Alternatives
Products that do what Model2Vec: make sentence transformers 500x faster on CPU, 15x smaller does
Hi HN! We (Thomas and Stéphan, hello!) recently released Model2Vec, a Python library for distilling any sentence transformer into a small set of static embeddings. This makes inference with such a model up to 500x faster, and reduces model size by a factor of 15 (7.5M params or 15/30MB on disk, depending on whether you use float16 or float32). This reduction of course comes at a cost: distilled models are a lot worse than their parent models. Even so, they are actually a lot better than large sets of conventional static embeddings, such as GLoVe or word2vec-based models, which are many…
- 1MM
Hi HN! We (Thomas and Stéphan, hello!) recently released Model2Vec, a Python library for distilling any sentence transformer into a small set of static embeddings. This makes inference with such a model up to 500x faster, and reduces model size by a factor of 15 (7.5M params or 15/30MB on disk, depending on whether you use float16 or float32). This allows you to embed 50-100k documents per second on a cpu on a macbook. This reduction of course comes at a cost: distilled models are worse than their parent models. Even so, they are actually a lot better than large sets of conventional…
2024 · github.com
- 2
- 3WA
2018 · wikipedia2vec.github.io
- 4
- 5LR
I just noticed it takes literally ~5 minutes to train millions parameters on slow CPU...but before you call Yudkowsky that "it's over", an important note: the main bottleneck is the corpus size, params are just 'cleverness' but given limited info it's powerless. Anyway, here is the project: https://github.com/bggb7781-collab/lrnnsmdds/tree/main couple of notes: 1. single C file, no dependencies. Below are literally all the "dependencies", not even custom header (copy paste from the top of the single c file): #define _POSIX_C_SOURCE 200809L #include #include…
Apr 2026 · raw.githubusercontent.com
- 6LI
2018 · languagemodels.io
- 7IE
Quick note on how it works and how I've done my batch embedding engine IgniteMS. The whole thing runs as one process using Rust, reading input, tokenizing, packing batches, keeping the queue full. TensorRT handles inference. Python is only as a wrapper. I built it this way because when you use more than couple of GPUs, the GPUs stop being the problem. CPU cannot feed them fast enough. One A100 can go through batches faster than Python can tokenize and feed, so the GPU just sits there idle waiting for work. Most of my time went into optimizing this. At 8 GPUs that was basically the entire…
Jun 2026 · github.com
- 8MA
I've been working on training this small vision language model for the last month - excited to release the first prototype today! It is based on SigLIP (image encoder), Phi-1.5 (text model) and trained using the LLaVa-1.5 training dataset. It runs reasonably fast on CPU with ~8GB of RAM in full 32-bit precision. There's plenty of room to speed it up and reduce memory consumption by quantizing the model. I posted a video of it running on my M2 Macbook Air (on CPU not MPS, so performance should be comparable on other hardware) on Twitter to demonstrate inference speed:…
2023 · github.com
- 9ML
Aug 2026 · github.com
- 10LM
2021 · os.cohere.ai
- 11CT
I had been looking to try <500M parameter language models but you wouldn't find an API to try them anywhere, so I built this cloudflare hosted static website that hosts weights and built an inference runtime for these models that uses WebGPU and runs inference from your browser. These are only so useful in a multi-turn conversation but it's still interesting to see what you can pack in a <250mb model. I tried using ONNX versions earlier, but there were too many quirks of using them with language models and the TPS wasn't too impressive. Inspired by svenflow/webgpu-gemma, I put my codex…
May 2026 · chonklm.com
- 12D0
We used our platform to fine-tune a tiny text-to-SQL model using distillation from DeepSeek V3. Repo has instructions for how to replicate this. This is definitely not the best-performing model like this out there! But I found it surprising we were able to get to this much out of it: stone's throw away from a teacher 1000x the size! We also ran the same thing using the 4B Qwen and matched the teacher accuracy, though here the difference is merely 100x :) I find this pretty cool - obviously our distilled models can only do this one task and don't generalize, but that's often exactly what you…
Jan 2026 · github.com
- 13MC
Hey everyone! Many of you might have come across the Mamba paper a few days ago, which introduced an LLM based on a state space model architecture. The Mamba architecture is quite useful as its complexity scales subquadratically with input length and is therefore way more efficient than transformer models: https://github.com/state-spaces/mamba I got really excited about the paper, so I decided to fine-tune the model on a chat dataset. It turns that this actually worked quite well! The model is quite suitable for casual chatting, which honestly surprised me given that it…
2023 · github.com
- 14BO
Read the full blogpost at https://rach.codes/blog/Introducing-Bhumi (click on reader to see the technical breakdown!) AI inference should be fast, but in practice it’s painfully slow. Inference bottlenecks slow down LLM-powered chatbots and AI workflows everywhere. I built Bhumi to fix that. Bhumi is a Python library designed for developers, yet its performance-critical core is implemented in Rust (via PyO3) for near-native speed. This hybrid approach delivers up to 2.5x faster response times across providers like OpenAI, Anthropic, and Gemini—without changing the…
2025 · bhumi.trilok.ai
- 15SV
2017 · github.com
- 16VA
VibeSolve is an open-source tool that turns a plain-English description of an optimisation problem into Timefold code. Mathematical optimisation is a branch of mathematics and computer science that searches for the minimum/maximum of objective functions, and has applications in transport, logistics, scheduling etc. We are exploring where LLMs can add value in optimisation algorithm development, and where they get in the way. Right now, it works well for rapid prototyping. It does not create production-ready code and requires technical skills to use. It is noticeably better at creating…
Jun 2026 · vibesolve.ai
- 17DP
2023 · github.com
- 18IT
Jul 2026 · inflect-tts.geronimo-labs.com
- 19UD
I've been working a fair bit with DSPy lately, and I did some work in combining the benefits of vector search and LLMs (via a DSPy pipeline) to disambiguate records with a high degree of accuracy to help enrich a dataset. The blog post shows how this approach scales well, is very cost-effective and super concise - all it takes is < 100 lines of DSPy code and it all runs async. The code to reproduce is in this repo if anyone's interested (all tools are 100% free and open source, and the methodology will work with open weight LLMs too).…
2025 · blog.kuzudb.com
- 20GR
Hi everyone, wanted to share about gline-rs, an inference engine for GLiNER models written in Rust. This family of lightweight language models proved to be efficient at zero-shot Named Entity Recognition (NER) and other tasks such as Relation Extraction, while consuming less resources than large generative models (LLMs). This implementation has been written from the ground up in Rust, and supports both span- and token-oriented variants (for inference only). The goal is to provide a production-grade and user-friendly API in a modern and safe programming language, including a clean and…
2025 · github.com
- 21
- 22FM
As the title and description of the GitHub repo suggest, I’m working on a small project for purely educational purposes, with the goal of implementing generative model inference (small models capable of modeling 2D distributions) based on the Flow Matching paradigm in C. I’ve worked on generative AI models based on Flow Matching from a more “abstract” perspective, using frameworks like PyTorch, and I wanted to understand what goes on behind the scenes. The repository is still a work in progress and is also one of my first "serious" projects in C.
Jul 2026 · github.com
- 23NL
Refuel LLM (84.2%) outperforms trained human annotators (80.4%), GPT-3-5-turbo (81.3%), PaLM-2 (82.3%) and Claude (79.3%) across a benchmark of 15 text labeling datasets. It is a Llama-v2-13b base model, trained on over 2500 unique datasets (5.24B tokens) spanning categories such as classification, entity resolution, matching, reading comprehension and information extraction. Here is the interactive demo: https://labs.refuel.ai/playground. Pretty fun to play with!
2023
- 24LS
2019 · github.com
Ranked by how close each launch is in meaning, then by votes. Refine with a description →