Hugging Face Models (Daily)

SmolVLA is a compact, efficient vision-language-action model that achieves competitive performance at reduced computational costs and can be deployed on consumer-grade hardware.



Open: https://huggingface.co/adroitLee/smolvla_260128_ep50_fullseq_Syringe_cam1_bs16_s100000

--- language: - en - es license: mit tags: - pytorch - text-classification - weather - minimalism - educational - overfit datasets: - synthetic metrics: - accuracy model-index: - name: atacama results: - task: type: text-classification name: Weather Classification metrics: - type: accuracy value: 0.999 name: Accuracy ---



Open: https://huggingface.co/AtacamaLLM/atacama

!Untitled design (2)



Open: https://huggingface.co/BeefyDoesAI/Number-E

Open: https://huggingface.co/bharatddd/Whisper-Marathi-Finetuned

LoRA adapters for the Humanities tutor in the Canis.teach suite.



Open: https://huggingface.co/CanisAI/teach-humanities-ministral-3b-r2

LoRA adapters for the Science tutor in the Canis.teach suite.



Open: https://huggingface.co/CanisAI/teach-science-ministral-3b-r2

Gemma-3-4B-VL-it-GLM-4.7-Flash-Heretic-Uncensored-Thinking



Open: https://huggingface.co/DavidAU/Gemma-3-4B-VL-it-GLM-4.7-Flash-Heretic-Uncensored-Thinking

name: Shelly AI



Open: https://huggingface.co/fingerclose/Shelly

!Sala



Open: https://huggingface.co/Finisha-F-scratch/Sala

**Easter Spirit 2B** is a compact seasonal model fine-tuned for **warm, cheerful, and family-friendly** text generation. This repository provides **GGUF** builds optimized for **local inference** using the **llama.cpp** ecosystem and compatible runtimes.



Open: https://huggingface.co/ghostai1/Easter-Spirit2b

**Halloween2B-Poltergeist v1.0** is a compact seasonal model tuned for **spooky, cinematic, and playful** text generation. This repository provides **GGUF** builds optimized for **local inference** with **llama.cpp** and compatible runtimes.



Open: https://huggingface.co/ghostai1/Halloween2B_Poltergeist

A 120B parameter LLM fine-tuned with GRPO (Group Relative Policy Optimization) for GPU kernel generation, evaluated on KernelBench L1.



Open: https://huggingface.co/Jarrodbarnes/KernelBench-RLVR-120b

```bash brew install llama.cpp



Open: https://huggingface.co/l-nmch/bge-reranker-v2-m3-Q8_0-GGUF

The MERaLiON-SpeechEncoder is a speech foundation model designed to support a wide range of downstream speech applications, like speech recognition, intent classification and speaker identification, among others. This version was trained on **200,000 hours of predominantly English data including 10,000 hours of Singapore-based speech**, to cater to the spee…



Open: https://huggingface.co/MERaLiON/MERaLiON-SpeechEncoder-v1

- **Developed by:** NamuTechnology - **License:** apache-2.0 - **Finetuned from model :** WorldOpenTechnology/Araptor-1



Open: https://huggingface.co/NamuTechnology/NamuLM

This repository contains only a LoRA adapter trained on GSM8K. Users must load the base model separately and attach the adapter.



Open: https://huggingface.co/Neural-Hacker/OpenMath

These are MXFP4\_MOE quantizations of the model GLM-4.7-Flash-Claude-Opus-4.5-High-Reasoning-Distill.



Open: https://huggingface.co/noctrex/GLM-4.7-Flash-Claude-Opus-4.5-High-Reasoning-Distill

A selectively quantized version of Qwen3-30B-A3B optimized for local inference on Apple Silicon, with preserved precision on code analysis experts.



Open: https://huggingface.co/PKSGIN/qwen3-30b-selective-quant

A selectively quantized version of Qwen3-30B-A3B optimized for local inference on Apple Silicon, with preserved precision on code analysis experts.



Open: https://huggingface.co/PKSGIN/qwen3-30b-selective-quant-mlx

Recently Added Z-Image Models:



Open: https://huggingface.co/pmczip/Z-Image-Turbo_Models

📟 PaGeR(Panoramic Geometry Estimation) Depth Estimation Model Card



Open: https://huggingface.co/prs-eth/PaGeR-metric-depth

农村低压配电网电压异常检测实验模型检查点。



Open: https://huggingface.co/Sheldon123z/rural-voltage-detection-models

GritHopper: Decomposition-Free Multi-Hop Dense Retrieval 🌐 Website | 🏗️ GitHub Repo | 📃 Paper



Open: https://huggingface.co/UKPLab/GritHopper-7B

+ Это кастомная модель (LLM), обученная на моем железе. Совпадения с датасетами или чем то другим это случайность. **АХТУНГ!!!** МОДЕЛЬ ПРЕДОБУЧЕНА ДО СРЕДНЕГО УРОВНЯ!!! ДООБУЧЕНИЕ НА ВАШИХ ДАННЫХ ЖЕЛАТЕ…



Open: https://huggingface.co/ViorikaAI/ViorikaLM-1.2B

👉 **Source Code & Training Scripts** | **Dataset**



Open: https://huggingface.co/YiMeng-SYSU/roberta-logic-sentiment-zh

👋 Join our WeChat and Discord community 📍 Use GLM-OCR's API



Open: https://huggingface.co/zai-org/GLM-OCR

The Llama Nemotron Embedding 1B model is optimized for **multilingual and cross-lingual** text question-answering retrieval with **support for long documents (up to 8192 tokens) and dynamic embedding size (Matryoshka Embeddings)**. This model was evaluated on 26 languages: English, Arabic, Bengali, Chinese, Czech, Danish, Dutch, Finnish, French, German, Heb…

Daily delta: +17,322 downloads / +8 likes

Open: https://huggingface.co/nvidia/llama-nemotron-embed-1b-v2

Try LFM • Documentation • LEAP

Daily delta: +13,160 downloads / +101 likes

Open: https://huggingface.co/LiquidAI/LFM2.5-1.2B-Thinking

**Best OCR model .** LightOnOCR-2-1B is our flagship OCR model, refined with RLVR training for maximum accuracy. We recommend this variant for most OCR tasks.

Daily delta: +6,872 downloads / +31 likes

Open: https://huggingface.co/lightonai/LightOnOCR-2-1B

XLM-RoBERTa 기반 한국어/영어 이름 성별 분류 모델

Daily delta: +6,807 downloads / +0 likes

Open: https://huggingface.co/solonsophy/name-gender-classifier-ko

Try LFM • Documentation • LEAP

Daily delta: +6,673 downloads / +26 likes

Open: https://huggingface.co/LiquidAI/LFM2.5-1.2B-Instruct

I added a small injection at the end of the original chat_template.jinja to support a GLM-like switch: you can try to disable the “thinking/reasoning” mode in your chat completion request via "chat_template_kwargs": {"enable_thinking": False}.

Daily delta: +4,167 downloads / +2 likes

Open: https://huggingface.co/QuantTrio/Step3-VL-10B-AWQ

weighted/imatrix quants of https://huggingface.co/Emperorizzis/ASTRA-14B-Thinking-v1

Daily delta: +3,679 downloads / +0 likes

Open: https://huggingface.co/mradermacher/ASTRA-14B-Thinking-v1-i1-GGUF

weighted/imatrix quants of https://huggingface.co/0xSero/MiniMax-M2.1-REAP-30

Daily delta: +3,674 downloads / +0 likes

Open: https://huggingface.co/mradermacher/MiniMax-M2.1-REAP-30-i1-GGUF

🚀 Spaces Demo, 🔧 Github

Daily delta: +2,993 downloads / +21 likes

Open: https://huggingface.co/neuphonic/neutts-nano

**A.X K1** is a large-scale Mixture-of-Experts (MoE) language model designed for efficient high-capacity reasoning and instruction following. The model contains **519 billion total parameters**, with **33 billion active parameters**, enabling strong performance while maintaining practical inference efficiency.

Daily delta: +1,406 downloads / +5 likes

Open: https://huggingface.co/skt/A.X-K1

Released under Apache 2.0 License with a 256K context window, Jamba2 3B enables developers to build reliable AI applications for edge environments. For more details, read the full release blog post.

Daily delta: +1,365 downloads / +16 likes

Open: https://huggingface.co/ai21labs/AI21-Jamba2-3B

This requirement arises because **GENERator** employs a 6-mer tokenizer. If the input sequence length is not a multiple of **6**, the tokenizer will append an '' (out-of-vocabulary) token to the end of the token sequence. This can result in uninformative subsequent generations, such as repeated 'AAAAAA'.

Daily delta: +1,123 downloads / +1 likes

Open: https://huggingface.co/GenerTeam/GENERator-v2-eukaryote-1.2b-base

!image

Daily delta: +885 downloads / +19 likes

Open: https://huggingface.co/janhq/Jan-v3-4B-base-instruct

'Make knowledge free for everyone'

Daily delta: +804 downloads / +0 likes

Open: https://huggingface.co/DevQuasar/deepseek-ai.DeepSeek-V3.2-GGUF

- As of January 23, 2025, **INF-Retriever-v1** ranks both **No.1** on the Automated Heterogeneous Information Retrieval Benchmark of version 24.04 & 24.05(AIR-Bench), showcasing its cutting-edge performance in heterogeneous information retrieval tasks.

Daily delta: +709 downloads / +1 likes

Open: https://huggingface.co/infly/inf-retriever-v1

History:

Daily delta: +651 downloads / +0 likes

Open: https://huggingface.co/nphSi/Z-Image-Lora

 

Daily delta: +597 downloads / +6 likes

Open: https://huggingface.co/briaai/FIBO

weighted/imatrix quants of https://huggingface.co/vanta-research/mox-tiny-1

Daily delta: +455 downloads / +1 likes

Open: https://huggingface.co/mradermacher/mox-tiny-1-i1-GGUF

Jamba2 Mini is an open source small language model built for enterprise reliability. With 12B active parameters (52B total), it delivers precise question answering without the computational overhead of reasoning models. The model's SSM-Transformer architecture provides a memory-efficient solution for production agent stacks where consistent, grounded output…

Daily delta: +396 downloads / +21 likes

Open: https://huggingface.co/ai21labs/AI21-Jamba2-Mini

!Word Error Rate (WER) across multiple Vietnamese datasets *Figure 1: ASR WER comparison across different model and CSPs. This does not include Punctuation and Capitalisation errors. Details about each testset are mentioned below.*

Daily delta: +388 downloads / +33 likes

Open: https://huggingface.co/nvidia/parakeet-ctc-0.6b-Vietnamese

weighted/imatrix quants of https://huggingface.co/vanta-research/atom-80b

Daily delta: +385 downloads / +0 likes

Open: https://huggingface.co/mradermacher/atom-80b-i1-GGUF

static quants of https://huggingface.co/beita6969/FlowSteer-8b

Daily delta: +381 downloads / +0 likes

Open: https://huggingface.co/mradermacher/FlowSteer-8b-GGUF

This model is a fine-tuned version of None. It has been trained using TRL.

Daily delta: +377 downloads / +0 likes

Open: https://huggingface.co/MohamedGomaa30/spark-tts-normazlied-masri-mega

'Make knowledge free for everyone'

Daily delta: +342 downloads / +0 likes

Open: https://huggingface.co/DevQuasar/arcee-ai.Trinity-Large-Preview-GGUF