mohsen1

26 ms·

361.
by mohsen1 2y ago
https://github.com/bodo-run/yek This is more sophisticated for serializing your repo. Please check it out and let me know what do you think?
362.
by mohsen1 2y ago
Yup. As an example extensions can not read the content of the terminal. The API is there but not allowed to be used in published extensions
363.
by mohsen1 2y ago
The guys at Unsloth did a great job making this workflow accessible: https://news.ycombinator.com/item?id=42969736
364.
Train your own R1 reasoning model with Unsloth(unsloth.ai)
6 points by mohsen1 2y ago|0 comments
365.
by mohsen1 2y ago
I don't know what those "needle in haystack" benchmarks are testing for because in my experience dumping a big amount of code in the context is not working as you'd expect. It works better if you keep the context small
366.
by mohsen1 2y ago
> available via the Gemini API in Google AI Studio and Vertex AI. > Gemini 2.0, 2.0 Pro and 2.0 Pro Experimental, Gemini 2.0 Flash, Gemini 2.0 Flash Lite 3 different ways of accessing the API, more than 5 different but extremely simi
367.
by mohsen1 2y ago
That explains it!
368.
by mohsen1 2y ago
I did a JS1K back in the day [1] such a fun concept! I wonder these days if anyone is using multi-character emoji (don't know what do you call those exactly) to compress more data in 1024 "bytes"? Since HackerNews does not al
369.
by mohsen1 2y ago
Pretty ironic that they use an automated system called CodeSignal that does the first round of the interviews
370.
by mohsen1 2y ago
With modified prompt to make less look like the original prompt it thought 6x more: https://chatgpt.com/share/679f086d-a758-8007-b240-38e6843037...
371.
by mohsen1 2y ago
https://chatgpt.com/share/679f0353-12bc-8007-91cf-dd63d52044... O1 Pro gets the answer with proper reasoning Can’t tell if this is dude to data contamination or it really figured it out? How can we form the question in
372.
by mohsen1 2y ago
tsc caches the interface of projects so when a project is not changed, its not compiled. see https://www.typescriptlang.org/docs/handbook/project-referen...
373.
Ask HN: Is anyone working on an LSP-based AI debugger?
1 points by mohsen1 2y ago|0 comments
374.
by mohsen1 2y ago
Where I worked full TypeScript compilation in the monorepo would take an hour or more. But that would be kinda rare. Using project references and heavy CI caching and remote caching we were able to type-check parts of the repo with confiden
375.
by mohsen1 2y ago
AI probably can do very nice landscaping design too. It's just that image generation models are not made for it. If we manage to give AI a more structured input, you'll get a nice design in return.
376.
by mohsen1 2y ago
if you are running locally, you can add <think> tokens to your own prompt to distract it. it works in many cases that I tried.
377.
by mohsen1 2y ago
It's funny because I asked it to fix my script to show deepseek 's chain of thoughts in the script but it refuses to answer hahaha
378.
by mohsen1 2y ago
I see! Now I understand why I need to create those useless `Modelfile` files... I'm glad there is a more open source alternative to Ollama now.
379.
by mohsen1 2y ago
It's self dealing in the context of taxes.
380.
by mohsen1 2y ago
See function calling being called out here https://huggingface.co/mistralai/Mistral-Small-24B-Instruct-...
381.
by mohsen1 2y ago
Not so subtle in function calling example[1] "role": "assistant", "content": "---\n\nOpenAI is a FOR-profit company.", [1] https://huggingface.co/mistralai/M
382.
by mohsen1 2y ago
every time you respond to an AI model "no, you got that wrong, do it this way" you provide a very valuable piece of data to train on. With reasoning tokens there is just a lot more of that data to train on now
383.
by mohsen1 2y ago
The idea that a lot of compute is moving towards inference has a huge consequence for the current "AI investments". This is bad news for NVDA particularly. The inference focused solutions have better economics than paying NVDA tho
384.
by mohsen1 2y ago
with 57 million(!!) tokens
385.
by mohsen1 2y ago
This is not the reasoning model. If they beat Deepseek V3 in benchmarks I think a 'reasoning' model would beat o1 Pro
386.
by mohsen1 2y ago
My guess is that OS vendors are the real winners in the long run. If Siri/Goolge can access my stuff and core of LLMs is this replicable then I don't see anyone downloading any apps for their typical AI usage. Specially that users
387.
by mohsen1 2y ago
if you publish only the binary it's not open source if open the source then it is open source if you write a book/blog about how you came up with the ideas but didn't publish the source it's not open source, even if you
388.
by mohsen1 2y ago
Yes! I've done something like this here in my repo. This was nice while lasted (Deepseek is practically useless through the API since yesterday) https://github.com/bodo-run/yek/blob/main/.github&#x2F
389.
by mohsen1 2y ago
I am subscribed to o1 Pro and am working on a little Rust crate. I asked both o1 Pro and Deepseek R1 to write e2e tests given all of the code in the repo (using yek[1]). o1 Pro code: https://github.com/bodo-run/clap-con
390.
by mohsen1 2y ago
Cursor can follow links

More ›