File size: 1,189 Bytes
6907d07
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
---
language:
- en
multilinguality:
- monolingual
size_categories:
- 1M<n<10M
task_categories:
- feature-extraction
- sentence-similarity
pretty_name: AllNLI
tags:
- sentence-transformers
dataset_info:
- config_name: pair-class-distill
  features:
  - name: premise
    dtype: string
  - name: hypothesis
    dtype: string
  - name: label
    dtype:
      class_label:
        names:
          '0': contradiction
          '1': entailment
          '2': neutral
---

# Dataset Card for AllNLI

This dataset is a concatenation of the [SNLI](https://huggingface.co/datasets/stanfordnlp/snli) and [MultiNLI](https://huggingface.co/datasets/nyu-mll/multi_nli) datasets.

This is the same dataset as `sentence-transformers/all-nli` `pair-class` split;
however, the label ids are not identical, and teacher scores have been added from `dleemiller/ModernCE-large-nli`.

I have also added hashes for score lookup, since a lookup must be added into a custom loss function,
if using the sentence transformers CrossEncoder trainer.

The hashes were computed straightforwardly as follows:
```
df.hash = df.apply(lambda x: hashlib.md5(f"{x.premise}\n{x.hypothesis}".encode()).hexdigest(), axis=1)
```