DeepSeek-R1-W4AFP8 / README.md
Barrrrry's picture
Update README.md
157b649 verified
---
license: mit
base_model:
- deepseek-ai/DeepSeek-R1
base_model_relation: quantized
---
# DeepSeek-R1-W4AFP8
This model is a mixed-precision quantized DeepSeek-R1, with dense layer using `FP8_BLOCK_SCALING`, MoE layers uses INT4 weights and FP8 activation.