Datasets:
Upload dataset_card.md with huggingface_hub
Browse files- dataset_card.md +67 -0
dataset_card.md
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Dataset Card: curatorkit_run
|
| 2 |
+
|
| 3 |
+
## Summary
|
| 4 |
+
|
| 5 |
+
- **Run timestamp**: 2026-09-01T05:33:22.078875Z
|
| 6 |
+
- **Wall clock time**: 50.8s
|
| 7 |
+
- **Pipeline config hash**: `3b235ce9a22decbf`
|
| 8 |
+
|
| 9 |
+
## Source Documents
|
| 10 |
+
|
| 11 |
+
- Unknown
|
| 12 |
+
|
| 13 |
+
## Pipeline Stages
|
| 14 |
+
|
| 15 |
+
| Step | Input | Passed | Recovered | Rejected | Forward |
|
| 16 |
+
|------|-------|--------|-----------|----------|---------|
|
| 17 |
+
| JSONLReader | 40 | 40 | — | 0 | 40 |
|
| 18 |
+
| SchemaGate | 40 | 39 | 0 | 1 | 39 |
|
| 19 |
+
| ExactDeduplicator | 39 | 39 | — | — | 39 |
|
| 20 |
+
| TextCleaner | 39 | 39 | — | — | 39 |
|
| 21 |
+
| SecretsGate | 39 | 38 | 0 | 1 | 38 |
|
| 22 |
+
| PIIPseudonymizer | 38 | 38 | — | — | 38 |
|
| 23 |
+
| ToxicityGate | 38 | 30 | 0 | 8 | 30 |
|
| 24 |
+
| AlpacaExporter | — | 30 | — | — | 30 |
|
| 25 |
+
|
| 26 |
+
## Gate Pass Rates and Rejection Breakdown
|
| 27 |
+
|
| 28 |
+
- **JSONLReader**: 40/0 passed (N/A), 0 rejected
|
| 29 |
+
- **SchemaGate**: 39/40 passed (97.5%), 1 rejected
|
| 30 |
+
- **SecretsGate**: 38/39 passed (97.4%), 1 rejected
|
| 31 |
+
- **ToxicityGate**: 30/38 passed (78.9%), 8 rejected
|
| 32 |
+
|
| 33 |
+
|
| 34 |
+
**Rejection reasons:**
|
| 35 |
+
- `toxic_content:classifier:0.996`: 4
|
| 36 |
+
- `below_min_tokens:9`: 1
|
| 37 |
+
- `secret_detected:AWS Access Key`: 1
|
| 38 |
+
- `toxic_content:classifier:0.993`: 1
|
| 39 |
+
- `toxic_content:classifier:0.995`: 1
|
| 40 |
+
- `toxic_content:classifier:0.997`: 1
|
| 41 |
+
- `toxic_content:classifier:0.992`: 1
|
| 42 |
+
|
| 43 |
+
## Deduplication
|
| 44 |
+
|
| 45 |
+
- exact_duplicates_removed: 0
|
| 46 |
+
|
| 47 |
+
## Known Limitations
|
| 48 |
+
|
| 49 |
+
- No MinHash deduplication applied.
|
| 50 |
+
- Table extraction requires `extract_tables: true` on the PDF reader and MinerU with GPU support. Disabled by default.
|
| 51 |
+
- Token counting uses a whitespace tokenizer by default. For exact LLM context window
|
| 52 |
+
enforcement, enable `use_tiktoken: true` in the SchemaGate config.
|
| 53 |
+
- Both within-run (MinHash/Diversity) and cross-run (EmbeddingDeduplicator) deduplication
|
| 54 |
+
are available. Cross-run dedup persists an embedding index across runs; use
|
| 55 |
+
`embedding_reset_index: true` to start fresh.
|
| 56 |
+
|
| 57 |
+
## Reproduction
|
| 58 |
+
|
| 59 |
+
```bash
|
| 60 |
+
pip install curatorkit
|
| 61 |
+
curatorkit run pipeline.yaml --output-dir ./out/
|
| 62 |
+
```
|
| 63 |
+
|
| 64 |
+
## Tool Versions
|
| 65 |
+
|
| 66 |
+
- **curatorkit**: 0.1.6
|
| 67 |
+
- **python**: 3.13.15
|