423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-08-04 12:42:51 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix #2346

Browse Source 
Former-commit-id: 7b8f5029018f0481f7da83cc5ee4408d95c9beb2
This commit is contained in:
hiyouga 2024-03-20 17:56:33 +08:00
parent a9d85cf3c6
commit cf149bf43c
1 changed files with 12 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
src/llmtuner/model/patcher.py
View File

@ -103,14 +103,19 @@ def _get_quantization_dataset(tokenizer: "PreTrainedTokenizer", model_args: "Mod
return samples return samples
def _configure_attn_implementation(model_args: "ModelArguments", init_kwargs: Dict[str, Any]) -> None: def _configure_attn_implementation(
config: "PretrainedConfig", model_args: "ModelArguments", init_kwargs: Dict[str, Any]
) -> None:
if model_args.flash_attn: if model_args.flash_attn:
if is_flash_attn2_available(): if not is_flash_attn2_available():
logger.info("Using FlashAttention-2 for faster training and inference.")
init_kwargs["attn_implementation"] = "flash_attention_2"
else:
logger.warning("FlashAttention2 is not installed.") logger.warning("FlashAttention2 is not installed.")
init_kwargs["attn_implementation"] = None return
logger.info("Using FlashAttention-2 for faster training and inference.")
if getattr(config, "model_type", None) == "internlm2": # special case for custom models
setattr(config, "attn_implementation", "flash_attention_2")
else:
init_kwargs["attn_implementation"] = "flash_attention_2"
else: else:
init_kwargs["attn_implementation"] = "eager" init_kwargs["attn_implementation"] = "eager"
@ -283,7 +288,7 @@ def patch_config(
for dtype_name, dtype in [("fp16", torch.float16), ("bf16", torch.bfloat16), ("fp32", torch.float32)]: for dtype_name, dtype in [("fp16", torch.float16), ("bf16", torch.bfloat16), ("fp32", torch.float32)]:
setattr(config, dtype_name, model_args.compute_dtype == dtype) setattr(config, dtype_name, model_args.compute_dtype == dtype)
_configure_attn_implementation(model_args, init_kwargs) _configure_attn_implementation(config, model_args, init_kwargs)
_configure_rope(config, model_args, is_trainable) _configure_rope(config, model_args, is_trainable)
_configure_longlora(config, model_args, is_trainable) _configure_longlora(config, model_args, is_trainable)
_configure_quantization(config, tokenizer, model_args, init_kwargs) _configure_quantization(config, tokenizer, model_args, init_kwargs)