GenZ-mental-health-toxic-content-classification-v2

This model is a fine-tuned version of vinai/phobert-base-v2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.7329
• Accuracy: 0.8816
• F1: 0.8088

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.2558	200	0.3588	0.8499	0.7643
No log	0.5115	400	0.3447	0.8649	0.7834
No log	0.7673	600	0.3380	0.8740	0.7832
0.3661	1.0230	800	0.3297	0.8771	0.8042
0.3661	1.2788	1000	0.2997	0.8829	0.8085
0.3661	1.5345	1200	0.3038	0.8841	0.7980
0.3661	1.7903	1400	0.3058	0.8859	0.8110
0.2685	2.0460	1600	0.3259	0.8867	0.8171
0.2685	2.3018	1800	0.3194	0.8861	0.8086
0.2685	2.5575	2000	0.3328	0.8792	0.8137
0.2685	2.8133	2200	0.3462	0.8823	0.7894
0.2176	3.0691	2400	0.4151	0.8867	0.8209
0.2176	3.3248	2600	0.3727	0.8830	0.8129
0.2176	3.5806	2800	0.3754	0.8807	0.8126
0.2176	3.8363	3000	0.3496	0.8905	0.8229
0.1825	4.0921	3200	0.4204	0.8839	0.8058
0.1825	4.3478	3400	0.4341	0.8890	0.8129
0.1825	4.6036	3600	0.3610	0.8908	0.8166
0.1825	4.8593	3800	0.3546	0.8883	0.8143
0.1592	5.1151	4000	0.4929	0.8740	0.8073
0.1592	5.3708	4200	0.4861	0.8865	0.8029
0.1592	5.6266	4400	0.4340	0.8856	0.8162
0.1592	5.8824	4600	0.4357	0.8724	0.8038
0.1375	6.1381	4800	0.4334	0.8885	0.8157
0.1375	6.3939	5000	0.4724	0.8760	0.8083
0.1375	6.6496	5200	0.4504	0.8899	0.8162
0.1375	6.9054	5400	0.3867	0.8854	0.8096
0.1255	7.1611	5600	0.5133	0.8756	0.8070
0.1255	7.4169	5800	0.4806	0.8883	0.8163
0.1255	7.6726	6000	0.4748	0.8816	0.8114
0.1255	7.9284	6200	0.5101	0.8803	0.8084
0.115	8.1841	6400	0.5017	0.8832	0.8068
0.115	8.4399	6600	0.4820	0.8823	0.8041
0.115	8.6957	6800	0.5131	0.8865	0.8089
0.115	8.9514	7000	0.4742	0.8858	0.8145
0.1005	9.2072	7200	0.5905	0.8870	0.8108
0.1005	9.4629	7400	0.5393	0.8796	0.8067
0.1005	9.7187	7600	0.5595	0.8776	0.8077
0.1005	9.9744	7800	0.5101	0.8859	0.8079
0.0918	10.2302	8000	0.6249	0.8781	0.8067
0.0918	10.4859	8200	0.5490	0.8825	0.8077
0.0918	10.7417	8400	0.5394	0.8769	0.8040
0.0818	10.9974	8600	0.6048	0.8807	0.8099
0.0818	11.2532	8800	0.5951	0.8745	0.8011
0.0818	11.5090	9000	0.6220	0.8819	0.8077
0.0818	11.7647	9200	0.6505	0.8785	0.8063
0.078	12.0205	9400	0.6327	0.8785	0.8048
0.078	12.2762	9600	0.6260	0.8796	0.8084
0.078	12.5320	9800	0.5645	0.8800	0.8075
0.078	12.7877	10000	0.6264	0.8823	0.8083
0.071	13.0435	10200	0.6611	0.8798	0.8118
0.071	13.2992	10400	0.6474	0.8845	0.8125
0.071	13.5550	10600	0.6508	0.8819	0.8125
0.071	13.8107	10800	0.6394	0.8823	0.8089
0.0652	14.0665	11000	0.6261	0.8783	0.8069
0.0652	14.3223	11200	0.6541	0.8809	0.8070
0.0652	14.5780	11400	0.7019	0.8778	0.8088
0.0652	14.8338	11600	0.6469	0.8830	0.8091
0.0606	15.0895	11800	0.7078	0.8767	0.8049
0.0606	15.3453	12000	0.6889	0.8809	0.8070
0.0606	15.6010	12200	0.7316	0.8787	0.8090
0.0606	15.8568	12400	0.6827	0.8801	0.8033
0.0575	16.1125	12600	0.7547	0.8812	0.8094
0.0575	16.3683	12800	0.7358	0.8825	0.8055
0.0575	16.6240	13000	0.7128	0.8794	0.8045
0.0575	16.8798	13200	0.7322	0.8818	0.8061
0.054	17.1355	13400	0.7335	0.8814	0.8072
0.054	17.3913	13600	0.7275	0.8818	0.8058
0.054	17.6471	13800	0.7316	0.8810	0.8063
0.054	17.9028	14000	0.7090	0.8823	0.8056
0.052	18.1586	14200	0.7444	0.8781	0.8051
0.052	18.4143	14400	0.7201	0.8810	0.8066
0.052	18.6701	14600	0.7200	0.8825	0.8103
0.052	18.9258	14800	0.7259	0.8800	0.8073
0.049	19.1816	15000	0.7419	0.8830	0.8085
0.049	19.4373	15200	0.7344	0.8823	0.8085
0.049	19.6931	15400	0.7400	0.8816	0.8085
0.049	19.9488	15600	0.7329	0.8816	0.8088

Framework versions

• Transformers 4.44.0
• Pytorch 2.1.2
• Datasets 2.20.0
• Tokenizers 0.19.1