transformers-qa-kaggle-tpu

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

• Train Loss: 0.2278
• Train End Logits Accuracy: 0.9244
• Train Start Logits Accuracy: 0.9207
• Validation Loss: 3.8999
• Validation End Logits Accuracy: 0.4812
• Validation Start Logits Accuracy: 0.4542
• Epoch: 14

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:

• optimizer: {'name': 'Adam', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': False, 'is_legacy_optimizer': False, 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 122160, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False}
• training_precision: float32

Training results

Train Loss	Train End Logits Accuracy	Train Start Logits Accuracy	Validation Loss	Validation End Logits Accuracy	Validation Start Logits Accuracy	Epoch
2.2837	0.4519	0.4182	2.1117	0.4890	0.4658	0
1.7361	0.5642	0.5326	2.0268	0.5035	0.4788	1
1.4664	0.6186	0.5893	2.0023	0.5093	0.4833	2
1.2479	0.6661	0.6379	2.1252	0.5057	0.4744	3
1.0596	0.7076	0.6832	2.2703	0.4975	0.4690	4
0.8999	0.7434	0.7214	2.3834	0.4968	0.4714	5
0.7661	0.7760	0.7557	2.5503	0.4906	0.4654	6
0.6520	0.8042	0.7892	2.7740	0.4922	0.4540	7
0.5549	0.8313	0.8156	3.0625	0.4884	0.4607	8
0.4739	0.8512	0.8405	3.1365	0.4862	0.4535	9
0.4072	0.8691	0.8620	3.2969	0.4830	0.4509	10
0.3515	0.8863	0.8786	3.4301	0.4852	0.4530	11
0.3025	0.9010	0.8954	3.5350	0.4814	0.4548	12
0.2646	0.9127	0.9083	3.7923	0.4832	0.4539	13
0.2278	0.9244	0.9207	3.8999	0.4812	0.4542	14

Framework versions

• Transformers 4.31.0.dev0
• TensorFlow 2.12.0
• Datasets 2.13.1
• Tokenizers 0.13.3