vijil-bias-detector-v5

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

• Loss: 0.2223
• Accuracy: 0.9026
• F1: 0.9073
• Precision: 0.8943
• Recall: 0.9207

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: 1e-05
• train_batch_size: 32
• eval_batch_size: 64
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 0.1
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
1.0866	0.1610	100	1.0051	0.5906	0.5164	0.6646	0.4222
0.9161	0.3221	200	0.8349	0.6932	0.6872	0.7278	0.6509
0.6550	0.4831	300	0.5942	0.7331	0.7702	0.6948	0.8639
0.5327	0.6441	400	0.5399	0.7246	0.7005	0.8016	0.6221
0.4826	0.8052	500	0.4923	0.7085	0.6506	0.8575	0.5241
0.4280	0.9662	600	0.4257	0.7637	0.7957	0.7202	0.8888
0.3756	1.1272	700	0.3708	0.7746	0.7794	0.7899	0.7691
0.3532	1.2882	800	0.3466	0.7810	0.7870	0.7926	0.7815
0.3614	1.4493	900	0.4325	0.7291	0.6722	0.8996	0.5365
0.3559	1.6103	1000	0.3456	0.7721	0.7731	0.7980	0.7496
0.3307	1.7713	1100	0.3258	0.8144	0.8273	0.7983	0.8585
0.3022	1.9324	1200	0.2971	0.8281	0.8408	0.8074	0.8771
0.2884	2.0934	1300	0.2958	0.8362	0.8393	0.8524	0.8266
0.3009	2.2544	1400	0.2928	0.8325	0.8411	0.8266	0.8561
0.2763	2.4155	1500	0.2940	0.8321	0.8333	0.8576	0.8103
0.2700	2.5765	1600	0.2985	0.8092	0.8241	0.7884	0.8631
0.2703	2.7375	1700	0.2720	0.8414	0.8492	0.8363	0.8624
0.2763	2.8986	1800	0.2551	0.8579	0.8607	0.8735	0.8484
0.2272	3.0596	1900	0.2455	0.8724	0.8802	0.8565	0.9051
0.2185	3.2206	2000	0.2309	0.8792	0.8867	0.8620	0.9129
0.1805	3.3816	2100	0.2335	0.8792	0.8902	0.8409	0.9456
0.1908	3.5427	2200	0.2063	0.8965	0.9018	0.8866	0.9176
0.1835	3.7037	2300	0.2301	0.8756	0.8762	0.9041	0.8499
0.1828	3.8647	2400	0.2217	0.8893	0.8977	0.8608	0.9378
0.1746	4.0258	2500	0.2087	0.8933	0.9020	0.8603	0.9479
0.1557	4.1868	2600	0.2168	0.8909	0.8939	0.9006	0.8872
0.1630	4.3478	2700	0.2032	0.8981	0.9037	0.8852	0.9230
0.1592	4.5089	2800	0.2131	0.8849	0.8869	0.9026	0.8717
0.1607	4.6699	2900	0.2305	0.8724	0.8711	0.9130	0.8328
0.1413	4.8309	3000	0.2142	0.8965	0.9033	0.8753	0.9331
0.1693	4.9919	3100	0.2149	0.8945	0.9029	0.8626	0.9471
0.1129	5.1530	3200	0.2177	0.8937	0.8974	0.8974	0.8974
0.1257	5.3140	3300	0.2158	0.9014	0.9082	0.8764	0.9425
0.1484	5.4750	3400	0.2174	0.8990	0.9067	0.8689	0.9479
0.1352	5.6361	3500	0.2116	0.8905	0.8942	0.8949	0.8935
0.1394	5.7971	3600	0.2131	0.8981	0.9067	0.8625	0.9557
0.1322	5.9581	3700	0.2072	0.8969	0.9006	0.8992	0.9020
0.1021	6.1192	3800	0.2119	0.8994	0.9044	0.8901	0.9191
0.1003	6.2802	3900	0.2146	0.9010	0.9057	0.8933	0.9184
0.1162	6.4412	4000	0.2158	0.9026	0.9090	0.8805	0.9393
0.1013	6.6023	4100	0.2141	0.9006	0.9050	0.8951	0.9152
0.1013	6.7633	4200	0.2214	0.8998	0.9059	0.8810	0.9323
0.1179	6.9243	4300	0.2279	0.8953	0.8967	0.9171	0.8771
0.0795	7.0853	4400	0.2296	0.8986	0.9009	0.9116	0.8904
0.0747	7.2464	4500	0.2315	0.9038	0.9077	0.9018	0.9137
0.0811	7.4074	4600	0.2297	0.8969	0.9009	0.8968	0.9051
0.0824	7.5684	4700	0.2237	0.9030	0.9073	0.8973	0.9176
0.0844	7.7295	4800	0.2223	0.9026	0.9073	0.8943	0.9207

Framework versions

• Transformers 5.5.0
• Pytorch 2.11.0+cu130
• Datasets 4.8.4
• Tokenizers 0.22.2