cleanup video code

Signed-off-by: Vladimir Mandic <mandic00@live.com>

TODO.md

@@ -8,9 +8,7 @@ Main ToDo list can be found at [GitHub projects](https://github.com/users/vladma
 
 - VLM Gemma3: requires `transformers==git+https://github.com/huggingface/transformers@v4.49.0-Gemma-3`  
 - VAE Remote encode: SD15 and Flux.1 issues: <https://github.com/huggingface/diffusers/issues/11069>  
-- Video: API support is TBD  
-- Video: Hunyuan Video I2V: transformers incompatibility <https://github.com/huggingface/diffusers/issues/11118>  
-- Video: Hunyuan Video I2V: add 16ch vs 33ch processing <https://github.com/huggingface/diffusers/pull/11066>  
+- Video: Hunyuan Video I2V: requires `transformers==4.47.1` <https://github.com/huggingface/diffusers/issues/11118>  
 - Video: Latte 1 T2V: dtype mismatch <https://github.com/huggingface/diffusers/issues/11137>  
 - Video: WAN 2.1 14B I2V 480p/720p: broken offload  
 - Video: CogVideoX 1.5 5B T2V/I2V: all-gray output  
@@ -24,8 +22,8 @@ Main ToDo list can be found at [GitHub projects](https://github.com/users/vladma
 - Video: add generate context menu
 - Video: FasterCache and PyramidAttentionBroadcast granular config  
 - Video: FasterCache and PyramidAttentionBroadcast for LTX and WAN <https://github.com/huggingface/diffusers/issues/11134>  
-- Video: OponSora v2 https://huggingface.co/hpcai-tech/Open-Sora-v2
-- Video: STG: https://github.com/huggingface/diffusers/blob/main/examples/community/README.md#spatiotemporal-skip-guidance
+- Video: API support
+- Video: STG: <https://github.com/huggingface/diffusers/blob/main/examples/community/README.md#spatiotemporal-skip-guidance>
 - Video SmoothCache: https://github.com/huggingface/diffusers/issues/11135
 - FasterCache, PyramidAttentionBroadcast, SmoothCache general support
 

installer.py

@@ -536,9 +536,9 @@ def check_python(supported_minors=[9, 10, 11, 12], reason=None):
 # check diffusers version
 def check_diffusers():
     t_start = time.time()
-    if args.skip_all or args.skip_git:
+    if args.skip_all or args.skip_git or args.experimental:
         return
-    sha = '5dbe4f5de6398159f8c2bedd371bc116683edbd3' # diffusers commit hash
+    sha = '1ddf3f3a19095344166ad7207ebc5be7a862d17e' # diffusers commit hash
     pkg = pkg_resources.working_set.by_key.get('diffusers', None)
     minor = int(pkg.version.split('.')[1] if pkg is not None else 0)
     cur = opts.get('diffusers_version', '') if minor > 0 else ''

modules/video_models/video_run.py

@@ -66,6 +66,7 @@ def generate(*args, **kwargs):
     video_vae.set_vae_params(p)
     video_cache.set_cache(faster_cache=faster_cache, pyramid_attention_broadcast=pyramid_attention)
     video_utils.set_prompt(p)
+    p.task_args['num_inference_steps'] = p.steps
     p.task_args['width'] = p.width
     p.task_args['height'] = p.height
     p.task_args['output_type'] = 'latent' if (p.vae_type == 'Remote') else 'pil'