codeparrot-ds

This model is a fine-tuned version of gpt2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 5.1374

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: 0.0005
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
9.5079	0.2186	10	8.1216
7.1151	0.4372	20	7.1928
6.6137	0.6557	30	6.8735
6.2329	0.8743	40	6.5471
5.8587	1.0929	50	6.2769
5.5239	1.3115	60	6.0781
5.3796	1.5301	70	5.9107
5.234	1.7486	80	5.7720
5.1009	1.9672	90	5.6632
4.8991	2.1858	100	5.5841
4.807	2.4044	110	5.5124
4.7083	2.6230	120	5.4414
4.6195	2.8415	130	5.3711
4.5424	3.0601	140	5.3166
4.3707	3.2787	150	5.2790
4.3599	3.4973	160	5.2316
4.3317	3.7158	170	5.1933
4.2907	3.9344	180	5.1684
4.21	4.1530	190	5.1599
4.2023	4.3716	200	5.1457
4.1793	4.5902	210	5.1398
4.1662	4.8087	220	5.1374

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.0
• Tokenizers 0.19.1