dgrauet
/

void-model-mlx-q8

void-transformer

Model card Files Files and versions

dgrauet/void-model-mlx-q8

Int8 quantization (group_size 64, transformer Linear weights only) of dgrauet/void-model-mlx, the MLX conversion of netflix/void-model.

Quantized with mlx-forge (mlx-forge convert void-model --quantize --bits 8).

Good quality/memory balance (~48 GB RAM recommended for the full two-pass pipeline). On 32 GB Macs use the q4 variant instead.

Usage

These weights are loaded by void-model-mlx:

python -m void_mlx.infer \
    --sample sample/BigBen \
    --pass1 weights/q8/void_pass1.safetensors \
    --pass2 weights/q8/void_pass2.safetensors \
    --base-model /path/to/CogVideoX-Fun-V1.5-5b-InP-mlx-q8 \
    --steps 30 --max-frames 13 --height 352 --width 624 \
    --output result.gif

Keep quantize_config.json next to the weights (the loader also infers bits/group_size from the weight shapes if it is missing).

Related Projects

void-model-mlx (inference): https://github.com/dgrauet/void-model-mlx
VideoX-Fun-mlx (engine): https://github.com/dgrauet/VideoX-Fun-mlx
Base model weights (q8): https://huggingface.co/dgrauet/CogVideoX-Fun-V1.5-5b-InP-mlx-q8
bf16 variant: https://huggingface.co/dgrauet/void-model-mlx
q4 variant: https://huggingface.co/dgrauet/void-model-mlx-q4

Files

config.json
quantize_config.json (bits=8, group_size=64)
void_pass1.safetensors (6.68 GB)
void_pass2.safetensors (6.68 GB)

Downloads last month: 27

MLX

Hardware compatibility

Log In to add your hardware

Quantized

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for dgrauet/void-model-mlx-q8

Base model

netflix/void-model

Finetuned

(4)

this model

Collection including dgrauet/void-model-mlx-q8

Void

3 items • Updated Apr 11