4 ms·
From the article: I've been seeing some very promising results from DeepSeek R1 for code as well. Here's a recent transcript where I used it to rewrite the l
by rahimnathwani 2y ago
From the article:
I've been seeing some very promising results from DeepSeek R1 for code as well. Here's a recent transcript where I used it to rewrite the llm_groq.py plugin to imitate the cached model JSON pattern used by llm_mistral.py, resulting in this PR.
But the transcript mentioned was not with Deepseek R1 (not the original, and not even the 1.58 quantized version), but with a Llama model finetuned on R1 output: deepseek-r1-distill-llama-70b
So perhaps it's doubly impressive?
- simonw 2y agoYeah, I was using the lightning fast Groq-hosted 70B distilled version.
- rahimnathwani 2y agoDid you happen to try the same thing on Deepseek R1 on https://chat.deepseek.com/ https://chat.deepseek.com/ ?
- simonw 2y agoNo. I tried it just now with the same prompt and got a similar looking response (with some different design decisions but I'd expect that for even the exact same model). https://gist.github.com/simonw/115620647028336e3a1edfe8a48e139c https://gist.github.com/simonw/115620647028336e3a1edfe8a48e1...