Ernie Image Sft Mlx Q4

Usage

These weights can be used with ernie-image-mlx:

pip install ernie-image-mlx
ernie-image-mlx generate \
    -p "一只黑白相间的中华田园犬" \
    --repo-id dgrauet/ernie-image-sft-mlx-q4 \
    -o dog.png

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

• mlx-forge (conversion): https://github.com/dgrauet/mlx-forge
• mlx-arsenal (reusable MLX ops): https://github.com/dgrauet/mlx-arsenal
• mlx-porting (Claude Code skill): https://github.com/dgrauet/claude-skill-mlx-porting
• bf16 variant: https://huggingface.co/dgrauet/ernie-image-sft-mlx
• q8 variant: https://huggingface.co/dgrauet/ernie-image-sft-mlx-q8

Files

• model_index.json (547.00 B)
• pe_tokenizer_tokenizer_config.json (20.63 KB)
• quantize_config.json (107.00 B)
• scheduler_scheduler_config.json (482.00 B)
• special_tokens_map.json (414.00 B)
• split_model.json (1.07 KB)
• text_encoder.safetensors (1.80 GB)
• text_encoder_config.json (1.64 KB)
• tokenizer.json (8.83 MB)
• tokenizer_tokenizer_config.json (379.00 B)
• transformer.safetensors (4.45 GB)
• transformer_config.json (383.00 B)
• vae.safetensors (160.33 MB)
• vae_config.json (831.00 B)

--- library_name: mlx license: apache-2.0 base_model: baidu/ERNIE-Image tags: - mlx - mlx-forge - apple-silicon - safetensors - quantized - int4 --- # dgrauet/ernie-image-sft-mlx-q4 Int4 quantization (group_size 64, attention and FFN Linear weights only, leaving the small projections and the VAE in bf16) of [dgrauet/ernie-image-sft-mlx](https://huggingface.co/dgrauet/ernie-image-sft-mlx), the MLX conversion of [baidu/ERNIE-Image](https://huggingface.co/baidu/ERNIE-Image). Quantized with [mlx-forge](https://github.com/dgrauet/mlx-forge) (`mlx-forge convert ernie-image --variant sft --quantize --bits 4`). ## Usage These weights can be used with [ernie-image-mlx](https://github.com/dgrauet/ernie-image-mlx): ```bash pip install ernie-image-mlx ernie-image-mlx generate \ -p "一只黑白相间的中华田园犬" \ --repo-id dgrauet/ernie-image-sft-mlx-q4 \ -o dog.png ``` 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 - **mlx-forge (conversion):** https://github.com/dgrauet/mlx-forge - **mlx-arsenal (reusable MLX ops):** https://github.com/dgrauet/mlx-arsenal - **mlx-porting (Claude Code skill):** https://github.com/dgrauet/claude-skill-mlx-porting - **bf16 variant:** https://huggingface.co/dgrauet/ernie-image-sft-mlx - **q8 variant:** https://huggingface.co/dgrauet/ernie-image-sft-mlx-q8 ## Files - `model_index.json` (547.00 B) - `pe_tokenizer_tokenizer_config.json` (20.63 KB) - `quantize_config.json` (107.00 B) - `scheduler_scheduler_config.json` (482.00 B) - `special_tokens_map.json` (414.00 B) - `split_model.json` (1.07 KB) - `text_encoder.safetensors` (1.80 GB) - `text_encoder_config.json` (1.64 KB) - `tokenizer.json` (8.83 MB) - `tokenizer_tokenizer_config.json` (379.00 B) - `transformer.safetensors` (4.45 GB) - `transformer_config.json` (383.00 B) - `vae.safetensors` (160.33 MB) - `vae_config.json` (831.00 B)