{"data":{"kind":"file","path":"README.md","version_id":"qghm85if3hf1ayimqj8w1etr","entry":{"name":"README.md","path":"README.md","is_directory":false,"size":2962,"modified_at":"2026-09-13T04:06:06.956000","content_hash":"7c278928d377be5a29c22f4bf81df082ada08ce10fdbf0f4350aa558289d67fe"},"entries":[],"content":"# scaleswe\n\nScale-SWE issue-resolving tasks, each with a per-task image, a base commit, and FAIL_TO_PASS/PASS_TO_PASS pytest ids. Scored 1.0 iff every expected test id passes after the held-out tests are restored to base and the failing test is re-applied.\n\n## Taskset\n\n- **Source:** [`PrimeIntellect/Scale-SWE-Verified`](https://huggingface.co/datasets/PrimeIntellect/Scale-SWE-Verified)\n- **Size:** 17,202 tasks (`train` split)\n\n## Changelog\n\n- 2026-09-04: Add a solver system prompt (adapted from FrontierCode 1.1's fair-internet-use prompt) that allows documentation lookups but forbids retrieving the task's upstream fix online or from non-current-branch git history; complements #795, which restored network access.\n- 2026-09-03: Restore default solver network access by removing the `network_allow=[]` override introduced in #780; training rollouts need outbound network.\n- 2026-08-31: Yield task records on demand so bounded evaluations construct only the requested prefix.\n- 2026-07-17: `patch_capture` now imports from `verifiers.v1` (`capture_patch`/`resolve_head` upstreamed in verifiers#2054); the copied module is removed. Requires `verifiers>=0.2.2.dev5`.\n- 2026-07-17: The test-drop sweep in `RESTORE` now lists untracked files too (`git ls-files --cached --others --exclude-standard`), so agent-added tests are dropped whether staged, unstaged, or never tracked.\n- 2026-07-16: `finalize` persists the agent's final patch to `trace.info[\"patch\"]` (truncated at 2 MB with `patch_truncated`; failures record `patch_error` instead of failing the rollout) via the copied `patch_capture` helper.\n- 2026-07-16: Removed `use_prime_registry` — the Prime platform auto-builds and caches sandbox images from the dataset's Docker Hub `image_url` refs on first use, so no registry mapping is needed. Tasks whose image is missing from the Docker Hub mirror are dropped by the default `filter_unavailable_images` (as before).\n- 2026-07-10: Ported to the task-centric verifiers API: rewards and lifecycle hooks live on the `Task` (a `TaskData` row + behavior split), and task-facing config knobs (judges, tool/user placement, scoring parameters) moved from `--env.taskset.*` to `--env.taskset.task.*`. Requires `verifiers>=0.2.0` and Python `>=3.11`.\n- 2026-07-08: `use_prime_registry` now resolves images from the public Prime platform registry (`prime/primeintellect/scaleswe:<tag>`, `aweaiteam/` namespace stripped) instead of the private GCP Artifact Registry mirror; any Prime user can pull them.\n- 2026-07-06: Default dataset renamed to `PrimeIntellect/Scale-SWE-Verified` (formerly `Scale-SWE`); the old name redirects.\n- 2026-07-01: Added `filter_fn`, applied directly with `datasets.Dataset.filter` to raw HF rows before task construction.\n- 2026-07-01: Default dataset is now the PrimeIntellect reupload `PrimeIntellect/Scale-SWE`. The upstream `AweAI-Team/Scale-SWE` dataset remains available as a typed override via `ScaleSWEConfig.dataset_name`.\n","encoding":"utf-8","truncated":false,"total_bytes":2962},"status":null}