DanielusG
DanielusG
> thanks DanielusG, I tried increasing the layers, the timings didn't change much > > ``` > 24 layers > llama_print_timings: load time = 16568.45 ms > llama_print_timings: sample time...
> Haven't tried this one but I was able to run the original(privateGPT) without problem on my mac M1 with 8G. Question: what would be the configuration in running this...
> I'm not a maintainer but I think it would be super helpful if you separate out all your changes and create separate PRs. That'll make it easier to test/evaluate...
I don't know if it is related, but i have a similar problem on linux, when i start the recording and stop, the file continue to grow after the stop,...
The -coder version is an old version of deepseek (V2). The team discovered that by merging the v2 + v2 coder version they obtained a more powerful model (deepseek V2.5)...
Why downvoting this feature? This can be really helpful for collaboration with AI to create documents and notes
I spent the weekend yesterday investigating this problem! The result is that 99% deepseek API is throttling us. This is because within a very few days they have had a...
I have noticed the same thing myself, it seems to happen when working with files longer than the maximum output in tocken that the model can generate, so when writing...
> I think the correct fix is adding the reasoning_details support to the openai compatible provider. We should track the interleveaned thinking boolean per model tho but that should first...