Skip to content

Commit

Permalink
new folder
Browse files Browse the repository at this point in the history
  • Loading branch information
JinheTang committed Aug 16, 2024
1 parent 4cf56bd commit dc2bd2f
Show file tree
Hide file tree
Showing 2 changed files with 2 additions and 2 deletions.
4 changes: 2 additions & 2 deletions python/llm/example/GPU/HuggingFace/LLM/codegeex2/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -29,15 +29,15 @@ pip install --pre --upgrade ipex-llm[xpu] --extra-index-url https://pytorch-exte
```

### 2. Download Model and Replace File
If you select the codegeex2-6b model ([THUDM/codegeex-6b](https://huggingface.co/THUDM/codegeex2-6b)), please note that their code (`tokenization_chatglm.py`) initialized tokenizer after the call of `__init__` of its parent class, which may result in error during loading tokenizer. To address issue, we have provided an updated file ([tokenization_chatglm.py](./tokenization_chatglm.py))
If you select the codegeex2-6b model ([THUDM/codegeex-6b](https://huggingface.co/THUDM/codegeex2-6b)), please note that their code (`tokenization_chatglm.py`) initialized tokenizer after the call of `__init__` of its parent class, which may result in error during loading tokenizer. To address issue, we have provided an updated file ([tokenization_chatglm.py](./codegeex2-6b/tokenization_chatglm.py))

```python
def __init__(self, vocab_file, padding_side="left", clean_up_tokenization_spaces=False, **kwargs):
self.tokenizer = SPTokenizer(vocab_file)
super().__init__(padding_side=padding_side, clean_up_tokenization_spaces=clean_up_tokenization_spaces, **kwargs)
```

You could download the model from [THUDM/codegeex-6b](https://huggingface.co/THUDM/codegeex2-6b), and replace the file `tokenization_chatglm.py` with [tokenization_chatglm.py](./tokenization_chatglm.py).
You could download the model from [THUDM/codegeex-6b](https://huggingface.co/THUDM/codegeex2-6b), and replace the file `tokenization_chatglm.py` with [tokenization_chatglm.py](./codegeex2-6b/tokenization_chatglm.py).

### 3. Configures OneAPI environment variables for Linux

Expand Down

0 comments on commit dc2bd2f

Please sign in to comment.