NLP-at-home

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

• Loss: 1.1836
• F1: 0.8284

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: 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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	F1
1.0833	1.0	24	0.6880	0.6592
0.5737	2.0	48	0.5213	0.8235
0.3165	3.0	72	0.5841	0.7939
0.2087	4.0	96	0.6425	0.7846
0.1377	5.0	120	0.7989	0.7536
0.0835	6.0	144	0.7516	0.8257
0.0963	7.0	168	0.7099	0.8243
0.0558	8.0	192	0.7994	0.7905
0.0389	9.0	216	0.7475	0.8494
0.057	10.0	240	0.7741	0.8265
0.0435	11.0	264	1.1178	0.7826
0.0458	12.0	288	0.9156	0.7847
0.033	13.0	312	0.9355	0.8198
0.0556	14.0	336	0.9090	0.8152
0.0139	15.0	360	0.9933	0.7950
0.0137	16.0	384	1.1050	0.7889
0.0173	17.0	408	1.1387	0.7836
0.0221	18.0	432	1.0860	0.8072
0.0206	19.0	456	1.0689	0.8120
0.0125	20.0	480	0.9927	0.8087
0.0034	21.0	504	1.1213	0.7919
0.0052	22.0	528	1.2096	0.7969
0.0005	23.0	552	1.1544	0.8132
0.0004	24.0	576	1.1947	0.8143
0.0017	25.0	600	1.2692	0.7957
0.0003	26.0	624	1.2705	0.7981
0.0004	27.0	648	1.4028	0.7617
0.0003	28.0	672	1.2183	0.8193
0.0038	29.0	696	1.2414	0.8106
0.0002	30.0	720	1.3012	0.8022
0.0008	31.0	744	1.1945	0.8120
0.0002	32.0	768	1.1859	0.8125
0.0002	33.0	792	1.1988	0.8078
0.0004	34.0	816	1.2846	0.8144
0.0016	35.0	840	1.2518	0.8121
0.0002	36.0	864	1.2062	0.8254
0.0002	37.0	888	1.2049	0.8197
0.0002	38.0	912	1.2056	0.8254
0.0001	39.0	936	1.2062	0.8254
0.0001	40.0	960	1.1666	0.8378
0.0001	41.0	984	1.1612	0.8378
0.0001	42.0	1008	1.1614	0.8378
0.0001	43.0	1032	1.1739	0.8277
0.0001	44.0	1056	1.1778	0.8277
0.0001	45.0	1080	1.1808	0.8284
0.0001	46.0	1104	1.1816	0.8284
0.0001	47.0	1128	1.1825	0.8284
0.0001	48.0	1152	1.1833	0.8284
0.0001	49.0	1176	1.1836	0.8284
0.0001	50.0	1200	1.1836	0.8284

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.18.0
• Tokenizers 0.19.1