Spaces:
Paused
Paused
Commit ·
e2ac79e
1
Parent(s): 3541fc5
update
Browse files
examples/tutorials/dpo/ultrafeedback-dpo-unsloth/step_4_merge.py
CHANGED
|
@@ -19,7 +19,6 @@ else:
|
|
| 19 |
temp_directory = Path("/root/autodl-tmp/OpenMiniMind/temp")
|
| 20 |
|
| 21 |
from unsloth import FastLanguageModel
|
| 22 |
-
from transformers import TextStreamer
|
| 23 |
|
| 24 |
|
| 25 |
def get_args():
|
|
@@ -72,7 +71,7 @@ def main():
|
|
| 72 |
)
|
| 73 |
|
| 74 |
# 注入lora适配器
|
| 75 |
-
model.load_adapter(args.lora_adapter_path)
|
| 76 |
|
| 77 |
trained_models_dir = temp_directory / "trained_models/Qwen2.5-0.5B-ultrafeedback-dpo-ddp-qlora-unsloth-fp16"
|
| 78 |
trained_models_dir.mkdir(parents=True, exist_ok=True)
|
|
|
|
| 19 |
temp_directory = Path("/root/autodl-tmp/OpenMiniMind/temp")
|
| 20 |
|
| 21 |
from unsloth import FastLanguageModel
|
|
|
|
| 22 |
|
| 23 |
|
| 24 |
def get_args():
|
|
|
|
| 71 |
)
|
| 72 |
|
| 73 |
# 注入lora适配器
|
| 74 |
+
model.load_adapter(args.lora_adapter_path, adapter_name="default")
|
| 75 |
|
| 76 |
trained_models_dir = temp_directory / "trained_models/Qwen2.5-0.5B-ultrafeedback-dpo-ddp-qlora-unsloth-fp16"
|
| 77 |
trained_models_dir.mkdir(parents=True, exist_ok=True)
|