metadata

library_name: transformers
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - bleu
model-index:
  - name: parallel-mean-bottleneck-gpt2-medium-wikitext
    results: []

parallel-mean-bottleneck-gpt2-medium-wikitext

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

Loss: 3.1864
Accuracy: 0.4195
Perplexity: 24.2005
Bleu: 0.1476

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: 0.0001
train_batch_size: 64
eval_batch_size: 64
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Perplexity	Bleu
6.0443	0.2806	500	5.9164	0.1901	371.0844	0.0350
5.0429	0.5612	1000	4.8947	0.2638	133.5839	0.0647
4.3531	0.8418	1500	4.2426	0.3176	69.5891	0.0829
3.9503	1.1223	2000	3.8874	0.3517	48.7842	0.1050
3.7613	1.4029	2500	3.7124	0.3672	40.9504	0.1211
3.6548	1.6835	3000	3.5911	0.3780	36.2753	0.1308
3.5531	1.9641	3500	3.5068	0.3860	33.3428	0.1340
3.4344	2.2447	4000	3.4411	0.3920	31.2224	0.1356
3.3743	2.5253	4500	3.3875	0.3972	29.5917	0.1389
3.3443	2.8058	5000	3.3429	0.4016	28.3017	0.1373
3.225	3.0864	5500	3.3080	0.4055	27.3310	0.1419
3.2185	3.3670	6000	3.2781	0.4090	26.5258	0.1463
3.1972	3.6476	6500	3.2500	0.4121	25.7899	0.1453
3.1719	3.9282	7000	3.2268	0.4144	25.1990	0.1465
3.1052	4.2088	7500	3.2109	0.4162	24.8018	0.1472
3.0672	4.4893	8000	3.1978	0.4179	24.4788	0.1469
3.0773	4.7699	8500	3.1864	0.4195	24.2005	0.1476

Framework versions

Transformers 4.49.0
Pytorch 2.6.0+cu124
Datasets 3.3.2
Tokenizers 0.21.0