Crumb
c/localllama · by noneabove1182@sh.itjust.works · 2y

Orca 2: Teaching Small Language Models How to Reason

https://www.microsoft.com/en-us/research/blog/orca-2-teaching-small-language-models-how-to-reason/

Orca 2 released by Microsoft!

Full weights here:

https://huggingface.co/microsoft/Orca-2-7b

https://huggingface.co/microsoft/Orca-2-13b

My own exllamav2 quants here:

https://huggingface.co/bartowski/Orca-2-7b-exl2

https://huggingface.co/bartowski/Orca-2-13b-exl2

GGUF from TheBloke (links to GPTQ/AWQ inside it):

https://huggingface.co/TheBloke/Orca-2-7B-GGUF

https://huggingface.co/TheBloke/Orca-2-13B-GGUF

20 points · 6 comments · view on lemmy.world

6 Comments

Hanabie@sh.itjust.works · 4 pts · 2y (4 replies)

What's the context window size?

noneabove1182@sh.itjust.works · 2 pts · 2y (3 replies)

according to the config it looks like it's only 4096, and they specify in the arxiv that they kept the training data under that value so it must be 4096.. i'm sure people will expand it soon like they have with others

Hanabie@sh.itjust.works · 2 pts · 2y (2 replies)

I'll wait a little then I guess. I need 16k for what I'm doing. Thanks for the heads up 🙂

noneabove1182@sh.itjust.works · 2 pts · 2y (1 reply)

Btw there's a 16k tune available now:

https://huggingface.co/bartowski/Orca-2-13B-16k-exl2

Hanabie@sh.itjust.works · 2 pts · 2y

Nice, will give it a try. Thank you 😊

pennomi@lemmy.world · 3 pts · 2y (1 reply)
[ removed ]
noneabove1182@sh.itjust.works · 2 pts · 2y

I'm looking forward to trying it today, I think this might make a good RAG model based on the orca 2 paper, but testing will be needed