README
FastMetal-1.3B-QAD
3-step text-to-video, INT8 pre-quantized for Apple Silicon.
The entry-tier FastMetal model — a DMD2-distilled 1.3B Wan video model with a quantization-aware-trained INT8 DiT. Pre-quantized: no startup quantization, tiny download, runs on 16 GB-class Macs.
What's inside
| Path | Contents |
|---|---|
mlx_dit.safetensors / mlx_dit.json | INT8 (affine, group-64) DiT |
text_encoder/, vae/, tokenizer/, scheduler/ | everything needed to run standalone (fp16 UMT5 text encoder) |
Quickstart
Requires macOS with Apple silicon (MPS) and Python 3.11+:
pip install torch transformers mlx safetensors av imageio imageio-ffmpeg
git clone https://github.com/FastVideo/FastVideo.git
cd FastVideo
python examples/inference/basic/mlx_wan_prompt_to_video.py \
--model-root ./FastMetal-1.3B-QAD \
--mlx-checkpoint ./FastMetal-1.3B-QAD \
--prompt "a misty mountain river valley at sunrise"
Model details
| Base model | FastWan 2.1 T2V 1.3B |
| Distillation | DMD2, 3 denoising steps |
| Quantization | affine INT8, group size 64, QAT-trained |
| Resolution | 448×832 (480p), 77 frames |
| Flow shift | 8.0 |
| DiT weights | ~1.5 GB (INT8) |
Training
DMD2 distillation of the FastWan 2.1 T2V 1.3B teacher onto an INT8 student
on NVIDIA GB200 clusters, with quantization-aware training (affine INT8,
group 64). Training corpus: FastVideo/Wan-Syn_77x448x832_600k.
FastMetal family
| Model | Tier |
|---|---|
| FastMetal-1.3B-QAD | Entry — 16 GB+ Macs |
| FastMetal-5B-QAD | Mid — 720p |
| FastMetal-14B-QAD | Quality — 16 GB+ Macs |
Similar repos
README
FastMetal-1.3B-QAD
3-step text-to-video, INT8 pre-quantized for Apple Silicon.
The entry-tier FastMetal model — a DMD2-distilled 1.3B Wan video model with a quantization-aware-trained INT8 DiT. Pre-quantized: no startup quantization, tiny download, runs on 16 GB-class Macs.
What's inside
| Path | Contents |
|---|---|
mlx_dit.safetensors / mlx_dit.json | INT8 (affine, group-64) DiT |
text_encoder/, vae/, tokenizer/, scheduler/ | everything needed to run standalone (fp16 UMT5 text encoder) |
Quickstart
Requires macOS with Apple silicon (MPS) and Python 3.11+:
pip install torch transformers mlx safetensors av imageio imageio-ffmpeg
git clone https://github.com/FastVideo/FastVideo.git
cd FastVideo
python examples/inference/basic/mlx_wan_prompt_to_video.py \
--model-root ./FastMetal-1.3B-QAD \
--mlx-checkpoint ./FastMetal-1.3B-QAD \
--prompt "a misty mountain river valley at sunrise"
Model details
| Base model | FastWan 2.1 T2V 1.3B |
| Distillation | DMD2, 3 denoising steps |
| Quantization | affine INT8, group size 64, QAT-trained |
| Resolution | 448×832 (480p), 77 frames |
| Flow shift | 8.0 |
| DiT weights | ~1.5 GB (INT8) |
Training
DMD2 distillation of the FastWan 2.1 T2V 1.3B teacher onto an INT8 student
on NVIDIA GB200 clusters, with quantization-aware training (affine INT8,
group 64). Training corpus: FastVideo/Wan-Syn_77x448x832_600k.
FastMetal family
| Model | Tier |
|---|---|
| FastMetal-1.3B-QAD | Entry — 16 GB+ Macs |
| FastMetal-5B-QAD | Mid — 720p |
| FastMetal-14B-QAD | Quality — 16 GB+ Macs |