tarruda

12 ms·

151.
by tarruda 1y ago
> The escape hatch is to use the FDroid version rather than the Play Store version. As long as Google doesn't remove the ability to sideload apps, Android users are fine.
152.
by tarruda 1y ago
I wonder if it is possible to have some kind of P2P protocol similar to BitTorrent where one can seed incremental snapshots of subsets of the internet. Something like the internet archive, but fully decentralized.
153.
by tarruda 1y ago
> I’m always amused when people don’t want to send fractions of their code to a LLM but happily host it on GitHub What amuses me even more is people thinking their code is too unique and precious, and that GitHub/Microsoft wants to
154.
by tarruda 2y ago
> I'm a little unimpressed by its instruction following Been trying the 109b version on Groq and it seems less capable than Gemma 3 27b
155.
by tarruda 2y ago
I read somewhere that ryzen AI 370 chip can run gemma 3 14b at 7 tokens/second, so I would expect the performance to be somewhere in that range for llama 4 scout with 17b active
156.
by tarruda 2y ago
Llama.com has the blog post
157.
by tarruda 2y ago
AFAIK CPython doesn't JIT compile.
158.
by tarruda 2y ago
I tested the full fp16 gguf
159.
by tarruda 2y ago
Thanks for this, but I'm still unable to reproduce the results from Google AI studio. I tried your version and when I ask it to create a tetris game in python, the resulting file has syntax errors. I see strange things like a space in
160.
by tarruda 2y ago
I ran the same prompt on google AI studio it had the same behavior of talking about improvements as if the code it wrote was not the first version. Other than that, the experience was completely different: - The game worked on first try - I
161.
by tarruda 2y ago
Can you share the all the recommended settings to run this LLM? It is clear that the performance is very good when running on AI studio. If possible, I'd like to use the all the same settings (temp, top-k, top-p, etc) on Ollama. AI stu
162.
by tarruda 2y ago
I tried ollama fp16 and it had the same issues.
163.
by tarruda 2y ago
Same experience here: On AI Studio, this is easily one of the strongest models I have used, including when compared to proprietary LLMs. But ollama and openwebui performance is very bad, even when running the FP16 version. I also tried to m
164.
by tarruda 2y ago
I checked this, the whole conversation was about 1000 tokens. I suspect the Ollama version might have wrong default settings, such as conversation delimiters. The experience of Gemma 3 in AI studio is completely different.
165.
by tarruda 2y ago
The new M3 Ultra Mac Studio (512GB version) seems to be capable of running DeepSeek R1 in Q4
166.
by tarruda 2y ago
My usual non-scientific benchmark is asking it to implement the game Tetris in python, and then iterating with the LLM to fix/tweak it. My prompt to Gemma 27b (q4) on open webui + ollama: "Can you create the game tetris in python?
167.
by tarruda 2y ago
In my experience, Gemma models were always bad at coding (but good at other tasks).
168.
by tarruda 2y ago
Is "OpenAI" the only AI company that hasn't released any model weights?
169.
by tarruda 2y ago
> How well does it cite the source? I don't know about the OP tool, but open webui has its own document database which you can integrate with LLMs, and when answering questions it always cites the source with a link for you to veri
170.
by tarruda 2y ago
> We ain't even solved garbage collection yet Can you elaborate on that?
171.
by tarruda 2y ago
> What is the purpose of @dataclass on Task class? No purpose. I think I added in the initial implementation and ended up not being required, but I forgot to remove
172.
by tarruda 2y ago
If you count only the library code, it is less than 250 LOC. I then asked for Claude to write the docstrings and examples which increased by 500
173.
Show HN: Python micro event loop library (~250 LOC)(gist.github.com)
59 points by tarruda 2y ago|12 comments
174.
by tarruda 2y ago
Would love another MoE that fits in 120GB VRAM for the 128gb Mac owners
175.
by tarruda 2y ago
> And it's quite out of character for Linus not to have a blazingly clear opinion. (We all know his stance on C++, for instance.) People change. As you get older, you might find you no longer care that much about subjects you previo
176.
by tarruda 2y ago
> What is a bit weird about AI currently is that you basically always want to run the best model, I think the problem is thinking that you always need to use the best LLM. Consider this: - When you don't need correct output (such as
177.
by tarruda 2y ago
Interesting. A common libuv binding for lua is "luv": https://github.com/luvit/luv
178.
by tarruda 2y ago
It is still unconfirmed since no one outside of deepseek reproduced it. If confirmed, Nvidia could go down even more
179.
by tarruda 2y ago
I remember that Llama 3 was trained on data curated by Llama 2 and it resulted in a model with a significant performance boost (even though it was trained by a previous generation model of the same size). Maybe using a strong reasoning mode
180.
by tarruda 2y ago
Would be great if the next generation of base models was designed to be inferred with 128GB of VRAM while 8bit quantized (which would fit in the consumer hardware class). For example, I imagine a strong MoE base with 16 billion active param

More ›