Hugging Face's logo

amd
/
DeepSeek-R1-MXFP4-Preview

Safetensors
deepseek_v3
custom_code
8-bit precision
quark
Model card Files
xet
Community
2
linzhao-amd commited on
Commit
c59c670
·
verified ·
1 Parent(s): d39da30

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -37,7 +37,7 @@ python3 quantize_quark.py --model_dir $MODEL_DIR \
37
  --quant_scheme w_mxfp4_a_mxfp4 \
38
  --group_size 32 \
39
  --num_calib_data 128 \
40
- --exclude_layers "*lm_head" \
41
  --multi_gpu \
42
  --model_export hf_format \
43
  --output_dir amd/DeepSeek-R1-MXFP4
 
37
  --quant_scheme w_mxfp4_a_mxfp4 \
38
  --group_size 32 \
39
  --num_calib_data 128 \
40
+ --exclude_layers "lm_head" \
41
  --multi_gpu \
42
  --model_export hf_format \
43
  --output_dir amd/DeepSeek-R1-MXFP4