adding hunyuan hf (support lora finetuning); unified hunyuan hf inference with quantization #135
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This reverts commit 580575c.
jzhang38
reviewed
Jan 13, 2025
| initialize_sequence_parallel_state(world_size) | ||
|
|
||
|
|
||
| def inference(args): |
There was a problem hiding this comment.
Why separate inference and inference quantization functions?
jzhang38
reviewed
Jan 13, 2025
fastvideo/train.py
Outdated
| # print("2222222222222222222222222222222222222222222222") | ||
| # print(type(latents_attention_mask)) | ||
| # print(latents_attention_mask) | ||
| training_guidance = torch.tensor([1000.0], |
There was a problem hiding this comment.
training guidance only works with Hunyuan. Adding this would make Mochi mal function,
fastvideo/utils/checkpoint.py
Outdated
| @@ -46,6 +45,14 @@ def save_checkpoint_optimizer(model, | |||
| "diffusion_pytorch_model.safetensors") | |||
| save_file(cpu_state, weight_path) | |||
| config_dict = dict(model.config) | |||
| config_dict.pop('dtype') | |||
| # dtype = config_dict['dtype'] | |||
| @@ -1,19 +1,19 @@ | |||
| #!/bin/bash | |||
|
|
|||
| num_gpus=1 | |||
| export MODEL_BASE="data/FastHunyuan-diffusers" | |||
| export MODEL_BASE="data/FastHunyuan" | |||
There was a problem hiding this comment.
Should be FastHunyuan-diffusers?
| --pretrained_model_name_or_path data/hunyuan_diffusers \ | ||
| --model_type hunyuan_hf \ | ||
| --cache_dir data/.cache \ | ||
| --data_json_path data/Image-Vid-Finetune-HunYuan/videos2caption.json \ |
There was a problem hiding this comment.
This dataset path should be set to the 22 videos?
| --cfg 0.0 \ | ||
| --ema_decay 0.999 \ | ||
| --log_validation \ | ||
| --output_dir data/outputs/HSH-Taylor-Finetune-Hunyuan \ |
jzhang38
approved these changes
Jan 14, 2025
bobboli
pushed a commit
to bobboli/FastVideo
that referenced
this pull request
Mar 5, 2025
…ence with quantization (hao-ai-lab#135)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
adding hunyuan hf (support lora finetuning); unified hunyuan hf inference with quantization