Mirrors/text-generation-webui
1
0
Fork 0
mirror of https://github.com/oobabooga/text-generation-webui.git synced 2024-12-23 21:18:00 +01:00
Code Issues Packages Projects Releases Wiki Activity

No need to have bf16 support here

Browse Source
This commit is contained in:
oobabooga 2023-02-20 15:12:42 -03:00
parent 8123c5b01a
commit c1de491c63
1 changed files with 1 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
convert-to-safetensors.py
View File

@ -23,7 +23,6 @@ parser = argparse.ArgumentParser(formatter_class=lambda prog: argparse.HelpForma
parser.add_argument('MODEL', type=str, default=None, nargs='?', help="Path to the input model.")
parser.add_argument('--output', type=str, default=None, help='Path to the output folder (default: models/{model_name}_safetensors).')
parser.add_argument("--max-shard-size", type=str, default="2GB", help="Maximum size of a shard in GB or MB (default: %(default)s).")
parser.add_argument('--bf16', action='store_true', help='Load the model with bfloat16 precision. Requires NVIDIA Ampere GPU.')
args = parser.parse_args()
if __name__ == '__main__':
@ -31,7 +30,7 @@ if __name__ == '__main__':
model_name = path.name
print(f"Loading {model_name}...")
model = AutoModelForCausalLM.from_pretrained(path, low_cpu_mem_usage=True, torch_dtype=torch.bfloat16 if args.bf16 else torch.float16)
model = AutoModelForCausalLM.from_pretrained(path, low_cpu_mem_usage=True, torch_dtype=torch.float16)
tokenizer = AutoTokenizer.from_pretrained(path)
out_folder = args.output or Path(f"models/{model_name}_safetensors")