Codestral: Mistral's Code Model
mistral.ai
Codestral: Mistral's Code Model
1–10 of 225 posts
Re: Codestral: Mistral's Code Model
#2How does this compare to Github Copilot? It's not shown in their comparison
Re: Codestral: Mistral's Code Model
#3Link to the huggingface page: https://huggingface.co/mistralai/Codestral-22B-v0.1
Re: Codestral: Mistral's Code Model
#4How does this compare to Github Copilot? It's not shown in their comparison
Not sure how much current Copilot varies from the original Codex, but another set of benchmarks here: https://paperswithcode.com/sota/code-generation-on-humaneval
Re: Codestral: Mistral's Code Model
#5How does this compare to Github Copilot? It's not shown in their comparison
Knowing the training data GH has I doubt it's comparable, then again I don't have the benchmarks
Re: Codestral: Mistral's Code Model
#6How does this compare to Github Copilot? It's not shown in their comparison
Knowing the training data GH has I doubt it's comparable, then again I don't have the benchmarks
After typing this I tried the live chat out and it honestly seems a lot more promising than current GH Copilot, very nice!
Re: Codestral: Mistral's Code Model
#7How does this compare to Github Copilot? It's not shown in their comparison
Knowing the training data GH has I doubt it's comparable, then again I don't have the benchmarks
Are you saying GH has more than Codestral and therefore GH has a better model? Or that Codestral would be better because Codestral is not littered with "bad" code?
Re: Codestral: Mistral's Code Model
#8Earlier quoted context omitted.
Knowing the training data GH has I doubt it's comparable, then again I don't have the benchmarks
Are you saying GH has more than Codestral and therefore GH has a better model? Or that Codestral would be better because Codestral is not littered with "bad" code?
Bad code is obviously very subjective, but I would wager that GH places a much higher value on feedback mechanisms like stars, issues, PRs, velocity, etc. Their ubiquity likely allows them to automatically cherry-pick less "bad code."
Re: Codestral: Mistral's Code Model
#9Does anyone know of a link to a codegen comparison page? In other words, you write your request, and it's submitted to multiple codegen engines, so you can compare the output.
Re: Codestral: Mistral's Code Model
#10This is an open weights 22B model. The download on Huggingface is 44GB.
Is there a rule-of-thumb estimate for how much RAM this would need to be used locally?
Is the RAM requirement the same for a GPU and "unified" RAM like Apple silicon?