PhobertLexicalMeta-v2

This model is a fine-tuned version of gechim/metadata-cls-no-gov-8k-v3 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3926
• Accuracy: 0.9062
• F1: 0.8781

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: 2e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.8772	100	0.2699	0.9080	0.8801
0.1564	1.7544	200	0.2984	0.9011	0.8723
0.073	2.6316	300	0.3218	0.8987	0.8705
0.0502	3.5088	400	0.3472	0.8927	0.8641
0.0326	4.3860	500	0.3627	0.8941	0.8635
0.0285	5.2632	600	0.3752	0.8964	0.8685
0.0179	6.1404	700	0.3666	0.9025	0.8734
0.0156	7.0175	800	0.3759	0.9043	0.8748
0.0156	7.8947	900	0.3830	0.9080	0.8788
0.011	8.7719	1000	0.3917	0.9039	0.8746
0.0092	9.6491	1100	0.3926	0.9062	0.8781

Framework versions

• Transformers 4.43.3
• Pytorch 2.1.2
• Datasets 2.20.0
• Tokenizers 0.19.1