전체 검색

AI 모델 84760

• qwen3vl8b-finetune-GGUF↗

  mradermacher

• QwQ-32B-Mlem-i1-GGUF↗

  mradermacher

• Salus-7B-GGUF↗

  mradermacher

• SketchThinker-R1-3B-GGUF↗

  mradermacher

• typhoon-ocr1.5-2b-i1-GGUF↗

  mradermacher

• typhoon2-qwen2vl-7b-vision-instruct-GGUF↗

  mradermacher

• ViAble-1.7-9B-i1-GGUF↗

  mradermacher

• indic-parler-tts↗

  naklitechie

  It is a fine-tuned version of Indic Parler-TTS Pretrained, trained on a **1,806 hours** multilingual Indic and English dataset. **Indic Parler-TTS Mini** can officially speak in 20 Indic languages, making it comprehensive for regional language technologies, and in English. The **21 languages** supported are: Assamese, Bengali, Bodo, Dogri, English, Gujarati, Hindi, Kannada, Konkani, Maithili, Malayalam, Manipuri, Marathi, Nepali, Odia, Sanskrit, Santali, Sindhi, Tamil, Telugu, and Urdu. Thanks to its **better prompt tokenizer**, it can easily be extended to other languages. This tokenizer has a larger vocabulary and handles byte fallback, which simplifies multilingual training.

• nemu-ocr-coreml↗

  nemu-pm

• Qwen3VL-8B-Abliterated-INT8-ConvRot↗

  NidAll

  A complete multimodal **Qwen3-VL text encoder for Qwen Image 2.1 Turbo in ComfyUI**, reconstructed from AtomicChat's BF16 GGUF release. Choose the **9.35 GB INT8 ConvRot** checkpoint for reduced storage, or **17.53 GB BF16** for the original reconstructed precision. Both include the entire BF16 vision tower and image-to-language mergers needed for image conditioning. **No separate mmproj is required.** The source modification affects the language encoder; the diffusion transformer and VAE are unchanged. The INT8 checkpoint matches the official Comfy-Org encoder's **exact file size and tensor precision policy**. It is **46.67% smaller than BF16**. Each file is a complete encoder; download one for ordinary use.