423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-10-14 23:58:11 +08:00
Code Issues Packages Projects Releases Wiki Activity

Fix slow model initialization in bfloat16 dtype.

Browse Source 
Former-commit-id: cf2e2f6f9b7f09b1e2faf6fbc413e3f62e3846c7
This commit is contained in:
ShaneTian 2023-12-21 21:25:20 +08:00
parent 67f7034a21
commit d05febe5de
1 changed files with 1 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/llmtuner/model/loader.py
View File

@ -67,10 +67,10 @@ def load_model_and_tokenizer(
model = AutoModelForCausalLM.from_pretrained(
model_args.model_name_or_path,
config=config,
torch_dtype=model_args.compute_dtype,
low_cpu_mem_usage=(not is_deepspeed_zero3_enabled()),
**config_kwargs
)
model = model.to(model_args.compute_dtype) if not getattr(model, "quantization_method", None) else model
patcher.patch_model(model)
register_autoclass(config, model, tokenizer)
if not is_deepspeed_zero3_enabled():
@ -95,7 +95,6 @@ def load_model_and_tokenizer(
if not is_trainable:
model.requires_grad_(False) # fix all model params
model = model.to(model_args.compute_dtype) if not getattr(model, "quantization_method", None) else model
model.eval()
else:
model.train()