Search FoxAI
AI Models 2119
- Phi-3-small-8k-instruct↗
Microsoft
🎉 **Phi-3.5**: [[mini-instruct]](https://huggingface.co/microsoft/Phi-3.5-mini-instruct); [[MoE-instruct]](https://huggingface.co/microsoft/Phi-3.5-MoE-instruct) ; [[vision-instruct]](https://huggingface.co/microsoft/Phi-3.5-vision-instruct) The Phi-3-Small-8K-Instruct is a 7B parameters, lightweight, state-of-the-art open model trained with the Phi-3 datasets that includes both synthetic data and the filtered publicly available websites data with a focus on high-quality and reasoning dense properties. The model belongs to the Phi-3 family with the Small version in two variants 8K and 128K which is the context length (in tokens) that it can support. The model has underwent a post-training process that incorporates both supervised fine-tuning and direct preference optimization for the instruction following and safety measures. When assessed against benchmarks testing common sense, language understanding, math, code, long context and logical reasoning, Phi-3-Small-8K-Instruct showcased a robust and state-of-the-art performance among models of the same-size and next-size-up.
- Lumimaid-v0.2-12B-GGUF↗
NeverSleep
NOTE: As explained on Mistral-Nemo-Instruct-2407 repo, it's recommended to use a low temperature, please experiment! As some people have told us our models are sloppy, Ikari decided to say fuck it and literally nuke all chats out with most slop. Our dataset stayed the same since day one, we added data over time, cleaned them, and repeat. After not releasing model for a while because we were never satisfied, we think it's time to come back!
- NVIDIA-Nemotron-3.5-Lightning-30B-A3B-MLX-4bit↗
TensorFold
A native Apple-silicon conversion of nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 , quantized with stock 4-bit affine weights and packaged for MLX-LM and oMLX. This repository contains a stock 4-bit affine MLX conversion of NVIDIA Nemotron 3.5 Lightning. The source is a 30B-total / 3B-active hybrid mixture-of-experts model that interleaves Mamba-2, sparse MoE, and attention layers. The upstream tokenizer, chat template, and generation configuration are preserved. The decode figure is the median of three greedy 256-token runs after a 32-token Metal-kernel warm-up. It is a practical local reference, not a controlled cross-platform benchmark. Prompt length, context growth, sampler settings, memory pressure, thermal state, and MLX/oMLX versions can materially change performance.
- sam2.1-hiera-small↗
facebook
Repository for SAM 2: Segment Anything in Images and Videos, a foundation model towards solving promptable visual segmentation in images and videos from FAIR. See the SAM 2 paper for more information. The official code is publicly release in this repo. Refer to the demo notebooks for details.
- pegasus-xsum↗
Google
See Docs: here Authors: Jingqing Zhang, Yao Zhao, Mohammad Saleh and Peter J. Liu on Dec 18, 2019 We train a pegasus model with sampled gap sentence ratios on both C4 and HugeNews, and stochastically sample important sentences. The updated the results are reported in this table.
- csm-1b↗
sesame
- Wan2.2-T2V-A14B-Diffusers↗
Wan-AI
💜 Wan    ｜    🖥️ GitHub    |   🤗 Hugging Face    |   🤖 ModelScope    |    📑 Technical Report    |    📑 Blog    |   💬 WeChat Group    |    📖 Discord    We are excited to introduce **Wan2.2**, a major upgrade to our foundational video models. With **Wan2.2**, we have focused on incorporating the following innovations: - 👍 **Effective MoE Architecture**: Wan2.2 introduces a Mixture-of-Experts (MoE) architecture into video diffusion models. By separating the denoising process cross timesteps with specialized powerful expert models, this enlarges the overall model capacity while maintaining the same computational cost.
- Qwen3-Coder-30B-A3B-Instruct-AWQ↗
dark-side-of-the-code
AWQ **W4A16** (4-bit weights, group size 128) quantization of `Qwen/Qwen3-Coder-30B-A3B-Instruct`, produced with llm-compressor for serving on a single 24 GB GPU with vLLM. Qwen3-Coder-30B-A3B is a Mixture-of-Experts model (128 experts, 8 active per token) using standard **GQA** attention — so it serves on consumer Ada GPUs with the stock attention backends, no special kernels required. Notes: - At `--max-model-len 65536` on a 24 GB card the KV pool holds ~118K tokens (~1.8x concurrency). The base model's native context is 262K; 64K is the practical cap at 24 GB. - Tool calling works via vLLM's first-class `qwen3_coder` parser. - Non-thinking model — no reasoning parser needed. - Pass `--override-generation-config` if your client doesn't send sampling params, or vLLM will default to `temperature=1.0`, which is far too hot for coding.
- DeepSeek-V4-Pro-Qwen3.5-9B-MTP-GGUF↗
Jackrong
🔬 DeepSeek-V4-Pro Distillation 🧠 9B Parameters 📐 ~250K Math & STEM Samples ⚡ MTP & NVFP4 Evaluated 💡 What is DeepSeek-V4-Pro-Qwen3.5-9B? DeepSeek-V4-Pro-Qwen3.5-9B is a reasoning-focused fine-tune of Qwen3.5-9B , distilled from responses generated by DeepSeek-V4-Pro in Max Effect mode . The release concentrates its supervised signal on mathematics and STEM problem solving while preserving the efficiency and practical deployability of the 9B parameter class. 🧩 Structured Reasoning Learns rigorous problem decomposition and multi-step solution patterns from a strong teacher.
- sepformer-whamr16k↗
speechbrain
This repository provides all the necessary tools to perform audio source separation with a SepFormer model, implemented with SpeechBrain, and pretrained on WHAMR! dataset with 16k sampling frequency, which is basically a version of WSJ0-Mix dataset with environmental noise and reverberation in 16k. For a better experience we encourage you to learn more about SpeechBrain. The given model performance is 13.5 dB SI-SNRi on the test set of WHAMR! dataset. The system expects input recordings sampled at 16kHz (single channel). If your signal has a different sample rate, resample it (e.g, using torchaudio or sox) before using the interface. To perform inference on the GPU, add `run_opts={"device":"cuda"}` when calling the `from_hparams` method.