feat: disable token counting in GPT2Tokenizer by returning 0.

Signed-off-by: -LAN- <laipz8200@outlook.com>

api/core/model_runtime/model_providers/__base/tokenizers/gpt2_tokenzier.py

@@ -25,7 +25,8 @@ class GPT2Tokenizer:
         # future = _executor.submit(GPT2Tokenizer._get_num_tokens_by_gpt2, text)
         # result = future.result()
         # return cast(int, result)
-        return GPT2Tokenizer._get_num_tokens_by_gpt2(text)
+        # return GPT2Tokenizer._get_num_tokens_by_gpt2(text)
+        return 0
 
     @staticmethod
     def get_encoder() -> Any: