Hugging Face Models (Daily)

This repository provides a fully merged model. The LoRA adapter has already been merged into the base model.



Open: https://huggingface.co/HamadaMayu/qwen2.5-7b-agent-trajectory-lora-marged-mixed_db_alf_1to1

This repository provides a **LoRA adapter** fine-tuned from **Qwen/Qwen3-4B-Instruct-2507** using **LoRA + Unsloth**.



Open: https://huggingface.co/HamadaMayu/qwen3-4b-agent-trajectory-lora-marged-alfworld_v4

This repository provides a **LoRA adapter** fine-tuned from **Qwen/Qwen3-4B-Instruct-2507** using **LoRA + Unsloth**.



Open: https://huggingface.co/HamadaMayu/qwen3-4b-agent-trajectory-lora-marged-alfworld_v5

This repository provides a fully merged model fine-tuned from Qwen/Qwen3-4B-Instruct-2507 using LoRA + Unsloth.



Open: https://huggingface.co/HamadaMayu/qwen3-4b-agent-trajectory-lora-marged-dbbench_v4

This repository provides a **LoRA adapter** fine-tuned from **Qwen/Qwen3-4B-Instruct-2507**.



Open: https://huggingface.co/HamadaMayu/qwen3-4b-agent-trajectory-lora-marged-mixed_db_alf_1to1

| Parameter | Value | | :-------- | :---: | | **direction_index** | 17.48 | | **attn.o_proj.max_weight** | 3.99 | | **attn.o_proj.max_weight_position** | 20.00 | | **attn.o_proj.min_weight** | 3.24 | | **attn.o_proj.min_weight_distance** | 13.70 | | **mlp.down_proj.max_weight** | 3.18 | | **mlp.down_proj.max_weight_position** | 28.59 | | **mlp.down_proj.min…



Open: https://huggingface.co/heretic-org/Nanbeige4.1-3B-heretic

| Parameter | Value | | :-------- | :---: | | **direction_index** | 21.70 | | **attn.o_proj.max_weight** | 1.49 | | **attn.o_proj.max_weight_position** | 23.54 | | **attn.o_proj.min_weight** | 1.41 | | **attn.o_proj.min_weight_distance** | 14.03 | | **mlp.down_proj.max_weight** | 1.36 | | **mlp.down_proj.max_weight_position** | 38.71 | | **mlp.down_proj.min…



Open: https://huggingface.co/heretic-org/XortronCriminalComputingConfig-heretic

**See Qwen3.5-397B-A17B MLX in action - demonstration video - coming soon**



Open: https://huggingface.co/inferencerlabs/Qwen3.5-397B-A17B-MLX-9bit

https://hf.co/vinai/PhoWhisper-tiny with ONNX weights to be compatible with Transformers.js.



Open: https://huggingface.co/ivanthepevt/PhoWhisper-tiny-for-L-N

**Inventor:** Konstantin Vladimirovich Grabko **Email:** grabko@cmsmanhattan.com **Date:** December 21, 2025 **Needed:** A sponsor for Llama 405b Distilation to high quality as origonal Llama 405b or Data Center with cooperation



Open: https://huggingface.co/kgrabko/JiRackTernary_405b

Bio-8B-it is an 8B parameter biomedical instruction-tuned language model built on top of Qwen 3-8B. The model was fine-tuned using Supervised Fine-Tuning (SFT) with QLoRA via the PEFT framework.



Open: https://huggingface.co/khazarai/Bio-8B-it

Bio-8B-it is an 8B parameter biomedical instruction-tuned language model built on top of Qwen 3-8B. The model was fine-tuned using Supervised Fine-Tuning (SFT) with QLoRA via the PEFT framework.



Open: https://huggingface.co/khazarai/Bio-8B-it-8bit

```bash brew install llama.cpp



Open: https://huggingface.co/KnutJaegersberg/JoyAI-LLM-Flash-Q6_K-GGUF

Try LFM • Documentation • LEAP • Blog



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

- Ninja-v1



Open: https://huggingface.co/Local-Novel-LLM-project/Vecteus-v1

Converted to MLX for Apple Silicon.



Open: https://huggingface.co/McG-221/GLM-4.7-Flash-absolute-heresy-mxfp8-gs32

4-bit MLX quantized version of the **text** model from Qwen/Qwen3.5-397B-A17B.



Open: https://huggingface.co/mlx-community/Qwen3.5-397B-A17B-4bit

4-bit nvfp4 MLX quantized version of the **text** model from Qwen/Qwen3.5-397B-A17B.



Open: https://huggingface.co/mlx-community/Qwen3.5-397B-A17B-nvfp4

Introducing Kimi-Dev: A Strong and Open-source Coding LLM for Issue Resolution Kimi-Dev Team



Open: https://huggingface.co/mrx171/Kimi-Dev-72B

This is a **Nanbeige4.1-3B** fine-tune, produced at the request of redaihf through P-E-W's Heretic (v1.2.0) abliteration engine with Magnitude-Preserving Orthogonal Ablation enabled.



Open: https://huggingface.co/MuXodious/Nanbeige4.1-3B-absolute-heresy

Open: https://huggingface.co/Nono25evfzvsf/jzkjfkzf

Music Flamingo: Scaling Music Understaning in Audio Language Models



Open: https://huggingface.co/nvidia/music-flamingo-think-2601-hf

!1



Open: https://huggingface.co/prithivMLmods/gemma-3-27b-it-abliterated-FP8

> [!Note] > This repository contains model weights and configuration files for the post-trained model in the Hugging Face Transformers format. > > These artifacts are compatible with Hugging Face Transformers, vLLM, SGLang, etc.



Open: https://huggingface.co/Qwen/Qwen3.5-397B-A17B

- **Developed by:** rikunarita - **License:** apache-2.0 - **Finetuned from model :** rikunarita/Qwen3-4B-Thinking-2507-Genius



Open: https://huggingface.co/rikunarita/Qwen3-4B-Thinking-2507-Genius-Buddy

| Parameter | Value | | :-------- | :---: | | **direction_index** | per layer | | **attn.o_proj.max_weight** | 1.44 | | **attn.o_proj.max_weight_position** | 16.31 | | **attn.o_proj.min_weight** | 1.38 | | **attn.o_proj.min_weight_distance** | 13.41 |



Open: https://huggingface.co/servic/gpt-oss-20b-reap-0.4-mxfp4-heretic

This repository provides **pre-converted GGUF weights** for running **microsoft/Phi-4-multimodal-instruct** with a **quantized language model** and a **multimodal projector (mmproj)** on top of a specialized llama.cpp fork.



Open: https://huggingface.co/ShayanCyan/phi4-multimodal-quantisized-gguf

Open: https://huggingface.co/shefgan/real-me

This is a domain-adapted version of Google's EmbeddingGemma-300M model, fine-tuned specifically for Norwegian health and medical information retrieval tasks.



Open: https://huggingface.co/thivy/embeddinggemma-300m-norwegian-health

Website | Docs | API Key



Open: https://huggingface.co/tonyassi/face-swap-video

There is not yet support in ik_llama.cpp, though an open issue and work on qwen3next coming in now.



Open: https://huggingface.co/ubergarm/Qwen3.5-397B-A17B-GGUF

> [!Note] > This repository contains model weights and configuration files for the post-trained model in the Hugging Face Transformers format. > > These artifacts are compatible with Hugging Face Transformers, vLLM, SGLang, etc.



Open: https://huggingface.co/unsloth/Qwen3.5-397B-A17B

To run Qwen3.5 locally - Read our Guide! Unsloth Dynamic 2.0 achieves superior accuracy & outperforms other leading quants.



Open: https://huggingface.co/unsloth/Qwen3.5-397B-A17B-GGUF

Open: https://huggingface.co/veeceey/Huihui-Devstral-Small-2-24B-Instruct-2512-abliterated-mlx-4bit

Open: https://huggingface.co/veeceey/Huihui-Qwen3-32B-abliterated-mlx-4bit

Open: https://huggingface.co/veeceey/Huihui-Qwen3-Coder-Next-abliterated-mlx-4bit

**!!!АХТУНГ!!!** Данная модель не создана компанией **"СБЕР"** это кастомный LM.



Open: https://huggingface.co/ViorikaAI-org/RuGPT-5-small-v1

**!!!АХТУНГ!!!** Данная модель не создана компанией **"СБЕР"** это кастомный LM.



Open: https://huggingface.co/ViorikaAI/RuGPT-5-small-v1

**Model ID:** webxos/microclaw-for-openclaw-version-2026.2.16 **Tags:** openclaw, fallback-agent, grpo, vae, kv-cache, soul.md, huggingface-space, gguf, llm-distillation



Open: https://huggingface.co/webxos/microclaw-for-openclaw-2026.2.16

Open: https://huggingface.co/webxos/microclaw-for-openclaw-version-2026.2.16

This repo contains a LeRobot SmolVLA checkpoint exported from: outputs/train/smolvla_so101_test_ft_2cam_v1b/checkpoints/last



Open: https://huggingface.co/Zekai-Chen/smolvla_so101_test_ft_2cam_v1b

We are releasing **BCLM-1 Preview**, an open family of flagship language models that deliver strong, predictable performance across a range of tasks.



Open: https://huggingface.co/bcml/bclm-1-small-preview

**BharatGen** presents **Param-2-17B-MoE-A2.4B**, a large-scale Mixture-of-Experts (MoE) language model designed to deliver high model capacity while retaining the inference efficiency of a much smaller dense model. It uses a **Hybrid MoE architecture** with **17B total parameters**, while activating only **2.4B parameters per token**.



Open: https://huggingface.co/bharatgenai/Param2-Thinking

Join Our 💬 WeChat | 🧩 Discord community. MiniMax Agent | ⚡️ API | MCP | MiniMax Website 🤗 Hugging Face | 🚀 Hugging Face API | 🐙 GitHub | 🤖️ ModelScope | 📄 License: Modified-MIT



Open: https://huggingface.co/cyankiwi/MiniMax-M2.5-AWQ-4bit

Multi-task DistilBERT model for detecting Personally Identifiable Information (PII) in text with coreference resolution. Fine-tuned from distilbert-base-cased.



Open: https://huggingface.co/DataikuNLP/kiji-pii-model

INT8-quantized ONNX version of the Kiji PII detection model for efficient CPU inference. Detects Personally Identifiable Information (PII) in text with coreference resolution.



Open: https://huggingface.co/DataikuNLP/kiji-pii-model-onnx

LFM2.5-1.2B-Thinking-Claude-4.6-Opus-Heretic-Uncensored-DISTILL



Open: https://huggingface.co/DavidAU/LFM2.5-1.2B-Thinking-Claude-4.6-Opus-Heretic-Uncensored-DISTILL

LFM2.5-1.2B-Thinking-Deepseek-3.1-Heretic-Uncensored-DISTILL



Open: https://huggingface.co/DavidAU/LFM2.5-1.2B-Thinking-Deepseek-3.1-Heretic-Uncensored-DISTILL

LFM2.5-1.2B-Thinking-Gemini-Pro-1000-Heretic-Uncensored-DISTILL



Open: https://huggingface.co/DavidAU/LFM2.5-1.2B-Thinking-Gemini-Pro-1000-Heretic-Uncensored-DISTILL

LFM2.5-1.2B-Thinking-Gemini-Pro-Heretic-Uncensored-DISTILL



Open: https://huggingface.co/DavidAU/LFM2.5-1.2B-Thinking-Gemini-Pro-Heretic-Uncensored-DISTILL