amd
/

DeepSeek-R1-0528-MXFP4

8-bit precision

Model card Files Files and versions

linzhao-amd commited on Oct 21, 2025

Commit

1e68a0f

·

verified ·

1 Parent(s): e182145

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -100,7 +100,7 @@ The model was evaluated on AIME24, GPQA Diamond, and MATH-500 benchmarks using t
 ### Reproduction
-The results were obtained using [vLLM](https://docs.vllm.ai/en/latest/) and [lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness) framworks. We took 10 rounds with different random seeds for reliable performance estimation.
 ```
 MODEL_ARGS="model_name=amd/DeepSeek-R1-0528-MXFP4-Preview,dtype=bfloat16,tensor_parallel_size=8,max_model_length=71536,max_num_batched_tokens=32768,gpu_memory_utilization=0.85,generation_parameters={max_new_tokens:65536,temperature:0.6,top_p:0.95,seed:$SEED}"

 ### Reproduction
+The results can be reproduced using vLLM emulation docker `rocmshared/pytorch:vllm-gfx950-mxfp4-mxfp6-v3` and [lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness) evaluation framwork. We took 10 rounds with different random seeds for reliable performance estimation.
 ```
 MODEL_ARGS="model_name=amd/DeepSeek-R1-0528-MXFP4-Preview,dtype=bfloat16,tensor_parallel_size=8,max_model_length=71536,max_num_batched_tokens=32768,gpu_memory_utilization=0.85,generation_parameters={max_new_tokens:65536,temperature:0.6,top_p:0.95,seed:$SEED}"