Hugging Face's logo

Join the conversation

Join the community of Machine Learners and AI enthusiasts.

Sign Up

All HF Hub posts
Banaxi-Techย 
posted an update 1 day ago
view post
Post
5733
BananaMind 2 Pro is training!
The current checkpoint (ONLY 20% DONE) GETS #6 On the entire Open SLM Leaderboard.
We are going to release the first public preview on August 2-4 (estimated from speed)

Give us a follow to know when it releases!
  • 2 replies
ยท
FredyRivera-devย 
posted an update about 17 hours ago
view post
Post
2820
We wrote a full technical guide on how to train a bilingual (ES/EN) LLM from scratch: TinyQwen.

Covers:
- Hybrid architecture based on Qwen3.5
- Pre-training with 15B tokens
- Cost benchmark between H200 and B200
- Post-training with SFT + LoRA
- Full code and data, open source

With ~$11 of compute on an H200 we ran an initial training run, enough to validate the full architecture and pipeline.

Blog post: https://aquiles-ai.vercel.app/blog/tinyqwen-from-scratch

Technical feedback welcome, especially from anyone looking to replicate the pipeline with more compute.
  • 2 replies
ยท
owensongย 
posted an update 2 days ago
view post
Post
6281
I just released Inflect v2, pushing complete TTS below 4M parameters.

Inflect-Nano-v2 fits every learned inference component, including waveform generation, into 3.96M parameters (15.97 MB FP32), making it one of the smallest publicly available complete TTS models. Inflect-Micro-v2 stays under 10M at 9.36M parameters (37.53 MB), using the extra capacity for clearer, more natural speech.

This was not about building a toy that technically produces audio. I wanted to see how much useful TTS could actually fit at this scale.

Nano stayed surprisingly close to Micro across the reported release metrics, but if you want to hear Inflect at its best, start with Micro:
owensong/Inflect-Micro-v2

Text goes in and 24 kHz speech comes out without a second learned model. Both models have PyTorch and ONNX releases: PyTorch runs on CPU or CUDA, while the browser demo runs ONNX locally through WebGPU with a WASM fallback. The Inflect code and weights are released under Apache 2.0.

Try Nano and Micro side by side:
owensong/Inflect-v2

Some people may remember the much rougher Inflect-Nano-v1 experiment. V2 was rebuilt and retrained to address its biggest weaknesses: metallic audio, unstable timing, weak prosody, and difficulty with unseen text. Nano-v2 is smaller while sounding substantially cleaner and more stable.

In an anonymous community blind test across nine voices, Micro and Nano placed second and third by normalized preference rate. Full results, samples, limitations, and inference instructions are on the model pages.

Inflect-Nano-v2:
owensong/Inflect-Nano-v2

I built Inflect independently with a limited training budget. If you test it, give it something difficult: unusual names, numbers, abbreviations, symbols, or awkward punctuation. Send me the exact text when something breaks. If you build something with it, I would genuinely love to see it.
  • 1 reply
ยท
Nymboย 
posted an update 3 days ago
view post
Post
5154
Introducing Inflect-v2, two exceptionally small, open-weight English TTS models at just 3.9M and 9.3M parameters. Both generate speech multiple times faster than real-time on CPU. Despite their size, Inflect-v2 delivers quality that is competitive with much larger lightweight TTS systems, including KittenTTS, Piper, and Supertonic-3.

CPU, CUDA, PyTorch, and ONNX are supported. Apache 2.0.

See it for yourselves:
owensong/Inflect-Micro-v2
owensong/Inflect-Nano-v2

Try the Demos:
Nymbo/Inflect-TTS (unlimited CPU usage)
owensong/Inflect-v2 (ultra-fast ZeroGPU usage)
  • 5 replies
ยท
SeaWolf-AIย 
posted an update 2 days ago
view post
Post
6217
๐Ÿ–ผ๏ธ POCKET-Image โ€” the POCKET series goes visual: character-perfect text in any language, on-device

A new model in VIDRAFT's POCKET family. POCKET put 35B-class models on phones and no-GPU PCs. POCKET-Image carries the same "big capability, small hardware" idea into image generation โ€” and fixes the one thing nearly every image model gets wrong: text.

Type "์•ˆ๋…•ํ•˜์„ธ์š”" into a typical model and you get "์•ˆใ…๊ธฐ." Hangul alone composes 11,172 syllable blocks; Arabic connects its letters; Thai stacks marks. Diffusion models draw scripts as shapes, so they smear. POCKET-Image renders every glyph exactly โ€” ํ•œ๊ตญ์–ด ยท ไธญๆ–‡ ยท ๆ—ฅๆœฌ่ชž ยท ุงู„ุนุฑุจูŠุฉ (RTL) ยท เน„เธ—เธข ยท Latin and more โ€” onto any scene you describe.

What it is:
โ€ข 100% accurate text, any language โ€” where global models produce gibberish
โ€ข Any background from a prompt โ€” text is optional (empty โ†’ a pure image)
โ€ข No GPU, no NPU โ€” runs on plain CPU + RAM via the POCKET-Core engine
โ€ข Measured footprint: 8.6 GB (RTX 3050/4060) ยท 4.5 GB (offloaded, 6 GB cards) ยท 13.4 GB (MacBook, 16 GB+)
โ€ข Windows ยท macOS ยท Linux ยท fully local, no cloud

Built on the open, commercial-friendly Z-Image (Apache-2.0) foundation.

Honest note: the text is the guaranteed-correct part โ€” the surrounding scene is ordinary generation, so a busy foreground can crowd the letters. We say so; clean backgrounds stay razor-sharp.

๐ŸŽจ Studio โ€” generate right here, any language:
FINAL-Bench/POCKET-Image-Studio

๐Ÿงฉ Model card:
FINAL-Bench/POCKET-Image-Zimage

๐Ÿ“š The POCKET collection:
https://huggingface.co/collections/FINAL-Bench/pocket-models
ProCreationsย 
posted an update 3 days ago
view post
Post
2014
what should i do for 200 followers... hmmmm... top reply to this post (most emojis) wins, just not anything too crazy
  • 11 replies
ยท
SeaWolf-AIย 
posted an update 4 days ago
view post
Post
3063
POCKET now speaks Gemma 4 โ€” a 26B model that loads in every app, and runs on your PC with no GPU

We're adding a Gemma-4 sibling to POCKET: POCKET-26B, built from Google's Gemma-4-26B-A4B (Apache-2.0). Our flagship POCKET-35B is a Qwen-family MoE and needs a recent llama.cpp; POCKET-26B trades a little size for the thing people kept asking for โ€” it just loads, everywhere, today: Ollama, LM Studio, PocketPal, MLX, any stock llama.cpp. No fork, no bleeding-edge runtime, no CUDA, no cloud.

It's a sparse Mixture-of-Experts (25.2B total, ~4B active per token), so the work per token stays small โ€” a real 26B that generates on a CPU with no graphics card.

Two things make it stand out:

1) Universal compatibility. Gemma 4 is a standard, widely-supported architecture, so POCKET-26B runs on the tools you already have โ€” no waiting for your app to add a new model type.

2) Quality that survives compression. Measured GPQA-Diamond (198 q, greedy):
โ€ข Full base: 67.7%
โ€ข POCKET-26B Q4_K_M (17 GB): 67.7% โ€” lossless
โ€ข POCKET-26B Q2_K (11 GB): 67.2% โ€” near-lossless, at 11 GB

Live, on a CPU-only box (our demo Space โ€” POCKET-26B vs Bonsai-27B, same machine, same stock llama.cpp): POCKET-26B โ‰ˆ 19 tok/s vs Bonsai โ‰ˆ 6 tok/s โ†’ about 3ร— faster generation, no GPU. (Honest notes: shared CPU box, sequential race; a dedicated machine is faster.)

Where it fits in the family:
โ€ข POCKET-35B (Qwen MoE) โ€” bigger, top-tier, needs a recent llama.cpp.
โ€ข POCKET-26B (Gemma 4) โ€” loads in any app, quality-robust when compressed. The demo runs the Q4_K_M build; Q2_K (11 GB) is the smallest footprint. For a true โ‰ค8 GB phone, the 5 GB POCKET-KR (Qwen) is still the pick.

Try it and grab it:
๐Ÿ–ฅ๏ธ Live demo (Gemma4-based, answering on a CPU, no GPU): FINAL-Bench/POCKET-26B-CPU
๐Ÿ“ฆ POCKET-26B-GGUF (Q4_K_M 17 GB ยท Q2_K 11 GB): FINAL-Bench/POCKET-26B-GGUF
๐Ÿ“š POCKET collection: https://huggingface.co/collections/FINAL-Bench/pocket-models
  • 3 replies
ยท
sergiopaniegoย 
posted an update 1 day ago
view post
Post
420
quick reminder! ๐Ÿšจ

tomorrow (Tuesday, July 28), we're back with Class 3 of the Training Agents live series

๐Ÿง  what: reinforcement learning for training agents (GRPO): how it works, how to implement it in TRL, and end-to-end examples
๐Ÿ—“๏ธ when: Tuesday, July 28 - ๐Ÿ•” 5:00 PM CEST / 8:30 PM IST
๐Ÿ“ where: Live on @huggingface 's X, YouTube, and LinkedIn

live: https://www.youtube.com/watch?v=ztdTed5egrM

class 1: https://x.com/SergioPaniego/status/2069382207618379813
class 2: https://x.com/SergioPaniego/status/2075180665184686187
Banaxi-Techย 
posted an update 3 days ago
view post
Post
3514
We're excited to release BananaMindBench Leaderboard, our leaderboard for BananaMind Base Bench 1.1.
It measures model performance on a variety of different tasks:
Language Completion
Common sense too
World Knowledge
Context Tracking
Quantitative
Logical Reasoning
Code Completion
Each has a different score and 1 overall score.
Submit your own model:
BananaMind/BananaMindBench-Leaderboard
Check it out:
BananaMind/BananaMindBench-Leaderboard
  • 10 replies
ยท
ProCreationsย 
posted an update 2 days ago
Recently active users