, -4.6997e-03, ..., -5.3131e-02, 4.3732e-02, -5.2795e-02],\n ...,\n [-6.7383e-02, -1.9684e-02, -3.8330e-02, ..., -2.5360e-02, -5.8502e-02, 1.3870e-02],\n [ 5.0163e-03, -2.1484e-02, -5.3942e-05, ..., 3.0899e-02, 1.5106e-03, 6.2683e-02],\n [-3.3905e-02, -5.5023e-02, -3.8757e-02, ..., -3.5492e-02, -8.9417e-03, 1.5230e-03]], dtype=torch.float16), '24.to_v_lora.up.weight': tensor([[ 0.0071, 0.0110, 0.0340, ..., 0.0063, 0.0074, -0.0640],\n [-0.0109, 0.0073, -0.0092, ..., 0.0205, -0.0129, 0.0420],\n [ 0.0303, -0.0199, -0.0391, ..., -0.0221, 0.0143, 0.0414],\n ...,\n [-0.0095, -0.0249, -0.0307, ..., 0.0176, 0.0133, -0.0108],\n [-0.0226, 0.0249, -0.0083, ..., -0.0179, 0.0563, 0.0075],\n [ 0.0131, -0.0543, 0.0578, ..., 0.0470, -0.0007, -0.0102]], dtype=torch.float16), '24.to_out_lora.down.weight': tensor([[-0.0127, 0.0248, -0.0058, ..., -0.0381, 0.0115, 0.0291],\n [-0.0170, -0.0317, -0.0237, ..., -0.0257, -0.0094, -0.0204],\n [ 0.0523, -0.0211, -0.0247, ..., -0.0207, -0.0638, -0.0727],\n ...,\n [-0.0051, 0.0147, 0.0204, ..., 0.0025, -0.0168, -0.0149],\n [ 0.0314, -0.0588, 0.0008, ..., -0.0087, 0.0094, 0.0095],\n [-0.0017, -0.0187, -0.0432, ..., 0.0047, -0.0271, -0.0214]], dtype=torch.float16), '24.to_out_lora.up.weight': tensor([[ 0.0057, 0.0183, 0.0089, ..., -0.0319, 0.0145, 0.0210],\n [-0.0104, 0.0077, -0.0107, ..., 0.0219, 0.0206, -0.0160],\n [ 0.0019, 0.0109, -0.0135, ..., 0.0338, 0.0223, 0.0064],\n ...,\n [ 0.0181, 0.0286, -0.0618, ..., 0.0225, -0.0283, -0.0263],\n [ 0.0190, -0.0254, 0.0090, ..., -0.0035, 0.0088, 0.0126],\n [-0.0009, -0.0276, -0.0144, ..., 0.0062, 0.0056, -0.0138]], dtype=torch.float16), '25.to_q_lora.down.weight': tensor([[-0.0213, -0.0325, 0.0391, ..., 0.0065, 0.0449, 0.0103],\n [-0.0105, -0.0616, -0.0487, ..., -0.0076, -0.0215, 0.0258],\n [ 0.0530, -0.0090, 0.0749, ..., -0.0355, 0.0374, 0.0180],\n ...,\n [ 0.0157, 0.0085, -0.0578, ..., -0.0255, -0.0021, 0.0728],\n [ 0.0401, -0.0664, -0.0239, ..., -0.0060, 0.0522, -0.0050],\n [ 0.0085, -0.0622, 0.0365, ..., 0.0595, 0.0650, -0.0281]], dtype=torch.float16), '25.to_q_lora.up.weight': tensor([[-0.0086, 0.0165, 0.0607, ..., 0.0381, -0.0032, -0.0505],\n [ 0.0033, 0.0511, -0.0978, ..., 0.0075, 0.0034, -0.0484],\n [ 0.0181, 0.0373, 0.0088, ..., 0.0317, -0.0133, 0.0217],\n ...,\n [-0.0048, 0.0101, -0.0069, ..., -0.0060, -0.0110, -0.0001],\n [-0.0270, -0.0161, 0.0067, ..., -0.0016, -0.0170, -0.0187],\n [ 0.0151, 0.0547, 0.0074, ..., -0.0137, 0.0222, -0.0080]], dtype=torch.float16), '25.to_k_lora.down.weight': tensor([[-0.0240, -0.0397, -0.0194, ..., 0.0377, 0.0179, -0.0151],\n [-0.0210, 0.0358, -0.0010, ..., 0.0045, 0.0056, 0.0089],\n [ 0.0065, 0.0103, 0.0261, ..., -0.0153, 0.0707, -0.0369],\n ...,\n [ 0.0307, 0.0182, 0.0197, ..., -0.0125, -0.0259, -0.0143],\n [-0.0240, -0.0055, 0.0410, ..., -0.0053, -0.0202, 0.0091],\n [ 0.0194, 0.0041, -0.0244, ..., 0.0066, 0.0092, -0.0513]], dtype=torch.float16), '25.to_k_lora.up.weight': tensor([[-0.0210, -0.0226, -0.0670, ..., 0.0383, 0.0292, -0.0049],\n [-0.0397, -0.0499, -0.0620, ..., 0.0350, 0.0170, 0.0146],\n [ 0.0176, 0.0108, -0.0300, ..., 0.0356, -0.0241, -0.0224],\n ...,\n [-0.0121, -0.0091, -0.0089, ..., 0.0026, -0.0087, 0.0157],\n [ 0.0092, 0.0156, -0.0035, ..., -0.0293, 0.0071, 0.0156],\n [-0.0313, 0.0311, -0.0410, ..., -0.0141, 0.0282, 0.0115]], dtype=torch.float16), '25.to_v_lora.down.weight': tensor([[ 0.0105, 0.0531, -0.0123, ..., 0.0140, -0.0289, 0.0266],\n [-0.0263, 0.0149, 0.0073, ..., -0.0167, -0.0068, 0.0143],\n [-0.0083, -0.0231, -0.0116, ..., 0.0032, -0.0168, 0.0122],\n ...,\n [-0.0277, -0.0497, 0.0048, ..., -0.0247, -0.0054, -0.0010],\n [-0.0322, 0.0424, -0.0175, ..., -0.0153, -0.0011, 0.0006],\n [-0.0048, 0.0112, -0.0098, ..., 0.0179, 0.0136, 0.0331]], dtype=torch.float16), '25.to_v_lora.up.weight': tensor([[ 0.0114, 0.0195, -0.0179, ..., -0.0174, -0.0172, -0.0003],\n [ 0.0251, -0.0011, 0.0098, ..., -0.0187, -0.0335, 0.0123],\n [-0.0015, 0.0023, -0.0103, ..., 0.0366, -0.0017, 0.0200],\n ...,\n [-0.0193, -0.0257, 0.0123, ..., -0.0210, 0.0010, 0.0161],\n [-0.0226, 0.0119, -0.0164, ..., -0.0170, -0.0014, 0.0123],\n [-0.0160, 0.0005, 0.0240, ..., -0.0360, 0.0147, -0.0119]], dtype=torch.float16), '25.to_out_lora.down.weight': tensor([[ 0.0175, -0.0015, -0.0122, ..., 0.0278, 0.0402, -0.0099],\n [ 0.0134, -0.0122, 0.0039, ..., -0.0060, 0.0077, -0.0271],\n [ 0.0007, -0.0270, -0.0012, ..., 0.0130, -0.0386, -0.0580],\n ...,\n [ 0.0145, -0.0049, 0.0413, ..., 0.0044, -0.0244, 0.0001],\n [ 0.0020, 0.0475, -0.0280, ..., 0.0352, -0.0423, -0.0031],\n [-0.0024, 0.0282, 0.0210, ..., 0.0131, -0.0020, -0.0151]], dtype=torch.float16), '25.to_out_lora.up.weight': tensor([[-0.0059, 0.0022, -0.0076, ..., 0.0123, -0.0026, -0.0415],\n [ 0.0073, 0.0223, -0.0168, ..., -0.0510, 0.0299, -0.0553],\n [-0.0527, -0.0071, 0.0110, ..., 0.0142, -0.0020, 0.0016],\n ...,\n [-0.0114, 0.0203, -0.0240, ..., 0.0061, 0.0019, 0.0017],\n [ 0.0037, -0.0017, 0.0166, ..., -0.0004, 0.0226, -0.0253],\n [ 0.0010, 0.0045, -0.0347, ..., -0.0375, -0.0359, 0.0202]], dtype=torch.float16), '25.to_k_ip.weight': tensor([[-1.8372e-02, 3.4271e-02, 1.0437e-04, ..., -7.0686e-03, -8.1421e-02, -4.2389e-02],\n [-3.3295e-02, 5.0140e-02, 1.7776e-02, ..., -6.8176e-02, -6.2500e-02, 3.3569e-02],\n [ 5.2887e-02, -1.2598e-01, 1.7700e-02, ..., 3.2104e-02, -1.1292e-02, 9.3079e-03],\n ...,\n [-6.2012e-02, -8.1711e-03, 3.8086e-02, ..., 3.5973e-03, 2.8915e-02, -1.7090e-02],\n [ 2.1515e-02, 4.2480e-02, -9.5398e-02, ..., -7.8613e-02, 5.1003e-03, -5.9174e-02],\n [ 6.7688e-02, -1.2598e-01, -3.6011e-02, ..., 3.7140e-02, -5.1666e-02, 2.5360e-02]], dtype=torch.float16), '25.to_v_ip.weight': tensor([[-0.0172, -0.0059, 0.0124, ..., -0.0181, -0.0182, -0.0110],\n [-0.0059, -0.0151, -0.0214, ..., 0.0361, -0.0385, 0.0008],\n [-0.0268, -0.0545, -0.0420, ..., 0.0011, -0.0677, -0.0341],\n ...,\n [ 0.0248, 0.0039, 0.0264, ..., -0.0052, -0.0199, 0.0103],\n [-0.0421, 0.0344, -0.0388, ..., -0.0205, 0.0197, 0.0472],\n [-0.0005, 0.0272, 0.0099, ..., -0.0002, -0.0421, 0.0517]], dtype=torch.float16), '26.to_q_lora.down.weight': tensor([[-0.0019, 0.0508, 0.0072, ..., -0.0303, 0.0100, 0.0084],\n [-0.0315, 0.0409, 0.0303, ..., -0.0208, -0.0251, 0.0242],\n [ 0.0271, -0.0149, 0.0230, ..., -0.0073, -0.0181, -0.0045],\n ...,\n [-0.0084, 0.0580, 0.0598, ..., 0.0597, -0.0082, 0.0216],\n [-0.0138, -0.0645, 0.0323, ..., 0.0043, 0.0373, -0.0255],\n [-0.0216, 0.0068, 0.0342, ..., -0.0309, -0.0173, -0.0453]], dtype=torch.float16), '26.to_q_lora.up.weight': tensor([[ 0.0911, 0.0204, 0.0147, ..., 0.0227, 0.1025, 0.0155],\n [-0.0129, -0.0168, 0.0406, ..., -0.0011, 0.0530, -0.0372],\n [ 0.0170, 0.0125, 0.0764, ..., -0.0218, 0.0324, 0.0036],\n ...,\n [-0.0190, -0.1039, -0.0581, ..., 0.0282, 0.0845, -0.0101],\n [-0.0018, 0.0874, -0.0300, ..., 0.0312, -0.0417, -0.0664],\n [-0.0186, 0.0168, 0.0461, ..., 0.0384, -0.0149, -0.0032]], dtype=torch.float16), '26.to_k_lora.down.weight': tensor([[ 0.0119, 0.0019, 0.0361, ..., -0.0024, -0.0038, -0.0143],\n [ 0.0066, -0.0278, -0.0579, ..., -0.0119, 0.0087, 0.0101],\n [-0.0245, -0.0520, 0.0728, ..., 0.0185, -0.0106, -0.0525],\n ...,\n [-0.0236, -0.0056, -0.0109, ..., -0.0220, 0.0164, 0.0180],\n [-0.0046, -0.0677, -0.0161, ..., -0.0116, 0.0302, 0.0364],\n [ 0.0047, -0.0243, -0.0397, ..., -0.0002, 0.0131, -0.0026]], dtype=torch.float16), '26.to_k_lora.up.weight': tensor([[ 0.0285, -0.0619, -0.0249, ..., 0.0445, 0.0037, -0.0068],\n [-0.0539, 0.0340, -0.0280, ..., -0.0095, 0.0529, 0.0216],\n [ 0.0138, 0.0573, 0.0303, ..., 0.0302, -0.0084, -0.0238],\n ...,\n [ 0.0235, 0.1015, -0.0677, ..., 0.0112, 0.0729, -0.0410],\n [ 0.0082, 0.0504, -0.0009, ..., 0.0080, -0.0061, 0.0162],\n [ 0.0389, -0.0530, 0.0453, ..., 0.0350, 0.0580, -0.0080]], dtype=torch.float16), '26.to_v_lora.down.weight': tensor([[ 0.0338, -0.0208, 0.0351, ..., 0.0063, -0.0010, -0.0317],\n [-0.0011, 0.0554, 0.0027, ..., 0.0132, 0.0080, 0.0125],\n [-0.0066, 0.0144, -0.0344, ..., -0.0204, 0.0367, -0.0166],\n ...,\n [-0.0396, 0.0055, -0.0126, ..., -0.0168, 0.0342, 0.0164],\n [ 0.0278, -0.0150, -0.0046, ..., 0.0032, -0.0226, 0.0012],\n [ 0.0032, -0.0087, 0.0272, ..., 0.0272, -0.0117, 0.0214]], dtype=torch.float16), '26.to_v_lora.up.weight': tensor([[ 0.0173, -0.0047, -0.0108, ..., 0.0017, -0.0375, 0.0175],\n [ 0.0145, -0.0112, 0.0140, ..., 0.0046, 0.0197, -0.0165],\n [ 0.0030, -0.0086, 0.0360, ..., 0.0090, -0.0214, 0.0144],\n ...,\n [-0.0086, 0.0070, 0.0076, ..., -0.0424, -0.0197, 0.0097],\n [-0.0479, 0.0430, 0.0209, ..., -0.0173, -0.0126, -0.0461],\n [ 0.0285, 0.0022, 0.0129, ..., 0.0160, 0.0217, 0.0034]], dtype=torch.float16), '26.to_out_lora.down.weight': tensor([[-2.0554e-02, 8.8577e-03, -9.4757e-03, ..., -1.9272e-02, 4.2511e-02, -7.6790e-03],\n [-2.2690e-02, -2.6646e-03, -2.8443e-04, ..., -1.4946e-02, -1.7990e-02, -2.0447e-02],\n [-2.0142e-03, 9.8648e-03, -1.7868e-02, ..., 2.2781e-02, -2.2110e-02, -2.7054e-02],\n ...,\n [-2.3441e-03, 1.3802e-02, 1.0669e-05, ..., 2.1942e-02, 3.1433e-02, -4.3602e-03],\n [ 4.7394e-02, -2.4368e-02, -2.0401e-02, ..., 4.6082e-02, 4.1565e-02, 4.8767e-02],\n [ 3.4241e-02, 1.5671e-02, -3.0289e-02, ..., -1.9653e-02, 3.3302e-03, 3.4546e-02]], dtype=torch.float16), '26.to_out_lora.up.weight': tensor([[-0.0284, 0.0296, 0.0085, ..., 0.0082, 0.0251, 0.0152],\n [-0.0026, 0.0298, -0.0088, ..., -0.0065, 0.0068, 0.0238],\n [ 0.0100, -0.0028, -0.0104, ..., 0.0222, 0.0547, -0.0135],\n ...,\n [-0.0196, -0.0318, 0.0341, ..., 0.0064, -0.0001, 0.0029],\n [ 0.0228, 0.0143, 0.0102, ..., 0.0134, -0.0386, -0.0066],\n [ 0.0091, -0.0359, 0.0193, ..., -0.0098, 0.0091, 0.0115]], dtype=torch.float16), '27.to_q_lora.down.weight': tensor([[-0.0544, -0.0182, -0.0276, ..., 0.0827, 0.0142, -0.0082],\n [-0.0202, 0.0037, -0.0267, ..., 0.0081, 0.0245, -0.0440],\n [-0.0422, 0.0004, 0.0742, ..., 0.0265, 0.0479, 0.0203],\n ...,\n [ 0.0287, -0.0467, -0.0611, ..., -0.0493, 0.0082, -0.0106],\n [-0.0381, -0.0151, 0.0937, ..., 0.0283, 0.0191, -0.0443],\n [-0.0493, 0.0137, 0.0145, ..., 0.0302, 0.0431, 0.0850]], dtype=torch.float16), '27.to_q_lora.up.weight': tensor([[-0.0219, -0.0130, 0.0113, ..., -0.0558, 0.0191, -0.0517],\n [-0.0205, 0.0676, -0.0301, ..., 0.0432, -0.0291, -0.0715],\n [-0.0151, -0.0230, 0.0034, ..., -0.0129, 0.0064, 0.0142],\n ...,\n [ 0.0404, -0.0520, 0.0178, ..., -0.0403, -0.0163, 0.0810],\n [ 0.0335, -0.0591, 0.0241, ..., 0.0657, -0.0145, -0.0359],\n [-0.0049, 0.0332, -0.0420, ..., 0.0020, -0.0274, -0.0303]], dtype=torch.float16), '27.to_k_lora.down.weight': tensor([[-0.0339, 0.0046, -0.0149, ..., 0.0045, 0.0108, -0.0233],\n [ 0.0058, 0.0125, 0.0048, ..., 0.0023, 0.0623, -0.0038],\n [-0.0150, 0.0157, 0.0004, ..., -0.0116, 0.0523, -0.0092],\n ...,\n [ 0.0064, 0.0430, 0.0024, ..., 0.0019, 0.0357, -0.0048],\n [-0.0149, 0.0324, 0.0047, ..., 0.0036, 0.0093, 0.0019],\n [-0.0198, 0.0007, -0.0109, ..., -0.0115, -0.0245, -0.0254]], dtype=torch.float16), '27.to_k_lora.up.weight': tensor([[ 0.0135, 0.0384, 0.0201, ..., -0.0449, 0.0192, -0.0485],\n [-0.0115, 0.0398, -0.0042, ..., -0.0198, 0.0146, -0.0777],\n [-0.0017, 0.0559, 0.0416, ..., 0.0014, 0.0397, -0.0569],\n ...,\n [ 0.0231, -0.0324, -0.0394, ..., -0.0544, -0.0252, -0.0219],\n [ 0.0296, 0.0626, 0.0363, ..., -0.0040, -0.0163, -0.0205],\n [-0.0161, -0.0181, 0.0047, ..., 0.0109, -0.0027, 0.0072]], dtype=torch.float16), '27.to_v_lora.down.weight': tensor([[-2.7191e-02, 1.1414e-02, -2.2552e-02, ..., -7.7934e-03, -4.4037e-02, -9.6207e-03],\n [ 4.6349e-03, -6.1859e-02, 9.4604e-03, ..., 1.6266e-02, 1.7319e-02, 4.0833e-02],\n [-1.0910e-02, -7.9117e-03, -6.9542e-03, ..., 2.2068e-03, 2.2842e-02, -4.2915e-03],\n ...,\n [ 3.0502e-02, -2.1774e-02, 8.3694e-03, ..., -7.1411e-03, -1.7639e-02, -9.3460e-05],\n [ 1.1848e-02, 2.3117e-02, -8.4534e-03, ..., 1.6251e-02, 1.8646e-02, 1.8402e-02],\n [-6.0692e-03, -1.9531e-03, 6.1569e-03, ..., -1.1272e-03, -5.6229e-03, -1.3237e-02]], dtype=torch.float16), '27.to_v_lora.up.weight': tensor([[ 2.6817e-03, 2.3689e-03, -4.9286e-03, ..., -1.7227e-02, 3.0762e-02, 3.2166e-02],\n [-3.2440e-02, 1.3924e-02, -1.3741e-02, ..., -1.0216e-02, -1.8477e-06, -1.9440e-02],\n [ 1.9417e-03, -5.5428e-03, -1.0948e-02, ..., 3.8269e-02, -3.5877e-03, 8.5068e-03],\n ...,\n [-7.3395e-03, 2.2339e-02, 5.7888e-04, ..., 7.2289e-03, 3.4866e-03, 2.7802e-02],\n [-1.1147e-02, 1.2360e-02, -7.5798e-03, ..., 2.6760e-03, -2.2640e-03, -2.3537e-03],\n [-2.2186e-02, 3.1952e-02, 2.1576e-02, ..., 1.6373e-02, 1.3161e-02, -4.9171e-03]], dtype=torch.float16), '27.to_out_lora.down.weight': tensor([[ 0.0496, -0.0152, 0.0197, ..., 0.0093, -0.0102, 0.0219],\n [ 0.0227, -0.0055, 0.0275, ..., -0.0051, 0.0630, 0.0063],\n [ 0.0083, 0.0382, -0.0088, ..., 0.0103, 0.0006, 0.0015],\n ...,\n [ 0.0056, -0.0023, 0.0122, ..., 0.0181, -0.0066, -0.0258],\n [ 0.0312, -0.0241, -0.0162, ..., -0.0088, -0.0242, 0.0128],\n [-0.0319, 0.0019, -0.0100, ..., -0.0085, 0.0332, 0.0493]], dtype=torch.float16), '27.to_out_lora.up.weight': tensor([[-0.0325, -0.0009, 0.0194, ..., -0.0073, 0.0350, 0.0236],\n [-0.0101, 0.0248, 0.0247, ..., 0.0177, -0.0073, 0.0147],\n [ 0.0172, 0.0239, -0.0204, ..., -0.0257, -0.0311, -0.0127],\n ...,\n [-0.0159, 0.0190, -0.0163, ..., -0.0368, 0.0115, 0.0360],\n [ 0.0049, -0.0161, 0.0461, ..., -0.0159, 0.0494, 0.0092],\n [-0.0202, -0.0153, -0.0043, ..., 0.0006, 0.0019, -0.0165]], dtype=torch.float16), '27.to_k_ip.weight': tensor([[ 4.4586e-02, -3.0197e-02, -3.8544e-02, ..., -5.2887e-02, -4.8615e-02, -3.5736e-02],\n [ 6.2378e-02, 5.8380e-02, -4.3640e-02, ..., 1.0735e-02, 2.2919e-02, 2.9129e-02],\n [-3.7060e-03, 4.6158e-03, 2.0813e-02, ..., 6.4026e-02, 4.6326e-02, 4.3488e-02],\n ...,\n [-5.5695e-02, 8.0261e-02, 4.9744e-02, ..., 4.6492e-05, -7.3853e-02, -6.4774e-03],\n [-2.9678e-02, -5.3284e-02, -3.4821e-02, ..., 8.7585e-03, 1.6556e-02, 3.4180e-02],\n [-4.0436e-02, 9.8877e-03, 3.9917e-02, ..., 1.2985e-02, 4.8309e-02, -2.9510e-02]], dtype=torch.float16), '27.to_v_ip.weight': tensor([[ 0.0072, -0.0116, -0.0477, ..., -0.0168, 0.0001, -0.0435],\n [-0.0298, 0.0138, -0.0273, ..., -0.0038, 0.0180, -0.0169],\n [-0.0094, 0.0216, 0.0231, ..., 0.0282, -0.0226, -0.0131],\n ...,\n [-0.0323, -0.0262, 0.0348, ..., 0.0321, -0.0073, -0.0154],\n [ 0.0075, 0.0729, -0.0945, ..., -0.0545, -0.0510, 0.0308],\n [ 0.0583, -0.0287, -0.0220, ..., -0.0347, 0.0339, -0.0188]], dtype=torch.float16), '28.to_q_lora.down.weight': tensor([[ 0.0226, 0.0342, 0.0180, ..., -0.0287, -0.0205, -0.0105],\n [-0.0206, -0.0077, 0.0099, ..., -0.0543, 0.0088, 0.0331],\n [-0.0045, -0.0143, 0.0033, ..., 0.0219, 0.0643, -0.0119],\n ...,\n [-0.0033, -0.0088, -0.0220, ..., 0.0634, 0.0363, -0.0558],\n [-0.0100, -0.0512, 0.0463, ..., -0.0532, -0.0071, 0.0337],\n [-0.0297, 0.0191, 0.0454, ..., -0.0097, 0.0793, -0.0317]], dtype=torch.float16), '28.to_q_lora.up.weight': tensor([[ 0.0279, 0.0066, 0.0074, ..., 0.0285, -0.0079, -0.0054],\n [-0.0410, -0.0078, -0.0461, ..., 0.0112, 0.0308, -0.0261],\n [ 0.0632, 0.0494, 0.0327, ..., -0.0217, 0.0493, -0.0269],\n ...,\n [-0.0789, 0.0369, -0.0192, ..., -0.0439, 0.0279, -0.0334],\n [-0.0127, -0.0068, 0.0340, ..., -0.0014, 0.0033, -0.0110],\n [ 0.1110, -0.0485, 0.0013, ..., 0.0196, -0.0020, -0.0438]], dtype=torch.float16), '28.to_k_lora.down.weight': tensor([[-0.0510, 0.0254, -0.0430, ..., 0.0085, 0.0109, 0.0187],\n [-0.0170, 0.0212, 0.0268, ..., 0.0314, -0.0377, 0.0047],\n [ 0.0168, 0.0135, 0.0253, ..., -0.0114, 0.0446, -0.0059],\n ...,\n [ 0.0219, -0.0344, 0.0366, ..., -0.0338, 0.0532, -0.0038],\n [ 0.0101, -0.0591, -0.0192, ..., -0.0102, -0.0067, 0.0255],\n [-0.0204, 0.0173, -0.0331, ..., -0.0019, 0.1027, 0.0031]], dtype=torch.float16), '28.to_k_lora.up.weight': tensor([[ 0.0188, 0.0175, 0.0090, ..., 0.0316, 0.0020, -0.0278],\n [ 0.0229, -0.0748, -0.0692, ..., 0.0538, -0.0143, -0.0046],\n [-0.0078, -0.0437, 0.0159, ..., 0.0043, -0.0089, 0.0609],\n ...,\n [-0.0018, 0.0592, -0.0450, ..., -0.0206, 0.0026, 0.0311],\n [ 0.0296, -0.0187, -0.0034, ..., 0.0352, -0.0381, -0.0527],\n [ 0.0160, -0.0676, -0.0066, ..., -0.0076, 0.0400, 0.0456]], dtype=torch.float16), '28.to_v_lora.down.weight': tensor([[ 1.9684e-02, 1.5617e-02, 5.5466e-03, ..., 1.6571e-02, 1.4397e-02, -2.9144e-02],\n [-4.7455e-02, 2.0477e-02, 1.6693e-02, ..., 1.0948e-02, -5.7936e-05, -6.6185e-03],\n [-8.8577e-03, -2.9572e-02, -8.9417e-03, ..., 2.3270e-02, 3.6983e-03, -2.1011e-02],\n ...,\n [-2.2430e-02, 9.6893e-03, 3.4515e-02, ..., 4.3640e-03, -4.4342e-02, -2.5497e-02],\n [-4.4373e-02, 1.2674e-03, 6.9237e-03, ..., -3.0060e-02, 4.0741e-02, -1.9264e-03],\n [-2.6016e-02, -3.1555e-02, 3.2257e-02, ..., -8.4610e-03, -1.1497e-02, -1.6296e-02]], dtype=torch.float16), '28.to_v_lora.up.weight': tensor([[-0.0349, -0.0410, 0.0047, ..., 0.0102, -0.0048, 0.0147],\n [ 0.0186, -0.0035, 0.0393, ..., -0.0324, -0.0077, -0.0101],\n [ 0.0245, -0.0083, 0.0532, ..., -0.0094, -0.0259, 0.0783],\n ...,\n [ 0.0053, -0.0099, -0.0149, ..., -0.0087, 0.0076, -0.0009],\n [ 0.0044, 0.0249, 0.0087, ..., -0.0505, -0.0011, -0.0250],\n [-0.0077, 0.0221, 0.0323, ..., 0.0605, 0.0464, 0.0018]], dtype=torch.float16), '28.to_out_lora.down.weight': tensor([[ 0.0143, -0.0031, 0.0006, ..., 0.0127, -0.0053, -0.0472],\n [-0.0246, 0.0181, -0.0059, ..., -0.0317, -0.0107, 0.0659],\n [-0.0313, -0.0095, -0.0010, ..., -0.0112, 0.0250, 0.0335],\n ...,\n [-0.0230, -0.0072, 0.0188, ..., -0.0386, -0.0159, -0.0052],\n [-0.0230, -0.0202, 0.0244, ..., -0.0474, 0.0337, 0.0080],\n [-0.0061, 0.0034, -0.0110, ..., -0.0153, -0.0440, 0.0994]], dtype=torch.float16), '28.to_out_lora.up.weight': tensor([[ 6.2065e-03, 4.4495e-02, 2.6154e-02, ..., 2.7679e-02, 4.4647e-02, -1.8661e-02],\n [ 7.0839e-03, -3.6743e-02, -5.3596e-03, ..., -5.4016e-03, -5.0201e-02, 3.3722e-02],\n [ 6.7368e-03, 5.1544e-02, 2.7199e-03, ..., 1.6708e-02, 7.3471e-03, 9.8724e-03],\n ...,\n [-4.4098e-02, 1.7441e-02, 1.6052e-02, ..., 4.7684e-05, 4.4918e-04, -5.5237e-03],\n [-8.3389e-03, -6.3400e-03, -1.1187e-03, ..., -1.0475e-02, 2.1683e-02, -6.2447e-03],\n [ 8.9111e-03, 3.7811e-02, 3.1891e-02, ..., -3.7628e-02, -1.7776e-02, 3.6377e-02]], dtype=torch.float16), '29.to_q_lora.down.weight': tensor([[ 0.0203, 0.0068, 0.0011, ..., -0.0235, -0.0029, 0.0126],\n [-0.0314, 0.0240, 0.0365, ..., 0.0324, -0.0277, 0.0077],\n [-0.0208, -0.0059, 0.0298, ..., -0.0110, -0.0045, -0.0186],\n ...,\n [-0.0325, 0.0202, -0.0112, ..., -0.0388, 0.0660, -0.0382],\n [ 0.0312, -0.0230, 0.0378, ..., -0.0393, 0.0211, -0.0147],\n [-0.0004, -0.0260, 0.0468, ..., 0.0002, 0.0256, 0.0356]], dtype=torch.float16), '29.to_q_lora.up.weight': tensor([[ 0.0721, 0.0240, 0.0502, ..., 0.0646, 0.0224, -0.0037],\n [ 0.0797, -0.0316, 0.0753, ..., 0.0198, -0.0217, -0.0327],\n [ 0.0609, 0.0485, 0.0741, ..., 0.0641, -0.0038, -0.0385],\n ...,\n [ 0.0078, 0.0124, -0.0117, ..., -0.0015, -0.0059, 0.0070],\n [ 0.0333, 0.0663, 0.0265, ..., 0.0134, 0.0388, 0.0104],\n [ 0.0510, -0.0699, -0.0249, ..., -0.0042, 0.0204, 0.0254]], dtype=torch.float16), '29.to_k_lora.down.weight': tensor([[-0.0354, 0.0076, 0.0046, ..., -0.0113, 0.0208, -0.0355],\n [ 0.0261, -0.0320, -0.0095, ..., -0.0020, -0.0217, -0.0192],\n [-0.0199, 0.0155, 0.0024, ..., -0.0421, 0.0193, -0.0102],\n ...,\n [-0.0278, -0.0167, 0.0257, ..., 0.0336, 0.0135, -0.0291],\n [ 0.0414, -0.0193, 0.0106, ..., -0.0414, 0.0061, 0.0121],\n [ 0.0242, 0.0382, 0.0054, ..., -0.0208, -0.0154, -0.0016]], dtype=torch.float16), '29.to_k_lora.up.weight': tensor([[-0.0202, 0.0189, 0.0883, ..., 0.0072, -0.0012, 0.0418],\n [ 0.0105, 0.0047, 0.0303, ..., 0.0054, 0.0159, -0.0565],\n [-0.0107, 0.0307, 0.0385, ..., 0.0259, -0.0101, -0.0093],\n ...,\n [-0.0146, 0.0003, 0.0020, ..., -0.0166, 0.0100, -0.0046],\n [ 0.0120, -0.0010, 0.0180, ..., 0.0015, 0.0094, 0.0232],\n [-0.0072, 0.0383, 0.0240, ..., -0.0319, 0.0353, 0.0534]], dtype=torch.float16), '29.to_v_lora.down.weight': tensor([[ 0.0104, 0.0333, -0.0178, ..., 0.0093, -0.0017, 0.0149],\n [-0.0237, -0.0164, 0.0079, ..., -0.0054, 0.0564, 0.0369],\n [ 0.0318, -0.0214, 0.0023, ..., -0.0293, -0.0311, 0.0186],\n ...,\n [ 0.0127, -0.0185, 0.0047, ..., -0.0163, 0.0217, -0.0127],\n [-0.0124, -0.0076, -0.0114, ..., 0.0237, -0.0112, 0.0458],\n [-0.0134, -0.0633, -0.0179, ..., 0.0080, 0.0348, -0.0075]], dtype=torch.float16), '29.to_v_lora.up.weight': tensor([[-0.0282, -0.0098, 0.0056, ..., -0.0121, -0.0060, -0.0074],\n [-0.0252, 0.0226, -0.0171, ..., 0.0073, 0.0008, -0.0066],\n [-0.0028, 0.0059, -0.0142, ..., -0.0080, -0.0227, -0.0106],\n ...,\n [-0.0062, 0.0196, -0.0047, ..., -0.0280, 0.0125, -0.0021],\n [ 0.0414, -0.0105, 0.0043, ..., -0.0030, 0.0097, -0.0326],\n [ 0.0163, 0.0001, 0.0069, ..., 0.0080, -0.0060, -0.0326]], dtype=torch.float16), '29.to_out_lora.down.weight': tensor([[-1.7090e-02, -3.2837e-02, -1.3809e-03, ..., -4.1847e-03, 6.1493e-03, -6.7890e-05],\n [-8.3694e-03, -1.8295e-02, 1.1749e-02, ..., 1.5854e-02, 6.0692e-03, -2.6703e-02],\n [ 1.7395e-02, 7.6866e-03, 9.6970e-03, ..., 1.9882e-02, -1.3611e-02, -2.9312e-02],\n ...,\n [-1.4900e-02, -1.5671e-02, -9.6436e-03, ..., -2.8290e-02, -7.0343e-03, 1.5053e-02],\n [ 1.3184e-02, 5.1994e-03, -1.3977e-02, ..., 1.3634e-02, 1.5228e-02, -1.5900e-02],\n [ 1.3023e-02, 8.1177e-03, 6.7062e-03, ..., 1.2779e-02, -5.0049e-03, -4.9591e-03]], dtype=torch.float16), '29.to_out_lora.up.weight': tensor([[ 0.0252, -0.0216, -0.0138, ..., 0.0034, -0.0314, 0.0128],\n [ 0.0077, -0.0092, -0.0070, ..., -0.0309, 0.0227, -0.0548],\n [ 0.0250, 0.0043, -0.0293, ..., -0.0108, 0.0241, 0.0218],\n ...,\n [-0.0197, 0.0232, 0.0074, ..., 0.0154, 0.0124, -0.0260],\n [ 0.0041, 0.0278, -0.0258, ..., 0.0093, -0.0032, -0.0144],\n [ 0.0318, 0.0219, -0.0126, ..., -0.0263, 0.0117, 0.0170]], dtype=torch.float16), '29.to_k_ip.weight': tensor([[-0.0441, -0.0419, 0.0092, ..., 0.0760, 0.0058, 0.0152],\n [ 0.0023, -0.0952, 0.0375, ..., 0.1295, 0.0614, -0.0092],\n [-0.0532, -0.0547, -0.0304, ..., 0.0066, -0.0600, 0.0097],\n ...,\n [-0.0296, 0.0001, 0.0258, ..., -0.0125, 0.0092, -0.0119],\n [ 0.0044, -0.0307, 0.0297, ..., 0.0597, -0.0182, -0.0834],\n [ 0.0475, 0.0523, 0.0378, ..., -0.1425, -0.1231, 0.1145]], dtype=torch.float16), '29.to_v_ip.weight': tensor([[ 0.0130, -0.0206, 0.0003, ..., -0.0223, -0.0155, -0.0117],\n [ 0.0087, 0.0067, -0.0287, ..., -0.0274, -0.0119, -0.0053],\n [ 0.0249, 0.0022, -0.0012, ..., -0.0035, 0.0767, 0.0762],\n ...,\n [ 0.0167, 0.0198, 0.0194, ..., -0.0094, -0.0330, 0.0088],\n [-0.0292, 0.0543, 0.0176, ..., 0.0240, -0.0252, -0.0535],\n [-0.0315, 0.0269, -0.0136, ..., -0.0118, -0.0146, -0.0058]], dtype=torch.float16), '30.to_q_lora.down.weight': tensor([[ 0.0316, 0.0324, -0.0615, ..., -0.0235, -0.0332, -0.0285],\n [-0.0233, 0.0401, -0.0380, ..., -0.0112, 0.0013, -0.0081],\n [-0.0219, -0.0080, 0.0137, ..., 0.0323, 0.0574, 0.0634],\n ...,\n [ 0.0277, 0.0216, -0.0540, ..., -0.0215, 0.0184, -0.0242],\n [-0.0112, -0.0286, 0.0446, ..., -0.0061, 0.0134, 0.0116],\n [ 0.0469, -0.0003, -0.0354, ..., 0.0331, -0.0005, 0.0167]], dtype=torch.float16), '30.to_q_lora.up.weight': tensor([[-0.0122, 0.0053, 0.0106, ..., -0.0257, 0.0218, 0.0321],\n [ 0.0079, 0.0583, -0.0039, ..., -0.0156, 0.0344, 0.0141],\n [ 0.0307, -0.0122, -0.0086, ..., 0.0054, -0.0307, 0.0183],\n ...,\n [-0.0023, -0.0006, -0.0063, ..., 0.0213, -0.0155, -0.0251],\n [-0.0338, -0.0092, 0.0250, ..., 0.0284, -0.0370, -0.0051],\n [ 0.0078, 0.0064, -0.0378, ..., 0.0496, -0.0222, -0.0012]], dtype=torch.float16), '30.to_k_lora.down.weight': tensor([[ 0.0232, 0.0088, 0.0137, ..., 0.0460, -0.0380, -0.0140],\n [ 0.0149, -0.0225, 0.0068, ..., 0.0369, 0.0180, 0.0035],\n [-0.0097, 0.0109, 0.0435, ..., -0.0558, -0.0002, 0.0032],\n ...,\n [-0.0325, -0.0533, 0.0152, ..., 0.0214, 0.0342, 0.0572],\n [-0.0004, 0.0340, 0.0177, ..., -0.0026, -0.0526, -0.0043],\n [-0.0304, 0.0040, -0.0529, ..., 0.0247, 0.0477, 0.0186]], dtype=torch.float16), '30.to_k_lora.up.weight': tensor([[ 0.0102, -0.0131, -0.0124, ..., -0.0021, -0.0115, 0.0124],\n [ 0.0227, -0.0026, 0.0433, ..., -0.0295, -0.0188, -0.0044],\n [ 0.0046, -0.0187, 0.0102, ..., -0.0157, 0.0073, 0.0388],\n ...,\n [ 0.0127, -0.0192, 0.0238, ..., 0.0132, 0.0176, 0.0023],\n [ 0.0279, 0.0420, -0.0372, ..., 0.0393, 0.0396, 0.0216],\n [-0.0244, -0.0277, -0.0200, ..., 0.0034, -0.0067, -0.0036]], dtype=torch.float16), '30.to_v_lora.down.weight': tensor([[-1.4969e-02, 4.0283e-02, 1.3828e-05, ..., -2.6665e-03, 7.4280e-02, -3.3722e-02],\n [-8.4686e-03, 1.6603e-03, -2.1072e-02, ..., -1.5244e-02, 4.2664e-02, 1.8829e-02],\n [ 3.2135e-02, 8.2016e-03, -1.8845e-02, ..., -3.0930e-02, -5.4932e-02, -1.9028e-02],\n ...,\n [ 3.3417e-02, 1.6586e-02, 2.3193e-02, ..., -2.0996e-02, 1.6815e-02, -7.4585e-02],\n [ 3.6285e-02, -8.5510e-02, -4.1077e-02, ..., -2.4246e-02, -9.9945e-03, 3.2532e-02],\n [-5.3520e-03, -1.4410e-03, -1.8204e-02, ..., 2.0981e-02, -9.5367e-03, -1.6983e-02]], dtype=torch.float16), '30.to_v_lora.up.weight': tensor([[ 0.0190, 0.0376, -0.0367, ..., 0.0098, -0.0392, -0.0243],\n [ 0.0575, 0.0259, -0.0006, ..., -0.0076, 0.0296, -0.0238],\n [ 0.0258, 0.0030, 0.0029, ..., 0.0025, -0.0458, -0.0006],\n ...,\n [ 0.0296, 0.0016, -0.0030, ..., -0.0096, -0.0322, 0.0568],\n [ 0.0124, 0.0308, -0.0264, ..., -0.0596, 0.0056, -0.0025],\n [ 0.0302, -0.0407, 0.0059, ..., 0.0053, -0.0202, 0.0122]], dtype=torch.float16), '30.to_out_lora.down.weight': tensor([[ 0.0061, 0.0521, -0.0174, ..., -0.0188, -0.0831, -0.0355],\n [-0.0251, -0.0253, -0.0049, ..., -0.0486, 0.0086, 0.0154],\n [ 0.0434, -0.0019, -0.0280, ..., 0.0334, 0.0135, 0.0076],\n ...,\n [-0.0063, 0.0023, 0.0386, ..., -0.0032, 0.0326, 0.0469],\n [-0.0428, 0.0206, 0.0440, ..., -0.1021, -0.0104, 0.0112],\n [-0.0296, 0.0316, 0.0138, ..., -0.0183, -0.0468, -0.0168]], dtype=torch.float16), '30.to_out_lora.up.weight': tensor([[-0.0073, -0.0680, 0.0253, ..., -0.0345, 0.0790, 0.0191],\n [ 0.0210, -0.0059, 0.0226, ..., 0.0055, -0.0410, -0.0365],\n [ 0.0155, -0.0303, -0.0597, ..., -0.0028, -0.0518, 0.0206],\n ...,\n [ 0.0390, 0.0822, 0.0956, ..., -0.0121, -0.0020, 0.0082],\n [-0.0212, -0.0058, -0.0145, ..., 0.0709, 0.0287, 0.0121],\n [-0.0360, 0.0591, -0.0493, ..., -0.0440, 0.0015, 0.0625]], dtype=torch.float16), '31.to_q_lora.down.weight': tensor([[ 0.0208, -0.0194, 0.0135, ..., -0.0457, -0.0275, -0.0258],\n [-0.0187, 0.0101, 0.0047, ..., 0.0016, 0.0153, -0.0005],\n [-0.0070, 0.0113, 0.0500, ..., -0.0069, -0.0384, -0.0123],\n ...,\n [ 0.0168, -0.0129, -0.0019, ..., 0.0042, -0.0057, 0.0265],\n [-0.0253, 0.0160, 0.0005, ..., 0.0048, 0.0112, -0.0584],\n [-0.0088, -0.0155, 0.0052, ..., -0.0306, -0.0007, -0.0747]], dtype=torch.float16), '31.to_q_lora.up.weight': tensor([[ 0.0069, -0.0114, 0.0467, ..., -0.0030, -0.0005, 0.0429],\n [ 0.0059, 0.0077, -0.0105, ..., 0.0450, -0.0035, 0.0206],\n [ 0.0166, -0.0377, -0.0334, ..., -0.0446, -0.0215, 0.0027],\n ...,\n [-0.0743, -0.0023, 0.0207, ..., 0.0203, -0.0308, -0.0828],\n [-0.0189, -0.0177, 0.0281, ..., -0.0208, 0.0150, 0.0324],\n [ 0.0047, -0.0127, -0.0131, ..., -0.0381, -0.0324, 0.0323]], dtype=torch.float16), '31.to_k_lora.down.weight': tensor([[-0.0146, 0.0450, -0.0230, ..., -0.0078, -0.0648, -0.0323],\n [ 0.0805, -0.0456, -0.0308, ..., 0.0092, -0.0133, -0.0430],\n [-0.0150, -0.0246, -0.0231, ..., -0.0182, 0.0540, -0.0053],\n ...,\n [ 0.0065, 0.0035, -0.0402, ..., 0.0574, -0.0556, 0.0374],\n [-0.0354, 0.0100, -0.0275, ..., -0.0035, -0.0212, 0.0159],\n [-0.0851, -0.0264, -0.0223, ..., 0.0272, 0.0479, 0.0502]], dtype=torch.float16), '31.to_k_lora.up.weight': tensor([[-0.0706, -0.0347, 0.0592, ..., -0.0358, -0.0094, -0.0143],\n [-0.0121, -0.0789, 0.0913, ..., -0.0283, -0.0693, 0.0118],\n [ 0.0894, 0.0361, -0.0095, ..., 0.0409, 0.0145, -0.0218],\n ...,\n [-0.0039, 0.0064, -0.0399, ..., -0.0381, -0.0363, -0.0668],\n [ 0.0453, 0.0495, 0.0252, ..., -0.0039, 0.0028, 0.0190],\n [ 0.0484, -0.0098, -0.0850, ..., -0.0160, 0.0759, 0.0077]], dtype=torch.float16), '31.to_v_lora.down.weight': tensor([[ 0.0083, -0.0222, 0.0027, ..., -0.0130, 0.0019, -0.0265],\n [ 0.0163, -0.0045, -0.0605, ..., 0.0014, -0.0017, -0.0229],\n [-0.0030, -0.0059, -0.0309, ..., -0.0199, -0.0636, 0.0432],\n ...,\n [ 0.0079, 0.0235, 0.0162, ..., 0.0117, 0.0027, 0.0516],\n [ 0.0559, 0.0740, 0.0221, ..., -0.0158, 0.0148, 0.0044],\n [ 0.0275, -0.0196, -0.0178, ..., -0.0285, -0.0625, 0.0179]], dtype=torch.float16), '31.to_v_lora.up.weight': tensor([[ 0.0419, 0.0218, 0.0022, ..., -0.0328, 0.0112, -0.0506],\n [ 0.0101, -0.0071, -0.0379, ..., 0.0291, -0.0058, 0.0107],\n [ 0.0117, 0.0198, -0.0271, ..., -0.0016, 0.0225, 0.0233],\n ...,\n [ 0.0203, 0.0037, -0.0452, ..., 0.0179, -0.0094, -0.0230],\n [ 0.0414, 0.0260, 0.0065, ..., 0.0022, 0.0145, -0.0369],\n [-0.0486, 0.0873, -0.0355, ..., -0.0515, -0.0012, -0.0127]], dtype=torch.float16), '31.to_out_lora.down.weight': tensor([[-0.0070, 0.0296, 0.0073, ..., -0.0003, -0.0012, -0.0081],\n [-0.0233, -0.0076, 0.0025, ..., 0.0105, 0.0256, -0.0375],\n [ 0.0172, 0.0568, -0.0243, ..., -0.0433, 0.0066, -0.0700],\n ...,\n [ 0.0132, -0.0240, 0.0129, ..., -0.0522, 0.0804, -0.0233],\n [ 0.0114, 0.0236, -0.0640, ..., -0.0231, 0.0026, 0.0458],\n [ 0.0374, -0.0029, -0.0574, ..., 0.0020, -0.0343, 0.0119]], dtype=torch.float16), '31.to_out_lora.up.weight': tensor([[-0.0780, 0.0345, -0.0841, ..., -0.0178, 0.0144, 0.0300],\n [ 0.0507, 0.0289, 0.0127, ..., -0.0489, -0.0277, -0.0297],\n [ 0.0022, 0.0027, 0.0120, ..., 0.0003, -0.0123, 0.0334],\n ...,\n [-0.0435, -0.0129, -0.0224, ..., 0.0444, 0.0199, -0.0397],\n [-0.0134, 0.0269, 0.0223, ..., 0.0140, -0.0728, 0.0269],\n [ 0.0129, -0.0266, -0.0266, ..., -0.0012, 0.0427, 0.0534]], dtype=torch.float16), '31.to_k_ip.weight': tensor([[ 0.0117, -0.0105, -0.0403, ..., 0.0399, -0.0995, 0.0184],\n [ 0.0714, -0.0170, 0.0360, ..., -0.0345, 0.0458, 0.0044],\n [-0.0070, -0.0508, -0.0340, ..., -0.0429, -0.0795, 0.0382],\n ...,\n [-0.0045, -0.0867, -0.0748, ..., -0.0863, -0.0276, 0.0065],\n [-0.0663, 0.0357, -0.0175, ..., 0.0252, 0.0253, -0.0147],\n [ 0.0314, 0.0681, -0.0164, ..., -0.0091, 0.0026, 0.0118]], dtype=torch.float16), '31.to_v_ip.weight': tensor([[-0.0550, -0.1074, -0.0609, ..., -0.0365, -0.0549, -0.0112],\n [-0.0108, -0.0092, 0.0294, ..., 0.0155, 0.0335, 0.0136],\n [ 0.0311, -0.0522, 0.0442, ..., -0.0022, 0.0081, 0.0778],\n ...,\n [ 0.0018, 0.0224, 0.0701, ..., 0.0236, -0.0257, -0.0486],\n [ 0.0560, -0.0087, -0.0007, ..., 0.0106, 0.0070, -0.0557],\n [ 0.0225, 0.0718, -0.0681, ..., 0.0032, -0.0220, 0.0679]], dtype=torch.float16)}, 'faceidplusv2': True}"]}, {"id": 10, "type": "Anything Everywhere", "pos": [4800, 60], "size": {"0": 240, "1": 60}, "flags": {}, "order": 18, "mode": 0, "inputs": [{"name": "CLIP_VISION", "type": "*", "link": 7, "color_on": "#A8DADC"}], "title": "Clip Vision Everywhere", "properties": {"group_restricted": false, "color_restricted": false, "Node name for S&R": "Anything Everywhere"}, "widgets_values": ["<comfy.clip_vision.ClipVisionModel object at 0x000001E2112836D0>"]}, {"id": 8, "type": "IPAdapterModelLoader", "pos": [4320, -60], "size": {"0": 315, "1": 58}, "flags": {}, "order": 0, "mode": 0, "outputs": [{"name": "IPADAPTER", "type": "IPADAPTER", "links": [6], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "IPAdapterModelLoader"}, "widgets_values": ["ip-adapter-faceid-plusv2_sd15.bin"]}, {"id": 5, "type": "CLIPVisionLoader", "pos": [4320, 60], "size": {"0": 315, "1": 58}, "flags": {}, "order": 1, "mode": 0, "outputs": [{"name": "CLIP_VISION", "type": "CLIP_VISION", "links": [7], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "CLIPVisionLoader"}, "widgets_values": ["CLIP-ViT-H-14-laion2B-s32B-b79K.safetensors"]}, {"id": 6, "type": "IPAdapterInsightFaceLoader", "pos": [4320, 180], "size": {"0": 315, "1": 58}, "flags": {}, "order": 2, "mode": 0, "outputs": [{"name": "INSIGHTFACE", "type": "INSIGHTFACE", "links": [8], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "IPAdapterInsightFaceLoader"}, "widgets_values": ["CUDA"]}, {"id": 13, "type": "PreviewImage", "pos": [5280, 120], "size": [240, 240], "flags": {}, "order": 27, "mode": 0, "inputs": [{"name": "images", "type": "IMAGE", "link": 10}], "properties": {"Node name for S&R": "PreviewImage"}}, {"id": 21, "type": "PreviewImage", "pos": [6240, 120], "size": [240, 240], "flags": {}, "order": 29, "mode": 0, "inputs": [{"name": "images", "type": "IMAGE", "link": 17}], "properties": {"Node name for S&R": "PreviewImage"}}, {"id": 17, "type": "PreviewImage", "pos": [5760, 120], "size": [240, 240], "flags": {}, "order": 28, "mode": 0, "inputs": [{"name": "images", "type": "IMAGE", "link": 14}], "properties": {"Node name for S&R": "PreviewImage"}}, {"id": 2, "type": "IPAdapterFaceID", "pos": [5280, -300], "size": {"0": 315, "1": 322}, "flags": {}, "order": 30, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": 2}, {"name": "ipadapter", "type": "IPADAPTER", "link": null}, {"name": "image", "type": "IMAGE", "link": 9}, {"name": "image_negative", "type": "IMAGE", "link": null}, {"name": "attn_mask", "type": "MASK", "link": 11}, {"name": "clip_vision", "type": "CLIP_VISION", "link": null}, {"name": "insightface", "type": "INSIGHTFACE", "link": null}], "outputs": [{"name": "MODEL", "type": "MODEL", "links": [18], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "IPAdapterFaceID"}, "widgets_values": [1, 1, "linear", "concat", 0, 1, "V only"]}, {"id": 15, "type": "IPAdapterFaceID", "pos": [5760, -300], "size": {"0": 315, "1": 322}, "flags": {}, "order": 32, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": 18}, {"name": "ipadapter", "type": "IPADAPTER", "link": null}, {"name": "image", "type": "IMAGE", "link": 12}, {"name": "image_negative", "type": "IMAGE", "link": null}, {"name": "attn_mask", "type": "MASK", "link": 13}, {"name": "clip_vision", "type": "CLIP_VISION", "link": null}, {"name": "insightface", "type": "INSIGHTFACE", "link": null}], "outputs": [{"name": "MODEL", "type": "MODEL", "links": [19], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "IPAdapterFaceID"}, "widgets_values": [1, 1, "linear", "concat", 0, 1, "V only"]}, {"id": 14, "type": "LoadImageMask", "pos": [5280, 480], "size": [300, 300], "flags": {}, "order": 3, "mode": 0, "outputs": [{"name": "MASK", "type": "MASK", "links": [11], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "LoadImageMask"}, "widgets_values": ["left 3.png", "alpha", "image"]}, {"id": 18, "type": "LoadImageMask", "pos": [5760, 480], "size": [300, 300], "flags": {}, "order": 4, "mode": 0, "outputs": [{"name": "MASK", "type": "MASK", "links": [13], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "LoadImageMask"}, "widgets_values": ["middle 4.png", "alpha", "image"]}, {"id": 22, "type": "LoadImageMask", "pos": [6240, 480], "size": [300, 300], "flags": {}, "order": 5, "mode": 0, "outputs": [{"name": "MASK", "type": "MASK", "links": [16], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "LoadImageMask"}, "widgets_values": ["right 3.png", "alpha", "image"]}, {"id": 25, "type": "CR LoRA Stack", "pos": [6660, -300], "size": {"0": 315, "1": 342}, "flags": {}, "order": 6, "mode": 0, "inputs": [{"name": "lora_stack", "type": "LORA_STACK", "link": null}], "outputs": [{"name": "LORA_STACK", "type": "LORA_STACK", "links": [21], "shape": 3, "slot_index": 0}, {"name": "show_help", "type": "STRING", "links": null, "shape": 3}], "properties": {"Node name for S&R": "CR LoRA Stack"}, "widgets_values": ["Off", "skin_tone_slider_v1.safetensors", 2, 2, "Off", "None", 1, 1, "Off", "None", 1, 1]}, {"id": 29, "type": "Prompts Everywhere", "pos": [4800, 660], "size": {"0": 210, "1": 46}, "flags": {}, "order": 22, "mode": 0, "inputs": [{"name": "CONDITIONING", "type": "*", "link": 23, "color_on": "#FFA931"}, {"name": "CONDITIONING", "type": "*", "link": 24, "color_on": "#FFA931"}], "properties": {"group_restricted": false, "color_restricted": false, "Node name for S&R": "Prompts Everywhere"}}, {"id": 26, "type": "Anything Everywhere", "pos": [4800, 300], "size": {"0": 240, "1": 60}, "flags": {}, "order": 38, "mode": 0, "inputs": [{"name": "MODEL", "type": "*", "link": 22, "color_on": "#B39DDB", "slot_index": 0}], "title": "Model Everywhere", "properties": {"group_restricted": false, "color_restricted": false, "Node name for S&R": "Anything Everywhere"}, "widgets_values": ["<comfy.model_patcher.ModelPatcher object at 0x000001E20F831B90>"]}, {"id": 30, "type": "VAELoader", "pos": [4260, 1080], "size": {"0": 315, "1": 58}, "flags": {}, "order": 7, "mode": 0, "outputs": [{"name": "VAE", "type": "VAE", "links": [25], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "VAELoader"}, "widgets_values": ["AnimateEveryone\\diffusion_pytorch_model.bin"]}, {"id": 32, "type": "Anything Everywhere", "pos": [4800, 1080], "size": {"0": 240, "1": 60}, "flags": {}, "order": 20, "mode": 0, "inputs": [{"name": "VAE", "type": "*", "link": 25, "slot_index": 0, "color_on": "#FF6E6E"}], "title": "VAE Everywhere", "properties": {"group_restricted": false, "color_restricted": false, "Node name for S&R": "Anything Everywhere"}, "widgets_values": ["<comfy.sd.VAE object at 0x000001E4BD2A35D0>"]}, {"id": 34, "type": "EmptyLatentImage", "pos": [4260, 1260], "size": {"0": 315, "1": 106}, "flags": {}, "order": 8, "mode": 0, "outputs": [{"name": "LATENT", "type": "LATENT", "links": [26], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "EmptyLatentImage"}, "widgets_values": [768, 512, 1]}, {"id": 28, "type": "GR Prompt Selector", "pos": [4260, 480], "size": {"0": 400, "1": 489.9999694824219}, "flags": {}, "order": 9, "mode": 0, "inputs": [{"name": "clip", "type": "CLIP", "link": null}], "outputs": [{"name": "positive", "type": "CONDITIONING", "links": [23], "shape": 3, "slot_index": 0}, {"name": "negative", "type": "CONDITIONING", "links": [24], "shape": 3, "slot_index": 1}, {"name": "prompts", "type": "STRING", "links": null, "shape": 3}], "properties": {"Node name for S&R": "GR Prompt Selector"}, "widgets_values": [",(((3 women))), standing, outside, white top, black coat", "", "", "", "", "", ",skinny,", "nsfw, blurry", 1]}, {"id": 11, "type": "Anything Everywhere", "pos": [4800, 180], "size": {"0": 240, "1": 60}, "flags": {}, "order": 19, "mode": 0, "inputs": [{"name": "INSIGHTFACE", "type": "*", "link": 8, "color_on": ""}], "title": "Insightface Everywhere", "properties": {"group_restricted": false, "color_restricted": false, "Node name for S&R": "Anything Everywhere"}, "widgets_values": ["<insightface.app.face_analysis.FaceAnalysis object at 0x000001E2111B7550>"]}, {"id": 38, "type": "Anything Everywhere", "pos": [4800, 780], "size": {"0": 240, "1": 60}, "flags": {}, "order": 24, "mode": 0, "inputs": [{"name": "CLIP", "type": "*", "link": 29, "color_on": "#FFD500"}], "title": "Clip Everywhere", "properties": {"group_restricted": false, "color_restricted": false, "Node name for S&R": "Anything Everywhere"}, "widgets_values": ["<comfy.sd.CLIP object at 0x000001E20E53D650>"]}, {"id": 27, "type": "Seed Everywhere", "pos": [4800, 480], "size": {"0": 300, "1": 120}, "flags": {}, "order": 10, "mode": 0, "outputs": [{"name": "INT", "type": "INT", "links": null, "shape": 3}], "properties": {"group_restricted": false, "color_restricted": false, "Node name for S&R": "Seed Everywhere"}, "widgets_values": [1056660035176881, "randomize", "930994575722780"]}, {"id": 3, "type": "CheckpointLoaderSimple", "pos": [4320, -240], "size": {"0": 315, "1": 98}, "flags": {}, "order": 11, "mode": 0, "outputs": [{"name": "MODEL", "type": "MODEL", "links": [1], "shape": 3, "slot_index": 0}, {"name": "CLIP", "type": "CLIP", "links": [29], "shape": 3, "slot_index": 1}, {"name": "VAE", "type": "VAE", "links": null, "shape": 3}], "properties": {"Node name for S&R": "CheckpointLoaderSimple"}, "widgets_values": ["Juggernaut\\juggernaut_reborn.safetensors"]}, {"id": 33, "type": "KSampler", "pos": [7080, -300], "size": [300, 480], "flags": {}, "order": 12, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": null}, {"name": "positive", "type": "CONDITIONING", "link": null}, {"name": "negative", "type": "CONDITIONING", "link": null}, {"name": "latent_image", "type": "LATENT", "link": null}], "outputs": [{"name": "LATENT", "type": "LATENT", "links": [30], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "KSampler"}, "widgets_values": [125512508779993, "randomize", 50, 10, "euler", "normal", 1]}, {"id": 35, "type": "Anything Everywhere", "pos": [4800, 1260], "size": {"0": 240, "1": 60}, "flags": {}, "order": 21, "mode": 0, "inputs": [{"name": "LATENT", "type": "*", "link": 26, "slot_index": 0, "color_on": "#FF9CF9"}], "title": "Empty Latent Everywhere", "properties": {"group_restricted": false, "color_restricted": false, "Node name for S&R": "Anything Everywhere"}, "widgets_values": ["Latent shape torch.Size([1, 4, 64, 96])"]}, {"id": 42, "type": "Anything Everywhere", "pos": [4800, 1440], "size": {"0": 240, "1": 60}, "flags": {}, "order": 26, "mode": 0, "inputs": [{"name": "UPSCALE_MODEL", "type": "*", "link": 34, "slot_index": 0, "color_on": ""}], "title": "Upscale Model Everywhere", "properties": {"group_restricted": false, "color_restricted": false, "Node name for S&R": "Anything Everywhere"}, "widgets_values": ["RRDBNet(\n (model): Sequential(\n (0): Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n (1): Identity + \n |Sequential(\n | (0): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (1): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (2): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (3): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (4): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (5): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (6): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (7): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (8): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (9): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (10): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (11): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (12): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (13): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (14): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (15): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (16): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (17): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (18): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (19): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (20): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (21): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (22): RRDB(\n | (RDB1): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB2): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | (RDB3): ResidualDenseBlock_5C(\n | (conv1): Sequential(\n | (0): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv2): Sequential(\n | (0): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv3): Sequential(\n | (0): Conv2d(128, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv4): Sequential(\n | (0): Conv2d(160, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | (1): LeakyReLU(negative_slope=0.2, inplace=True)\n | )\n | (conv5): Sequential(\n | (0): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n | )\n | )\n | )\n | (23): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n |)\n (2): Upsample(scale_factor=2.0, mode='nearest')\n (3): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n (4): LeakyReLU(negative_slope=0.2, inplace=True)\n (5): Upsample(scale_factor=2.0, mode='nearest')\n (6): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n (7): LeakyReLU(negative_slope=0.2, inplace=True)\n (8): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n (9): LeakyReLU(negative_slope=0.2, inplace=True)\n (10): Conv2d(64, 3, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n )\n)"]}, {"id": 43, "type": "UpscaleModelLoader", "pos": [4260, 1440], "size": {"0": 315, "1": 58}, "flags": {}, "order": 13, "mode": 0, "outputs": [{"name": "UPSCALE_MODEL", "type": "UPSCALE_MODEL", "links": [34], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "UpscaleModelLoader"}, "widgets_values": ["RealESRGAN_x4.pth"]}, {"id": 40, "type": "KSampler", "pos": [7920, -300], "size": {"0": 300, "1": 480}, "flags": {}, "order": 31, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": null}, {"name": "positive", "type": "CONDITIONING", "link": null}, {"name": "negative", "type": "CONDITIONING", "link": null}, {"name": "latent_image", "type": "LATENT", "link": 31}], "outputs": [{"name": "LATENT", "type": "LATENT", "links": [32], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "KSampler"}, "widgets_values": [91775796555809, "randomize", 50, 8, "euler", "normal", 0.5]}, {"id": 45, "type": "VAEEncode", "pos": [9180, -420], "size": {"0": 210, "1": 46}, "flags": {}, "order": 37, "mode": 0, "inputs": [{"name": "pixels", "type": "IMAGE", "link": 35}, {"name": "vae", "type": "VAE", "link": null}], "outputs": [{"name": "LATENT", "type": "LATENT", "links": [36], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "VAEEncode"}}, {"id": 36, "type": "VAEDecode", "pos": [8400, -360], "size": {"0": 210, "1": 46}, "flags": {}, "order": 33, "mode": 0, "inputs": [{"name": "samples", "type": "LATENT", "link": 32, "slot_index": 0}, {"name": "vae", "type": "VAE", "link": null}], "outputs": [{"name": "IMAGE", "type": "IMAGE", "links": [33], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "VAEDecode"}}, {"id": 37, "type": "PreviewImage", "pos": [9960, -420], "size": [240, 240], "flags": {}, "order": 41, "mode": 0, "inputs": [{"name": "images", "type": "IMAGE", "link": 38}], "properties": {"Node name for S&R": "PreviewImage"}}, {"id": 44, "type": "KSampler", "pos": [9240, -240], "size": {"0": 300, "1": 480}, "flags": {}, "order": 39, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": null}, {"name": "positive", "type": "CONDITIONING", "link": null}, {"name": "negative", "type": "CONDITIONING", "link": null}, {"name": "latent_image", "type": "LATENT", "link": 36}], "outputs": [{"name": "LATENT", "type": "LATENT", "links": [37], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "KSampler"}, "widgets_values": [519227386807816, "randomize", 50, 5, "euler", "normal", 0.2]}, {"id": 47, "type": "Image Comparer (rgthree)", "pos": {"0": 9960, "1": -60, "2": 0, "3": 0, "4": 0, "5": 0, "6": 0, "7": 0, "8": 0, "9": 0}, "size": [1140, 540], "flags": {}, "order": 42, "mode": 0, "inputs": [{"name": "image_a", "type": "IMAGE", "link": 42, "dir": 3}, {"name": "image_b", "type": "IMAGE", "link": 41, "dir": 3}], "outputs": [], "properties": {"comparer_mode": "Slide"}, "widgets_values": [["/view?filename=rgthree.compare._temp_krgzt_00007_.png&type=temp&subfolder=&rand=0.3813708940269229", "/view?filename=rgthree.compare._temp_krgzt_00008_.png&type=temp&subfolder=&rand=0.05611420278314272"]]}, {"id": 46, "type": "VAEDecode", "pos": [9600, -360], "size": {"0": 210, "1": 46}, "flags": {}, "order": 40, "mode": 0, "inputs": [{"name": "samples", "type": "LATENT", "link": 37, "slot_index": 0}, {"name": "vae", "type": "VAE", "link": null}], "outputs": [{"name": "IMAGE", "type": "IMAGE", "links": [38, 41], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "VAEDecode"}}, {"id": 39, "type": "KSampler", "pos": [7560, -300], "size": {"0": 300, "1": 480}, "flags": {}, "order": 25, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": null}, {"name": "positive", "type": "CONDITIONING", "link": null}, {"name": "negative", "type": "CONDITIONING", "link": null}, {"name": "latent_image", "type": "LATENT", "link": 30}], "outputs": [{"name": "LATENT", "type": "LATENT", "links": [31], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "KSampler"}, "widgets_values": [950626952914904, "randomize", 50, 8, "euler", "normal", 0.7000000000000001]}, {"id": 19, "type": "IPAdapterFaceID", "pos": [6240, -300], "size": {"0": 315, "1": 322}, "flags": {}, "order": 34, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": 19}, {"name": "ipadapter", "type": "IPADAPTER", "link": null}, {"name": "image", "type": "IMAGE", "link": 15}, {"name": "image_negative", "type": "IMAGE", "link": null}, {"name": "attn_mask", "type": "MASK", "link": 16}, {"name": "clip_vision", "type": "CLIP_VISION", "link": null}, {"name": "insightface", "type": "INSIGHTFACE", "link": null}], "outputs": [{"name": "MODEL", "type": "MODEL", "links": [48], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "IPAdapterFaceID"}, "widgets_values": [1, 1, "linear", "concat", 0, 1, "V only"]}, {"id": 41, "type": "UltimateSDUpscale", "pos": [8760, -300], "size": [300, 840], "flags": {}, "order": 35, "mode": 0, "inputs": [{"name": "image", "type": "IMAGE", "link": 33}, {"name": "model", "type": "MODEL", "link": null}, {"name": "positive", "type": "CONDITIONING", "link": null}, {"name": "negative", "type": "CONDITIONING", "link": null}, {"name": "vae", "type": "VAE", "link": null}, {"name": "upscale_model", "type": "UPSCALE_MODEL", "link": null}], "outputs": [{"name": "IMAGE", "type": "IMAGE", "links": [35, 42], "shape": 3, "slot_index": 0}], "properties": {"Node name for S&R": "UltimateSDUpscale"}, "widgets_values": [4, 1021561766371649, "randomize", 20, 8, "euler", "normal", 0.2, "Linear", 512, 512, 8, 32, "None", 1, 64, 8, 16, true, false]}, {"id": 23, "type": "CR Apply LoRA Stack", "pos": [6660, -480], "size": {"0": 254.40000915527344, "1": 66}, "flags": {}, "order": 36, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": 48}, {"name": "clip", "type": "CLIP", "link": null}, {"name": "lora_stack", "type": "LORA_STACK", "link": 21}], "outputs": [{"name": "MODEL", "type": "MODEL", "links": [22], "shape": 3}, {"name": "CLIP", "type": "CLIP", "links": null, "shape": 3}, {"name": "show_help", "type": "STRING", "links": null, "shape": 3}], "properties": {"Node name for S&R": "CR Apply LoRA Stack"}}, {"id": 12, "type": "LoadImagesFromDir //Inspire", "pos": [5280, -540], "size": {"0": 315, "1": 170}, "flags": {}, "order": 14, "mode": 0, "outputs": [{"name": "IMAGE", "type": "IMAGE", "links": [9, 10], "shape": 3, "slot_index": 0}, {"name": "MASK", "type": "MASK", "links": null, "shape": 3}, {"name": "INT", "type": "INT", "links": null, "shape": 3}], "properties": {"Node name for S&R": "LoadImagesFromDir //Inspire"}, "widgets_values": ["", 0, 0, false]}, {"id": 16, "type": "LoadImagesFromDir //Inspire", "pos": [5760, -540], "size": {"0": 315, "1": 170}, "flags": {}, "order": 15, "mode": 0, "outputs": [{"name": "IMAGE", "type": "IMAGE", "links": [12, 14], "shape": 3, "slot_index": 0}, {"name": "MASK", "type": "MASK", "links": null, "shape": 3}, {"name": "INT", "type": "INT", "links": null, "shape": 3}], "properties": {"Node name for S&R": "LoadImagesFromDir //Inspire"}, "widgets_values": ["", 0, 0, false]}, {"id": 20, "type": "LoadImagesFromDir //Inspire", "pos": [6240, -540], "size": {"0": 315, "1": 170}, "flags": {}, "order": 16, "mode": 0, "outputs": [{"name": "IMAGE", "type": "IMAGE", "links": [15, 17], "shape": 3, "slot_index": 0}, {"name": "MASK", "type": "MASK", "links": null, "shape": 3}, {"name": "INT", "type": "INT", "links": null, "shape": 3}], "properties": {"Node name for S&R": "LoadImagesFromDir //Inspire"}, "widgets_values": ["", 15, 0, false]}], "links": [[1, 3, 0, 4, 0, "MODEL"], [2, 4, 0, 2, 0, "MODEL"], [6, 8, 0, 9, 0, "*"], [7, 5, 0, 10, 0, "*"], [8, 6, 0, 11, 0, "*"], [9, 12, 0, 2, 2, "IMAGE"], [10, 12, 0, 13, 0, "IMAGE"], [11, 14, 0, 2, 4, "MASK"], [12, 16, 0, 15, 2, "IMAGE"], [13, 18, 0, 15, 4, "MASK"], [14, 16, 0, 17, 0, "IMAGE"], [15, 20, 0, 19, 2, "IMAGE"], [16, 22, 0, 19, 4, "MASK"], [17, 20, 0, 21, 0, "IMAGE"], [18, 2, 0, 15, 0, "MODEL"], [19, 15, 0, 19, 0, "MODEL"], [21, 25, 0, 23, 2, "LORA_STACK"], [22, 23, 0, 26, 0, "*"], [23, 28, 0, 29, 0, "*"], [24, 28, 1, 29, 1, "*"], [25, 30, 0, 32, 0, "*"], [26, 34, 0, 35, 0, "*"], [29, 3, 1, 38, 0, "*"], [30, 33, 0, 39, 3, "LATENT"], [31, 39, 0, 40, 3, "LATENT"], [32, 40, 0, 36, 0, "LATENT"], [33, 36, 0, 41, 0, "IMAGE"], [34, 43, 0, 42, 0, "*"], [35, 41, 0, 45, 0, "IMAGE"], [36, 45, 0, 44, 3, "LATENT"], [37, 44, 0, 46, 0, "LATENT"], [38, 46, 0, 37, 0, "IMAGE"], [41, 46, 0, 47, 1, "IMAGE"], [42, 41, 0, 47, 0, "IMAGE"], [48, 19, 0, 23, 0, "MODEL"]], "groups": [], "config": {}, "extra": {}, "version": 0.4} |