| { | |
| "@context": { | |
| "croissant": "http://mlcommons.org/croissant/", | |
| "schema": "https://schema.org/", | |
| "dct": "http://purl.org/dc/terms/" | |
| }, | |
| "@graph": [ | |
| { | |
| "@type": "croissant:TaskProblem", | |
| "@id": "https://huggingface.co/datasets/gaia-hazlab/repere-codameter#task-param_recommendation", | |
| "dct:conformsTo": "http://mlcommons.org/croissant/1.1", | |
| "schema:name": "repere-codameter / param_recommendation", | |
| "schema:description": "Repère code_generation task. Each row of record set 'param_recommendation' is one problem instance: give the prompt to the system under test, then score its answer with the row's scorer_spec.", | |
| "croissant:input": { | |
| "@type": "schema:Dataset", | |
| "@id": "https://huggingface.co/datasets/gaia-hazlab/repere-codameter", | |
| "dct:conformsTo": "http://mlcommons.org/croissant/1.1", | |
| "schema:name": "gaia-hazlab/repere-codameter" | |
| }, | |
| "croissant:output": { | |
| "@type": "croissant:OutputSpec", | |
| "@id": "https://huggingface.co/datasets/gaia-hazlab/repere-codameter#task-param_recommendation/output", | |
| "schema:description": "A JSON processing configuration (estimator, frequency band, coda window, stack length, reference, quality gate)." | |
| }, | |
| "croissant:implementation": { | |
| "@type": "croissant:ImplementationSpec", | |
| "@id": "https://huggingface.co/datasets/gaia-hazlab/repere-codameter#task-param_recommendation/implementation", | |
| "schema:description": "The model or agent under test. Repère runs it with and without a domain skill file; tools, budget and model are recorded per run in the leaderboard row." | |
| }, | |
| "croissant:execution": { | |
| "@type": "croissant:ExecutionSpec", | |
| "@id": "https://huggingface.co/datasets/gaia-hazlab/repere-codameter#task-param_recommendation/execution", | |
| "schema:description": "Repeated runs per configuration (default 5, temperature 0.7); report mean score with a bootstrap interval." | |
| }, | |
| "croissant:evaluation": { | |
| "@type": "croissant:EvaluationSpec", | |
| "@id": "https://huggingface.co/datasets/gaia-hazlab/repere-codameter#task-param_recommendation/evaluation", | |
| "schema:name": "dvv_recovery", | |
| "schema:description": "Scorer dvv_recovery runs the configuration through codameter on the case's synthetic data and grades the RMS error of the recovered dv/v against the imposed one, from 1 at or below rms_ceiling to 0 at rms_bad. Missing core keys or an invalid band, window or estimator score 0.", | |
| "schema:url": "https://github.com/mdenolle/repere", | |
| "schema:version": "0.5.2" | |
| } | |
| }, | |
| { | |
| "@type": "croissant:TaskProblem", | |
| "@id": "https://huggingface.co/datasets/gaia-hazlab/repere-codameter#task-dvv_series", | |
| "dct:conformsTo": "http://mlcommons.org/croissant/1.1", | |
| "schema:name": "repere-codameter / dvv_series", | |
| "schema:description": "Repère code_generation task. Each row of record set 'dvv_series' is one problem instance: give the prompt to the system under test, then score its answer with the row's scorer_spec.", | |
| "croissant:input": { | |
| "@type": "schema:Dataset", | |
| "@id": "https://huggingface.co/datasets/gaia-hazlab/repere-codameter", | |
| "dct:conformsTo": "http://mlcommons.org/croissant/1.1", | |
| "schema:name": "gaia-hazlab/repere-codameter" | |
| }, | |
| "croissant:output": { | |
| "@type": "croissant:OutputSpec", | |
| "@id": "https://huggingface.co/datasets/gaia-hazlab/repere-codameter#task-dvv_series/output", | |
| "schema:description": "The recovered dv/v(t) series, one value per day." | |
| }, | |
| "croissant:implementation": { | |
| "@type": "croissant:ImplementationSpec", | |
| "@id": "https://huggingface.co/datasets/gaia-hazlab/repere-codameter#task-dvv_series/implementation", | |
| "schema:description": "The model or agent under test. Repère runs it with and without a domain skill file; tools, budget and model are recorded per run in the leaderboard row." | |
| }, | |
| "croissant:execution": { | |
| "@type": "croissant:ExecutionSpec", | |
| "@id": "https://huggingface.co/datasets/gaia-hazlab/repere-codameter#task-dvv_series/execution", | |
| "schema:description": "Repeated runs per configuration (default 5, temperature 0.7); report mean score with a bootstrap interval." | |
| }, | |
| "croissant:evaluation": { | |
| "@type": "croissant:EvaluationSpec", | |
| "@id": "https://huggingface.co/datasets/gaia-hazlab/repere-codameter#task-dvv_series/evaluation", | |
| "schema:name": "dvv_series_regression", | |
| "schema:description": "Scorer dvv_series_regression grades the RMS error of the returned series against the imposed dv/v on the same scale (rms_ceiling to rms_bad); rms_bad is set so a no-change series scores about 0.", | |
| "schema:url": "https://github.com/mdenolle/repere", | |
| "schema:version": "0.5.2" | |
| } | |
| } | |
| ] | |
| } | |