InsecureErasure/Z-Image-Turbo-NVFP4

A bust portrait of a woman in her mid-twenties with messy dark hair tied in a loose bun, wearing a worn denim jacket over a gray hoodie.
She is leaning her elbows on a washing machine, her chin resting on her folded hands. Behind her, a row of industrial dryers against a tiled wall,
with one dryer door hanging open. Above the dryers, a handwritten sign taped to the wall says 'OUT OF ORDER' in black marker,
with a small smiley face drawn on it. To her left, a plastic basket overflows with unfolded clothes. To her right, a vending machine glows green,
displaying 'SOAP $1.50' on a small digital screen. The light is cool and buzzing, like fluorescent tubes overhead. She looks tired but amused
with a faint smirk.

#!/bin/bash
# NVFP4 baseline + MXFP8 for sensitive layers + BF16 at critical points.
# Refiners: block 0 fully MXFP8, block 1 outputs kept in BF16.
# Last QKV (layer 29), late adaLN (22-29), and refiner outputs in BF16.
# All main-trunk w1 (gate) projections in MXFP8.
convert_to_quant -i $1 \
  --nvfp4 --zimage --comfy_quant --save-quant-metadata \
  --custom-type mxfp8 \
  --custom-layers "layers\.(10|16|26)\.attention\.qkv\.weight|layers\.(27|28)\.attention\.qkv\.weight|layers\.(0|1)\.attention\.out\.weight|layers\.(3|6|9|11|12|13|14|19|20|26)\.attention\.out\.weight|layers\.(27|28|29)\.attention\.out\.weight|layers\.(3|4|5|6|7|8|9|10|11|12|13|14|15|16|17|18|19|20|21|22|23|24|25|26)\.feed_forward\.w1\.weight|layers\.(27|28|29)\.feed_forward\.w1\.weight|layers\.(16|17|18|19|20|21)\.adaLN_modulation\.0\.weight|context_refiner\.(0|1)\.attention\.qkv\.weight|context_refiner\.(0|1)\.feed_forward\.w1\.weight|context_refiner\.(0|1)\.feed_forward\.w2\.weight|context_refiner\.(0|1)\.feed_forward\.w3\.weight|noise_refiner\.(0)\.attention\.(qkv|out)\.weight|noise_refiner\.(0)\.feed_forward\.(w1|w2)\.weight|noise_refiner\.(1)\.feed_forward\.w1\.weight" \
  --exclude-layers "layers\.(29)\.attention\.qkv\.weight|layers\.(22|23|24|25|26)\.adaLN_modulation\.0\.weight|layers\.(27|28|29)\.adaLN_modulation\.0\.weight|context_refiner\.(0|1)\.attention\.out\.weight|context_refiner\.(1)\.feed_forward\.w2\.weight|noise_refiner\.(1)\.attention\.qkv\.weight|noise_refiner\.(1)\.attention\.out\.weight|noise_refiner\.(1)\.feed_forward\.w2\.weight|noise_refiner\.(0|1)\.feed_forward\.w3\.weight" \
  --num-iter 6000 --top-p 0.35 --calib-samples 8192 \
  --scale-optimization iterative --scale-refinement-rounds 2 \
  --extract-lora --lora-rank 32 \
  -o "${1%%.safetensors}-nvfp4.safetensors"