EvalHarness

__init__.py

EvalHarness data layer: 28 dataset plugins, lazy materialize cache (raw/ + samples.jsonl + meta.json), ModelScope native loader, CLI list/fetch/unload/stats/show

2026-08-24 03:35:11 +00:00

_bbh_cot_prompts.py

Paper-faithful eval: P1 logprob scoring (completions_echo + chat_first_token fallback, sglang top_logprobs parser), few_shot_num dataset defaults (mmlu5/cmmlu5/mmlu_pro5/gsm8k4/math4/drop3), official BBH 3-shot CoT vendored, strict_letter default prompt (evalscope parity), official simpleqa grader prompt

2026-08-25 02:21:59 +00:00

aime24.py

EvalHarness data layer: 28 dataset plugins, lazy materialize cache (raw/ + samples.jsonl + meta.json), ModelScope native loader, CLI list/fetch/unload/stats/show

2026-08-24 03:35:11 +00:00

aime25.py

EvalHarness data layer: 28 dataset plugins, lazy materialize cache (raw/ + samples.jsonl + meta.json), ModelScope native loader, CLI list/fetch/unload/stats/show

2026-08-24 03:35:11 +00:00

aime26.py

Add evaluation layer + visualization: extract/score/aggregate plugins, 28 recipes, official-aligned scorers (PRM800K math, DROP Hungarian EM/F1, SimpleQA A/B/C judge), report artifacts, console renderers, CLI eval/viz, regression tests

2026-08-24 06:09:38 +00:00

arc.py

EvalHarness data layer: 28 dataset plugins, lazy materialize cache (raw/ + samples.jsonl + meta.json), ModelScope native loader, CLI list/fetch/unload/stats/show

2026-08-24 03:35:11 +00:00

bbh.py

Paper-faithful eval: P1 logprob scoring (completions_echo + chat_first_token fallback, sglang top_logprobs parser), few_shot_num dataset defaults (mmlu5/cmmlu5/mmlu_pro5/gsm8k4/math4/drop3), official BBH 3-shot CoT vendored, strict_letter default prompt (evalscope parity), official simpleqa grader prompt

2026-08-25 02:21:59 +00:00

bfcl_v3.py

Add evaluation layer + visualization: extract/score/aggregate plugins, 28 recipes, official-aligned scorers (PRM800K math, DROP Hungarian EM/F1, SimpleQA A/B/C judge), report artifacts, console renderers, CLI eval/viz, regression tests

2026-08-24 06:09:38 +00:00

bigcodebench.py

EvalHarness data layer: 28 dataset plugins, lazy materialize cache (raw/ + samples.jsonl + meta.json), ModelScope native loader, CLI list/fetch/unload/stats/show

2026-08-24 03:35:11 +00:00

cmmlu.py

Paper-faithful eval: P1 logprob scoring (completions_echo + chat_first_token fallback, sglang top_logprobs parser), few_shot_num dataset defaults (mmlu5/cmmlu5/mmlu_pro5/gsm8k4/math4/drop3), official BBH 3-shot CoT vendored, strict_letter default prompt (evalscope parity), official simpleqa grader prompt

2026-08-25 02:21:59 +00:00

competition_math.py

Paper-faithful eval: P1 logprob scoring (completions_echo + chat_first_token fallback, sglang top_logprobs parser), few_shot_num dataset defaults (mmlu5/cmmlu5/mmlu_pro5/gsm8k4/math4/drop3), official BBH 3-shot CoT vendored, strict_letter default prompt (evalscope parity), official simpleqa grader prompt

2026-08-25 02:21:59 +00:00

drop.py

Paper-faithful eval: P1 logprob scoring (completions_echo + chat_first_token fallback, sglang top_logprobs parser), few_shot_num dataset defaults (mmlu5/cmmlu5/mmlu_pro5/gsm8k4/math4/drop3), official BBH 3-shot CoT vendored, strict_letter default prompt (evalscope parity), official simpleqa grader prompt

2026-08-25 02:21:59 +00:00

general_fc.py

EvalHarness data layer: 28 dataset plugins, lazy materialize cache (raw/ + samples.jsonl + meta.json), ModelScope native loader, CLI list/fetch/unload/stats/show

2026-08-24 03:35:11 +00:00

gpqa_diamond.py

Add evaluation layer + visualization: extract/score/aggregate plugins, 28 recipes, official-aligned scorers (PRM800K math, DROP Hungarian EM/F1, SimpleQA A/B/C judge), report artifacts, console renderers, CLI eval/viz, regression tests

2026-08-24 06:09:38 +00:00

gsm8k.py

Paper-faithful eval: P1 logprob scoring (completions_echo + chat_first_token fallback, sglang top_logprobs parser), few_shot_num dataset defaults (mmlu5/cmmlu5/mmlu_pro5/gsm8k4/math4/drop3), official BBH 3-shot CoT vendored, strict_letter default prompt (evalscope parity), official simpleqa grader prompt

2026-08-25 02:21:59 +00:00

hellaswag.py

EvalHarness data layer: 28 dataset plugins, lazy materialize cache (raw/ + samples.jsonl + meta.json), ModelScope native loader, CLI list/fetch/unload/stats/show

2026-08-24 03:35:11 +00:00

hle.py

EvalHarness data layer: 28 dataset plugins, lazy materialize cache (raw/ + samples.jsonl + meta.json), ModelScope native loader, CLI list/fetch/unload/stats/show

2026-08-24 03:35:11 +00:00

hmmt26.py

Add evaluation layer + visualization: extract/score/aggregate plugins, 28 recipes, official-aligned scorers (PRM800K math, DROP Hungarian EM/F1, SimpleQA A/B/C judge), report artifacts, console renderers, CLI eval/viz, regression tests

2026-08-24 06:09:38 +00:00

humaneval.py

EvalHarness data layer: 28 dataset plugins, lazy materialize cache (raw/ + samples.jsonl + meta.json), ModelScope native loader, CLI list/fetch/unload/stats/show

2026-08-24 03:35:11 +00:00

imo_answerbench.py

Add evaluation layer + visualization: extract/score/aggregate plugins, 28 recipes, official-aligned scorers (PRM800K math, DROP Hungarian EM/F1, SimpleQA A/B/C judge), report artifacts, console renderers, CLI eval/viz, regression tests

2026-08-24 06:09:38 +00:00

live_code_bench.py

Add evaluation layer + visualization: extract/score/aggregate plugins, 28 recipes, official-aligned scorers (PRM800K math, DROP Hungarian EM/F1, SimpleQA A/B/C judge), report artifacts, console renderers, CLI eval/viz, regression tests

2026-08-24 06:09:38 +00:00

longbench_v2.py

EvalHarness data layer: 28 dataset plugins, lazy materialize cache (raw/ + samples.jsonl + meta.json), ModelScope native loader, CLI list/fetch/unload/stats/show

2026-08-24 03:35:11 +00:00

mmlu_pro.py

Paper-faithful eval: P1 logprob scoring (completions_echo + chat_first_token fallback, sglang top_logprobs parser), few_shot_num dataset defaults (mmlu5/cmmlu5/mmlu_pro5/gsm8k4/math4/drop3), official BBH 3-shot CoT vendored, strict_letter default prompt (evalscope parity), official simpleqa grader prompt

2026-08-25 02:21:59 +00:00

mmlu.py

Paper-faithful eval: P1 logprob scoring (completions_echo + chat_first_token fallback, sglang top_logprobs parser), few_shot_num dataset defaults (mmlu5/cmmlu5/mmlu_pro5/gsm8k4/math4/drop3), official BBH 3-shot CoT vendored, strict_letter default prompt (evalscope parity), official simpleqa grader prompt

2026-08-25 02:21:59 +00:00

openai_mrcr.py

Add evaluation layer + visualization: extract/score/aggregate plugins, 28 recipes, official-aligned scorers (PRM800K math, DROP Hungarian EM/F1, SimpleQA A/B/C judge), report artifacts, console renderers, CLI eval/viz, regression tests

2026-08-24 06:09:38 +00:00

simple_qa.py

EvalHarness data layer: 28 dataset plugins, lazy materialize cache (raw/ + samples.jsonl + meta.json), ModelScope native loader, CLI list/fetch/unload/stats/show

2026-08-24 03:35:11 +00:00

swe_bench_verified.py

swe_bench_verified end-to-end verified (oracle patch -> official sweb image -> resolved); image naming mapped to local swebench/ set; ladder2 complete 28/28

2026-08-26 06:40:55 +00:00

tau2_bench.py

tau2 via OFFICIAL engine: self-running env plugin (run_task hook + needs_adapter generic dispatch, no name hardcoding), deep generate() patch, official ToolCall shape, data plugin keeps verbatim Task json, official reward scoring; checkpoint plugin (per-sample resume) + summary csv with time/categories

2026-08-25 11:06:08 +00:00

trivia_qa.py

Multi-endpoint model pool (openai-pool/{lo..hi} round-robin with failover), nothink/textools spec flags (chat_template_kwargs enable_thinking=false; text-protocol tool calls for backends without --enable-auto-tool-choice), global cross-bench queue in ladder runs

2026-08-26 06:26:52 +00:00

winogrande.py

EvalHarness data layer: 28 dataset plugins, lazy materialize cache (raw/ + samples.jsonl + meta.json), ModelScope native loader, CLI list/fetch/unload/stats/show

2026-08-24 03:35:11 +00:00