mirror of
https://github.com/hiyouga/LLaMA-Factory.git
synced 2025-08-19 12:12:48 +08:00
[feature] add support for dft loss (#8917)
This commit is contained in:
parent
2b66b4df43
commit
0454c10456
43
examples/extras/dft/qwen2_full_sft.yaml
Normal file
43
examples/extras/dft/qwen2_full_sft.yaml
Normal file
@ -0,0 +1,43 @@
|
|||||||
|
### model
|
||||||
|
model_name_or_path: Qwen/Qwen2-1.5B-Instruct
|
||||||
|
trust_remote_code: true
|
||||||
|
|
||||||
|
### method
|
||||||
|
stage: sft
|
||||||
|
do_train: true
|
||||||
|
finetuning_type: full
|
||||||
|
use_dft_loss: true
|
||||||
|
|
||||||
|
### dataset
|
||||||
|
dataset: identity,alpaca_en_demo
|
||||||
|
template: qwen
|
||||||
|
cutoff_len: 2048
|
||||||
|
max_samples: 1000
|
||||||
|
overwrite_cache: true
|
||||||
|
preprocessing_num_workers: 16
|
||||||
|
dataloader_num_workers: 4
|
||||||
|
|
||||||
|
### output
|
||||||
|
output_dir: saves/qwen2-1_5b/full/sft
|
||||||
|
logging_steps: 10
|
||||||
|
save_steps: 500
|
||||||
|
plot_loss: true
|
||||||
|
overwrite_output_dir: true
|
||||||
|
save_only_model: false
|
||||||
|
report_to: none # choices: [none, wandb, tensorboard, swanlab, mlflow]
|
||||||
|
|
||||||
|
### train
|
||||||
|
per_device_train_batch_size: 1
|
||||||
|
gradient_accumulation_steps: 8
|
||||||
|
learning_rate: 1.0e-5
|
||||||
|
num_train_epochs: 3.0
|
||||||
|
lr_scheduler_type: cosine
|
||||||
|
warmup_ratio: 0.1
|
||||||
|
bf16: true
|
||||||
|
ddp_timeout: 180000000
|
||||||
|
|
||||||
|
### eval
|
||||||
|
# val_size: 0.1
|
||||||
|
# per_device_eval_batch_size: 1
|
||||||
|
# eval_strategy: steps
|
||||||
|
# eval_steps: 500
|
@ -428,6 +428,10 @@ class FinetuningArguments(
|
|||||||
default=False,
|
default=False,
|
||||||
metadata={"help": "Whether or not to use the Muon optimizer."},
|
metadata={"help": "Whether or not to use the Muon optimizer."},
|
||||||
)
|
)
|
||||||
|
use_dft_loss: bool = field(
|
||||||
|
default=False,
|
||||||
|
metadata={"help": "Whether to use the DFT loss."},
|
||||||
|
)
|
||||||
freeze_vision_tower: bool = field(
|
freeze_vision_tower: bool = field(
|
||||||
default=True,
|
default=True,
|
||||||
metadata={"help": "Whether ot not to freeze the vision tower in MLLM training."},
|
metadata={"help": "Whether ot not to freeze the vision tower in MLLM training."},
|
||||||
|
@ -78,6 +78,11 @@ class CustomSeq2SeqTrainer(Seq2SeqTrainer):
|
|||||||
self.accelerator.clip_grad_norm_ = MethodType(clip_grad_norm_old_version, self.accelerator)
|
self.accelerator.clip_grad_norm_ = MethodType(clip_grad_norm_old_version, self.accelerator)
|
||||||
self.add_callback(BAdamCallback)
|
self.add_callback(BAdamCallback)
|
||||||
|
|
||||||
|
if finetuning_args.use_dft_loss:
|
||||||
|
from ..trainer_utils import dft_loss_func
|
||||||
|
|
||||||
|
self.compute_loss_func = dft_loss_func
|
||||||
|
|
||||||
@override
|
@override
|
||||||
def create_optimizer(self) -> "torch.optim.Optimizer":
|
def create_optimizer(self) -> "torch.optim.Optimizer":
|
||||||
if self.optimizer is None:
|
if self.optimizer is None:
|
||||||
|
@ -631,6 +631,51 @@ def get_batch_logps(
|
|||||||
return logps, valid_length
|
return logps, valid_length
|
||||||
|
|
||||||
|
|
||||||
|
def dft_loss_func(outputs, labels, num_items_in_batch=None):
|
||||||
|
logits = outputs.get("logits")
|
||||||
|
if logits is None:
|
||||||
|
return outputs.get("loss", torch.tensor(0.0))
|
||||||
|
|
||||||
|
logits = logits.float()
|
||||||
|
vocab_size = logits.size(-1)
|
||||||
|
labels = torch.nn.functional.pad(labels, (0, 1), value=-100)
|
||||||
|
shift_labels = labels[..., 1:].contiguous()
|
||||||
|
logits = logits.view(-1, vocab_size)
|
||||||
|
shift_labels = shift_labels.view(-1)
|
||||||
|
shift_labels = shift_labels.to(logits.device)
|
||||||
|
|
||||||
|
loss = _dft_cross_entropy(logits, shift_labels, num_items_in_batch)
|
||||||
|
return loss
|
||||||
|
|
||||||
|
|
||||||
|
def _dft_cross_entropy(
|
||||||
|
source: torch.Tensor,
|
||||||
|
target: torch.Tensor,
|
||||||
|
num_items_in_batch: Optional[torch.Tensor] = None,
|
||||||
|
ignore_index: int = -100,
|
||||||
|
) -> torch.Tensor:
|
||||||
|
per_token_loss = torch.nn.functional.cross_entropy(source, target, ignore_index=ignore_index, reduction="none")
|
||||||
|
valid_mask = target != ignore_index
|
||||||
|
if not valid_mask.any():
|
||||||
|
return torch.tensor(0.0, device=source.device, dtype=source.dtype)
|
||||||
|
|
||||||
|
valid_losses = per_token_loss[valid_mask]
|
||||||
|
|
||||||
|
with torch.no_grad():
|
||||||
|
target_probs = torch.exp(-valid_losses)
|
||||||
|
|
||||||
|
weighted_losses = valid_losses * target_probs
|
||||||
|
|
||||||
|
if num_items_in_batch is not None:
|
||||||
|
total_loss = weighted_losses.sum()
|
||||||
|
if torch.is_tensor(num_items_in_batch):
|
||||||
|
num_items_in_batch = num_items_in_batch.to(total_loss.device)
|
||||||
|
loss = total_loss / num_items_in_batch
|
||||||
|
else:
|
||||||
|
loss = weighted_losses.mean()
|
||||||
|
return loss
|
||||||
|
|
||||||
|
|
||||||
def nested_detach(
|
def nested_detach(
|
||||||
tensors: Union["torch.Tensor", list["torch.Tensor"], tuple["torch.Tensor"], dict[str, "torch.Tensor"]],
|
tensors: Union["torch.Tensor", list["torch.Tensor"], tuple["torch.Tensor"], dict[str, "torch.Tensor"]],
|
||||||
clone: bool = False,
|
clone: bool = False,
|
||||||
|
Loading…
x
Reference in New Issue
Block a user