Views
No views yet
mlx-video Wan2.1 substrate; the only net-new surface is the
reference-injection path (refs encoded as trailing temporal latent frames + dual-scale chained CFG).| File | What |
|---|---|
transformer-bf16.safetensors | DiT, bf16 (2.84 GB) |
transformer-4bit.safetensors | DiT, int4 group-size 64 (0.98 GB) — embeds/time/head kept hi-precision, cosine 0.99633 vs bf16 |
vae-encoder.safetensors / vae-decoder.safetensors | Wan2.1 16-ch VAE (bf16) |
t5_encoder.safetensors | umT5-XXL text encoder (bf16) |
config.json | architecture + sampling defaults |
guide_img=5.0, guide_text=7.5), FlowUniPC scheduler (shift 5, 50 steps).1from phantom_wan_mlx import pipeline_mlx as P # github.com/xocialize/phantom-wan-mlx
2P.s2v(
3 "two friends walking together in a park",
4 ["subjectA.png", "subjectB.png"],
5 "out.mp4",
6 size=(832, 480), frame_num=81,
7)WanModel (825/826 keys; the gap is the computed RoPE buffer);
forward + VAE numerics inherited from the published mlx-video / Bernini-R Wan2.1 substrate. Streaming
decode is bit-exact to whole-sequence decode on mx.cpu.