timestamp string | end_timestamp string | stage_name string | stage_number int64 | level string | message string | stdout_content string | stderr_content string | experiment_name string | elapsed_time_seconds float64 | stage_complete bool |
|---|---|---|---|---|---|---|---|---|---|---|
2025-11-13T03:40:40.886114 | 2025-11-13T04:10:38.672910 | evaluation_eval_rl | 1 | INFO | Complete log capture for stage: evaluation_eval_rl | "[INFO] Starting stage: Evaluation - eval_rl\n[INFO] Starting evaluation pipeline for eval_rl\n[INFO(...TRUNCATED) | "\u001b[2;36m[11/13/25 03:40:42]\u001b[0m\u001b[2;36m \u001b[0m\u001b[34mINFO \u001b[0m Getting r(...TRUNCATED) | FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg | 1,797.786796 | true |
2025-11-13T04:11:45.063320 | 2025-11-13T04:11:57.373073 | evaluation_eval_rl | 1 | INFO | Complete log capture for stage: evaluation_eval_rl | "[INFO] Starting stage: Evaluation - eval_rl\n[INFO] Starting evaluation pipeline for eval_rl\n[INFO(...TRUNCATED) | "\u001b[2;36m[11/13/25 04:11:55]\u001b[0m\u001b[2;36m \u001b[0m\u001b[34mINFO \u001b[0m Getting r(...TRUNCATED) | FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg | 12.309753 | true |
2025-11-13T04:13:15.363240 | 2025-11-13T04:13:28.947044 | evaluation_eval_rl | 1 | INFO | Complete log capture for stage: evaluation_eval_rl | "[INFO] Starting stage: Evaluation - eval_rl\n[INFO] Starting evaluation pipeline for eval_rl\n[INFO(...TRUNCATED) | "\u001b[2;36m[11/13/25 04:13:26]\u001b[0m\u001b[2;36m \u001b[0m\u001b[34mINFO \u001b[0m Getting r(...TRUNCATED) | FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg | 13.583804 | true |
Experiment Tracker: FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg
Experiment Description: Evaluation experiment for task countdown_4arg from FinEval_16k_fulleval_AT_OursFixed-RL
Start Time: 2025-11-14T16:32:51.894450
Tracker Dataset: TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg__v1
Stages Completed
Total stages: 1
Models Created
Dataset Configurations
This tracker dataset contains the following configurations with immediate upload as stages complete:
Training Data (Complete Datasets)
Hyperparameters (Complete Configurations)
Logs (Stage-Specific)
Evaluation Results (Complete with Annotations)
Metadata
- experiment_metadata: Timeline and stage information
Usage
Load specific configurations with:
from datasets import load_dataset
# Load experiment metadata
metadata = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg__v1', 'experiment_metadata')
# Load complete training datasets
sft_data = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg__v1', 'training_data__sft')
sft_metadata = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg__v1', 'training_data__sft_metadata')
# Load complete configurations
sft_hyperparams = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg__v1', 'hyperparameters__sft')
rl_hyperparams = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg__v1', 'hyperparameters__rl')
# Load stage-specific logs
sft_logs = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg__v1', 'logs__sft')
rl_logs = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg__v1', 'logs__rl')
# Load evaluation results with annotations
sft_eval_results = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg__v1', 'evals_eval_sft')
rl_eval_results = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg__v1', 'evals_eval_rl')
Models
Registry
All models from this experiment are automatically registered in the SkillFactory Model Registry with:
- Complete training configuration (hyperparameters, datasets, methods)
- Experiment lineage (links back to this tracker dataset)
- Stage-specific metadata (SFT vs RL training details)
- Structured input data references (training datasets and configurations)
Registry entries follow the naming pattern: Model - FinEval_16k_fulleval_AT_OursFixed-RL-countdown_4arg - {stage_name} - {SFT/RL}
Generated by SkillFactory Experiment Management System All artifacts uploaded immediately as stages complete with perfect data provenance
- Downloads last month
- 4