gpt2-ear_1_migrants-hs_cn

This model is a fine-tuned version of gpt2-medium on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5331

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 4
• seed: 21
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss
69.995	0.03	10	64.5661
28.3476	0.06	20	18.0685
5.2324	0.08	30	6.3912
0.7104	0.11	40	2.7627
-0.8669	0.14	50	1.2027
-1.7912	0.17	60	0.8591
-1.7957	0.2	70	0.7706
-2.1587	0.23	80	0.7197
-2.1961	0.25	90	0.6167
-2.1901	0.28	100	0.6059
-2.0514	0.31	110	0.5845
-2.0839	0.34	120	0.5717
-2.1579	0.37	130	0.5667
-2.0266	0.4	140	0.5601
-2.2698	0.42	150	0.5582
-2.1635	0.45	160	0.5694
-2.1359	0.48	170	0.5584
-2.1628	0.51	180	0.5510
-2.0485	0.54	190	0.5520
-2.1333	0.57	200	0.5431
-2.2908	0.59	210	0.5438
-2.1131	0.62	220	0.5545
-2.1988	0.65	230	0.5371
-2.187	0.68	240	0.5349
-2.0381	0.71	250	0.5504
-2.1413	0.74	260	0.5293
-2.0951	0.76	270	0.5257
-2.2314	0.79	280	0.5263
-2.1851	0.82	290	0.5291
-2.2318	0.85	300	0.5331

Framework versions

• Transformers 4.29.0.dev0
• Pytorch 1.12.0a0+bd13bc6
• Datasets 2.12.0
• Tokenizers 0.13.3