codet5-small-Generate-docstrings-for-Python-bs-32

This model is a fine-tuned version of Salesforce/codet5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.1105
• Rouge1: 0.3307
• Rouge2: 0.16
• Rougel: 0.297
• Rougelsum: 0.3149
• Gen Len: 16.7441

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: 32
• eval_batch_size: 32
• 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	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
2.7701	1.0	4472	2.3322	0.3225	0.1525	0.2894	0.3067	16.3153
2.4907	2.0	8944	2.2464	0.328	0.1555	0.293	0.3119	17.0097
2.405	3.0	13416	2.2004	0.3267	0.1562	0.2934	0.311	16.4531
2.3512	4.0	17888	2.1696	0.3292	0.1571	0.2944	0.3134	17.3872
2.3144	5.0	22360	2.1503	0.3293	0.1586	0.2954	0.3137	16.932
2.2862	6.0	26832	2.1355	0.3307	0.1588	0.2962	0.3149	17.0269
2.2666	7.0	31304	2.1246	0.33	0.1594	0.2962	0.3144	16.7064
2.2514	8.0	35776	2.1163	0.3305	0.1595	0.2968	0.3145	16.4765
2.2401	9.0	40248	2.1120	0.3305	0.1595	0.2967	0.3147	16.763
2.2333	10.0	44720	2.1105	0.3307	0.16	0.297	0.3149	16.7441

Framework versions

• Transformers 4.33.0.dev0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4.dev0
• Tokenizers 0.13.3