Downloads · 30 days
0
mmh3turbo-bundles
yunfengwang/mmh3turbo-bundles
mmh3turbo-bundles is a text-to-video model from yunfengwang. Use it when you need video from a text prompt. It is set up for mlx. The card lists the license as mit.
Prepared weights for mmh3turbo, an MLX port of MiniMax-H3 that generates video and audio on Apple Silicon using hand-written int8 Metal kernels.
Access
Public
Updated Aug 7, 2026
Repo size
36.2 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin21 GB · 58%
At a glance
- Task
- Text-to-Video
- Library
- mlx
- License
- mit
- Access
- Public
- Created
- Aug 7, 2026
- Updated
- Aug 7, 2026
- SHA
- 1a3d56ee
Base models
Model card
From the Hugging Face model README
mmh3turbo bundles
Prepared weights for mmh3turbo, an MLX port of MiniMax-H3 that generates video and audio on Apple Silicon using hand-written int8 Metal kernels.
You do not need to download these by hand:
uvx mmh3turbo "a rainbow colored skunk leaps over a mossy log in a supermarket"
Contents
| File | Size | What it is |
|---|---|---|
dit.bin + dit.idx | 19.5 GiB | The 50-block DiT as per-output-channel int8, requantized from the reference fp8-E4M3 checkpoint. Laid out for this port's fused Metal kernels, so it is not interchangeable with MLX-native quantized repos. |
qwen3vl_4bit.safetensors | 14.2 GiB | The Qwen3-VL-32B text tower's first 50 layers in MLX 4-bit (group 64). Cuts text encoding from 43.8 s to 3.3 s. |
Why a 4-bit text tower
The tower is consumed from a 7.9 GB GGUF rather than the 66.7 GB unquantized encoder, and streaming it costs ~52 s per new prompt because its 50 layers are 25.16 B elements and numpy k-quant dequantization runs at ~487 M/s. Quantizing once to MLX 4-bit moves the matmuls onto the GPU and turns that into an mmap'd load. The source is Q2_K — 2.6 effective bits — so 4 bits loses nothing against it: cosine similarity against the streamed path is 0.9994.
What the DiT bundle is not
This is int8 with per-output-channel scales plus per-token activation scales,
which is what gemm_i8 and flash_i8 consume. It will not load in
implementations built on mx.quantize. SmoothQuant was tried and rejected: it
made the end-to-end velocity error worse (16.75% -> 63.43%), because the source
is already fp8-E4M3 with 3 mantissa bits and has no range left to migrate.
License
MIT for the packaging. The underlying weights are MiniMax-H3's; see MiniMaxAI/MiniMax-H3.
- Task
- Text-to-Video
- Library
- mlx
- License
- mit
- Created
- Aug 7, 2026
- Updated
- Aug 7, 2026