Open NearCai opened 6 months ago
We do not use multi-threading when using other LLMs. If you would like to run open-sourced LLMs, you should refer to run_open_LLM.py
. Thank you for the feedback.
Thanks, but if I want to access the API interface of another model, what should I do? This problem occurs at this time
The HTTP request methods are provided differently across diverse service providers. You may need to re-write the request body by checking their official documents. However, I think multi-threading is still helpful in accelerating the requests.
When calling an external large model, such as ChatGPT, it seems that the oneline file starts multithreading here, even if it is changed to another large model, it is useless.Especially the sentence between: for thread in tqdm(threads): thread.join()