Update README.md
Browse files
README.md
CHANGED
|
@@ -9,4 +9,52 @@ metrics:
|
|
| 9 |
- bertscore
|
| 10 |
pipeline_tag: text-classification
|
| 11 |
---
|
| 12 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
- bertscore
|
| 10 |
pipeline_tag: text-classification
|
| 11 |
---
|
| 12 |
+
# AlignScoreCS
|
| 13 |
+
MultiTask multilingual model for assessing facticity in various diverse tasks. We followed the initial paper AlignScore https://arxiv.org/abs/2305.16739.
|
| 14 |
+
|
| 15 |
+
|
| 16 |
+
|
| 17 |
+
# Training datasets
|
| 18 |
+
The following table shows datasets that has been utilized for training the model. We translated these english datasets to Czech using seamLessM4t.
|
| 19 |
+
|
| 20 |
+
| NLP Task | Dataset | Training Task | Context (n words) | Claim (n words) | Sample Count |
|
| 21 |
+
|-----------------------|-------------------|---------------|-------------------|-----------------|--------------|
|
| 22 |
+
| NLI | SNLI | 3-way | 10 | 13 | Cs: 500k |
|
| 23 |
+
| | | | | | En: 550k |
|
| 24 |
+
| | MultiNLI | 3-way | 16 | 20 | Cs: 393k |
|
| 25 |
+
| | | | | | En: 393k |
|
| 26 |
+
| | Adversarial NLI | 3-way | 48 | 54 | Cs: 163k |
|
| 27 |
+
| | | | | | En: 163k |
|
| 28 |
+
| | DocNLI | 2-way | 97 | 285 | Cs: 200k |
|
| 29 |
+
| | | | | | En: 942k |
|
| 30 |
+
| Fact Verification | NLI-style FEVER | 3-way | 48 | 50 | Cs: 208k |
|
| 31 |
+
| | | | | | En: 208k |
|
| 32 |
+
| | Vitamin C | 3-way | 23 | 25 | Cs: 371k |
|
| 33 |
+
| | | | | | En: 371k |
|
| 34 |
+
| Paraphrase | QQP | 2-way | 9 | 11 | Cs: 162k |
|
| 35 |
+
| | | | | | En: 364k |
|
| 36 |
+
| | PAWS | 2-way | - | 18 | Cs: - |
|
| 37 |
+
| | | | | | En: 707k |
|
| 38 |
+
| | PAWS labeled | 2-way | 18 | - | Cs: 49k |
|
| 39 |
+
| | | | | | En: - |
|
| 40 |
+
| | PAWS unlabeled | 2-way | 18 | - | Cs: 487k |
|
| 41 |
+
| | | | | | En: - |
|
| 42 |
+
| STS | SICK | reg | - | 10 | Cs: - |
|
| 43 |
+
| | | | | | En: 4k |
|
| 44 |
+
| | STS Benchmark | reg | - | 10 | Cs: - |
|
| 45 |
+
| | | | | | En: 6k |
|
| 46 |
+
| | Free-N1 | reg | 18 | - | Cs: 20k |
|
| 47 |
+
| | | | | | En: - |
|
| 48 |
+
| QA | SQuAD v2 | 2-way | 105 | 119 | Cs: 130k |
|
| 49 |
+
| | | | | | En: 130k |
|
| 50 |
+
| | RACE | 2-way | 266 | 273 | Cs: 200k |
|
| 51 |
+
| | | | | | En: 351k |
|
| 52 |
+
| Information Retrieval| MS MARCO | 2-way | 49 | 56 | Cs: 200k |
|
| 53 |
+
| | | | | | En: 5M |
|
| 54 |
+
| Summarization | WikiHow | 2-way | 434 | 508 | Cs: 157k |
|
| 55 |
+
| | | | | | En: 157k |
|
| 56 |
+
| | SumAug | 2-way | - | - | Cs: - |
|
| 57 |
+
| | | | | | En: - |
|
| 58 |
+
|
| 59 |
+
|
| 60 |
+
|