Update README.md
Browse files
README.md
CHANGED
|
@@ -24,6 +24,16 @@ pipeline_tag: text-generation
|
|
| 24 |
--group_size 128 \
|
| 25 |
--quant_algo awq \
|
| 26 |
--exclude_layers
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
- ## Perplexity Numbers
|
| 28 |
- HF Model - 8.56457710266113
|
| 29 |
- Quark-Quantized Model - 9.78733825683593
|
|
|
|
| 24 |
--group_size 128 \
|
| 25 |
--quant_algo awq \
|
| 26 |
--exclude_layers
|
| 27 |
+
|
| 28 |
+
- ## For DML Export
|
| 29 |
+
|
| 30 |
+
- Packages required transformers-4.50.0, OGA-0.7.1, ORT-1.21.1
|
| 31 |
+
- python -m onnxruntime_genai.models.builder \
|
| 32 |
+
-i "path/to/quark-quantized/model/" \
|
| 33 |
+
-o "path/to/output/dir" \
|
| 34 |
+
-p int4 \
|
| 35 |
+
-e dml
|
| 36 |
+
|
| 37 |
- ## Perplexity Numbers
|
| 38 |
- HF Model - 8.56457710266113
|
| 39 |
- Quark-Quantized Model - 9.78733825683593
|