Skip to main content
For a supported downstream release, select a versioned consumer profile. Run sediment export rlvr --target <target> --out <directory> to project Sediment Rollouts into reinforcement learning from verifiable rewards (RLVR) rows. The command writes tasks, Rollouts, or both according to the selected target. Sediment supports its audit format, SWE-bench tasks, and NeMo Gym rollouts. The command requires an explicit target because RLVR has no universal interchange schema. For shared preparation, schema inspection, and holdout guidance, see Choose a training export.

​
Choose a target

The exporter projects captured Rollouts and mirror evidence. It doesn’t run a Verifier or fill missing evidence. Every emitted row names version 1 of the rlvr_ci Evidence recipe. A resolved CI pass records reward_source: resolved_ci_pass. A resolved CI failure records reward_source: resolved_ci_fail. A non-verdict omits reward_source and doesn’t create a directional label or fractional Reward.

​
Export one target

Configure SEDIMENT_MIRROR_PATH, then mirror each repository that can contribute a Reference patch. Run one of these commands:
If you reviewed a derived bundle, pass it with --from:
The bundle freezes canonical Rollouts. The sediment and swe-bench targets still read local mirrors for Reference patches, so they require SEDIMENT_MIRROR_PATH even from a bundle. The nemo-gym target projects purely from captured segments, so sediment export rlvr --target nemo-gym --from <bundle> doesn’t read a mirror and runs with SEDIMENT_MIRROR_PATH unset. Run derivations covers the review workflow. A Reference patch ends at the attributed commit named by its exact Verifier result. Later attributed commits without Verifier evidence never extend that patch.

​
Export from Python

As a library caller, pass the target explicitly:

​
Inspect the output

If evaluation splitting is enabled, the writer produces <name>.train.jsonl and <name>.eval.jsonl. Otherwise, it produces one <name>.jsonl. An empty row list doesn’t create or truncate a file. The built-in split is Session-grained. It keeps every segment and task derived from one Session on one side. It doesn’t isolate repositories, exact prompts, or external task identifiers across Sessions. If a benchmark uses task-keyed isolation, audit both files against the versioned task manifest before training or evaluation. Warning: Use a fresh output directory for each export. Stale split, task, Rollout, or manifest files could otherwise create a mixed-generation consumer contract. The first export writes a hidden .sediment-rlvr-target claim in its output directory. If the directory contains an RLVR artifact, another export fails before writing, including an export for the same target. A directory with unclaimed RLVR artifacts also fails closed. An empty export can reuse its claimed directory because it wrote no artifact. A hidden generation lock rejects concurrent exports for the same directory. If a stopped process leaves the lock behind, the directory remains fail-closed. Use a fresh directory for the next export.

​
Understand target contracts

Compare repeated exports within one mode by JSONL bytes and row order. Across direct and bundle exports, compare parsed values and order: bundles sort nested keys while direct exports retain source order. Inspect representation skips such as non_finite_number and unrepresentable_unicode. Each declined task or Segment counts once. File replacement is atomic per file, not across all partitions. If publication fails or stops, retain its diagnostics and export to a fresh directory. Empty partitions don’t clear existing files. Validate rows against the Schema reference before consumer adaptation. A target contract doesn’t establish compatibility with an external trainer release; use a qualified consumer profile. Canonical schema, Evidence recipe, and consumer-profile versions identify separate contracts.

​
Configure verification

Verification configuration tells a consumer how a Verifier can run again. A Verifier result is the exact recorded CIOutcome Fact. Every target keeps these values separate. If a repository has a known verification command, create a TOML file:
Set the configuration path:
If the setting or repository entry is absent, the projection omits verification. The projection retains recorded Verifier results. Sediment never derives a command from workflow YAML, and it never runs the configured command.

​
Read Sediment target rows

The target writes task rows and Rollout rows. Task rows preserve a passing or failing historical patch, recorded Verifier results, CI resolution, Attribution source, observation IDs, split, and Provenance. The Reference patch starts at the parent of the first attributed commit in the selected CI resolution’s repository and ends at its exact verified commit. Later unverified commits and commits in other repositories don’t extend it.

​
Resolve attempts

Outcomes group by (org_id, provider, run_id). The resolver orders them only by provider run_attempt. Null sorts as attempt 0 when mixed with numbered attempts. Capture time, ingest order, URL, and generated IDs never select a verdict. The last ordered passed or failed result supplies the workflow verdict. Non-verdicts remain in verifier_results without creating Reward. Conflicting workflow verdicts produce no aggregate Reward and count ambiguous_workflow_verdicts. Agreeing workflows use minimum reliability. A fail-to-pass retry therefore resolves to passed with suspected_flake: true and default reliability: 0.0. A pass-to-fail retry resolves to failed with the same reliability treatment. NeMo Gym Reward remains exactly 1.0, 0.0, or absent. When consumers choose a downstream sample weight, they must honor reliability metadata. A categorical verdict beside zero reliability is intentional. verifier_results preserves every recorded CIOutcome exactly.

​
Read task prompts

Task exporters retain each canonical TextPart.content verbatim, including JSON-looking examples, whitespace, and Unicode. They join explicit parts with newlines and mark each non-text part as [non-text content omitted]. They don’t decode text as legacy provider content. Turns in rollouts.jsonl keep the full typed message structure.

​
Read Rollout rows

Rollout implementation version 4 continues a Segment only when typed input replays the prior input and captured output. Missing or contradictory echoes, rewritten input, missing output, and unsupported message regrouping retain the Turn in a separate Segment. The Derivation counts these boundaries in fragmented, separately from projection skips. Each Rollout segment becomes one row. The row preserves Session identifiers, segment index, structured messages, completion text, tool calls, decisions, split, Provenance, and exact verifier_results. Session-terminal Verifier results repeat on each segment because Sediment has no Fact that assigns a result to one segment. ci_resolutions carries the recomputed commit-level interpretations. ci_resolution identifies the resolution that supplied reward_source, including its reliability. Sediment task skip reasons form a closed vocabulary: Sediment Rollout rows use conflicting_run_identity, ambiguous_workflow_verdicts, and no_segments.

​
Read SWE-bench target rows

See SWEBenchTaskRow for the complete shape. The row carries repository, base commit, problem statement, patch, and Sediment evidence in metadata. The target populates patch only from a recorded terminal pass. It omits issue identifiers, test patches, versions, FAIL_TO_PASS, and PASS_TO_PASS because Sediment Facts don’t contain them. SWE-bench adds failed_reference_patch to the Sediment task skip vocabulary. It counts that reason when the recorded terminal Verifier result is a failure. This count remains distinct from missing_verifier_evidence.

​
Read NeMo Gym target rows

Each captured Segment maps to a NemoGymRolloutRow with responses_create_params, response, optional reward, and evidence in metadata. A resolved pass maps to numeric Reward 1.0. A resolved failure maps to 0.0. Without a resolved pass or failure, the row omits both reward and reward_source. Metadata retains every exact terminal Verifier result, including non-label outcomes. It keeps the selected ci_resolution separate from optional Verification configuration. The target doesn’t infer a NeMo Gym resource server or executable environment. Its closed skip vocabulary contains conflicting_run_identity, ambiguous_workflow_verdicts, and no_segments. This mapping names NeMo Gym’s rollout boundary, but it doesn’t claim direct compatibility. Sediment messages and captured responses remain canonical evidence, not fabricated OpenAI response objects.

​
Interpret the experimental environment manifest

Only the Sediment target can write environment.yaml. The document contains experimental: true because its OpenEnv shape follows a proposal rather than a compatibility-tested contract. SWE-bench and NeMo Gym exports never write this file. The manifest can include an informational CI workflow reference when every task agrees on one workflow. It includes frameworks.openenv only from SEDIMENT_OPENENV_IMAGE or SEDIMENT_OPENENV_PACKAGE. It includes frameworks.nemo_gym only from an operator-supplied SEDIMENT_NEMO_GYM_RESOURCES_SERVER, with optional SEDIMENT_NEMO_GYM_CONFIG. It never derives either runtime from workflow YAML. The OpenEnv field mapping documents intent, not compatibility. An OpenEnv dataset needs a runtime that implements the expected protocol. The Verifier-result-to-Reward mapping doesn’t provide a Verifiers taskset or Verifier implementation.

​
Understand capture limits

Rollout rows require gateway-routed inference calls. OpenTelemetry-only capture can produce decisions and CI Facts without a trajectory. Copilot model calls can’t route through the Sediment gateway, so Copilot contributes decisions but not Rollouts. For field-level types, see the generated Schema reference.