codet5-base-v1

This model is a fine-tuned version of Salesforce/codet5-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1484
• Bleu Score: 0.0628
• Gen Len: 12.8108

Model description

Trained,

• on: chathuranga-jayanath/context-5-predict-token-for-fine-tune.
• epochs: 3
• metric: bleu
• tokens added: [BUG], [CONTEXT], [DELETED]

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: 2e-05
• train_batch_size: 20
• eval_batch_size: 20
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu Score	Gen Len
No log	1.0	16	0.3666	0.0607	12.9189
No log	2.0	32	0.1967	0.0618	12.9189
No log	3.0	48	0.1484	0.0628	12.8108

Framework versions

• Transformers 4.38.0.dev0
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1