dresser flex gen chunk push: chunk_001

3drawers/gen_flex/chunk_001/_logs/b1_ep00000000.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a residential mudroom with an entryway area, with soft warm gray walls and white wainscot panels, light yellow pine plank floor slightly worn, and a row of black iron coat hooks holding a denim jacket on the wall. In the foreground, there is a warm taupe-painted three-drawer dresser with matte gold bar pulls, and a pair of leather boots and a folded scarf rest on a low wooden bench beside it. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=274  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.10s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.52it/s]
  6%|▌         | 3/50 [00:02<00:39,  1.18it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.54it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.55it/s]
 14%|█▍        | 7/50 [00:05<00:33,  1.27it/s]
 16%|█▌        | 8/50 [00:05<00:28,  1.49it/s]
 18%|█▊        | 9/50 [00:05<00:20,  2.03it/s]
 20%|██        | 10/50 [00:06<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.71it/s]
 24%|██▍       | 12/50 [00:07<00:17,  2.20it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.75it/s]
 28%|██▊       | 14/50 [00:07<00:11,  3.05it/s]
 30%|███       | 15/50 [00:09<00:21,  1.66it/s]
 32%|███▏      | 16/50 [00:09<00:17,  1.91it/s]
 34%|███▍      | 17/50 [00:09<00:13,  2.42it/s]
 36%|███▌      | 18/50 [00:09<00:11,  2.75it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.12it/s]
 40%|████      | 20/50 [00:10<00:08,  3.62it/s]
 42%|████▏     | 21/50 [00:11<00:15,  1.90it/s]
 44%|████▍     | 22/50 [00:11<00:13,  2.05it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.08it/s]
 50%|█████     | 25/50 [00:12<00:07,  3.50it/s]
 52%|█████▏    | 26/50 [00:13<00:12,  1.94it/s]
 54%|█████▍    | 27/50 [00:13<00:10,  2.13it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.50it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.53it/s]
 64%|██████▍   | 32/50 [00:15<00:06,  2.62it/s]
 70%|███████   | 35/50 [00:16<00:05,  2.57it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.58it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.23it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.33it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  2.09it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.22it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.68it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.87it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.47it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.70it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.36it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.60it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.31it/s]
+
+  clip 1/4: 54.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.39it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 18%|█▊        | 9/50 [00:06<00:21,  1.90it/s]
 20%|██        | 10/50 [00:07<00:29,  1.36it/s]
 22%|██▏       | 11/50 [00:07<00:24,  1.60it/s]
 28%|██▊       | 14/50 [00:08<00:10,  3.37it/s]
 30%|███       | 15/50 [00:09<00:16,  2.06it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.11it/s]
 34%|███▍      | 17/50 [00:09<00:12,  2.57it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.03it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.42it/s]
 40%|████      | 20/50 [00:10<00:08,  3.51it/s]
 42%|████▏     | 21/50 [00:11<00:16,  1.77it/s]
 44%|████▍     | 22/50 [00:12<00:15,  1.83it/s]
 46%|████▌     | 23/50 [00:12<00:11,  2.35it/s]
 48%|████▊     | 24/50 [00:12<00:09,  2.81it/s]
 50%|█████     | 25/50 [00:12<00:07,  3.27it/s]
 52%|█████▏    | 26/50 [00:13<00:13,  1.73it/s]
 54%|█████▍    | 27/50 [00:14<00:11,  1.94it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.31it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.30it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.38it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.36it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.41it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.06it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.17it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.94it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.05it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.55it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.73it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.36it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.57it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.47it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.21it/s]
+
+  clip 2/4: 55.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.39it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.35it/s]
 10%|█         | 5/50 [00:04<00:42,  1.07it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.35it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.11it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 20%|██        | 10/50 [00:07<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.67it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.68it/s]
 30%|███       | 15/50 [00:09<00:16,  2.13it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.23it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.88it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.71it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.71it/s]
 48%|████▊     | 24/50 [00:11<00:06,  3.83it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.61it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.62it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.60it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.56it/s]
 64%|██████▍   | 32/50 [00:15<00:06,  2.59it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.47it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.50it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.12it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.21it/s]
 82%|█████��██▏ | 41/50 [00:19<00:04,  1.95it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.07it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.56it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.73it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.36it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.23it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.40it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.16it/s]
+
+  clip 3/4: 52.1s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.39it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.39it/s]
 10%|█         | 5/50 [00:04<00:40,  1.11it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.68it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.73it/s]
 30%|███       | 15/50 [00:09<00:16,  2.18it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.25it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.38it/s]
 40%|████      | 20/50 [00:09<00:06,  4.76it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.49it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.46it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.55it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.55it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.16it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.50it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.36it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.22it/s]
 74%|███████▍  | 37/50 [00:16<00:05,  2.57it/s]
 76%|███████▌  | 38/50 [00:17<00:06,  1.81it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  1.90it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.75it/s]
 84%|████████▍ | 42/50 [00:19<00:04,  1.90it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.46it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.66it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.54it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.21it/s]
+
+  clip 4/4: 48.9s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000000/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000000/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 11283.7ms  total=  11.3s
+  clip_total         n=  4  mean= 52761.3ms  total= 211.0s
 
+Done in 222.5s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000001.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a baby nursery with a changing area, with soft white walls and a gentle pale gray painted arch, light yellow maple plank floor, and a knit mobile turning slowly above a crib in warm diffused light. In the foreground, there is a weathered driftwood-gray three-drawer dresser with rope pull handles, and a folded muslin blanket and a small woven basket of toys sit on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=276  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:53,  1.09s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.54it/s]
  6%|▌         | 3/50 [00:02<00:39,  1.19it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.55it/s]
 10%|█         | 5/50 [00:03<00:35,  1.25it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.56it/s]
 14%|█▍        | 7/50 [00:05<00:33,  1.28it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:06<00:24,  1.64it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.75it/s]
 24%|██▍       | 12/50 [00:07<00:17,  2.22it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.42it/s]
 30%|███       | 15/50 [00:08<00:16,  2.08it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.12it/s]
 34%|███▍      | 17/50 [00:09<00:13,  2.51it/s]
 36%|███▌      | 18/50 [00:09<00:11,  2.80it/s]
 38%|███▊      | 19/50 [00:09<00:10,  3.02it/s]
 40%|████      | 20/50 [00:09<00:08,  3.59it/s]
 42%|████▏     | 21/50 [00:11<00:15,  1.93it/s]
 44%|████▍     | 22/50 [00:11<00:14,  1.87it/s]
 46%|████▌     | 23/50 [00:11<00:11,  2.36it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.69it/s]
 52%|█████▏    | 26/50 [00:13<00:11,  2.07it/s]
 54%|█████▍    | 27/50 [00:13<00:10,  2.20it/s]
 56%|█████▌    | 28/50 [00:13<00:08,  2.69it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.87it/s]
 62%|██████▏   | 31/50 [00:14<00:08,  2.17it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.25it/s]
 66%|██████▌   | 33/50 [00:15<00:06,  2.77it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.16it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.27it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.01it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.16it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.94it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.10it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.62it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.83it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.45it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.69it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.37it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.62it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.33it/s]
+
+  clip 1/4: 56.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.17s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.34it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.04it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.32it/s]
 10%|█         | 5/50 [00:04<00:41,  1.09it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.34it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.12it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 20%|██        | 10/50 [00:07<00:27,  1.48it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.68it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.20it/s]
 30%|███       | 15/50 [00:09<00:17,  2.05it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.01it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.05it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.70it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.40it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.64it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.63it/s]
 52%|█████▏    | 26/50 [00:13<00:11,  2.13it/s]
 54%|█████▍    | 27/50 [00:14<00:11,  2.05it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.48it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.93it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.19it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.15it/s]
 66%|██████▌   | 33/50 [00:16<00:06,  2.64it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.07it/s]
 72%|███████▏  | 36/50 [00:18<00:06,  2.08it/s]
 74%|███████▍  | 37/50 [00:18<00:05,  2.57it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.67it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.87it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.76it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.93it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.46it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.67it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.55it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.22it/s]
+
+  clip 2/4: 54.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:02,  1.27s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.35it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.38it/s]
 10%|█         | 5/50 [00:04<00:40,  1.12it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.40it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 20%|██        | 10/50 [00:07<00:26,  1.49it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.65it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.16it/s]
 30%|███       | 15/50 [00:09<00:17,  2.05it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.06it/s]
 36%|███▌      | 18/50 [00:09<00:10,  3.10it/s]
 40%|████      | 20/50 [00:10<00:06,  4.41it/s]
 44%|████▍     | 22/50 [00:11<00:12,  2.32it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.64it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.34it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.15it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.48it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.47it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.48it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.25it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.01it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.11it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.92it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.03it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.55it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.68it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.50it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.23it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.39it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.15it/s]
+
+  clip 3/4: 50.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.69it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.74it/s]
 30%|███       | 15/50 [00:09<00:16,  2.17it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.27it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.37it/s]
 40%|████      | 20/50 [00:09<00:06,  4.62it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.49it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.89it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.48it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.32it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.53it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.57it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.49it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.17it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.18it/s]
 76%|███████▌  | 38/50 [00:17<00:06,  1.89it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  1.97it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.39it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.62it/s]
 84%|████████▍ | 42/50 [00:19<00:04,  1.81it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.38it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.61it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.28it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.47it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.21it/s]
+
+  clip 4/4: 51.1s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000001/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000001/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 11778.6ms  total=  11.8s
+  clip_total         n=  4  mean= 52956.3ms  total= 211.8s
 
+Done in 224.0s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000002.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a tidy classroom with a learning area, with bright white walls and a long pale gray pinboard, light gray speckled vinyl floor, and rows of small desks under even fluorescent-style ceiling light. In the foreground, there is a classic white three-drawer dresser with faceted crystal glass knobs, and a globe and a neat cup of colored pencils are placed on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=276  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.32it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:31,  1.46it/s]
 10%|█         | 5/50 [00:04<00:37,  1.21it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.52it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:07<00:24,  1.65it/s]
 22%|██▏       | 11/50 [00:07<00:20,  1.86it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.92it/s]
 28%|██▊       | 14/50 [00:07<00:11,  3.20it/s]
 30%|███       | 15/50 [00:08<00:18,  1.91it/s]
 32%|███▏      | 16/50 [00:09<00:17,  1.99it/s]
 34%|███▍      | 17/50 [00:09<00:13,  2.43it/s]
 36%|███▌      | 18/50 [00:09<00:12,  2.56it/s]
 38%|███▊      | 19/50 [00:10<00:10,  2.98it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.28it/s]
 44%|████▍     | 22/50 [00:11<00:12,  2.29it/s]
 46%|████▌     | 23/50 [00:11<00:09,  2.84it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.37it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.69it/s]
 52%|█████▏    | 26/50 [00:13<00:12,  1.91it/s]
 54%|█████▍    | 27/50 [00:13<00:10,  2.13it/s]
 56%|█████▌    | 28/50 [00:13<00:07,  2.75it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.03it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.12it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.29it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.42it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.51it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.17it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.30it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  2.04it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.17it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.66it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.86it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.47it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.70it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.37it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.62it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.33it/s]
+
+  clip 1/4: 47.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:01,  1.25s/it]
  4%|▍         | 2/50 [00:01<00:40,  1.19it/s]
  6%|▌         | 3/50 [00:03<00:49,  1.05s/it]
  8%|▊         | 4/50 [00:03<00:38,  1.20it/s]
 10%|█         | 5/50 [00:04<00:44,  1.02it/s]
 12%|█▏        | 6/50 [00:05<00:34,  1.28it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.08it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.32it/s]
 20%|██        | 10/50 [00:08<00:27,  1.44it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.63it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.63it/s]
 30%|███       | 15/50 [00:09<00:16,  2.11it/s]
 32%|███▏      | 16/50 [00:10<00:15,  2.18it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.18it/s]
 40%|████      | 20/50 [00:10<00:06,  4.30it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.37it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.31it/s]
 46%|████▌     | 23/50 [00:12<00:09,  2.73it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.27it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.30it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.29it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.40it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.86it/s]
 62%|██████▏   | 31/50 [00:15<00:09,  2.03it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.16it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.39it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.00it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.06it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.86it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.93it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.28it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.54it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.71it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.31it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.52it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.24it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.47it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.21it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.44it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.19it/s]
+
+  clip 2/4: 50.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.32it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.04it/s]
  8%|▊         | 4/50 [00:03<00:35,  1.31it/s]
 10%|█         | 5/50 [00:04<00:41,  1.08it/s]
 12%|█▏        | 6/50 [00:04<00:33,  1.32it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.11it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.29it/s]
 18%|█▊        | 9/50 [00:06<00:23,  1.74it/s]
 20%|██        | 10/50 [00:08<00:31,  1.25it/s]
 22%|██▏       | 11/50 [00:08<00:27,  1.40it/s]
 24%|██▍       | 12/50 [00:08<00:22,  1.72it/s]
 26%|██▌       | 13/50 [00:09<00:16,  2.23it/s]
 30%|███       | 15/50 [00:10<00:18,  1.86it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.02it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.22it/s]
 40%|████      | 20/50 [00:10<00:06,  4.68it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.44it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.92it/s]
 54%|█████▍    | 27/50 [00:14<00:09,  2.44it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.22it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.43it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.42it/s]
 66%|██████▌   | 33/50 [00:16<00:05,  2.85it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.21it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.27it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.96it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.02it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.44it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.61it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.79it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.36it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.55it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.25it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.45it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.19it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.42it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.19it/s]
+
+  clip 3/4: 49.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:38,  1.26it/s]
  6%|▌         | 3/50 [00:02<00:47,  1.02s/it]
  8%|▊         | 4/50 [00:03<00:37,  1.23it/s]
 10%|█         | 5/50 [00:04<00:43,  1.04it/s]
 12%|█▏        | 6/50 [00:05<00:35,  1.25it/s]
 14%|█▍        | 7/50 [00:06<00:40,  1.06it/s]
 16%|█▌        | 8/50 [00:06<00:34,  1.23it/s]
 18%|█▊        | 9/50 [00:07<00:25,  1.59it/s]
 20%|██        | 10/50 [00:08<00:33,  1.20it/s]
 22%|██▏       | 11/50 [00:08<00:28,  1.37it/s]
 24%|██▍       | 12/50 [00:09<00:22,  1.72it/s]
 26%|██▌       | 13/50 [00:09<00:18,  2.03it/s]
 28%|██▊       | 14/50 [00:09<00:14,  2.44it/s]
 30%|███       | 15/50 [00:11<00:23,  1.49it/s]
 32%|███▏      | 16/50 [00:11<00:20,  1.64it/s]
 34%|███▍      | 17/50 [00:11<00:16,  2.06it/s]
 36%|███▌      | 18/50 [00:11<00:13,  2.44it/s]
 40%|████      | 20/50 [00:12<00:08,  3.74it/s]
 42%|████▏     | 21/50 [00:13<00:14,  1.98it/s]
 44%|████▍     | 22/50 [00:13<00:13,  2.05it/s]
 46%|████▌     | 23/50 [00:13<00:10,  2.59it/s]
 48%|████▊     | 24/50 [00:13<00:07,  3.27it/s]
 52%|█████▏    | 26/50 [00:15<00:10,  2.21it/s]
 54%|█████▍    | 27/50 [00:15<00:10,  2.23it/s]
 56%|█████▌    | 28/50 [00:15<00:08,  2.64it/s]
 58%|█████▊    | 29/50 [00:16<00:07,  2.93it/s]
 60%|██████    | 30/50 [00:16<00:06,  3.26it/s]
 62%|██████▏   | 31/50 [00:17<00:11,  1.71it/s]
 64%|██████▍   | 32/50 [00:18<00:09,  1.85it/s]
 66%|██████▌   | 33/50 [00:18<00:07,  2.38it/s]
 68%|██████▊   | 34/50 [00:18<00:05,  2.78it/s]
 70%|███████   | 35/50 [00:19<00:09,  1.56it/s]
 72%|███████▏  | 36/50 [00:20<00:07,  1.79it/s]
 76%|███████▌  | 38/50 [00:21<00:06,  1.71it/s]
 78%|███████▊  | 39/50 [00:21<00:05,  1.88it/s]
 82%|████████▏ | 41/50 [00:22<00:05,  1.77it/s]
 84%|████████▍ | 42/50 [00:23<00:04,  1.92it/s]
 86%|████████▌ | 43/50 [00:24<00:04,  1.47it/s]
 88%|████████▊ | 44/50 [00:24<00:03,  1.67it/s]
 90%|█████████ | 45/50 [00:25<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:26<00:02,  1.54it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:29<00:00,  1.21it/s]
+
+  clip 4/4: 44.8s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000002/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000002/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 11532.2ms  total=  11.5s
+  clip_total         n=  4  mean= 47943.6ms  total= 191.8s
 
+Done in 203.4s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000003.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a cozy cafe with a counter area, with warm white walls and a small chalkboard menu, light yellow oak plank floor, and a row of hanging Edison bulbs glowing over a wooden service counter. In the foreground, there is a cherry-wood three-drawer dresser with brushed-brass handles, and a ceramic mug and a small jar of sugar sit on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=267  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.11s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.52it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.17it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.54it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.54it/s]
 14%|█▍        | 7/50 [00:05<00:33,  1.27it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.51it/s]
 18%|█▊        | 9/50 [00:05<00:21,  1.95it/s]
 20%|██        | 10/50 [00:07<00:28,  1.40it/s]
 22%|██▏       | 11/50 [00:07<00:24,  1.61it/s]
 24%|██▍       | 12/50 [00:07<00:18,  2.05it/s]
 26%|██▌       | 13/50 [00:07<00:14,  2.61it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.92it/s]
 30%|███       | 15/50 [00:09<00:21,  1.65it/s]
 32%|███▏      | 16/50 [00:09<00:20,  1.63it/s]
 34%|███▍      | 17/50 [00:10<00:17,  1.92it/s]
 36%|███▌      | 18/50 [00:10<00:13,  2.32it/s]
 40%|████      | 20/50 [00:10<00:08,  3.64it/s]
 42%|████▏     | 21/50 [00:11<00:13,  2.10it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.06it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.52it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.08it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.81it/s]
 52%|█████▏    | 26/50 [00:13<00:12,  1.95it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.11it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.70it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.41it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.05it/s]
 62%|██████▏   | 31/50 [00:15<00:09,  1.98it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.06it/s]
 66%|██████▌   | 33/50 [00:16<00:06,  2.69it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.12it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.21it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.98it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.09it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.92it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.08it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.60it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.81it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.44it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.68it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.36it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.61it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.32it/s]
+
+  clip 1/4: 56.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.17s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:42,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.15it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:05<00:38,  1.13it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.35it/s]
 18%|█▊        | 9/50 [00:06<00:23,  1.77it/s]
 20%|██        | 10/50 [00:07<00:31,  1.27it/s]
 22%|██▏       | 11/50 [00:08<00:26,  1.46it/s]
 24%|██▍       | 12/50 [00:08<00:21,  1.73it/s]
 26%|██▌       | 13/50 [00:08<00:18,  1.96it/s]
 28%|██▊       | 14/50 [00:09<00:16,  2.22it/s]
 30%|███       | 15/50 [00:10<00:23,  1.46it/s]
 32%|███▏      | 16/50 [00:10<00:20,  1.63it/s]
 34%|███▍      | 17/50 [00:11<00:16,  2.02it/s]
 36%|███▌      | 18/50 [00:11<00:13,  2.29it/s]
 38%|███▊      | 19/50 [00:11<00:12,  2.56it/s]
 40%|████      | 20/50 [00:11<00:09,  3.24it/s]
 42%|████▏     | 21/50 [00:13<00:17,  1.70it/s]
 44%|████▍     | 22/50 [00:13<00:15,  1.86it/s]
 48%|████▊     | 24/50 [00:13<00:08,  2.97it/s]
 50%|█████     | 25/50 [00:13<00:07,  3.45it/s]
 52%|█████▏    | 26/50 [00:15<00:12,  1.86it/s]
 54%|█████▍    | 27/50 [00:15<00:11,  2.00it/s]
 56%|█████▌    | 28/50 [00:15<00:08,  2.53it/s]
 58%|█████▊    | 29/50 [00:15<00:06,  3.15it/s]
 60%|██████    | 30/50 [00:15<00:05,  3.87it/s]
 62%|██████▏   | 31/50 [00:17<00:10,  1.88it/s]
 64%|██████▍   | 32/50 [00:17<00:08,  2.07it/s]
 68%|██████▊   | 34/50 [00:17<00:04,  3.48it/s]
 70%|███████   | 35/50 [00:18<00:07,  1.97it/s]
 72%|███████▏  | 36/50 [00:19<00:06,  2.13it/s]
 76%|███████▌  | 38/50 [00:20<00:06,  1.89it/s]
 78%|███████▊  | 39/50 [00:20<00:05,  2.03it/s]
 82%|████████▏ | 41/50 [00:21<00:04,  1.84it/s]
 84%|████████▍ | 42/50 [00:22<00:04,  1.97it/s]
 86%|████████▌ | 43/50 [00:23<00:04,  1.50it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.69it/s]
 90%|█████████ | 45/50 [00:25<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.56it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.22it/s]
+
+  clip 2/4: 55.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.32it/s]
 10%|█         | 5/50 [00:04<00:40,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.34it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.11it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 20%|██        | 10/50 [00:07<00:26,  1.48it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.62it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.51it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.01it/s]
 30%|███       | 15/50 [00:09<00:19,  1.77it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.91it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.10it/s]
 40%|████      | 20/50 [00:10<00:06,  4.49it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.38it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.36it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.49it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.53it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.13it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.48it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.34it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.26it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.03it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.12it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.92it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.03it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.56it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.72it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.35it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.23it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.41it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.18it/s]
+
+  clip 3/4: 51.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.76it/s]
 30%|███       | 15/50 [00:09<00:16,  2.16it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.26it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.35it/s]
 40%|████      | 20/50 [00:09<00:06,  4.60it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.48it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.44it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.53it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.50it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.49it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.52it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.45it/s]
 66%|██████▌   | 33/50 [00:15<00:05,  2.86it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.37it/s]
 70%|███████   | 35/50 [00:16<00:07,  1.90it/s]
 72%|███████▏  | 36/50 [00:16<00:07,  1.97it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.76it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  1.91it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.78it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.93it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.48it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.67it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.54it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.22it/s]
+
+  clip 4/4: 49.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000003/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000003/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 11946.6ms  total=  11.9s
+  clip_total         n=  4  mean= 53174.2ms  total= 212.7s
 
+Done in 224.9s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000004.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a medical clinic waiting area with a reception area, with clean white walls and a soft gray accent panel, light gray sealed-vinyl floor, and calm even light from a wide ceiling diffuser. In the foreground, there is a light gray three-drawer dresser with simple round white knobs, and a small potted plant and a neat fan of magazines rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=266  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.10s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.53it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.17it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.54it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.54it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:06<00:24,  1.64it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.85it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.51it/s]
 30%|███       | 15/50 [00:08<00:15,  2.25it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.37it/s]
 36%|███▌      | 18/50 [00:08<00:08,  3.59it/s]
 42%|████▏     | 21/50 [00:10<00:09,  3.02it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.81it/s]
 46%|████▌     | 23/50 [00:10<00:08,  3.14it/s]
 48%|████▊     | 24/50 [00:10<00:08,  3.23it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.60it/s]
 52%|█████▏    | 26/50 [00:12<00:12,  2.00it/s]
 54%|█████▍    | 27/50 [00:12<00:10,  2.11it/s]
 56%|█████▌    | 28/50 [00:12<00:08,  2.67it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.96it/s]
 62%|██████▏   | 31/50 [00:14<00:08,  2.14it/s]
 64%|██████▍   | 32/50 [00:14<00:08,  2.16it/s]
 66%|██████▌   | 33/50 [00:14<00:06,  2.69it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.33it/s]
 70%|███████   | 35/50 [00:16<00:08,  1.85it/s]
 72%|███████▏  | 36/50 [00:16<00:07,  1.81it/s]
 74%|███████▍  | 37/50 [00:16<00:06,  2.11it/s]
 76%|███████▌  | 38/50 [00:18<00:08,  1.45it/s]
 78%|███████▊  | 39/50 [00:18<00:07,  1.57it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.03it/s]
 82%|████████▏ | 41/50 [00:19<00:06,  1.45it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.70it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.35it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.56it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.29it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.54it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.28it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.52it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.27it/s]
+
+  clip 1/4: 41.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.20s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.41it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.35it/s]
 14%|█▍        | 7/50 [00:05<00:38,  1.13it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 20%|██        | 10/50 [00:07<00:27,  1.48it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.61it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.04it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.55it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.98it/s]
 30%|███       | 15/50 [00:09<00:21,  1.62it/s]
 32%|███▏      | 16/50 [00:10<00:20,  1.66it/s]
 34%|███▍      | 17/50 [00:10<00:15,  2.12it/s]
 36%|███▌      | 18/50 [00:10<00:12,  2.62it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.24it/s]
 40%|████      | 20/50 [00:11<00:07,  3.78it/s]
 42%|████▏     | 21/50 [00:12<00:16,  1.76it/s]
 44%|████▍     | 22/50 [00:12<00:15,  1.75it/s]
 46%|████▌     | 23/50 [00:13<00:12,  2.10it/s]
 48%|████▊     | 24/50 [00:13<00:11,  2.32it/s]
 50%|█████     | 25/50 [00:13<00:09,  2.61it/s]
 52%|█████▏    | 26/50 [00:15<00:15,  1.54it/s]
 54%|█████▍    | 27/50 [00:15<00:13,  1.65it/s]
 56%|█████▌    | 28/50 [00:15<00:11,  2.00it/s]
 58%|█████▊    | 29/50 [00:15<00:08,  2.43it/s]
 60%|██████    | 30/50 [00:16<00:07,  2.77it/s]
 62%|██████▏   | 31/50 [00:17<00:12,  1.58it/s]
 64%|██████▍   | 32/50 [00:17<00:10,  1.67it/s]
 66%|██████▌   | 33/50 [00:18<00:08,  2.06it/s]
 68%|██████▊   | 34/50 [00:18<00:06,  2.56it/s]
 70%|███████   | 35/50 [00:19<00:09,  1.58it/s]
 72%|███████▏  | 36/50 [00:20<00:08,  1.59it/s]
 74%|███████▍  | 37/50 [00:20<00:06,  2.07it/s]
 76%|███████▌  | 38/50 [00:21<00:08,  1.43it/s]
 78%|███████▊  | 39/50 [00:22<00:06,  1.58it/s]
 80%|████████  | 40/50 [00:22<00:04,  2.06it/s]
 82%|████████▏ | 41/50 [00:23<00:06,  1.39it/s]
 84%|████████▍ | 42/50 [00:23<00:05,  1.54it/s]
 86%|████████▌ | 43/50 [00:25<00:05,  1.23it/s]
 88%|████████▊ | 44/50 [00:25<00:04,  1.45it/s]
 90%|█████████ | 45/50 [00:26<00:04,  1.18it/s]
 92%|█████████▏| 46/50 [00:27<00:03,  1.25it/s]
 94%|█████████▍| 47/50 [00:28<00:02,  1.07it/s]
 96%|█████████▌| 48/50 [00:29<00:01,  1.31it/s]
 98%|█████████▊| 49/50 [00:30<00:00,  1.13it/s]
+
+  clip 2/4: 50.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.39it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.40it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.40it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.76it/s]
 30%|███       | 15/50 [00:09<00:15,  2.19it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.28it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.38it/s]
 40%|████      | 20/50 [00:09<00:06,  4.60it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.43it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.36it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.45it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.38it/s]
 56%|█████▌    | 28/50 [00:13<00:08,  2.59it/s]
 58%|█████▊    | 29/50 [00:13<00:08,  2.54it/s]
 60%|██████    | 30/50 [00:14<00:07,  2.68it/s]
 62%|██████▏   | 31/50 [00:15<00:11,  1.66it/s]
 64%|██████▍   | 32/50 [00:16<00:10,  1.72it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.20it/s]
 70%|███████   | 35/50 [00:17<00:08,  1.86it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.88it/s]
 74%|███████▍  | 37/50 [00:18<00:05,  2.25it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.52it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.65it/s]
 80%|████████  | 40/50 [00:20<00:05,  1.96it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.39it/s]
 84%|████████▍ | 42/50 [00:21<00:05,  1.52it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.22it/s]
 88%|████████▊ | 44/50 [00:23<00:04,  1.32it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.10it/s]
 92%|█████████▏| 46/50 [00:25<00:03,  1.26it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.08it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.27it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.08it/s]
+
+  clip 3/4: 45.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.39it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.77it/s]
 30%|███       | 15/50 [00:09<00:16,  2.17it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.27it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.21it/s]
 40%|████      | 20/50 [00:09<00:06,  4.47it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.41it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.46it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.22it/s]
 52%|█████▏    | 26/50 [00:12<00:10,  2.38it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.37it/s]
 56%|█████▌    | 28/50 [00:13<00:07,  2.85it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.57it/s]
 64%|██████▍   | 32/50 [00:15<00:06,  2.60it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.48it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.51it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.13it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.17it/s]
 80%|████████  | 40/50 [00:18<00:03,  2.61it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.73it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.74it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.32it/s]
 88%|████████▊ | 44/50 [00:21<00:04,  1.48it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.21it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.33it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.11it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.25it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.06it/s]
+
+  clip 4/4: 46.8s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000004/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000004/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 10880.6ms  total=  10.9s
+  clip_total         n=  4  mean= 45938.9ms  total= 183.8s
 
+Done in 194.9s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000005.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a modern furniture showroom with a display area, with clean white walls and recessed ceiling spotlights washing them evenly, light gray sealed concrete floor in a smooth wide pour, and a single linen-upholstered armchair set back as a display piece. In the foreground, there is a smooth matte white three-drawer dresser with brushed-steel bar handles, and a small folded brochure and a square ceramic vase are placed on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=266  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:56,  1.16s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.40it/s]
  6%|▌         | 3/50 [00:02<00:42,  1.11it/s]
  8%|▊         | 4/50 [00:03<00:31,  1.45it/s]
 10%|█         | 5/50 [00:04<00:37,  1.19it/s]
 12%|█▏        | 6/50 [00:04<00:29,  1.49it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.23it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.51it/s]
 18%|█▊        | 9/50 [00:06<00:19,  2.05it/s]
 20%|██        | 10/50 [00:07<00:27,  1.48it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.74it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.98it/s]
 30%|███       | 15/50 [00:08<00:15,  2.29it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.41it/s]
 36%|███▌      | 18/50 [00:09<00:08,  3.58it/s]
 42%|████▏     | 21/50 [00:10<00:09,  3.00it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.98it/s]
 48%|████▊     | 24/50 [00:10<00:06,  4.13it/s]
 52%|█████▏    | 26/50 [00:12<00:08,  2.72it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.66it/s]
 56%|█████▌    | 28/50 [00:12<00:07,  3.09it/s]
 58%|█████▊    | 29/50 [00:12<00:05,  3.58it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.25it/s]
 62%|██████▏   | 31/50 [00:14<00:09,  2.10it/s]
 64%|██████▍   | 32/50 [00:14<00:08,  2.19it/s]
 66%|██████▌   | 33/50 [00:14<00:06,  2.79it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.51it/s]
 70%|███████   | 35/50 [00:15<00:07,  1.90it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.03it/s]
 74%|███████▍  | 37/50 [00:16<00:05,  2.58it/s]
 76%|███████▌  | 38/50 [00:17<00:07,  1.62it/s]
 78%|███████▊  | 39/50 [00:18<00:06,  1.74it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.24it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.51it/s]
 84%|████████▍ | 42/50 [00:19<00:04,  1.71it/s]
 86%|████████▌ | 43/50 [00:20<00:05,  1.35it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.55it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.26it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.51it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.51it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.26it/s]
+
+  clip 1/4: 52.5s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.41it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.56it/s]
 24%|██▍       | 12/50 [00:08<00:21,  1.80it/s]
 26%|██▌       | 13/50 [00:08<00:17,  2.17it/s]
 28%|██▊       | 14/50 [00:08<00:14,  2.57it/s]
 30%|███       | 15/50 [00:10<00:22,  1.57it/s]
 32%|███▏      | 16/50 [00:10<00:20,  1.66it/s]
 34%|███▍      | 17/50 [00:10<00:17,  1.89it/s]
 36%|███▌      | 18/50 [00:11<00:14,  2.19it/s]
 38%|███▊      | 19/50 [00:11<00:11,  2.65it/s]
 40%|████      | 20/50 [00:11<00:10,  2.94it/s]
 42%|████▏     | 21/50 [00:12<00:18,  1.57it/s]
 44%|████▍     | 22/50 [00:13<00:16,  1.67it/s]
 46%|████▌     | 23/50 [00:13<00:12,  2.10it/s]
 48%|████▊     | 24/50 [00:13<00:10,  2.50it/s]
 50%|█████     | 25/50 [00:14<00:08,  3.06it/s]
 52%|█████▏    | 26/50 [00:15<00:14,  1.67it/s]
 54%|█████▍    | 27/50 [00:15<00:12,  1.78it/s]
 56%|█████▌    | 28/50 [00:15<00:09,  2.32it/s]
 58%|█████▊    | 29/50 [00:16<00:07,  2.86it/s]
 60%|██████    | 30/50 [00:16<00:05,  3.58it/s]
 62%|██████▏   | 31/50 [00:17<00:10,  1.79it/s]
 64%|██████▍   | 32/50 [00:17<00:09,  1.86it/s]
 66%|██████▌   | 33/50 [00:17<00:06,  2.44it/s]
 68%|██████▊   | 34/50 [00:18<00:05,  3.11it/s]
 70%|███████   | 35/50 [00:19<00:08,  1.72it/s]
 72%|███████▏  | 36/50 [00:19<00:07,  1.88it/s]
 76%|███████▌  | 38/50 [00:20<00:06,  1.75it/s]
 78%|███████▊  | 39/50 [00:21<00:05,  1.92it/s]
 82%|████████▏ | 41/50 [00:22<00:05,  1.78it/s]
 84%|████████▍ | 42/50 [00:22<00:04,  1.93it/s]
 86%|████████▌ | 43/50 [00:24<00:04,  1.46it/s]
 88%|████████▊ | 44/50 [00:24<00:03,  1.66it/s]
 90%|█████████ | 45/50 [00:25<00:03,  1.31it/s]
 92%|█████████▏| 46/50 [00:26<00:02,  1.54it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.47it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.21it/s]
+
+  clip 2/4: 53.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.39it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:35,  1.29it/s]
 10%|█         | 5/50 [00:04<00:42,  1.06it/s]
 12%|█▏        | 6/50 [00:05<00:34,  1.27it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.08it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.30it/s]
 20%|██        | 10/50 [00:07<00:28,  1.41it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.56it/s]
 24%|██▍       | 12/50 [00:08<00:20,  1.90it/s]
 26%|██▌       | 13/50 [00:08<00:15,  2.36it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.95it/s]
 30%|███       | 15/50 [00:10<00:20,  1.70it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.87it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.79it/s]
 42%|████▏     | 21/50 [00:11<00:11,  2.59it/s]
 44%|████▍     | 22/50 [00:12<00:10,  2.57it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.70it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.53it/s]
 54%|█████▍    | 27/50 [00:14<00:08,  2.57it/s]
 58%|█████▊    | 29/50 [00:14<00:05,  3.60it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.54it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.56it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.71it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.19it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.25it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.95it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.08it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.87it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.01it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.51it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.68it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.55it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.21it/s]
+
+  clip 3/4: 50.5s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.37it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.39it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 18%|█▊        | 9/50 [00:06<00:21,  1.86it/s]
 20%|██        | 10/50 [00:07<00:29,  1.34it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.58it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.59it/s]
 30%|███       | 15/50 [00:09<00:17,  2.03it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.09it/s]
 34%|███▍      | 17/50 [00:10<00:12,  2.57it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.14it/s]
 40%|████      | 20/50 [00:10<00:06,  4.80it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.30it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.26it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.57it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.47it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.51it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.32it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.50it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.43it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.48it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.12it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.21it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.96it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.08it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.57it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.74it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.37it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.57it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.21it/s]
+
+  clip 4/4: 49.6s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000005/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000005/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 11886.3ms  total=  11.9s
+  clip_total         n=  4  mean= 51582.8ms  total= 206.3s
 
+Done in 218.4s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000006.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a bright sunroom with a relaxing area, with white-painted walls and white window mullions, light yellow pine plank floor, and a wall of tall windows pouring warm sunlight over a few leafy potted plants. In the foreground, there is a natural rattan-and-wood three-drawer dresser with small wooden knobs, and a watering can and a small terracotta pot sit on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=277  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.12s/it]
  4%|▍         | 2/50 [00:01<00:32,  1.48it/s]
  6%|▌         | 3/50 [00:02<00:41,  1.14it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.50it/s]
 10%|█         | 5/50 [00:04<00:36,  1.22it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.52it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:06<00:24,  1.64it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.85it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.42it/s]
 30%|███       | 15/50 [00:08<00:16,  2.16it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.28it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.99it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.83it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.85it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.61it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.68it/s]
 56%|█████▌    | 28/50 [00:12<00:07,  3.06it/s]
 58%|█████▊    | 29/50 [00:12<00:06,  3.29it/s]
 60%|██████    | 30/50 [00:12<00:05,  3.65it/s]
 62%|██████▏   | 31/50 [00:13<00:09,  2.07it/s]
 64%|██████▍   | 32/50 [00:14<00:08,  2.17it/s]
 66%|██████▌   | 33/50 [00:14<00:06,  2.71it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.31it/s]
 70%|███████   | 35/50 [00:15<00:08,  1.87it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.06it/s]
 76%|███████▌  | 38/50 [00:17<00:06,  1.91it/s]
 78%|███████▊  | 39/50 [00:17<00:05,  2.08it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  1.94it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.10it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.60it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.80it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.43it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.58it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.31it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.55it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.29it/s]
+
+  clip 1/4: 50.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.17s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.41it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.39it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 20%|██        | 10/50 [00:07<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.57it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.95it/s]
 26%|██▌       | 13/50 [00:08<00:16,  2.28it/s]
 28%|██▊       | 14/50 [00:08<00:14,  2.57it/s]
 30%|███       | 15/50 [00:10<00:22,  1.55it/s]
 32%|███▏      | 16/50 [00:10<00:22,  1.54it/s]
 34%|███▍      | 17/50 [00:11<00:18,  1.83it/s]
 36%|███▌      | 18/50 [00:11<00:15,  2.13it/s]
 38%|███▊      | 19/50 [00:11<00:12,  2.39it/s]
 40%|████      | 20/50 [00:11<00:10,  2.76it/s]
 42%|████▏     | 21/50 [00:13<00:18,  1.58it/s]
 44%|████▍     | 22/50 [00:13<00:16,  1.74it/s]
 46%|████▌     | 23/50 [00:13<00:11,  2.25it/s]
 48%|████▊     | 24/50 [00:13<00:09,  2.77it/s]
 50%|█████     | 25/50 [00:14<00:07,  3.31it/s]
 52%|█████▏    | 26/50 [00:15<00:13,  1.73it/s]
 54%|█████▍    | 27/50 [00:15<00:12,  1.88it/s]
 58%|█████▊    | 29/50 [00:15<00:06,  3.01it/s]
 60%|██████    | 30/50 [00:16<00:05,  3.34it/s]
 62%|██████▏   | 31/50 [00:17<00:10,  1.82it/s]
 64%|██████▍   | 32/50 [00:17<00:09,  1.93it/s]
 66%|██████▌   | 33/50 [00:17<00:07,  2.41it/s]
 68%|██████▊   | 34/50 [00:18<00:05,  2.94it/s]
 70%|███████   | 35/50 [00:19<00:08,  1.67it/s]
 72%|███████▏  | 36/50 [00:19<00:07,  1.82it/s]
 76%|███████▌  | 38/50 [00:20<00:07,  1.69it/s]
 78%|███████▊  | 39/50 [00:21<00:05,  1.86it/s]
 82%|████████▏ | 41/50 [00:22<00:05,  1.76it/s]
 84%|████████▍ | 42/50 [00:22<00:04,  1.91it/s]
 86%|████████▌ | 43/50 [00:24<00:04,  1.47it/s]
 88%|████████▊ | 44/50 [00:24<00:03,  1.67it/s]
 90%|█████████ | 45/50 [00:25<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:26<00:02,  1.55it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.22it/s]
+
+  clip 2/4: 55.5s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.17s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.44it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.34it/s]
 10%|█         | 5/50 [00:04<00:41,  1.09it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.35it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.11it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.35it/s]
 20%|██        | 10/50 [00:07<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.62it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.47it/s]
 30%|███       | 15/50 [00:09<00:16,  2.07it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.13it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.49it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.75it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.25it/s]
 40%|████      | 20/50 [00:10<00:07,  3.76it/s]
 42%|████▏     | 21/50 [00:11<00:15,  1.86it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.05it/s]
 50%|█████     | 25/50 [00:12<00:06,  4.07it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.28it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.37it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.19it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.40it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.42it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.39it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.45it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.09it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.19it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.95it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.06it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.56it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.74it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.37it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.55it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.45it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.18it/s]
+
+  clip 3/4: 52.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.17s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.41it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.15it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.43it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.43it/s]
 20%|██        | 10/50 [00:07<00:26,  1.52it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.69it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.73it/s]
 30%|███       | 15/50 [00:09<00:16,  2.19it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.29it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.97it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.74it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.73it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.34it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.54it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.58it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.70it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.60it/s]
 64%|██████▍   | 32/50 [00:14<00:06,  2.62it/s]
 70%|███████   | 35/50 [00:15<00:05,  2.50it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.54it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.14it/s]
 78%|███████▊  | 39/50 [00:17<00:04,  2.23it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.98it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.10it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.58it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.76it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.38it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.60it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.29it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.51it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.24it/s]
+
+  clip 4/4: 47.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000006/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000006/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 12218.3ms  total=  12.2s
+  clip_total         n=  4  mean= 51484.6ms  total= 205.9s
 
+Done in 218.5s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000007.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a walk-in dressing room with a wardrobe area, with soft white walls and slim gray trim, white-painted herringbone wood floor, and a tall mirror reflecting a row of neatly hung clothes under warm rail lighting. In the foreground, there is a soft greige-painted three-drawer dresser with polished brass knobs, and a folded silk scarf and a small jewelry tray rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=301  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:55,  1.14s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.41it/s]
  6%|▌         | 3/50 [00:02<00:42,  1.12it/s]
  8%|▊         | 4/50 [00:03<00:31,  1.46it/s]
 10%|█         | 5/50 [00:04<00:37,  1.21it/s]
 12%|█▏        | 6/50 [00:04<00:29,  1.50it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.24it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.50it/s]
 20%|██        | 10/50 [00:07<00:24,  1.60it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.74it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.75it/s]
 30%|███       | 15/50 [00:08<00:15,  2.26it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.26it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.93it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.83it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.85it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.59it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.65it/s]
 56%|█████▌    | 28/50 [00:12<00:07,  2.97it/s]
 58%|█████▊    | 29/50 [00:12<00:06,  3.33it/s]
 60%|██████    | 30/50 [00:12<00:05,  3.90it/s]
 62%|██████▏   | 31/50 [00:14<00:08,  2.14it/s]
 64%|██████▍   | 32/50 [00:14<00:08,  2.13it/s]
 66%|██████▌   | 33/50 [00:14<00:06,  2.64it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.30it/s]
 70%|███████   | 35/50 [00:15<00:07,  1.88it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.03it/s]
 76%|███████▌  | 38/50 [00:17<00:06,  1.84it/s]
 78%|███████▊  | 39/50 [00:17<00:05,  2.02it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  1.88it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.04it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.57it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.78it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.42it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.64it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.33it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.56it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.30it/s]
+
+  clip 1/4: 49.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:40,  1.12it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.40it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 20%|██        | 10/50 [00:07<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.61it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.99it/s]
 26%|██▌       | 13/50 [00:08<00:15,  2.39it/s]
 28%|██▊       | 14/50 [00:08<00:13,  2.75it/s]
 30%|███       | 15/50 [00:09<00:21,  1.62it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.80it/s]
 34%|███▍      | 17/50 [00:10<00:14,  2.31it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.81it/s]
 38%|███▊      | 19/50 [00:10<00:10,  3.01it/s]
 40%|████      | 20/50 [00:11<00:10,  2.90it/s]
 42%|████▏     | 21/50 [00:12<00:18,  1.57it/s]
 44%|████▍     | 22/50 [00:13<00:16,  1.72it/s]
 46%|████▌     | 23/50 [00:13<00:12,  2.18it/s]
 48%|████▊     | 24/50 [00:13<00:09,  2.71it/s]
 50%|█████     | 25/50 [00:13<00:07,  3.23it/s]
 52%|█████▏    | 26/50 [00:14<00:14,  1.66it/s]
 54%|█████▍    | 27/50 [00:15<00:12,  1.82it/s]
 58%|█████▊    | 29/50 [00:15<00:06,  3.01it/s]
 62%|██████▏   | 31/50 [00:16<00:08,  2.20it/s]
 64%|██████▍   | 32/50 [00:17<00:07,  2.29it/s]
 68%|██████▊   | 34/50 [00:17<00:04,  3.48it/s]
 70%|███████   | 35/50 [00:18<00:07,  2.08it/s]
 72%|███████▏  | 36/50 [00:18<00:06,  2.20it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.93it/s]
 78%|███████▊  | 39/50 [00:20<00:05,  2.06it/s]
 82%|████████▏ | 41/50 [00:21<00:04,  1.85it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  2.00it/s]
 86%|████████▌ | 43/50 [00:23<00:04,  1.51it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.70it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.56it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.47it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.21it/s]
+
+  clip 2/4: 52.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.35it/s]
 14%|█▍        | 7/50 [00:05<00:38,  1.12it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.30it/s]
 20%|██        | 10/50 [00:07<00:28,  1.41it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.56it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.54it/s]
 30%|███       | 15/50 [00:09<00:16,  2.08it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.18it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.21it/s]
 40%|████      | 20/50 [00:10<00:07,  4.27it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.33it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.40it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.35it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.73it/s]
 52%|█████▏    | 26/50 [00:13<00:12,  1.98it/s]
 54%|█████▍    | 27/50 [00:13<00:11,  2.08it/s]
 56%|█████▌    | 28/50 [00:13<00:08,  2.63it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.03it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.13it/s]
 64%|██████▍   | 32/50 [00:15<00:08,  2.24it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.30it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.37it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.03it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.15it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.92it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.05it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.54it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.70it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.35it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.23it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.46it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.21it/s]
+
+  clip 3/4: 49.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:37,  1.28it/s]
  6%|▌         | 3/50 [00:02<00:46,  1.01it/s]
  8%|▊         | 4/50 [00:03<00:35,  1.28it/s]
 10%|█         | 5/50 [00:04<00:42,  1.06it/s]
 12%|█▏        | 6/50 [00:05<00:33,  1.32it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.11it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.36it/s]
 20%|██        | 10/50 [00:07<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.66it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.19it/s]
 30%|███       | 15/50 [00:09<00:17,  2.03it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.16it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.22it/s]
 40%|████      | 20/50 [00:10<00:06,  4.39it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.37it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.40it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.51it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.44it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.46it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.45it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.43it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.44it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.53it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.11it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.19it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.92it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.05it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.86it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.99it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.51it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.70it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.55it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.22it/s]
+
+  clip 4/4: 47.2s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000007/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000007/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 12942.6ms  total=  12.9s
+  clip_total         n=  4  mean= 49885.9ms  total= 199.5s
 
+Done in 212.7s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000008.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a quiet reading nook with a window area, with warm gray walls and a white window seat, light yellow oak plank floor, and a soft pool of lamplight beside a cushioned bench under a tall bookshelf. In the foreground, there is an olive-green painted three-drawer dresser with aged-brass knobs, and an open paperback and a steaming mug of tea sit on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=260  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:53,  1.09s/it]
  4%|▍         | 2/50 [00:01<00:30,  1.55it/s]
  6%|▌         | 3/50 [00:02<00:39,  1.18it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.55it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.55it/s]
 14%|█▍        | 7/50 [00:05<00:33,  1.27it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.55it/s]
 20%|██        | 10/50 [00:06<00:24,  1.64it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.85it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.51it/s]
 30%|███       | 15/50 [00:08<00:15,  2.25it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.37it/s]
 38%|███▊      | 19/50 [00:08<00:07,  4.18it/s]
 42%|████▏     | 21/50 [00:10<00:09,  2.91it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.92it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.71it/s]
 54%|█████▍    | 27/50 [00:11<00:08,  2.76it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.15it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.71it/s]
 70%|███████   | 35/50 [00:14<00:05,  2.64it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.68it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.31it/s]
 78%|███████▊  | 39/50 [00:16<00:04,  2.40it/s]
 82%|████████▏ | 41/50 [00:17<00:04,  2.13it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.25it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.71it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.90it/s]
 90%|█████████ | 45/50 [00:20<00:03,  1.49it/s]
 92%|█████████▏| 46/50 [00:20<00:02,  1.71it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.38it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.62it/s]
 98%|█████████▊| 49/50 [00:23<00:00,  1.33it/s]
+
+  clip 1/4: 42.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.40it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.49it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.69it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.25it/s]
 30%|███       | 15/50 [00:09<00:16,  2.08it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.19it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.24it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.54it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.40it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.46it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.17it/s]
 52%|█████▏    | 26/50 [00:12<00:10,  2.39it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.36it/s]
 56%|█████▌    | 28/50 [00:13<00:07,  2.82it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.27it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.34it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.33it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.18it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.96it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.07it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.89it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.01it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.54it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.65it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.30it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.46it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.21it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.36it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.15it/s]
+
+  clip 2/4: 48.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.37it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.35it/s]
 10%|█         | 5/50 [00:04<00:41,  1.09it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.63it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.65it/s]
 30%|███       | 15/50 [00:09<00:16,  2.15it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.23it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.26it/s]
 40%|████      | 20/50 [00:10<00:06,  4.53it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.38it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.33it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.48it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.49it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.43it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.51it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.50it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.54it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.16it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.27it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.96it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.09it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.88it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.98it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.51it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.70it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.55it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.43it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.16it/s]
+
+  clip 3/4: 50.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.33it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.04it/s]
  8%|▊         | 4/50 [00:03<00:35,  1.30it/s]
 10%|█         | 5/50 [00:04<00:41,  1.08it/s]
 12%|█▏        | 6/50 [00:04<00:33,  1.32it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.10it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.33it/s]
 20%|██        | 10/50 [00:07<00:27,  1.46it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.58it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.52it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.96it/s]
 30%|███       | 15/50 [00:09<00:19,  1.78it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.95it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.90it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.25it/s]
 40%|████      | 20/50 [00:10<00:08,  3.70it/s]
 42%|████▏     | 21/50 [00:11<00:15,  1.93it/s]
 44%|████▍     | 22/50 [00:12<00:14,  1.94it/s]
 46%|████▌     | 23/50 [00:12<00:11,  2.38it/s]
 48%|████▊     | 24/50 [00:12<00:08,  2.98it/s]
 52%|█████▏    | 26/50 [00:14<00:11,  2.15it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.21it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.41it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.39it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.46it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.55it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.12it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.24it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.95it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.08it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.88it/s]
 84%|████████▍ | 42/50 [00:21<00:03,  2.02it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.52it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.68it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.50it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.45it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.21it/s]
+
+  clip 4/4: 50.3s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000008/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000008/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13400.8ms  total=  13.4s
+  clip_total         n=  4  mean= 47983.4ms  total= 191.9s
 
+Done in 205.5s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000009.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a photography studio with a shooting area, with seamless white walls and a white sweep backdrop, light gray smooth concrete floor, and a softbox light stand glowing evenly to one side. In the foreground, there is an industrial-style three-drawer dresser with a metal frame and reclaimed-wood drawer fronts and black pipe handles, and a folded gray reflector and a small clapperboard rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=249  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:39,  1.20it/s]
  6%|▌         | 3/50 [00:02<00:46,  1.00it/s]
  8%|▊         | 4/50 [00:03<00:38,  1.21it/s]
 10%|█         | 5/50 [00:04<00:42,  1.05it/s]
 12%|█▏        | 6/50 [00:05<00:33,  1.30it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.13it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.34it/s]
 18%|█▊        | 9/50 [00:06<00:22,  1.79it/s]
 20%|██        | 10/50 [00:07<00:29,  1.36it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.55it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.03it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.57it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.25it/s]
 30%|███       | 15/50 [00:09<00:19,  1.79it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.94it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.25it/s]
 40%|████      | 20/50 [00:10<00:06,  4.83it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.52it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.50it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.63it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.62it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.34it/s]
 64%|██████▍   | 32/50 [00:15<00:06,  2.65it/s]
 70%|███████   | 35/50 [00:16<00:05,  2.59it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.62it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.25it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.34it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  2.10it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.12it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.62it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.79it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.44it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.65it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.35it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.58it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.31it/s]
+
+  clip 1/4: 47.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:02,  1.28s/it]
  4%|▍         | 2/50 [00:01<00:42,  1.12it/s]
  6%|▌         | 3/50 [00:03<00:51,  1.09s/it]
  8%|▊         | 4/50 [00:03<00:39,  1.16it/s]
 10%|█         | 5/50 [00:05<00:45,  1.01s/it]
 12%|█▏        | 6/50 [00:05<00:34,  1.26it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.08it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.34it/s]
 20%|██        | 10/50 [00:08<00:27,  1.46it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.65it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.03it/s]
 30%|███       | 15/50 [00:10<00:16,  2.18it/s]
 32%|███▏      | 16/50 [00:10<00:14,  2.27it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.35it/s]
 40%|████      | 20/50 [00:10<00:06,  4.57it/s]
 42%|████▏     | 21/50 [00:11<00:11,  2.43it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.43it/s]
 46%|████▌     | 23/50 [00:12<00:09,  2.90it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.59it/s]
 54%|█████▍    | 27/50 [00:14<00:08,  2.60it/s]
 58%|█████▊    | 29/50 [00:14<00:05,  3.73it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.58it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.57it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.72it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.20it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.18it/s]
 74%|███████▍  | 37/50 [00:17<00:04,  2.67it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.69it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.81it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.28it/s]
 82%|████████▏ | 41/50 [00:20<00:06,  1.49it/s]
 84%|████████▍ | 42/50 [00:21<00:05,  1.58it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.24it/s]
 88%|████████▊ | 44/50 [00:22<00:04,  1.47it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.20it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.43it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.18it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.39it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.16it/s]
+
+  clip 2/4: 48.5s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.24s/it]
  4%|▍         | 2/50 [00:01<00:43,  1.11it/s]
  6%|▌         | 3/50 [00:03<00:49,  1.06s/it]
  8%|▊         | 4/50 [00:03<00:39,  1.15it/s]
 10%|█         | 5/50 [00:04<00:45,  1.00s/it]
 12%|█▏        | 6/50 [00:05<00:36,  1.21it/s]
 14%|█▍        | 7/50 [00:06<00:40,  1.05it/s]
 16%|█▌        | 8/50 [00:07<00:34,  1.23it/s]
 20%|██        | 10/50 [00:08<00:29,  1.36it/s]
 22%|██▏       | 11/50 [00:08<00:26,  1.49it/s]
 24%|██▍       | 12/50 [00:09<00:20,  1.82it/s]
 26%|██▌       | 13/50 [00:09<00:16,  2.26it/s]
 30%|███       | 15/50 [00:10<00:18,  1.88it/s]
 32%|███▏      | 16/50 [00:11<00:17,  1.92it/s]
 34%|███▍      | 17/50 [00:11<00:13,  2.38it/s]
 40%|████      | 20/50 [00:11<00:06,  4.41it/s]
 42%|████▏     | 21/50 [00:12<00:12,  2.39it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.45it/s]
 50%|█████     | 25/50 [00:13<00:05,  4.29it/s]
 54%|█████▍    | 27/50 [00:14<00:09,  2.45it/s]
 56%|█████▌    | 28/50 [00:14<00:07,  2.82it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.29it/s]
 62%|██████▏   | 31/50 [00:16<00:08,  2.34it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.36it/s]
 66%|██████▌   | 33/50 [00:16<00:05,  2.85it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.18it/s]
 72%|███████▏  | 36/50 [00:18<00:06,  2.20it/s]
 74%|███████▍  | 37/50 [00:18<00:04,  2.69it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.70it/s]
 78%|███████▊  | 39/50 [00:20<00:05,  1.88it/s]
 82%|████████▏ | 41/50 [00:21<00:05,  1.70it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.85it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.42it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.57it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.27it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.49it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.38it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.14it/s]
+
+  clip 3/4: 49.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:03,  1.30s/it]
  4%|▍         | 2/50 [00:01<00:42,  1.14it/s]
  6%|▌         | 3/50 [00:03<00:49,  1.05s/it]
  8%|▊         | 4/50 [00:03<00:37,  1.22it/s]
 10%|█         | 5/50 [00:04<00:43,  1.04it/s]
 12%|█▏        | 6/50 [00:05<00:35,  1.26it/s]
 14%|█▍        | 7/50 [00:06<00:40,  1.07it/s]
 16%|█▌        | 8/50 [00:06<00:33,  1.26it/s]
 18%|█▊        | 9/50 [00:07<00:23,  1.72it/s]
 20%|██        | 10/50 [00:08<00:30,  1.30it/s]
 22%|██▏       | 11/50 [00:08<00:26,  1.49it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.61it/s]
 30%|███       | 15/50 [00:10<00:16,  2.10it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.06it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.49it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.67it/s]
 42%|████▏     | 21/50 [00:12<00:11,  2.49it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.49it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.56it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.49it/s]
 54%|█████▍    | 27/50 [00:14<00:09,  2.50it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.21it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.45it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.50it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.43it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.47it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.10it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.20it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.95it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.06it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.55it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.68it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.50it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.21it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.43it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.19it/s]
+
+  clip 4/4: 45.2s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000009/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000009/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 10232.5ms  total=  10.2s
+  clip_total         n=  4  mean= 47851.2ms  total= 191.4s
 
+Done in 201.8s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000010.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a coastal vacation rental kitchenette with a breakfast area, with soft white walls and horizontal shiplap on the lower half, light yellow weathered pine plank floor, and warm morning sun filtering through a sheer linen curtain on the right. In the foreground, there is a light ash three-drawer dresser with smooth recessed cut-out handles, and a small ceramic bowl of seashells and a stack of striped dish towels rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=274  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:56,  1.15s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.40it/s]
  6%|▌         | 3/50 [00:02<00:41,  1.12it/s]
  8%|▊         | 4/50 [00:03<00:31,  1.48it/s]
 10%|█         | 5/50 [00:04<00:37,  1.21it/s]
 12%|█▏        | 6/50 [00:04<00:29,  1.52it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.25it/s]
 16%|█▌        | 8/50 [00:05<00:28,  1.47it/s]
 18%|█▊        | 9/50 [00:06<00:20,  2.00it/s]
 20%|██        | 10/50 [00:07<00:27,  1.46it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.73it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.94it/s]
 30%|███       | 15/50 [00:08<00:15,  2.30it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.38it/s]
 38%|███▊      | 19/50 [00:09<00:07,  4.21it/s]
 42%|████▏     | 21/50 [00:10<00:09,  2.92it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.89it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.65it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.74it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.15it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.69it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.54it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.46it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.20it/s]
 78%|███████▊  | 39/50 [00:16<00:04,  2.30it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  2.09it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.12it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.64it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.74it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.38it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.60it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.33it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.57it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.31it/s]
+
+  clip 1/4: 53.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.17s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.39it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 28%|██▊       | 14/50 [00:07<00:11,  3.23it/s]
 30%|███       | 15/50 [00:09<00:16,  2.08it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.19it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.89it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.67it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.71it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.35it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.51it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.49it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.46it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.96it/s]
 62%|██████▏   | 31/50 [00:14<00:09,  2.08it/s]
 64%|██████▍   | 32/50 [00:14<00:08,  2.20it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.43it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.00it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.02it/s]
 74%|███████▍  | 37/50 [00:16<00:05,  2.45it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.55it/s]
 78%|███████▊  | 39/50 [00:18<00:06,  1.76it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.29it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.51it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.74it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.33it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.56it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.25it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.49it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.45it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.20it/s]
+
+  clip 2/4: 51.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.32it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.04it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.37it/s]
 10%|█         | 5/50 [00:04<00:40,  1.12it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.39it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 20%|██        | 10/50 [00:07<00:27,  1.45it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.65it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.63it/s]
 30%|███       | 15/50 [00:09<00:16,  2.10it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.15it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.12it/s]
 40%|████      | 20/50 [00:10<00:07,  4.27it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.36it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.44it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.70it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.53it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.41it/s]
 56%|█████▌    | 28/50 [00:13<00:07,  2.86it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.59it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.51it/s]
 66%|██████▌   | 33/50 [00:15<00:05,  2.91it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.39it/s]
 70%|███████   | 35/50 [00:16<00:07,  1.89it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.95it/s]
 74%|███████▍  | 37/50 [00:17<00:05,  2.45it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.55it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.77it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.71it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.89it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.44it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.65it/s]
 90%|██████��██ | 45/50 [00:23<00:03,  1.31it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.47it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.21it/s]
+
+  clip 3/4: 50.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.17s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.45it/s]
  6%|▌         | 3/50 [00:02<00:42,  1.10it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.44it/s]
 10%|█         | 5/50 [00:04<00:38,  1.15it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.44it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.18it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.44it/s]
 20%|██        | 10/50 [00:07<00:26,  1.53it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.72it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.29it/s]
 30%|███       | 15/50 [00:09<00:16,  2.09it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.19it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.86it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.69it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.64it/s]
 48%|████▊     | 24/50 [00:11<00:06,  3.74it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.62it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.60it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.66it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.59it/s]
 64%|██████▍   | 32/50 [00:14<00:06,  2.62it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.71it/s]
 70%|███████   | 35/50 [00:15<00:06,  2.19it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.27it/s]
 76%|███████▌  | 38/50 [00:17<00:06,  1.97it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.05it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.87it/s]
 84%|████████▍ | 42/50 [00:19<00:04,  2.00it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.52it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.63it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.30it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.48it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.21it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.44it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.20it/s]
+
+  clip 4/4: 46.7s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000010/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000010/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 11708.5ms  total=  11.7s
+  clip_total         n=  4  mean= 50481.8ms  total= 201.9s
 
+Done in 213.9s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000011.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a modern kitchen with a cooking area, with clean white walls, light yellow oak hardwood floor, and a row of white quartz countertops along the back wall under bright daylight. In the foreground, there is a white three-drawer dresser with slim black bar handles, and a stainless steel kettle and a small vase of fresh herbs sit on a side counter. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=279  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:56,  1.15s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.36it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.34it/s]
 10%|█         | 5/50 [00:04<00:40,  1.12it/s]
 12%|█▏        | 6/50 [00:04<00:33,  1.33it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.72it/s]
 28%|██▊       | 14/50 [00:08<00:10,  3.30it/s]
 30%|███       | 15/50 [00:09<00:16,  2.17it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.29it/s]
 38%|███▊      | 19/50 [00:09<00:07,  4.05it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.86it/s]
 44%|████▍     | 22/50 [00:11<00:09,  2.87it/s]
 48%|████▊     | 24/50 [00:11<00:06,  4.07it/s]
 52%|█████▏    | 26/50 [00:12<00:08,  2.84it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.86it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.65it/s]
 64%|██████▍   | 32/50 [00:14<00:06,  2.74it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.69it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.42it/s]
 74%|███████▍  | 37/50 [00:15<00:04,  2.79it/s]
 76%|███████▌  | 38/50 [00:17<00:06,  1.95it/s]
 78%|███████▊  | 39/50 [00:17<00:05,  2.11it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  1.95it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.10it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.61it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.80it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.43it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.66it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.35it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.59it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.31it/s]
+
+  clip 1/4: 51.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.35it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.33it/s]
 10%|█         | 5/50 [00:04<00:41,  1.08it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.36it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.13it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 20%|██        | 10/50 [00:07<00:26,  1.49it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.67it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.70it/s]
 30%|███       | 15/50 [00:09<00:16,  2.15it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.25it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.37it/s]
 40%|████      | 20/50 [00:10<00:06,  4.64it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.50it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.38it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.50it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.52it/s]
 56%|█████▌    | 28/50 [00:13<00:07,  2.91it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.45it/s]
 62%|██████▏   | 31/50 [00:14<00:08,  2.29it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.27it/s]
 66%|██████▌   | 33/50 [00:15<00:06,  2.78it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.10it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.22it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.94it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.04it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.86it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  2.00it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.51it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.70it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.56it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.22it/s]
+
+  clip 2/4: 53.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.40it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 18%|█▊        | 9/50 [00:06<00:21,  1.89it/s]
 20%|██        | 10/50 [00:07<00:29,  1.36it/s]
 22%|██▏       | 11/50 [00:07<00:24,  1.61it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.80it/s]
 30%|███       | 15/50 [00:09<00:16,  2.11it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.22it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.37it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.92it/s]
 42%|████▏     | 21/50 [00:11<00:11,  2.43it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.41it/s]
 46%|████▌     | 23/50 [00:11<00:09,  2.86it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.33it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.77it/s]
 52%|█████▏    | 26/50 [00:13<00:12,  1.85it/s]
 54%|█████▍    | 27/50 [00:13<00:11,  1.93it/s]
 56%|█████▌    | 28/50 [00:14<00:09,  2.38it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.92it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.37it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.79it/s]
 64%|██████▍   | 32/50 [00:15<00:09,  1.93it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.41it/s]
 68%|██████▊   | 34/50 [00:16<00:05,  3.01it/s]
 70%|███████   | 35/50 [00:17<00:08,  1.68it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.90it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.72it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.89it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.77it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.93it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.47it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.67it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.54it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.44it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.18it/s]
+
+  clip 3/4: 53.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.40it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.71it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.75it/s]
 30%|███       | 15/50 [00:09<00:16,  2.15it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.24it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.89it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.66it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.58it/s]
 46%|████▌     | 23/50 [00:11<00:08,  3.04it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.43it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.40it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.85it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.45it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.41it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.45it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.12it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.21it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.97it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.08it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.58it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.71it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.51it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.23it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.46it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.21it/s]
+
+  clip 4/4: 44.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000011/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000011/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 12006.4ms  total=  12.0s
+  clip_total         n=  4  mean= 50740.5ms  total= 203.0s
 
+Done in 215.3s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000012.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a farmhouse kitchen with a baking area, with warm white shiplap walls, light yellow weathered pine plank floor, and an open shelf of ceramic crockery beside a window letting in soft morning sun. In the foreground, there is a warm cream-painted three-drawer dresser with small brushed-brass knobs, and a wooden bowl of lemons and a folded gingham towel rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=277  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.11s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.52it/s]
  6%|▌         | 3/50 [00:02<00:39,  1.18it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.55it/s]
 10%|█         | 5/50 [00:03<00:36,  1.25it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.55it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:06<00:24,  1.64it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.85it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.99it/s]
 30%|███       | 15/50 [00:08<00:14,  2.38it/s]
 32%|███▏      | 16/50 [00:08<00:13,  2.48it/s]
 36%|███▌      | 18/50 [00:08<00:08,  3.68it/s]
 42%|████▏     | 21/50 [00:10<00:09,  3.07it/s]
 44%|████▍     | 22/50 [00:10<00:09,  3.05it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.80it/s]
 54%|█████▍    | 27/50 [00:11<00:08,  2.82it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.21it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.75it/s]
 68%|██████▊   | 34/50 [00:13<00:04,  3.60it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.48it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.17it/s]
 78%|███████▊  | 39/50 [00:16<00:04,  2.25it/s]
 80%|████████  | 40/50 [00:16<00:03,  2.65it/s]
 82%|████████▏ | 41/50 [00:17<00:04,  1.86it/s]
 84%|████████▍ | 42/50 [00:18<00:04,  1.98it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.52it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.72it/s]
 90%|█████████ | 45/50 [00:20<00:03,  1.37it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.58it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.30it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.51it/s]
 98%|█████████▊| 49/50 [00:23<00:00,  1.27it/s]
+
+  clip 1/4: 43.1s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.39it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.69it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.70it/s]
 30%|███       | 15/50 [00:09<00:16,  2.17it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.18it/s]
 34%|███▍      | 17/50 [00:09<00:12,  2.59it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.83it/s]
 42%|████▏     | 21/50 [00:11<00:11,  2.53it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.42it/s]
 46%|████▌     | 23/50 [00:11<00:09,  2.80it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.08it/s]
 50%|█████     | 25/50 [00:12<00:07,  3.25it/s]
 52%|█████▏    | 26/50 [00:13<00:13,  1.79it/s]
 54%|█████▍    | 27/50 [00:14<00:13,  1.75it/s]
 56%|█████▌    | 28/50 [00:14<00:09,  2.23it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.86it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.46it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.78it/s]
 64%|██████▍   | 32/50 [00:16<00:09,  1.88it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.38it/s]
 68%|██████▊   | 34/50 [00:16<00:05,  2.98it/s]
 70%|███████   | 35/50 [00:17<00:08,  1.67it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.80it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.72it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.89it/s]
 82%|████████▏ | 41/50 [00:21<00:05,  1.77it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.92it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.47it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.67it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.53it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.44it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.19it/s]
+
+  clip 2/4: 48.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.39it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.40it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 20%|██        | 10/50 [00:07<00:27,  1.45it/s]
 22%|██▏       | 11/50 [00:08<00:26,  1.49it/s]
 24%|██▍       | 12/50 [00:08<00:21,  1.81it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.84it/s]
 30%|███       | 15/50 [00:09<00:19,  1.82it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.79it/s]
 34%|███▍      | 17/50 [00:10<00:14,  2.25it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.83it/s]
 40%|████      | 20/50 [00:10<00:07,  4.07it/s]
 42%|████▏     | 21/50 [00:12<00:13,  2.13it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.17it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.70it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.37it/s]
 52%|█████▏    | 26/50 [00:14<00:10,  2.20it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.24it/s]
 56%|█████▌    | 28/50 [00:14<00:07,  2.77it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.41it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.25it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.33it/s]
 66%|██████▌   | 33/50 [00:16<00:05,  2.86it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.51it/s]
 70%|███████   | 35/50 [00:17<00:08,  1.85it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.90it/s]
 74%|███████▍  | 37/50 [00:18<00:05,  2.40it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.52it/s]
 78%|��██████▊  | 39/50 [00:20<00:06,  1.70it/s]
 80%|████████  | 40/50 [00:20<00:04,  2.23it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.48it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.65it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.28it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.52it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.23it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.47it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.20it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.45it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.20it/s]
+
+  clip 3/4: 47.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.41it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.40it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.34it/s]
 14%|█▍        | 7/50 [00:05<00:38,  1.12it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.36it/s]
 18%|█▊        | 9/50 [00:06<00:22,  1.84it/s]
 20%|██        | 10/50 [00:07<00:30,  1.33it/s]
 22%|██▏       | 11/50 [00:08<00:26,  1.48it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.50it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.95it/s]
 30%|███       | 15/50 [00:09<00:19,  1.76it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.84it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.37it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.75it/s]
 42%|████▏     | 21/50 [00:11<00:11,  2.49it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.47it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.68it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.54it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.50it/s]
 58%|█████▊    | 29/50 [00:14<00:05,  3.60it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.09it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.16it/s]
 64%|██████▍   | 32/50 [00:15<00:08,  2.18it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.40it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.02it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.99it/s]
 74%|███████▍  | 37/50 [00:17<00:05,  2.47it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.61it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.82it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.72it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.89it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.43it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.63it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.29it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.51it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.23it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.45it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.20it/s]
+
+  clip 4/4: 47.6s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000012/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000012/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13802.3ms  total=  13.8s
+  clip_total         n=  4  mean= 46557.3ms  total= 186.2s
 
+Done in 200.3s

3drawers/gen_flex/chunk_001/_logs/b1_ep00000013.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a narrow galley kitchen with a prep area, with bright white walls and a white subway-tile backsplash, light gray vinyl plank floor, and slim under-cabinet lights glowing along both runs of cabinetry. In the foreground, there is a light oak three-drawer dresser with natural wood grain and recessed finger pulls, and a chrome utensil holder and a small potted basil plant sit nearby. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=261  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:56,  1.15s/it]
  4%|▍         | 2/50 [00:01<00:32,  1.45it/s]
  6%|▌         | 3/50 [00:02<00:41,  1.13it/s]
  8%|▊         | 4/50 [00:03<00:31,  1.47it/s]
 10%|█         | 5/50 [00:04<00:37,  1.20it/s]
 12%|█▏        | 6/50 [00:04<00:29,  1.50it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.24it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.51it/s]
 20%|██        | 10/50 [00:07<00:24,  1.61it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.82it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.47it/s]
 30%|███       | 15/50 [00:08<00:15,  2.23it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.21it/s]
 34%|███▍      | 17/50 [00:09<00:12,  2.66it/s]
 36%|███▌      | 18/50 [00:09<00:10,  3.19it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.82it/s]
 40%|████      | 20/50 [00:09<00:07,  4.15it/s]
 42%|████▏     | 21/50 [00:10<00:14,  1.99it/s]
 44%|████▍     | 22/50 [00:11<00:14,  1.93it/s]
 46%|████▌     | 23/50 [00:11<00:12,  2.09it/s]
 48%|████▊     | 24/50 [00:11<00:10,  2.55it/s]
 50%|█████     | 25/50 [00:12<00:08,  2.86it/s]
 52%|█████▏    | 26/50 [00:13<00:14,  1.71it/s]
 54%|█████▍    | 27/50 [00:13<00:12,  1.84it/s]
 56%|█████▌    | 28/50 [00:13<00:09,  2.33it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.76it/s]
 60%|██████    | 30/50 [00:14<00:06,  3.11it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.75it/s]
 64%|██████▍   | 32/50 [00:15<00:09,  1.91it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.42it/s]
 68%|██████▊   | 34/50 [00:16<00:05,  2.72it/s]
 70%|███████   | 35/50 [00:17<00:09,  1.56it/s]
 72%|███████▏  | 36/50 [00:18<00:08,  1.73it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.69it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.76it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.23it/s]
 82%|████████▏ | 41/50 [00:21<00:05,  1.59it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.82it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.41it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.65it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.59it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.30it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.56it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.28it/s]
+
+  clip 1/4: 43.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.41it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.38it/s]
 18%|█▊        | 9/50 [00:06<00:21,  1.88it/s]
 20%|██        | 10/50 [00:07<00:29,  1.36it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.56it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.02it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.59it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.24it/s]
 30%|███       | 15/50 [00:09<00:20,  1.71it/s]
 32%|███▏      | 16/50 [00:10<00:19,  1.77it/s]
 34%|███▍      | 17/50 [00:10<00:14,  2.32it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.54it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.34it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.24it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.63it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.19it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.61it/s]
 52%|█████▏    | 26/50 [00:14<00:13,  1.84it/s]
 54%|█████▍    | 27/50 [00:14<00:13,  1.74it/s]
 56%|█████▌    | 28/50 [00:14<00:10,  2.01it/s]
 58%|█████▊    | 29/50 [00:15<00:09,  2.30it/s]
 60%|██████    | 30/50 [00:15<00:07,  2.56it/s]
 62%|██████▏   | 31/50 [00:16<00:12,  1.50it/s]
 64%|██████▍   | 32/50 [00:17<00:10,  1.65it/s]
 66%|██████▌   | 33/50 [00:17<00:07,  2.16it/s]
 70%|███████   | 35/50 [00:18<00:08,  1.84it/s]
 72%|███████▏  | 36/50 [00:19<00:07,  1.91it/s]
 76%|███████▌  | 38/50 [00:20<00:06,  1.74it/s]
 78%|███████▊  | 39/50 [00:20<00:05,  1.84it/s]
 82%|████████▏ | 41/50 [00:22<00:05,  1.68it/s]
 84%|████████▍ | 42/50 [00:22<00:04,  1.79it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.40it/s]
 88%|████████▊ | 44/50 [00:24<00:03,  1.54it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.25it/s]
 92%|█████████▏| 46/50 [00:26<00:02,  1.38it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.16it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.37it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.16it/s]
+
+  clip 2/4: 46.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.20s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.33it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.04it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.37it/s]
 10%|█         | 5/50 [00:04<00:40,  1.11it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.39it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.40it/s]
 20%|██        | 10/50 [00:07<00:27,  1.46it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.55it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.96it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.49it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.92it/s]
 30%|███       | 15/50 [00:09<00:20,  1.68it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.79it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.90it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.38it/s]
 40%|████      | 20/50 [00:10<00:07,  3.99it/s]
 42%|████▏     | 21/50 [00:12<00:15,  1.93it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.02it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.57it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.16it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.94it/s]
 52%|█████▏    | 26/50 [00:14<00:12,  1.85it/s]
 54%|█████▍    | 27/50 [00:14<00:12,  1.89it/s]
 56%|█████▌    | 28/50 [00:14<00:10,  2.16it/s]
 58%|█████▊    | 29/50 [00:15<00:08,  2.49it/s]
 60%|██████    | 30/50 [00:15<00:07,  2.81it/s]
 62%|██████▏   | 31/50 [00:16<00:12,  1.58it/s]
 64%|██████▍   | 32/50 [00:17<00:10,  1.69it/s]
 66%|██████▌   | 33/50 [00:17<00:08,  2.10it/s]
 68%|██████▊   | 34/50 [00:17<00:06,  2.48it/s]
 70%|███████   | 35/50 [00:18<00:09,  1.54it/s]
 72%|███████▏  | 36/50 [00:19<00:08,  1.67it/s]
 74%|███████▍  | 37/50 [00:19<00:05,  2.19it/s]
 76%|███████▌  | 38/50 [00:20<00:08,  1.46it/s]
 78%|███████▊  | 39/50 [00:21<00:06,  1.61it/s]
 80%|████████  | 40/50 [00:21<00:04,  2.02it/s]
 82%|████████▏ | 41/50 [00:22<00:06,  1.36it/s]
 84%|████████▍ | 42/50 [00:23<00:05,  1.52it/s]
 86%|████████▌ | 43/50 [00:24<00:05,  1.21it/s]
 88%|████████▊ | 44/50 [00:24<00:04,  1.40it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.16it/s]
 92%|█████████▏| 46/50 [00:26<00:02,  1.35it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.13it/s]
 96%|█████████▌| 48/50 [00:28<00:01,  1.35it/s]
 98%|█████████▊| 49/50 [00:29<00:00,  1.14it/s]
+
+  clip 3/4: 47.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.70it/s]
 28%|██▊       | 14/50 [00:07<00:11,  3.25it/s]
 30%|███       | 15/50 [00:09<00:16,  2.07it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.17it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.85it/s]
 42%|████▏     | 21/50 [00:10<00:11,  2.63it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.57it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.46it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.69it/s]
 52%|█████▏    | 26/50 [00:13<00:11,  2.01it/s]
 54%|█████▍    | 27/50 [00:13<00:11,  2.03it/s]
 56%|█████▌    | 28/50 [00:13<00:09,  2.36it/s]
 58%|█████▊    | 29/50 [00:13<00:07,  2.77it/s]
 60%|██████    | 30/50 [00:14<00:06,  3.28it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.77it/s]
 64%|██████▍   | 32/50 [00:15<00:09,  1.89it/s]
 68%|██████▊   | 34/50 [00:16<00:05,  2.97it/s]
 70%|███████   | 35/50 [00:17<00:08,  1.82it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.82it/s]
 74%|███████▍  | 37/50 [00:18<00:06,  2.13it/s]
 76%|███████▌  | 38/50 [00:19<00:08,  1.39it/s]
 78%|███████▊  | 39/50 [00:19<00:07,  1.53it/s]
 80%|████████  | 40/50 [00:20<00:05,  1.91it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.33it/s]
 84%|████████▍ | 42/50 [00:21<00:05,  1.49it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.19it/s]
 88%|████████��� | 44/50 [00:23<00:04,  1.34it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.12it/s]
 92%|█████████▏| 46/50 [00:25<00:03,  1.32it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.12it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.36it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.15it/s]
+
+  clip 4/4: 43.9s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000013/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b1_ep00000013/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 11064.7ms  total=  11.1s
+  clip_total         n=  4  mean= 45293.5ms  total= 181.2s
 
+Done in 192.4s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000000.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, an industrial loft kitchen with a dining area, with white-painted brick walls, polished gray concrete floor, and a black-framed factory window flooding the room with cool afternoon light. In the foreground, there is a natural pine three-drawer dresser with round wooden knobs, and a cast-iron pan and a stack of enamel mugs are arranged on the counter. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=262  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:55,  1.14s/it]
  4%|▍         | 2/50 [00:01<00:32,  1.47it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.15it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.52it/s]
 10%|█         | 5/50 [00:04<00:36,  1.23it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.53it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.25it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.53it/s]
 20%|██        | 10/50 [00:06<00:24,  1.63it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.84it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.94it/s]
 30%|███       | 15/50 [00:08<00:14,  2.35it/s]
 32%|███▏      | 16/50 [00:08<00:13,  2.45it/s]
 38%|███▊      | 19/50 [00:09<00:07,  4.22it/s]
 42%|████▏     | 21/50 [00:10<00:09,  2.94it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.95it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.74it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.79it/s]
 58%|█████▊    | 29/50 [00:12<00:05,  3.72it/s]
 62%|██████▏   | 31/50 [00:13<00:06,  2.80it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.77it/s]
 66%|██████▌   | 33/50 [00:13<00:05,  3.23it/s]
 70%|███████   | 35/50 [00:14<00:06,  2.43it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.42it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.09it/s]
 78%|███████▊  | 39/50 [00:17<00:05,  2.09it/s]
 80%|████████  | 40/50 [00:17<00:04,  2.43it/s]
 82%|████████▏ | 41/50 [00:18<00:05,  1.69it/s]
 84%|████████▍ | 42/50 [00:18<00:04,  1.73it/s]
 86%|████████▌ | 43/50 [00:20<00:05,  1.37it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.50it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.25it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.50it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.52it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.28it/s]
+
+  clip 1/4: 44.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.24s/it]
  4%|▍         | 2/50 [00:01<00:39,  1.22it/s]
  6%|▌         | 3/50 [00:02<00:47,  1.00s/it]
  8%|▊         | 4/50 [00:03<00:35,  1.30it/s]
 10%|█         | 5/50 [00:04<00:41,  1.09it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.35it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.13it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 20%|██        | 10/50 [00:07<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.65it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.67it/s]
 30%|███       | 15/50 [00:09<00:16,  2.16it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.26it/s]
 38%|███▊      | 19/50 [00:10<00:07,  3.91it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.72it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.70it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.62it/s]
 50%|█████     | 25/50 [00:11<00:06,  4.04it/s]
 52%|█████▏    | 26/50 [00:13<00:11,  2.15it/s]
 54%|█████▍    | 27/50 [00:13<00:10,  2.16it/s]
 56%|█████▌    | 28/50 [00:13<00:08,  2.65it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.90it/s]
 62%|██████▏   | 31/50 [00:15<00:09,  2.07it/s]
 64%|██████▍   | 32/50 [00:15<00:08,  2.09it/s]
 66%|██████▌   | 33/50 [00:15<00:06,  2.60it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.02it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.10it/s]
 74%|███████▍  | 37/50 [00:17<00:05,  2.58it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.67it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.84it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.74it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.85it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.41it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.55it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.24it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.44it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.19it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.41it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.19it/s]
+
+  clip 2/4: 49.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.33it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.05it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.33it/s]
 10%|█         | 5/50 [00:04<00:41,  1.09it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.37it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.75it/s]
 30%|███       | 15/50 [00:09<00:16,  2.13it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.20it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.22it/s]
 40%|████      | 20/50 [00:10<00:06,  4.50it/s]
 44%|████▍     | 22/50 [00:11<00:12,  2.28it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.48it/s]
 48%|████▊     | 24/50 [00:12<00:09,  2.79it/s]
 50%|█████     | 25/50 [00:12<00:08,  3.02it/s]
 52%|█████▏    | 26/50 [00:13<00:13,  1.77it/s]
 54%|█████▍    | 27/50 [00:14<00:12,  1.86it/s]
 56%|█████▌    | 28/50 [00:14<00:09,  2.31it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.82it/s]
 60%|██████    | 30/50 [00:14<00:06,  3.19it/s]
 62%|██████▏   | 31/50 [00:15<00:11,  1.71it/s]
 64%|██████▍   | 32/50 [00:16<00:10,  1.77it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.16it/s]
 68%|██████▊   | 34/50 [00:16<00:05,  2.71it/s]
 70%|███████   | 35/50 [00:18<00:09,  1.64it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.79it/s]
 74%|███████▍  | 37/50 [00:18<00:05,  2.34it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.52it/s]
 78%|███████▊  | 39/50 [00:20<00:06,  1.71it/s]
 80%|████████  | 40/50 [00:20<00:04,  2.26it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.47it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.71it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.31it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.55it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.24it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.49it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.21it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.46it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.20it/s]
+
+  clip 3/4: 52.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.34it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.05it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.39it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.38it/s]
 20%|██        | 10/50 [00:07<00:27,  1.44it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.57it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.00it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.11it/s]
 30%|███       | 15/50 [00:09<00:18,  1.91it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.08it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.31it/s]
 40%|████      | 20/50 [00:10<00:06,  4.81it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.46it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.47it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.55it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.59it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.72it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.63it/s]
 64%|██████▍   | 32/50 [00:15<00:06,  2.65it/s]
 70%|███████   | 35/50 [00:16<00:05,  2.51it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.55it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.15it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.22it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.97it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.07it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.57it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.75it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.36it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.56it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.46it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.21it/s]
+
+  clip 4/4: 45.6s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000000/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000000/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 12147.6ms  total=  12.1s
+  clip_total         n=  4  mean= 48007.5ms  total= 192.0s
 
+Done in 204.4s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000001.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a home office with a desk area, with warm gray walls, light yellow maple hardwood floor, and sheer curtains diffusing soft afternoon sun across the room. In the foreground, there is a honey-toned maple three-drawer dresser with brushed-nickel cup pulls, and a closed silver laptop and a small potted succulent are placed on the desk surface. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=281  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.12s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.39it/s]
  6%|▌         | 3/50 [00:02<00:42,  1.10it/s]
  8%|▊         | 4/50 [00:03<00:31,  1.44it/s]
 10%|█         | 5/50 [00:04<00:37,  1.20it/s]
 12%|█▏        | 6/50 [00:04<00:29,  1.48it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.23it/s]
 16%|█▌        | 8/50 [00:05<00:28,  1.50it/s]
 20%|██        | 10/50 [00:07<00:24,  1.61it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.81it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.91it/s]
 30%|███       | 15/50 [00:08<00:15,  2.31it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.35it/s]
 38%|███▊      | 19/50 [00:09<00:07,  4.07it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.89it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.88it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.63it/s]
 54%|█████▍    | 27/50 [00:12<00:09,  2.52it/s]
 56%|█████▌    | 28/50 [00:12<00:07,  2.75it/s]
 60%|██████    | 30/50 [00:12<00:05,  3.62it/s]
 62%|██████▏   | 31/50 [00:14<00:08,  2.30it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.30it/s]
 66%|██████▌   | 33/50 [00:14<00:06,  2.72it/s]
 68%|██████▊   | 34/50 [00:14<00:05,  3.00it/s]
 70%|███████   | 35/50 [00:16<00:08,  1.77it/s]
 72%|███████▏  | 36/50 [00:16<00:07,  1.92it/s]
 74%|███████▍  | 37/50 [00:16<00:05,  2.37it/s]
 76%|███████▌  | 38/50 [00:17<00:07,  1.58it/s]
 78%|███████▊  | 39/50 [00:18<00:06,  1.71it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.13it/s]
 82%|████████▏ | 41/50 [00:19<00:06,  1.50it/s]
 84%|████████▍ | 42/50 [00:19<00:04,  1.73it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.36it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.61it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.57it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.30it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.56it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.29it/s]
+
+  clip 1/4: 48.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.37it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.76it/s]
 30%|███       | 15/50 [00:09<00:15,  2.19it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.28it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.37it/s]
 40%|████      | 20/50 [00:09<00:06,  4.53it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.41it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.35it/s]
 46%|████▌     | 23/50 [00:11<00:09,  2.84it/s]
 50%|█████     | 25/50 [00:11<00:06,  4.08it/s]
 52%|█████▏    | 26/50 [00:12<00:11,  2.18it/s]
 54%|█████▍    | 27/50 [00:13<00:10,  2.22it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.15it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.55it/s]
 62%|██████▏   | 31/50 [00:15<00:09,  1.94it/s]
 64%|██████▍   | 32/50 [00:15<00:08,  2.03it/s]
 66%|██████▌   | 33/50 [00:15<00:06,  2.51it/s]
 70%|███████   | 35/50 [00:16<00:07,  1.97it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.04it/s]
 74%|███████▍  | 37/50 [00:17<00:05,  2.51it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.59it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.68it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.07it/s]
 82%|████████▏ | 41/50 [00:20<00:06,  1.46it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.63it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.27it/s]
 88%|████████▊ | 44/50 [00:22<00:04,  1.49it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.21it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.45it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.18it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.41it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.18it/s]
+
+  clip 2/4: 53.5s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.41it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:24,  1.61it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.05it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.04it/s]
 30%|███       | 15/50 [00:09<00:19,  1.82it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.87it/s]
 34%|███▍      | 17/50 [00:10<00:15,  2.17it/s]
 36%|███▌      | 18/50 [00:10<00:12,  2.64it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.24it/s]
 42%|████▏     | 21/50 [00:11<00:13,  2.19it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.18it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.24it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.35it/s]
 54%|█████▍    | 27/50 [00:14<00:09,  2.42it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.47it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.45it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.39it/s]
 66%|██████▌   | 33/50 [00:16<00:06,  2.81it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.16it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.24it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.95it/s]
 78%|███��███▊  | 39/50 [00:19<00:05,  2.08it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.88it/s]
 84%|████████▍ | 42/50 [00:21<00:03,  2.01it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.52it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.71it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.35it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.56it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.22it/s]
+
+  clip 3/4: 49.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.13it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 18%|█▊        | 9/50 [00:06<00:22,  1.85it/s]
 20%|██        | 10/50 [00:07<00:30,  1.33it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.58it/s]
 28%|██▊       | 14/50 [00:08<00:10,  3.32it/s]
 30%|███       | 15/50 [00:09<00:17,  2.05it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.18it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.96it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.72it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.72it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.45it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.58it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.91it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.53it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.46it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.50it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.13it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.21it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.97it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.08it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.58it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.76it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.38it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.59it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.28it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.50it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.23it/s]
+
+  clip 4/4: 41.7s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000001/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000001/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13707.4ms  total=  13.7s
+  clip_total         n=  4  mean= 48253.5ms  total= 193.0s
 
+Done in 206.9s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000002.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a corporate office with a workstation area, with crisp white walls and a glass partition, medium gray carpet-tile floor, and even cool light from recessed ceiling panels. In the foreground, there is a dark walnut three-drawer dresser with matte black bar pulls, and a black desk monitor and a neat tray of pens sit on the surface. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=264  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:55,  1.13s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.50it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.17it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.53it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.54it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:28,  1.46it/s]
 18%|█▊        | 9/50 [00:05<00:20,  1.96it/s]
 20%|██        | 10/50 [00:07<00:27,  1.44it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.64it/s]
 24%|██▍       | 12/50 [00:07<00:17,  2.15it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.72it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.31it/s]
 30%|███       | 15/50 [00:09<00:18,  1.86it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.09it/s]
 36%|███▌      | 18/50 [00:09<00:08,  3.56it/s]
 42%|████▏     | 21/50 [00:10<00:09,  2.97it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.95it/s]
 48%|████▊     | 24/50 [00:11<00:06,  4.22it/s]
 52%|█████▏    | 26/50 [00:12<00:08,  2.88it/s]
 54%|█████▍    | 27/50 [00:12<00:07,  2.88it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.72it/s]
 64%|██████▍   | 32/50 [00:14<00:06,  2.75it/s]
 70%|███████   | 35/50 [00:15<00:05,  2.65it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.69it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.28it/s]
 78%|███████▊  | 39/50 [00:17<00:04,  2.37it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  2.05it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.18it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.67it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.86it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.47it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.70it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.38it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.62it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.32it/s]
+
+  clip 1/4: 38.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.35it/s]
 10%|█         | 5/50 [00:04<00:40,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:34,  1.29it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.09it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.28it/s]
 18%|█▊        | 9/50 [00:06<00:23,  1.73it/s]
 20%|██        | 10/50 [00:07<00:30,  1.31it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.53it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.51it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.84it/s]
 30%|███       | 15/50 [00:10<00:20,  1.68it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.86it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.66it/s]
 42%|████▏     | 21/50 [00:11<00:11,  2.57it/s]
 44%|████▍     | 22/50 [00:12<00:10,  2.61it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.68it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.53it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.55it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.24it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.48it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.49it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.60it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.13it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.22it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.94it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.04it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.84it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.96it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.50it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.69it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.55it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.22it/s]
+
+  clip 2/4: 42.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.20s/it]
  4%|▍         | 2/50 [00:01<00:38,  1.25it/s]
  6%|▌         | 3/50 [00:02<00:46,  1.00it/s]
  8%|▊         | 4/50 [00:03<00:36,  1.24it/s]
 10%|█         | 5/50 [00:04<00:43,  1.04it/s]
 12%|█▏        | 6/50 [00:05<00:35,  1.22it/s]
 14%|█▍        | 7/50 [00:06<00:41,  1.04it/s]
 16%|█▌        | 8/50 [00:07<00:34,  1.22it/s]
 18%|█▊        | 9/50 [00:07<00:26,  1.58it/s]
 20%|██        | 10/50 [00:08<00:33,  1.20it/s]
 22%|██▏       | 11/50 [00:08<00:27,  1.41it/s]
 26%|██▌       | 13/50 [00:09<00:15,  2.44it/s]
 30%|███       | 15/50 [00:10<00:17,  2.02it/s]
 32%|███▏      | 16/50 [00:10<00:15,  2.13it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.80it/s]
 42%|████▏     | 21/50 [00:12<00:10,  2.67it/s]
 44%|████▍     | 22/50 [00:12<00:10,  2.68it/s]
 48%|████▊     | 24/50 [00:12<00:06,  3.82it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.64it/s]
 54%|█████▍    | 27/50 [00:14<00:08,  2.64it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.31it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.48it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.52it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.52it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.12it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.23it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.92it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.02it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.48it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.67it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.79it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.36it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.53it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.23it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.42it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.18it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.41it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.18it/s]
+
+  clip 3/4: 47.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:38,  1.26it/s]
  6%|▌         | 3/50 [00:02<00:46,  1.01it/s]
  8%|▊         | 4/50 [00:03<00:36,  1.27it/s]
 10%|█         | 5/50 [00:04<00:42,  1.07it/s]
 12%|█▏        | 6/50 [00:05<00:33,  1.30it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.09it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.32it/s]
 18%|█▊        | 9/50 [00:06<00:23,  1.72it/s]
 20%|██        | 10/50 [00:08<00:31,  1.26it/s]
 22%|██▏       | 11/50 [00:08<00:27,  1.44it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.93it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.24it/s]
 30%|███       | 15/50 [00:10<00:18,  1.89it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.93it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.40it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.90it/s]
 40%|████      | 20/50 [00:11<00:07,  4.17it/s]
 42%|████▏     | 21/50 [00:12<00:13,  2.08it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.21it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.17it/s]
 52%|█████▏    | 26/50 [00:14<00:10,  2.32it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.28it/s]
 56%|█████▌    | 28/50 [00:14<00:07,  2.80it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.23it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.23it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.33it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.59it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.06it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.18it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.89it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.03it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.84it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.98it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.50it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.69it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.54it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.21it/s]
+
+  clip 4/4: 45.1s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000002/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000002/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13203.1ms  total=  13.2s
+  clip_total         n=  4  mean= 43353.8ms  total= 173.4s
 
+Done in 186.7s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000003.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, an open-plan coworking space with a lounge area, with bright white walls and pale wood acoustic panels, light yellow birch plank floor, and a long row of pendant lights over a shared table. In the foreground, there is an espresso-brown three-drawer dresser with antique bronze handles, and a ceramic mug and a small stack of notebooks rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=276  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.11s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.52it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.17it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.50it/s]
 10%|█         | 5/50 [00:04<00:36,  1.22it/s]
 12%|█▏        | 6/50 [00:04<00:29,  1.49it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.23it/s]
 16%|█▌        | 8/50 [00:05<00:28,  1.47it/s]
 20%|██        | 10/50 [00:07<00:25,  1.57it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.76it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.85it/s]
 30%|███       | 15/50 [00:08<00:15,  2.32it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.42it/s]
 38%|███▊      | 19/50 [00:09<00:07,  4.16it/s]
 42%|████▏     | 21/50 [00:10<00:09,  2.91it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.90it/s]
 48%|████▊     | 24/50 [00:10<00:06,  4.00it/s]
 52%|█████▏    | 26/50 [00:11<00:08,  2.81it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.83it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.61it/s]
 62%|██████▏   | 31/50 [00:13<00:07,  2.69it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.73it/s]
 70%|███████   | 35/50 [00:15<00:05,  2.64it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.68it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.27it/s]
 78%|███████▊  | 39/50 [00:16<00:04,  2.37it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  2.10it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.19it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.66it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.76it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.41it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.62it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.33it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.56it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.28it/s]
+
+  clip 1/4: 42.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 20%|██        | 10/50 [00:07<00:26,  1.49it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.67it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.23it/s]
 30%|███       | 15/50 [00:09<00:16,  2.06it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.18it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.30it/s]
 40%|████      | 20/50 [00:09<00:06,  4.70it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.46it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.90it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.47it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.30it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.52it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.51it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.49it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.18it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.24it/s]
 76%|███████▌  | 38/50 [00:17<00:06,  1.91it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  1.97it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.28it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.58it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.70it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.31it/s]
 88%|████████▊ | 44/50 [00:21<00:04,  1.48it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.20it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.36it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.12it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.31it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.11it/s]
+
+  clip 2/4: 46.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.35it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.36it/s]
 10%|█         | 5/50 [00:04<00:40,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.37it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.12it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.38it/s]
 20%|██        | 10/50 [00:07<00:27,  1.45it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.61it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.98it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.17it/s]
 30%|███       | 15/50 [00:09<00:18,  1.93it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.08it/s]
 38%|███▊      | 19/50 [00:10<00:07,  3.90it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.66it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.63it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.33it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.54it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.86it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.49it/s]
 68%|██████▊   | 34/50 [00:15<00:05,  3.19it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.12it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.08it/s]
 74%|███████▍  | 37/50 [00:17<00:05,  2.49it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.67it/s]
 78%|███████▊  | 39/50 [00:18<00:06,  1.81it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.69it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.85it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.43it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.60it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.27it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.38it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.15it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.34it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.12it/s]
+
+  clip 3/4: 47.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.20s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.34it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.04it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.37it/s]
 10%|█         | 5/50 [00:04<00:40,  1.12it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.39it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.31it/s]
 20%|██        | 10/50 [00:07<00:28,  1.41it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.58it/s]
 26%|██▌       | 13/50 [00:08<00:15,  2.47it/s]
 30%|███       | 15/50 [00:09<00:17,  2.00it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.06it/s]
 36%|███▌      | 18/50 [00:10<00:10,  2.94it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.45it/s]
 40%|████      | 20/50 [00:10<00:07,  4.06it/s]
 42%|████▏     | 21/50 [00:11<00:14,  2.02it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.06it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.53it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.16it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.21it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.29it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.18it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.35it/s]
 64%|██████▍   | 32/50 [00:15<00:08,  2.25it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.21it/s]
 70%|███████   | 35/50 [00:17<00:07,  1.98it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.02it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.80it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.95it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.80it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.93it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.47it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.65it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.51it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.42it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.17it/s]
+
+  clip 4/4: 45.7s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000003/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000003/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 12547.6ms  total=  12.5s
+  clip_total         n=  4  mean= 45591.6ms  total= 182.4s
 
+Done in 195.1s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000004.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, an office breakroom with a coffee area, with bright white walls, medium gray polished concrete floor, and overhead LED panels casting cool even light across the room. In the foreground, there is a matte black three-drawer dresser with slim leather pull tabs, and a chrome coffee machine and a neat stack of paper cups are arranged on the laminate counter. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=257  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.35it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.32it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.65it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.01it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.12it/s]
 30%|███       | 15/50 [00:09<00:17,  1.97it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.06it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.12it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.68it/s]
 40%|████      | 20/50 [00:10<00:07,  4.19it/s]
 42%|████▏     | 21/50 [00:11<00:13,  2.14it/s]
 44%|████▍     | 22/50 [00:11<00:12,  2.30it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.67it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.52it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.61it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.71it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.27it/s]
 62%|██████▏   | 31/50 [00:14<00:08,  2.27it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.37it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.64it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.15it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.30it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.04it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.19it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  2.00it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.15it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.63it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.80it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.44it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.67it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.35it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.60it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.32it/s]
+
+  clip 1/4: 44.5s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:02,  1.28s/it]
  4%|▍         | 2/50 [00:01<00:40,  1.20it/s]
  6%|▌         | 3/50 [00:02<00:47,  1.00s/it]
  8%|▊         | 4/50 [00:03<00:37,  1.22it/s]
 10%|█         | 5/50 [00:04<00:43,  1.02it/s]
 12%|█▏        | 6/50 [00:05<00:35,  1.22it/s]
 14%|█▍        | 7/50 [00:06<00:41,  1.04it/s]
 16%|█▌        | 8/50 [00:07<00:33,  1.25it/s]
 18%|█▊        | 9/50 [00:07<00:24,  1.70it/s]
 20%|██        | 10/50 [00:08<00:31,  1.26it/s]
 22%|██▏       | 11/50 [00:09<00:29,  1.34it/s]
 24%|██▍       | 12/50 [00:09<00:21,  1.78it/s]
 26%|██▌       | 13/50 [00:09<00:16,  2.29it/s]
 30%|███       | 15/50 [00:10<00:18,  1.91it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.06it/s]
 36%|███▌      | 18/50 [00:11<00:09,  3.24it/s]
 40%|████      | 20/50 [00:11<00:06,  4.65it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.46it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.95it/s]
 54%|█████▍    | 27/50 [00:14<00:09,  2.48it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.75it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.49it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.24it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.26it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.04it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.12it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.93it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.05it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.58it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.75it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.38it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.59it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.29it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.43it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.19it/s]
+
+  clip 2/4: 49.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:39,  1.23it/s]
  6%|▌         | 3/50 [00:03<00:47,  1.01s/it]
  8%|▊         | 4/50 [00:03<00:37,  1.24it/s]
 10%|█         | 5/50 [00:04<00:42,  1.05it/s]
 12%|█▏        | 6/50 [00:05<00:33,  1.32it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.12it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.35it/s]
 20%|██        | 10/50 [00:07<00:27,  1.46it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.63it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.06it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.20it/s]
 30%|███       | 15/50 [00:09<00:18,  1.90it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.03it/s]
 34%|███▍      | 17/50 [00:10<00:12,  2.57it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.23it/s]
 40%|████      | 20/50 [00:10<00:05,  5.00it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.31it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.37it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.40it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.41it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.30it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.49it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.51it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.43it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.45it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.10it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.00it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.27it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.63it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.72it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.33it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.54it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.25it/s]
 92%|██��██████▏| 46/50 [00:24<00:02,  1.41it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.17it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.36it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.14it/s]
+
+  clip 3/4: 46.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:01,  1.25s/it]
  4%|▍         | 2/50 [00:01<00:40,  1.19it/s]
  6%|▌         | 3/50 [00:03<00:49,  1.05s/it]
  8%|▊         | 4/50 [00:03<00:36,  1.25it/s]
 10%|█         | 5/50 [00:04<00:42,  1.05it/s]
 12%|█▏        | 6/50 [00:05<00:34,  1.28it/s]
 14%|█▍        | 7/50 [00:06<00:40,  1.07it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.32it/s]
 20%|██        | 10/50 [00:08<00:27,  1.46it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.65it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.65it/s]
 30%|███       | 15/50 [00:09<00:16,  2.14it/s]
 32%|███▏      | 16/50 [00:10<00:15,  2.25it/s]
 38%|███▊      | 19/50 [00:10<00:07,  3.91it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.73it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.71it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.38it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.57it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.90it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.53it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.47it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.50it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.16it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.23it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.98it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.09it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.59it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.76it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.39it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.59it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.28it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.50it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.23it/s]
+
+  clip 4/4: 43.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000004/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000004/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 14156.4ms  total=  14.2s
+  clip_total         n=  4  mean= 46061.3ms  total= 184.2s
 
+Done in 198.7s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000005.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a modern bedroom with a sleeping area, with muted dove gray walls and white baseboards, light yellow ash hardwood floor glowing warm in late-afternoon light, and a queen bed dressed in a soft linen duvet against the far wall. In the foreground, there is a charcoal-gray three-drawer dresser with long chrome handles, and a small ceramic lamp and a folded knit blanket are placed on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=285  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.11s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.44it/s]
  6%|▌         | 3/50 [00:02<00:42,  1.11it/s]
  8%|▊         | 4/50 [00:03<00:31,  1.47it/s]
 10%|█         | 5/50 [00:04<00:37,  1.21it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.45it/s]
 14%|█▍        | 7/50 [00:05<00:35,  1.22it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.44it/s]
 20%|██        | 10/50 [00:07<00:25,  1.56it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.73it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.65it/s]
 30%|███       | 15/50 [00:08<00:15,  2.22it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.27it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.33it/s]
 40%|████      | 20/50 [00:09<00:06,  4.69it/s]
 44%|████▍     | 22/50 [00:10<00:10,  2.63it/s]
 50%|█████     | 25/50 [00:11<00:06,  4.13it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.66it/s]
 58%|█████▊    | 29/50 [00:12<00:05,  3.54it/s]
 62%|██████▏   | 31/50 [00:13<00:06,  2.72it/s]
 64%|██████▍   | 32/50 [00:14<00:06,  2.73it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.77it/s]
 70%|███████   | 35/50 [00:15<00:06,  2.37it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.47it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.15it/s]
 78%|███████▊  | 39/50 [00:17<00:04,  2.28it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  2.05it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.18it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.65it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.83it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.44it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.60it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.30it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.25it/s]
+
+  clip 1/4: 46.5s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.30it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.04it/s]
  8%|▊         | 4/50 [00:03<00:36,  1.26it/s]
 10%|█         | 5/50 [00:04<00:43,  1.04it/s]
 12%|█▏        | 6/50 [00:05<00:33,  1.31it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.11it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 20%|██        | 10/50 [00:07<00:26,  1.48it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.68it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.23it/s]
 30%|███       | 15/50 [00:09<00:16,  2.07it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.19it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.88it/s]
 40%|████      | 20/50 [00:10<00:06,  4.39it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.31it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.40it/s]
 46%|████▌     | 23/50 [00:11<00:09,  2.96it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.54it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.34it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.85it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.42it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.39it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.44it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.11it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.20it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.96it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.04it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.52it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.65it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.30it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.46it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.20it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.39it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.17it/s]
+
+  clip 2/4: 47.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.20s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.37it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.37it/s]
 10%|█         | 5/50 [00:04<00:40,  1.12it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.39it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.49it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.68it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.73it/s]
 30%|███       | 15/50 [00:09<00:16,  2.17it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.26it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.39it/s]
 40%|████      | 20/50 [00:09<00:06,  4.31it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.30it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.34it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.36it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.91it/s]
 52%|█████▏    | 26/50 [00:13<00:11,  2.08it/s]
 54%|█████▍    | 27/50 [00:13<00:10,  2.17it/s]
 56%|█████▌    | 28/50 [00:13<00:08,  2.74it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.23it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.76it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.80it/s]
 64%|██████▍   | 32/50 [00:15<00:09,  1.97it/s]
 66%|██████▌   | 33/50 [00:15<00:06,  2.51it/s]
 68%|██████▊   | 34/50 [00:15<00:05,  3.15it/s]
 70%|███████   | 35/50 [00:17<00:08,  1.72it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.91it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.76it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.84it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.30it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.53it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.69it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.30it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.52it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.23it/s]
 92%|��████████▏| 46/50 [00:24<00:02,  1.47it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.20it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.44it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.19it/s]
+
+  clip 3/4: 48.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.17s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 28%|██▊       | 14/50 [00:07<00:11,  3.26it/s]
 30%|███       | 15/50 [00:09<00:16,  2.08it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.19it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.88it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.69it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.69it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.40it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.56it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.89it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.52it/s]
 70%|███████   | 35/50 [00:15<00:06,  2.45it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.49it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.15it/s]
 78%|███████▊  | 39/50 [00:17<00:04,  2.23it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.99it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.10it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.59it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.76it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.39it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.57it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.22it/s]
+
+  clip 4/4: 42.4s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000005/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000005/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 18770.4ms  total=  18.8s
+  clip_total         n=  4  mean= 46297.5ms  total= 185.2s
 
+Done in 204.1s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000006.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a children's bedroom with a play area, with soft white walls and a single pale gray accent stripe, light yellow pine plank floor, and a low bookshelf of picture books beside a round rug. In the foreground, there is a soft sage-green painted three-drawer dresser with white ceramic knobs, and a plush teddy bear and a small wooden toy box sit on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=264  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:56,  1.15s/it]
  4%|▍         | 2/50 [00:01<00:32,  1.47it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.16it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.51it/s]
 10%|█         | 5/50 [00:04<00:36,  1.23it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.53it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.52it/s]
 20%|██        | 10/50 [00:06<00:24,  1.63it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.84it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.52it/s]
 30%|███       | 15/50 [00:08<00:15,  2.25it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.37it/s]
 38%|███▊      | 19/50 [00:08<00:07,  4.19it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.89it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.89it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.68it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.73it/s]
 56%|█████▌    | 28/50 [00:12<00:07,  3.10it/s]
 58%|█████▊    | 29/50 [00:12<00:06,  3.43it/s]
 60%|██████    | 30/50 [00:12<00:05,  3.91it/s]
 62%|██████▏   | 31/50 [00:13<00:08,  2.16it/s]
 64%|██████▍   | 32/50 [00:14<00:08,  2.23it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.33it/s]
 70%|███████   | 35/50 [00:15<00:07,  2.05it/s]
 72%|███████▏  | 36/50 [00:15<00:06,  2.21it/s]
 76%|███████▌  | 38/50 [00:16<00:06,  1.99it/s]
 78%|███████▊  | 39/50 [00:17<00:05,  2.14it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  1.97it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.12it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.61it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.83it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.44it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.68it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.36it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.60it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.31it/s]
+
+  clip 1/4: 40.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.15it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.43it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.40it/s]
 18%|█▊        | 9/50 [00:06<00:21,  1.91it/s]
 20%|██        | 10/50 [00:07<00:29,  1.37it/s]
 22%|██▏       | 11/50 [00:07<00:24,  1.56it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.92it/s]
 26%|██▌       | 13/50 [00:08<00:16,  2.26it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.78it/s]
 30%|███       | 15/50 [00:09<00:21,  1.61it/s]
 32%|███▏      | 16/50 [00:10<00:19,  1.73it/s]
 34%|███▍      | 17/50 [00:10<00:15,  2.19it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.69it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.20it/s]
 40%|████      | 20/50 [00:11<00:08,  3.61it/s]
 42%|████▏     | 21/50 [00:12<00:16,  1.77it/s]
 44%|████▍     | 22/50 [00:12<00:14,  1.88it/s]
 46%|████▌     | 23/50 [00:12<00:11,  2.43it/s]
 48%|████▊     | 24/50 [00:13<00:09,  2.86it/s]
 50%|█████     | 25/50 [00:13<00:07,  3.33it/s]
 52%|█████▏    | 26/50 [00:14<00:13,  1.77it/s]
 54%|█████▍    | 27/50 [00:14<00:12,  1.84it/s]
 56%|█████▌    | 28/50 [00:15<00:09,  2.30it/s]
 58%|█████▊    | 29/50 [00:15<00:07,  2.68it/s]
 60%|██████    | 30/50 [00:15<00:06,  3.31it/s]
 62%|██████▏   | 31/50 [00:16<00:11,  1.72it/s]
 64%|██████▍   | 32/50 [00:17<00:09,  1.93it/s]
 68%|██████▊   | 34/50 [00:17<00:05,  3.17it/s]
 70%|███████   | 35/50 [00:18<00:08,  1.85it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.87it/s]
 74%|███████▍  | 37/50 [00:19<00:05,  2.36it/s]
 76%|███████▌  | 38/50 [00:20<00:07,  1.54it/s]
 78%|███████▊  | 39/50 [00:20<00:06,  1.65it/s]
 80%|████████  | 40/50 [00:20<00:04,  2.17it/s]
 82%|████████▏ | 41/50 [00:22<00:06,  1.49it/s]
 84%|████████▍ | 42/50 [00:22<00:04,  1.67it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.28it/s]
 88%|████████▊ | 44/50 [00:24<00:03,  1.50it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.21it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.45it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.18it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.41it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.18it/s]
+
+  clip 2/4: 44.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.39it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.37it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.13it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 20%|██        | 10/50 [00:07<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.64it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.03it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.52it/s]
 30%|███       | 15/50 [00:09<00:17,  2.03it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.07it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.53it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.12it/s]
 38%|███▊      | 19/50 [00:10<00:10,  3.02it/s]
 40%|████      | 20/50 [00:10<00:09,  3.14it/s]
 42%|████▏     | 21/50 [00:12<00:17,  1.70it/s]
 44%|████▍     | 22/50 [00:12<00:15,  1.85it/s]
 46%|████▌     | 23/50 [00:12<00:11,  2.41it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.88it/s]
 52%|█████▏    | 26/50 [00:13<00:11,  2.07it/s]
 54%|█████▍    | 27/50 [00:14<00:11,  2.01it/s]
 56%|█████▌    | 28/50 [00:14<00:09,  2.40it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.78it/s]
 60%|██████    | 30/50 [00:15<00:06,  3.20it/s]
 62%|██████▏   | 31/50 [00:16<00:10,  1.75it/s]
 64%|██████▍   | 32/50 [00:16<00:09,  1.83it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.37it/s]
 68%|██████▊   | 34/50 [00:17<00:05,  2.97it/s]
 70%|███████   | 35/50 [00:18<00:09,  1.58it/s]
 72%|███████▏  | 36/50 [00:18<00:08,  1.66it/s]
 74%|███████▍  | 37/50 [00:19<00:06,  2.05it/s]
 76%|███████▌  | 38/50 [00:20<00:08,  1.40it/s]
 78%|███████▊  | 39/50 [00:21<00:07,  1.48it/s]
 80%|████████  | 40/50 [00:21<00:05,  1.79it/s]
 82%|████████▏ | 41/50 [00:22<00:06,  1.30it/s]
 84%|████████▍ | 42/50 [00:23<00:05,  1.47it/s]
 86%|████████▌ | 43/50 [00:24<00:06,  1.16it/s]
 88%|████████▊ | 44/50 [00:24<00:04,  1.40it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.17it/s]
 92%|█████████▏| 46/50 [00:26<00:02,  1.35it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.14it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.38it/s]
 98%|█████████▊| 49/50 [00:29<00:00,  1.16it/s]
+
+  clip 3/4: 48.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.38it/s]
 10%|█         | 5/50 [00:04<00:40,  1.12it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.40it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.52it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.71it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.76it/s]
 30%|███       | 15/50 [00:09<00:15,  2.19it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.29it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.43it/s]
 40%|████      | 20/50 [00:09<00:06,  4.77it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.47it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.32it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.48it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.41it/s]
 56%|█████▌    | 28/50 [00:13<00:08,  2.64it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.09it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.66it/s]
 62%|██████▏   | 31/50 [00:14<00:09,  1.95it/s]
 64%|██████▍   | 32/50 [00:15<00:09,  1.96it/s]
 66%|██████▌   | 33/50 [00:15<00:07,  2.33it/s]
 68%|██████▊   | 34/50 [00:15<00:05,  2.76it/s]
 70%|███████   | 35/50 [00:17<00:09,  1.59it/s]
 72%|███████▏  | 36/50 [00:17<00:08,  1.62it/s]
 74%|███████▍  | 37/50 [00:17<00:06,  1.96it/s]
 76%|███████▌  | 38/50 [00:19<00:08,  1.36it/s]
 78%|███████▊  | 39/50 [00:19<00:07,  1.51it/s]
 80%|████████  | 40/50 [00:19<00:05,  1.93it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.33it/s]
 84%|████████▍ | 42/50 [00:21<00:05,  1.46it/s]
 86%|████████▌ | 43/50 [00:22<00:06,  1.16it/s]
 88%|████████▊ | 44/50 [00:23<00:04,  1.36it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.15it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.38it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.16it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.35it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.14it/s]
+
+  clip 4/4: 46.7s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000006/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000006/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 11786.2ms  total=  11.8s
+  clip_total         n=  4  mean= 44777.2ms  total= 179.1s
 
+Done in 191.0s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000007.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a guest bedroom with a reading area, with warm gray walls and white crown molding, white-painted wide-plank wood floor, and a small armchair under a window with a linen roman shade. In the foreground, there is a dusty-blue painted three-drawer dresser with brushed-brass bar pulls, and a folded quilt and a glass of water on a coaster rest on the surface. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=292  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:55,  1.12s/it]
  4%|▍         | 2/50 [00:01<00:32,  1.50it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.16it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.52it/s]
 10%|█         | 5/50 [00:04<00:36,  1.23it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.53it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.25it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.53it/s]
 20%|██        | 10/50 [00:06<00:24,  1.63it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.84it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.50it/s]
 30%|███       | 15/50 [00:08<00:15,  2.24it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.36it/s]
 36%|███▌      | 18/50 [00:08<00:09,  3.54it/s]
 40%|████      | 20/50 [00:09<00:05,  5.00it/s]
 44%|████▍     | 22/50 [00:10<00:10,  2.67it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.21it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.67it/s]
 60%|██████    | 30/50 [00:12<00:05,  3.98it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.66it/s]
 70%|███████   | 35/50 [00:14<00:05,  2.61it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.58it/s]
 74%|███████▍  | 37/50 [00:15<00:04,  2.95it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.01it/s]
 78%|███████▊  | 39/50 [00:16<00:05,  2.09it/s]
 80%|████████  | 40/50 [00:17<00:03,  2.58it/s]
 82%|████████▏ | 41/50 [00:18<00:05,  1.71it/s]
 84%|████████▍ | 42/50 [00:18<00:04,  1.92it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.47it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.70it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.36it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.61it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.31it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.57it/s]
 98%|█████████▊| 49/50 [00:23<00:00,  1.29it/s]
+
+  clip 1/4: 42.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.40it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.40it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.26it/s]
 30%|███       | 15/50 [00:09<00:16,  2.08it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.18it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.87it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.68it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.49it/s]
 46%|████▌     | 23/50 [00:11<00:09,  2.72it/s]
 48%|████▊     | 24/50 [00:11<00:08,  2.97it/s]
 50%|█████     | 25/50 [00:12<00:07,  3.20it/s]
 52%|█████▏    | 26/50 [00:13<00:13,  1.75it/s]
 54%|█████▍    | 27/50 [00:13<00:12,  1.86it/s]
 56%|█████▌    | 28/50 [00:14<00:09,  2.38it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.71it/s]
 60%|██████    | 30/50 [00:14<00:06,  3.16it/s]
 62%|██████▏   | 31/50 [00:15<00:11,  1.69it/s]
 64%|██████▍   | 32/50 [00:16<00:10,  1.76it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.26it/s]
 70%|███████   | 35/50 [00:17<00:07,  1.91it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.98it/s]
 74%|███████▍  | 37/50 [00:18<00:05,  2.39it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.57it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.75it/s]
 80%|████████  | 40/50 [00:20<00:04,  2.18it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.47it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.63it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.27it/s]
 88%|████████▊ | 44/50 [00:23<00:04,  1.39it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.14it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.34it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.12it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.32it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.12it/s]
+
+  clip 2/4: 52.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.37it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.40it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.59it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.50it/s]
 30%|███       | 15/50 [00:09<00:17,  2.02it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.06it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.52it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.71it/s]
 42%|████▏     | 21/50 [00:11<00:11,  2.43it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.48it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.62it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.49it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.41it/s]
 56%|█████▌    | 28/50 [00:13<00:08,  2.69it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.83it/s]
 60%|██████    | 30/50 [00:14<00:06,  2.88it/s]
 62%|██████▏   | 31/50 [00:15<00:11,  1.61it/s]
 64%|██████▍   | 32/50 [00:16<00:10,  1.71it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.17it/s]
 68%|██████▊   | 34/50 [00:16<00:06,  2.57it/s]
 70%|███████   | 35/50 [00:18<00:09,  1.54it/s]
 72%|███████▏  | 36/50 [00:18<00:08,  1.70it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.62it/s]
 78%|███████▊  | 39/50 [00:20<00:06,  1.75it/s]
 80%|████████  | 40/50 [00:20<00:04,  2.20it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.49it/s]
 84%|████████▍ | 42/50 [00:22<00:04,  1.67it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.30it/s]
 88%|████████▊ | 44/50 [00:23<00:04,  1.43it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.18it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.42it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.18it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.42it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.18it/s]
+
+  clip 3/4: 52.1s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.38it/s]
 18%|█▊        | 9/50 [00:06<00:21,  1.89it/s]
 20%|██        | 10/50 [00:07<00:29,  1.35it/s]
 22%|██▏       | 11/50 [00:08<00:26,  1.49it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.95it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.50it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.19it/s]
 30%|███       | 15/50 [00:09<00:20,  1.71it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.86it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.19it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.70it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.54it/s]
 46%|████▌     | 23/50 [00:12<00:09,  2.91it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.49it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.33it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.27it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.25it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.34it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.41it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.38it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.44it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.08it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.17it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.93it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.02it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.54it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.71it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.35it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.57it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.22it/s]
+
+  clip 4/4: 46.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000007/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000007/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 14668.0ms  total=  14.7s
+  clip_total         n=  4  mean= 48287.1ms  total= 193.1s
 
+Done in 208.0s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000008.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a contemporary living room with a seating area, with light gray walls, light yellow honey-oak hardwood floor, and a low sofa with neutral cushions facing a flat console under soft lamplight. In the foreground, there is a pale gray three-drawer dresser with smooth handle-less integrated edge pulls, and a stack of coffee-table books and a small fern in a pot sit on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=278  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:56,  1.15s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.41it/s]
  6%|▌         | 3/50 [00:02<00:41,  1.12it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.38it/s]
 10%|█         | 5/50 [00:04<00:38,  1.16it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.39it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 18%|█▊        | 9/50 [00:06<00:22,  1.83it/s]
 20%|██        | 10/50 [00:07<00:29,  1.38it/s]
 22%|██▏       | 11/50 [00:07<00:25,  1.54it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.64it/s]
 30%|███       | 15/50 [00:09<00:15,  2.21it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.15it/s]
 34%|███▍      | 17/50 [00:09<00:13,  2.53it/s]
 36%|███▌      | 18/50 [00:10<00:10,  2.92it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.50it/s]
 40%|████      | 20/50 [00:10<00:07,  3.89it/s]
 42%|████▏     | 21/50 [00:11<00:14,  1.97it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.05it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.60it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.23it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.92it/s]
 52%|█████▏    | 26/50 [00:13<00:13,  1.84it/s]
 54%|█████▍    | 27/50 [00:14<00:11,  2.04it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.41it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.41it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.50it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.71it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.22it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.35it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.07it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.19it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  2.00it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.15it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.64it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.84it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.46it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.67it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.36it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.61it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.32it/s]
+
+  clip 1/4: 43.5s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.17s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.36it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.33it/s]
 10%|█         | 5/50 [00:04<00:42,  1.06it/s]
 12%|█▏        | 6/50 [00:05<00:35,  1.26it/s]
 14%|█▍        | 7/50 [00:06<00:41,  1.04it/s]
 16%|█▌        | 8/50 [00:06<00:33,  1.24it/s]
 18%|█▊        | 9/50 [00:07<00:25,  1.64it/s]
 20%|██        | 10/50 [00:08<00:32,  1.24it/s]
 22%|██▏       | 11/50 [00:08<00:27,  1.42it/s]
 24%|██▍       | 12/50 [00:08<00:21,  1.80it/s]
 26%|██▌       | 13/50 [00:09<00:16,  2.21it/s]
 28%|██▊       | 14/50 [00:09<00:13,  2.76it/s]
 30%|███       | 15/50 [00:10<00:21,  1.60it/s]
 32%|███▏      | 16/50 [00:11<00:19,  1.73it/s]
 34%|███▍      | 17/50 [00:11<00:14,  2.23it/s]
 36%|███▌      | 18/50 [00:11<00:11,  2.72it/s]
 38%|███▊      | 19/50 [00:11<00:09,  3.12it/s]
 40%|████      | 20/50 [00:11<00:08,  3.42it/s]
 42%|████▏     | 21/50 [00:12<00:16,  1.74it/s]
 44%|████▍     | 22/50 [00:13<00:15,  1.82it/s]
 46%|████▌     | 23/50 [00:13<00:12,  2.20it/s]
 48%|████▊     | 24/50 [00:13<00:09,  2.66it/s]
 50%|█████     | 25/50 [00:14<00:08,  2.96it/s]
 52%|█████▏    | 26/50 [00:15<00:14,  1.64it/s]
 54%|█████▍    | 27/50 [00:15<00:13,  1.76it/s]
 56%|█████▌    | 28/50 [00:16<00:10,  2.17it/s]
 58%|█████▊    | 29/50 [00:16<00:08,  2.40it/s]
 60%|██████    | 30/50 [00:16<00:07,  2.83it/s]
 62%|██████▏   | 31/50 [00:17<00:11,  1.61it/s]
 64%|██████▍   | 32/50 [00:18<00:10,  1.65it/s]
 66%|██████▌   | 33/50 [00:18<00:08,  2.04it/s]
 68%|██████▊   | 34/50 [00:18<00:06,  2.30it/s]
 70%|███████   | 35/50 [00:20<00:10,  1.48it/s]
 72%|███████▏  | 36/50 [00:20<00:08,  1.62it/s]
 74%|███████▍  | 37/50 [00:20<00:06,  2.10it/s]
 76%|███████▌  | 38/50 [00:22<00:08,  1.42it/s]
 78%|███████▊  | 39/50 [00:22<00:06,  1.65it/s]
 82%|████████▏ | 41/50 [00:23<00:05,  1.64it/s]
 84%|████████▍ | 42/50 [00:24<00:04,  1.81it/s]
 86%|████████▌ | 43/50 [00:25<00:05,  1.40it/s]
 88%|████████▊ | 44/50 [00:25<00:03,  1.61it/s]
 90%|█████████ | 45/50 [00:26<00:03,  1.29it/s]
 92%|█████████▏| 46/50 [00:27<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:28<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:28<00:01,  1.45it/s]
 98%|█████████▊| 49/50 [00:29<00:00,  1.20it/s]
+
+  clip 2/4: 48.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.15it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.43it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.43it/s]
 20%|██        | 10/50 [00:07<00:26,  1.52it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.68it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.65it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.19it/s]
 30%|███       | 15/50 [00:09<00:18,  1.87it/s]
 32%|███▏      | 16/50 [00:09<00:18,  1.87it/s]
 34%|███▍      | 17/50 [00:09<00:13,  2.39it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.41it/s]
 40%|████      | 20/50 [00:10<00:08,  3.68it/s]
 42%|████▏     | 21/50 [00:11<00:14,  1.96it/s]
 44%|████▍     | 22/50 [00:12<00:14,  1.97it/s]
 46%|████▌     | 23/50 [00:12<00:12,  2.17it/s]
 48%|████▊     | 24/50 [00:12<00:10,  2.48it/s]
 50%|█████     | 25/50 [00:12<00:09,  2.76it/s]
 52%|█████▏    | 26/50 [00:14<00:14,  1.62it/s]
 54%|█████▍    | 27/50 [00:14<00:13,  1.75it/s]
 56%|█████▌    | 28/50 [00:14<00:09,  2.22it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.68it/s]
 60%|██████    | 30/50 [00:15<00:06,  3.00it/s]
 62%|██████▏   | 31/50 [00:16<00:11,  1.66it/s]
 64%|██████▍   | 32/50 [00:16<00:10,  1.75it/s]
 66%|██████▌   | 33/50 [00:17<00:07,  2.14it/s]
 68%|██████▊   | 34/50 [00:17<00:06,  2.46it/s]
 70%|███████   | 35/50 [00:18<00:09,  1.53it/s]
 72%|███████▏  | 36/50 [00:19<00:08,  1.68it/s]
 74%|███████▍  | 37/50 [00:19<00:06,  2.15it/s]
 76%|███████▌  | 38/50 [00:20<00:08,  1.44it/s]
 78%|███████▊  | 39/50 [00:21<00:07,  1.53it/s]
 80%|████████  | 40/50 [00:21<00:05,  1.99it/s]
 82%|████████▏ | 41/50 [00:22<00:06,  1.39it/s]
 84%|████████▍ | 42/50 [00:23<00:05,  1.41it/s]
 86%|████████▌ | 43/50 [00:24<00:06,  1.13it/s]
 88%|████████▊ | 44/50 [00:24<00:04,  1.34it/s]
 90%|█████████ | 45/50 [00:26<00:04,  1.13it/s]
 92%|█████████▏| 46/50 [00:26<00:02,  1.35it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.14it/s]
 96%|█████████▌| 48/50 [00:28<00:01,  1.38it/s]
 98%|█████████▊| 49/50 [00:29<00:00,  1.16it/s]
+
+  clip 3/4: 45.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.15it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.43it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.43it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.68it/s]
 28%|██▊       | 14/50 [00:07<00:11,  3.22it/s]
 30%|███       | 15/50 [00:09<00:16,  2.07it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.20it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.34it/s]
 40%|████      | 20/50 [00:09<00:06,  4.74it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.47it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.29it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.67it/s]
 52%|█████▏    | 26/50 [00:12<00:11,  2.10it/s]
 54%|█████▍    | 27/50 [00:13<00:10,  2.14it/s]
 56%|█████▌    | 28/50 [00:13<00:08,  2.55it/s]
 58%|█████▊    | 29/50 [00:13<00:07,  2.96it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.56it/s]
 62%|██████▏   | 31/50 [00:14<00:10,  1.84it/s]
 64%|██████▍   | 32/50 [00:15<00:09,  1.86it/s]
 66%|██████▌   | 33/50 [00:15<00:08,  2.10it/s]
 68%|██████▊   | 34/50 [00:16<00:06,  2.49it/s]
 70%|███████   | 35/50 [00:17<00:09,  1.54it/s]
 72%|███████▏  | 36/50 [00:17<00:08,  1.66it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.57it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.76it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.68it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.77it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.38it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.52it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.25it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.47it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.45it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.20it/s]
+
+  clip 4/4: 41.7s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000008/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000008/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13676.7ms  total=  13.7s
+  clip_total         n=  4  mean= 44880.1ms  total= 179.5s
 
+Done in 193.4s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000009.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a dining room with a table area, with off-white walls and simple white wainscoting, light gray engineered-wood floor, and a slim linear pendant hanging over a wooden dining table. In the foreground, there is a mid-century teak three-drawer dresser with sculpted wooden pulls, and a ceramic fruit bowl and a folded table runner are placed on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=266  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.11s/it]
  4%|▍         | 2/50 [00:01<00:32,  1.46it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.16it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.52it/s]
 10%|█         | 5/50 [00:04<00:36,  1.22it/s]
 12%|█▏        | 6/50 [00:04<00:29,  1.52it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.25it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.53it/s]
 20%|██        | 10/50 [00:06<00:24,  1.63it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.81it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.47it/s]
 30%|███       | 15/50 [00:08<00:15,  2.23it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.34it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.51it/s]
 40%|████      | 20/50 [00:09<00:06,  4.94it/s]
 44%|████▍     | 22/50 [00:10<00:10,  2.58it/s]
 46%|████▌     | 23/50 [00:10<00:08,  3.00it/s]
 50%|█████     | 25/50 [00:10<00:06,  4.11it/s]
 52%|█████▏    | 26/50 [00:11<00:10,  2.40it/s]
 54%|█████▍    | 27/50 [00:12<00:09,  2.34it/s]
 56%|█████▌    | 28/50 [00:12<00:08,  2.60it/s]
 58%|█████▊    | 29/50 [00:12<00:07,  2.77it/s]
 60%|██████    | 30/50 [00:13<00:06,  3.32it/s]
 62%|██████▏   | 31/50 [00:14<00:10,  1.88it/s]
 64%|██████▍   | 32/50 [00:14<00:09,  1.98it/s]
 66%|██████▌   | 33/50 [00:14<00:06,  2.47it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.03it/s]
 72%|███████▏  | 36/50 [00:16<00:07,  1.90it/s]
 74%|███████▍  | 37/50 [00:16<00:05,  2.17it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.55it/s]
 78%|███████▊  | 39/50 [00:18<00:07,  1.53it/s]
 80%|████████  | 40/50 [00:19<00:05,  1.85it/s]
 82%|████████▏ | 41/50 [00:20<00:06,  1.38it/s]
 84%|████████▍ | 42/50 [00:20<00:05,  1.57it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.26it/s]
 88%|████████▊ | 44/50 [00:22<00:04,  1.42it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.21it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.39it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.19it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.37it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.17it/s]
+
+  clip 1/4: 50.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.39it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.59it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.02it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.52it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.17it/s]
 30%|███       | 15/50 [00:09<00:20,  1.73it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.85it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.39it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.05it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.42it/s]
 40%|████      | 20/50 [00:10<00:07,  4.12it/s]
 42%|████▏     | 21/50 [00:11<00:15,  1.89it/s]
 44%|████▍     | 22/50 [00:12<00:14,  1.89it/s]
 46%|████▌     | 23/50 [00:12<00:11,  2.36it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.94it/s]
 52%|█████▏    | 26/50 [00:13<00:11,  2.06it/s]
 54%|█████▍    | 27/50 [00:14<00:12,  1.89it/s]
 56%|█████▌    | 28/50 [00:14<00:10,  2.14it/s]
 58%|█████▊    | 29/50 [00:14<00:08,  2.61it/s]
 60%|██████    | 30/50 [00:15<00:06,  3.20it/s]
 62%|██████▏   | 31/50 [00:16<00:10,  1.76it/s]
 64%|██████▍   | 32/50 [00:16<00:10,  1.74it/s]
 66%|██████▌   | 33/50 [00:17<00:07,  2.15it/s]
 70%|███████   | 35/50 [00:18<00:08,  1.85it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.85it/s]
 74%|███████▍  | 37/50 [00:19<00:05,  2.30it/s]
 76%|███████▌  | 38/50 [00:20<00:07,  1.51it/s]
 78%|███████▊  | 39/50 [00:20<00:06,  1.73it/s]
 82%|████████▏ | 41/50 [00:21<00:05,  1.68it/s]
 84%|████████▍ | 42/50 [00:22<00:04,  1.85it/s]
 86%|████████▌ | 43/50 [00:23<00:04,  1.42it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.62it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.30it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.47it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.21it/s]
+
+  clip 2/4: 49.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:37,  1.28it/s]
  6%|▌         | 3/50 [00:02<00:46,  1.01it/s]
  8%|▊         | 4/50 [00:03<00:36,  1.27it/s]
 10%|█         | 5/50 [00:04<00:42,  1.06it/s]
 12%|█▏        | 6/50 [00:05<00:33,  1.33it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.12it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.35it/s]
 18%|█▊        | 9/50 [00:06<00:22,  1.85it/s]
 20%|██        | 10/50 [00:07<00:29,  1.34it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.52it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.01it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.64it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.19it/s]
 30%|███       | 15/50 [00:10<00:20,  1.68it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.84it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.08it/s]
 40%|████      | 20/50 [00:10<00:06,  4.60it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.29it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.17it/s]
 50%|█████     | 25/50 [00:12<00:07,  3.56it/s]
 52%|█████▏    | 26/50 [00:13<00:11,  2.06it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.12it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.61it/s]
 60%|█���████    | 30/50 [00:14<00:05,  3.77it/s]
 62%|██████▏   | 31/50 [00:15<00:09,  2.07it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.07it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.22it/s]
 70%|███████   | 35/50 [00:17<00:07,  1.96it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.11it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.88it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.02it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.85it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.98it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.50it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.69it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.54it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.22it/s]
+
+  clip 3/4: 48.1s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.77it/s]
 30%|███       | 15/50 [00:09<00:16,  2.19it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.20it/s]
 34%|███▍      | 17/50 [00:09<00:12,  2.68it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.93it/s]
 40%|████      | 20/50 [00:09<00:06,  4.53it/s]
 42%|████▏     | 21/50 [00:11<00:13,  2.18it/s]
 44%|████▍     | 22/50 [00:11<00:12,  2.23it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.41it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.41it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.47it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.28it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.48it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.43it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.37it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.06it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.15it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.94it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.05it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.55it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.73it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.37it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.57it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.50it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.22it/s]
+
+  clip 4/4: 41.1s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000009/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000009/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 17031.0ms  total=  17.0s
+  clip_total         n=  4  mean= 47428.7ms  total= 189.7s
 
+Done in 206.9s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000010.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a residential entryway with a foyer area, with crisp white walls, light yellow oak parquet floor in a herringbone pattern, and a round mirror catching daylight from a glass front door. In the foreground, there is a two-tone three-drawer dresser with a white frame and natural-wood drawer fronts and round black knobs, and a small dish of keys and a folded pair of gloves rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=275  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.11s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.53it/s]
  6%|▌         | 3/50 [00:02<00:39,  1.18it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.54it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.55it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:06<00:24,  1.65it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.85it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.52it/s]
 30%|███       | 15/50 [00:08<00:15,  2.25it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.37it/s]
 38%|███▊      | 19/50 [00:08<00:07,  4.17it/s]
 42%|████▏     | 21/50 [00:10<00:09,  2.91it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.91it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.69it/s]
 54%|█████▍    | 27/50 [00:11<00:08,  2.76it/s]
 58%|█████▊    | 29/50 [00:12<00:05,  3.67it/s]
 62%|██████▏   | 31/50 [00:13<00:06,  2.72it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.70it/s]
 66%|██████▌   | 33/50 [00:13<00:05,  3.11it/s]
 68%|██████▊   | 34/50 [00:13<00:04,  3.48it/s]
 70%|███████   | 35/50 [00:15<00:07,  2.04it/s]
 72%|███████▏  | 36/50 [00:15<00:06,  2.17it/s]
 76%|███████▌  | 38/50 [00:16<00:06,  1.92it/s]
 78%|███████▊  | 39/50 [00:16<00:05,  2.09it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  1.94it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.10it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.60it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.81it/s]
 90%|█████████ | 45/50 [00:20<00:03,  1.44it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.68it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.36it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.61it/s]
 98%|█████████▊| 49/50 [00:23<00:00,  1.32it/s]
+
+  clip 1/4: 39.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.20s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.31it/s]
  6%|▌         | 3/50 [00:02<00:46,  1.01it/s]
  8%|▊         | 4/50 [00:03<00:36,  1.25it/s]
 10%|█         | 5/50 [00:04<00:42,  1.06it/s]
 12%|█▏        | 6/50 [00:05<00:34,  1.29it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.08it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.31it/s]
 18%|█▊        | 9/50 [00:06<00:23,  1.74it/s]
 20%|██        | 10/50 [00:08<00:30,  1.31it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.56it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.25it/s]
 30%|███       | 15/50 [00:09<00:17,  2.02it/s]
 32%|███▏      | 16/50 [00:10<00:15,  2.15it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.34it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.79it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.78it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.36it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.55it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.58it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.74it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.59it/s]
 64%|██████▍   | 32/50 [00:15<00:06,  2.61it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.74it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.15it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.10it/s]
 74%|███████▍  | 37/50 [00:17<00:05,  2.59it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.66it/s]
 78%|███████▊  | 39/50 [00:18<00:06,  1.83it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.71it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.84it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.41it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.57it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.25it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.44it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.16it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.35it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.15it/s]
+
+  clip 2/4: 42.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.31it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.02it/s]
  8%|▊         | 4/50 [00:03<00:36,  1.24it/s]
 10%|█         | 5/50 [00:04<00:43,  1.03it/s]
 12%|█▏        | 6/50 [00:05<00:36,  1.21it/s]
 14%|█▍        | 7/50 [00:06<00:42,  1.02it/s]
 16%|█▌        | 8/50 [00:07<00:35,  1.19it/s]
 20%|██        | 10/50 [00:08<00:29,  1.34it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.54it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.51it/s]
 30%|███       | 15/50 [00:10<00:16,  2.07it/s]
 32%|███▏      | 16/50 [00:10<00:15,  2.16it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.24it/s]
 40%|████      | 20/50 [00:10<00:06,  4.53it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.48it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.92it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.48it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.30it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.53it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.46it/s]
 66%|██████▌   | 33/50 [00:15<00:05,  2.89it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.26it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.28it/s]
 74%|███████▍  | 37/50 [00:17<00:04,  2.77it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.76it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.87it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.73it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.76it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.36it/s]
 88%|████████▊ | 44/50 [00:22<00:04,  1.43it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.17it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.34it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.13it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.32it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.12it/s]
+
+  clip 3/4: 44.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:39,  1.21it/s]
  6%|▌         | 3/50 [00:03<00:47,  1.02s/it]
  8%|▊         | 4/50 [00:03<00:36,  1.25it/s]
 10%|█         | 5/50 [00:04<00:42,  1.06it/s]
 12%|█▏        | 6/50 [00:05<00:33,  1.30it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.11it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.32it/s]
 20%|██        | 10/50 [00:07<00:27,  1.44it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.58it/s]
 26%|██▌       | 13/50 [00:08<00:15,  2.43it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.93it/s]
 30%|███       | 15/50 [00:09<00:19,  1.78it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.94it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.16it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.72it/s]
 44%|████▍     | 22/50 [00:12<00:10,  2.72it/s]
 48%|████▊     | 24/50 [00:12<00:06,  3.78it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.63it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.64it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.81it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.62it/s]
 64%|██████▍   | 32/50 [00:15<00:06,  2.64it/s]
 70%|███████   | 35/50 [00:16<00:05,  2.50it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.53it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.13it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.20it/s]
 80%|████████  | 40/50 [00:18<00:03,  2.63it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.75it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.92it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.44it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.59it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.26it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.38it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.14it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.30it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.10it/s]
+
+  clip 4/4: 45.4s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000010/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000010/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 12692.1ms  total=  12.7s
+  clip_total         n=  4  mean= 42992.7ms  total= 172.0s
 
+Done in 184.9s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000011.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a bright residential hallway with a transitional area, with smooth white walls and a row of framed black-and-white prints, light gray laminate floor, and warm wash light from evenly spaced ceiling fixtures. In the foreground, there is a distressed white farmhouse three-drawer dresser with oil-rubbed bronze cup pulls, and a slim table lamp and a small woven tray sit on the surface. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=261  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.12s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.51it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.17it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.53it/s]
 10%|█         | 5/50 [00:03<00:36,  1.23it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.54it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:28,  1.47it/s]
 18%|█▊        | 9/50 [00:05<00:20,  1.99it/s]
 20%|██        | 10/50 [00:07<00:27,  1.44it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.66it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.82it/s]
 30%|███       | 15/50 [00:08<00:15,  2.21it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.34it/s]
 36%|███▌      | 18/50 [00:09<00:08,  3.57it/s]
 40%|████      | 20/50 [00:09<00:05,  5.04it/s]
 44%|████▍     | 22/50 [00:10<00:10,  2.68it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.22it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.67it/s]
 58%|█████▊    | 29/50 [00:12<00:06,  3.38it/s]
 62%|██████▏   | 31/50 [00:13<00:07,  2.59it/s]
 64%|██████▍   | 32/50 [00:14<00:06,  2.61it/s]
 70%|███████   | 35/50 [00:15<00:05,  2.57it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.51it/s]
 74%|███████▍  | 37/50 [00:15<00:04,  2.88it/s]
 76%|███████▌  | 38/50 [00:17<00:06,  1.94it/s]
 78%|███████▊  | 39/50 [00:17<00:05,  2.10it/s]
 80%|████████  | 40/50 [00:17<00:03,  2.62it/s]
 82%|████████▏ | 41/50 [00:18<00:05,  1.71it/s]
 84%|████████▍ | 42/50 [00:18<00:04,  1.93it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.46it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.70it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.36it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.60it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.31it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.57it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.29it/s]
+
+  clip 1/4: 39.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.34it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.05it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.36it/s]
 10%|█         | 5/50 [00:04<00:40,  1.11it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.35it/s]
 18%|█▊        | 9/50 [00:06<00:22,  1.84it/s]
 20%|██        | 10/50 [00:07<00:29,  1.34it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.52it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.03it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.60it/s]
 30%|███       | 15/50 [00:09<00:17,  1.98it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.09it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.27it/s]
 40%|████      | 20/50 [00:10<00:06,  4.41it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.31it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.39it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.59it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.90it/s]
 52%|█████▏    | 26/50 [00:13<00:12,  1.99it/s]
 54%|█████▍    | 27/50 [00:14<00:11,  2.09it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.85it/s]
 60%|██████    | 30/50 [00:14<00:06,  2.98it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.76it/s]
 64%|██████▍   | 32/50 [00:16<00:09,  1.85it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.33it/s]
 68%|██████▊   | 34/50 [00:16<00:05,  2.74it/s]
 70%|███████   | 35/50 [00:17<00:09,  1.63it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.77it/s]
 74%|███████▍  | 37/50 [00:18<00:05,  2.28it/s]
 76%|███████▌  | 38/50 [00:19<00:08,  1.48it/s]
 78%|███████▊  | 39/50 [00:20<00:06,  1.64it/s]
 80%|████████  | 40/50 [00:20<00:04,  2.09it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.39it/s]
 84%|████████▍ | 42/50 [00:22<00:05,  1.53it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.22it/s]
 88%|████████▊ | 44/50 [00:23<00:04,  1.40it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.14it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.38it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.14it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.38it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.16it/s]
+
+  clip 2/4: 48.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:38,  1.25it/s]
  6%|▌         | 3/50 [00:02<00:46,  1.01it/s]
  8%|▊         | 4/50 [00:03<00:36,  1.26it/s]
 10%|█         | 5/50 [00:04<00:42,  1.05it/s]
 12%|█▏        | 6/50 [00:05<00:34,  1.27it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.08it/s]
 16%|█▌        | 8/50 [00:06<00:33,  1.27it/s]
 18%|█▊        | 9/50 [00:07<00:24,  1.65it/s]
 20%|██        | 10/50 [00:08<00:31,  1.26it/s]
 22%|██▏       | 11/50 [00:08<00:26,  1.46it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.97it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.17it/s]
 30%|███       | 15/50 [00:10<00:18,  1.87it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.99it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.12it/s]
 40%|████      | 20/50 [00:10<00:06,  4.37it/s]
 42%|████▏     | 21/50 [00:12<00:12,  2.27it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.28it/s]
 46%|████▌     | 23/50 [00:12<00:09,  2.82it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.26it/s]
 52%|█████▏    | 26/50 [00:14<00:10,  2.18it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.20it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.71it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.95it/s]
 62%|██████▏   | 31/50 [00:16<00:09,  2.08it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.20it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.25it/s]
 72%|████��██▏  | 36/50 [00:18<00:06,  2.33it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.98it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.09it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.88it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.99it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.52it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.70it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.56it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.22it/s]
+
+  clip 3/4: 47.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.31it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.04it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.37it/s]
 10%|█         | 5/50 [00:04<00:40,  1.12it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.40it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.49it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.62it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.06it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.63it/s]
 30%|███       | 15/50 [00:09<00:17,  2.03it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.09it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.09it/s]
 40%|████      | 20/50 [00:10<00:07,  4.26it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.30it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.40it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.26it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.40it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.46it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.26it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.47it/s]
 66%|██████▌   | 33/50 [00:15<00:05,  2.86it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.25it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.24it/s]
 74%|███████▍  | 37/50 [00:17<00:05,  2.59it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.68it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  1.85it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.73it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.87it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.44it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.63it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.30it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.44it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.19it/s]
+
+  clip 4/4: 45.1s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000011/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000011/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 12010.9ms  total=  12.0s
+  clip_total         n=  4  mean= 45144.0ms  total= 180.6s
 
+Done in 192.7s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000012.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a boutique retail shop with a display area, with matte white walls and a single horizontal black accent line at picture height, medium gray polished concrete floor, and warm spotlights from an exposed ceiling track. In the foreground, there is a glossy white three-drawer dresser with polished gold handles, and a folded cashmere sweater and a small leather price tag rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=263  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:55,  1.13s/it]
  4%|▍         | 2/50 [00:01<00:32,  1.50it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.17it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.53it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.54it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:28,  1.47it/s]
 20%|██        | 10/50 [00:07<00:25,  1.56it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.66it/s]
 26%|██▌       | 13/50 [00:07<00:14,  2.55it/s]
 28%|██▊       | 14/50 [00:07<00:11,  3.06it/s]
 30%|███       | 15/50 [00:08<00:18,  1.89it/s]
 32%|███▏      | 16/50 [00:09<00:17,  1.89it/s]
 34%|███▍      | 17/50 [00:09<00:14,  2.29it/s]
 36%|███▌      | 18/50 [00:09<00:11,  2.79it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.23it/s]
 40%|████      | 20/50 [00:10<00:07,  3.81it/s]
 42%|████▏     | 21/50 [00:11<00:14,  1.96it/s]
 44%|████▍     | 22/50 [00:11<00:13,  2.14it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.60it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.55it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.62it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.92it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.68it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.54it/s]
 66%|██████▌   | 33/50 [00:15<00:05,  2.90it/s]
 68%|██████▊   | 34/50 [00:15<00:05,  3.16it/s]
 70%|███████   | 35/50 [00:16<00:07,  1.90it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.90it/s]
 74%|███████▍  | 37/50 [00:17<00:06,  2.17it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.51it/s]
 78%|███████▊  | 39/50 [00:18<00:06,  1.70it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.20it/s]
 82%|████████▏ | 41/50 [00:20<00:06,  1.48it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.63it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.31it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.57it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.29it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.55it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.29it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.47it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.23it/s]
+
+  clip 1/4: 44.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.17s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.44it/s]
  6%|▌         | 3/50 [00:02<00:42,  1.10it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.39it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.71it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.76it/s]
 30%|███       | 15/50 [00:09<00:15,  2.19it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.19it/s]
 34%|███▍      | 17/50 [00:09<00:12,  2.62it/s]
 36%|███▌      | 18/50 [00:09<00:10,  3.14it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.82it/s]
 40%|████      | 20/50 [00:10<00:07,  4.20it/s]
 42%|████▏     | 21/50 [00:11<00:15,  1.93it/s]
 44%|████▍     | 22/50 [00:11<00:13,  2.01it/s]
 46%|████▌     | 23/50 [00:11<00:10,  2.49it/s]
 48%|████▊     | 24/50 [00:12<00:09,  2.89it/s]
 50%|█████     | 25/50 [00:12<00:07,  3.27it/s]
 52%|█████▏    | 26/50 [00:13<00:14,  1.68it/s]
 54%|█████▍    | 27/50 [00:14<00:13,  1.69it/s]
 56%|█████▌    | 28/50 [00:14<00:10,  2.11it/s]
 58%|█████▊    | 29/50 [00:14<00:08,  2.59it/s]
 60%|██████    | 30/50 [00:14<00:06,  3.22it/s]
 62%|██████▏   | 31/50 [00:16<00:11,  1.70it/s]
 64%|██████▍   | 32/50 [00:16<00:09,  1.83it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.30it/s]
 68%|██████▊   | 34/50 [00:16<00:05,  2.71it/s]
 70%|███████   | 35/50 [00:18<00:09,  1.60it/s]
 72%|███████▏  | 36/50 [00:18<00:08,  1.65it/s]
 74%|███████▍  | 37/50 [00:18<00:06,  2.14it/s]
 76%|███████▌  | 38/50 [00:20<00:08,  1.44it/s]
 78%|███████▊  | 39/50 [00:20<00:07,  1.54it/s]
 80%|████████  | 40/50 [00:20<00:04,  2.01it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.41it/s]
 84%|████████▍ | 42/50 [00:22<00:05,  1.51it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.21it/s]
 88%|████████▊ | 44/50 [00:24<00:04,  1.39it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.16it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.34it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.13it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.33it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.12it/s]
+
+  clip 2/4: 46.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.15it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.43it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.43it/s]
 20%|██        | 10/50 [00:07<00:26,  1.52it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.72it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.29it/s]
 30%|███       | 15/50 [00:09<00:16,  2.09it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.21it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.36it/s]
 40%|████      | 20/50 [00:09<00:06,  4.69it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.48it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.37it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.46it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.34it/s]
 56%|█████▌    | 28/50 [00:13<00:08,  2.68it/s]
 58%|��████▊    | 29/50 [00:13<00:07,  2.91it/s]
 60%|██████    | 30/50 [00:13<00:06,  3.13it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.77it/s]
 64%|██████▍   | 32/50 [00:15<00:09,  1.82it/s]
 66%|██████▌   | 33/50 [00:15<00:07,  2.28it/s]
 68%|██████▊   | 34/50 [00:15<00:05,  2.80it/s]
 70%|███████   | 35/50 [00:17<00:09,  1.65it/s]
 72%|███████▏  | 36/50 [00:17<00:08,  1.71it/s]
 74%|███████▍  | 37/50 [00:17<00:06,  2.13it/s]
 76%|███████▌  | 38/50 [00:19<00:08,  1.42it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.57it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.01it/s]
 82%|████████▏ | 41/50 [00:20<00:06,  1.40it/s]
 84%|████████▍ | 42/50 [00:21<00:05,  1.56it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.22it/s]
 88%|████████▊ | 44/50 [00:23<00:04,  1.42it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.17it/s]
 92%|█████████▏| 46/50 [00:24<00:03,  1.30it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.08it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.22it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.04it/s]
+
+  clip 3/4: 43.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:01,  1.26s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.34it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.05it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.39it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.69it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.70it/s]
 30%|███       | 15/50 [00:09<00:16,  2.17it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.27it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.93it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.70it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.70it/s]
 48%|████▊     | 24/50 [00:11<00:06,  3.83it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.66it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.67it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.85it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.65it/s]
 64%|██████▍   | 32/50 [00:14<00:06,  2.58it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.55it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.13it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.14it/s]
 74%|███████▍  | 37/50 [00:16<00:04,  2.60it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.69it/s]
 78%|███████▊  | 39/50 [00:18<00:06,  1.74it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.08it/s]
 82%|████████▏ | 41/50 [00:20<00:06,  1.43it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.60it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.26it/s]
 88%|████████▊ | 44/50 [00:22<00:04,  1.46it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.20it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.42it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.17it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.40it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.17it/s]
+
+  clip 4/4: 43.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000012/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000012/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13413.6ms  total=  13.4s
+  clip_total         n=  4  mean= 44414.6ms  total= 177.7s
 
+Done in 191.3s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000013.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a boutique hotel room with a lounge area, with crisp white walls and slim brass picture rails, bleached white oak plank floor, and a single arc floor lamp curving over a side table. In the foreground, there is a light gray-washed oak three-drawer dresser with slim black handles, and a folded linen throw and a small porcelain ring tray sit on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=258  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:56,  1.14s/it]
  4%|▍         | 2/50 [00:01<00:32,  1.49it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.16it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.51it/s]
 10%|█         | 5/50 [00:04<00:37,  1.20it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.43it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.19it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.44it/s]
 18%|█▊        | 9/50 [00:06<00:21,  1.95it/s]
 20%|██        | 10/50 [00:07<00:28,  1.41it/s]
 22%|██▏       | 11/50 [00:07<00:25,  1.53it/s]
 26%|██▌       | 13/50 [00:07<00:14,  2.62it/s]
 30%|███       | 15/50 [00:09<00:16,  2.18it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.23it/s]
 34%|███▍      | 17/50 [00:09<00:12,  2.64it/s]
 36%|███▌      | 18/50 [00:09<00:10,  3.05it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.72it/s]
 42%|████▏     | 21/50 [00:11<00:11,  2.42it/s]
 44%|████▍     | 22/50 [00:11<00:12,  2.32it/s]
 46%|████▌     | 23/50 [00:11<00:09,  2.82it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.06it/s]
 50%|█████     | 25/50 [00:12<00:07,  3.28it/s]
 52%|█████▏    | 26/50 [00:13<00:12,  1.85it/s]
 54%|█████▍    | 27/50 [00:13<00:11,  1.94it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.46it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.11it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.66it/s]
 62%|██████▏   | 31/50 [00:15<00:09,  1.91it/s]
 64%|██████▍   | 32/50 [00:15<00:09,  1.90it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.38it/s]
 68%|██████▊   | 34/50 [00:16<00:05,  2.92it/s]
 70%|███████   | 35/50 [00:17<00:08,  1.73it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.81it/s]
 74%|███████▍  | 37/50 [00:17<00:05,  2.38it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.59it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.65it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.12it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.52it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.66it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.30it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.53it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.27it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.53it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.51it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.25it/s]
+
+  clip 1/4: 47.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.20s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.35it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.05it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.35it/s]
 10%|█         | 5/50 [00:04<00:41,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 20%|██        | 10/50 [00:07<00:26,  1.49it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.67it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.63it/s]
 30%|███       | 15/50 [00:09<00:16,  2.12it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.17it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.10it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.31it/s]
 40%|████      | 20/50 [00:10<00:07,  3.79it/s]
 42%|████▏     | 21/50 [00:11<00:14,  2.02it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.02it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.48it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.70it/s]
 52%|█████▏    | 26/50 [00:13<00:11,  2.01it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.14it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.69it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.17it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.82it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.87it/s]
 64%|██████▍   | 32/50 [00:16<00:09,  1.91it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.35it/s]
 68%|██████▊   | 34/50 [00:16<00:05,  2.87it/s]
 70%|███████   | 35/50 [00:17<00:09,  1.66it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.86it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.74it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.89it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.76it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.89it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.44it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.64it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.31it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.53it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.47it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.21it/s]
+
+  clip 2/4: 47.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.39it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.13it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.34it/s]
 18%|█▊        | 9/50 [00:06<00:22,  1.82it/s]
 20%|██        | 10/50 [00:07<00:30,  1.33it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.56it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.64it/s]
 30%|███       | 15/50 [00:09<00:16,  2.10it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.11it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.07it/s]
 40%|████      | 20/50 [00:10<00:07,  4.22it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.31it/s]
 44%|████▍     | 22/50 [00:11<00:12,  2.30it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.40it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.45it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.42it/s]
 56%|█████▌    | 28/50 [00:13<00:07,  2.86it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.32it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.76it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.87it/s]
 64%|██████▍   | 32/50 [00:15<00:08,  2.05it/s]
 66%|██████▌   | 33/50 [00:15<00:06,  2.64it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.22it/s]
 70%|███████   | 35/50 [00:17<00:08,  1.73it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.94it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.78it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.95it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.81it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.95it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.48it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.68it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.54it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.22it/s]
+
+  clip 3/4: 46.1s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.40it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 20%|██        | 10/50 [00:07<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.63it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.08it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.57it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.24it/s]
 30%|███       | 15/50 [00:09<00:19,  1.76it/s]
 32%|███▏      | 16/50 [00:09<00:17,  1.96it/s]
 38%|███▊      | 19/50 [00:10<00:07,  3.96it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.65it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.64it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.41it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.55it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.90it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.51it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.45it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.49it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.11it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.19it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.96it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.07it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.58it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.75it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.37it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.58it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.22it/s]
+
+  clip 4/4: 40.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000013/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000013/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 12765.2ms  total=  12.8s
+  clip_total         n=  4  mean= 45416.9ms  total= 181.7s
 
+Done in 194.6s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000014.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a small home library with a reading area, with off-white walls and chair-rail trim, light yellow honey-toned oak parquet floor, and a built-in white bookshelf packed with hardcover books along the back wall. In the foreground, there is a navy-blue painted three-drawer dresser with brass ring pulls, and a green-shaded reading lamp and a small stack of leather-bound books are placed on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=301  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.10s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.53it/s]
  6%|▌         | 3/50 [00:02<00:39,  1.18it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.54it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.55it/s]
 14%|█▍        | 7/50 [00:05<00:33,  1.27it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.55it/s]
 20%|██        | 10/50 [00:06<00:24,  1.64it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.85it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.99it/s]
 30%|███       | 15/50 [00:08<00:14,  2.37it/s]
 32%|███▏      | 16/50 [00:08<00:13,  2.47it/s]
 38%|███▊      | 19/50 [00:08<00:07,  4.27it/s]
 42%|████▏     | 21/50 [00:10<00:09,  2.97it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.96it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.77it/s]
 54%|█████▍    | 27/50 [00:11<00:08,  2.79it/s]
 58%|█████▊    | 29/50 [00:12<00:05,  3.74it/s]
 62%|██████▏   | 31/50 [00:13<00:06,  2.79it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.80it/s]
 68%|██████▊   | 34/50 [00:13<00:04,  3.92it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.40it/s]
 74%|███████▍  | 37/50 [00:15<00:04,  2.75it/s]
 76%|███████▌  | 38/50 [00:16<00:06,  1.93it/s]
 78%|███████▊  | 39/50 [00:16<00:05,  1.98it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  1.83it/s]
 84%|████████▍ | 42/50 [00:18<00:04,  1.99it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.55it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.77it/s]
 90%|█████████ | 45/50 [00:20<00:03,  1.42it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.65it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.34it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.59it/s]
 98%|█████████▊| 49/50 [00:23<00:00,  1.31it/s]
+
+  clip 1/4: 38.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.40it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.40it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.49it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.67it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.63it/s]
 30%|███       | 15/50 [00:09<00:16,  2.13it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.23it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.88it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.70it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.70it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.36it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.53it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.57it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.29it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.51it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.40it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.28it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.03it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.13it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.93it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.01it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.54it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.68it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.41it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.17it/s]
+
+  clip 2/4: 41.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.39it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.34it/s]
 18%|█▊        | 9/50 [00:06<00:23,  1.78it/s]
 20%|██        | 10/50 [00:07<00:31,  1.27it/s]
 22%|██▏       | 11/50 [00:08<00:26,  1.46it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.48it/s]
 30%|███       | 15/50 [00:09<00:17,  2.02it/s]
 32%|███▏      | 16/50 [00:10<00:15,  2.14it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.27it/s]
 40%|████      | 20/50 [00:10<00:06,  4.37it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.35it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.40it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.61it/s]
 50%|█████     | 25/50 [00:12<00:06,  4.16it/s]
 52%|█████▏    | 26/50 [00:13<00:11,  2.15it/s]
 54%|█████▍    | 27/50 [00:13<00:10,  2.19it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.48it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.44it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.49it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.42it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.47it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.06it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.08it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.83it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.93it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.48it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.61it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.29it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.46it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.18it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.31it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.12it/s]
+
+  clip 3/4: 46.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.38it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.35it/s]
 10%|█         | 5/50 [00:04<00:41,  1.09it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.36it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.13it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 20%|██        | 10/50 [00:07<00:26,  1.48it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.68it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.66it/s]
 30%|███       | 15/50 [00:09<00:16,  2.09it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.09it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.00it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.52it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.39it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.45it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.18it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.40it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.43it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.61it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.54it/s]
 64%|██████▍   | 32/50 [00:15<00:06,  2.58it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.47it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.49it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.11it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.20it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.95it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.08it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.56it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.75it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.37it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.48it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.21it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.35it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.12it/s]
+
+  clip 4/4: 44.9s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000014/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000014/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 14071.2ms  total=  14.1s
+  clip_total         n=  4  mean= 42783.6ms  total= 171.1s
 
+Done in 185.3s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000015.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, an artist studio loft with a painting area, with bare white painted brick walls, white-painted concrete floor with faint paint splatters, and a wooden easel beside a tall stool catching bright north-facing daylight. In the foreground, there is a natural birch three-drawer dresser with minimalist black knobs, and a few jars of brushes and a folded paint-stained cloth rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=279  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.12s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.51it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.17it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.53it/s]
 10%|█         | 5/50 [00:03<00:36,  1.23it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.47it/s]
 14%|█▍        | 7/50 [00:05<00:35,  1.21it/s]
 16%|█▌        | 8/50 [00:05<00:28,  1.46it/s]
 20%|██        | 10/50 [00:07<00:25,  1.54it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.74it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.78it/s]
 30%|███       | 15/50 [00:08<00:16,  2.18it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.24it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.24it/s]
 40%|████      | 20/50 [00:09<00:06,  4.53it/s]
 42%|████▏     | 21/50 [00:10<00:11,  2.53it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.51it/s]
 46%|████▌     | 23/50 [00:11<00:08,  3.01it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.61it/s]
 50%|█████     | 25/50 [00:11<00:06,  4.04it/s]
 52%|█████▏    | 26/50 [00:12<00:11,  2.03it/s]
 54%|█████▍    | 27/50 [00:12<00:10,  2.15it/s]
 56%|█████▌    | 28/50 [00:13<00:07,  2.76it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.21it/s]
 62%|██████▏   | 31/50 [00:14<00:08,  2.22it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.30it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.46it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.05it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.08it/s]
 74%|███████▍  | 37/50 [00:16<00:04,  2.61it/s]
 76%|███████▌  | 38/50 [00:17<00:07,  1.71it/s]
 78%|███████▊  | 39/50 [00:18<00:06,  1.82it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.78it/s]
 84%|████████▍ | 42/50 [00:19<00:04,  1.96it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.51it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.73it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.38it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.62it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.32it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.54it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.28it/s]
+
+  clip 1/4: 40.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.41it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.32it/s]
 10%|█         | 5/50 [00:04<00:42,  1.06it/s]
 12%|█▏        | 6/50 [00:04<00:33,  1.30it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.08it/s]
 16%|█▌        | 8/50 [00:06<00:33,  1.24it/s]
 18%|█▊        | 9/50 [00:06<00:25,  1.62it/s]
 20%|██        | 10/50 [00:08<00:32,  1.22it/s]
 22%|██▏       | 11/50 [00:08<00:27,  1.44it/s]
 26%|██▌       | 13/50 [00:08<00:15,  2.42it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.98it/s]
 30%|███       | 15/50 [00:10<00:19,  1.77it/s]
 32%|███▏      | 16/50 [00:10<00:20,  1.70it/s]
 34%|███▍      | 17/50 [00:11<00:16,  2.01it/s]
 36%|███▌      | 18/50 [00:11<00:12,  2.52it/s]
 38%|███▊      | 19/50 [00:11<00:10,  3.08it/s]
 40%|████      | 20/50 [00:11<00:08,  3.46it/s]
 42%|████▏     | 21/50 [00:12<00:16,  1.72it/s]
 44%|████▍     | 22/50 [00:13<00:15,  1.78it/s]
 46%|████▌     | 23/50 [00:13<00:12,  2.19it/s]
 48%|████▊     | 24/50 [00:13<00:09,  2.73it/s]
 50%|█████     | 25/50 [00:13<00:07,  3.32it/s]
 52%|█████▏    | 26/50 [00:15<00:14,  1.69it/s]
 54%|█████▍    | 27/50 [00:15<00:13,  1.66it/s]
 56%|█████▌    | 28/50 [00:16<00:11,  1.94it/s]
 58%|█████▊    | 29/50 [00:16<00:09,  2.22it/s]
 60%|██████    | 30/50 [00:16<00:08,  2.29it/s]
 62%|██████▏   | 31/50 [00:18<00:13,  1.45it/s]
 64%|██████▍   | 32/50 [00:18<00:12,  1.50it/s]
 66%|██████▌   | 33/50 [00:18<00:08,  1.91it/s]
 68%|██████▊   | 34/50 [00:18<00:06,  2.45it/s]
 70%|███████   | 35/50 [00:20<00:09,  1.52it/s]
 72%|███████▏  | 36/50 [00:20<00:08,  1.67it/s]
 76%|███████▌  | 38/50 [00:21<00:07,  1.62it/s]
 78%|███████▊  | 39/50 [00:22<00:06,  1.73it/s]
 80%|████████  | 40/50 [00:22<00:04,  2.18it/s]
 82%|████████▏ | 41/50 [00:23<00:05,  1.52it/s]
 84%|████████▍ | 42/50 [00:24<00:04,  1.61it/s]
 86%|████████▌ | 43/50 [00:25<00:05,  1.26it/s]
 88%|████████▊ | 44/50 [00:25<00:04,  1.48it/s]
 90%|█████████ | 45/50 [00:27<00:04,  1.21it/s]
 92%|█████████▏| 46/50 [00:27<00:02,  1.40it/s]
 94%|█████████▍| 47/50 [00:28<00:02,  1.15it/s]
 96%|█████████▌| 48/50 [00:29<00:01,  1.37it/s]
 98%|█████████▊| 49/50 [00:30<00:00,  1.14it/s]
+
+  clip 2/4: 48.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:01,  1.25s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.37it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.40it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 20%|██        | 10/50 [00:07<00:27,  1.48it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.60it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.01it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.05it/s]
 30%|███       | 15/50 [00:09<00:18,  1.86it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.90it/s]
 34%|███▍      | 17/50 [00:10<00:14,  2.32it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.91it/s]
 40%|████      | 20/50 [00:10<00:06,  4.44it/s]
 42%|████▏     | 21/50 [00:11<00:13,  2.17it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.18it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.63it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.01it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.59it/s]
 52%|█████▏    | 26/50 [00:13<00:13,  1.83it/s]
 54%|█████▍    | 27/50 [00:14<00:11,  1.93it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.24it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.29it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.37it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.49it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.06it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.18it/s]
 74%|███████▍  | 37/50 [00:17<00:04,  2.70it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.70it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.82it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.71it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.84it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.41it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.61it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.28it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.51it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.23it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.46it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.20it/s]
+
+  clip 3/4: 48.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.37it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.34it/s]
 18%|█▊        | 9/50 [00:06<00:22,  1.83it/s]
 20%|██        | 10/50 [00:07<00:30,  1.32it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.51it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.52it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.03it/s]
 30%|███       | 15/50 [00:09<00:19,  1.78it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.92it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.47it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.78it/s]
 40%|████      | 20/50 [00:10<00:06,  4.42it/s]
 42%|████▏     | 21/50 [00:11<00:14,  2.06it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.13it/s]
 46%|████▌     | 23/50 [00:12<00:09,  2.70it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.36it/s]
 54%|█████▍    | 27/50 [00:13<00:10,  2.27it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.79it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.39it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.21it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.21it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  2.00it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.09it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.90it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.02it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.55it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.72it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.36it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.57it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.22it/s]
+
+  clip 4/4: 45.8s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000015/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000015/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13045.7ms  total=  13.0s
+  clip_total         n=  4  mean= 45858.2ms  total= 183.4s
 
+Done in 196.7s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000016.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a small residential laundry room with a folding area, with bright white walls and subway-tile wainscoting at the bottom, white hex porcelain tile floor with thin gray grout lines, and soft daylight filtering through a small window. In the foreground, there is a cream three-drawer dresser with woven rattan drawer fronts and small wooden knobs, and a wicker basket and a folded stack of white towels sit on a low bench beside it. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=301  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:53,  1.10s/it]
  4%|▍         | 2/50 [00:01<00:32,  1.48it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.15it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.51it/s]
 10%|█         | 5/50 [00:04<00:36,  1.23it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.53it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:06<00:24,  1.64it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.85it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.94it/s]
 30%|███       | 15/50 [00:08<00:14,  2.35it/s]
 32%|███▏      | 16/50 [00:08<00:13,  2.45it/s]
 36%|███▌      | 18/50 [00:08<00:08,  3.59it/s]
 40%|████      | 20/50 [00:09<00:06,  5.00it/s]
 44%|████▍     | 22/50 [00:10<00:10,  2.70it/s]
 46%|████▌     | 23/50 [00:10<00:08,  3.10it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.27it/s]
 52%|█████▏    | 26/50 [00:11<00:09,  2.45it/s]
 54%|█████▍    | 27/50 [00:12<00:09,  2.46it/s]
 58%|█████▊    | 29/50 [00:12<00:05,  3.52it/s]
 62%|██████▏   | 31/50 [00:13<00:07,  2.50it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.48it/s]
 66%|██████▌   | 33/50 [00:14<00:05,  2.97it/s]
 70%|███████   | 35/50 [00:15<00:06,  2.27it/s]
 72%|███████▏  | 36/50 [00:15<00:06,  2.30it/s]
 74%|███████▍  | 37/50 [00:15<00:04,  2.76it/s]
 76%|███████▌  | 38/50 [00:17<00:06,  1.80it/s]
 78%|███████▊  | 39/50 [00:17<00:05,  1.96it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  1.82it/s]
 84%|████████▍ | 42/50 [00:19<00:04,  1.94it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.50it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.65it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.53it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.52it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.27it/s]
+
+  clip 1/4: 44.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.15it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.67it/s]
 30%|███       | 15/50 [00:09<00:16,  2.16it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.06it/s]
 34%|███▍      | 17/50 [00:09<00:13,  2.45it/s]
 36%|███▌      | 18/50 [00:09<00:10,  3.01it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.64it/s]
 40%|████      | 20/50 [00:10<00:06,  4.30it/s]
 42%|████▏     | 21/50 [00:11<00:14,  1.99it/s]
 44%|████▍     | 22/50 [00:12<00:14,  1.88it/s]
 46%|████▌     | 23/50 [00:12<00:12,  2.14it/s]
 48%|████▊     | 24/50 [00:12<00:11,  2.28it/s]
 50%|█████     | 25/50 [00:12<00:09,  2.55it/s]
 52%|█████▏    | 26/50 [00:14<00:15,  1.50it/s]
 54%|█████▍    | 27/50 [00:14<00:14,  1.64it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.69it/s]
 62%|██████▏   | 31/50 [00:16<00:08,  2.14it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.15it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.26it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.03it/s]
 72%|███████▏  | 36/50 [00:18<00:06,  2.05it/s]
 74%|███████▍  | 37/50 [00:18<00:05,  2.54it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.66it/s]
 78%|███████▊  | 39/50 [00:20<00:06,  1.67it/s]
 80%|████████  | 40/50 [00:20<00:05,  1.94it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.35it/s]
 84%|████████▍ | 42/50 [00:22<00:05,  1.46it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.18it/s]
 88%|████████▊ | 44/50 [00:24<00:04,  1.34it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.11it/s]
 92%|█████████▏| 46/50 [00:25<00:03,  1.26it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.08it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.30it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.12it/s]
+
+  clip 2/4: 48.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 28%|██▊       | 14/50 [00:07<00:11,  3.20it/s]
 30%|███       | 15/50 [00:09<00:16,  2.06it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.18it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.84it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.68it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.61it/s]
 46%|████▌     | 23/50 [00:11<00:08,  3.06it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.50it/s]
 50%|█████     | 25/50 [00:11<00:06,  4.04it/s]
 52%|█████▏    | 26/50 [00:13<00:12,  1.91it/s]
 54%|█████▍    | 27/50 [00:13<00:12,  1.91it/s]
 56%|█████▌    | 28/50 [00:13<00:08,  2.45it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.08it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.16it/s]
 64%|██████▍   | 32/50 [00:15<00:08,  2.19it/s]
 66%|██████▌   | 33/50 [00:15<00:06,  2.72it/s]
 68%|██████▊   | 34/50 [00:15<00:05,  3.15it/s]
 70%|���██████   | 35/50 [00:17<00:08,  1.72it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.85it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.68it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.80it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.71it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.71it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.34it/s]
 88%|████████▊ | 44/50 [00:22<00:04,  1.39it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.15it/s]
 92%|█████████▏| 46/50 [00:24<00:03,  1.33it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.13it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.25it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.03it/s]
+
+  clip 3/4: 43.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.15it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.43it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.43it/s]
 20%|██        | 10/50 [00:07<00:26,  1.52it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.71it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.28it/s]
 30%|███       | 15/50 [00:09<00:16,  2.10it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.21it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.37it/s]
 40%|████      | 20/50 [00:09<00:06,  4.69it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.45it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.38it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.52it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.56it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.66it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.59it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.52it/s]
 66%|██████▌   | 33/50 [00:14<00:05,  2.95it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.32it/s]
 70%|███████   | 35/50 [00:16<00:08,  1.86it/s]
 72%|███████▏  | 36/50 [00:16<00:07,  1.96it/s]
 74%|███████▍  | 37/50 [00:16<00:05,  2.51it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.62it/s]
 78%|███████▊  | 39/50 [00:18<00:06,  1.76it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.31it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.52it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.66it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.28it/s]
 88%|████████▊ | 44/50 [00:21<00:04,  1.49it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.22it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.45it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.20it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.44it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.19it/s]
+
+  clip 4/4: 41.4s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000016/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000016/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 16020.2ms  total=  16.0s
+  clip_total         n=  4  mean= 44520.6ms  total= 178.1s
 
+Done in 194.3s

3drawers/gen_flex/chunk_001/_logs/b2_ep00000017.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a residential mudroom with an entryway area, with soft warm gray walls and white wainscot panels, light yellow pine plank floor slightly worn, and a row of black iron coat hooks holding a denim jacket on the wall. In the foreground, there is a warm taupe-painted three-drawer dresser with matte gold bar pulls, and a pair of leather boots and a folded scarf rest on a low wooden bench beside it. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=294  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.11s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.35it/s]
  6%|▌         | 3/50 [00:02<00:42,  1.10it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.40it/s]
 10%|█         | 5/50 [00:04<00:38,  1.17it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.34it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.10it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.29it/s]
 18%|█▊        | 9/50 [00:06<00:25,  1.63it/s]
 20%|██        | 10/50 [00:07<00:32,  1.25it/s]
 22%|██▏       | 11/50 [00:08<00:27,  1.42it/s]
 24%|██▍       | 12/50 [00:08<00:20,  1.88it/s]
 26%|██▌       | 13/50 [00:08<00:15,  2.40it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.91it/s]
 30%|███       | 15/50 [00:10<00:20,  1.70it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.89it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.46it/s]
 36%|███▌      | 18/50 [00:10<00:10,  2.96it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.60it/s]
 42%|████▏     | 21/50 [00:12<00:12,  2.31it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.34it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.57it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.52it/s]
 54%|█████▍    | 27/50 [00:14<00:08,  2.57it/s]
 58%|█████▊    | 29/50 [00:14<00:05,  3.75it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.66it/s]
 64%|██████▍   | 32/50 [00:15<00:06,  2.70it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.93it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.34it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.44it/s]
 74%|███████▍  | 37/50 [00:17<00:04,  2.98it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.79it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.99it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.88it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.95it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.50it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.66it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.54it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.52it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.28it/s]
+
+  clip 1/4: 43.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:39,  1.22it/s]
  6%|▌         | 3/50 [00:02<00:47,  1.01s/it]
  8%|▊         | 4/50 [00:03<00:35,  1.30it/s]
 10%|█         | 5/50 [00:04<00:41,  1.09it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.37it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.49it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.60it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.92it/s]
 26%|██▌       | 13/50 [00:08<00:15,  2.39it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.95it/s]
 30%|███       | 15/50 [00:09<00:20,  1.71it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.92it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.87it/s]
 42%|████▏     | 21/50 [00:11<00:11,  2.61it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.54it/s]
 46%|████▌     | 23/50 [00:12<00:09,  2.99it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.25it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.25it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.22it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.72it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.16it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.17it/s]
 64%|██████▍   | 32/50 [00:15<00:08,  2.24it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.51it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.05it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.08it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.86it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.96it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.79it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.88it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.44it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.61it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.28it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.50it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.45it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.20it/s]
+
+  clip 2/4: 47.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:42,  1.14it/s]
  6%|▌         | 3/50 [00:03<00:51,  1.10s/it]
  8%|▊         | 4/50 [00:03<00:39,  1.16it/s]
 10%|█         | 5/50 [00:04<00:44,  1.01it/s]
 12%|█▏        | 6/50 [00:05<00:34,  1.29it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.10it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.34it/s]
 20%|██        | 10/50 [00:08<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.66it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.19it/s]
 30%|███       | 15/50 [00:09<00:17,  2.05it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.10it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.08it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.62it/s]
 40%|████      | 20/50 [00:10<00:07,  4.11it/s]
 42%|████▏     | 21/50 [00:11<00:14,  2.04it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.18it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.35it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.32it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.30it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.69it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.21it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.21it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.21it/s]
 66%|██████▌   | 33/50 [00:16<00:06,  2.72it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.26it/s]
 70%|███���███   | 35/50 [00:17<00:08,  1.78it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.90it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.73it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.90it/s]
 82%|████████▏ | 41/50 [00:21<00:05,  1.78it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.93it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.47it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.66it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.54it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.21it/s]
+
+  clip 3/4: 45.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.65it/s]
 24%|██▍       | 12/50 [00:07<00:17,  2.11it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.67it/s]
 30%|███       | 15/50 [00:09<00:17,  2.02it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.15it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.31it/s]
 40%|████      | 20/50 [00:10<00:07,  4.22it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.24it/s]
 44%|████▍     | 22/50 [00:11<00:13,  2.09it/s]
 46%|████▌     | 23/50 [00:12<00:11,  2.45it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.00it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.20it/s]
 54%|█████▍    | 27/50 [00:13<00:10,  2.30it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.16it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.37it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.44it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.39it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.44it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.05it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.15it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.93it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.05it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.55it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.73it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.36it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.57it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.22it/s]
+
+  clip 4/4: 43.0s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000017/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b2_ep00000017/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 14531.9ms  total=  14.5s
+  clip_total         n=  4  mean= 44860.6ms  total= 179.4s
 
+Done in 194.1s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000000.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a baby nursery with a changing area, with soft white walls and a gentle pale gray painted arch, light yellow maple plank floor, and a knit mobile turning slowly above a crib in warm diffused light. In the foreground, there is a weathered driftwood-gray three-drawer dresser with rope pull handles, and a folded muslin blanket and a small woven basket of toys sit on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=278  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.10s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.45it/s]
  6%|▌         | 3/50 [00:02<00:41,  1.13it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.49it/s]
 10%|█         | 5/50 [00:04<00:36,  1.22it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.53it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:06<00:24,  1.64it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.84it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.53it/s]
 30%|███       | 15/50 [00:08<00:15,  2.26it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.39it/s]
 36%|███▌      | 18/50 [00:08<00:08,  3.61it/s]
 42%|████▏     | 21/50 [00:10<00:09,  3.05it/s]
 44%|████▍     | 22/50 [00:10<00:09,  3.03it/s]
 48%|████▊     | 24/50 [00:10<00:06,  4.23it/s]
 52%|█████▏    | 26/50 [00:11<00:08,  2.92it/s]
 54%|█████▍    | 27/50 [00:12<00:07,  2.93it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.77it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.72it/s]
 66%|██████▌   | 33/50 [00:13<00:05,  3.10it/s]
 70%|███████   | 35/50 [00:14<00:06,  2.47it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.55it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.20it/s]
 78%|███████▊  | 39/50 [00:16<00:04,  2.27it/s]
 80%|████████  | 40/50 [00:16<00:03,  2.72it/s]
 82%|████████▏ | 41/50 [00:17<00:04,  1.83it/s]
 84%|████████▍ | 42/50 [00:18<00:04,  1.96it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.49it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.71it/s]
 90%|█████████ | 45/50 [00:20<00:03,  1.37it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.62it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.33it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.58it/s]
 98%|█████████▊| 49/50 [00:23<00:00,  1.30it/s]
+
+  clip 1/4: 39.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.67it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.21it/s]
 30%|███       | 15/50 [00:09<00:16,  2.07it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.19it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.32it/s]
 40%|████      | 20/50 [00:09<00:06,  4.62it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.47it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.92it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.48it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.71it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.46it/s]
 66%|██████▌   | 33/50 [00:14<00:06,  2.77it/s]
 68%|██████▊   | 34/50 [00:14<00:05,  3.16it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.01it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.13it/s]
 76%|███████▌  | 38/50 [00:17<00:06,  1.90it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.04it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.86it/s]
 84%|████████▍ | 42/50 [00:19<00:04,  1.98it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.49it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.68it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.50it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.38it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.14it/s]
+
+  clip 2/4: 45.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.34it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.33it/s]
 10%|█         | 5/50 [00:04<00:41,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:34,  1.28it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.08it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.33it/s]
 20%|██        | 10/50 [00:07<00:27,  1.45it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.65it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.65it/s]
 30%|███       | 15/50 [00:09<00:16,  2.14it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.24it/s]
 38%|███▊      | 19/50 [00:10<00:07,  3.89it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.71it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.71it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.33it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.45it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.42it/s]
 56%|█████▌    | 28/50 [00:13<00:08,  2.72it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.80it/s]
 60%|██████    | 30/50 [00:14<00:07,  2.83it/s]
 62%|██████▏   | 31/50 [00:15<00:11,  1.65it/s]
 64%|██████▍   | 32/50 [00:16<00:10,  1.64it/s]
 66%|██████▌   | 33/50 [00:16<00:08,  1.92it/s]
 68%|██████▊   | 34/50 [00:16<00:07,  2.24it/s]
 70%|███████   | 35/50 [00:18<00:10,  1.46it/s]
 72%|███████▏  | 36/50 [00:18<00:08,  1.58it/s]
 74%|███████▍  | 37/50 [00:18<00:06,  2.02it/s]
 76%|███████▌  | 38/50 [00:20<00:08,  1.39it/s]
 78%|███████▊  | 39/50 [00:20<00:07,  1.49it/s]
 80%|████████  | 40/50 [00:20<00:05,  1.91it/s]
 82%|████████▏ | 41/50 [00:22<00:06,  1.36it/s]
 84%|████████▍ | 42/50 [00:22<00:05,  1.53it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.21it/s]
 88%|████████▊ | 44/50 [00:24<00:04,  1.39it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.14it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.37it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.15it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.39it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.16it/s]
+
+  clip 3/4: 51.5s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.40it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.40it/s]
 18%|█▊        | 9/50 [00:06<00:21,  1.87it/s]
 20%|██        | 10/50 [00:07<00:29,  1.35it/s]
 22%|██▏       | 11/50 [00:07<00:24,  1.60it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.79it/s]
 30%|███       | 15/50 [00:09<00:16,  2.10it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.13it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.46it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.80it/s]
 40%|████      | 20/50 [00:10<00:07,  4.07it/s]
 42%|████▏     | 21/50 [00:11<00:13,  2.11it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.08it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.52it/s]
 48%|████▊     | 24/50 [00:12<00:08,  2.93it/s]
 50%|█████     | 25/50 [00:12<00:07,  3.22it/s]
 52%|█████▏    | 26/50 [00:13<00:13,  1.76it/s]
 54%|█████▍    | 27/50 [00:14<00:12,  1.87it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.00it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.25it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.34it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.35it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.41it/s]
 76%|███████▌  | 38/50 [00:19<00:05,  2.07it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.17it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.92it/s]
 84%|████████▍ | 42/50 [00:21<00:03,  2.04it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.55it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.70it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.35it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.56it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.22it/s]
+
+  clip 4/4: 47.9s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000000/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000000/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 14118.5ms  total=  14.1s
+  clip_total         n=  4  mean= 46157.3ms  total= 184.6s
 
+Done in 198.9s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000001.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a tidy classroom with a learning area, with bright white walls and a long pale gray pinboard, light gray speckled vinyl floor, and rows of small desks under even fluorescent-style ceiling light. In the foreground, there is a classic white three-drawer dresser with faceted crystal glass knobs, and a globe and a neat cup of colored pencils are placed on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=263  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:55,  1.13s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.37it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.40it/s]
 10%|█         | 5/50 [00:04<00:38,  1.16it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.18it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 18%|█▊        | 9/50 [00:06<00:22,  1.84it/s]
 20%|██        | 10/50 [00:07<00:29,  1.37it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.63it/s]
 28%|██▊       | 14/50 [00:08<00:10,  3.40it/s]
 30%|███       | 15/50 [00:09<00:16,  2.15it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.28it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.44it/s]
 40%|████      | 20/50 [00:09<00:06,  4.75it/s]
 42%|████▏     | 21/50 [00:10<00:11,  2.53it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.59it/s]
 48%|████▊     | 24/50 [00:11<00:06,  3.90it/s]
 52%|█████▏    | 26/50 [00:12<00:08,  2.70it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.74it/s]
 58%|█████▊    | 29/50 [00:12<00:05,  4.02it/s]
 62%|██████▏   | 31/50 [00:14<00:06,  2.72it/s]
 64%|██████▍   | 32/50 [00:14<00:06,  2.75it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.99it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.43it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.16it/s]
 78%|███████▊  | 39/50 [00:17<00:04,  2.22it/s]
 80%|████████  | 40/50 [00:17<00:03,  2.63it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  1.82it/s]
 84%|████████▍ | 42/50 [00:19<00:04,  1.99it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.52it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.75it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.39it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.61it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.32it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.56it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.28it/s]
+
+  clip 1/4: 41.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:37,  1.26it/s]
  6%|▌         | 3/50 [00:02<00:47,  1.01s/it]
  8%|▊         | 4/50 [00:03<00:38,  1.21it/s]
 10%|█         | 5/50 [00:04<00:44,  1.01it/s]
 12%|█▏        | 6/50 [00:05<00:34,  1.26it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.08it/s]
 16%|█▌        | 8/50 [00:06<00:33,  1.27it/s]
 18%|█▊        | 9/50 [00:07<00:23,  1.73it/s]
 20%|██        | 10/50 [00:08<00:31,  1.28it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.53it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.19it/s]
 30%|███       | 15/50 [00:10<00:17,  1.98it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.12it/s]
 34%|███▍      | 17/50 [00:10<00:12,  2.62it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.04it/s]
 40%|████      | 20/50 [00:10<00:06,  4.70it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.27it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.37it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.32it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.38it/s]
 54%|█████▍    | 27/50 [00:14<00:09,  2.43it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.26it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.43it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.38it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.33it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.04it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.00it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.17it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.50it/s]
 84%|████████▍ | 42/50 [00:21<00:05,  1.53it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.22it/s]
 88%|████████▊ | 44/50 [00:23<00:04,  1.39it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.17it/s]
 92%|█████████▏| 46/50 [00:25<00:03,  1.33it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.13it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.32it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.11it/s]
+
+  clip 2/4: 48.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:37,  1.28it/s]
  6%|▌         | 3/50 [00:02<00:46,  1.01it/s]
  8%|▊         | 4/50 [00:03<00:35,  1.29it/s]
 10%|█         | 5/50 [00:04<00:42,  1.06it/s]
 12%|█▏        | 6/50 [00:05<00:37,  1.18it/s]
 14%|█▍        | 7/50 [00:06<00:42,  1.01it/s]
 16%|█▌        | 8/50 [00:07<00:34,  1.21it/s]
 18%|█▊        | 9/50 [00:07<00:24,  1.67it/s]
 20%|██        | 10/50 [00:08<00:31,  1.27it/s]
 22%|██▏       | 11/50 [00:08<00:27,  1.41it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.90it/s]
 26%|██▌       | 13/50 [00:09<00:14,  2.51it/s]
 28%|██▊       | 14/50 [00:09<00:11,  3.24it/s]
 30%|███       | 15/50 [00:10<00:20,  1.72it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.93it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.18it/s]
 40%|████      | 20/50 [00:11<00:06,  4.74it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.39it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.76it/s]
 52%|█████▏    | 26/50 [00:14<00:10,  2.34it/s]
 54%|█████▍    | 27/50 [00:14<00:09,  2.35it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.34it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.44it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.36it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.43it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.11it/s]
 72%|███████▏  | 36/50 [00:18<00:06,  2.11it/s]
 74%|███████▍  | 37/50 [00:18<00:05,  2.47it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.59it/s]
 78%|███████▊  | 39/50 [00:20<00:06,  1.71it/s]
 80%|████████  | 40/50 [00:20<00:04,  2.17it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.46it/s]
 84%|████████��� | 42/50 [00:21<00:04,  1.62it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.26it/s]
 88%|████████▊ | 44/50 [00:23<00:04,  1.50it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.20it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.41it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.17it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.27it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.05it/s]
+
+  clip 3/4: 48.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:37,  1.28it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.03it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.35it/s]
 10%|█         | 5/50 [00:04<00:40,  1.11it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.39it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:27,  1.46it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.66it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.21it/s]
 30%|███       | 15/50 [00:09<00:17,  2.06it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.17it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.26it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.74it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.72it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.36it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.55it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.58it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.63it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.56it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.54it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.53it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.10it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.11it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.84it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  1.96it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.77it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.82it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.40it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.54it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.26it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.34it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.10it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.32it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.13it/s]
+
+  clip 4/4: 45.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000001/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000001/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 12796.2ms  total=  12.8s
+  clip_total         n=  4  mean= 46149.5ms  total= 184.6s
 
+Done in 197.6s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000002.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a cozy cafe with a counter area, with warm white walls and a small chalkboard menu, light yellow oak plank floor, and a row of hanging Edison bulbs glowing over a wooden service counter. In the foreground, there is a cherry-wood three-drawer dresser with brushed-brass handles, and a ceramic mug and a small jar of sugar sit on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=264  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:53,  1.10s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.54it/s]
  6%|▌         | 3/50 [00:02<00:39,  1.18it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.54it/s]
 10%|█         | 5/50 [00:03<00:36,  1.23it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.53it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:06<00:24,  1.63it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.84it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.42it/s]
 30%|███       | 15/50 [00:08<00:15,  2.22it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.33it/s]
 36%|███▌      | 18/50 [00:08<00:09,  3.49it/s]
 40%|████      | 20/50 [00:09<00:06,  4.90it/s]
 44%|████▍     | 22/50 [00:10<00:10,  2.64it/s]
 48%|████▊     | 24/50 [00:10<00:07,  3.64it/s]
 52%|█████▏    | 26/50 [00:11<00:08,  2.72it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.72it/s]
 58%|█████▊    | 29/50 [00:12<00:05,  3.82it/s]
 62%|██████▏   | 31/50 [00:13<00:06,  2.76it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.76it/s]
 68%|██████▊   | 34/50 [00:13<00:04,  3.90it/s]
 70%|███████   | 35/50 [00:14<00:06,  2.36it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.46it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.11it/s]
 78%|███████▊  | 39/50 [00:16<00:04,  2.25it/s]
 82%|████████▏ | 41/50 [00:17<00:04,  2.02it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.11it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.61it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.76it/s]
 90%|█████████ | 45/50 [00:20<00:03,  1.41it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.55it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.47it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.23it/s]
+
+  clip 1/4: 46.1s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 20%|██        | 10/50 [00:07<00:27,  1.48it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.65it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.08it/s]
 30%|███       | 15/50 [00:09<00:17,  1.97it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.10it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.74it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.65it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.61it/s]
 46%|████▌     | 23/50 [00:11<00:09,  2.86it/s]
 48%|████▊     | 24/50 [00:11<00:08,  3.20it/s]
 50%|█████     | 25/50 [00:12<00:07,  3.45it/s]
 52%|█████▏    | 26/50 [00:13<00:12,  1.87it/s]
 54%|█████▍    | 27/50 [00:13<00:12,  1.90it/s]
 56%|█████▌    | 28/50 [00:14<00:09,  2.26it/s]
 58%|█████▊    | 29/50 [00:14<00:08,  2.62it/s]
 60%|██████    | 30/50 [00:14<00:06,  3.19it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.73it/s]
 64%|██████▍   | 32/50 [00:16<00:09,  1.83it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.36it/s]
 68%|██████▊   | 34/50 [00:16<00:05,  2.86it/s]
 70%|███████   | 35/50 [00:17<00:09,  1.66it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.78it/s]
 74%|███████▍  | 37/50 [00:18<00:05,  2.28it/s]
 76%|███████▌  | 38/50 [00:19<00:08,  1.49it/s]
 78%|███████▊  | 39/50 [00:20<00:06,  1.62it/s]
 80%|████████  | 40/50 [00:20<00:04,  2.10it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.43it/s]
 84%|████████▍ | 42/50 [00:21<00:05,  1.54it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.23it/s]
 88%|████████▊ | 44/50 [00:23<00:04,  1.46it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.20it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.42it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.19it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.38it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.15it/s]
+
+  clip 2/4: 49.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.39it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.40it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.40it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.57it/s]
 24%|██▍       | 12/50 [00:08<00:20,  1.87it/s]
 26%|██▌       | 13/50 [00:08<00:15,  2.37it/s]
 30%|███       | 15/50 [00:09<00:18,  1.91it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.91it/s]
 34%|███▍      | 17/50 [00:10<00:14,  2.24it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.71it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.34it/s]
 40%|████      | 20/50 [00:10<00:07,  4.03it/s]
 42%|████▏     | 21/50 [00:12<00:15,  1.86it/s]
 44%|████▍     | 22/50 [00:12<00:14,  1.99it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.37it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.38it/s]
 54%|█████▍    | 27/50 [00:14<00:09,  2.45it/s]
 58%|█████▊    | 29/50 [00:14<00:05,  3.68it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.55it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.54it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.45it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.48it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.06it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.17it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.90it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.98it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.50it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.69it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.35it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.50it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.41it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.17it/s]
+
+  clip 3/4: 47.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.15it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.43it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.44it/s]
 20%|██        | 10/50 [00:07<00:26,  1.52it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.68it/s]
 24%|██▍       | 12/50 [00:07<00:18,  2.09it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.49it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.12it/s]
 30%|███       | 15/50 [00:09<00:20,  1.72it/s]
 32%|███▏      | 16/50 [00:09<00:18,  1.88it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.40it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.63it/s]
 40%|████      | 20/50 [00:10<00:06,  4.30it/s]
 42%|████▏     | 21/50 [00:11<00:14,  2.03it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.13it/s]
 50%|█████     | 25/50 [00:12<00:06,  4.07it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.28it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.34it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.39it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.37it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.33it/s]
 66%|██████▌   | 33/50 [00:16<00:06,  2.46it/s]
 68%|██████▊   | 34/50 [00:16<00:06,  2.64it/s]
 70%|███████   | 35/50 [00:17<00:08,  1.67it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.87it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.76it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.92it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.79it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.94it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.48it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.68it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.56it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.39it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.16it/s]
+
+  clip 4/4: 41.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000002/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000002/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13512.4ms  total=  13.5s
+  clip_total         n=  4  mean= 46099.2ms  total= 184.4s
 
+Done in 198.1s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000003.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a medical clinic waiting area with a reception area, with clean white walls and a soft gray accent panel, light gray sealed-vinyl floor, and calm even light from a wide ceiling diffuser. In the foreground, there is a light gray three-drawer dresser with simple round white knobs, and a small potted plant and a neat fan of magazines rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=301  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.35it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.40it/s]
 10%|█         | 5/50 [00:04<00:38,  1.17it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.31it/s]
 18%|█▊        | 9/50 [00:06<00:24,  1.65it/s]
 20%|██        | 10/50 [00:07<00:31,  1.28it/s]
 22%|██▏       | 11/50 [00:08<00:28,  1.38it/s]
 24%|██▍       | 12/50 [00:08<00:23,  1.65it/s]
 26%|██▌       | 13/50 [00:09<00:18,  2.02it/s]
 28%|██▊       | 14/50 [00:09<00:14,  2.47it/s]
 30%|███       | 15/50 [00:10<00:22,  1.56it/s]
 32%|███▏      | 16/50 [00:10<00:20,  1.70it/s]
 34%|███▍      | 17/50 [00:11<00:15,  2.16it/s]
 38%|███▊      | 19/50 [00:11<00:08,  3.49it/s]
 40%|████      | 20/50 [00:11<00:07,  3.89it/s]
 42%|████▏     | 21/50 [00:12<00:14,  2.01it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.14it/s]
 46%|████▌     | 23/50 [00:13<00:10,  2.65it/s]
 50%|█████     | 25/50 [00:13<00:06,  3.99it/s]
 52%|█████▏    | 26/50 [00:14<00:11,  2.18it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.25it/s]
 56%|█████▌    | 28/50 [00:14<00:07,  2.84it/s]
 60%|██████    | 30/50 [00:15<00:04,  4.14it/s]
 62%|██████▏   | 31/50 [00:16<00:08,  2.21it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.33it/s]
 66%|██████▌   | 33/50 [00:16<00:05,  2.90it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.24it/s]
 72%|███████▏  | 36/50 [00:18<00:06,  2.27it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.96it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.11it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.94it/s]
 84%|████████▍ | 42/50 [00:21<00:03,  2.07it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.59it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.80it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.43it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.61it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.31it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.55it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.29it/s]
+
+  clip 1/4: 44.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.35it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.35it/s]
 10%|█         | 5/50 [00:04<00:40,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.36it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.13it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.30it/s]
 20%|██        | 10/50 [00:07<00:28,  1.40it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.54it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.96it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.03it/s]
 30%|███       | 15/50 [00:09<00:18,  1.87it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.91it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.01it/s]
 40%|████      | 20/50 [00:10<00:07,  3.94it/s]
 42%|████▏     | 21/50 [00:11<00:13,  2.17it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.10it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.50it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.00it/s]
 50%|█████     | 25/50 [00:12<00:07,  3.51it/s]
 52%|█████▏    | 26/50 [00:14<00:13,  1.83it/s]
 54%|█████▍    | 27/50 [00:14<00:12,  1.89it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.05it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.62it/s]
 62%|██████▏   | 31/50 [00:16<00:09,  1.90it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.00it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.37it/s]
 68%|██████▊   | 34/50 [00:17<00:06,  2.60it/s]
 70%|███████   | 35/50 [00:18<00:09,  1.57it/s]
 72%|███████▏  | 36/50 [00:18<00:08,  1.70it/s]
 74%|███████▍  | 37/50 [00:18<00:05,  2.20it/s]
 76%|███████▌  | 38/50 [00:20<00:08,  1.44it/s]
 78%|███████▊  | 39/50 [00:20<00:06,  1.59it/s]
 82%|████████▏ | 41/50 [00:21<00:05,  1.60it/s]
 84%|████████▍ | 42/50 [00:22<00:04,  1.71it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.30it/s]
 88%|████████▊ | 44/50 [00:24<00:04,  1.43it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.16it/s]
 92%|█████████▏| 46/50 [00:25<00:03,  1.33it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.12it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.34it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.14it/s]
+
+  clip 2/4: 46.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.35it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.32it/s]
 10%|█         | 5/50 [00:04<00:41,  1.09it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.37it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.32it/s]
 18%|█▊        | 9/50 [00:06<00:22,  1.81it/s]
 20%|██        | 10/50 [00:07<00:30,  1.33it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.50it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.01it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.52it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.21it/s]
 30%|███       | 15/50 [00:09<00:20,  1.71it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.86it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.45it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.06it/s]
 40%|████      | 20/50 [00:10<00:06,  4.66it/s]
 42%|████▏     | 21/50 [00:11<00:13,  2.20it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.31it/s]
 46%|████▌     | 23/50 [00:12<00:09,  2.90it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.59it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.24it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.18it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.56it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.08it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.72it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.81it/s]
 64%|██████▍   | 32/50 [00:16<00:09,  1.94it/s]
 66%|██████▌   | 33/50 [00:16<00:07,  2.40it/s]
 68%|██████▊   | 34/50 [00:16<00:06,  2.58it/s]
 70%|███████   | 35/50 [00:18<00:09,  1.52it/s]
 72%|███████▏  | 36/50 [00:18<00:08,  1.64it/s]
 74%|███████▍  | 37/50 [00:18<00:06,  2.03it/s]
 76%|███████▌  | 38/50 [00:20<00:08,  1.36it/s]
 78%|███████▊  | 39/50 [00:20<00:07,  1.53it/s]
 80%|████████  | 40/50 [00:20<00:04,  2.04it/s]
 82%|████████▏ | 41/50 [00:21<00:06,  1.41it/s]
 84%|████████▍ | 42/50 [00:22<00:04,  1.64it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.28it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.51it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.22it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.46it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.20it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.44it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.19it/s]
+
+  clip 3/4: 44.1s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.66it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.57it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.07it/s]
 30%|███       | 15/50 [00:09<00:19,  1.84it/s]
 32%|███▏      | 16/50 [00:09<00:17,  1.95it/s]
 34%|███▍      | 17/50 [00:09<00:13,  2.47it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.11it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.83it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.26it/s]
 44%|████▍     | 22/50 [00:11<00:12,  2.32it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.25it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.37it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.44it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.29it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.48it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.43it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.38it/s]
 74%|███████▍  | 37/50 [00:17<00:04,  2.76it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.85it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.00it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.83it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.97it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.49it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.69it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.55it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.21it/s]
+
+  clip 4/4: 42.4s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000003/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000003/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13506.5ms  total=  13.5s
+  clip_total         n=  4  mean= 44302.1ms  total= 177.2s
 
+Done in 190.9s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000004.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a modern furniture showroom with a display area, with clean white walls and recessed ceiling spotlights washing them evenly, light gray sealed concrete floor in a smooth wide pour, and a single linen-upholstered armchair set back as a display piece. In the foreground, there is a smooth matte white three-drawer dresser with brushed-steel bar handles, and a small folded brochure and a square ceramic vase are placed on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=268  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:55,  1.13s/it]
  4%|▍         | 2/50 [00:01<00:32,  1.46it/s]
  6%|▌         | 3/50 [00:02<00:41,  1.15it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.51it/s]
 10%|█         | 5/50 [00:04<00:36,  1.23it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.53it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:06<00:24,  1.64it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.85it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.53it/s]
 30%|███       | 15/50 [00:08<00:15,  2.26it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.38it/s]
 38%|███▊      | 19/50 [00:08<00:07,  4.21it/s]
 42%|████▏     | 21/50 [00:10<00:09,  2.92it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.93it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.73it/s]
 54%|█████▍    | 27/50 [00:11<00:08,  2.78it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.19it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.73it/s]
 70%|███████   | 35/50 [00:14<00:05,  2.65it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.69it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.32it/s]
 78%|███████▊  | 39/50 [00:16<00:04,  2.41it/s]
 82%|████████▏ | 41/50 [00:17<00:04,  2.15it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.27it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.73it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.91it/s]
 90%|█████████ | 45/50 [00:20<00:03,  1.50it/s]
 92%|█████████▏| 46/50 [00:20<00:02,  1.72it/s]
 94%|█████████▍| 47/50 [00:21<00:02,  1.39it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.63it/s]
 98%|█████████▊| 49/50 [00:23<00:00,  1.33it/s]
+
+  clip 1/4: 38.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:38,  1.26it/s]
  6%|▌         | 3/50 [00:02<00:47,  1.00s/it]
  8%|▊         | 4/50 [00:03<00:35,  1.29it/s]
 10%|█         | 5/50 [00:04<00:41,  1.08it/s]
 12%|█▏        | 6/50 [00:05<00:33,  1.32it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.11it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 20%|██        | 10/50 [00:07<00:26,  1.48it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.68it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.73it/s]
 30%|███       | 15/50 [00:09<00:16,  2.18it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.25it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.26it/s]
 40%|████      | 20/50 [00:10<00:06,  4.43it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.36it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.41it/s]
 46%|████▌     | 23/50 [00:11<00:09,  2.94it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.56it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.28it/s]
 52%|█████▏    | 26/50 [00:13<00:12,  1.97it/s]
 54%|█████▍    | 27/50 [00:13<00:11,  2.04it/s]
 56%|█████▌    | 28/50 [00:13<00:08,  2.63it/s]
 58%|█████▊    | 29/50 [00:14<00:07,  2.97it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.56it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.81it/s]
 64%|██████▍   | 32/50 [00:15<00:09,  1.93it/s]
 66%|██████▌   | 33/50 [00:16<00:06,  2.45it/s]
 68%|██████▊   | 34/50 [00:16<00:05,  2.95it/s]
 70%|███████   | 35/50 [00:17<00:09,  1.66it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.84it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.74it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.91it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.78it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.87it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.42it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.56it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.26it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.43it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.18it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.41it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.18it/s]
+
+  clip 2/4: 46.1s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.34it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.05it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.36it/s]
 10%|█         | 5/50 [00:04<00:40,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.37it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.69it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.24it/s]
 30%|███       | 15/50 [00:09<00:16,  2.07it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.18it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.26it/s]
 40%|████      | 20/50 [00:09<00:06,  4.32it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.34it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.38it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.56it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.50it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.54it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.73it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.59it/s]
 64%|██████▍   | 32/50 [00:15<00:06,  2.61it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.49it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.43it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.04it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.08it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.84it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.90it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.45it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.63it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.32it/s]
 92%|█████████���| 46/50 [00:23<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.41it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.17it/s]
+
+  clip 3/4: 46.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.20s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.31it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.04it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.34it/s]
 10%|█         | 5/50 [00:04<00:40,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.37it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.39it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.69it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.23it/s]
 30%|███       | 15/50 [00:09<00:16,  2.07it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.16it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.84it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.69it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.69it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.37it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.57it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.47it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.59it/s]
 64%|██████▍   | 32/50 [00:14<00:06,  2.62it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.50it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.47it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.09it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.15it/s]
 80%|████████  | 40/50 [00:18<00:03,  2.55it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.69it/s]
 84%|████████▍ | 42/50 [00:19<00:04,  1.87it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.41it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.62it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.28it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.49it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.45it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.20it/s]
+
+  clip 4/4: 47.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000004/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000004/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 12206.1ms  total=  12.2s
+  clip_total         n=  4  mean= 44870.6ms  total= 179.5s
 
+Done in 192.0s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000005.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a bright sunroom with a relaxing area, with white-painted walls and white window mullions, light yellow pine plank floor, and a wall of tall windows pouring warm sunlight over a few leafy potted plants. In the foreground, there is a natural rattan-and-wood three-drawer dresser with small wooden knobs, and a watering can and a small terracotta pot sit on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=269  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.12s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.34it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.39it/s]
 10%|█         | 5/50 [00:04<00:38,  1.16it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.45it/s]
 14%|█▍        | 7/50 [00:05<00:35,  1.22it/s]
 16%|█▌        | 8/50 [00:06<00:28,  1.48it/s]
 18%|█▊        | 9/50 [00:06<00:20,  1.96it/s]
 20%|██        | 10/50 [00:07<00:28,  1.42it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.65it/s]
 24%|██▍       | 12/50 [00:07<00:17,  2.17it/s]
 28%|██▊       | 14/50 [00:08<00:10,  3.51it/s]
 30%|███       | 15/50 [00:09<00:17,  2.04it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.12it/s]
 34%|███▍      | 17/50 [00:09<00:12,  2.65it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.80it/s]
 40%|████      | 20/50 [00:10<00:06,  4.43it/s]
 42%|████▏     | 21/50 [00:11<00:13,  2.22it/s]
 44%|████▍     | 22/50 [00:11<00:12,  2.30it/s]
 46%|████▌     | 23/50 [00:11<00:09,  2.87it/s]
 50%|█████     | 25/50 [00:11<00:06,  4.07it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.18it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.32it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.63it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.51it/s]
 64%|██████▍   | 32/50 [00:15<00:06,  2.59it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.81it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.24it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.36it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.07it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.21it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  2.01it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.16it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.63it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.78it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.41it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.61it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.31it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.43it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.19it/s]
+
+  clip 1/4: 40.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.32it/s]
 10%|█         | 5/50 [00:04<00:42,  1.07it/s]
 12%|█▏        | 6/50 [00:04<00:34,  1.27it/s]
 14%|█▍        | 7/50 [00:06<00:40,  1.07it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.29it/s]
 20%|██        | 10/50 [00:07<00:28,  1.38it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.53it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.94it/s]
 26%|██▌       | 13/50 [00:08<00:15,  2.37it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.01it/s]
 30%|███       | 15/50 [00:10<00:20,  1.70it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.86it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.07it/s]
 40%|████      | 20/50 [00:10<00:07,  4.11it/s]
 42%|████▏     | 21/50 [00:12<00:13,  2.14it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.26it/s]
 50%|█████     | 25/50 [00:12<00:06,  4.12it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.35it/s]
 54%|█████▍    | 27/50 [00:14<00:09,  2.32it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.32it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.36it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.36it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.27it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.03it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.12it/s]
 74%|███████▍  | 37/50 [00:18<00:04,  2.62it/s]
 76%|███████▌  | 38/50 [00:19<00:07,  1.67it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.81it/s]
 82%|████████▏ | 41/50 [00:21<00:05,  1.67it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.80it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.38it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.58it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.27it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.49it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.46it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.20it/s]
+
+  clip 2/4: 44.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:33,  1.30it/s]
 14%|█▍        | 7/50 [00:06<00:40,  1.07it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.30it/s]
 20%|██        | 10/50 [00:07<00:28,  1.42it/s]
 22%|██▏       | 11/50 [00:08<00:26,  1.45it/s]
 24%|██▍       | 12/50 [00:08<00:21,  1.78it/s]
 26%|██▌       | 13/50 [00:08<00:17,  2.06it/s]
 28%|██▊       | 14/50 [00:09<00:15,  2.39it/s]
 30%|███       | 15/50 [00:10<00:22,  1.52it/s]
 32%|███▏      | 16/50 [00:11<00:21,  1.58it/s]
 34%|███▍      | 17/50 [00:11<00:16,  2.01it/s]
 36%|███▌      | 18/50 [00:11<00:12,  2.50it/s]
 38%|███▊      | 19/50 [00:11<00:10,  3.05it/s]
 40%|████      | 20/50 [00:11<00:08,  3.41it/s]
 42%|████▏     | 21/50 [00:12<00:16,  1.73it/s]
 44%|████▍     | 22/50 [00:13<00:14,  1.88it/s]
 48%|████▊     | 24/50 [00:13<00:08,  3.01it/s]
 50%|█████     | 25/50 [00:13<00:07,  3.47it/s]
 52%|█████▏    | 26/50 [00:15<00:13,  1.84it/s]
 54%|█████▍    | 27/50 [00:15<00:11,  1.94it/s]
 56%|█████▌    | 28/50 [00:15<00:08,  2.49it/s]
 58%|█████▊    | 29/50 [00:15<00:07,  2.96it/s]
 60%|██████    | 30/50 [00:15<00:05,  3.34it/s]
 62%|██████▏   | 31/50 [00:17<00:11,  1.72it/s]
 64%|██���███▍   | 32/50 [00:17<00:10,  1.79it/s]
 66%|██████▌   | 33/50 [00:17<00:08,  2.11it/s]
 68%|██████▊   | 34/50 [00:18<00:06,  2.55it/s]
 70%|███████   | 35/50 [00:19<00:09,  1.57it/s]
 72%|███████▏  | 36/50 [00:19<00:08,  1.71it/s]
 74%|███████▍  | 37/50 [00:20<00:06,  2.11it/s]
 76%|███████▌  | 38/50 [00:21<00:08,  1.42it/s]
 78%|███████▊  | 39/50 [00:21<00:06,  1.59it/s]
 80%|████████  | 40/50 [00:21<00:04,  2.10it/s]
 82%|████████▏ | 41/50 [00:23<00:06,  1.42it/s]
 84%|████████▍ | 42/50 [00:23<00:04,  1.63it/s]
 86%|████████▌ | 43/50 [00:24<00:05,  1.26it/s]
 88%|████████▊ | 44/50 [00:25<00:04,  1.47it/s]
 90%|█████████ | 45/50 [00:26<00:04,  1.19it/s]
 92%|█████████▏| 46/50 [00:26<00:02,  1.38it/s]
 94%|█████████▍| 47/50 [00:28<00:02,  1.15it/s]
 96%|█████████▌| 48/50 [00:28<00:01,  1.35it/s]
 98%|█████████▊| 49/50 [00:29<00:00,  1.12it/s]
+
+  clip 3/4: 46.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.37it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.05it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.35it/s]
 10%|█         | 5/50 [00:04<00:40,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.37it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.13it/s]
 16%|█▌        | 8/50 [00:06<00:33,  1.27it/s]
 18%|█▊        | 9/50 [00:06<00:24,  1.66it/s]
 20%|██        | 10/50 [00:08<00:32,  1.24it/s]
 22%|██▏       | 11/50 [00:08<00:28,  1.38it/s]
 24%|██▍       | 12/50 [00:08<00:21,  1.78it/s]
 26%|██▌       | 13/50 [00:09<00:17,  2.13it/s]
 28%|██▊       | 14/50 [00:09<00:13,  2.60it/s]
 30%|███       | 15/50 [00:10<00:22,  1.56it/s]
 32%|███▏      | 16/50 [00:10<00:20,  1.65it/s]
 34%|███▍      | 17/50 [00:11<00:15,  2.16it/s]
 38%|███▊      | 19/50 [00:11<00:08,  3.59it/s]
 40%|████      | 20/50 [00:11<00:07,  4.24it/s]
 42%|████▏     | 21/50 [00:12<00:14,  2.01it/s]
 44%|████▍     | 22/50 [00:13<00:14,  2.00it/s]
 48%|████▊     | 24/50 [00:13<00:07,  3.29it/s]
 52%|█████▏    | 26/50 [00:14<00:10,  2.36it/s]
 54%|█████▍    | 27/50 [00:14<00:09,  2.43it/s]
 58%|█████▊    | 29/50 [00:14<00:05,  3.64it/s]
 62%|██████▏   | 31/50 [00:16<00:07,  2.52it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.56it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.67it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.15it/s]
 72%|███████▏  | 36/50 [00:18<00:06,  2.10it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.82it/s]
 78%|███████▊  | 39/50 [00:20<00:05,  1.89it/s]
 80%|████████  | 40/50 [00:20<00:04,  2.30it/s]
 82%|████████▏ | 41/50 [00:21<00:05,  1.59it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.79it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.36it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.59it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.27it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.50it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.21it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.41it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.17it/s]
+
+  clip 4/4: 43.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000005/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000005/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 11650.2ms  total=  11.7s
+  clip_total         n=  4  mean= 43794.9ms  total= 175.2s
 
+Done in 187.0s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000006.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a walk-in dressing room with a wardrobe area, with soft white walls and slim gray trim, white-painted herringbone wood floor, and a tall mirror reflecting a row of neatly hung clothes under warm rail lighting. In the foreground, there is a soft greige-painted three-drawer dresser with polished brass knobs, and a folded silk scarf and a small jewelry tray rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=276  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.10s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.52it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.17it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.53it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.54it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:06<00:24,  1.64it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.76it/s]
 24%|██▍       | 12/50 [00:07<00:16,  2.24it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.85it/s]
 30%|███       | 15/50 [00:08<00:15,  2.23it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.22it/s]
 34%|███▍      | 17/50 [00:09<00:12,  2.71it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.32it/s]
 40%|████      | 20/50 [00:09<00:05,  5.01it/s]
 42%|████▏     | 21/50 [00:10<00:11,  2.45it/s]
 44%|████▍     | 22/50 [00:10<00:11,  2.53it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.52it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.75it/s]
 52%|█████▏    | 26/50 [00:12<00:11,  2.04it/s]
 54%|█████▍    | 27/50 [00:12<00:10,  2.20it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.38it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.64it/s]
 62%|██████▏   | 31/50 [00:14<00:09,  2.00it/s]
 64%|██████▍   | 32/50 [00:14<00:08,  2.12it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.33it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.06it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.22it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.01it/s]
 78%|███████▊  | 39/50 [00:17<00:05,  2.16it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.99it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.08it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.58it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.70it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.35it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.48it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.45it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.23it/s]
+
+  clip 1/4: 46.1s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.30it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.03it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.33it/s]
 10%|█         | 5/50 [00:04<00:41,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.34it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.12it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.33it/s]
 20%|██        | 10/50 [00:07<00:27,  1.45it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.61it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.05it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.21it/s]
 30%|███       | 15/50 [00:09<00:18,  1.94it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.02it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.17it/s]
 40%|████      | 20/50 [00:10<00:06,  4.53it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.43it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.92it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.48it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.75it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.49it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.44it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.43it/s]
 74%|███████▍  | 37/50 [00:17<00:04,  2.74it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.87it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  1.96it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.79it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.88it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.45it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.60it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.27it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.49it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.23it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.43it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.19it/s]
+
+  clip 2/4: 47.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:01,  1.25s/it]
  4%|▍         | 2/50 [00:01<00:38,  1.24it/s]
  6%|▌         | 3/50 [00:02<00:47,  1.01s/it]
  8%|▊         | 4/50 [00:03<00:34,  1.32it/s]
 10%|█         | 5/50 [00:04<00:40,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.36it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.13it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.38it/s]
 20%|██        | 10/50 [00:07<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.65it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.61it/s]
 30%|███       | 15/50 [00:09<00:16,  2.10it/s]
 32%|███▏      | 16/50 [00:10<00:15,  2.13it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.11it/s]
 40%|████      | 20/50 [00:10<00:06,  4.30it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.38it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.45it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.32it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.48it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.40it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.55it/s]
 64%|██████▍   | 32/50 [00:15<00:06,  2.58it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.48it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.43it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.03it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.06it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.88it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.86it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.44it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.63it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.53it/s]
 94%|█████████��| 47/50 [00:24<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.22it/s]
+
+  clip 3/4: 46.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.35it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.05it/s]
  8%|▊         | 4/50 [00:03<00:35,  1.31it/s]
 10%|█         | 5/50 [00:04<00:41,  1.09it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.37it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.25it/s]
 30%|███       | 15/50 [00:09<00:16,  2.08it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.20it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.86it/s]
 40%|████      | 20/50 [00:09<00:07,  4.23it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.30it/s]
 44%|████▍     | 22/50 [00:11<00:12,  2.31it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.55it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.49it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.54it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.32it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.52it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.43it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.21it/s]
 76%|███████▌  | 38/50 [00:17<00:06,  1.95it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  1.93it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.17it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.59it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.77it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.38it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.60it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.29it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.51it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.23it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.47it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.21it/s]
+
+  clip 4/4: 40.7s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000006/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000006/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13819.9ms  total=  13.8s
+  clip_total         n=  4  mean= 45012.0ms  total= 180.0s
 
+Done in 194.0s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000007.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a quiet reading nook with a window area, with warm gray walls and a white window seat, light yellow oak plank floor, and a soft pool of lamplight beside a cushioned bench under a tall bookshelf. In the foreground, there is an olive-green painted three-drawer dresser with aged-brass knobs, and an open paperback and a steaming mug of tea sit on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=279  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:55,  1.14s/it]
  4%|▍         | 2/50 [00:01<00:32,  1.50it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.17it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.53it/s]
 10%|█         | 5/50 [00:03<00:36,  1.23it/s]
 12%|█▏        | 6/50 [00:04<00:29,  1.47it/s]
 14%|█▍        | 7/50 [00:05<00:35,  1.23it/s]
 16%|█▌        | 8/50 [00:05<00:29,  1.44it/s]
 18%|█▊        | 9/50 [00:06<00:21,  1.93it/s]
 20%|██        | 10/50 [00:07<00:28,  1.41it/s]
 22%|██▏       | 11/50 [00:07<00:24,  1.60it/s]
 24%|██▍       | 12/50 [00:07<00:18,  2.08it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.67it/s]
 30%|███       | 15/50 [00:09<00:16,  2.09it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.19it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.28it/s]
 40%|████      | 20/50 [00:09<00:06,  4.41it/s]
 42%|████▏     | 21/50 [00:10<00:11,  2.43it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.52it/s]
 48%|████▊     | 24/50 [00:11<00:06,  3.85it/s]
 52%|█████▏    | 26/50 [00:12<00:08,  2.67it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.71it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.58it/s]
 64%|██████▍   | 32/50 [00:14<00:06,  2.69it/s]
 70%|███████   | 35/50 [00:15<00:05,  2.63it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.54it/s]
 74%|███████▍  | 37/50 [00:16<00:04,  2.87it/s]
 76%|███████▌  | 38/50 [00:17<00:06,  1.94it/s]
 78%|███████▊  | 39/50 [00:17<00:05,  2.00it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.40it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.60it/s]
 84%|████████▍ | 42/50 [00:19<00:04,  1.77it/s]
 86%|████████▌ | 43/50 [00:20<00:05,  1.38it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.50it/s]
 90%|█████████ | 45/50 [00:22<00:04,  1.24it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.39it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.16it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.36it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.18it/s]
+
+  clip 1/4: 42.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.39it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.37it/s]
 14%|█▍        | 7/50 [00:05<00:38,  1.12it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.33it/s]
 18%|█▊        | 9/50 [00:06<00:22,  1.81it/s]
 20%|██        | 10/50 [00:07<00:30,  1.31it/s]
 22%|██▏       | 11/50 [00:08<00:27,  1.44it/s]
 24%|██▍       | 12/50 [00:08<00:21,  1.76it/s]
 26%|██▌       | 13/50 [00:08<00:18,  2.02it/s]
 28%|██▊       | 14/50 [00:09<00:14,  2.44it/s]
 30%|███       | 15/50 [00:10<00:23,  1.51it/s]
 32%|███▏      | 16/50 [00:10<00:21,  1.62it/s]
 34%|███▍      | 17/50 [00:11<00:17,  1.84it/s]
 36%|███▌      | 18/50 [00:11<00:16,  1.99it/s]
 38%|███▊      | 19/50 [00:12<00:15,  1.94it/s]
 40%|████      | 20/50 [00:12<00:12,  2.47it/s]
 42%|████▏     | 21/50 [00:13<00:18,  1.56it/s]
 44%|████▍     | 22/50 [00:13<00:15,  1.77it/s]
 48%|████▊     | 24/50 [00:14<00:08,  2.99it/s]
 52%|█████▏    | 26/50 [00:15<00:10,  2.20it/s]
 54%|█████▍    | 27/50 [00:15<00:10,  2.28it/s]
 58%|█████▊    | 29/50 [00:15<00:06,  3.46it/s]
 62%|██████▏   | 31/50 [00:17<00:07,  2.47it/s]
 64%|██████▍   | 32/50 [00:17<00:07,  2.51it/s]
 70%|███████   | 35/50 [00:18<00:06,  2.42it/s]
 72%|███████▏  | 36/50 [00:19<00:05,  2.46it/s]
 76%|███████▌  | 38/50 [00:20<00:05,  2.10it/s]
 78%|███████▊  | 39/50 [00:20<00:05,  2.12it/s]
 80%|████████  | 40/50 [00:21<00:04,  2.43it/s]
 82%|████████▏ | 41/50 [00:22<00:05,  1.63it/s]
 84%|████████▍ | 42/50 [00:22<00:04,  1.74it/s]
 86%|████████▌ | 43/50 [00:23<00:05,  1.33it/s]
 88%|████████▊ | 44/50 [00:24<00:04,  1.36it/s]
 90%|█████████ | 45/50 [00:25<00:04,  1.11it/s]
 92%|█████████▏| 46/50 [00:26<00:03,  1.25it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.08it/s]
 96%|█████████▌| 48/50 [00:28<00:01,  1.29it/s]
 98%|█████████▊| 49/50 [00:29<00:00,  1.10it/s]
+
+  clip 2/4: 46.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.36it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.39it/s]
 10%|█         | 5/50 [00:04<00:40,  1.12it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.40it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.15it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.37it/s]
 18%|█▊        | 9/50 [00:06<00:21,  1.88it/s]
 20%|██        | 10/50 [00:07<00:29,  1.34it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.56it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.08it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.70it/s]
 28%|██▊       | 14/50 [00:08<00:10,  3.42it/s]
 30%|███       | 15/50 [00:09<00:19,  1.77it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.88it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.01it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.62it/s]
 40%|████      | 20/50 [00:10<00:07,  4.18it/s]
 42%|████▏     | 21/50 [00:11<00:14,  2.02it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.17it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.25it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.32it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.39it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.24it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.40it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.46it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.40it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.03it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.13it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.86it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.97it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.80it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.92it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.47it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.57it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.25it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.37it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.13it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.28it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.08it/s]
+
+  clip 3/4: 43.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.34it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.04it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.34it/s]
 10%|█         | 5/50 [00:04<00:40,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.61it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.05it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.99it/s]
 30%|███       | 15/50 [00:09<00:19,  1.80it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.89it/s]
 34%|███▍      | 17/50 [00:10<00:14,  2.36it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.76it/s]
 40%|████      | 20/50 [00:10<00:06,  4.33it/s]
 42%|████▏     | 21/50 [00:11<00:13,  2.18it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.29it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.28it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.37it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.44it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.27it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.39it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.23it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.13it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.24it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.96it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.08it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.88it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.01it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.52it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.71it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.23it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.40it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.16it/s]
+
+  clip 4/4: 40.8s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000007/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000007/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 14093.6ms  total=  14.1s
+  clip_total         n=  4  mean= 43260.8ms  total= 173.0s
 
+Done in 187.3s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000008.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a photography studio with a shooting area, with seamless white walls and a white sweep backdrop, light gray smooth concrete floor, and a softbox light stand glowing evenly to one side. In the foreground, there is an industrial-style three-drawer dresser with a metal frame and reclaimed-wood drawer fronts and black pipe handles, and a folded gray reflector and a small clapperboard rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=253  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.10s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.53it/s]
  6%|▌         | 3/50 [00:02<00:39,  1.18it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.53it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.54it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.55it/s]
 20%|██        | 10/50 [00:06<00:24,  1.65it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.85it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.53it/s]
 30%|███       | 15/50 [00:08<00:15,  2.26it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.38it/s]
 38%|███▊      | 19/50 [00:08<00:07,  4.18it/s]
 42%|████▏     | 21/50 [00:10<00:09,  2.92it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.92it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.72it/s]
 54%|█████▍    | 27/50 [00:11<00:08,  2.77it/s]
 58%|█████▊    | 29/50 [00:12<00:05,  3.72it/s]
 62%|██████▏   | 31/50 [00:13<00:06,  2.80it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.82it/s]
 70%|███████   | 35/50 [00:14<00:05,  2.69it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.73it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.31it/s]
 78%|███████▊  | 39/50 [00:16<00:04,  2.40it/s]
 82%|████████▏ | 41/50 [00:17<00:04,  2.12it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.25it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.70it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.89it/s]
 90%|█████████ | 45/50 [00:20<00:03,  1.49it/s]
 92%|█████████▏| 46/50 [00:20<00:02,  1.71it/s]
 94%|█████████▍| 47/50 [00:21<00:02,  1.38it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.62it/s]
 98%|█████████▊| 49/50 [00:23<00:00,  1.33it/s]
+
+  clip 1/4: 38.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.15it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.43it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:24,  1.57it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.97it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.52it/s]
 30%|███       | 15/50 [00:09<00:17,  1.99it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.02it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.04it/s]
 40%|████      | 20/50 [00:10<00:07,  4.20it/s]
 42%|████▏     | 21/50 [00:11<00:12,  2.29it/s]
 44%|████▍     | 22/50 [00:11<00:12,  2.24it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.68it/s]
 48%|████▊     | 24/50 [00:12<00:09,  2.82it/s]
 50%|█████     | 25/50 [00:12<00:09,  2.74it/s]
 52%|█████▏    | 26/50 [00:14<00:14,  1.65it/s]
 54%|█████▍    | 27/50 [00:14<00:13,  1.73it/s]
 56%|█████▌    | 28/50 [00:14<00:09,  2.24it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.66it/s]
 62%|██████▏   | 31/50 [00:15<00:09,  2.02it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.17it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.48it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.02it/s]
 72%|███████▏  | 36/50 [00:18<00:06,  2.13it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.86it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.91it/s]
 82%|████████▏ | 41/50 [00:21<00:05,  1.77it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.92it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.48it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.67it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.50it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.21it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.40it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.17it/s]
+
+  clip 2/4: 46.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.20s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.41it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.35it/s]
 10%|█         | 5/50 [00:04<00:40,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:34,  1.29it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.09it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.31it/s]
 18%|█▊        | 9/50 [00:06<00:23,  1.77it/s]
 20%|██        | 10/50 [00:07<00:30,  1.30it/s]
 22%|██▏       | 11/50 [00:08<00:26,  1.47it/s]
 24%|██▍       | 12/50 [00:08<00:19,  1.90it/s]
 26%|██▌       | 13/50 [00:08<00:15,  2.46it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.04it/s]
 30%|███       | 15/50 [00:10<00:20,  1.67it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.80it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.36it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.90it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.43it/s]
 40%|████      | 20/50 [00:11<00:08,  3.61it/s]
 42%|████▏     | 21/50 [00:12<00:17,  1.70it/s]
 44%|████▍     | 22/50 [00:12<00:15,  1.84it/s]
 46%|████▌     | 23/50 [00:13<00:11,  2.41it/s]
 48%|████▊     | 24/50 [00:13<00:08,  3.00it/s]
 50%|█████     | 25/50 [00:13<00:07,  3.40it/s]
 52%|█████▏    | 26/50 [00:14<00:13,  1.73it/s]
 54%|█████▍    | 27/50 [00:15<00:11,  1.94it/s]
 58%|█████▊    | 29/50 [00:15<00:06,  3.11it/s]
 60%|██████    | 30/50 [00:15<00:05,  3.62it/s]
 62%|██████▏   | 31/50 [00:16<00:09,  1.93it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.09it/s]
 70%|███████   | 35/50 [00:18<00:06,  2.21it/s]
 72%|███████▏  | 36/50 [00:18<00:06,  2.27it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.99it/s]
 78%|███���███▊  | 39/50 [00:20<00:05,  2.09it/s]
 82%|████████▏ | 41/50 [00:21<00:04,  1.86it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.99it/s]
 86%|████████▌ | 43/50 [00:23<00:04,  1.52it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.70it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.56it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.48it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.21it/s]
+
+  clip 3/4: 47.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.24s/it]
  4%|▍         | 2/50 [00:01<00:39,  1.20it/s]
  6%|▌         | 3/50 [00:03<00:47,  1.02s/it]
  8%|▊         | 4/50 [00:03<00:36,  1.25it/s]
 10%|█         | 5/50 [00:04<00:42,  1.05it/s]
 12%|█▏        | 6/50 [00:05<00:34,  1.27it/s]
 14%|█▍        | 7/50 [00:06<00:40,  1.07it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.28it/s]
 20%|██        | 10/50 [00:08<00:28,  1.39it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.58it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.50it/s]
 28%|██▊       | 14/50 [00:08<00:12,  2.96it/s]
 30%|███       | 15/50 [00:10<00:19,  1.76it/s]
 32%|███▏      | 16/50 [00:10<00:17,  1.92it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.13it/s]
 40%|████      | 20/50 [00:10<00:06,  4.56it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.40it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.88it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.46it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.31it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.54it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.49it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.43it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.47it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.12it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.21it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.97it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.09it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.58it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.74it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.37it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.57it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.21it/s]
+
+  clip 4/4: 46.4s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000008/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000008/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 11914.5ms  total=  11.9s
+  clip_total         n=  4  mean= 44681.4ms  total= 178.7s
 
+Done in 190.9s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000009.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a coastal vacation rental kitchenette with a breakfast area, with soft white walls and horizontal shiplap on the lower half, light yellow weathered pine plank floor, and warm morning sun filtering through a sheer linen curtain on the right. In the foreground, there is a light ash three-drawer dresser with smooth recessed cut-out handles, and a small ceramic bowl of seashells and a stack of striped dish towels rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=262  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.12s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.52it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.17it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.53it/s]
 10%|█         | 5/50 [00:03<00:36,  1.23it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.54it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:06<00:24,  1.63it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.84it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.48it/s]
 30%|███       | 15/50 [00:08<00:15,  2.24it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.36it/s]
 36%|███▌      | 18/50 [00:08<00:08,  3.58it/s]
 40%|████      | 20/50 [00:09<00:05,  5.04it/s]
 44%|████▍     | 22/50 [00:10<00:10,  2.67it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.21it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.66it/s]
 58%|█████▊    | 29/50 [00:12<00:05,  3.55it/s]
 62%|██████▏   | 31/50 [00:13<00:06,  2.72it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.70it/s]
 68%|██████▊   | 34/50 [00:13<00:04,  3.56it/s]
 70%|███████   | 35/50 [00:15<00:06,  2.28it/s]
 72%|███████▏  | 36/50 [00:15<00:06,  2.31it/s]
 74%|███████▍  | 37/50 [00:15<00:04,  2.78it/s]
 76%|███████▌  | 38/50 [00:16<00:06,  1.82it/s]
 78%|███████▊  | 39/50 [00:16<00:05,  2.02it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  1.89it/s]
 84%|████████▍ | 42/50 [00:18<00:04,  1.99it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.52it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.67it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.36it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.25it/s]
+
+  clip 1/4: 41.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.77it/s]
 30%|███       | 15/50 [00:09<00:15,  2.19it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.27it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.94it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.72it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.72it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.40it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.52it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.29it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.73it/s]
 62%|██████▏   | 31/50 [00:14<00:08,  2.17it/s]
 64%|██████▍   | 32/50 [00:14<00:08,  2.17it/s]
 66%|██████▌   | 33/50 [00:15<00:06,  2.64it/s]
 68%|██████▊   | 34/50 [00:15<00:05,  2.89it/s]
 70%|███████   | 35/50 [00:16<00:08,  1.69it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.78it/s]
 74%|███████▍  | 37/50 [00:17<00:05,  2.25it/s]
 76%|███████▌  | 38/50 [00:18<00:08,  1.47it/s]
 78%|███████▊  | 39/50 [00:18<00:06,  1.65it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.08it/s]
 82%|████████▏ | 41/50 [00:20<00:06,  1.42it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.63it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.25it/s]
 88%|████████▊ | 44/50 [00:22<00:04,  1.49it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.20it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.43it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.18it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.40it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.17it/s]
+
+  clip 2/4: 46.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.41it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.70it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.76it/s]
 30%|███       | 15/50 [00:09<00:15,  2.19it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.28it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.42it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.83it/s]
 44%|████▍     | 22/50 [00:11<00:09,  2.80it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.43it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.50it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.43it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.46it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.86it/s]
 62%|██████▏   | 31/50 [00:14<00:09,  2.10it/s]
 64%|██████▍   | 32/50 [00:15<00:08,  2.02it/s]
 66%|██████▌   | 33/50 [00:15<00:06,  2.50it/s]
 68%|██████▊   | 34/50 [00:15<00:05,  2.97it/s]
 70%|███████   | 35/50 [00:16<00:08,  1.70it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.80it/s]
 74%|███████▍  | 37/50 [00:17<00:05,  2.28it/s]
 76%|███████▌  | 38/50 [00:18<00:08,  1.48it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.60it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.02it/s]
 82%|████████▏ | 41/50 [00:20<00:06,  1.36it/s]
 84%|████���███▍ | 42/50 [00:21<00:05,  1.56it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.22it/s]
 88%|████████▊ | 44/50 [00:22<00:04,  1.40it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.15it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.36it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.15it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.35it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.14it/s]
+
+  clip 3/4: 46.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.38it/s]
 10%|█         | 5/50 [00:04<00:40,  1.12it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.49it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.68it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.73it/s]
 30%|███       | 15/50 [00:09<00:16,  2.18it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.27it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.94it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.70it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.71it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.35it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.50it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.54it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.60it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.03it/s]
 62%|██████▏   | 31/50 [00:14<00:08,  2.16it/s]
 64%|██████▍   | 32/50 [00:15<00:08,  2.20it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.45it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.03it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.07it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.81it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  1.92it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.77it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.87it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.45it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.58it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.27it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.48it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.45it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.20it/s]
+
+  clip 4/4: 44.9s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000009/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000009/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 11865.7ms  total=  11.9s
+  clip_total         n=  4  mean= 44944.0ms  total= 179.8s
 
+Done in 191.9s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000010.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a modern kitchen with a cooking area, with clean white walls, light yellow oak hardwood floor, and a row of white quartz countertops along the back wall under bright daylight. In the foreground, there is a white three-drawer dresser with slim black bar handles, and a stainless steel kettle and a small vase of fresh herbs sit on a side counter. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=265  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.11s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.53it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.16it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.51it/s]
 10%|█         | 5/50 [00:03<00:36,  1.23it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.54it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:28,  1.47it/s]
 18%|█▊        | 9/50 [00:06<00:21,  1.90it/s]
 20%|██        | 10/50 [00:07<00:29,  1.37it/s]
 22%|██▏       | 11/50 [00:07<00:25,  1.56it/s]
 24%|██▍       | 12/50 [00:07<00:18,  2.03it/s]
 26%|██▌       | 13/50 [00:08<00:15,  2.43it/s]
 28%|██▊       | 14/50 [00:08<00:13,  2.72it/s]
 30%|███       | 15/50 [00:09<00:20,  1.70it/s]
 32%|███▏      | 16/50 [00:09<00:19,  1.76it/s]
 34%|███▍      | 17/50 [00:10<00:15,  2.10it/s]
 36%|███▌      | 18/50 [00:10<00:13,  2.43it/s]
 38%|███▊      | 19/50 [00:10<00:11,  2.58it/s]
 40%|████      | 20/50 [00:10<00:09,  3.14it/s]
 42%|████▏     | 21/50 [00:12<00:15,  1.82it/s]
 44%|████▍     | 22/50 [00:12<00:14,  1.95it/s]
 46%|████▌     | 23/50 [00:12<00:11,  2.31it/s]
 48%|████▊     | 24/50 [00:12<00:09,  2.82it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.57it/s]
 52%|█████▏    | 26/50 [00:14<00:12,  1.86it/s]
 54%|█████▍    | 27/50 [00:14<00:11,  2.04it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.68it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.41it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.05it/s]
 62%|██████▏   | 31/50 [00:16<00:10,  1.89it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.11it/s]
 70%|███████   | 35/50 [00:17<00:06,  2.31it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.42it/s]
 76%|███████▌  | 38/50 [00:19<00:05,  2.11it/s]
 78%|███████▊  | 39/50 [00:19<00:04,  2.24it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  2.00it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.12it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.61it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.77it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.40it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.64it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.35it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.60it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.32it/s]
+
+  clip 1/4: 46.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.17s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.31it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.04it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.33it/s]
 10%|█         | 5/50 [00:04<00:41,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.34it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.12it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.28it/s]
 18%|█▊        | 9/50 [00:06<00:24,  1.67it/s]
 20%|██        | 10/50 [00:08<00:32,  1.25it/s]
 22%|██▏       | 11/50 [00:08<00:27,  1.42it/s]
 24%|██▍       | 12/50 [00:08<00:20,  1.88it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.49it/s]
 28%|██▊       | 14/50 [00:09<00:12,  2.92it/s]
 30%|███       | 15/50 [00:10<00:21,  1.64it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.80it/s]
 34%|███▍      | 17/50 [00:10<00:14,  2.34it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.90it/s]
 40%|████      | 20/50 [00:11<00:07,  4.07it/s]
 42%|████▏     | 21/50 [00:12<00:14,  2.05it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.10it/s]
 48%|████▊     | 24/50 [00:13<00:07,  3.25it/s]
 50%|█████     | 25/50 [00:13<00:06,  3.85it/s]
 52%|█████▏    | 26/50 [00:14<00:11,  2.04it/s]
 54%|█████▍    | 27/50 [00:14<00:11,  2.08it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.56it/s]
 58%|█████▊    | 29/50 [00:15<00:06,  3.15it/s]
 62%|██████▏   | 31/50 [00:16<00:08,  2.16it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.10it/s]
 66%|██████▌   | 33/50 [00:17<00:06,  2.55it/s]
 68%|██████▊   | 34/50 [00:17<00:05,  3.12it/s]
 70%|███████   | 35/50 [00:18<00:08,  1.80it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.96it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.81it/s]
 78%|███████▊  | 39/50 [00:20<00:05,  1.95it/s]
 82%|████████▏ | 41/50 [00:21<00:04,  1.80it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.88it/s]
 86%|████████▌ | 43/50 [00:23<00:04,  1.45it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.62it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.30it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.47it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.21it/s]
+
+  clip 2/4: 48.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:38,  1.26it/s]
  6%|▌         | 3/50 [00:02<00:46,  1.01it/s]
  8%|▊         | 4/50 [00:03<00:38,  1.20it/s]
 10%|█         | 5/50 [00:04<00:44,  1.01it/s]
 12%|█▏        | 6/50 [00:05<00:37,  1.18it/s]
 14%|█▍        | 7/50 [00:06<00:42,  1.02it/s]
 16%|█▌        | 8/50 [00:07<00:35,  1.18it/s]
 18%|█▊        | 9/50 [00:07<00:26,  1.55it/s]
 20%|██        | 10/50 [00:08<00:33,  1.20it/s]
 22%|██▏       | 11/50 [00:09<00:27,  1.42it/s]
 26%|██▌       | 13/50 [00:09<00:15,  2.38it/s]
 28%|██▊       | 14/50 [00:09<00:12,  2.79it/s]
 30%|███       | 15/50 [00:10<00:20,  1.70it/s]
 32%|███▏      | 16/50 [00:11<00:19,  1.75it/s]
 34%|███▍      | 17/50 [00:11<00:14,  2.27it/s]
 38%|███▊      | 19/50 [00:11<00:08,  3.57it/s]
 42%|████▏     | 21/50 [00:12<00:11,  2.44it/s]
 44%|████▍     | 22/50 [00:13<00:11,  2.41it/s]
 46%|████▌     | 23/50 [00:13<00:09,  2.89it/s]
 48%|████▊     | 24/50 [00:13<00:07,  3.30it/s]
 50%|█████     | 25/50 [00:13<00:06,  3.83it/s]
 52%|█████▏    | 26/50 [00:14<00:12,  1.91it/s]
 54%|█████▍    | 27/50 [00:15<00:11,  2.08it/s]
 60%|██████    | 30/50 [00:15<00:04,  4.09it/s]
 62%|██████▏   | 31/50 [00:16<00:08,  2.29it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.38it/s]
 70%|███████   | 35/50 [00:18<00:06,  2.37it/s]
 72%|███████▏  | 36/50 [00:18<00:05,  2.42it/s]
 76%|███████▌  | 38/50 [00:19<00:05,  2.07it/s]
 78%|███████▊  | 39/50 [00:20<00:05,  2.18it/s]
 82%|████████▏ | 41/50 [00:21<00:04,  1.95it/s]
 84%|████████▍ | 42/50 [00:21<00:03,  2.05it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.54it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.72it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.36it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.57it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.50it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.23it/s]
+
+  clip 3/4: 47.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.31it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.02it/s]
  8%|▊         | 4/50 [00:03<00:36,  1.27it/s]
 10%|█         | 5/50 [00:04<00:43,  1.03it/s]
 12%|█▏        | 6/50 [00:05<00:34,  1.27it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.08it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.29it/s]
 20%|██        | 10/50 [00:08<00:28,  1.42it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.59it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.58it/s]
 30%|███       | 15/50 [00:09<00:16,  2.12it/s]
 32%|███▏      | 16/50 [00:10<00:15,  2.22it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.87it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.72it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.72it/s]
 50%|█████     | 25/50 [00:12<00:05,  4.39it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.58it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.91it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.53it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.47it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.50it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.16it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.25it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.98it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.09it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.59it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.76it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.39it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.60it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.29it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.51it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.24it/s]
+
+  clip 4/4: 41.1s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000010/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000010/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13184.5ms  total=  13.2s
+  clip_total         n=  4  mean= 45711.6ms  total= 182.8s
 
+Done in 196.3s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000011.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a farmhouse kitchen with a baking area, with warm white shiplap walls, light yellow weathered pine plank floor, and an open shelf of ceramic crockery beside a window letting in soft morning sun. In the foreground, there is a warm cream-painted three-drawer dresser with small brushed-brass knobs, and a wooden bowl of lemons and a folded gingham towel rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=301  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.11s/it]
  4%|▍         | 2/50 [00:01<00:32,  1.47it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.15it/s]
  8%|▊         | 4/50 [00:02<00:31,  1.46it/s]
 10%|█         | 5/50 [00:04<00:37,  1.20it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.46it/s]
 14%|█▍        | 7/50 [00:05<00:35,  1.22it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.44it/s]
 20%|██        | 10/50 [00:07<00:25,  1.57it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.77it/s]
 24%|██▍       | 12/50 [00:07<00:16,  2.24it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.58it/s]
 30%|███       | 15/50 [00:08<00:16,  2.13it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.27it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.53it/s]
 42%|████▏     | 21/50 [00:10<00:09,  2.97it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.97it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.85it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.80it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.21it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.73it/s]
 68%|██████▊   | 34/50 [00:14<00:04,  3.55it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.42it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.11it/s]
 78%|███████▊  | 39/50 [00:17<00:05,  2.14it/s]
 80%|████████  | 40/50 [00:17<00:03,  2.53it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  1.81it/s]
 84%|████████▍ | 42/50 [00:19<00:04,  1.83it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.42it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.61it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.32it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.50it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.25it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.43it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.20it/s]
+
+  clip 1/4: 44.4s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.32it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.04it/s]
  8%|▊         | 4/50 [00:03<00:35,  1.30it/s]
 10%|█         | 5/50 [00:04<00:42,  1.06it/s]
 12%|█▏        | 6/50 [00:04<00:33,  1.31it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.10it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.36it/s]
 20%|██        | 10/50 [00:07<00:27,  1.48it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.67it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.17it/s]
 30%|███       | 15/50 [00:09<00:17,  2.05it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.16it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.18it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.74it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.74it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.40it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.56it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.58it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.30it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.46it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.34it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.25it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.01it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.11it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.91it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.93it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.47it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.59it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.26it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.39it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.17it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.35it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.14it/s]
+
+  clip 2/4: 47.7s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.24s/it]
  4%|▍         | 2/50 [00:01<00:37,  1.27it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.02it/s]
  8%|▊         | 4/50 [00:03<00:35,  1.31it/s]
 10%|█         | 5/50 [00:04<00:41,  1.09it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.36it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.69it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.73it/s]
 30%|███       | 15/50 [00:09<00:16,  2.15it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.25it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.22it/s]
 40%|████      | 20/50 [00:10<00:06,  4.55it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.48it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.92it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.48it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.72it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.46it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.23it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.23it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.01it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.08it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.44it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.71it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.85it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.41it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.54it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.25it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.42it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.18it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.35it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.12it/s]
+
+  clip 3/4: 44.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.40it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.42it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 28%|██▊       | 14/50 [00:07<00:11,  3.27it/s]
 30%|███       | 15/50 [00:09<00:16,  2.08it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.19it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.27it/s]
 40%|████      | 20/50 [00:09<00:06,  4.66it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.44it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.41it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.52it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.56it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.15it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.50it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.44it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.48it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.13it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.22it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.98it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.09it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.58it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.75it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.38it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.51it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.40it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.16it/s]
+
+  clip 4/4: 42.7s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000011/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000011/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 15995.8ms  total=  16.0s
+  clip_total         n=  4  mean= 44849.6ms  total= 179.4s
 
+Done in 195.6s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000012.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a narrow galley kitchen with a prep area, with bright white walls and a white subway-tile backsplash, light gray vinyl plank floor, and slim under-cabinet lights glowing along both runs of cabinetry. In the foreground, there is a light oak three-drawer dresser with natural wood grain and recessed finger pulls, and a chrome utensil holder and a small potted basil plant sit nearby. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=301  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.12s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.52it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.17it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.54it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.54it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.26it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.54it/s]
 20%|██        | 10/50 [00:06<00:24,  1.64it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.76it/s]
 24%|██▍       | 12/50 [00:07<00:17,  2.19it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.53it/s]
 30%|███       | 15/50 [00:08<00:16,  2.12it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.23it/s]
 38%|███▊      | 19/50 [00:09<00:07,  4.16it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.88it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.89it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.73it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.75it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.16it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.70it/s]
 70%|███████   | 35/50 [00:14<00:05,  2.64it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.68it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.31it/s]
 78%|███████▊  | 39/50 [00:16<00:04,  2.40it/s]
 82%|████████▏ | 41/50 [00:17<00:04,  2.13it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.25it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.71it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.90it/s]
 90%|█████████ | 45/50 [00:20<00:03,  1.49it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.71it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.38it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.62it/s]
 98%|█████████▊| 49/50 [00:23<00:00,  1.33it/s]
+
+  clip 1/4: 39.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.42it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:33,  1.33it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.09it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.30it/s]
 18%|█▊        | 9/50 [00:06<00:24,  1.70it/s]
 20%|██        | 10/50 [00:07<00:31,  1.25it/s]
 22%|██▏       | 11/50 [00:08<00:28,  1.37it/s]
 24%|██▍       | 12/50 [00:08<00:21,  1.74it/s]
 26%|██▌       | 13/50 [00:08<00:17,  2.17it/s]
 28%|██▊       | 14/50 [00:09<00:13,  2.73it/s]
 30%|███       | 15/50 [00:10<00:22,  1.59it/s]
 32%|███▏      | 16/50 [00:10<00:20,  1.69it/s]
 34%|███▍      | 17/50 [00:11<00:15,  2.10it/s]
 36%|███▌      | 18/50 [00:11<00:12,  2.65it/s]
 38%|███▊      | 19/50 [00:11<00:09,  3.30it/s]
 40%|████      | 20/50 [00:11<00:07,  3.97it/s]
 42%|████▏     | 21/50 [00:12<00:15,  1.86it/s]
 44%|████▍     | 22/50 [00:13<00:15,  1.83it/s]
 46%|████▌     | 23/50 [00:13<00:11,  2.27it/s]
 48%|████▊     | 24/50 [00:13<00:09,  2.86it/s]
 52%|█████▏    | 26/50 [00:14<00:11,  2.07it/s]
 54%|█████▍    | 27/50 [00:15<00:11,  2.05it/s]
 56%|█████▌    | 28/50 [00:15<00:08,  2.51it/s]
 60%|██████    | 30/50 [00:15<00:05,  3.88it/s]
 62%|██████▏   | 31/50 [00:16<00:09,  2.11it/s]
 64%|██████▍   | 32/50 [00:17<00:08,  2.16it/s]
 70%|███████   | 35/50 [00:18<00:06,  2.26it/s]
 72%|███████▏  | 36/50 [00:18<00:05,  2.34it/s]
 76%|███████▌  | 38/50 [00:20<00:05,  2.03it/s]
 78%|███████▊  | 39/50 [00:20<00:05,  2.14it/s]
 82%|████████▏ | 41/50 [00:21<00:04,  1.92it/s]
 84%|████████▍ | 42/50 [00:22<00:03,  2.04it/s]
 86%|████████▌ | 43/50 [00:23<00:04,  1.54it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.72it/s]
 90%|█████████ | 45/50 [00:24<00:03,  1.36it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.57it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.27it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.50it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.23it/s]
+
+  clip 2/4: 46.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:39,  1.21it/s]
  6%|▌         | 3/50 [00:02<00:47,  1.00s/it]
  8%|▊         | 4/50 [00:03<00:37,  1.24it/s]
 10%|█         | 5/50 [00:04<00:42,  1.06it/s]
 12%|█▏        | 6/50 [00:05<00:36,  1.22it/s]
 14%|█▍        | 7/50 [00:06<00:41,  1.03it/s]
 16%|█▌        | 8/50 [00:07<00:34,  1.22it/s]
 18%|█▊        | 9/50 [00:07<00:24,  1.66it/s]
 20%|██        | 10/50 [00:08<00:31,  1.26it/s]
 22%|██▏       | 11/50 [00:08<00:28,  1.38it/s]
 24%|██▍       | 12/50 [00:09<00:20,  1.86it/s]
 26%|██▌       | 13/50 [00:09<00:15,  2.41it/s]
 28%|██▊       | 14/50 [00:09<00:11,  3.12it/s]
 30%|███       | 15/50 [00:10<00:20,  1.70it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.83it/s]
 36%|███▌      | 18/50 [00:11<00:10,  3.02it/s]
 40%|████      | 20/50 [00:11<00:06,  4.55it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.32it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.28it/s]
 52%|█████▏    | 26/50 [00:14<00:09,  2.46it/s]
 54%|█████▍    | 27/50 [00:14<00:09,  2.48it/s]
 60%|██████    | 30/50 [00:14<00:04,  4.09it/s]
 64%|██████▍   | 32/50 [00:16<00:07,  2.47it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.33it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.26it/s]
 76%|███████▌  | 38/50 [00:19<00:05,  2.02it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.12it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.90it/s]
 84%|████████▍ | 42/50 [00:21<00:03,  2.01it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.55it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.69it/s]
 90%|████��████ | 45/50 [00:23<00:03,  1.35it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.53it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.46it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.20it/s]
+
+  clip 3/4: 48.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:02,  1.28s/it]
  4%|▍         | 2/50 [00:01<00:40,  1.19it/s]
  6%|▌         | 3/50 [00:03<00:49,  1.06s/it]
  8%|▊         | 4/50 [00:03<00:38,  1.19it/s]
 10%|█         | 5/50 [00:04<00:44,  1.01it/s]
 12%|█▏        | 6/50 [00:05<00:36,  1.22it/s]
 14%|█▍        | 7/50 [00:06<00:41,  1.04it/s]
 16%|█▌        | 8/50 [00:07<00:33,  1.25it/s]
 18%|█▊        | 9/50 [00:07<00:24,  1.69it/s]
 20%|██        | 10/50 [00:08<00:31,  1.28it/s]
 22%|██▏       | 11/50 [00:08<00:26,  1.47it/s]
 24%|██▍       | 12/50 [00:09<00:20,  1.90it/s]
 26%|██▌       | 13/50 [00:09<00:17,  2.10it/s]
 28%|██▊       | 14/50 [00:09<00:16,  2.18it/s]
 30%|███       | 15/50 [00:11<00:25,  1.39it/s]
 32%|███▏      | 16/50 [00:11<00:21,  1.59it/s]
 36%|███▌      | 18/50 [00:11<00:11,  2.75it/s]
 40%|████      | 20/50 [00:11<00:07,  4.00it/s]
 42%|████▏     | 21/50 [00:13<00:14,  2.03it/s]
 44%|████▍     | 22/50 [00:13<00:13,  2.11it/s]
 46%|████▌     | 23/50 [00:13<00:10,  2.49it/s]
 48%|████▊     | 24/50 [00:14<00:09,  2.74it/s]
 50%|█████     | 25/50 [00:14<00:08,  2.97it/s]
 52%|█████▏    | 26/50 [00:15<00:14,  1.70it/s]
 54%|█████▍    | 27/50 [00:15<00:12,  1.90it/s]
 58%|█████▊    | 29/50 [00:16<00:06,  3.15it/s]
 62%|██████▏   | 31/50 [00:17<00:08,  2.31it/s]
 64%|██████▍   | 32/50 [00:17<00:07,  2.39it/s]
 70%|███████   | 35/50 [00:18<00:06,  2.38it/s]
 72%|███████▏  | 36/50 [00:19<00:05,  2.44it/s]
 76%|███████▌  | 38/50 [00:20<00:05,  2.05it/s]
 78%|███████▊  | 39/50 [00:20<00:05,  2.16it/s]
 82%|████████▏ | 41/50 [00:22<00:04,  1.93it/s]
 84%|████████▍ | 42/50 [00:22<00:03,  2.06it/s]
 86%|████████▌ | 43/50 [00:23<00:04,  1.56it/s]
 88%|████████▊ | 44/50 [00:24<00:03,  1.67it/s]
 90%|█████████ | 45/50 [00:25<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:25<00:02,  1.50it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.22it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.44it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.20it/s]
+
+  clip 4/4: 50.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000012/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000012/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13228.1ms  total=  13.2s
+  clip_total         n=  4  mean= 46298.7ms  total= 185.2s
 
+Done in 198.6s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000013.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, an industrial loft kitchen with a dining area, with white-painted brick walls, polished gray concrete floor, and a black-framed factory window flooding the room with cool afternoon light. In the foreground, there is a natural pine three-drawer dresser with round wooden knobs, and a cast-iron pan and a stack of enamel mugs are arranged on the counter. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=301  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.11s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.52it/s]
  6%|▌         | 3/50 [00:02<00:40,  1.17it/s]
  8%|▊         | 4/50 [00:02<00:30,  1.50it/s]
 10%|█         | 5/50 [00:04<00:37,  1.21it/s]
 12%|█▏        | 6/50 [00:04<00:29,  1.50it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.24it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.51it/s]
 20%|██        | 10/50 [00:06<00:24,  1.60it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.82it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.94it/s]
 30%|███       | 15/50 [00:08<00:14,  2.35it/s]
 32%|███▏      | 16/50 [00:08<00:13,  2.45it/s]
 38%|███▊      | 19/50 [00:09<00:07,  4.25it/s]
 42%|████▏     | 21/50 [00:10<00:09,  2.91it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.91it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.69it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.75it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.14it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.70it/s]
 68%|██████▊   | 34/50 [00:13<00:04,  3.51it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.44it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.18it/s]
 78%|███████▊  | 39/50 [00:16<00:04,  2.24it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  2.01it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.04it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.60it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.74it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.40it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.56it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.29it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.44it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.22it/s]
+
+  clip 1/4: 42.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.36it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.39it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.37it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.36it/s]
 18%|█▊        | 9/50 [00:06<00:22,  1.82it/s]
 20%|██        | 10/50 [00:07<00:30,  1.32it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.54it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.68it/s]
 30%|███       | 15/50 [00:09<00:16,  2.12it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.23it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.37it/s]
 40%|████      | 20/50 [00:10<00:06,  4.71it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.46it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.37it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.51it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.55it/s]
 58%|█████▊    | 29/50 [00:13<00:05,  3.60it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.53it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.46it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.32it/s]
 70%|███████   | 35/50 [00:16<00:07,  2.00it/s]
 72%|███████▏  | 36/50 [00:17<00:06,  2.00it/s]
 74%|███████▍  | 37/50 [00:17<00:05,  2.27it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.55it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.65it/s]
 80%|████████  | 40/50 [00:19<00:04,  2.02it/s]
 82%|████████▏ | 41/50 [00:20<00:06,  1.41it/s]
 84%|████████▍ | 42/50 [00:21<00:05,  1.58it/s]
 86%|████████▌ | 43/50 [00:22<00:05,  1.25it/s]
 88%|████████▊ | 44/50 [00:22<00:04,  1.41it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.15it/s]
 92%|█████████▏| 46/50 [00:24<00:03,  1.30it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.10it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.25it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.08it/s]
+
+  clip 2/4: 47.3s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:01,  1.26s/it]
  4%|▍         | 2/50 [00:01<00:38,  1.25it/s]
  6%|▌         | 3/50 [00:02<00:47,  1.00s/it]
  8%|▊         | 4/50 [00:03<00:34,  1.33it/s]
 10%|█         | 5/50 [00:04<00:40,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.39it/s]
 14%|█▍        | 7/50 [00:06<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.69it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.24it/s]
 30%|███       | 15/50 [00:09<00:16,  2.07it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.19it/s]
 34%|███▍      | 17/50 [00:09<00:12,  2.68it/s]
 38%|███▊      | 19/50 [00:09<00:07,  4.07it/s]
 42%|████▏     | 21/50 [00:11<00:11,  2.62it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.61it/s]
 48%|████▊     | 24/50 [00:11<00:06,  3.86it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.61it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.62it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.40it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.53it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.45it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.43it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.05it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.06it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.32it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.56it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.66it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.29it/s]
 88%|████████▊ | 44/50 [00:22<00:04,  1.41it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.15it/s]
 92%|█████████▏| 46/50 [00:24<00:03,  1.28it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.09it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.26it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.08it/s]
+
+  clip 3/4: 47.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.33it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.03it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.37it/s]
 10%|█         | 5/50 [00:04<00:40,  1.11it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.39it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.38it/s]
 20%|██        | 10/50 [00:07<00:27,  1.48it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.67it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.18it/s]
 30%|███       | 15/50 [00:09<00:17,  2.03it/s]
 32%|███▏      | 16/50 [00:09<00:16,  2.11it/s]
 36%|███▌      | 18/50 [00:09<00:10,  3.17it/s]
 40%|████      | 20/50 [00:10<00:06,  4.51it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.45it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.89it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.44it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.20it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.48it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.52it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.44it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.48it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.12it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.21it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.95it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.01it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.51it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.60it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.25it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.39it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.16it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.30it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.10it/s]
+
+  clip 4/4: 46.7s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000013/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000013/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 14282.3ms  total=  14.3s
+  clip_total         n=  4  mean= 45719.8ms  total= 182.9s
 
+Done in 197.3s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000014.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a home office with a desk area, with warm gray walls, light yellow maple hardwood floor, and sheer curtains diffusing soft afternoon sun across the room. In the foreground, there is a honey-toned maple three-drawer dresser with brushed-nickel cup pulls, and a closed silver laptop and a small potted succulent are placed on the desk surface. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=256  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:53,  1.10s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.53it/s]
  6%|▌         | 3/50 [00:02<00:39,  1.18it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.54it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.43it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.18it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.38it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:24,  1.57it/s]
 24%|██▍       | 12/50 [00:08<00:20,  1.87it/s]
 26%|██▌       | 13/50 [00:08<00:16,  2.28it/s]
 28%|██▊       | 14/50 [00:08<00:13,  2.76it/s]
 30%|███       | 15/50 [00:09<00:20,  1.68it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.82it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.38it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.86it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.18it/s]
 40%|████      | 20/50 [00:10<00:08,  3.54it/s]
 42%|████▏     | 21/50 [00:11<00:15,  1.85it/s]
 44%|████▍     | 22/50 [00:12<00:14,  1.92it/s]
 46%|████▌     | 23/50 [00:12<00:10,  2.50it/s]
 48%|████▊     | 24/50 [00:12<00:08,  3.16it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.87it/s]
 52%|█████▏    | 26/50 [00:13<00:12,  1.94it/s]
 54%|█████▍    | 27/50 [00:14<00:11,  2.05it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.64it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.22it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.60it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.83it/s]
 64%|██████▍   | 32/50 [00:16<00:09,  1.96it/s]
 66%|██████▌   | 33/50 [00:16<00:06,  2.56it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.28it/s]
 70%|███████   | 35/50 [00:17<00:08,  1.78it/s]
 72%|███████▏  | 36/50 [00:18<00:06,  2.02it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.89it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  2.04it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.91it/s]
 84%|████████▍ | 42/50 [00:21<00:03,  2.08it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.59it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.80it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.43it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.60it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.31it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.50it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.25it/s]
+
+  clip 1/4: 45.5s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.36it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:35,  1.31it/s]
 10%|█         | 5/50 [00:04<00:41,  1.07it/s]
 12%|█▏        | 6/50 [00:04<00:33,  1.32it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.11it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.35it/s]
 20%|██        | 10/50 [00:07<00:27,  1.46it/s]
 22%|██▏       | 11/50 [00:08<00:24,  1.59it/s]
 24%|██▍       | 12/50 [00:08<00:18,  2.03it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.60it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.15it/s]
 30%|███       | 15/50 [00:09<00:20,  1.68it/s]
 32%|███▏      | 16/50 [00:10<00:18,  1.86it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.14it/s]
 40%|████      | 20/50 [00:10<00:06,  4.67it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.40it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.33it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.81it/s]
 52%|█████▏    | 26/50 [00:13<00:11,  2.13it/s]
 54%|█████▍    | 27/50 [00:13<00:10,  2.24it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.39it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.37it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.40it/s]
 66%|██████▌   | 33/50 [00:15<00:05,  2.89it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.37it/s]
 70%|███████   | 35/50 [00:17<00:07,  1.89it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.95it/s]
 74%|███████▍  | 37/50 [00:17<00:05,  2.45it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.58it/s]
 78%|███████▊  | 39/50 [00:19<00:06,  1.74it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.64it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.82it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.41it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.62it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.29it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.24it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.46it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.19it/s]
+
+  clip 2/4: 48.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.24s/it]
  4%|▍         | 2/50 [00:01<00:39,  1.21it/s]
  6%|▌         | 3/50 [00:03<00:47,  1.01s/it]
  8%|▊         | 4/50 [00:03<00:36,  1.24it/s]
 10%|█         | 5/50 [00:04<00:42,  1.05it/s]
 12%|█▏        | 6/50 [00:05<00:35,  1.25it/s]
 14%|█▍        | 7/50 [00:06<00:40,  1.07it/s]
 16%|█▌        | 8/50 [00:06<00:33,  1.27it/s]
 18%|█▊        | 9/50 [00:07<00:23,  1.71it/s]
 20%|██        | 10/50 [00:08<00:30,  1.30it/s]
 22%|██▏       | 11/50 [00:08<00:27,  1.42it/s]
 24%|██▍       | 12/50 [00:09<00:23,  1.65it/s]
 26%|██▌       | 13/50 [00:09<00:20,  1.77it/s]
 28%|██▊       | 14/50 [00:09<00:17,  2.11it/s]
 30%|███       | 15/50 [00:11<00:25,  1.40it/s]
 32%|███▏      | 16/50 [00:11<00:24,  1.42it/s]
 34%|███▍      | 17/50 [00:12<00:19,  1.72it/s]
 36%|███▌      | 18/50 [00:12<00:15,  2.08it/s]
 38%|███▊      | 19/50 [00:12<00:12,  2.41it/s]
 40%|████      | 20/50 [00:12<00:10,  2.81it/s]
 42%|████▏     | 21/50 [00:14<00:18,  1.59it/s]
 44%|████▍     | 22/50 [00:14<00:15,  1.79it/s]
 48%|████▊     | 24/50 [00:14<00:09,  2.88it/s]
 52%|█████▏    | 26/50 [00:15<00:11,  2.18it/s]
 54%|█████▍    | 27/50 [00:16<00:10,  2.28it/s]
 58%|█████▊    | 29/50 [00:16<00:06,  3.31it/s]
 62%|██████▏   | 31/50 [00:17<00:07,  2.43it/s]
 64%|██████▍   | 32/50 [00:18<00:07,  2.46it/s]
 70%|███████   | 35/50 [00:19<00:06,  2.42it/s]
 72%|███████▏  | 36/50 [00:19<00:05,  2.46it/s]
 76%|███████▌  | 38/50 [00:21<00:05,  2.10it/s]
 78%|███████▊  | 39/50 [00:21<00:05,  2.19it/s]
 82%|████████▏ | 41/50 [00:22<00:04,  1.95it/s]
 84%|████████▍ | 42/50 [00:23<00:03,  2.03it/s]
 86%|████████▌ | 43/50 [00:24<00:04,  1.53it/s]
 88%|████████▊ | 44/50 [00:24<00:03,  1.68it/s]
 90%|█████████ | 45/50 [00:25<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:26<00:02,  1.54it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:27<00:01,  1.49it/s]
 98%|█████████▊| 49/50 [00:28<00:00,  1.22it/s]
+
+  clip 3/4: 50.9s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:38,  1.24it/s]
  6%|▌         | 3/50 [00:02<00:47,  1.01s/it]
  8%|▊         | 4/50 [00:03<00:36,  1.26it/s]
 10%|█         | 5/50 [00:04<00:43,  1.05it/s]
 12%|█▏        | 6/50 [00:05<00:34,  1.28it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.09it/s]
 16%|█▌        | 8/50 [00:06<00:31,  1.35it/s]
 20%|██        | 10/50 [00:07<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.67it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.20it/s]
 30%|███       | 15/50 [00:09<00:16,  2.07it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.19it/s]
 38%|███▊      | 19/50 [00:10<00:08,  3.73it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.64it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.66it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.32it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.53it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.57it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.31it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.53it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.46it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.50it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.15it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.24it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.99it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.10it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.59it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.76it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.38it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.59it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.28it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.51it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.23it/s]
+
+  clip 4/4: 42.5s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000014/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000014/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 12676.3ms  total=  12.7s
+  clip_total         n=  4  mean= 46877.4ms  total= 187.5s
 
+Done in 200.4s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000015.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, a corporate office with a workstation area, with crisp white walls and a glass partition, medium gray carpet-tile floor, and even cool light from recessed ceiling panels. In the foreground, there is a dark walnut three-drawer dresser with matte black bar pulls, and a black desk monitor and a neat tray of pens sit on the surface. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=268  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.16s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.36it/s]
  6%|▌         | 3/50 [00:02<00:42,  1.10it/s]
  8%|▊         | 4/50 [00:03<00:31,  1.46it/s]
 10%|█         | 5/50 [00:04<00:37,  1.21it/s]
 12%|█▏        | 6/50 [00:04<00:29,  1.49it/s]
 14%|█▍        | 7/50 [00:05<00:35,  1.22it/s]
 16%|█▌        | 8/50 [00:05<00:28,  1.48it/s]
 20%|██        | 10/50 [00:07<00:24,  1.60it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.81it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.87it/s]
 30%|███       | 15/50 [00:08<00:15,  2.30it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.41it/s]
 36%|███▌      | 18/50 [00:09<00:08,  3.60it/s]
 40%|████      | 20/50 [00:09<00:06,  4.85it/s]
 44%|████▍     | 22/50 [00:10<00:10,  2.63it/s]
 48%|████▊     | 24/50 [00:10<00:07,  3.64it/s]
 52%|█████▏    | 26/50 [00:12<00:08,  2.67it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.72it/s]
 58%|█████▊    | 29/50 [00:12<00:05,  3.78it/s]
 62%|██████▏   | 31/50 [00:13<00:06,  2.74it/s]
 64%|██████▍   | 32/50 [00:14<00:06,  2.78it/s]
 70%|███████   | 35/50 [00:15<00:05,  2.67it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.71it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.30it/s]
 78%|███████▊  | 39/50 [00:17<00:04,  2.40it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  2.11it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.11it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.60it/s]
 88%|████████▊ | 44/50 [00:20<00:03,  1.68it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.38it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.56it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.28it/s]
 96%|█████████▌| 48/50 [00:23<00:01,  1.44it/s]
 98%|█████████▊| 49/50 [00:24<00:00,  1.21it/s]
+
+  clip 1/4: 46.5s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.22s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.30it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.03it/s]
  8%|▊         | 4/50 [00:03<00:35,  1.30it/s]
 10%|█         | 5/50 [00:04<00:41,  1.08it/s]
 12%|█▏        | 6/50 [00:05<00:33,  1.30it/s]
 14%|█▍        | 7/50 [00:06<00:39,  1.09it/s]
 16%|█▌        | 8/50 [00:06<00:32,  1.29it/s]
 18%|█▊        | 9/50 [00:06<00:23,  1.74it/s]
 20%|██        | 10/50 [00:08<00:30,  1.31it/s]
 22%|██▏       | 11/50 [00:08<00:25,  1.52it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.56it/s]
 30%|███       | 15/50 [00:09<00:16,  2.08it/s]
 32%|███▏      | 16/50 [00:10<00:15,  2.19it/s]
 36%|███▌      | 18/50 [00:10<00:09,  3.32it/s]
 40%|████      | 20/50 [00:10<00:06,  4.64it/s]
 44%|████▍     | 22/50 [00:12<00:11,  2.47it/s]
 50%|█████     | 25/50 [00:12<00:06,  3.91it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.43it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.12it/s]
 62%|██████▏   | 31/50 [00:15<00:07,  2.44it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.49it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.43it/s]
 72%|███████▏  | 36/50 [00:17<00:05,  2.46it/s]
 76%|███████▌  | 38/50 [00:18<00:05,  2.11it/s]
 78%|███████▊  | 39/50 [00:18<00:04,  2.20it/s]
 82%|████████▏ | 41/50 [00:20<00:04,  1.94it/s]
 84%|████████▍ | 42/50 [00:20<00:03,  2.06it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.56it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.69it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.33it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.49it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.21it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.35it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.11it/s]
+
+  clip 2/4: 49.2s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.35it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.05it/s]
  8%|▊         | 4/50 [00:03<00:33,  1.36it/s]
 10%|█         | 5/50 [00:04<00:40,  1.10it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.38it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.14it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.49it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.69it/s]
 26%|██▌       | 13/50 [00:08<00:13,  2.74it/s]
 30%|███       | 15/50 [00:09<00:16,  2.14it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.21it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.33it/s]
 40%|████      | 20/50 [00:09<00:06,  4.61it/s]
 44%|████▍     | 22/50 [00:11<00:11,  2.45it/s]
 50%|█████     | 25/50 [00:11<00:06,  3.89it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.47it/s]
 60%|██████    | 30/50 [00:13<00:05,  3.65it/s]
 64%|██████▍   | 32/50 [00:15<00:07,  2.43it/s]
 66%|██████▌   | 33/50 [00:15<00:06,  2.76it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.25it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.31it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.02it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.13it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.89it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.02it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.53it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.72it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.35it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.56it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.26it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.42it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.18it/s]
+
+  clip 3/4: 45.6s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:57,  1.18s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.43it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.09it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.43it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:30,  1.43it/s]
 14%|█▍        | 7/50 [00:05<00:36,  1.17it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.42it/s]
 20%|██        | 10/50 [00:07<00:26,  1.51it/s]
 22%|██▏       | 11/50 [00:07<00:22,  1.70it/s]
 28%|██▊       | 14/50 [00:07<00:11,  3.26it/s]
 30%|███       | 15/50 [00:09<00:16,  2.08it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.20it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.35it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.79it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.78it/s]
 48%|████▊     | 24/50 [00:11<00:06,  3.87it/s]
 52%|█████▏    | 26/50 [00:12<00:08,  2.67it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.67it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.38it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.55it/s]
 70%|███████   | 35/50 [00:15<00:06,  2.47it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.50it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.14it/s]
 78%|███████▊  | 39/50 [00:17<00:04,  2.23it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.97it/s]
 84%|████████▍ | 42/50 [00:19<00:03,  2.09it/s]
 86%|████████▌ | 43/50 [00:20<00:04,  1.58it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.75it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.38it/s]
 92%|█████████▏| 46/50 [00:22<00:02,  1.59it/s]
 94%|█████████▍| 47/50 [00:23<00:02,  1.28it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.50it/s]
 98%|█████████▊| 49/50 [00:25<00:00,  1.23it/s]
+
+  clip 4/4: 41.9s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000015/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000015/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 14951.2ms  total=  15.0s
+  clip_total         n=  4  mean= 45772.5ms  total= 183.1s
 
+Done in 198.2s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000016.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, an open-plan coworking space with a lounge area, with bright white walls and pale wood acoustic panels, light yellow birch plank floor, and a long row of pendant lights over a shared table. In the foreground, there is an espresso-brown three-drawer dresser with antique bronze handles, and a ceramic mug and a small stack of notebooks rest on top. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=275  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.11s/it]
  4%|▍         | 2/50 [00:01<00:31,  1.52it/s]
  6%|▌         | 3/50 [00:02<00:39,  1.18it/s]
  8%|▊         | 4/50 [00:02<00:29,  1.54it/s]
 10%|█         | 5/50 [00:03<00:36,  1.24it/s]
 12%|█▏        | 6/50 [00:04<00:28,  1.55it/s]
 14%|█▍        | 7/50 [00:05<00:33,  1.27it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.55it/s]
 20%|██        | 10/50 [00:06<00:24,  1.63it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.83it/s]
 26%|██▌       | 13/50 [00:07<00:12,  2.87it/s]
 30%|███       | 15/50 [00:08<00:15,  2.29it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.36it/s]
 38%|███▊      | 19/50 [00:09<00:07,  4.08it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.89it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.90it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.63it/s]
 52%|█████▏    | 26/50 [00:11<00:08,  2.73it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.76it/s]
 60%|██████    | 30/50 [00:12<00:04,  4.47it/s]
 62%|██████▏   | 31/50 [00:13<00:07,  2.64it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.69it/s]
 68%|██████▊   | 34/50 [00:13<00:04,  3.89it/s]
 70%|███████   | 35/50 [00:14<00:06,  2.32it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.44it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.11it/s]
 78%|███████▊  | 39/50 [00:16<00:04,  2.24it/s]
 82%|████████▏ | 41/50 [00:17<00:04,  2.03it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.15it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.63it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.78it/s]
 90%|█████████ | 45/50 [00:20<00:03,  1.40it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.63it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.33it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.57it/s]
 98%|█████████▊| 49/50 [00:23<00:00,  1.30it/s]
+
+  clip 1/4: 38.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:58,  1.19s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.36it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:35,  1.30it/s]
 10%|█         | 5/50 [00:04<00:43,  1.02it/s]
 12%|█▏        | 6/50 [00:05<00:33,  1.30it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.11it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.36it/s]
 20%|██        | 10/50 [00:07<00:27,  1.47it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.64it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.15it/s]
 30%|███       | 15/50 [00:09<00:17,  2.04it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.05it/s]
 34%|███▍      | 17/50 [00:10<00:13,  2.43it/s]
 36%|███▌      | 18/50 [00:10<00:11,  2.91it/s]
 38%|███▊      | 19/50 [00:10<00:09,  3.43it/s]
 40%|████      | 20/50 [00:10<00:07,  3.94it/s]
 42%|████▏     | 21/50 [00:11<00:15,  1.89it/s]
 44%|████▍     | 22/50 [00:12<00:13,  2.01it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.27it/s]
 52%|█████▏    | 26/50 [00:13<00:10,  2.25it/s]
 54%|█████▍    | 27/50 [00:14<00:10,  2.18it/s]
 56%|█████▌    | 28/50 [00:14<00:08,  2.53it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.09it/s]
 62%|██████▏   | 31/50 [00:15<00:08,  2.19it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.21it/s]
 66%|██████▌   | 33/50 [00:16<00:06,  2.70it/s]
 68%|██████▊   | 34/50 [00:16<00:04,  3.33it/s]
 70%|███████   | 35/50 [00:17<00:08,  1.82it/s]
 72%|███████▏  | 36/50 [00:18<00:07,  1.90it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.76it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.91it/s]
 82%|████████▏ | 41/50 [00:21<00:05,  1.77it/s]
 84%|████████▍ | 42/50 [00:21<00:04,  1.86it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.44it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.53it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.24it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.38it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.12it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.34it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.14it/s]
+
+  clip 2/4: 45.5s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:38,  1.25it/s]
  6%|▌         | 3/50 [00:02<00:46,  1.01it/s]
  8%|▊         | 4/50 [00:03<00:36,  1.28it/s]
 10%|█         | 5/50 [00:04<00:43,  1.04it/s]
 12%|█▏        | 6/50 [00:05<00:35,  1.22it/s]
 14%|█▍        | 7/50 [00:06<00:41,  1.04it/s]
 16%|█▌        | 8/50 [00:06<00:33,  1.26it/s]
 18%|█▊        | 9/50 [00:07<00:23,  1.73it/s]
 20%|██        | 10/50 [00:08<00:31,  1.28it/s]
 22%|██▏       | 11/50 [00:08<00:26,  1.46it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.52it/s]
 30%|███       | 15/50 [00:10<00:17,  2.04it/s]
 32%|███▏      | 16/50 [00:10<00:16,  2.10it/s]
 34%|███▍      | 17/50 [00:10<00:12,  2.55it/s]
 36%|███▌      | 18/50 [00:10<00:10,  3.02it/s]
 40%|████      | 20/50 [00:10<00:06,  4.47it/s]
 42%|████▏     | 21/50 [00:12<00:13,  2.20it/s]
 44%|████▍     | 22/50 [00:12<00:12,  2.24it/s]
 48%|████▊     | 24/50 [00:12<00:07,  3.52it/s]
 52%|█████▏    | 26/50 [00:13<00:09,  2.45it/s]
 54%|█████▍    | 27/50 [00:14<00:09,  2.34it/s]
 58%|█████▊    | 29/50 [00:14<00:06,  3.34it/s]
 60%|██████    | 30/50 [00:14<00:05,  3.87it/s]
 62%|██████▏   | 31/50 [00:15<00:09,  2.08it/s]
 64%|██████▍   | 32/50 [00:16<00:08,  2.19it/s]
 66%|██████▌   | 33/50 [00:16<00:06,  2.74it/s]
 70%|███████   | 35/50 [00:17<00:07,  2.07it/s]
 72%|███████▏  | 36/50 [00:18<00:06,  2.12it/s]
 76%|███████▌  | 38/50 [00:19<00:06,  1.85it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.93it/s]
 82%|████████▏ | 41/50 [00:21<00:05,  1.79it/s]
 84%|██████��█▍ | 42/50 [00:21<00:04,  1.87it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.45it/s]
 88%|████████▊ | 44/50 [00:23<00:03,  1.55it/s]
 90%|█████████ | 45/50 [00:24<00:04,  1.25it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.44it/s]
 94%|█████████▍| 47/50 [00:26<00:02,  1.19it/s]
 96%|█████████▌| 48/50 [00:26<00:01,  1.42it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.18it/s]
+
+  clip 3/4: 46.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<01:00,  1.23s/it]
  4%|▍         | 2/50 [00:01<00:36,  1.31it/s]
  6%|▌         | 3/50 [00:02<00:45,  1.03it/s]
  8%|▊         | 4/50 [00:03<00:36,  1.28it/s]
 10%|█         | 5/50 [00:04<00:42,  1.05it/s]
 12%|█▏        | 6/50 [00:05<00:35,  1.25it/s]
 14%|█▍        | 7/50 [00:06<00:40,  1.06it/s]
 16%|█▌        | 8/50 [00:06<00:34,  1.21it/s]
 18%|█▊        | 9/50 [00:07<00:26,  1.56it/s]
 20%|██        | 10/50 [00:08<00:33,  1.20it/s]
 22%|██▏       | 11/50 [00:08<00:28,  1.39it/s]
 24%|██▍       | 12/50 [00:09<00:20,  1.83it/s]
 26%|██▌       | 13/50 [00:09<00:15,  2.41it/s]
 28%|██▊       | 14/50 [00:09<00:11,  3.05it/s]
 30%|███       | 15/50 [00:10<00:21,  1.64it/s]
 32%|███▏      | 16/50 [00:11<00:19,  1.76it/s]
 34%|███▍      | 17/50 [00:11<00:14,  2.24it/s]
 36%|███▌      | 18/50 [00:11<00:11,  2.69it/s]
 38%|███▊      | 19/50 [00:11<00:10,  2.99it/s]
 40%|████      | 20/50 [00:11<00:09,  3.14it/s]
 42%|████▏     | 21/50 [00:13<00:17,  1.63it/s]
 44%|████▍     | 22/50 [00:13<00:16,  1.69it/s]
 46%|████▌     | 23/50 [00:13<00:12,  2.18it/s]
 48%|████▊     | 24/50 [00:14<00:09,  2.83it/s]
 52%|█████▏    | 26/50 [00:15<00:11,  2.06it/s]
 54%|█████▍    | 27/50 [00:15<00:10,  2.09it/s]
 56%|█████▌    | 28/50 [00:15<00:08,  2.63it/s]
 58%|█████▊    | 29/50 [00:15<00:06,  3.23it/s]
 60%|██████    | 30/50 [00:16<00:05,  3.63it/s]
 62%|██████▏   | 31/50 [00:17<00:10,  1.81it/s]
 64%|██████▍   | 32/50 [00:17<00:09,  1.85it/s]
 66%|██████▌   | 33/50 [00:18<00:07,  2.34it/s]
 68%|██████▊   | 34/50 [00:18<00:05,  2.88it/s]
 70%|███████   | 35/50 [00:19<00:09,  1.64it/s]
 72%|███████▏  | 36/50 [00:19<00:07,  1.85it/s]
 76%|███████▌  | 38/50 [00:21<00:06,  1.73it/s]
 78%|███████▊  | 39/50 [00:21<00:05,  1.90it/s]
 82%|████████▏ | 41/50 [00:22<00:05,  1.78it/s]
 84%|████████▍ | 42/50 [00:23<00:04,  1.93it/s]
 86%|████████▌ | 43/50 [00:24<00:04,  1.47it/s]
 88%|████████▊ | 44/50 [00:24<00:03,  1.62it/s]
 90%|█████████ | 45/50 [00:25<00:03,  1.30it/s]
 92%|█████████▏| 46/50 [00:26<00:02,  1.44it/s]
 94%|█████████▍| 47/50 [00:27<00:02,  1.16it/s]
 96%|█████████▌| 48/50 [00:28<00:01,  1.39it/s]
 98%|█████████▊| 49/50 [00:29<00:00,  1.16it/s]
+
+  clip 4/4: 46.7s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000016/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000016/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 13074.8ms  total=  13.1s
+  clip_total         n=  4  mean= 44059.6ms  total= 176.2s
 
+Done in 189.4s

3drawers/gen_flex/chunk_001/_logs/b3_ep00000017.log

@@ -1,9 +1,3 @@
-/svl/u/ravenh/miniconda3/envs/mlspaces/lib/python3.11/site-packages/torch/cuda/__init__.py:287: UserWarning: 
-NVIDIA RTX PRO 6000 Blackwell Server Edition with CUDA capability sm_120 is not compatible with the current PyTorch installation.
-The current PyTorch install supports CUDA capabilities sm_50 sm_60 sm_70 sm_75 sm_80 sm_86 sm_90.
-If you want to use the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
-
-  warnings.warn(
 Views: ['ext1', 'wrist']
 Prompt: This video shows in the background, an office breakroom with a coffee area, with bright white walls, medium gray polished concrete floor, and overhead LED panels casting cool even light across the room. In the foreground, there is a matte black three-drawer dresser with slim leather pull tabs, and a chrome coffee machine and a neat stack of paper cups are arranged on the laminate counter. The robot is instructed to open the top drawer of the dresser.
 TeaCache thresh: 0.1
@@ -40,34 +34,76 @@ Loaded model: {
 }
 Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth"
 Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
-Traceback (most recent call last):
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 154, in <module>
-    main()
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/sim2realgen/inference_long_video_fast.py", line 90, in main
-    pipe = WanVideoPipeline.from_pretrained(
-           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/pipelines/wan_video.py", line 134, in from_pretrained
-    model_pool = pipe.download_and_load_models(model_configs, vram_limit)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/diffusion/base_pipeline.py", line 293, in download_and_load_models
-    model_pool.auto_load_model(
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 70, in auto_load_model
-    model = self.load_model_file(config, path, vram_config, vram_limit=vram_limit)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/models/model_loader.py", line 40, in load_model_file
-    model = load_model(
-            ^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/loader/model.py", line 46, in load_model
-    state_dict = state_dict_converter(state_dict)
-                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/utils/state_dict_converters/wan_video_vae.py", line 6, in WanVideoVAEStateDictConverter
-    state_dict_['model.' + name] = state_dict[name]
-                                   ~~~~~~~~~~^^^^^^
-  File "/svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/diffsynth/core/vram/disk_map.py", line 64, in __getitem__
-    param = param.to(self.torch_dtype)
-            ^^^^^^^^^^^^^^^^^^^^^^^^^^
-RuntimeError: CUDA error: no kernel image is available for execution on the device
-CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
-For debugging consider passing CUDA_LAUNCH_BLOCKING=1
-Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
+Loaded model: {
+    "model_name": "wan_video_vae",
+    "model_class": "diffsynth.models.wan_video_vae.WanVideoVAE",
+    "extra_kwargs": null
+}
+Loading models from: "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth"
+Detected non-safetensors files, which may cause slower loading. It's recommended to convert it to a safetensors file.
+Loaded model: {
+    "model_name": "wan_video_image_encoder",
+    "model_class": "diffsynth.models.wan_video_image_encoder.WanImageEncoder",
+    "extra_kwargs": null
+}
+Using wan_video_text_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_t5_umt5-xxl-enc-bf16.pth".
+Using wan_video_dit from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/diffusion_pytorch_model.safetensors".
+Using wan_video_vae from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/Wan2.1_VAE.pth".
+Using wan_video_image_encoder from "./models/PAI/Wan2.1-Fun-V1.1-1.3B-Control/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth".
+No wan_video_motion_controller models available. This is not an error.
+No wan_video_vace models available. This is not an error.
+No wan_video_vap models available. This is not an error.
+No wans2v_audio_encoder models available. This is not an error.
+No wan_video_animate_adapter models available. This is not an error.
+Loading LoRA from: /svl/u/ravenh/renderscale/Sim2RealGen_Video_Model/lora_ckpts/Wan2.1-Fun-V1.1-1.3B-Mixed-DROID-AGIBOT-3View-DualControl-PrevFrame-droidnew_continue_lora/step-5294.safetensors
+300 tensors are fused by LoRA. Fused LoRA layers cannot be cleared by `pipe.clear_lora()`.
+Setting up multi-view pipeline...
+
+===== source: video0 =====
+  total_frames=271  num_clips=4
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:54,  1.11s/it]
  4%|▍         | 2/50 [00:01<00:33,  1.42it/s]
  6%|▌         | 3/50 [00:02<00:42,  1.11it/s]
  8%|▊         | 4/50 [00:03<00:31,  1.45it/s]
 10%|█         | 5/50 [00:04<00:37,  1.19it/s]
 12%|█▏        | 6/50 [00:04<00:29,  1.49it/s]
 14%|█▍        | 7/50 [00:05<00:34,  1.23it/s]
 16%|█▌        | 8/50 [00:05<00:27,  1.51it/s]
 20%|██        | 10/50 [00:07<00:24,  1.62it/s]
 22%|██▏       | 11/50 [00:07<00:21,  1.82it/s]
 28%|██▊       | 14/50 [00:07<00:10,  3.46it/s]
 30%|███       | 15/50 [00:08<00:15,  2.23it/s]
 32%|███▏      | 16/50 [00:08<00:14,  2.35it/s]
 38%|███▊      | 19/50 [00:09<00:07,  4.14it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.88it/s]
 44%|████▍     | 22/50 [00:10<00:09,  2.89it/s]
 50%|█████     | 25/50 [00:10<00:05,  4.65it/s]
 54%|█████▍    | 27/50 [00:12<00:08,  2.74it/s]
 58%|█████▊    | 29/50 [00:12<00:05,  3.68it/s]
 62%|██████▏   | 31/50 [00:13<00:06,  2.75it/s]
 64%|██████▍   | 32/50 [00:13<00:06,  2.78it/s]
 68%|██████▊   | 34/50 [00:13<00:04,  3.74it/s]
 70%|███████   | 35/50 [00:15<00:06,  2.32it/s]
 72%|███████▏  | 36/50 [00:15<00:05,  2.38it/s]
 76%|███████▌  | 38/50 [00:16<00:05,  2.09it/s]
 78%|███████▊  | 39/50 [00:16<00:04,  2.23it/s]
 82%|████████▏ | 41/50 [00:18<00:04,  2.01it/s]
 84%|████████▍ | 42/50 [00:18<00:03,  2.15it/s]
 86%|████████▌ | 43/50 [00:19<00:04,  1.63it/s]
 88%|████████▊ | 44/50 [00:19<00:03,  1.84it/s]
 90%|█████████ | 45/50 [00:21<00:03,  1.45it/s]
 92%|█████████▏| 46/50 [00:21<00:02,  1.68it/s]
 94%|█████████▍| 47/50 [00:22<00:02,  1.36it/s]
 96%|█████████▌| 48/50 [00:22<00:01,  1.61it/s]
 98%|█████████▊| 49/50 [00:23<00:00,  1.32it/s]
+
+  clip 1/4: 43.8s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:35,  1.34it/s]
  6%|▌         | 3/50 [00:02<00:44,  1.06it/s]
  8%|▊         | 4/50 [00:03<00:34,  1.32it/s]
 10%|█         | 5/50 [00:04<00:41,  1.08it/s]
 12%|█▏        | 6/50 [00:04<00:32,  1.35it/s]
 14%|█▍        | 7/50 [00:06<00:38,  1.12it/s]
 16%|█▌        | 8/50 [00:06<00:30,  1.38it/s]
 20%|██        | 10/50 [00:07<00:26,  1.49it/s]
 22%|██▏       | 11/50 [00:08<00:23,  1.68it/s]
 28%|██▊       | 14/50 [00:08<00:11,  3.23it/s]
 30%|███       | 15/50 [00:09<00:16,  2.07it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.17it/s]
 38%|███▊      | 19/50 [00:09<00:08,  3.82it/s]
 42%|████▏     | 21/50 [00:11<00:11,  2.59it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.60it/s]
 46%|████▌     | 23/50 [00:11<00:08,  3.07it/s]
 48%|████▊     | 24/50 [00:11<00:07,  3.64it/s]
 50%|█████     | 25/50 [00:11<00:06,  4.06it/s]
 52%|█████▏    | 26/50 [00:13<00:12,  1.99it/s]
 54%|█████▍    | 27/50 [00:13<00:10,  2.14it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.18it/s]
 60%|██████    | 30/50 [00:14<00:06,  3.20it/s]
 62%|██████▏   | 31/50 [00:15<00:10,  1.79it/s]
 64%|██████▍   | 32/50 [00:15<00:09,  1.89it/s]
 68%|██████▊   | 34/50 [00:16<00:05,  2.94it/s]
 70%|███████   | 35/50 [00:17<00:08,  1.80it/s]
 72%|███████▏  | 36/50 [00:17<00:07,  1.90it/s]
 76%|███████▌  | 38/50 [00:18<00:06,  1.75it/s]
 78%|███████▊  | 39/50 [00:19<00:05,  1.90it/s]
 82%|████████▏ | 41/50 [00:20<00:05,  1.78it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.92it/s]
 86%|████████▌ | 43/50 [00:22<00:04,  1.46it/s]
 88%|████████▊ | 44/50 [00:22<00:03,  1.62it/s]
 90%|█████████ | 45/50 [00:23<00:03,  1.28it/s]
 92%|█████████▏| 46/50 [00:24<00:02,  1.45it/s]
 94%|█████████▍| 47/50 [00:25<00:02,  1.18it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.40it/s]
 98%|█████████▊| 49/50 [00:27<00:00,  1.16it/s]
+
+  clip 2/4: 50.0s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.40it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.08it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.41it/s]
 10%|█         | 5/50 [00:04<00:39,  1.14it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.40it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:24,  1.62it/s]
 26%|██▌       | 13/50 [00:08<00:14,  2.62it/s]
 30%|███       | 15/50 [00:09<00:16,  2.12it/s]
 32%|███▏      | 16/50 [00:09<00:15,  2.20it/s]
 36%|███▌      | 18/50 [00:09<00:09,  3.28it/s]
 42%|████▏     | 21/50 [00:11<00:10,  2.78it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.71it/s]
 48%|████▊     | 24/50 [00:11<00:06,  3.75it/s]
 52%|█████▏    | 26/50 [00:12<00:09,  2.62it/s]
 54%|█████▍    | 27/50 [00:13<00:08,  2.62it/s]
 60%|██████    | 30/50 [00:13<00:04,  4.28it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.50it/s]
 64%|██████▍   | 32/50 [00:14<00:07,  2.54it/s]
 68%|██████▊   | 34/50 [00:15<00:04,  3.69it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.16it/s]
 72%|███████▏  | 36/50 [00:16<00:06,  2.20it/s]
 74%|███████▍  | 37/50 [00:16<00:05,  2.48it/s]
 76%|███████▌  | 38/50 [00:18<00:07,  1.65it/s]
 78%|███████▊  | 39/50 [00:18<00:06,  1.76it/s]
 80%|████████  | 40/50 [00:18<00:04,  2.25it/s]
 82%|████████▏ | 41/50 [00:19<00:05,  1.51it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.69it/s]
 86%|████████▌ | 43/50 [00:21<00:05,  1.29it/s]
 88%|████████▊ | 44/50 [00:22<00:04,  1.48it/s]
 90%|█████████ | 45/50 [00:23<00:04,  1.20it/s]
 92%|█████��███▏| 46/50 [00:23<00:02,  1.40it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.15it/s]
 96%|█████████▌| 48/50 [00:25<00:01,  1.32it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.08it/s]
+
+  clip 3/4: 48.1s
+
+
+
+
+
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:59,  1.21s/it]
  4%|▍         | 2/50 [00:01<00:34,  1.39it/s]
  6%|▌         | 3/50 [00:02<00:43,  1.07it/s]
  8%|▊         | 4/50 [00:03<00:32,  1.40it/s]
 10%|█         | 5/50 [00:04<00:39,  1.13it/s]
 12%|█▏        | 6/50 [00:04<00:31,  1.41it/s]
 14%|█▍        | 7/50 [00:05<00:37,  1.16it/s]
 16%|█▌        | 8/50 [00:06<00:29,  1.41it/s]
 20%|██        | 10/50 [00:07<00:26,  1.50it/s]
 22%|██▏       | 11/50 [00:07<00:23,  1.69it/s]
 26%|██▌       | 13/50 [00:07<00:13,  2.74it/s]
 30%|███       | 15/50 [00:09<00:16,  2.18it/s]
 32%|███▏      | 16/50 [00:09<00:14,  2.27it/s]
 38%|███▊      | 19/50 [00:09<00:07,  3.93it/s]
 42%|████▏     | 21/50 [00:10<00:10,  2.72it/s]
 44%|████▍     | 22/50 [00:11<00:10,  2.71it/s]
 50%|█████     | 25/50 [00:11<00:05,  4.37it/s]
 54%|█████▍    | 27/50 [00:13<00:09,  2.49it/s]
 58%|█████▊    | 29/50 [00:13<00:06,  3.35it/s]
 62%|██████▏   | 31/50 [00:14<00:07,  2.54it/s]
 64%|██████▍   | 32/50 [00:14<00:06,  2.57it/s]
 70%|███████   | 35/50 [00:16<00:06,  2.47it/s]
 72%|███████▏  | 36/50 [00:16<00:05,  2.41it/s]
 76%|███████▌  | 38/50 [00:17<00:05,  2.05it/s]
 78%|███████▊  | 39/50 [00:18<00:05,  2.04it/s]
 82%|████████▏ | 41/50 [00:19<00:04,  1.82it/s]
 84%|████████▍ | 42/50 [00:20<00:04,  1.94it/s]
 86%|████████▌ | 43/50 [00:21<00:04,  1.51it/s]
 88%|████████▊ | 44/50 [00:21<00:03,  1.69it/s]
 90%|█████████ | 45/50 [00:22<00:03,  1.34it/s]
 92%|█████████▏| 46/50 [00:23<00:02,  1.52it/s]
 94%|█████████▍| 47/50 [00:24<00:02,  1.23it/s]
 96%|█████████▌| 48/50 [00:24<00:01,  1.44it/s]
 98%|█████████▊| 49/50 [00:26<00:00,  1.18it/s]
+
+  clip 4/4: 46.7s
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000017/generated_ext1.mp4
+
+To disable this warning, you can either:
+	- Avoid using `tokenizers` before the fork if possible
+	- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)
+
+  saved /scr/ravenh/3drawers_flex_w1/_workdir_gen/chunk_001/_inf_tmp/b3_ep00000017/generated_wrist.mp4
+
+=== timings ===
+  load_source        n=  1  mean= 14363.1ms  total=  14.4s
+  clip_total         n=  4  mean= 47139.6ms  total= 188.6s
 
+Done in 203.1s