tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-06-28 12:25:03 +00:00
Code Issues Packages Projects Releases Wiki Activity

convert : fix remote option in Windows (#14100)

Browse Source
This commit is contained in:
pqnet
2025-06-19 12:21:40 +02:00
committed by GitHub
parent faed5a5f5d
commit 5fc7856815
1 changed files with 11 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
convert_hf_to_gguf.py
View File

@ -6389,8 +6389,8 @@ def parse_args() -> argparse.Namespace:
help="model is executed on big endian machine", help="model is executed on big endian machine",
) )
parser.add_argument( parser.add_argument(
"model", type=Path, "model", type=str,
help="directory containing model file", help="directory containing model file or huggingface repository ID (if --remote)",
nargs="?", nargs="?",
) )
parser.add_argument( parser.add_argument(
@ -6493,18 +6493,20 @@ def main() -> None:
else: else:
logging.basicConfig(level=logging.INFO) logging.basicConfig(level=logging.INFO)
dir_model = args.model
if args.remote: if args.remote:
hf_repo_id = args.model
from huggingface_hub import snapshot_download from huggingface_hub import snapshot_download
local_dir = snapshot_download( local_dir = snapshot_download(
repo_id=str(dir_model), repo_id=hf_repo_id,
allow_patterns=["LICENSE", "*.json", "*.md", "*.txt", "tokenizer.model"]) allow_patterns=["LICENSE", "*.json", "*.md", "*.txt", "tokenizer.model"])
dir_model = Path(local_dir) dir_model = Path(local_dir)
logger.info(f"Downloaded config and tokenizer to {local_dir}") logger.info(f"Downloaded config and tokenizer to {local_dir}")
else:
hf_repo_id = None
dir_model = Path(args.model)
if not dir_model.is_dir(): if not dir_model.is_dir():
logger.error(f'Error: {args.model} is not a directory') logger.error(f'Error: {dir_model} is not a directory')
sys.exit(1) sys.exit(1)
ftype_map: dict[str, gguf.LlamaFileType] = { ftype_map: dict[str, gguf.LlamaFileType] = {
@ -6524,9 +6526,9 @@ def main() -> None:
if args.outfile is not None: if args.outfile is not None:
fname_out = args.outfile fname_out = args.outfile
elif args.remote: elif hf_repo_id:
# if remote, use the model ID as the output file name # if remote, use the model ID as the output file name
fname_out = Path("./" + str(args.model).replace("/", "-") + "-{ftype}.gguf") fname_out = Path("./" + hf_repo_id.replace("/", "-") + "-{ftype}.gguf")
else: else:
fname_out = dir_model fname_out = dir_model
@ -6555,7 +6557,7 @@ def main() -> None:
split_max_tensors=args.split_max_tensors, split_max_tensors=args.split_max_tensors,
split_max_size=split_str_to_n_bytes(args.split_max_size), dry_run=args.dry_run, split_max_size=split_str_to_n_bytes(args.split_max_size), dry_run=args.dry_run,
small_first_shard=args.no_tensor_first_split, small_first_shard=args.no_tensor_first_split,
remote_hf_model_id=str(args.model) if args.remote else None) remote_hf_model_id=hf_repo_id)
if args.vocab_only: if args.vocab_only:
logger.info("Exporting model vocab...") logger.info("Exporting model vocab...")