Search FoxAI

AI Models 2119

• Qwen-Image-2.1-PE-MTP-GGUF↗

  mozophe

  GGUF builds of the Qwen-Image 2.1 prompt enhancers (PE), with the Qwen3.5 multi-token prediction (MTP) head included, so llama.cpp can use speculative decoding (`--spec-type draft-mtp`). These files are made for the llama.cpp backend of ComfyUI-Qwen-Image-2.1-PromptEnhancer-MTP, which downloads them automatically. They also work with a plain `llama-server` (see Usage). The PE models don't ship an MTP head. It is grafted from the base model they were fine-tuned from, `Qwen/Qwen3.5-9B` (the same 15 `mtp.*` tensors the ComfyUI node uses). Community GGUFs of these models have no MTP head and no Q8_0.

• 01-ai_-_Yi-1.5-34B-Chat-gguf↗

  RichardErkhov

  Yi-1.5-34B-Chat - GGUF - Model creator: https://huggingface.co/01-ai/ - Original model: https://huggingface.co/01-ai/Yi-1.5-34B-Chat/ Yi-1.5 is an upgraded version of Yi. It is continuously pre-trained on Yi with a high-quality corpus of 500B tokens and fine-tuned on 3M diverse fine-tuning samples. Compared with Yi, Yi-1.5 delivers stronger performance in coding, math, reasoning, and instruction-following capability, while still maintaining excellent capabilities in language understanding, commonsense reasoning, and reading comprehension.

• Midnight-Miqu-70B-v1.5↗

  sophosympatheia

  This is a DARE Linear merge between sophosympatheia/Midnight-Miqu-70B-v1.0 and migtissera/Tess-70B-v1.6. This version is close in feel and performance to Midnight Miqu v1.0 but I think it picked up some goodness from Tess. Their EQ Bench scores are virtually the same and their post-EXL2 quant perplexity scores were the same too. However, Midnight Miqu v1.5 passes some tests I use that Midnight Miqu v1.0 fails, without sacrificing writing quality. This model was designed for roleplaying and storytelling and I think it does well at both. It may also perform well at other tasks but I have not tested its performance in other areas. Try the following context template for use in SillyTavern. It might help, although it's a little heavy on tokens. If you save the text as a .json file, you can import it directly.

• stable-diffusion-xl-refiner-1.0↗

  Stability AI

  Alternatively, we can use a two-stage pipeline as follows: First, the base model is used to generate latents of the desired output size. In the second step, we use a specialized high-resolution model and apply a technique called SDEdit (https://arxiv.org/abs/2108.01073, also known as "img2img") to the latents generated in the first step, using the same prompt. This technique is slightly slower than the first one, as it requires more function evaluations. - **Developed by:** Stability AI - **Model type:** Diffusion-based text-to-image generative model - **License:** CreativeML Open RAIL++-M License - **Model Description:** This is a model that can be used to generate and modify images based on text prompts. It is a Latent Diffusion Model that uses two fixed, pretrained text encoders (OpenCLIP-ViT/G and CLIP-ViT/L). - **Resources for more information:** Check out our GitHub Repository and the SDXL report on arXiv. For research purposes, we recommned our `generative-models` Github repository (https://github.com/Stability-AI/generative-models), which implements the most popoular diffusion frameworks (both training and inference) and for which new functionalities like distillation will be added over time. Clipdrop provides free SDXL inference.

• StartLux-Decision-v1-9B-Q4_K_M-GGUF↗

  startlux-models

  StartLux-Decision-v1-9B in Q4_K_M (4-bit) as a GGUF file for llama.cpp. The decision server in this repository turns it into the same typed decisions, with a probability for every option, as the original model. "Same decision" is the share of the 231 public JevBench items on which the file picks the same answer as the original weights run through the `startlux_decision` package, with the same prompts, readout and temperatures. Q8_0 and Q4_K_M are llama.cpp's standard quantizations of BF16, without an importance matrix. `-ngl 99` puts every layer on the GPU (CUDA or Metal); leave it out on a CPU-only machine. llama.cpp has to be recent enough for this model (build b10454 or newer).

• SamudrACE-CM4-piControl↗

  allenai

• musicgen-small↗

  facebook

  MusicGen is a text-to-music model capable of genreating high-quality music samples conditioned on text descriptions or audio prompts. It is a single stage auto-regressive Transformer model trained over a 32kHz EnCodec tokenizer with 4 codebooks sampled at 50 Hz. Unlike existing methods, like MusicLM, MusicGen doesn't require a self-supervised semantic representation, and it generates all 4 codebooks in one pass. By introducing a small delay between the codebooks, we show we can predict them in parallel, thus having only 50 auto-regressive steps per second of audio. MusicGen was published in Simple and Controllable Music Generation by *Jade Copet, Felix Kreuk, Itai Gat, Tal Remez, David Kant, Gabriel Synnaeve, Yossi Adi, Alexandre Défossez*. Four checkpoints are released: - **small** (this checkpoint) - medium - large - melody

• Ternary-Bonsai-2-27B-MTP↗

  ProCreations

  Updated September 18, 2026 with the best verified continued-training checkpoint: **r3-mtp**. The original Ternary Bonsai 2 27B PQ2_0 target is unchanged. The head is available as BF16 master weights and **integer Q8_0** inference weights, not FP8. **The +1.26% below is the extra gain over our previous trained MTP head, not the total gain from enabling MTP.** For an independent MTP-on versus MTP-off comparison, see the community reproduction below. Matched testing against the previous release measured **167.14 → 169.25 decode tokens/sec (+1.26%)** on eight unseen prompts with two repeats. This is a small measured gain, not a large speedup. Both arms used the same runtime and maximum 2 draft tokens.

• mms-tts-tam↗

  facebook

  This model is part of Facebook's Massively Multilingual Speech project, aiming to provide speech technology across a diverse range of languages. You can find more details about the supported languages and their ISO 639-3 codes in the MMS Language Coverage Overview, and see all MMS-TTS checkpoints on the Hugging Face Hub: facebook/mms-tts. VITS (**V**ariational **I**nference with adversarial learning for end-to-end **T**ext-to-**S**peech) is an end-to-end speech synthesis model that predicts a speech waveform conditional on an input text sequence. It is a conditional variational autoencoder (VAE) comprised of a posterior encoder, decoder, and conditional prior. A set of spectrogram-based acoustic features are predicted by the flow-based module, which is formed of a Transformer-based text encoder and multiple coupling layers. The spectrogram is decoded using a stack of transposed convolutional layers, much in the same style as the HiFi-GAN vocoder. Motivated by the one-to-many nature of the TTS problem, where the same text input can be spoken in multiple ways, the model also includes a stochastic duration predictor, which allows the model to synthesise speech with different rhythms from the same input text.

• Qwen3.8-Flash-Next-IQ4_XS-GGUF↗

  julianmb

  second-generation quant: same trunk type (iq4_xs), but calibrated with a 926-entry imatrix (1,024 chunks) via the ROCmFPX banded quantizer, and the 51B PLE lookup table **left at q8_0** (no `--tensor-type` cut). 5.56 bpw, 115 giB — 24 giB bigger than the 91g PLE file. pick M2 when you want the **best measured quality** and don't mind the 24 giB: at ≤32k plain it matches PLE, and at 32k MTP it measured fastest. for shallow-depth MTP speed take the 91g PLE; for deep 128k+ MTP the static 116g had a small in-sweep edge (17.4 vs 13.3/13.5 — within the same-config spread, n=1 caveat). every quant descends from an F16 that was byte-verified against the official `Qwen/Qwen3.8-Flash-Next-FP8` checkpoint: hyper-connection norms folded to (1 + w) (97/97 tensors — the converter bug that produces deterministic garbage is fixed in our pipeline), PLE fp8 scale applied, expert stacking identity probed 512x3, GDN v-head reorder checked. details: https://github.com/julianmb/haloq38flash