Fix openai eval bug (#78)

* fix bug

* make blue happy

---------

Co-authored-by: wangyuxin <[email protected]>

mteb-zh/mteb_zh/models.py

@@ -107,13 +107,16 @@ def __init__(
  self,
  api_key: Optional[str] = None,
  model_name: str = 'text-embedding-ada-002',
+ max_length: int = 4000,
  ) -> None:
  if api_key is not None:
  openai.api_key = api_key
  self._client = openai.Embedding.create
  self.model_name = model_name
+ self.max_length = max_length
 
  def encode(self, sentences: list[str], batch_size: int = 32, **kwargs) -> list[np.ndarray]:
+ sentences = [sentence[: self.max_length] for sentence in sentences]
  all_embeddings = []
  for batch in tqdm(
  generate_batch(sentences, batch_size),
@@ -128,15 +131,17 @@ def encode(self, sentences: list[str], batch_size: int = 32, **kwargs) -> list[n
 
 
 class AzureModel:
- def __init__(self, model_name: str = 'text-embedding-ada-002') -> None:
+ def __init__(self, model_name: str = 'text-embedding-ada-002', max_length: int = 4000) -> None:
  openai.api_type = 'azure'
  openai.api_key = os.environ['AZURE_API_KEY']
  openai.api_base = os.environ['AZURE_API_BASE']
  openai.api_version = '2023-03-15-preview'
  self._client = openai.Embedding.create
  self.model_name = model_name
+ self.max_length = max_length
 
  def encode(self, sentences: list[str], batch_size: int = 32, **kwargs) -> list[np.ndarray]:
+ sentences = [sentence[: self.max_length] for sentence in sentences]
  all_embeddings = []
  for text in tqdm(sentences):
  output = self._client(input=text, engine=self.model_name)