YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
LongCat-Video-Avatar 1.5 — MLX (int8_fused)
MLX-format weights for LongCat-Video-Avatar 1.5, converted for Apple-silicon (MLX) inference with a fused int8 DiT.
Contents
base_model_int8_fused/— DiT (int8, fused) +audio_adapter.safetensorslora/dmd_lora.safetensors— DMD distillation LoRA (few-step generation)vae/— WAN-VAE decoder
Companion (not included here): text encoder + tokenizer → shraey/umt5-xxl-mlx; Whisper → mlx-community/whisper-large-v3-mlx; upscaler → shraey/flashvsr-v1.1-mlx.
Attribution & license
- Original model: meituan-longcat/LongCat-Video-Avatar-1.5 and meituan-longcat/LongCat-Video — MIT License, © Meituan LongCat team.
- This repository is a derivative work: format conversion to MLX + int8 quantization. Distributed under the same MIT terms.
- Downloads last month
- -
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support