this post was submitted on 31 Oct 2023
1 points (100.0% liked)

LocalLLaMA

1 readers
1 users here now

Community to discuss about Llama, the family of large language models created by Meta AI.

founded 10 months ago
MODERATORS
communick@poweruser.forum
1
Phind Model beats GPT-4 at coding, with GPT-3.5 speed and 16k context (news.ycombinator.com)
submitted 10 months ago by sleeper-2@alien.top to c/localllama@poweruser.forum
7 comments fedilink hide all child comments
top 7 comments
sorted by: hot top controversial new old
[–] StrangeImagination5@alien.top 1 points 10 months ago (1 children)

Is the Phind model here on this page https://www.phind.com/ the Phind/Phind-CodeLlama-34B-v2 model?

Or is a different one? On the page it says Phind v7

[–] Combinatorilliance@alien.top 1 points 10 months ago

I'm not very familiar with phind, do you mean that this model that's competitive with GPT-4 is up for downloading right here?

https://huggingface.co/Phind/Phind-CodeLlama-34B-v2

That's insane

[–] You_Wen_AzzHu@alien.top 1 points 10 months ago

I know phind codellama 34b is the best local LLM. But it is not close to GPT4.

[–] Blobbloblaw@alien.top 1 points 10 months ago

There's like five of these a week. These benchmark scam news need to fuck right off

[–] Jean-Porte@alien.top 1 points 10 months ago

At coding benchmark* (singular)

[–] MyFest@alien.top 1 points 10 months ago

we need a test for whether models have been pretrained on a dataset

[–] Snoo98445@alien.top 1 points 10 months ago

study with open example doesn't guarantee better learning result