Ben Heckmann

Hi there! I am using LlamaIndex with german data and german prompts. As far as I can tell the default prompts are only in English and there is no option...

Extend token count to Embeddings and (non-streaming) BaseChatModel (#945)

5

This PR addresses issue #945 that embeddings do not count toward the callback's token count. I have added a callback handler to the base `Embeddings` class and implemented the calling...

feat: PreProcessor split by token (tiktoken & Hugging Face)

4

### Related Issues Implements #4983. This allows a user to split documents by token count using the `PreProcessor` class. ### Proposed Changes: #### Interface: The user can choose any of...

DRAFT add Data Health Checker

2

## Description First draft for a data health checker as discussed in #854. The checker receives a path to the data in CSV or qlib format (not implemented yet). It...

**Is your feature request related to a problem? Please describe.** I am building a chatbot for my project and require something similar to [this](https://www.npmjs.com/package/react-chat-widget). **Describe the solution you'd like** A...

It would be great to be able to set a shortcut for a kind of lock that generates hints whenever a site refreshes / changes. Especially while using the browser,...

bug: zrange rev flag causes empty return list / flaky zrange

2

Hi, When using `kv.zrange`, the same input sometimes returns the list of values, and sometimes an empty list. When using the cli on Vercel, I always get the list. ```js...