saaz

This model is a fine-tuned version of distilgpt2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2110

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: 0.0002
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
No log	1.0	1	0.9750
No log	2.0	2	0.6813
No log	3.0	3	0.4624
No log	4.0	4	0.3387
No log	5.0	5	0.2877
No log	6.0	6	0.2399
No log	7.0	7	0.2235
No log	8.0	8	0.2259
No log	9.0	9	0.2267
No log	10.0	10	0.2259
No log	11.0	11	0.2259
No log	12.0	12	0.2173
No log	13.0	13	0.2126
No log	14.0	14	0.2077
No log	15.0	15	0.2066
No log	16.0	16	0.2067
No log	17.0	17	0.2089
No log	18.0	18	0.2103
No log	19.0	19	0.2108
No log	20.0	20	0.2110

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2