p04798526
/
LLaMA-Factory-Mirror
1
0
Fork 1
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Merge pull request #4080 from MengqingCao/npu 

Add npu option for model exporting
This commit is contained in:
hoshi-hiyouga 2024-06-06 03:15:44 +08:00 committed by GitHub
parent 0e740aa463 feaee36c46
commit ca459f67eb
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
src/llamafactory/hparams/model_args.py
View File

@ -153,9 +153,9 @@ class ModelArguments:
default=1,
metadata={"help": "The file shard size (in GB) of the exported model."},
)
export_device: Literal["cpu", "cuda"] = field(
export_device: Literal["cpu", "auto"] = field(
default="cpu",
metadata={"help": "The device used in model export, use cuda to avoid addmm errors."},
metadata={"help": "The device used in model export, use `auto` to accelerate exporting."},
)
export_quantization_bit: Optional[int] = field(
default=None,

2
src/llamafactory/webui/components/export.py
View File

@ -89,7 +89,7 @@ def create_export_tab(engine: "Engine") -> Dict[str, "Component"]:
export_size = gr.Slider(minimum=1, maximum=100, value=1, step=1)
export_quantization_bit = gr.Dropdown(choices=["none"] + GPTQ_BITS, value="none")
export_quantization_dataset = gr.Textbox(value="data/c4_demo.json")
export_device = gr.Radio(choices=["cpu", "cuda"], value="cpu")
export_device = gr.Radio(choices=["cpu", "auto"], value="cpu")
export_legacy_format = gr.Checkbox()
with gr.Row():