Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
## Llama-3-8B-Instruct-FP8-v1
|
| 2 |
+
|
| 3 |
+
* Weights and activations are per-tensor quantized to float8_e4m3.
|
| 4 |
+
* Quantization with AutoFP8.
|
| 5 |
+
* Calibration dataset: Ultrachat (mgoin/ultrachat_2k)
|
| 6 |
+
* Samples: 512
|
| 7 |
+
* Sequence length: 8192
|
| 8 |
+
|
| 9 |
+
## Evaluation
|
| 10 |
+
|
| 11 |
+
TBA
|