Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -1,25 +1,57 @@
|
|
| 1 |
---
|
| 2 |
-
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
dtype: int64
|
| 8 |
-
- name: tool
|
| 9 |
-
dtype: string
|
| 10 |
-
- name: prompt
|
| 11 |
-
dtype: string
|
| 12 |
-
- name: response
|
| 13 |
-
dtype: string
|
| 14 |
-
splits:
|
| 15 |
-
- name: train
|
| 16 |
-
num_bytes: 1325244
|
| 17 |
-
num_examples: 8000
|
| 18 |
-
download_size: 281564
|
| 19 |
-
dataset_size: 1325244
|
| 20 |
-
configs:
|
| 21 |
-
- config_name: default
|
| 22 |
-
data_files:
|
| 23 |
-
- split: train
|
| 24 |
-
path: data/train-*
|
| 25 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
license: mit
|
| 3 |
+
tags:
|
| 4 |
+
- grpo-tool-saturation
|
| 5 |
+
- sft
|
| 6 |
+
- llamafactory-ready
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
---
|
| 8 |
+
|
| 9 |
+
# grpo-tool-sat-sft-corpus-v1
|
| 10 |
+
|
| 11 |
+
SFT-only view of grpo-tool-sat-dataset-v1. 8000 demonstration rows with prompt/response columns ready for LlamaFactory. Overlap map-fraction 0.594; token-level shuffled across r-classes.
|
| 12 |
+
|
| 13 |
+
## Dataset Info
|
| 14 |
+
|
| 15 |
+
- **Rows**: 8000
|
| 16 |
+
- **Columns**: 5
|
| 17 |
+
|
| 18 |
+
## Columns
|
| 19 |
+
|
| 20 |
+
| Column | Type | Description |
|
| 21 |
+
|--------|------|-------------|
|
| 22 |
+
| k | Value('int64') | Key integer in [0, 9999] |
|
| 23 |
+
| r | Value('int64') | k mod 3 (0 = M_only, 1 = T_only, 2 = Overlap) |
|
| 24 |
+
| tool | Value('string') | Tool used in this demo (map or table) |
|
| 25 |
+
| prompt | Value('string') | User-side prompt: Key: <k> |
|
| 26 |
+
| response | Value('string') | Target assistant completion: prose + <tool_call> + <observation> + <answer> |
|
| 27 |
+
|
| 28 |
+
## Generation Parameters
|
| 29 |
+
|
| 30 |
+
```json
|
| 31 |
+
{
|
| 32 |
+
"script_name": "src/data_gen.py (post-filter)",
|
| 33 |
+
"model": "n/a",
|
| 34 |
+
"description": "SFT-only view of grpo-tool-sat-dataset-v1. 8000 demonstration rows with prompt/response columns ready for LlamaFactory. Overlap map-fraction 0.594; token-level shuffled across r-classes.",
|
| 35 |
+
"hyperparameters": {
|
| 36 |
+
"seed": 1,
|
| 37 |
+
"overlap_skew_map": 0.6,
|
| 38 |
+
"hash_slice": 6
|
| 39 |
+
},
|
| 40 |
+
"input_datasets": [
|
| 41 |
+
"raca-workspace-v1/grpo-tool-sat-dataset-v1"
|
| 42 |
+
]
|
| 43 |
+
}
|
| 44 |
+
```
|
| 45 |
+
|
| 46 |
+
## Usage
|
| 47 |
+
|
| 48 |
+
```python
|
| 49 |
+
from datasets import load_dataset
|
| 50 |
+
|
| 51 |
+
dataset = load_dataset("raca-workspace-v1/grpo-tool-sat-sft-corpus-v1", split="train")
|
| 52 |
+
print(f"Loaded {len(dataset)} rows")
|
| 53 |
+
```
|
| 54 |
+
|
| 55 |
+
---
|
| 56 |
+
|
| 57 |
+
*Uploaded via [RACA](https://github.com/Zayne-sprague/Dr-Claude-Code) hf_utility.*
|