File size: 7,662 Bytes
a708ccc 34bc863 a708ccc 35cdaef a708ccc 4954f85 b717ad8 34bc863 3a9a7f3 b0577be a708ccc f2c795e a708ccc 35cdaef a708ccc 35cdaef f2c795e a708ccc f2c795e a708ccc 267172a 4b15c4c 267172a c72e58a a708ccc 35cdaef 3f00ab8 35cdaef |
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 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 |
---
language:
- en
license: mit
tags:
- text-classification
- zero-shot-classification
metrics:
- accuracy
datasets:
- multi_nli
- anli
- fever
pipeline_tag: zero-shot-classification
model-index:
- name: MoritzLaurer/DeBERTa-v3-base-mnli-fever-anli
results:
- task:
type: natural-language-inference
name: Natural Language Inference
dataset:
name: anli
type: anli
config: plain_text
split: test_r3
metrics:
- name: Accuracy
type: accuracy
value: 0.495
verified: true
- name: Precision Macro
type: precision
value: 0.4984740618243923
verified: true
- name: Precision Micro
type: precision
value: 0.495
verified: true
- name: Precision Weighted
type: precision
value: 0.4984357572868885
verified: true
- name: Recall Macro
type: recall
value: 0.49461028192371476
verified: true
- name: Recall Micro
type: recall
value: 0.495
verified: true
- name: Recall Weighted
type: recall
value: 0.495
verified: true
- name: F1 Macro
type: f1
value: 0.4942810999491704
verified: true
- name: F1 Micro
type: f1
value: 0.495
verified: true
- name: F1 Weighted
type: f1
value: 0.4944671868893595
verified: true
- name: loss
type: loss
value: 1.8788293600082397
verified: true
- task:
type: natural-language-inference
name: Natural Language Inference
dataset:
name: anli
type: anli
config: plain_text
split: test_r1
metrics:
- name: Accuracy
type: accuracy
value: 0.712
verified: true
- name: Precision Macro
type: precision
value: 0.7134839439315348
verified: true
- name: Precision Micro
type: precision
value: 0.712
verified: true
- name: Precision Weighted
type: precision
value: 0.7134676028447461
verified: true
- name: Recall Macro
type: recall
value: 0.7119814425203647
verified: true
- name: Recall Micro
type: recall
value: 0.712
verified: true
- name: Recall Weighted
type: recall
value: 0.712
verified: true
- name: F1 Macro
type: f1
value: 0.7119226991285647
verified: true
- name: F1 Micro
type: f1
value: 0.712
verified: true
- name: F1 Weighted
type: f1
value: 0.7119242267218338
verified: true
- name: loss
type: loss
value: 1.0105403661727905
verified: true
- task:
type: natural-language-inference
name: Natural Language Inference
dataset:
name: multi_nli
type: multi_nli
config: default
split: validation_mismatched
metrics:
- name: Accuracy
type: accuracy
value: 0.902766476810415
verified: true
- name: Precision Macro
type: precision
value: 0.9023816542652491
verified: true
- name: Precision Micro
type: precision
value: 0.902766476810415
verified: true
- name: Precision Weighted
type: precision
value: 0.9034597464719761
verified: true
- name: Recall Macro
type: recall
value: 0.9024304801555488
verified: true
- name: Recall Micro
type: recall
value: 0.902766476810415
verified: true
- name: Recall Weighted
type: recall
value: 0.902766476810415
verified: true
- name: F1 Macro
type: f1
value: 0.9023086094638595
verified: true
- name: F1 Micro
type: f1
value: 0.902766476810415
verified: true
- name: F1 Weighted
type: f1
value: 0.9030161011457231
verified: true
- name: loss
type: loss
value: 0.3283354640007019
verified: true
---
# DeBERTa-v3-base-mnli-fever-anli
## Model description
This model was trained on the MultiNLI, Fever-NLI and Adversarial-NLI (ANLI) datasets, which comprise 763 913 NLI hypothesis-premise pairs. This base model outperforms almost all large models on the [ANLI benchmark](https://github.com/facebookresearch/anli).
The base model is [DeBERTa-v3-base from Microsoft](https://huggingface.co/microsoft/deberta-v3-base). The v3 variant of DeBERTa substantially outperforms previous versions of the model by including a different pre-training objective, see annex 11 of the original [DeBERTa paper](https://arxiv.org/pdf/2006.03654.pdf).
For highest performance (but less speed), I recommend using https://huggingface.co/MoritzLaurer/DeBERTa-v3-large-mnli-fever-anli-ling-wanli.
## Intended uses & limitations
#### How to use the model
```python
from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch
device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
model_name = "MoritzLaurer/DeBERTa-v3-base-mnli-fever-anli"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)
premise = "I first thought that I liked the movie, but upon second thought it was actually disappointing."
hypothesis = "The movie was good."
input = tokenizer(premise, hypothesis, truncation=True, return_tensors="pt")
output = model(input["input_ids"].to(device)) # device = "cuda:0" or "cpu"
prediction = torch.softmax(output["logits"][0], -1).tolist()
label_names = ["entailment", "neutral", "contradiction"]
prediction = {name: round(float(pred) * 100, 1) for pred, name in zip(prediction, label_names)}
print(prediction)
```
### Training data
DeBERTa-v3-base-mnli-fever-anli was trained on the MultiNLI, Fever-NLI and Adversarial-NLI (ANLI) datasets, which comprise 763 913 NLI hypothesis-premise pairs.
### Training procedure
DeBERTa-v3-base-mnli-fever-anli was trained using the Hugging Face trainer with the following hyperparameters.
```
training_args = TrainingArguments(
num_train_epochs=3, # total number of training epochs
learning_rate=2e-05,
per_device_train_batch_size=32, # batch size per device during training
per_device_eval_batch_size=32, # batch size for evaluation
warmup_ratio=0.1, # number of warmup steps for learning rate scheduler
weight_decay=0.06, # strength of weight decay
fp16=True # mixed precision training
)
```
### Eval results
The model was evaluated using the test sets for MultiNLI and ANLI and the dev set for Fever-NLI. The metric used is accuracy.
mnli-m | mnli-mm | fever-nli | anli-all | anli-r3
---------|----------|---------|----------|----------
0.903 | 0.903 | 0.777 | 0.579 | 0.495
## Limitations and bias
Please consult the original DeBERTa paper and literature on different NLI datasets for potential biases.
## Citation
If you use this model, please cite: Laurer, Moritz, Wouter van Atteveldt, Andreu Salleras Casas, and Kasper Welbers. 2022. ‘Less Annotating, More Classifying – Addressing the Data Scarcity Issue of Supervised Machine Learning with Deep Transfer Learning and BERT - NLI’. Preprint, June. Open Science Framework. https://osf.io/74b8k.
### Ideas for cooperation or questions?
If you have questions or ideas for cooperation, contact me at m{dot}laurer{at}vu{dot}nl or [LinkedIn](https://www.linkedin.com/in/moritz-laurer/)
### Debugging and issues
Note that DeBERTa-v3 was released on 06.12.21 and older versions of HF Transformers seem to have issues running the model (e.g. resulting in an issue with the tokenizer). Using Transformers>=4.13 might solve some issues.
|