Software Product @huggingface🤗
Joined January 2018
- Tweets 3,881
- Following 1,060
- Followers 3,956
- Likes 7,060
823 Photos and videos
13h
great listen!
Had a fun conversation with @HarryStebbings on all things Model Independence, Software Factories, pricing wars…
And sprinkled in some hot takes that I may or may not regret sharing 😳
4
273
19h
Pro Tip: you can upload your Fable traces to both PRIVATE datasets and buckets on @huggingface
1
23
2,173
Caleb retweeted
you can distill from hf fable traces btw
14
34
736
57,892
Caleb retweeted
Jun 12
MiniMax M3, Open-Weight, Now On Hugging Face , with only ~428B parameters and ~23B activated parameters
Weights:
huggingface.co/MiniMaxAI/Min…
MiniMax Sparse Attention:
huggingface.co/papers/2606.1…
Introducing MiniMax M3: The First Open-Weights Model to Combine Three Frontier Capabilities
- Coding & Agentic Frontier: 59.0% SWE-Bench Pro, 66.0% Terminal Bench 2.1, 34.8% SWE-fficiency, 28.8% KernelBench Hard, 74.2% MCP Atlas
- MiniMax Sparse Attention scales context to 1M
- Natively Multimodal from Step Zero
API: platform.minimax.io
Token Plan: platform.minimax.io/subscrib…
🚀New! MiniMax Code: code.minimax.io
Weights & Tech Report in ~10 Days
114
328
2,750
627,665
Caleb retweeted
Jun 11
.@dh7net, SVP of Image Research, said it best: "The HF infra is a no-brainer."
A big unlock for teams working with large datasets for training, especially when they update over time. Read how Jasper used @huggingface as the creation and storage backbone for MONET: huggingface.co/storage/testi…
2
11
24
15,423
Caleb retweeted
Jun 11
Asked a question so dangerous they sent my ass to haiku
79
394
8,987
209,344
Caleb retweeted
Jun 10
Fable is the new leader on CADGenBench!
Still long way to go:
6
46
568
36,148
Caleb retweeted
Jun 9
ok who put Hugging Face in Rick and Morty
8
18
363
47,604
Jun 9
Cohere drops North Mini Code - an open weights model optimized for code generation
> 30B-A3B
> Apache 2.0 Licensed
> Available on @huggingface
1
6
24
1,911
Today we're announcing a multi-million-dollar strategic partnership with @huggingface.
The Hub is now the exclusive home for everything we build: every open model we release, every private run, every proprietary dataset, and every agent trace. All of it lives there.
10
27
343
32,744
Caleb retweeted
Jun 9
Super excited to announce that @arcee_ai is the first major American AI lab to replace AWS S3 with Hugging Face for ALL their models and datasets, public AND private 🔥🔥🔥
Multi-million $ partnership to support American open-source AI, let’s go!
41
60
512
60,785
Caleb retweeted
Jun 8
Introducing FrontierCode: a coding eval that raises the bar for difficulty & quality. Each task took 40 hrs of work by leading open-source maintainers.
Models write sloppy code that works but isn’t maintainable. Our eval is first to measure: would you actually merge this code?
234
313
4,284
2,507,526
Jun 8
This is super cool! Benchmark measuring how well agents can use CAD to build a valid, geometrically correct 3D part!
We need more benchmarks that model real world tasks like this!
Jun 8
Introducing CADGenBench: measure how well AI systems produce engineering-grade 3D parts!
While current models can generate 3D parts, they are far from precise enough to build functional parts. We built a benchmark to systematically measure their capabilities on two tasks:
1. Generation from an engineering drawing of a part
2. Editing: given an existing STEP file and a requested change
The benchmark is tool-agnostic. It makes no assumptions about how you build the model. You can vary the LLM, and you can vary the environment. Use build123d, Onshape, Autodesk, or a model without an LLM entirely. We open sourced the scoring engine and a reference baseline on top of build123d.
A collaboration between Hugging Face and @mecadoinc!
Submission space: huggingface.co/spaces/Huggin…
Code repository: github.com/huggingface/cadge…
2
2
515
Caleb retweeted
Jun 5
We just dropped Gemma 4 Quantization-Aware Training (QAT) checkpoints on Hugging Face!
All Gemma 4 model sizes and their drafters are now optimized with QAT to cut memory requirements and maximize on-device performance!
96
283
2,856
500,324
Caleb retweeted
Agent traces are the new fuel.
Looking fw to announce `trl` official support for agent traces for training💥
(w/ `datasets` v5, coming out tmr?)
Pick your local, synthetic, or community traces and train your own specialized Agent
🔜trl sft --dataset-name julien-c/synthtraces
Jun 4
Today I'm launching a new project called SynthTraces 🔥
It is a minimal codebase to generate synthetic coding agent session traces using Pi (from @badlogicgames)
I wanted a large number of coding-agent traces, so I built a tiny harness where two models talk to each other:
- an open model (served via HF Inference Providers) plays the coding agent. It gets read bash access to a real open source codebase (the huggingface OSS projects)
- a small local model (llama.cpp) plays the human user, asking simple questions like "how do I run this?" or "how is CI set up?"
The result is more than 2,000 Pi session traces which can be used to train or fine-tune LLMs, and optimize them for Pi 🤯
And ofc everything is published on @huggingface ✅
3
9
41
36,274
Jun 4
imagine being able to resume one of @karpathy auto-research sessions to learn and try different things!
Jun 4
One cool hack of of agent sessions being a json file is you can just resume someone else's trace!
Great for extracting valuable context or long compacted sessions! Try the prompt with your agent below!
1
1
945