Upload README.md with huggingface_hub

README.md

@@ -1,50 +1,85 @@
 ---
-dataset_info:
-  features:
-  - name: question
-    dtype: string
-  - name: metadata
-    dtype: string
-  - name: task_source
-    dtype: string
-  - name: formatted_prompt
-    list:
-    - name: content
-      dtype: string
-    - name: role
-      dtype: string
-  - name: responses_by_sample
-    list:
-      list: string
-  - name: tokens_by_sample
-    list: int64
-  - name: num_rounds_by_sample
-    list: int64
-  - name: continuation_prompts_by_sample
-    list:
-      list: 'null'
-  - name: continuation_text_by_sample
-    list:
-      list: 'null'
-  - name: strategy_type
-    dtype: string
-  - name: termination_reason_by_sample
-    list: string
-  - name: correct_round_by_sample
-    list: int64
-  - name: smart_termination_enabled
-    dtype: bool
-  - name: retry_count_by_sample
-    list: int64
-  splits:
-  - name: train
-    num_bytes: 9238
-    num_examples: 5
-  download_size: 15196
-  dataset_size: 9238
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
+license: mit
+tags:
+- degen_test_1_countdown
+- budget-2k
+- dfs_baseline
+- dfs_baseline
 ---
+
+# test-dfs-baseline
+
+Countdown reasoning with dfs_baseline strategy to 2000 tokens
+
+## Dataset Info
+
+- **Rows**: 5
+- **Columns**: 14
+
+## Columns
+
+| Column | Type | Description |
+|--------|------|-------------|
+| question | Value('string') | The countdown problem statement |
+| metadata | Value('string') | JSON with target number, available numbers, and solution |
+| task_source | Value('string') | Task type identifier (countdown) |
+| formatted_prompt | List({'content': Value('string'), 'role': Value('string')}) | Initial prompt messages sent to the model |
+| responses_by_sample | List(List(Value('string'))) | Full model response [sample_idx][round_idx] = response string |
+| tokens_by_sample | List(Value('int64')) | Total token count for each sample (based on continuation_text) |
+| num_rounds_by_sample | List(Value('int64')) | Number of continuation rounds for each sample |
+| continuation_prompts_by_sample | List(List(Value('null'))) | Continuation prompts used [sample_idx][round_idx] = prompt string |
+| continuation_text_by_sample | List(List(Value('null'))) | Truncated text used for continuation [sample_idx][round_idx]. For budget_forcing: truncated at first </think> tag. For other strategies: same as responses_by_sample. |
+| strategy_type | Value('string') | Prompting strategy: 'iterative', 'budget_forcing', or 'random_noise' |
+| termination_reason_by_sample | List(Value('string')) | Why each sample terminated: 'max_rounds_reached', 'budget_reached', or 'answer_correct' |
+| correct_round_by_sample | List(Value('int64')) | Round number where correct answer was first found (None if never correct) |
+| smart_termination_enabled | Value('bool') | Whether smart termination was enabled for this run |
+| retry_count_by_sample | List(Value('int64')) | Total retry attempts per sample due to empty or too-short responses (0 if no retries needed) |
+
+
+## Generation Parameters
+
+```json
+{
+  "script_name": "stage_01_generate_reasoning.py",
+  "model": "dfs_baseline",
+  "hyperparameters": {
+    "temperature": 0.7,
+    "max_tokens_per_round": 2000,
+    "target_total_tokens": 2000,
+    "samples": 2,
+    "strategy": "dfs_baseline",
+    "max_rounds": null,
+    "min_tokens": null,
+    "max_retries_per_round": -1
+  },
+  "input_datasets": [],
+  "description": "Countdown reasoning with dfs_baseline strategy to 2000 tokens",
+  "experiment": "degen_test_1_countdown",
+  "strategy": "dfs_baseline",
+  "arg_count": 4,
+  "seed": 42,
+  "smart_termination": false,
+  "max_rounds": null,
+  "min_tokens": null,
+  "inference_backend": "curator",
+  "status": "completed",
+  "completed_at": "2026-01-01T23:15:22.836847"
+}
+```
+
+## Experiment Documentation
+
+For complete experiment details, see [https://github.com/TAUR-Lab/Reasoning-Horizons/tree/main/experiments/degen_test_1_countdown](https://github.com/TAUR-Lab/Reasoning-Horizons/tree/main/experiments/degen_test_1_countdown)
+
+## Usage
+
+```python
+from datasets import load_dataset
+
+dataset = load_dataset("reasoning-degeneration-dev/test-dfs-baseline", split="train")
+print(f"Loaded {len(dataset)} rows")
+```
+
+---
+
+*This dataset is tracked in [reasoning-degeneration-dev/PROJECT-MANIFEST](https://huggingface.co/datasets/reasoning-degeneration-dev/PROJECT-MANIFEST)*