Skip to content

BASELINE.md

Latest commit

 

History

History
59 lines (53 loc) · 1.49 KB

BASELINE.md

File metadata and controls

59 lines (53 loc) · 1.49 KB

Baselines Evaluation

We provide evaluation scripts for evaluating baselines on math and science benchmarks, such as AIME24, AMC23, Minerva-Math, MATH500 and OlympiadBench. For LiveCodeBench, a dedicated evaluation script is provided.

Base Model

  • For DeepSeek-R1-Distill-Qwen-1.5B on Math & Science Benchmarks:
./scripts/base/eval_base_deepseek_1_5b.sh
  • For DeepSeek-R1-Distill-Qwen-7B on Math & Science Benchmarks:
./scripts/base/eval_base_deepseek_7b.sh
  • For Qwen QwQ-32B on Math & Science Benchmarks:
./scripts/base/eval_base_qwq.sh
  • For LiveCodeBench:
./scripts/base/eval_base_code.sh

s1

  • For DeepSeek-R1-Distill-Qwen-1.5B on Math & Science Benchmarks:
./scripts/s1/eval_s1_deepseek_1_5b.sh
  • For DeepSeek-R1-Distill-Qwen-7B on Math & Science Benchmarks:
./scripts/s1/eval_s1_deepseek_7b.sh
  • For Qwen QwQ-32B on Math & Science Benchmarks:
./scripts/s1/eval_s1_qwq.sh
  • For LiveCodeBench:
./scripts/s1/eval_s1_code.sh

Chain of Draft (CoD)

  • For DeepSeek-R1-Distill-Qwen-1.5B on Math & Science Benchmarks:
./scripts/cod/eval_cod_deepseek_1_5b.sh
  • For DeepSeek-R1-Distill-Qwen-7B on Math & Science Benchmarks:
./scripts/cod/eval_cod_deepseek_7b.sh
  • For Qwen QwQ-32B on Math & Science Benchmarks:
./scripts/cod/eval_cod_qwq.sh
  • For LiveCodeBench:
./scripts/cod/eval_cod_code.sh