Chenhe Gu comments

Results 25 comments of


                                            Chenhe Gu

can not save in the bedrock setting page.

can you try starting afresh? can't seem to reproduce this

add support for AWS bedrock Cohere embeddings.

Hi there, thanks so much for the PR - I've tested the embedding and got the following error message: ``` [2024-04-12 17:03:51,589: ERROR/MainProcess] consume document failed Traceback (most recent call...

Gemini pro return error

Hi! This may have been an [issue](https://github.com/google/generative-ai-python/issues/196) on gemini's side. I suggest also follow up on that issue.

Increase the limit of tokens for openai compatible llms

You're free to modify both context length and max output token in the model config to suit your model. I tried mixtral on together.ai's endpoint and had it return 6000+...

Yi34B's 8k context capability on dify doesn't work

Hi @PhenixZhang, seems that I'm not able to reproduce this issue on my end. Can you upload the backend logs? If you're ok with it and your endpoint is accessible...

token limitation after some questions

Error seems self explanatory, what context size did you input when configuring the model?

why gpt-4 could change the vector search keywords in the second conversation , by other LLM model can not do this

Hi there, could you provide more info, like embedding model and prompt used?

multiple conversation, how to seach vector with all the user input

Retrieval is done on a per-query basis by design, so it's expected that your second search won't return documents relevant to your first search. However, in an actual conversation the...

Update bedrock.yaml add Region Asia Pacific (Sydney)

if you are able to, can you check to make sure a key from ap area works? we don't have access to one unfortunately.

Tool addition: Cloud Vision API

We're working on enabling file upload, a tool would be highly welcomed once that's released