distilgpt2-finetuned-finance

This model is a fine-tuned version of distilgpt2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.2821

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: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 64
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
3.737	0.96	20	3.4239
3.491	1.97	41	3.3857
3.459	2.97	62	3.3642
3.4388	3.98	83	3.3495
3.422	4.99	104	3.3379
3.4102	6.0	125	3.3277
3.5688	6.96	145	3.3192
3.3892	7.96	166	3.3134
3.382	8.97	187	3.3060
3.3749	9.98	208	3.3021
3.3683	10.99	229	3.2967
3.3636	11.99	250	3.2935
3.5269	12.95	270	3.2902
3.3552	13.96	291	3.2885
3.3523	14.97	312	3.2859
3.3492	15.98	333	3.2844
3.3472	16.98	354	3.2825
3.3461	17.99	375	3.2828
3.3456	19.0	396	3.2822
3.3392	19.19	400	3.2821

Framework versions

• Transformers 4.36.0
• Pytorch 2.0.0
• Datasets 2.1.0
• Tokenizers 0.15.0