支持gpt-4-v处理多张图片
This commit is contained in:
parent
9a21e13d33
commit
682898a3ba
@ -198,7 +198,7 @@ def predict(inputs, llm_kwargs, plugin_kwargs, chatbot, history=[], system_promp
|
||||
# 前者是API2D的结束条件,后者是OPENAI的结束条件
|
||||
if ('data: [DONE]' in chunk_decoded) or (len(chunkjson['choices'][0]["delta"]) == 0):
|
||||
# 判定为数据流的结束,gpt_replying_buffer也写完了
|
||||
lastmsg = chatbot[-1][-1] + f"<br/><br/>{llm_kwargs['llm_model']}调用结束,该模型不具备上下文对话能力,如需追问,请及时切换模型。"
|
||||
lastmsg = chatbot[-1][-1] + f"\n\n\n\n「{llm_kwargs['llm_model']}调用结束,该模型不具备上下文对话能力,如需追问,请及时切换模型。」"
|
||||
yield from update_ui_lastest_msg(lastmsg, chatbot, history, delay=1)
|
||||
logging.info(f'[response] {gpt_replying_buffer}')
|
||||
break
|
||||
|
Loading…
x
Reference in New Issue
Block a user