vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

gguf-py : fix some metadata name extraction edge cases (#8591)

Browse source 
* gguf-py : fix some metadata name extraction edge cases

* convert_lora : use the lora dir for the model card path

* gguf-py : more metadata edge cases fixes

Multiple finetune versions are now joined together,
and the removal of the basename annotation on trailing versions
is more robust.

* gguf-py : add more name metadata extraction tests

* convert_lora : fix default filename

The default filename was previously hardcoded.

* convert_hf : Model.fname_out can no longer be None

* gguf-py : do not use title case for naming convention

Some models use acronyms in lowercase,
which can't be title-cased like other words,
so it's best to simply use the same case
as in the original model name.

Note that the size label still has an uppercased suffix
to make it distinguishable from the context size of a finetune.
This commit is contained in:
compilade 2024-07-20 21:58:49 -04:00 • committed by GitHub
parent c69c63039c
commit 328884f421
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
5 changed files with 112 additions and 44 deletions
Download patch file Download diff file Expand all files Collapse all files

26
convert_lora_to_gguf.py
View file

@ -290,7 +290,7 @@ if __name__ == '__main__':
fname_out = args.outfile
else:
# output in the same directory as the model by default
fname_out = dir_lora / 'ggml-lora-{ftype}.gguf'
fname_out = dir_lora
if os.path.exists(input_model):
# lazy import load_file only if lora is in safetensors format.
@ -304,12 +304,6 @@ if __name__ == '__main__':
# load base model
logger.info(f"Loading base model: {dir_base_model.name}")
hparams = Model.load_hparams(dir_base_model)
with open(lora_config, "r") as f:
lparams: dict[str, Any] = json.load(f)
alpha: float = lparams["lora_alpha"]
with torch.inference_mode():
try:
model_class = Model.from_model_architecture(hparams["architectures"][0])
@ -320,12 +314,21 @@ if __name__ == '__main__':
class LoraModel(model_class):
model_arch = model_class.model_arch
lora_alpha: float
def __init__(self, *args, dir_lora_model: Path, lora_alpha: float, **kwargs):
super().__init__(*args, **kwargs)
self.dir_model_card = dir_lora_model
self.lora_alpha = float(lora_alpha)
def set_type(self):
self.gguf_writer.add_type(gguf.GGUFType.ADAPTER)
self.gguf_writer.add_string(gguf.Keys.Adapter.TYPE, "lora")
def set_gguf_parameters(self):
self.gguf_writer.add_float32(gguf.Keys.Adapter.LORA_ALPHA, float(alpha))
self.gguf_writer.add_float32(gguf.Keys.Adapter.LORA_ALPHA, self.lora_alpha)
super().set_gguf_parameters()
def get_tensors(self) -> Iterator[tuple[str, Tensor]]:
@ -368,6 +371,11 @@ if __name__ == '__main__':
yield (dest_name + ".lora_a", lora_a)
yield (dest_name + ".lora_b", lora_b)
with open(lora_config, "r") as f:
lparams: dict[str, Any] = json.load(f)
alpha: float = lparams["lora_alpha"]
model_instance = LoraModel(
dir_base_model,
ftype,
@ -376,6 +384,8 @@ if __name__ == '__main__':
use_temp_file=False,
eager=args.no_lazy,
dry_run=args.dry_run,
dir_lora_model=dir_lora,
lora_alpha=alpha,
)
logger.info("Exporting model...")