ros/LlamaFactory
1
0
Fork 0
mirror of https://github.com/hiyouga/LlamaFactory.git synced 2026-02-03 08:53:38 +00:00
Code Issues Packages Projects Releases Wiki Activity

[model] unsloth resume from checkpoint bug (#8423)

Browse Source 
Co-authored-by: viyer <vivek_iyer2@apple.com>
This commit is contained in:
Vivek Iyer
2025-06-23 04:43:54 -04:00
committed by GitHub
parent 88a92be808
commit 7b252b2368
2 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/llamafactory/model/adapter.py
View File

@@ -188,7 +188,7 @@ def _setup_lora_tuning(
if adapter_to_resume is not None: # resume lora training
if model_args.use_unsloth:
model = load_unsloth_peft_model(config, model_args, is_trainable=is_trainable)
model = load_unsloth_peft_model(config, model_args, finetuning_args, is_trainable=is_trainable)
else:
model = PeftModel.from_pretrained(model, adapter_to_resume, is_trainable=is_trainable, **init_kwargs)