metadata

datasets:
  - xnli
model-index:
  - name: mcontriever-msmarco-xnli
    results: []
pipeline_tag: zero-shot-classification
language:
  - multilingual
  - en
  - ar
  - bg
  - de
  - el
  - es
  - fr
  - ru
  - sw
  - th
  - tr
  - ur
  - vi
  - zh
license: mit
widget:
  - text: Angela Merkel ist eine Politikerin in Deutschland und Vorsitzende der CDU
    candidate_labels: politics, economy, entertainment, environment

mcontriever-msmarco-xnli

This model is a fine-tuned version of facebook/mcontriever-msmarco on the XNLI dataset.

Model description

Unsupervised Dense Information Retrieval with Contrastive Learning. Gautier Izacard, Mathilde Caron, Lucas Hosseini, Sebastian Riedel, Piotr Bojanowski, Armand Joulin, Edouard Grave, arXiv 2021

How to use the model

With the zero-shot classification pipeline

The model can be loaded with the zero-shot-classification pipeline like so:

from transformers import pipeline
classifier = pipeline("zero-shot-classification",
                      model="mjwong/mcontriever-msmarco-xnli")

You can then use this pipeline to classify sequences into any of the class names you specify.

sequence_to_classify = "Angela Merkel ist eine Politikerin in Deutschland und Vorsitzende der CDU"
candidate_labels = ["politics", "economy", "entertainment", "environment"]
classifier(sequence_to_classify, candidate_labels)

If more than one candidate label can be correct, pass multi_class=True to calculate each class independently:

candidate_labels = ["politics", "economy", "entertainment", "environment"]
classifier(sequence_to_classify, candidate_labels, multi_label=True)

With manual PyTorch

The model can also be applied on NLI tasks like so:

import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification

# device = "cuda:0" or "cpu"
device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")

model_name = "mjwong/mcontriever-msmarco-xnli"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)

premise = "But I thought you'd sworn off coffee."
hypothesis = "I thought that you vowed to drink more coffee."

input = tokenizer(premise, hypothesis, truncation=True, return_tensors="pt")
output = model(input["input_ids"].to(device))
prediction = torch.softmax(output["logits"][0], -1).tolist()
label_names = ["entailment", "neutral", "contradiction"]
prediction = {name: round(float(pred) * 100, 2) for pred, name in zip(prediction, label_names)}
print(prediction)

Eval results

The model was evaluated using the XNLI test sets on 14 languages: English (en), Arabic (ar), Bulgarian (bg), German (de), Greek (el), Spanish (es), French (fr), Russian (ru), Swahili (sw), Thai (th), Turkish (tr), Urdu (ur), Vietnam (vi) and Chinese (zh). The metric used is accuracy.

Datasets	en	ar	bg	de	el	es	fr	ru	sw	th	tr	ur	vi	zh
mcontriever-xnli	0.820	0.733	0.773	0.774	0.748	0.788	0.781	0.755	0.690	0.690	0.741	0.647	0.766	0.767
mcontriever-msmarco-xnli	0.822	0.731	0.763	0.775	0.752	0.785	0.778	0.749	0.694	0.682	0.738	0.641	0.759	0.768

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 2e-05
train_batch_size: 16
eval_batch_size: 16
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 2

Framework versions

Transformers 4.28.1
Pytorch 1.12.1+cu116
Datasets 2.11.0
Tokenizers 0.12.1