5 Commits

Author	SHA1	Message	Date
sora	ad61a2c44e	Fix DeepSWE Pier routing and Terminal-Bench context/apt setup. ... Keep generation max_tokens from being used as terminus-2 context, rewrite Debian/Ubuntu .sources to the Tsinghua mirror, and skip injecting truncation_tokens into DeepSWE extra_params. Co-authored-by: Cursor <cursoragent@cursor.com>	2026-09-08 07:32:07 +00:00
sora	13274243a0	Bump vendored EvalScope and add K3-ready DPV4 configs. ... Keep K3 suite selection and report-schema scoring in bash, merge K3/vision dataset_args into dpv4 yamls, and pin EvalScope at 735d920ee911 with local patches. Co-authored-by: Cursor <cursoragent@cursor.com>	2026-09-02 07:30:48 +00:00
sora	0f7c460f1b	Add SciCode sandbox preload and extra bench configs for dpv4. ... Keep existing yaml entries unchanged; enable SciCode sandbox in run.py and host-side image build without GDPval. Co-authored-by: Cursor <cursoragent@cursor.com>	2026-08-31 07:24:51 +00:00
sora	b1eb4e116a	chore: update evalscope runner, thinking config and webui	2026-07-31 02:36:51 +00:00
sora	b585c1e814	Update thinking config auto-selection, deploy script, README, and SWE-bench utils	2026-07-29 06:37:36 +00:00