metadata

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

gpt2-wikitext

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

Model description

More information needed

More information needed

More information needed

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 Loss	Epoch	Step	Validation Loss	Accuracy	Perplexity	Bleu
6.0797	0.2806	500	5.9513	0.1886	384.2374	0.0328
5.0682	0.5612	1000	4.9243	0.2616	137.5896	0.0640
4.3376	0.8418	1500	4.2176	0.3220	67.8687	0.0837
3.946	1.1223	2000	3.8835	0.3531	48.5917	0.1062
3.7578	1.4029	2500	3.7072	0.3683	40.7404	0.1239
3.6484	1.6835	3000	3.5824	0.3790	35.9600	0.1312
3.5434	1.9641	3500	3.4942	0.3874	32.9255	0.1354
3.4228	2.2447	4000	3.4280	0.3938	30.8151	0.1332
3.3604	2.5253	4500	3.3732	0.3989	29.1719	0.1409
3.3288	2.8058	5000	3.3268	0.4039	27.8495	0.1396
3.2073	3.0864	5500	3.2888	0.4079	26.8096	0.1410
3.2009	3.3670	6000	3.2614	0.4109	26.0870	0.1426
3.1787	3.6476	6500	3.2330	0.4141	25.3543	0.1394
3.1528	3.9282	7000	3.2094	0.4164	24.7635	0.1459
3.0849	4.2088	7500	3.1927	0.4182	24.3547	0.1420
3.0471	4.4893	8000	3.1799	0.4200	24.0448	0.1476
3.0571	4.7699	8500	3.1672	0.4214	23.7410	0.1474