vikp

7 ms·

31.
Show HN: Texify – OCR math images to LaTeX and Markdown(github.com)
21 points by vikp 3y ago|3 comments
32.
by vikp 3y ago
Yes, this is on my list of things to do :)
33.
by vikp 3y ago
Yes, nougat is used as part of the pipeline to convert the equations (basically marker detects the equations then passes those regions to nougat). It's a great model for this.
34.
by vikp 3y ago
(author) Please feel free to open an issue if you try again. Poetry can be painful, I might just switch to a requirements.txt file in the future. (you can skip poetry if you want by just pulling everything in pyproject.toml into a require
35.
by vikp 3y ago
I chose markdown because I wanted to preserve equations (fenced by $/$$), tables, bold/italic information, and headers. I haven't looked into epub output, but this ruled out plain text.
36.
by vikp 3y ago
Author here: for my use case (converting scientific PDFs in bulk), nougat was the best solution, so I compared to it as the default. I also compare to naive text extraction further down. Nougat is a great model, and converts a lot of PDFs
37.
by vikp 3y ago
Author here - this is one of the reasons I made this. Also see https://github.com/VikParuchuri/libgen_to_txt , although I haven't integrated marker with it yet (it uses naive text extraction).
38.
by vikp 3y ago
The deep learning book is a great choice, as many have mentioned. I've been making a course that has a little less theory, and a little more application here - https://github.com/VikParuchuri/zero_to_gpt . Videos
39.
by vikp 3y ago
I'm building a course that teaches deep learning from the ground up - https://github.com/VikParuchuri/zero_to_gpt . It balances theory and code, and builds from the foundation up, so you're never typing somet
40.
by vikp 3y ago
This is not a random website - together is a prominent AI research company. Their chief scientist invented flashattention, which is used to train most LLMs. They release open source datasets/models like RedPajama. And they've p
41.
by vikp 3y ago
This post is misleading, in a way that is hard to do accidentally. - They compare the performance of this model to the worst 7B code llama model. The base code llama 7B python model scores 38.4% on humaneval, versus the non-python mode
42.
by vikp 3y ago
There are plenty of great embedding models that are on the order of a few hundreds megs (even outperforming ada-002). See the leaderboard here - https://huggingface.co/spaces/mteb/leaderboard . Local/offline
43.
by vikp 3y ago
Got it, thanks - and thanks for the model! I'd be interested in the results if anyone benchmarks without sampling. Edit: it could also be misleading to directly compare humaneval pass@1 against codellama without the same generation me
44.
by vikp 3y ago
Did you use the same pass@1 generation method as in the code llama paper (greedy decoding)? I couldn't find this in the blog post.
45.
by vikp 3y ago
I'm writing a deep learning course called Zero to GPT - https://github.com/VikParuchuri/zero_to_gpt . It teaches you everything you need to train an LLM, including the basics of deep learning and linear algebra.
46.
by vikp 3y ago
I clicked because I thought they were defining LLM developer as "someone training LLMs", but instead they define it as "someone integrating LLMs into their application". If you also had the same initial thought as me, th
47.
by vikp 3y ago
That's right, although I used "tasks", which I think are higher level than individual search terms. I think there's also potential for personalization - learning the patterns that work specifically for you.
48.
by vikp 3y ago
I'm torn on this. Sharing your own personal website and thoughts is a core part of the web, I agree. But, the popularity of LLMs indicates that people want to complete tasks efficiently. I think these two goals are usually in conflic
49.
A vision for the AI web: the real web 3.0(vikas.sh)
15 points by vikp 3y ago|10 comments
50.
by vikp 3y ago
I would use textsynth ( https://bellard.org/ts_server/ ) or llama.cpp ( https://github.com/ggerganov/llama.cpp ) if you're running on CPU. - I wouldn't use anything higher than a 7B mode
51.
by vikp 3y ago
I was in a similar boat, and I built a project called Endless Academy - https://www.endless.academy/ . It helped me both brush up on some new tools, and scratch an itch to build something. To start, just using an LLM API (
52.
by vikp 3y ago
I've noticed that semantic search tutorials only use cloud vector databases and embedding APIs, so I figured I'd take a stab at writing a simpler guide. It's possible to get good accuracy and speed entirely locally.
53.
Build a semantic search engine in Python(vikas.sh)
1 points by vikp 3y ago|1 comments
54.
by vikp 3y ago
I actually agree, which is why I wrote the comment. It is easy to create background tasks. Dramatiq can even handle some of the cases you mentioned - multi-step, fan-out, and retries. It is hard to scale background tasks when you hit a hi
55.
by vikp 3y ago
I think this is an interesting service, and could be a nice way to write a SvelteKit or Next.js app with background workers. The press release is unnecessarily hyperbolic, which turned me off, though: > Deploying new jobs to production
56.
by vikp 3y ago
Yes, I use retrieval for Endless Academy [1] , and it works well. Some tips: - Most vector search is basically kNN under the hood, with some kind of compression. If you have too many embeddings in your DB, this starts to pull up irrele
57.
by vikp 3y ago
I was surprised, too, but then I realized they all work at Qdrant. But the general dialogue around AI-related tools is surprising to me. The production parts of the langchain, embeddings, etc tools can usually be built in a few hours with b
58.
by vikp 3y ago
I'd rather run ~10 lines of code locally than setup 3 cloud services and a lambda function, but to each their own...
59.
by vikp 3y ago
Just run it on CPU, on your own machine. That's the cheapest way. You could also rent a free/cheap VPS, and even parallelize across multiple machines/cores if you need it.
60.
by vikp 3y ago
This tutorial is very complex. Here's how to get free semantic search with much less complexity: 1. Install sentence-transformers [1] 2. Initialize the MiniLM model - `model = SentenceTransformer('all-MiniLM-L6-v2')`

More ›