README.md

metadata

model-index:
  - name: poltextlab/xlm-roberta-large-pooled-emotions10-v2
    results:
      - task:
          type: text-classification
        metrics:
          - name: Accuracy
            type: accuracy
            value: 81%
          - name: F1-Score
            type: f1
            value: 81%
tags:
  - text-classification
  - pytorch
metrics:
  - precision
  - recall
  - f1-score
language:
  - en
  - hu
  - fr
  - cs
  - sk
  - pl
  - de
base_model:
  - xlm-roberta-large
pipeline_tag: text-classification
library_name: transformers
license: cc-by-4.0
extra_gated_prompt: >-
  Our models are intended for academic projects and academic research only. If
  you are not affiliated with an academic institution, please reach out to us at
  huggingface [at] poltextlab [dot] com for further inquiry. If we cannot
  clearly determine your academic affiliation and use case based on your form
  data, your request may be rejected. Please allow us a few business days to
  manually review subscriptions.
extra_gated_fields:
  Country: country
  Institution: text
  Institution Email: text
  Full Name: text
  Please specify your academic project/use case you want to use the models for: text

xlm-roberta-large-pooled-emotions10-v2

An xlm-roberta-large model finetuned on sentence-level multilingual training data hand-annotated using the following labels:

• 0: "Neutral"
• 1: "Anger"
• 2: "Fear"
• 3: "Disgust"
• 4: "Sadness"
• 5: "Joy"
• 6: "Hope"
• 7: "Enthusiasm"
• 8: "Pride"
• 9: "Other emotion"

The training data we used was augmented translated texts. It covers 7 languages (English, German, French, Polish, Slovak, Czech and Hungarian) with nearly identical shares.

How to use the model

from transformers import AutoTokenizer, pipeline

tokenizer = AutoTokenizer.from_pretrained("xlm-roberta-large")
pipe = pipeline(
    model="poltextlab/xlm-roberta-large-pooled-emotions10-v2",
    task="text-classification",
    tokenizer=tokenizer,
    use_fast=False,
)

text = "We will place an immediate 6-month halt on the finance driven closure of beds and wards, and set up an independent audit of needs and facilities."
pipe(text)

Classification Report

Overall Performance:

• Accuracy: 81%
• Macro Avg: Precision: 0.82, Recall: 0.81, F1-score: 0.81
• Weighted Avg: Precision: 0.81, Recall: 0.81, F1-score: 0.81

Per-Class Metrics:

Label	Precision	Recall	F1-score	Support
Neutral (0)	0.81	0.88	0.85	9367
Anger (1)	0.73	0.70	0.72	5433
Fear (2)	0.86	0.84	0.85	5434
Disgust (3)	0.95	0.95	0.95	5437
Sadness (4)	0.90	0.85	0.88	5434
Joy (5)	0.84	0.85	0.85	5162
Hope (6)	0.59	0.63	0.61	5437
Enthusiasm (7)	0.70	0.63	0.67	5433
Pride (8)	0.82	0.82	0.82	5435
Other emotion (9)	0.98	0.95	0.97	2051

Total samples: 54,623

Inference platform

This model is used by the Babel Machine, an open-source and free natural language processing tool, designed to simplify and speed up projects for comparative research.

Cooperation

Model performance can be significantly improved by extending our training sets. We appreciate every submission of coded corpora (of any domain and language) at poltextlab{at}poltextlab{dot}com or by using the Babel Machine.

Debugging and issues

This architecture uses the sentencepiece tokenizer. In order to use the model before transformers==4.27 you need to install it manually.

If you encounter a RuntimeError when loading the model using the from_pretrained() method, adding ignore_mismatched_sizes=True should solve the issue.