bert_base_lda_20_v1_book

This model is a fine-tuned version of on the gokulsrinivasagan/processed_book_corpus-ld-20 dataset. It achieves the following results on the evaluation set:

• Loss: 3.4102
• Accuracy: 0.7596

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: 96
• eval_batch_size: 96
• seed: 10
• 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
• lr_scheduler_warmup_steps: 10000
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
8.1966	0.4215	10000	8.0049	0.1640
8.0473	0.8431	20000	7.8939	0.1656
6.9645	1.2646	30000	6.4681	0.3352
4.9295	1.6861	40000	4.5709	0.5785
4.5377	2.1077	50000	4.2350	0.6289
4.3734	2.5292	60000	4.0854	0.6527
4.2676	2.9507	70000	3.9894	0.6667
4.1937	3.3723	80000	3.9218	0.6772
4.1383	3.7938	90000	3.8696	0.6855
4.0903	4.2153	100000	3.8305	0.6918
4.0556	4.6369	110000	3.7986	0.6968
4.0202	5.0584	120000	3.7627	0.7016
3.9954	5.4799	130000	3.7404	0.7054
3.9765	5.9014	140000	3.7157	0.7093
3.9475	6.3230	150000	3.6981	0.7121
3.934	6.7445	160000	3.6797	0.7152
3.9051	7.1660	170000	3.6612	0.7182
3.8964	7.5876	180000	3.6575	0.7176
3.8816	8.0091	190000	3.6411	0.7198
3.866	8.4306	200000	3.6233	0.7243
3.8514	8.8522	210000	3.6129	0.7259
3.8458	9.2737	220000	3.6029	0.7278
3.8295	9.6952	230000	3.5904	0.7293
3.8176	10.1168	240000	3.5856	0.7307
3.8131	10.5383	250000	3.5751	0.7319
3.7999	10.9598	260000	3.5652	0.7337
3.7893	11.3814	270000	3.5577	0.7348
3.7844	11.8029	280000	3.5480	0.7366
3.7698	12.2244	290000	3.5411	0.7377
3.7647	12.6460	300000	3.5322	0.7390
3.752	13.0675	310000	3.5254	0.7403
3.7505	13.4890	320000	3.5207	0.7412
3.7444	13.9106	330000	3.5145	0.7421
3.737	14.3321	340000	3.5068	0.7435
3.728	14.7536	350000	3.5010	0.7442
3.7219	15.1751	360000	3.4959	0.7450
3.7175	15.5967	370000	3.4885	0.7461
3.7095	16.0182	380000	3.4826	0.7472
3.7041	16.4397	390000	3.4774	0.7481
3.6965	16.8613	400000	3.4750	0.7488
3.6936	17.2828	410000	3.4675	0.7496
3.6885	17.7043	420000	3.4642	0.7505
3.679	18.1259	430000	3.4608	0.7512
3.6735	18.5474	440000	3.4564	0.7517
3.6726	18.9689	450000	3.4499	0.7528
3.6668	19.3905	460000	3.4468	0.7534
3.663	19.8120	470000	3.4422	0.7541
3.656	20.2335	480000	3.4387	0.7546
3.6499	20.6551	490000	3.4352	0.7554
3.6462	21.0766	500000	3.4307	0.7559
3.6423	21.4981	510000	3.4273	0.7566
3.6374	21.9197	520000	3.4252	0.7569
3.6389	22.3412	530000	3.4225	0.7574
3.6316	22.7627	540000	3.4215	0.7577
3.6323	23.1843	550000	3.4175	0.7582
3.6287	23.6058	560000	3.4146	0.7589
3.623	24.0273	570000	3.4132	0.7591
3.6224	24.4488	580000	3.4097	0.7595
3.6193	24.8704	590000	3.4098	0.7595

Framework versions

• Transformers 4.46.3
• Pytorch 2.2.1+cu118
• Datasets 2.17.0
• Tokenizers 0.20.3