tool_choose

This model is a fine-tuned version of bert-base-multilingual-cased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0897
• Micro f1: 0.8434
• Macro f1: 0.7771

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: 3e-05
• train_batch_size: 4
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Micro f1	Macro f1
0.2257	1.0	223	0.1608	0.0691	0.0279
0.131	2.0	446	0.1085	0.6282	0.2107
0.0971	3.0	669	0.0913	0.6594	0.2450
0.0791	4.0	892	0.0813	0.7333	0.2889
0.0676	5.0	1115	0.0730	0.7719	0.3426
0.0598	6.0	1338	0.0681	0.7931	0.3759
0.0499	7.0	1561	0.0756	0.7658	0.3849
0.0442	8.0	1784	0.0688	0.7894	0.3936
0.0404	9.0	2007	0.0637	0.8145	0.4532
0.0334	10.0	2230	0.0593	0.8276	0.4813
0.0293	11.0	2453	0.0672	0.8084	0.4919
0.0282	12.0	2676	0.0683	0.7967	0.5406
0.0244	13.0	2899	0.0617	0.8297	0.5594
0.0212	14.0	3122	0.0624	0.8372	0.6604
0.0201	15.0	3345	0.0731	0.7950	0.5878
0.0188	16.0	3568	0.0651	0.8283	0.6192
0.0157	17.0	3791	0.0705	0.8252	0.6689
0.0152	18.0	4014	0.0726	0.8115	0.6558
0.0138	19.0	4237	0.0707	0.8318	0.7159
0.0126	20.0	4460	0.0677	0.8387	0.7002
0.0129	21.0	4683	0.0707	0.8269	0.7254
0.0098	22.0	4906	0.0689	0.8257	0.7111
0.0089	23.0	5129	0.0793	0.8127	0.6561
0.0089	24.0	5352	0.0731	0.8227	0.6963
0.009	25.0	5575	0.0783	0.8203	0.7076
0.0099	26.0	5798	0.0745	0.8348	0.7155
0.0089	27.0	6021	0.0685	0.8458	0.7208
0.0077	28.0	6244	0.0780	0.8197	0.6605
0.0081	29.0	6467	0.0803	0.8193	0.6366
0.0085	30.0	6690	0.0764	0.8259	0.7797
0.0074	31.0	6913	0.0809	0.8269	0.7182
0.0036	32.0	7136	0.0808	0.8283	0.7305
0.0083	33.0	7359	0.0810	0.8378	0.7481
0.0071	34.0	7582	0.0826	0.8329	0.7348
0.0058	35.0	7805	0.1001	0.8041	0.6292
0.0047	36.0	8028	0.0864	0.8296	0.7206
0.006	37.0	8251	0.0820	0.8388	0.7131
0.0053	38.0	8474	0.0858	0.8194	0.7486
0.0056	39.0	8697	0.0902	0.8219	0.6887
0.0044	40.0	8920	0.0897	0.8434	0.7771

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1