Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -13,7 +13,7 @@ task_categories:
|
|
| 13 |
language:
|
| 14 |
- en
|
| 15 |
size_categories:
|
| 16 |
-
-
|
| 17 |
configs:
|
| 18 |
- config_name: all
|
| 19 |
data_files:
|
|
@@ -61,21 +61,23 @@ dataset_info:
|
|
| 61 |
dtype: string
|
| 62 |
splits:
|
| 63 |
- name: train
|
| 64 |
-
num_examples:
|
| 65 |
---
|
| 66 |
|
| 67 |
# ML SWE Prompts
|
| 68 |
|
| 69 |
Unified collection of ML/training-related software engineering prompts for OPD distillation training. All prompts are in English.
|
| 70 |
|
|
|
|
|
|
|
| 71 |
## Splits
|
| 72 |
|
| 73 |
| Config | Source | Rows | Description |
|
| 74 |
|--------|--------|------|-------------|
|
| 75 |
-
| `all` | Combined |
|
| 76 |
-
| `swe_bench_ml` | SWE-bench train |
|
| 77 |
| `swe_dev_sft` | SWE-Dev-train | 3,054 | ML-related agent conversations (SFT format) |
|
| 78 |
-
| `swe_dev_rft` | SWE-Dev-train |
|
| 79 |
| `terminal_bench_verified` | Terminal-Bench 2 Verified | 89 | Task instructions from TB2 verified tasks |
|
| 80 |
| `terminal_bench_trajectories` | TB2 Leaderboard | 18 | Unique ML task prompts from agent trajectories (deduplicated) |
|
| 81 |
|
|
@@ -88,10 +90,6 @@ All rows have at minimum `source` and `prompt`. Additional fields vary by source
|
|
| 88 |
- **terminal_bench_verified**: `task_name`
|
| 89 |
- **terminal_bench_trajectories**: `task_name`, `model`, `agent`, `reward`
|
| 90 |
|
| 91 |
-
## ML Repos in SWE-bench
|
| 92 |
-
|
| 93 |
-
huggingface (1,058), Lightning-AI (377), ray-project (342), numpy (937), scipy (101), tensorflow (50), open-mmlab (111), explosion/spaCy (41), pandas-dev (5,049), Qiskit (1,406)
|
| 94 |
-
|
| 95 |
## Citation
|
| 96 |
|
| 97 |
```bibtex
|
|
|
|
| 13 |
language:
|
| 14 |
- en
|
| 15 |
size_categories:
|
| 16 |
+
- 1K<n<10K
|
| 17 |
configs:
|
| 18 |
- config_name: all
|
| 19 |
data_files:
|
|
|
|
| 61 |
dtype: string
|
| 62 |
splits:
|
| 63 |
- name: train
|
| 64 |
+
num_examples: 6220
|
| 65 |
---
|
| 66 |
|
| 67 |
# ML SWE Prompts
|
| 68 |
|
| 69 |
Unified collection of ML/training-related software engineering prompts for OPD distillation training. All prompts are in English.
|
| 70 |
|
| 71 |
+
Filtered to core ML repos: huggingface (1,058), numpy (937), Lightning-AI (377), ray-project (342). Excludes pandas-dev, qiskit, open-mmlab, scipy, tensorflow, spaCy.
|
| 72 |
+
|
| 73 |
## Splits
|
| 74 |
|
| 75 |
| Config | Source | Rows | Description |
|
| 76 |
|--------|--------|------|-------------|
|
| 77 |
+
| `all` | Combined | 6,220 | All prompts combined |
|
| 78 |
+
| `swe_bench_ml` | SWE-bench train | 2,714 | Problem statements from core ML repos (HF, numpy, Lightning, Ray) + keyword-matched from SWE-Dev |
|
| 79 |
| `swe_dev_sft` | SWE-Dev-train | 3,054 | ML-related agent conversations (SFT format) |
|
| 80 |
+
| `swe_dev_rft` | SWE-Dev-train | 345 | ML-related agent conversations (RFT format, with rewards) |
|
| 81 |
| `terminal_bench_verified` | Terminal-Bench 2 Verified | 89 | Task instructions from TB2 verified tasks |
|
| 82 |
| `terminal_bench_trajectories` | TB2 Leaderboard | 18 | Unique ML task prompts from agent trajectories (deduplicated) |
|
| 83 |
|
|
|
|
| 90 |
- **terminal_bench_verified**: `task_name`
|
| 91 |
- **terminal_bench_trajectories**: `task_name`, `model`, `agent`, `reward`
|
| 92 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 93 |
## Citation
|
| 94 |
|
| 95 |
```bibtex
|