From a8bfdcfbd229bdfb66f177cdc4a56f3c846ef93b Mon Sep 17 00:00:00 2001 From: City <125218114+city96@users.noreply.github.com> Date: Sun, 18 Aug 2024 09:44:46 +0200 Subject: [PATCH] Fix Q4_0 #41 --- dequant.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/dequant.py b/dequant.py index c4a02d2..3a9aa71 100644 --- a/dequant.py +++ b/dequant.py @@ -103,7 +103,7 @@ def dequantize_blocks_Q4_1(blocks, block_size, type_size, dtype=None): def dequantize_blocks_Q4_0(blocks, block_size, type_size, dtype=None): n_blocks = blocks.shape[0] - d, qs = split_block_dims(blocks, 2, 2) + d, qs = split_block_dims(blocks, 2) d = d.view(torch.float16).to(dtype) qs = qs.reshape((n_blocks, -1, 1, block_size // 2)) >> torch.tensor([0, 4], device=d.device, dtype=torch.uint8).reshape((1, 1, 2, 1))