Update README for experiment FinEval_16k_fulleval_3args_Random-RL-longmult_4dig

README.md

@@ -216,3 +216,84 @@ configs:
   - split: train
     path: metadata/train-*
 ---
+# Experiment Tracker: FinEval_16k_fulleval_3args_Random-RL-longmult_4dig
+
+**Experiment Description:** Evaluation experiment for task longmult_4dig from FinEval_16k_fulleval_3args_Random-RL
+
+**Start Time:** 2025-11-09T11:42:11.211595
+
+**Tracker Dataset:** [TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_3args_Random-RL-longmult_4dig__v1](https://huggingface.co/datasets/TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_3args_Random-RL-longmult_4dig__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:
+
+```python
+from datasets import load_dataset
+
+# Load experiment metadata
+metadata = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_3args_Random-RL-longmult_4dig__v1', 'experiment_metadata')
+
+# Load complete training datasets
+sft_data = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_3args_Random-RL-longmult_4dig__v1', 'training_data__sft')
+sft_metadata = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_3args_Random-RL-longmult_4dig__v1', 'training_data__sft_metadata')
+
+# Load complete configurations
+sft_hyperparams = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_3args_Random-RL-longmult_4dig__v1', 'hyperparameters__sft')
+rl_hyperparams = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_3args_Random-RL-longmult_4dig__v1', 'hyperparameters__rl')
+
+# Load stage-specific logs
+sft_logs = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_3args_Random-RL-longmult_4dig__v1', 'logs__sft')
+rl_logs = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_3args_Random-RL-longmult_4dig__v1', 'logs__rl')
+
+# Load evaluation results with annotations
+sft_eval_results = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_3args_Random-RL-longmult_4dig__v1', 'evals_eval_sft')
+rl_eval_results = load_dataset('TAUR-dev/D-ExpTracker__FinEval_16k_fulleval_3args_Random-RL-longmult_4dig__v1', 'evals_eval_rl')
+```
+
+## Models
+
+
+
+## Registry
+
+All models from this experiment are automatically registered in the [SkillFactory Model Registry](https://huggingface.co/datasets/TAUR-dev/SkillFactory-Registration) 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_3args_Random-RL-longmult_4dig - {stage_name} - {SFT/RL}`
+
+---
+*Generated by SkillFactory Experiment Management System*
+*All artifacts uploaded immediately as stages complete with perfect data provenance*