FastWan 2.2 TI2V-5B for MLX (Smarty Pants)

The video model that Smarty Pants owners' Macs download to make short clips from a text prompt.

  • model.safetensors: the 3-step transformer from FastVideo/FastWan2.2-TI2V-5B-FullAttn-Diffusers (commit 3e187042), renamed from Diffusers names to the original Wan names and laid out for MLX, bfloat16.
  • t5_encoder.safetensors, vae.safetensors: the UMT5 prompt reader and the VAE from Wan-AI/Wan2.2-TI2V-5B (commit 921dbaf3), converted for MLX with mlx-video-with-audio 0.1.37.
  • tokenizer/: the UMT5 tokenizer files from the FastWan repository, unchanged.

Changes from the originals: weight names and layouts only (no retraining). Both originals are Apache-2.0; this copy is too (see LICENSE).

Downloads last month

-

Downloads are not tracked for this model. How to track
Safetensors
Model size
5B params
Tensor type
BF16
ยท
MLX
Hardware compatibility
Log In to add your hardware

Quantized

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for smartypantsnet/FastWan2.2-TI2V-5B-mlx-bf16

Finetuned
(5)
this model