SeaWolf-AI commited on
Commit
c380de7
·
verified ·
1 Parent(s): ac4042d

Card: remove proprietary method details

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -98,12 +98,12 @@ We started at 244 seconds and ended at 46. Every step is measured:
98
 
99
  | File | Size | What |
100
  |---|---|---|
101
- | `z_image_turbo-Q4_0-pocket.gguf` | 3.51 GB | Diffusion model, all tensors Q4_0 |
102
  | *(bring your own)* Qwen3-4B-Instruct-2507-Q4_K_M | 2.58 GB | Text encoder — [download](https://huggingface.co/lmstudio-community/Qwen3-4B-Instruct-2507-GGUF) |
103
  | *(bring your own)* `ae.safetensors` | 0.16 GB | VAE — [download](https://huggingface.co/Tongyi-MAI/Z-Image-Turbo) |
104
  | **Total** | **≈ 6.25 GB** | |
105
 
106
- The diffusion model here is re-quantized: the upstream GGUF keeps 10 tensors in Q8_0, and moving them to Q4_0 saves 170 MB with no visible quality change.
107
 
108
  ## Run it
109
 
 
98
 
99
  | File | Size | What |
100
  |---|---|---|
101
+ | `z_image_turbo-Q4_0-pocket.gguf` | 3.51 GB | Diffusion model, 4-bit (VIDRAFT CPU build) |
102
  | *(bring your own)* Qwen3-4B-Instruct-2507-Q4_K_M | 2.58 GB | Text encoder — [download](https://huggingface.co/lmstudio-community/Qwen3-4B-Instruct-2507-GGUF) |
103
  | *(bring your own)* `ae.safetensors` | 0.16 GB | VAE — [download](https://huggingface.co/Tongyi-MAI/Z-Image-Turbo) |
104
  | **Total** | **≈ 6.25 GB** | |
105
 
106
+ The diffusion model here is VIDRAFT's own CPU build of Z-Image-Turbo: 3.51 GB, with no visible quality change.
107
 
108
  ## Run it
109