monatis

6 ms·

1.
Agentic ML engineer. works with Colab. Zero infra needed. 3x faster TurboQuant(github.com)
6 points by monatis 5mo ago|1 comments
2.
by monatis 5mo ago
isanagent is opensource! I implemented an always-on and autonomous ML engineer that works 24/7 to conduct research, generate datasets, train models, learn and evolve. Works on Colab, local GPU or over SSH. Zero infra needed! Whether yo
3.
by monatis 5mo ago
Yeah this is one possible way to generate grounded an"responses" in Afterimage. To accomplish context augmentation when generating a response, it allows to use different RAG strategies where retriever may be chosen for the specifi
4.
AfterImage – Generate synthetic multi-turn chat data from documents(github.com)
5 points by monatis 5mo ago|4 comments
5.
by monatis 5mo ago
We kept running into the same exact bottleneck with fine-tuning and evals: You have the source documents, and you have the base model, but you usually don’t have the actual conversations. If you’re working with internal docs, regulatory tex
6.
Show HN: Afterimage is now open-source for infra-grade dataset generation(github.com)
2 points by monatis 5mo ago|0 comments
7.
Show HN: 2.4x faster baai/bge-M3(huggingface.co)
2 points by monatis 1y ago|0 comments
8.
Show HN: One-liner CLI for batched PDF-to-Markdown at $1 per ~6k pages(github.com)
8 points by monatis 1y ago|0 comments
9.
I implemented CLIP inference in plain C/C++(github.com)
5 points by monatis 3y ago|1 comments
10.
by monatis 3y ago
I implemented CLIP inference in plain C/C++ with no extra dependencies thanks to the great work in GGML that powers llama.cpp. It can work with models from both OpenAI and LAION. It also supports 4-bit quantization for extremely constr
11.
by monatis 3y ago
Similarity search becomes a part of many technical stacks. Qdrant can be deployed as a service for fast, lightweight and filterable vector search. It's opensource, and you can also get up and running instantly and for free wit the clou
12.
by monatis 4y ago
Great work, congrats. You might be already familiar with Qdrant, but feel free to reach out to me if you'd like to get help testing or discuss anything around it.
13.
by monatis 4y ago
Some loss functions such as ArcFace loss and CosFace loss enforce the encoder model to organize their latent space in such a way that categories are placed with an angular margin from one another. Thus the model implicitly learns a continuo
14.
by monatis 4y ago
The main idea is to train a deep learning model to encode a high-dimensional sample to a low-dimensional vector in a latent space. Then it can be used in various downstream tasks such as KNN applications, semantic search, multimodal retriev
15.
by monatis 4y ago
There's study here: http://cs230.stanford.edu/projects_spring_2021/reports/57.pd... Basically, you can collect text from different authors, then you can use authors names as labels to train a similarity learn
16.
by monatis 4y ago
I'm one of the authors of Quaterion and also a contributor of TF Similarity. First of all, Quaterion is based on PyTorch. From a more technical perspective, TF Similarity is currently stronger on the self-supervised methods for images