metadata

license: apache-2.0
base_model: distilbert-base-uncased
tags:
  - generated_from_trainer
  - emotion_classfication
metrics:
  - accuracy
  - f1
model-index:
  - name: distilbert-base-uncased-finetuned-emotion-classification-10_epochs
    results: []
datasets:
  - dair-ai/emotion
language:
  - en

distilbert-base-uncased-finetuned-emotion-classification-10_epochs

This model is a fine-tuned version of distilbert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.1464
Accuracy: 0.9375
F1 score: 0.9376

Model description

This model is fine-tuned on the 'emotion' dataset to classify text into six emotion categories: sadness, joy, love, anger, fear, and surprise.

Use the Model

from transformers import pipeline
import pandas as pd
emt_clf = pipeline("text-classification", model="Swoodplays/emotion-classification")

text = "I saw a movie today and it was really good."

preds = emt_clf(text, return_all_scores=True)
labels = ['sadness', 'joy', 'love', 'anger', 'fear', 'surprise']

print(preds)

preds_df = pd.DataFrame(preds[0])
plt.bar(labels, 100 * preds_df["score"])
plt.title(f'"{text}"')
plt.xlabel("Classfied emotions")
plt.ylabel("Class probability (%)")
plt.show()

Training and evaluation data

dair-ai/emotion

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 2e-05
train_batch_size: 128
eval_batch_size: 128
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 score
No log	1.0	125	0.4449	0.873	0.8654
No log	2.0	250	0.2166	0.9265	0.9270
No log	3.0	375	0.1726	0.933	0.9339
No log	4.0	500	0.1552	0.9385	0.9386
No log	5.0	625	0.1439	0.938	0.9383
No log	6.0	750	0.1435	0.937	0.9370
No log	7.0	875	0.1481	0.9355	0.9356
No log	8.0	1000	0.1402	0.935	0.9352
No log	9.0	1125	0.1491	0.9355	0.9355
No log	10.0	1250	0.1464	0.9375	0.9376

Framework versions

Transformers 4.44.0
Pytorch 2.4.0
Datasets 2.21.0
Tokenizers 0.19.1