Open-source software/tools/datasets, news, research et al. regarding #machinelearning and #datascience, with a focus on #deeplearning.
Global
Joined February 2012
- Tweets3.9K
- Following825
- Followers8.2K
- Likes5.8K
DeepLearning.Hub retweeted
DeepSeek V4.1-Flash-1017 can now run locally! 🐳
We made 1-bit retain ~88.2% accuracy while shrinking size from 508GB to 250GB (−52%).
Run on a 128GB+ RAM device.
V4.1 Flash 1017 is the strongest open model to date
Guide: unsloth.ai/docs/models/deeps…
GGUF: huggingface.co/unsloth/DeepS…
🚀 DeepSeek-V4.1-Flash-1017 is here!
Stronger benchmark performance, faster task completion, and shorter outputs than the previous version—a Pareto improvement in performance and cost. Now live on our website, app, and API.
🛠️ Optimized for @DeepSeekHarness, Tencent CodeBuddy, WorkBuddy, and more. Upgrade to DeepSeek Harness v0.2.1-rc.1 or later for the best experience with the new model.
1/4
DeepLearning.Hub retweeted
🚀 Introducing DeepInfer, our in-house inference framework, now fully open source!
Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. We’re sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable.
🔗 github.com/deepseek-ai/DeepI…
1/4
DeepLearning.Hub retweeted
Beam is a token-efficient, frontier-scale model for your coding agents, comes with its own harness 🔥
A23B/501B model, also comes with NVFP4 & FP8 and day-0 transformers support 🤗
check out our blog ⤵️
Beam is now available across platforms, with open weights under the Apache 2.0 license - in collaboration with an ecosystem of 40+ partners.
We're releasing:
- Weights, tokenizer, and reference implementation
- BF16, FP8 and NVFP4 weights and a DFlash2 drafter
- The open source harness we built and trained Beam on: Mirror CLI
Weights and model card on Hugging Face: huggingface.co/reflection
DeepLearning.Hub retweeted
Hy Memory 🧠 New memory model for AI agents from @TencentHunyuan
- 30B MoE / 3B active
- Apache 2.0
- Extracts important memories from conversations
- Updates outdated memories automatically
DeepLearning.Hub retweeted
Meta releases Muse Spark Open, a new SOTA open model.
Run on a Mac Studio via Dynamic 2-bit Unsloth GGUFs.
Muse-Spark-Open-816B-A42B supports text, images, video and audio input with 1M context.
GGUF: huggingface.co/unsloth/Muse-…
Guide: unsloth.ai/docs/models/muse-…
Muse Spark Open -- weights releasing today -- runs great on new Mac Studios thanks to work with @johnternus and Apple.
DeepLearning.Hub retweeted
On device 35B/A3B MoE + 1B VLM 🔥
huggingface.co/collections/o…
🚀 Introducing MiniCPM-V 4.7, our new open-source MLLM series!
Advancing the Performance–Efficiency Pareto Frontier.
It brings strong multimodal capabilities to iOS and Android devices with extreme efficiency.
Two new models:
🔹 MiniCPM-V 4.7 35B-A3B — Bring SOTA performance to local devices with newly introduced LES (Lookahead Expert Scheduling) technique, a new milestone of on-device multimodal AI.
🔹 MiniCPM-V 4.7 1B — Lightweight, efficient multimodal AI. Easier to deploy.
⚡ What's new?
SOTA performance: Competitive results across 34 benchmarks spanning multimodal understanding, STEM reasoning, OCR and long-video comprehension.
Fast on-device inference: LES predicts and preloads expert weights to reduce inference stalls, boosting 5× speedup on iPad Pro M4 (16GB) with byte-for-byte identical outputs. It enables MiniCPM-V 4.7 35B-A3B to achieve 17+ tokens/s on iPad Pro M4 (16GB) and 13+ tokens/s on iPhone 17 Pro Max, while costing only 3GB memory 🤯
Efficient visual processing: 4× and 16× hybrid visual token compression. Even at 16× compression, 35B-A3B scores 90.4 on OCRBench, 5.2 points ahead of Qwen3.6 35B-A3B.
Easier customization: We introduce Self-Replay Fine-Tuning, a simple way to customize new target capabilities at low cost while maintaining general performance.
Broad hardware support: Day-0 adaptations for Apple, Intel, AMD, Arm and Rockchip, with many more platforms supported through FlagOS!
Download, deploy, and build something we haven't imagined yet. 👀👇
🔗 GitHub: github.com/OpenBMB/MiniCPM-V
🤗 Hugging Face: huggingface.co/collections/o…
DeepLearning.Hub retweeted
Tiny decision models you can run everywhere:
> d1-3B has impeccable performance in text and vision
> d1-omni-600M supports text, vision, and audio!
Try our @huggingface space with interactive games today
Today we release Open d1: two open-weight multimodal models in our d1 decision model family.
> d1-3B: text + vision
> d1-omni-600M: text + image or text + audio
> Real-time decision making anywhere, from data centers such as @nvidia DGX to RTX workstations to Jetson at the edge.
1/
DeepLearning.Hub retweeted
You can now train your own Decision model like Jev locally!
We increased Qwen3.5 0.8B’s aggregate accuracy from 20.7% to 74.3% across 3 decision benchmarks - on just 4GB VRAM.
Turn any LLM like Qwen3.8, Gemma 4 into decision models with our open-source Unsloth repo.
We fine-tuned with a Clef head using Unsloth and LoRA (r=64) for one epoch, increasing downstream accuracy from 30–37% to 78%.
GitHub: github.com/unslothai/unsloth
Guide and Notebooks: unsloth.ai/docs/basics/train…
DeepLearning.Hub retweeted
Google releases EmbeddingGemma 2, a new open model that runs locally on 0.5GB RAM.
The 740M parameter Apache 2.0 model combines a 270M text model with vision (170M) + audio (300M).
Run & train the model via Unsloth.
GGUF: huggingface.co/unsloth/embed…
Guide: unsloth.ai/docs/models/embed…
DeepLearning.Hub retweeted
We’re releasing a broad range of new mathematical results produced by an internal frontier model.
We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results.
github.com/openai/math
DeepLearning.Hub retweeted
We release Whistle: speech to text in a 16.9MB file, running on the CPU in the same engine as Needle. It mostly beats Whisper base with 9x less file size and 6x speed. Whistle supports English, German, French, Spanish, Italian, Dutch and Polish.
cactuscompute.com/blog/whist…
DeepLearning.Hub retweeted
We release Needle 3: A Sliceable 8-29MB automation foundation model that can match DeepSeek V4 Flash.
One set of weights, every depth from 2 to 20 layers a model of its own, 25-121M parameters at CQ2-bit, built on our Simple Attention Networks and running locally at up to 4k tokens/sec decode speed on a Raspberry Pi 5.
Needle does not chat. Every turn is a function call: give it the tools your app exposes and it picks the right ones and fills every argument from what the user said, or hand it a schema and it returns a typed record. Ask for something no tool covers and you get an empty list, not a guess.
That trade is lets 121M parameters trained on 360B tokens of structured data beat models 10x their size on mobile tool calls and match 2-3x bigger models on structured JSON extraction.
It runs on mobiles, wearables, smart home devices, small robots and microcontrollers, with prebuilt engines for macOS, Linux, Windows, Android, iOS, watchOS, tvOS, the browser and WASI hosts. Try it in your browser: cactuscompute.com/needle
DeepLearning.Hub retweeted
Ollama now supports Jev-like decision models all locally.
Use decision models like Nimble for tasks like ticket triaging, model routing, and content moderation.
ollama pull nimble
Here’s Nimble playing Ollama racer through the new local /v1/systemone API by making decisions in real-time. 🏎️
DeepLearning.Hub retweeted
1/ In RL environments, reward design is everything.
Here's what we learned from the 7,780 environments Xiaomi open-sourced for MiMo 🧵
Wake up, ppl 👀
Xiaomi just open-sourced ~7K of the RL environments it used to train MiMo on HuggingFace, across domains like code, cyber, general, music and web dev
I think this is one of the biggest things to happen in frontier open-source RL environment data.
Indepth analysis and learnings coming soon 👀
DeepLearning.Hub retweeted
1/ Today we're introducing Cua-S1-4B-0.2, the first multimodal decision model trained with RLOO on live computer-use tasks, using task-completion rewards.
Text and multimodal adapters are available under Apache-2.0: github.com/trycua/cua
huggingface.co/cua-ai/cua-s1…
DeepLearning.Hub retweeted
All open-source datasets, specs, firmware, and 3d models:
- Eidon Glove: github.com/Eidon-AI/eidon-gl…
- Eidon Trackers: github.com/Eidon-AI/eidon-tr… + whitepaper drive.google.com/drive/u/5/f…
- Eidon Simulation Environment: github.com/Eidon-AI/eidon-si…
- Tracker + POV Video Data (~1,200h): huggingface.co/datasets/eido… + huggingface.co/datasets/eido…
- Video-Only Data: huggingface.co/buckets/eidon…
DeepLearning.Hub retweeted
MiMo-V2.6 is out and the results seem to put it right after GPT-5.6 Sol 🥹
same arch as V2.5, comes with 1M context window and MTP drafter
waiting for AA results 👀
huggingface.co/XiaomiMiMo/Mi…
DeepLearning.Hub retweeted
1,274 hours of egocentric video with paired 7-IMU arm tracking, now CC-BY-4.0 on @huggingface.
@eidon_ai, a robotics company, is winding down and has released all their data: people doing laundry, cleaning, dishes and cooking. 13,451 recordings, 9 TB.
huggingface.co/eidon-ai
DeepLearning.Hub retweeted
Today, we’re announcing Ternary Bonsai 2 27B.
Based on Qwen3.8 27B, Bonsai 2 27B is 9x smaller than its full-precision counterpart while retaining 98.2% of its aggregate benchmark performance.
Two months after the first Bonsai 27B release, the biggest change is quality. The footprint remains 5.9 GB, but the gap to full precision has narrowed materially, with particularly strong gains in agentic coding, multimodal reasoning, and long-horizon tool use.
Ternary Bonsai 2 27B is available today under Apache 2.0.
DeepLearning.Hub retweeted
Astra from @OpenAI took control of the SO-ARM101 and started painting. Very impressive results!
SO-ARM is a 3D printed robot arm you can build yourself for around $150. Connected with LeRobot library it lets you collect data and train models at home.
