github
/
ChatGLM-6B
mirror of https://github.com/THUDM/ChatGLM-6B
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

Merge 5ec5f04515 into 401bf3a8a7

pull/1034/merge
Mingdao Liu 2024-09-30 05:07:38 -04:00 committed by GitHub
parent 401bf3a8a7 5ec5f04515
commit 0e8edb133f
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 5 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
ptuning/README.md
View File

@ -8,7 +8,7 @@
## 软件依赖 ## 软件依赖
运行微调需要4.27.1版本的`transformers`。除 ChatGLM-6B 的依赖之外，还需要安装以下依赖 运行微调需要4.27.1版本的`transformers`。除 ChatGLM-6B 的依赖之外，还需要安装以下依赖
``` ```
pip install rouge_chinese nltk jieba datasets pip install rouge_chinese nltk jieba datasets filelock
``` ```
## 使用方法 ## 使用方法

5
ptuning/main.py
View File

@ -44,6 +44,8 @@ from trainer_seq2seq import Seq2SeqTrainer
from arguments import ModelArguments, DataTrainingArguments from arguments import ModelArguments, DataTrainingArguments
from filelock import FileLock
logger = logging.getLogger(__name__) logger = logging.getLogger(__name__)
def main(): def main():
@ -121,7 +123,8 @@ def main():
new_prefix_state_dict[k[len("transformer.prefix_encoder."):]] = v new_prefix_state_dict[k[len("transformer.prefix_encoder."):]] = v
model.transformer.prefix_encoder.load_state_dict(new_prefix_state_dict) model.transformer.prefix_encoder.load_state_dict(new_prefix_state_dict)
else: else:
model = AutoModel.from_pretrained(model_args.model_name_or_path, config=config, trust_remote_code=True) with FileLock("model.lock"):
model = AutoModel.from_pretrained(model_args.model_name_or_path, config=config, trust_remote_code=True)
if model_args.quantization_bit is not None: if model_args.quantization_bit is not None:
print(f"Quantized to {model_args.quantization_bit} bit") print(f"Quantized to {model_args.quantization_bit} bit")