サイト内検索

AI モデル 2119

• lfm2-electronics-vlm-v2↗

  erikku-sama

• Ternary-Bonsai-2-27B-Heretic-Satellite↗

  HybridRMD

  > ### Teacher model > > This model is a **LoRA decision-distillation of > `convaiinnovations/laya`** > by **Convai Innovations**. laya supplies every training label — the > `should_compact` / `preserve` / `level` / `mode` decisions this model emits. > No laya weights are redistributed here; the teacher is credited as the source > of the supervision, which it is. > > If you want the same decisions without this model, go and use laya. A **Qwen3.8-27B Bonsai 2 (Heretic)** derivative carrying a small LoRA that turns a long working context into one structured compaction decision. **Text only.** No vision tower, no projector. If you feed it images it will not do anything useful with them.

• Phi-3-medium-4k-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-Medium-4K-Instruct is a 14B 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 Medium version in two variants 4K and 128K which is the context length (in tokens) that it can support. The model has undergone 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-Medium-4K-Instruct showcased a robust and state-of-the-art performance among models of the same-size and next-size-up.

• flux-3-image-klein-base-FP8↗

  unsloth

  This is an FP8 / INT8 quantized version of FLUX 3 Image Klein Base (the undistilled checkpoint, 50 steps with CFG 3.5). - The INT8 file is Black Forest Labs' own `flux3_image_klein_int8_convrot.safetensors` from black-forest-labs/flux-3-image-klein-comfy, unchanged (INT8 weights with ConvRot rotation). The FP8 file is ours. - Optimized for efficient inference with reduced memory footprint: the transformer is 5.81 GB (INT8) or 6.42 GB (FP8) instead of 11.1 GB. Same-seed LPIPS vs the bf16 model (lower is better): 0.022 INT8, 0.036 FP8. - Loads in Unsloth Desktop. The text encoder (Qwen3-VL-4B-Instruct) and the VAE are unchanged and identical to the distilled release; take them from black-forest-labs/flux-3-image-klein-base or black-forest-labs/flux-3-image-klein. The text encoder rows are scored against the bf16 text encoder with the bf16 transformer, so they measure only the encoder. The VAE stays bf16. This repo also carries the upstream diffusers folders from black-forest-labs/flux-3-image-klein-base, unchanged (`modular_model_index.json`, `text_encoder/`, `transformer/`, `vae/`), so Unsloth Desktop downloads every component from here.

• musicgen-large↗

  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 - medium - **large** (this checkpoint) - melody

• sam2-hiera-base-plus↗

  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.

• MEETING_SUMMARY↗

  knkarthick

  Model obtained by Fine Tuning 'facebook/bart-large-xsum' using AMI Meeting Corpus, SAMSUM Dataset, DIALOGSUM Dataset, XSUM Dataset!

• VoxCPM1.5↗

  openbmb

  **Key Improvements:** - 🔊 **Higher Quality**: 44.1kHz sampling rate preserves more high-frequency details for better voice cloning - ⚡ **More Efficient**: Reduced token rate (6.25Hz) lowers computational cost while maintaining performance - 🎓 **Fine-tuning Support**: Train personalized voice models with SFT or LoRA **Note**: Output quality depends on the prompt speech quality. VoxCPM-0.5B remains fully supported with backward compatibility. VoxCPM is a novel tokenizer-free Text-to-Speech (TTS) system that redefines realism in speech synthesis. By modeling speech in a continuous space, it overcomes the limitations of discrete tokenization and enables two flagship capabilities: context-aware speech generation and true-to-life zero-shot voice cloning.

• wav2vec2-large-xlsr-53-chinese-zh-cn↗

  jonatasgrosman

  Fine-tuned facebook/wav2vec2-large-xlsr-53 on Chinese using the train and validation splits of Common Voice 6.1, CSS10 and ST-CMDS. When using this model, make sure that your speech input is sampled at 16kHz. This model has been fine-tuned thanks to the GPU credits generously given by the OVHcloud :) The script used for training can be found here: https://github.com/jonatasgrosman/wav2vec2-sprint

• Krea2-Turbo-Distill-2step-LoRA↗

  lvladikov

  > 🆕 **New: a 3-step LoRA** takes Krea 2 Turbo from 8 steps down to 3, one step more than this LoRA, with more fine detail than both this and the 4-step LoRA (fine texture 1.13× the 8-step teacher's on its 22 test prompts). This 2-step LoRA stays the fastest of the three. **A quarter of the steps · 4× faster denoising (text encoding and VAE decode excluded) · fine detail at 0.92–1.02× the teacher's across all 12 trained resolutions, the grid bands at its level · colour close to the teacher's · 1 point missing of 352 on a blind prompt-adherence rubric · teacher preferred on 12 of 66 judged renders · 51,195 training samples on recorded teacher trajectories · 32 days on one RTX 3090 · released from settled weights; the project's final checkpoint.** A LoRA for **Krea 2 Turbo** that takes the model from **8 steps down to 2** — Turbo's own weights and sigmas, guidance 0.0, a quarter of the denoising passes — aiming at the best quality two steps can give. It is for **fast previews and drafts**; the **4-step LoRA** remains the recommendation for quality renders.