200f867 9cc7235
1
2
3
4
5
6
7
8
9
10
11
## Mixtral-8x7B-Instruct-v0.1-FP8-v2 * Weights and activations are per-tensor quantized to float8_e4m3. * Quantization with AutoFP8. * Calibration dataset: Ultrachat-200k * Samples: 4096 * Sequence length: 8192 ## Evaluation TBA
