diff --git a/example_workflows/wanvideo_I2V_FantasyTalking_example_01.json b/example_workflows/wanvideo_I2V_FantasyTalking_example_01.json index 62c1114..629156a 100644 --- a/example_workflows/wanvideo_I2V_FantasyTalking_example_01.json +++ b/example_workflows/wanvideo_I2V_FantasyTalking_example_01.json @@ -1,22 +1,22 @@ { "id": "206247b6-9fec-4ed2-8927-e4f388c674d4", "revision": 0, - "last_node_id": 75, - "last_link_id": 112, + "last_node_id": 86, + "last_link_id": 122, "nodes": [ { "id": 46, "type": "WanVideoTextEmbedBridge", "pos": [ - 1204.152587890625, - 707.1484985351562 + 1015.870849609375, + 941.0142822265625 ], "size": [ 315, 46 ], "flags": {}, - "order": 23, + "order": 32, "mode": 2, "inputs": [ { @@ -48,15 +48,15 @@ "id": 50, "type": "CLIPTextEncode", "pos": [ - 754.153076171875, - 967.1485595703125 + 565.871337890625, + 1201.014404296875 ], "size": [ 400, 200 ], "flags": {}, - "order": 19, + "order": 26, "mode": 2, "inputs": [ { @@ -88,8 +88,8 @@ "id": 48, "type": "CLIPLoader", "pos": [ - 394.15313720703125, - 717.1484985351562 + 205.87127685546875, + 951.0142822265625 ], "size": [ 315, @@ -125,15 +125,15 @@ "id": 49, "type": "CLIPTextEncode", "pos": [ - 754.153076171875, - 717.1484985351562 + 565.871337890625, + 951.0142822265625 ], "size": [ 400, 200 ], "flags": {}, - "order": 18, + "order": 25, "mode": 2, "inputs": [ { @@ -165,8 +165,8 @@ "id": 42, "type": "Note", "pos": [ - -760, - -940 + -1010, + -1250 ], "size": [ 314.96246337890625, @@ -188,8 +188,8 @@ "id": 45, "type": "WanVideoVRAMManagement", "pos": [ - -390, - -760 + -640, + -1070 ], "size": [ 315, @@ -217,42 +217,19 @@ "color": "#223", "bgcolor": "#335" }, - { - "id": 36, - "type": "Note", - "pos": [ - -20, - -1190 - ], - "size": [ - 374.3061828613281, - 171.9547576904297 - ], - "flags": {}, - "order": 3, - "mode": 0, - "inputs": [], - "outputs": [], - "properties": {}, - "widgets_values": [ - "fp8_fast seems to cause huge quality degradation\n\nfp_16_fast enables \"Full FP16 Accmumulation in FP16 GEMMs\" feature available in the very latest pytorch nightly, this is around 20% speed boost. \n\nSageattn if you have it installed can be used for almost double inference speed" - ], - "color": "#432", - "bgcolor": "#653" - }, { "id": 33, "type": "Note", "pos": [ - -10, - -1330 + -260, + -1640 ], "size": [ 359.0753479003906, 88 ], "flags": {}, - "order": 4, + "order": 3, "mode": 0, "inputs": [], "outputs": [], @@ -267,15 +244,15 @@ "id": 51, "type": "Note", "pos": [ - 424.153076171875, - 547.1480712890625 + 235.87120056152344, + 781.0138549804688 ], "size": [ 253.16725158691406, 88 ], "flags": {}, - "order": 5, + "order": 4, "mode": 0, "inputs": [], "outputs": [], @@ -290,15 +267,15 @@ "id": 44, "type": "Note", "pos": [ - -800, - -1230 + -1050, + -1540 ], "size": [ 303.0501403808594, 88 ], "flags": {}, - "order": 6, + "order": 5, "mode": 0, "inputs": [], "outputs": [], @@ -313,15 +290,15 @@ "id": 59, "type": "CLIPVisionLoader", "pos": [ - -583.8698120117188, - -168.6510009765625 + -990, + -160 ], "size": [ 315, 58 ], "flags": {}, - "order": 7, + "order": 6, "mode": 0, "inputs": [], "outputs": [ @@ -356,13 +333,13 @@ 174 ], "flags": {}, - "order": 28, + "order": 39, "mode": 0, "inputs": [ { "name": "vae", "type": "WANVAE", - "link": 43 + "link": 118 }, { "name": "samples", @@ -399,15 +376,15 @@ "id": 65, "type": "WanVideoClipVisionEncode", "pos": [ - -176.9497833251953, - -167.4617462158203 + -590, + -160 ], "size": [ 327.5999755859375, 262 ], "flags": {}, - "order": 25, + "order": 33, "mode": 0, "inputs": [ { @@ -459,56 +436,6 @@ "color": "#233", "bgcolor": "#355" }, - { - "id": 16, - "type": "WanVideoTextEncode", - "pos": [ - 787.8640747070312, - -91.52558898925781 - ], - "size": [ - 420.30511474609375, - 261.5306701660156 - ], - "flags": {}, - "order": 24, - "mode": 0, - "inputs": [ - { - "name": "t5", - "type": "WANTEXTENCODER", - "link": 15 - }, - { - "name": "model_to_offload", - "shape": 7, - "type": "WANVIDEOMODEL", - "link": 79 - } - ], - "outputs": [ - { - "name": "text_embeds", - "type": "WANVIDEOTEXTEMBEDS", - "slot_index": 0, - "links": [ - 86 - ] - } - ], - "properties": { - "cnr_id": "ComfyUI-WanVideoWrapper", - "ver": "d9b1f4d1a5aea91d101ae97a54714a5861af3f50", - "Node name for S&R": "WanVideoTextEncode" - }, - "widgets_values": [ - "A woman is talking.", - "色调艳丽,过曝,静态,细节模糊不清,字幕,风格,作品,画作,画面,静止,整体发灰,最差质量,低质量,JPEG压缩残留,丑陋的,残缺的,多余的手指,画得不好的手部,画得不好的脸部,畸形的,毁容的,形态畸形的肢体,手指融合,静止不动的画面,杂乱的背景,三条腿,背景人很多,倒着走", - true - ], - "color": "#332922", - "bgcolor": "#593930" - }, { "id": 69, "type": "WanVideoSampler", @@ -518,10 +445,10 @@ ], "size": [ 317.4000244140625, - 943.75 + 869.4000244140625 ], "flags": {}, - "order": 27, + "order": 38, "mode": 0, "inputs": [ { @@ -604,6 +531,14 @@ "shape": 7, "type": "FANTASYTALKING_EMBEDS", "link": 101 + }, + { + "name": "steps", + "type": "INT", + "widget": { + "name": "steps" + }, + "link": 115 } ], "outputs": [ @@ -635,240 +570,19 @@ "" ] }, - { - "id": 11, - "type": "LoadWanVideoT5TextEncoder", - "pos": [ - 36.36686325073242, - -638.5691528320312 - ], - "size": [ - 377.1661376953125, - 130 - ], - "flags": {}, - "order": 8, - "mode": 0, - "inputs": [], - "outputs": [ - { - "name": "wan_t5_model", - "type": "WANTEXTENCODER", - "slot_index": 0, - "links": [ - 15 - ] - } - ], - "properties": { - "cnr_id": "ComfyUI-WanVideoWrapper", - "ver": "d9b1f4d1a5aea91d101ae97a54714a5861af3f50", - "Node name for S&R": "LoadWanVideoT5TextEncoder" - }, - "widgets_values": [ - "umt5-xxl-enc-bf16.safetensors", - "bf16", - "offload_device", - "disabled" - ], - "color": "#332922", - "bgcolor": "#593930" - }, - { - "id": 38, - "type": "WanVideoVAELoader", - "pos": [ - -361.60174560546875, - -629.2520751953125 - ], - "size": [ - 372.7727966308594, - 82 - ], - "flags": {}, - "order": 9, - "mode": 0, - "inputs": [], - "outputs": [ - { - "name": "vae", - "type": "WANVAE", - "slot_index": 0, - "links": [ - 43, - 81 - ] - } - ], - "properties": { - "cnr_id": "ComfyUI-WanVideoWrapper", - "ver": "d9b1f4d1a5aea91d101ae97a54714a5861af3f50", - "Node name for S&R": "WanVideoVAELoader" - }, - "widgets_values": [ - "wanvideo\\Wan2_1_VAE_bf16.safetensors", - "bf16" - ], - "color": "#322", - "bgcolor": "#533" - }, - { - "id": 71, - "type": "DownloadAndLoadWav2VecModel", - "pos": [ - -427.73681640625, - -473.0342102050781 - ], - "size": [ - 355.20001220703125, - 106 - ], - "flags": {}, - "order": 10, - "mode": 0, - "inputs": [], - "outputs": [ - { - "name": "wav2vec_model", - "type": "WAV2VECMODEL", - "links": [ - 99 - ] - } - ], - "properties": { - "cnr_id": "ComfyUI-WanVideoWrapper", - "ver": "df95c85283d7625fbdf664d0133a2e1c114ba14a", - "Node name for S&R": "DownloadAndLoadWav2VecModel" - }, - "widgets_values": [ - "facebook/wav2vec2-base-960h", - "fp16", - "main_device" - ] - }, - { - "id": 68, - "type": "FantasyTalkingModelLoader", - "pos": [ - -423.2041931152344, - -309.32421875 - ], - "size": [ - 340.20001220703125, - 82 - ], - "flags": {}, - "order": 11, - "mode": 0, - "inputs": [], - "outputs": [ - { - "name": "model", - "type": "FANTASYTALKINGMODEL", - "links": [ - 84, - 100 - ] - } - ], - "properties": { - "cnr_id": "ComfyUI-WanVideoWrapper", - "ver": "df95c85283d7625fbdf664d0133a2e1c114ba14a", - "Node name for S&R": "FantasyTalkingModelLoader" - }, - "widgets_values": [ - "WanVideo\\fantasytalking_fp16.safetensors", - "fp16" - ] - }, - { - "id": 72, - "type": "LoadAudio", - "pos": [ - -585.1002197265625, - -30.80118751525879 - ], - "size": [ - 315, - 136 - ], - "flags": {}, - "order": 12, - "mode": 0, - "inputs": [], - "outputs": [ - { - "name": "AUDIO", - "type": "AUDIO", - "links": [ - 95, - 102 - ] - } - ], - "properties": { - "cnr_id": "comfy-core", - "ver": "0.3.29", - "Node name for S&R": "LoadAudio" - }, - "widgets_values": [ - "woman.wav", - null, - null - ] - }, - { - "id": 52, - "type": "WanVideoTeaCache", - "pos": [ - 1280.1697998046875, - -855.5034790039062 - ], - "size": [ - 315, - 178 - ], - "flags": {}, - "order": 13, - "mode": 0, - "inputs": [], - "outputs": [ - { - "name": "teacache_args", - "type": "TEACACHEARGS", - "links": [ - 89 - ] - } - ], - "properties": { - "cnr_id": "ComfyUI-WanVideoWrapper", - "ver": "d9b1f4d1a5aea91d101ae97a54714a5861af3f50", - "Node name for S&R": "WanVideoTeaCache" - }, - "widgets_values": [ - 0.225, - 6, - -1, - "offload_device", - "true", - "e" - ] - }, { "id": 35, "type": "WanVideoTorchCompileSettings", "pos": [ - -460, - -1230 + -710, + -1540 ], "size": [ 421.6000061035156, 202 ], "flags": {}, - "order": 14, + "order": 7, "mode": 0, "inputs": [], "outputs": [ @@ -897,64 +611,369 @@ "bgcolor": "#335" }, { - "id": 39, - "type": "WanVideoBlockSwap", + "id": 58, + "type": "LoadImage", "pos": [ - -390, - -940 + -1060, + 180 ], "size": [ - 315, - 154 + 413.10479736328125, + 498.3180847167969 ], "flags": {}, - "order": 15, + "order": 8, "mode": 0, "inputs": [], "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 104 + ] + }, + { + "name": "MASK", + "type": "MASK", + "links": null + } + ], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.3.26", + "Node name for S&R": "LoadImage" + }, + "widgets_values": [ + "woman.png", + "image" + ], + "color": "#2a363b", + "bgcolor": "#3f5159" + }, + { + "id": 22, + "type": "WanVideoModelLoader", + "pos": [ + -280, + -1270 + ], + "size": [ + 477.4410095214844, + 254 + ], + "flags": {}, + "order": 31, + "mode": 0, + "inputs": [ + { + "name": "compile_args", + "shape": 7, + "type": "WANCOMPILEARGS", + "link": null + }, { "name": "block_swap_args", + "shape": 7, "type": "BLOCKSWAPARGS", + "link": 96 + }, + { + "name": "lora", + "shape": 7, + "type": "WANVIDLORA", + "link": null + }, + { + "name": "vram_management_args", + "shape": 7, + "type": "VRAM_MANAGEMENTARGS", + "link": null + }, + { + "name": "vace_model", + "shape": 7, + "type": "VACEPATH", + "link": null + }, + { + "name": "fantasytalking_model", + "shape": 7, + "type": "FANTASYTALKINGMODEL", + "link": 84 + } + ], + "outputs": [ + { + "name": "model", + "type": "WANVIDEOMODEL", "slot_index": 0, "links": [ - 96 + 79, + 85 ] } ], "properties": { "cnr_id": "ComfyUI-WanVideoWrapper", "ver": "d9b1f4d1a5aea91d101ae97a54714a5861af3f50", - "Node name for S&R": "WanVideoBlockSwap" + "Node name for S&R": "WanVideoModelLoader" }, "widgets_values": [ - 15, - false, - false, - true, - 0 + "WanVideo\\Wan2_1-I2V-14B-720P_fp8_e4m3fn.safetensors", + "fp16", + "fp8_e4m3fn", + "offload_device", + "sdpa" ], "color": "#223", "bgcolor": "#335" }, + { + "id": 76, + "type": "MarkdownNote", + "pos": [ + -1070, + -630 + ], + "size": [ + 446.07086181640625, + 88 + ], + "flags": {}, + "order": 9, + "mode": 0, + "inputs": [], + "outputs": [], + "title": "Model link", + "properties": {}, + "widgets_values": [ + "[https://huggingface.co/Kijai/WanVideo_comfy/blob/main/fantasytalking_fp16.safetensors](https://huggingface.co/Kijai/WanVideo_comfy/blob/main/fantasytalking_fp16.safetensors)" + ], + "color": "#432", + "bgcolor": "#653" + }, + { + "id": 77, + "type": "Note", + "pos": [ + -990, + -820 + ], + "size": [ + 373.65032958984375, + 139.77639770507812 + ], + "flags": {}, + "order": 10, + "mode": 0, + "inputs": [], + "outputs": [], + "properties": {}, + "widgets_values": [ + "Wave2vec modele is autodownloaded from:\n\nhttps://huggingface.co/facebook/wav2vec2-base-960h\n\nto:\n\nComfyUI/models/transformers/facebook/wav2vec2-base-960h\n\nThe .safetensor file + config files" + ], + "color": "#432", + "bgcolor": "#653" + }, + { + "id": 11, + "type": "LoadWanVideoT5TextEncoder", + "pos": [ + -210, + -950 + ], + "size": [ + 377.1661376953125, + 130 + ], + "flags": {}, + "order": 11, + "mode": 0, + "inputs": [], + "outputs": [ + { + "name": "wan_t5_model", + "type": "WANTEXTENCODER", + "slot_index": 0, + "links": [ + 15 + ] + } + ], + "properties": { + "cnr_id": "ComfyUI-WanVideoWrapper", + "ver": "d9b1f4d1a5aea91d101ae97a54714a5861af3f50", + "Node name for S&R": "LoadWanVideoT5TextEncoder" + }, + "widgets_values": [ + "umt5-xxl-enc-bf16.safetensors", + "bf16", + "offload_device", + "disabled" + ], + "color": "#332922", + "bgcolor": "#593930" + }, + { + "id": 80, + "type": "PreviewImage", + "pos": [ + -5.626028060913086, + 274.6432800292969 + ], + "size": [ + 210, + 246 + ], + "flags": {}, + "order": 34, + "mode": 0, + "inputs": [ + { + "name": "images", + "type": "IMAGE", + "link": 116 + } + ], + "outputs": [], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.3.29", + "Node name for S&R": "PreviewImage" + }, + "widgets_values": [] + }, + { + "id": 38, + "type": "WanVideoVAELoader", + "pos": [ + -610, + -940 + ], + "size": [ + 372.7727966308594, + 82 + ], + "flags": {}, + "order": 12, + "mode": 0, + "inputs": [], + "outputs": [ + { + "name": "vae", + "type": "WANVAE", + "slot_index": 0, + "links": [ + 117 + ] + } + ], + "properties": { + "cnr_id": "ComfyUI-WanVideoWrapper", + "ver": "d9b1f4d1a5aea91d101ae97a54714a5861af3f50", + "Node name for S&R": "WanVideoVAELoader" + }, + "widgets_values": [ + "wanvideo\\Wan2_1_VAE_bf16.safetensors", + "bf16" + ], + "color": "#322", + "bgcolor": "#533" + }, + { + "id": 81, + "type": "SetNode", + "pos": [ + -206.83546447753906, + -771.7630004882812 + ], + "size": [ + 210, + 60 + ], + "flags": { + "collapsed": true + }, + "order": 28, + "mode": 0, + "inputs": [ + { + "name": "WANVAE", + "type": "WANVAE", + "link": 117 + } + ], + "outputs": [ + { + "name": "*", + "type": "*", + "links": null + } + ], + "title": "Set_WanVAE", + "properties": { + "previousName": "WanVAE" + }, + "widgets_values": [ + "WanVAE" + ], + "color": "#322", + "bgcolor": "#533" + }, + { + "id": 82, + "type": "GetNode", + "pos": [ + 1688.8663330078125, + -708.3416748046875 + ], + "size": [ + 210, + 60 + ], + "flags": { + "collapsed": true + }, + "order": 13, + "mode": 0, + "inputs": [], + "outputs": [ + { + "name": "WANVAE", + "type": "WANVAE", + "links": [ + 118 + ] + } + ], + "title": "Get_WanVAE", + "properties": {}, + "widgets_values": [ + "WanVAE" + ], + "color": "#322", + "bgcolor": "#533" + }, { "id": 63, "type": "WanVideoImageToVideoEncode", "pos": [ - 334.3843078613281, - -156.56114196777344 + -8.957036018371582, + -156.54031372070312 ], "size": [ 352.79998779296875, 346 ], "flags": {}, - "order": 26, + "order": 37, "mode": 0, "inputs": [ { "name": "vae", "type": "WANVAE", - "link": 81 + "link": 119 }, { "name": "clip_embeds", @@ -1044,78 +1063,215 @@ "bgcolor": "#3f5159" }, { - "id": 73, - "type": "FantasyTalkingWav2VecEmbeds", + "id": 83, + "type": "GetNode", "pos": [ - 535.35791015625, - -405.02032470703125 + -171.16091918945312, + -154.3962860107422 ], "size": [ - 531.5999755859375, - 170 + 210, + 34 + ], + "flags": { + "collapsed": true + }, + "order": 14, + "mode": 0, + "inputs": [], + "outputs": [ + { + "name": "WANVAE", + "type": "WANVAE", + "links": [ + 119 + ] + } + ], + "title": "Get_WanVAE", + "properties": {}, + "widgets_values": [ + "WanVAE" + ], + "color": "#322", + "bgcolor": "#533" + }, + { + "id": 72, + "type": "LoadAudio", + "pos": [ + -982.36962890625, + -373.3668212890625 + ], + "size": [ + 315, + 136 ], "flags": {}, - "order": 21, + "order": 15, + "mode": 0, + "inputs": [], + "outputs": [ + { + "name": "AUDIO", + "type": "AUDIO", + "links": [ + 120 + ] + } + ], + "properties": { + "cnr_id": "comfy-core", + "ver": "0.3.29", + "Node name for S&R": "LoadAudio" + }, + "widgets_values": [ + "woman.wav", + null, + null + ], + "color": "#323", + "bgcolor": "#535" + }, + { + "id": 84, + "type": "SetNode", + "pos": [ + -589.344970703125, + -313.9405517578125 + ], + "size": [ + 210, + 60 + ], + "flags": { + "collapsed": true + }, + "order": 29, "mode": 0, "inputs": [ { - "name": "wav2vec_model", - "type": "WAV2VECMODEL", - "link": 99 - }, - { - "name": "fantasytalking_model", - "type": "FANTASYTALKINGMODEL", - "link": 100 - }, - { - "name": "audio", + "name": "AUDIO", "type": "AUDIO", - "link": 102 - }, - { - "name": "num_frames", - "type": "INT", - "widget": { - "name": "num_frames" - }, - "link": 112 + "link": 120 } ], "outputs": [ { - "name": "fantasytalking_embeds", - "type": "FANTASYTALKING_EMBEDS", + "name": "AUDIO", + "type": "AUDIO", "links": [ - 101 + 121 + ] + } + ], + "title": "Set_InputAudio", + "properties": { + "previousName": "InputAudio" + }, + "widgets_values": [ + "InputAudio" + ], + "color": "#323", + "bgcolor": "#535" + }, + { + "id": 85, + "type": "GetNode", + "pos": [ + 2110.023193359375, + -866.8945922851562 + ], + "size": [ + 210, + 60 + ], + "flags": { + "collapsed": true + }, + "order": 16, + "mode": 0, + "inputs": [], + "outputs": [ + { + "name": "AUDIO", + "type": "AUDIO", + "links": [ + 122 + ] + } + ], + "title": "Get_InputAudio", + "properties": {}, + "widgets_values": [ + "InputAudio" + ], + "color": "#323", + "bgcolor": "#535" + }, + { + "id": 16, + "type": "WanVideoTextEncode", + "pos": [ + 688.0745849609375, + -439.3515625 + ], + "size": [ + 420.30511474609375, + 261.5306701660156 + ], + "flags": {}, + "order": 36, + "mode": 0, + "inputs": [ + { + "name": "t5", + "type": "WANTEXTENCODER", + "link": 15 + }, + { + "name": "model_to_offload", + "shape": 7, + "type": "WANVIDEOMODEL", + "link": 79 + } + ], + "outputs": [ + { + "name": "text_embeds", + "type": "WANVIDEOTEXTEMBEDS", + "slot_index": 0, + "links": [ + 86 ] } ], "properties": { "cnr_id": "ComfyUI-WanVideoWrapper", - "ver": "df95c85283d7625fbdf664d0133a2e1c114ba14a", - "Node name for S&R": "FantasyTalkingWav2VecEmbeds" + "ver": "d9b1f4d1a5aea91d101ae97a54714a5861af3f50", + "Node name for S&R": "WanVideoTextEncode" }, "widgets_values": [ - 81, - 23, - 1, - 1 - ] + "A woman is talking.", + "色调艳丽,过曝,静态,细节模糊不清,字幕,风格,作品,画作,画面,静止,整体发灰,最差质量,低质量,JPEG压缩残留,丑陋的,残缺的,多余的手指,画得不好的手部,画得不好的脸部,畸形的,毁容的,形态畸形的肢体,手指融合,静止不动的画面,杂乱的背景,三条腿,背景人很多,倒着走", + true + ], + "color": "#332922", + "bgcolor": "#593930" }, { "id": 75, "type": "INTConstant", "pos": [ - -29.905210494995117, - 498.43133544921875 + 356.9280700683594, + -347.07037353515625 ], "size": [ - 200, + 210, 58 ], "flags": {}, - "order": 16, + "order": 17, "mode": 0, "inputs": [], "outputs": [ @@ -1140,6 +1296,313 @@ "color": "#1b4669", "bgcolor": "#29699c" }, + { + "id": 79, + "type": "INTConstant", + "pos": [ + 362.801513671875, + -481.2966003417969 + ], + "size": [ + 200, + 58 + ], + "flags": {}, + "order": 18, + "mode": 0, + "inputs": [], + "outputs": [ + { + "name": "value", + "type": "INT", + "links": [ + 114, + 115 + ] + } + ], + "title": "Steps", + "properties": { + "cnr_id": "comfyui-kjnodes", + "ver": "c3dc82108a2a86c17094107ead61d63f8c76200e", + "Node name for S&R": "INTConstant" + }, + "widgets_values": [ + 30 + ], + "color": "#1b4669", + "bgcolor": "#29699c" + }, + { + "id": 36, + "type": "Note", + "pos": [ + -270, + -1500 + ], + "size": [ + 374.3061828613281, + 171.9547576904297 + ], + "flags": {}, + "order": 19, + "mode": 0, + "inputs": [], + "outputs": [], + "properties": {}, + "widgets_values": [ + "fp8_fast seems to cause huge quality degradation\n\nfp_16_fast enables \"Full FP16 Accmumulation in FP16 GEMMs\" feature available in the very latest pytorch nightly, this is around 20% speed boost. \n\nSageattn if you have it installed can be used for almost double inference speed" + ], + "color": "#432", + "bgcolor": "#653" + }, + { + "id": 39, + "type": "WanVideoBlockSwap", + "pos": [ + -639.009033203125, + -1276.7559814453125 + ], + "size": [ + 315, + 154 + ], + "flags": {}, + "order": 20, + "mode": 0, + "inputs": [], + "outputs": [ + { + "name": "block_swap_args", + "type": "BLOCKSWAPARGS", + "slot_index": 0, + "links": [ + 96 + ] + } + ], + "properties": { + "cnr_id": "ComfyUI-WanVideoWrapper", + "ver": "d9b1f4d1a5aea91d101ae97a54714a5861af3f50", + "Node name for S&R": "WanVideoBlockSwap" + }, + "widgets_values": [ + 15, + false, + false, + true, + 0 + ], + "color": "#223", + "bgcolor": "#335" + }, + { + "id": 71, + "type": "DownloadAndLoadWav2VecModel", + "pos": [ + -600, + -790 + ], + "size": [ + 355.20001220703125, + 106 + ], + "flags": {}, + "order": 21, + "mode": 0, + "inputs": [], + "outputs": [ + { + "name": "wav2vec_model", + "type": "WAV2VECMODEL", + "links": [ + 99 + ] + } + ], + "properties": { + "cnr_id": "ComfyUI-WanVideoWrapper", + "ver": "df95c85283d7625fbdf664d0133a2e1c114ba14a", + "Node name for S&R": "DownloadAndLoadWav2VecModel" + }, + "widgets_values": [ + "facebook/wav2vec2-base-960h", + "fp16", + "main_device" + ], + "color": "#323", + "bgcolor": "#535" + }, + { + "id": 68, + "type": "FantasyTalkingModelLoader", + "pos": [ + -600, + -620 + ], + "size": [ + 340.20001220703125, + 82 + ], + "flags": {}, + "order": 22, + "mode": 0, + "inputs": [], + "outputs": [ + { + "name": "model", + "type": "FANTASYTALKINGMODEL", + "links": [ + 84, + 100 + ] + } + ], + "properties": { + "cnr_id": "ComfyUI-WanVideoWrapper", + "ver": "df95c85283d7625fbdf664d0133a2e1c114ba14a", + "Node name for S&R": "FantasyTalkingModelLoader" + }, + "widgets_values": [ + "WanVideo\\fantasytalking_fp16.safetensors", + "fp16" + ], + "color": "#323", + "bgcolor": "#535" + }, + { + "id": 73, + "type": "FantasyTalkingWav2VecEmbeds", + "pos": [ + 579.4556884765625, + -113.97596740722656 + ], + "size": [ + 531.5999755859375, + 170 + ], + "flags": {}, + "order": 35, + "mode": 0, + "inputs": [ + { + "name": "wav2vec_model", + "type": "WAV2VECMODEL", + "link": 99 + }, + { + "name": "fantasytalking_model", + "type": "FANTASYTALKINGMODEL", + "link": 100 + }, + { + "name": "audio", + "type": "AUDIO", + "link": 121 + }, + { + "name": "num_frames", + "type": "INT", + "widget": { + "name": "num_frames" + }, + "link": 112 + }, + { + "name": "audio_cfg_scale", + "type": "FLOAT", + "widget": { + "name": "audio_cfg_scale" + }, + "link": 113 + } + ], + "outputs": [ + { + "name": "fantasytalking_embeds", + "type": "FANTASYTALKING_EMBEDS", + "links": [ + 101 + ] + } + ], + "properties": { + "cnr_id": "ComfyUI-WanVideoWrapper", + "ver": "df95c85283d7625fbdf664d0133a2e1c114ba14a", + "Node name for S&R": "FantasyTalkingWav2VecEmbeds" + }, + "widgets_values": [ + 81, + 23, + 1, + 1 + ], + "color": "#323", + "bgcolor": "#535" + }, + { + "id": 74, + "type": "ImageResizeKJv2", + "pos": [ + -570, + 180 + ], + "size": [ + 315, + 242 + ], + "flags": {}, + "order": 27, + "mode": 0, + "inputs": [ + { + "name": "image", + "type": "IMAGE", + "link": 104 + } + ], + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 107, + 108, + 116 + ] + }, + { + "name": "width", + "type": "INT", + "links": [ + 109 + ] + }, + { + "name": "height", + "type": "INT", + "links": [ + 110 + ] + } + ], + "properties": { + "cnr_id": "comfyui-kjnodes", + "ver": "c3dc82108a2a86c17094107ead61d63f8c76200e", + "Node name for S&R": "ImageResizeKJv2" + }, + "widgets_values": [ + 512, + 512, + "lanczos", + "crop", + "0, 0, 0", + "center", + 2 + ], + "color": "#2a363b", + "bgcolor": "#3f5159" + }, { "id": 30, "type": "VHS_VideoCombine", @@ -1152,7 +1615,7 @@ 1322.8939208984375 ], "flags": {}, - "order": 29, + "order": 40, "mode": 0, "inputs": [ { @@ -1164,7 +1627,7 @@ "name": "audio", "shape": 7, "type": "AUDIO", - "link": 95 + "link": 122 }, { "name": "meta_batch", @@ -1215,186 +1678,117 @@ "fullpath": "N:\\AI\\ComfyUI\\output\\WanVideoWrapper_I2V_FantasyTalking_00002-audio.mp4" } } - } - }, - { - "id": 58, - "type": "LoadImage", - "pos": [ - -650.719482421875, - 173.7036895751953 - ], - "size": [ - 413.10479736328125, - 498.3180847167969 - ], - "flags": {}, - "order": 17, - "mode": 0, - "inputs": [], - "outputs": [ - { - "name": "IMAGE", - "type": "IMAGE", - "links": [ - 104 - ] - }, - { - "name": "MASK", - "type": "MASK", - "links": null - } - ], - "properties": { - "cnr_id": "comfy-core", - "ver": "0.3.26", - "Node name for S&R": "LoadImage" }, - "widgets_values": [ - "woman.png", - "image" - ], "color": "#2a363b", "bgcolor": "#3f5159" }, { - "id": 74, - "type": "ImageResizeKJv2", + "id": 52, + "type": "WanVideoTeaCache", "pos": [ - -161.7025909423828, - 174.3883056640625 + 1307.4208984375, + -664.744384765625 ], "size": [ 315, - 242 + 178 ], "flags": {}, - "order": 22, + "order": 23, "mode": 0, - "inputs": [ - { - "name": "image", - "type": "IMAGE", - "link": 104 - } - ], + "inputs": [], "outputs": [ { - "name": "IMAGE", - "type": "IMAGE", + "name": "teacache_args", + "type": "TEACACHEARGS", "links": [ - 107, - 108 - ] - }, - { - "name": "width", - "type": "INT", - "links": [ - 109 - ] - }, - { - "name": "height", - "type": "INT", - "links": [ - 110 - ] - } - ], - "properties": { - "cnr_id": "comfyui-kjnodes", - "ver": "c3dc82108a2a86c17094107ead61d63f8c76200e", - "Node name for S&R": "ImageResizeKJv2" - }, - "widgets_values": [ - 512, - 512, - "lanczos", - "crop", - "0, 0, 0", - "center", - 2 - ] - }, - { - "id": 22, - "type": "WanVideoModelLoader", - "pos": [ - -30, - -960 - ], - "size": [ - 477.4410095214844, - 254 - ], - "flags": {}, - "order": 20, - "mode": 0, - "inputs": [ - { - "name": "compile_args", - "shape": 7, - "type": "WANCOMPILEARGS", - "link": null - }, - { - "name": "block_swap_args", - "shape": 7, - "type": "BLOCKSWAPARGS", - "link": 96 - }, - { - "name": "lora", - "shape": 7, - "type": "WANVIDLORA", - "link": null - }, - { - "name": "vram_management_args", - "shape": 7, - "type": "VRAM_MANAGEMENTARGS", - "link": null - }, - { - "name": "vace_model", - "shape": 7, - "type": "VACEPATH", - "link": null - }, - { - "name": "fantasytalking_model", - "shape": 7, - "type": "FANTASYTALKINGMODEL", - "link": 84 - } - ], - "outputs": [ - { - "name": "model", - "type": "WANVIDEOMODEL", - "slot_index": 0, - "links": [ - 79, - 85 + 89 ] } ], "properties": { "cnr_id": "ComfyUI-WanVideoWrapper", "ver": "d9b1f4d1a5aea91d101ae97a54714a5861af3f50", - "Node name for S&R": "WanVideoModelLoader" + "Node name for S&R": "WanVideoTeaCache" }, "widgets_values": [ - "WanVideo\\Wan2_1-I2V-14B-720P_fp8_e4m3fn.safetensors", - "fp16", - "fp8_e4m3fn", + 0.225, + 6, + -1, "offload_device", - "sdpa" + "true", + "e" + ] + }, + { + "id": 78, + "type": "CreateCFGScheduleFloatList", + "pos": [ + 690.325927734375, + -689.5537719726562 ], - "color": "#223", - "bgcolor": "#335" + "size": [ + 403.1999816894531, + 178 + ], + "flags": {}, + "order": 30, + "mode": 0, + "inputs": [ + { + "name": "steps", + "type": "INT", + "widget": { + "name": "steps" + }, + "link": 114 + } + ], + "outputs": [ + { + "name": "float_list", + "type": "FLOAT", + "links": [ + 113 + ] + } + ], + "properties": { + "cnr_id": "ComfyUI-WanVideoWrapper", + "ver": "e3afc7fc758add9ba0ca7e6e219c30f312758484", + "Node name for S&R": "CreateCFGScheduleFloatList" + }, + "widgets_values": [ + 30, + 5, + 5, + "linear", + 0, + 0.1 + ] + }, + { + "id": 86, + "type": "Note", + "pos": [ + 686.342529296875, + -834.5337524414062 + ], + "size": [ + 410.2611389160156, + 90.62958526611328 + ], + "flags": {}, + "order": 24, + "mode": 0, + "inputs": [], + "outputs": [], + "properties": {}, + "widgets_values": [ + "Applying the audio_cfg only at the beginning can give most of it's benefit while not slowing down rest of the generation." + ], + "color": "#432", + "bgcolor": "#653" } ], "links": [ @@ -1414,14 +1808,6 @@ 0, "IMAGE" ], - [ - 43, - 38, - 0, - 28, - 0, - "VAE" - ], [ 52, 48, @@ -1470,14 +1856,6 @@ 1, "WANVIDEOMODEL" ], - [ - 81, - 38, - 0, - 63, - 0, - "WANVAE" - ], [ 82, 65, @@ -1534,14 +1912,6 @@ 1, "LATENT" ], - [ - 95, - 72, - 0, - 30, - 1, - "AUDIO" - ], [ 96, 39, @@ -1574,14 +1944,6 @@ 13, "FANTASYTALKING_EMBEDS" ], - [ - 102, - 72, - 0, - 73, - 2, - "AUDIO" - ], [ 104, 58, @@ -1637,6 +1999,86 @@ 73, 3, "INT" + ], + [ + 113, + 78, + 0, + 73, + 4, + "FLOAT" + ], + [ + 114, + 79, + 0, + 78, + 0, + "INT" + ], + [ + 115, + 79, + 0, + 69, + 14, + "INT" + ], + [ + 116, + 74, + 0, + 80, + 0, + "IMAGE" + ], + [ + 117, + 38, + 0, + 81, + 0, + "*" + ], + [ + 118, + 82, + 0, + 28, + 0, + "WANVAE" + ], + [ + 119, + 83, + 0, + 63, + 0, + "WANVAE" + ], + [ + 120, + 72, + 0, + 84, + 0, + "*" + ], + [ + 121, + 84, + 0, + 73, + 2, + "AUDIO" + ], + [ + 122, + 85, + 0, + 30, + 1, + "AUDIO" ] ], "groups": [ @@ -1644,23 +2086,36 @@ "id": 1, "title": "ComfyUI text encoding alternative", "bounding": [ - 331.337158203125, - 403.2147216796875, + 143.05531311035156, + 637.08056640625, 1210.621337890625, 805.9080810546875 ], "color": "#3f789e", "font_size": 24, "flags": {} + }, + { + "id": 2, + "title": "Models", + "bounding": [ + -1171.969482421875, + -1750.5992431640625, + 1456.9825439453125, + 1256.2666015625 + ], + "color": "#88A", + "font_size": 24, + "flags": {} } ], "config": {}, "extra": { "ds": { - "scale": 0.672749994932602, + "scale": 0.6115909044841843, "offset": [ - 1362.2293179125581, - 992.3831336690441 + 1733.9972278664989, + 1317.637074941902 ] }, "frontendVersion": "1.17.3",