JLNTV_IPD/CogVideo
4
0
Fork 0
mirror of https://github.com/THUDM/CogVideo.git synced 2025-06-17 13:59:17 +08:00
Code Issues Packages Projects Releases Wiki Activity

chore: code cleanup and parameter optimization

Browse Source 
- Remove redundant comments and debug information
- Adjust default parameters in training scripts
- Clean up code in lora_trainer and trainer implementations
This commit is contained in:
OleehyO 2025-01-13 11:53:13 +00:00
parent 954ba28d3c
commit 455b44a7b5
4 changed files with 3 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
finetune/models/cogvideox_t2v/lora_trainer.py
View File

@ -197,7 +197,6 @@ class CogVideoXT2VLoraTrainer(Trainer):
base_num_frames = num_frames base_num_frames = num_frames
else: else:
base_num_frames = (num_frames + transformer_config.patch_size_t - 1) // transformer_config.patch_size_t base_num_frames = (num_frames + transformer_config.patch_size_t - 1) // transformer_config.patch_size_t
breakpoint()
freqs_cos, freqs_sin = get_3d_rotary_pos_embed( freqs_cos, freqs_sin = get_3d_rotary_pos_embed(
embed_dim=transformer_config.attention_head_dim, embed_dim=transformer_config.attention_head_dim,
crops_coords=None, crops_coords=None,

2
finetune/train_zero_i2v.sh
View File

@ -49,7 +49,7 @@ SYSTEM_ARGS=(
CHECKPOINT_ARGS=( CHECKPOINT_ARGS=(
--checkpointing_steps 10 --checkpointing_steps 10
--checkpointing_limit 2 --checkpointing_limit 2
--resume_from_checkpoint "/absolute/path/to/checkpoint_dir" # if you want to resume from a checkpoint, otherwise, comment this line # --resume_from_checkpoint "/absolute/path/to/checkpoint_dir" # if you want to resume from a checkpoint, otherwise, comment this line
) )
# Validation Configuration # Validation Configuration

2
finetune/train_zero_t2v.sh
View File

@ -48,7 +48,7 @@ SYSTEM_ARGS=(
CHECKPOINT_ARGS=( CHECKPOINT_ARGS=(
--checkpointing_steps 10 --checkpointing_steps 10
--checkpointing_limit 2 --checkpointing_limit 2
--resume_from_checkpoint "/absolute/path/to/checkpoint_dir" # if you want to resume from a checkpoint, otherwise, comment this line # --resume_from_checkpoint "/absolute/path/to/checkpoint_dir" # if you want to resume from a checkpoint, otherwise, comment this line
) )
# Validation Configuration # Validation Configuration

12
finetune/trainer.py
View File

@ -758,16 +758,6 @@ class Trainer:
self.accelerator.register_save_state_pre_hook(save_model_hook) self.accelerator.register_save_state_pre_hook(save_model_hook)
self.accelerator.register_load_state_pre_hook(load_model_hook) self.accelerator.register_load_state_pre_hook(load_model_hook)
# def __maybe_save_checkpoint(self, global_step: int, must_save: bool = False):
# if self.accelerator.distributed_type == DistributedType.DEEPSPEED or self.accelerator.is_main_process:
# if must_save or global_step % self.args.checkpointing_steps == 0:
# save_path = get_intermediate_ckpt_path(
# checkpointing_limit=self.args.checkpointing_limit,
# step=global_step,
# output_dir=self.args.output_dir,
# )
# self.accelerator.save_state(save_path, safe_serialization=True)
def __maybe_save_checkpoint(self, global_step: int, must_save: bool = False): def __maybe_save_checkpoint(self, global_step: int, must_save: bool = False):
if self.accelerator.distributed_type == DistributedType.DEEPSPEED or self.accelerator.is_main_process: if self.accelerator.distributed_type == DistributedType.DEEPSPEED or self.accelerator.is_main_process:
if must_save or global_step % self.args.checkpointing_steps == 0: if must_save or global_step % self.args.checkpointing_steps == 0:
@ -783,4 +773,4 @@ class Trainer:
pipe_save_path.mkdir(parents=True, exist_ok=True) pipe_save_path.mkdir(parents=True, exist_ok=True)
pipe.save_pretrained(pipe_save_path) pipe.save_pretrained(pipe_save_path)
del pipe del pipe
torch.cuda.empty_cache() free_memory()