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: 3.1859
Accuracy: 0.4194
Perplexity: 24.1889
Bleu: 0.1461

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	Accuracy	Bleu	Validation Loss	Perplexity
6.0432	0.2806	500	0.1909	0.0378	5.9180	371.6605
5.0476	0.5612	1000	0.2633	0.0612	4.8985	134.0910
4.3528	0.8418	1500	0.3182	0.0834	4.2398	69.3933
3.9497	1.1223	2000	0.3520	0.1054	3.8879	48.8078
3.7614	1.4029	2500	0.3674	0.1207	3.7128	40.9670
3.6543	1.6835	3000	0.3780	0.1310	3.5902	36.2404
3.5527	1.9641	3500	0.3864	0.1337	3.5048	33.2757
3.4348	2.2447	4000	0.3923	0.1361	3.4401	31.1898
3.3739	2.5253	4500	3.3868	0.3974	29.5718	0.1419
3.3441	2.8058	5000	3.3419	0.4020	28.2718	0.1394
3.2252	3.0864	5500	3.3067	0.4057	27.2940	0.1432
3.2188	3.3670	6000	3.2775	0.4088	26.5107	0.1421
3.1971	3.6476	6500	3.2502	0.4115	25.7958	0.1426
3.1722	3.9282	7000	3.2266	0.4143	25.1936	0.1446
3.1052	4.2088	7500	3.2103	0.4163	24.7864	0.1433
3.0672	4.4893	8000	3.1967	0.4180	24.4514	0.1438
3.0774	4.7699	8500	3.1859	0.4194	24.1889	0.1461

Framework versions

Transformers 4.49.0
Pytorch 2.6.0+cu124
Datasets 3.3.2
Tokenizers 0.21.0