mirror of
https://github.com/ggerganov/llama.cpp.git
synced 2024-12-25 13:58:46 +01:00
multimodal : add BakLLaVA conversion support (#3682)
This commit is contained in:
parent
60abea9798
commit
f3b25e4043
@ -16,13 +16,29 @@ checkpoint = torch.load(path)
|
|||||||
mm_tensors = [k for k, v in checkpoint.items() if k.startswith("model.mm_projector")]
|
mm_tensors = [k for k, v in checkpoint.items() if k.startswith("model.mm_projector")]
|
||||||
|
|
||||||
# store these tensors in a new dictionary and torch.save them
|
# store these tensors in a new dictionary and torch.save them
|
||||||
projector = {name: checkpoint[name] for name in mm_tensors}
|
projector = {name: checkpoint[name].float() for name in mm_tensors}
|
||||||
torch.save(projector, f"{args.model}/llava.projector")
|
torch.save(projector, f"{args.model}/llava.projector")
|
||||||
|
|
||||||
# remove these tensors from the checkpoint and save it again
|
# remove these tensors from the checkpoint and save it again
|
||||||
for name in mm_tensors:
|
for name in mm_tensors:
|
||||||
del checkpoint[name]
|
del checkpoint[name]
|
||||||
|
|
||||||
|
# BakLLaVA models contain CLIP tensors in it
|
||||||
|
clip_tensors = [k for k, v in checkpoint.items() if k.startswith("model.vision_tower")]
|
||||||
|
if len(clip_tensors) > 0:
|
||||||
|
clip = {name.replace("vision_tower.vision_tower.", ""): checkpoint[name].float() for name in clip_tensors}
|
||||||
|
torch.save(clip, f"{args.model}/llava.clip")
|
||||||
|
|
||||||
|
# remove these tensors
|
||||||
|
for name in clip_tensors:
|
||||||
|
del checkpoint[name]
|
||||||
|
|
||||||
|
# added tokens should be removed to be able to convert Mistral models
|
||||||
|
if os.path.exists(f"{args.model}/added_tokens.json"):
|
||||||
|
with open(f"{args.model}/added_tokens.json", "w") as f:
|
||||||
|
f.write("{}\n")
|
||||||
|
|
||||||
|
|
||||||
torch.save(checkpoint, path)
|
torch.save(checkpoint, path)
|
||||||
|
|
||||||
print("Done!")
|
print("Done!")
|
||||||
|
Loading…
Reference in New Issue
Block a user