#Mistral 7b instruct v0.3
19 messages · Page 1 of 1 (latest)
I hope it's the version Mistral had as "Mistral Tiny", it was quite a bit better than v0.2
Should we route to this separately from the other versions?
I guess safest is to route separately for now then aliast later :-?...
it has a totally different tokenizer it seems, sooo.... i guess?
I've been testing mistral q8_0 on my local machine and I really like it its really good for story writing/role playing
also for general q&a it seems really knowledgable and is uncensored
Colab Code - https://drp.li/K98Z7
🕵️ Interested in building LLM Agents? Fill out the form below
Building LLM Agents Form: https://drp.li/dIMes
👨💻Github:
https://github.com/samwit/langchain-tutorials (updated)
https://git hub.com/samwit/llm-tutorials
⏱️Time Stamps:
00:00 Intro
00:54 Mistral 7B-V0.3 Benchmarks
01:26 What's beenadded to Mistral...
According to https://docs.mistral.ai/getting-started/models/ the latest mistral-tiny is mistral-tiny-2312, but it doesn't say which version of Mistral-7B that refers to
Almost there:
btw there's apparently also an updated version of Mixtral-8x22B-Instruct out there
Idk if anyone hosted this yet
https://huggingface.co/mistral-community/mixtral-8x22B-Instruct-v0.3-original
A high-performing, industry-standard 7.3B parameter model, with optimizations for speed and context length.
An improved version of Mistral 7B Instruct v0.2, with the following changes:
- Extended vocabulary to 32768
- Supports v3 Tokenizer
- Supports function calling
NOTE: Support for function cal...
Also https://openrouter.ai/models/mistralai/mistral-7b-instruct now points to 0.3
. Try out API on the Web