Deberta V3 Large Finetuned Mnli

🚀 DeBERTa-v3-large fine-tuned on MNLI

This model is a fine-tuned version of microsoft/deberta-v3-large on the GLUE MNLI dataset. It achieves high accuracy in text classification and natural language inference tasks, providing reliable performance for NLU applications.

🚀 Quick Start

This model is a fine-tuned version of microsoft/deberta-v3-large on the GLUE MNLI dataset. It achieves the following results on the evaluation set:

• Loss: 0.6763
• Accuracy: 0.8949

✨ Features

• Advanced Architecture: DeBERTa improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder.
• Enhanced Efficiency: In DeBERTa V3, efficiency is further improved using ELECTRA - Style pre - training with Gradient Disentangled Embedding Sharing.
• High Performance: Outperforms RoBERTa on a majority of NLU tasks with 80GB training data, and the V3 version significantly improves performance on downstream tasks.

📚 Documentation

Model description

DeBERTa improves the BERT and RoBERTa models using disentangled attention and enhanced mask decoder. With those two improvements, DeBERTa out perform RoBERTa on a majority of NLU tasks with 80GB training data.

In DeBERTa V3, we further improved the efficiency of DeBERTa using ELECTRA - Style pre - training with Gradient Disentangled Embedding Sharing. Compared to DeBERTa, our V3 version significantly improves the model performance on downstream tasks. You can find more technique details about the new model from our paper.

Please check the official repository for more implementation details and updates.

The DeBERTa V3 large model comes with 24 layers and a hidden size of 1024. It has 304M backbone parameters with a vocabulary containing 128K tokens which introduces 131M parameters in the Embedding layer. This model was trained using the 160GB data as DeBERTa V2.

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 3e - 05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon = 1e - 08
• lr_scheduler_type: linear
• num_epochs: 5.0
• mixed_precision_training: Native AMP

Training results

Framework versions

• Transformers 4.13.0.dev0
• Pytorch 1.10.0+cu111
• Datasets 1.16.1
• Tokenizers 0.10.3

📄 License

This project is licensed under the MIT License.