Skip to content

PTQ-Level Efficiency, QAT-Level Accuracy | 训练后量化的效率,量化感知训练的精度

docs: clarify diffusion export format (#2367)

Wang, Changea838e562cee
diff --git a/auto_round/compressors/diffusion/README.md b/auto_round/compressors/diffusion/README.md
index 91a177bf1..ba6b53beb 100644
--- a/auto_round/compressors/diffusion/README.md
+++ b/auto_round/compressors/diffusion/README.md
@@ -9,7 +9,7 @@ AutoRound uses the new compressor and calibration architecture for diffusion qua
Quantization for diffusion models is limited:
1. Only the transformer module of diffusion models is quantized.
-2. Loading quantized diffusion models is not supported yet, so use `fake` format for quantization.
+2. Use `auto_round` (the default) for a packed, reloadable checkpoint; use `fake` only for research or debugging.
3. Calibration dataset currently supports `coco2014` and user customized `.tsv` files.
### API Usage (CPU/GPU) Recommended
@@ -31,8 +31,7 @@ autoround = AutoRound(
# Save the quantized model
output_dir = "./tmp_autoround"
-# Loading quantized diffusion models is not supported yet, so use fake format.
-autoround.quantize_and_save(output_dir, format="fake", inplace=True)
+autoround.quantize_and_save(output_dir, format="auto_round", inplace=True)
```
- `dataset`: the dataset for quantization training. Currently supports `coco2014` and user customized `.tsv` files.
@@ -56,7 +55,7 @@ A user guide detailing the full list of supported arguments is provided by calli
auto-round \
--model black-forest-labs/FLUX.1-dev \
--scheme MXFP8 \
- --format fake \
+ --format auto_round \
--batch_size 1 \
--dataset coco2014 \
--calib_num_inference_steps 8 \