codelion/optillm

Issues

• Some observations and questions on Google FRAMES Benchmark readurls&memory-gpt-4o-mini method evaluation

  #106 opened a month ago by RGSmirnov
  1

• I want to try to use this project for local llm, but I'm not sure how.

  #93 opened a month ago by matbee-eth
  2

• Warning: Do not support sampling multiple responses

  #99 opened a month ago by Mushoz
  4

• Add a web search plugin that can be used to ground the responses

  #97 opened a month ago by codelion
  0

• How to reproduce RTC Eval 100% locally?

  #96 opened a month ago by botelhorui
  1

• Implement cot decoding with llama.cpp

  #65 opened a month ago by codelion
  5

• parse conversation reports error

  #84 opened 2 months ago by femto
  0

• I get the following error: list index out of range

  #67 opened 2 months ago by ErykCh
  14

• Feature: easy to add new approaches

  #72 opened 2 months ago by ErykCh
  1

• Setting the default approach doesn't work

  #69 opened 2 months ago by ErykCh
  6

• Ambiguous configuration for mcts

  #68 opened 2 months ago by ErykCh
  6

• Resulting docker image size (6.36GB) is quite large - is there any opportunity to reduce this?

  #71 opened 2 months ago by sammcj
  3

• Thanks for adding the entropy based sampling, by any chance do you have a comprasion with other alternative methods?

  #70 opened 2 months ago by shamanez
  2

• (MOA) Fails with "List Index Out of Range" Error on OpenAI-Compatible Ollama API Endpoint

  #60 opened 2 months ago by chrisoutwright
  6

• Add a lighting template for running optillm

  #56 opened 3 months ago by codelion
  1

• Is there any possibility we align some interest?

  #57 opened 2 months ago by femto
  1

• Using llama-server issue with 'no_key' API key

  #61 opened 2 months ago by s-hironobu
  1

• Scripts to reproduce benchmark results

  #63 opened 2 months ago by zhxieml
  1

• Request for Reference Citations for CoT Prompting Methods

  #64 opened 2 months ago by qsunyuan
  1

• Implement routing

  #37 opened 2 months ago by codelion
  1

• I can see cot_decode method has implemented, but we can't use it with the proxy.

  #59 opened 2 months ago by shamanez
  13

• When I tried the optillm with my own openai API compatible hosted model I get this error

  #58 opened 2 months ago by shamanez
  6

• Integration with Gemini 1.5 models

  #54 opened 3 months ago by tranhoangnguyen03
  2

• token counting

  #52 opened 3 months ago by darkacorn
  2

• [Question]: Which paper is mcts.py based on?

  #51 opened 3 months ago by RomanKoshkin
  1

• Can't install z3-solver, is it possible to support lean4?

  #38 opened 3 months ago by femto
  20

• Error processing request: litellm.AuthenticationError: AuthenticationError

  #43 opened 3 months ago by meeen
  2

• Add support for logging with --log=debug

  #44 opened 3 months ago by codelion
  1

• Add support for sympy in solver approach

  #41 opened 3 months ago by codelion
  1

• Possible error in calculate_confidence() logic for cot_decoding.py

  #45 opened 3 months ago by jovanwongzixi
  1

• Add support to pass slug as extra_body argument instead of prefix of model name

  #39 opened 3 months ago by codelion
  0

• Response text missing when using third-party AI frontend with local endpoint

  #36 opened 3 months ago by ibagur
  4

• Streaming, Context, Port & Proxy vs Library

  #3 opened 3 months ago by Spiritdude
  7

• Clarification: proxy or library for cot_decoding??

  #35 opened 3 months ago by lee-b
  2

• Change api-key to optillm-api-key

  #24 opened 3 months ago by codelion
  0

• use with llama.cpp

  #8 opened 3 months ago by scalar27
  8

• Flask import fails

  #23 opened 3 months ago by vanetreg
  5

• Create a gradio based GUI to compare different approaches

  #9 opened 3 months ago by codelion
  0

• Support AzureOpenAI client

  #13 opened 3 months ago by codelion
  1

• Gsm8k bad test

  #16 opened 3 months ago by Tostino
  1

• Minimal working MCTS example

  #5 opened 3 months ago by RomanKoshkin
  3

• Too many tokens

  #4 opened 3 months ago by integral-llc
  2

• initial_query both in system message and user message

  #2 opened 3 months ago by nitinrathi
  2