distilbert-base-uncased-swag

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

distilbert-base-uncased-swag

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

• Loss: 0.8950
• Accuracy: 0.6426

Model description

More information needed

Intended uses & limitations

This model should be used as an expert in the Meteor-of-LoRA framework.

Training and evaluation data

The data were splitted based on HuggingFace default dataset:

dataset = load_dataset("swag")

Training procedure

Our approach focuses explicitly on adapting the Transformers weights' Wq (query) and Wv (value) in the attention module for parameter efficiency.

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-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
• num_epochs: 5

Training results

Framework versions

• PEFT 0.12.1.dev0
• Transformers 4.45.0.dev0
• Pytorch 2.3.1+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1