用ollama翻译，能行嘛

[602387193c]

这是看了这视频得来的技巧，安装了ollama，希望作者把ollama的支持加下？ Ollama+Gemma2:9b本地开源大模型输出OpenAI兼容API作为本地翻译引擎工作流

[602387193c]

抱歉，好像没有用，运行到中间会出现这个错误：

File "C:\Users\c6023\anaconda3\envs\videolingo\lib\site-packages\streamlit\runtime\scriptrunner\exec_code.py", line 88, in exec_func_with_error_handling result = func() File "C:\Users\c6023\anaconda3\envs\videolingo\lib\site-packages\streamlit\runtime\scriptrunner\script_runner.py", line 590, in code_to_exec exec(code, module.dict) File "D:\AI\VideoLingo\st.py", line 126, in main() File "D:\AI\VideoLingo\st.py", line 122, in main text_processing_section() File "D:\AI\VideoLingo\st.py", line 35, in text_processing_section process_text() File "D:\AI\VideoLingo\st.py", line 60, in process_text step4_2_translate_all.translate_all() File "D:\AI\VideoLingo\core\step4_2_translate_all.py", line 83, in translate_all results.append(future.result()) File "C:\Users\c6023\anaconda3\envs\videolingo\lib\concurrent\futures_base.py", line 438, in result return self.get_result() File "C:\Users\c6023\anaconda3\envs\videolingo\lib\concurrent\futures_base.py", line 390, in get_result raise self._exception File "C:\Users\c6023\anaconda3\envs\videolingo\lib\concurrent\futures\thread.py", line 52, in run result = self.fn(*self.args, **self.kwargs) File "D:\AI\VideoLingo\core\step4_2_translate_all.py", line 44, in translate_chunk things_to_note_prompt = search_things_to_note_in_prompt(chunk) File "D:\AI\VideoLingo\core\step4_1_summarize.py", line 20, in search_things_to_note_in_prompt prompt = '\n'.join( File "D:\AI\VideoLingo\core\step4_1_summarize.py", line 22, in f' meaning: {term["explanation"]}'

[TITC]

ollama是有OpenAI compatibility支持的。但是据我这两天的尝试，A100单卡qwen2.5:7b 20k上下文都无法成功跑完一个5分多的YouTube视频 ，因为上下文程度不够，但是显存已经接近80GB了

[602387193c]

原来如此，看来还是得API才行，主要上下文限制了，一般电脑跑不动

[Huanshere]

"能用别的模型吗？ ✅ 支持 OAI-Like 的 API 接口，需要自行在 streamlit 侧边栏更换。 ⚠️ 但其他模型（尤其是小模型）遵循指令要求能力弱，非常容易在翻译过程报错，强烈不推荐。"

文档中有提到~