COHeN_jer

This model is a fine-tuned version of martijn75/BERiT_2000_custom_architecture_150_epochs_2_heb_sentences_unvocalized on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4819
• Accuracy: 0.8101
• Precision: 0.8716
• Recall: 0.8101
• F1: 0.8319

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: 0.0001
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.6767	1.0	1447	0.6122	0.6696	0.8455	0.6696	0.7237
0.573	2.0	2894	0.4290	0.8143	0.8683	0.8143	0.8341
0.5015	3.0	4341	0.4526	0.8020	0.8727	0.8020	0.8264
0.4708	4.0	5788	0.4999	0.7996	0.8728	0.7996	0.8247
0.4561	5.0	7235	0.4998	0.7916	0.8713	0.7916	0.8186
0.4358	6.0	8682	0.4962	0.7969	0.8715	0.7969	0.8225
0.4308	7.0	10129	0.5183	0.7874	0.8696	0.7874	0.8153
0.4213	8.0	11576	0.5121	0.7952	0.8721	0.7952	0.8213
0.4119	9.0	13023	0.4910	0.7987	0.8712	0.7987	0.8237
0.4084	10.0	14470	0.5022	0.7838	0.8692	0.7838	0.8126
0.4032	11.0	15917	0.4656	0.7996	0.8707	0.7996	0.8243
0.3967	12.0	17364	0.4951	0.7877	0.8690	0.7877	0.8154
0.3912	13.0	18811	0.4746	0.8017	0.8705	0.8017	0.8257
0.3931	14.0	20258	0.4680	0.8029	0.8719	0.8029	0.8269
0.3887	15.0	21705	0.4710	0.8089	0.8716	0.8089	0.8310
0.3853	16.0	23152	0.4808	0.8008	0.8710	0.8008	0.8252
0.3782	17.0	24599	0.4706	0.8110	0.8704	0.8110	0.8323
0.3842	18.0	26046	0.4832	0.8047	0.8716	0.8047	0.8281
0.3836	19.0	27493	0.4867	0.8026	0.8714	0.8026	0.8266
0.3761	20.0	28940	0.4663	0.8140	0.8706	0.8140	0.8344
0.3677	21.0	30387	0.4654	0.8158	0.8707	0.8158	0.8357
0.3719	22.0	31834	0.4866	0.8083	0.8711	0.8083	0.8305
0.3661	23.0	33281	0.4782	0.8089	0.8716	0.8089	0.8310
0.3743	24.0	34728	0.4919	0.8041	0.8711	0.8041	0.8275
0.3636	25.0	36175	0.4779	0.8092	0.8717	0.8092	0.8313
0.3696	26.0	37622	0.4914	0.8086	0.8716	0.8086	0.8308
0.3659	27.0	39069	0.4944	0.8068	0.8711	0.8068	0.8295
0.368	28.0	40516	0.4873	0.8080	0.8710	0.8080	0.8303
0.3618	29.0	41963	0.4806	0.8104	0.8710	0.8104	0.8320
0.364	30.0	43410	0.4819	0.8101	0.8716	0.8101	0.8319

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu118
• Datasets 3.2.0
• Tokenizers 0.21.0