modelscope / modelscope/DiffSynth-Studio

关于保存模型

Open
#973 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
13.1k
Forks
1.3k
Avg merge
13h 12m
Merged PRs (30d)
45

Description

### ./diffsynth/trainers/utils.py

class ModelLogger:
    def __init__(self, output_path, remove_prefix_in_ckpt=None, state_dict_converter=lambda x:x):
        self.output_path = output_path
        self.remove_prefix_in_ckpt = remove_prefix_in_ckpt
        self.state_dict_converter = state_dict_converter
        self.num_steps = 0

    def on_step_end(self, accelerator, model, save_steps=None):
        self.num_steps += 1
        if save_steps is not None and self.num_steps % save_steps == 0:
            self.save_model(accelerator, model, f"step-{self.num_steps}.safetensors")

    def on_epoch_end(self, accelerator, model, epoch_id):
        accelerator.wait_for_everyone()
        if accelerator.is_main_process:
            state_dict = accelerator.get_state_dict(model)
            state_dict = accelerator.unwrap_model(model).export_trainable_state_dict(state_dict, remove_prefix=self.remove_prefix_in_ckpt)
            state_dict = self.state_dict_converter(state_dict)
            os.makedirs(self.output_path, exist_ok=True)
            path = os.path.join(self.output_path, f"epoch-{epoch_id}.safetensors")
            accelerator.save(state_dict, path, safe_serialization=True)

    def on_training_end(self, accelerator, model, save_steps=None):
        if save_steps is not None and self.num_steps % save_steps != 0:
            self.save_model(accelerator, model, f"step-{self.num_steps}.safetensors")

    def save_model(self, accelerator, model, file_name):
        accelerator.wait_for_everyone()
        if accelerator.is_main_process:
            state_dict = accelerator.get_state_dict(model)
            state_dict = accelerator.unwrap_model(model).export_trainable_state_dict(state_dict, remove_prefix=self.remove_prefix_in_ckpt)
            state_dict = self.state_dict_converter(state_dict)
            os.makedirs(self.output_path, exist_ok=True)
            path = os.path.join(self.output_path, file_name)
            accelerator.save(state_dict, path, safe_serialization=True)

保存模型的这段代码,目前好像仅支持保存DiT的safetensors,请问将来是否会增加保存text encoder的功能呢?

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start in diffsynth/trainers/utils.py with ModelLogger and trace how on_epoch_end and save_model obtain and export the model state. Determine whether text-encoder state is available through the current training and accelerator flow; done should include a defined way to save it alongside the existing model checkpoints, with relevant validation added.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.