francescofan
francescofan
@dosu I am using ragas 0.1.20. My LLM is Deepseek v2.0. I alse encoutered a problem like this. Error message is "Invalid n value (currently only n = 1 is...
@dosu can you list all the language parameter values could be set?
@dosu when I ran the code after modification of language, error occured as following: **error** Traceback (most recent call last): File "d:\Notes\2.Python\next_console\next_console_qa_evaluate\ragas\ragas_test.py", line 112, in generator.adapt(language, evolutions=[simple, reasoning, conditional, multi_context])...
@dosu based on the message above, where is the output field?
@dosu I havn't define any example manually. Where could be the example located?
@dosu could you tell the excat place or filename where i should check?
@dosu in 'Lib\site-packages\ragas' anaconda packages directory,no file or directory is named as synthesizers.
@dosu I solved this problem while a new problem occured. In this funcition, generator.adapt(language, evolutions=[simple, reasoning, conditional, multi_context],cache_dir='xxxx') if i don't set the value of cache_dir, the program may not...
@dosu Do I need to set then cache_dir value as a GLOBAL parameter? How can generator.generate_with_langchain_docs find which prompts it should use?
@dosu I mean how can generator.generate_with_langchain_docs always use the specific cache_dir prompts. for example, I have '.\ragas\cache\zh' directory in which a lot of prompts JSON file are created. How can...