absa-bert

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

• Loss: 1.0596
• Accuracy: 0.8281
• F1: 0.7632

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: 256
• eval_batch_size: 256
• seed: 42
• optimizer: Use OptimizerNames.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: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.0575	1.0	12	0.9426	0.5933	0.2483
0.9027	2.0	24	0.8538	0.5952	0.2537
0.8016	3.0	36	0.7493	0.6876	0.4561
0.7066	4.0	48	0.6882	0.7246	0.5049
0.6343	5.0	60	0.6381	0.7394	0.5289
0.5703	6.0	72	0.6112	0.7542	0.6037
0.5229	7.0	84	0.6028	0.7634	0.6459
0.4864	8.0	96	0.6066	0.7634	0.6432
0.4479	9.0	108	0.6055	0.7689	0.6703
0.4154	10.0	120	0.5939	0.7708	0.6736
0.3815	11.0	132	0.6057	0.7708	0.6768
0.3525	12.0	144	0.6229	0.7689	0.6807
0.3229	13.0	156	0.6119	0.7745	0.6874
0.2962	14.0	168	0.6206	0.7874	0.7034
0.2718	15.0	180	0.6303	0.7726	0.6865
0.2448	16.0	192	0.6469	0.7745	0.6858
0.2241	17.0	204	0.6496	0.7745	0.6806
0.1985	18.0	216	0.6656	0.7763	0.6827
0.1784	19.0	228	0.6906	0.7726	0.6771
0.1693	20.0	240	0.6967	0.7856	0.7027
0.148	21.0	252	0.7059	0.7874	0.6945
0.1336	22.0	264	0.7480	0.7837	0.6823
0.1286	23.0	276	0.7424	0.7856	0.6896
0.1166	24.0	288	0.7293	0.7893	0.6970
0.1043	25.0	300	0.7177	0.7985	0.7151
0.0879	26.0	312	0.7184	0.8004	0.7237
0.0872	27.0	324	0.7407	0.7911	0.7121
0.0827	28.0	336	0.7513	0.7967	0.7193
0.0702	29.0	348	0.7797	0.8059	0.7261
0.0618	30.0	360	0.7911	0.8133	0.7378
0.0641	31.0	372	0.7861	0.8115	0.7369
0.051	32.0	384	0.8209	0.8078	0.7249
0.0473	33.0	396	0.7894	0.8133	0.7423
0.0463	34.0	408	0.8047	0.8152	0.7460
0.0447	35.0	420	0.8356	0.8133	0.7389
0.0408	36.0	432	0.8416	0.8078	0.7330
0.0393	37.0	444	0.8392	0.8133	0.7377
0.0356	38.0	456	0.8373	0.8207	0.7511
0.0387	39.0	468	0.8539	0.8096	0.7363
0.0347	40.0	480	0.8908	0.8078	0.7258
0.032	41.0	492	0.8500	0.8170	0.7474
0.0336	42.0	504	0.8515	0.8133	0.7423
0.0336	43.0	516	0.8515	0.8152	0.7407
0.0273	44.0	528	0.8832	0.8244	0.7497
0.0283	45.0	540	0.8701	0.8115	0.7398
0.0242	46.0	552	0.8904	0.8189	0.7487
0.027	47.0	564	0.8972	0.8244	0.7518
0.0275	48.0	576	0.8824	0.8189	0.7529
0.0197	49.0	588	0.9016	0.8152	0.7433
0.0205	50.0	600	0.9236	0.8152	0.7446
0.0235	51.0	612	0.9441	0.8189	0.7473
0.0215	52.0	624	0.9369	0.8207	0.7492
0.0195	53.0	636	0.9186	0.8244	0.7567
0.0155	54.0	648	0.9427	0.8244	0.7564
0.0184	55.0	660	0.9445	0.8226	0.7540
0.0159	56.0	672	0.9560	0.8189	0.7491
0.0172	57.0	684	0.9785	0.8152	0.7386
0.0151	58.0	696	0.9750	0.8152	0.7442
0.0159	59.0	708	0.9631	0.8207	0.7516
0.0126	60.0	720	1.0032	0.8189	0.7471
0.0153	61.0	732	0.9948	0.8170	0.7458
0.018	62.0	744	0.9880	0.8170	0.7494
0.0173	63.0	756	0.9989	0.8207	0.7523
0.0161	64.0	768	1.0125	0.8096	0.7313
0.0165	65.0	780	1.0105	0.8189	0.7474
0.0125	66.0	792	1.0118	0.8133	0.7374
0.0139	67.0	804	0.9968	0.8189	0.7514
0.0112	68.0	816	1.0152	0.8170	0.7442
0.0168	69.0	828	1.0176	0.8189	0.7463
0.0127	70.0	840	1.0108	0.8133	0.7404
0.0124	71.0	852	1.0187	0.8207	0.7487
0.0129	72.0	864	1.0138	0.8207	0.7531
0.0087	73.0	876	1.0301	0.8207	0.7481
0.0112	74.0	888	1.0131	0.8281	0.7629
0.0122	75.0	900	1.0176	0.8299	0.7658
0.0086	76.0	912	1.0301	0.8226	0.7543
0.0103	77.0	924	1.0421	0.8207	0.7527
0.0102	78.0	936	1.0410	0.8189	0.7480
0.0114	79.0	948	1.0451	0.8226	0.7539
0.0102	80.0	960	1.0286	0.8281	0.7632
0.0117	81.0	972	1.0311	0.8262	0.7603
0.0117	82.0	984	1.0386	0.8262	0.7589
0.0136	83.0	996	1.0320	0.8262	0.7603
0.01	84.0	1008	1.0304	0.8299	0.7679
0.0103	85.0	1020	1.0362	0.8262	0.7617
0.0105	86.0	1032	1.0440	0.8262	0.7617
0.0099	87.0	1044	1.0555	0.8244	0.7581
0.0117	88.0	1056	1.0615	0.8244	0.7571
0.0098	89.0	1068	1.0553	0.8262	0.7601
0.0105	90.0	1080	1.0586	0.8262	0.7603
0.0106	91.0	1092	1.0583	0.8262	0.7603
0.01	92.0	1104	1.0558	0.8262	0.7603
0.0105	93.0	1116	1.0589	0.8262	0.7603
0.011	94.0	1128	1.0609	0.8281	0.7632
0.0093	95.0	1140	1.0614	0.8281	0.7632
0.0096	96.0	1152	1.0615	0.8281	0.7632
0.0118	97.0	1164	1.0630	0.8281	0.7632
0.0089	98.0	1176	1.0618	0.8281	0.7632
0.0122	99.0	1188	1.0601	0.8281	0.7632
0.0101	100.0	1200	1.0596	0.8281	0.7632

Framework versions

• Transformers 4.51.1
• Pytorch 2.5.1+cu124
• Datasets 3.5.0
• Tokenizers 0.21.0