codet5-small-java-buggy-to-fixed-code

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

• Loss: 0.1496
• Rouge1: 28.945
• Rouge2: 25.498
• Rougel: 28.8801
• Rougelsum: 28.9201
• Gen Len: 18.9938

Model description

More information needed

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: 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: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
0.26	1.0	3750	0.1799	28.745	25.2441	28.6738	28.7241	18.9872
0.212	2.0	7500	0.1589	28.8636	25.4151	28.7997	28.8427	18.9926
0.1975	3.0	11250	0.1510	28.9016	25.4349	28.8356	28.8809	18.989
0.1887	4.0	15000	0.1496	28.945	25.498	28.8801	28.9201	18.9938

Framework versions

• Transformers 4.36.0.dev0
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0