distilbert_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: 3.5714
• Accuracy: 0.7290

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
7.2789	0.4215	10000	6.5733	0.3678
4.955	0.8431	20000	4.5929	0.5844
4.6182	1.2646	30000	4.3014	0.6199
4.4751	1.6861	40000	4.1692	0.6387
4.3802	2.1077	50000	4.0856	0.6498
4.3162	2.5292	60000	4.0203	0.6598
4.2637	2.9507	70000	3.9798	0.6658
4.225	3.3723	80000	3.9419	0.6716
4.1939	3.7938	90000	3.9101	0.6761
4.1637	4.2153	100000	3.8892	0.6795
4.1428	4.6369	110000	3.8674	0.6830
4.1204	5.0584	120000	3.8432	0.6860
4.1028	5.4799	130000	3.8268	0.6887
4.091	5.9014	140000	3.8095	0.6912
4.0686	6.3230	150000	3.7973	0.6936
4.0613	6.7445	160000	3.7854	0.6952
4.0375	7.1660	170000	3.7712	0.6976
4.0301	7.5876	180000	3.7640	0.6984
4.0202	8.0091	190000	3.7519	0.7002
4.0102	8.4306	200000	3.7429	0.7018
3.9985	8.8522	210000	3.7357	0.7031
3.9959	9.2737	220000	3.7270	0.7044
3.9781	9.6952	230000	3.7161	0.7058
3.9676	10.1168	240000	3.7122	0.7069
3.9672	10.5383	250000	3.7028	0.7080
3.9573	10.9598	260000	3.6969	0.7093
3.9497	11.3814	270000	3.6923	0.7098
3.9483	11.8029	280000	3.6841	0.7111
3.9348	12.2244	290000	3.6786	0.7119
3.9304	12.6460	300000	3.6703	0.7132
3.9205	13.0675	310000	3.6666	0.7139
3.9202	13.4890	320000	3.6620	0.7146
3.9152	13.9106	330000	3.6581	0.7152
3.9105	14.3321	340000	3.6501	0.7167
3.9024	14.7536	350000	3.6466	0.7168
3.8976	15.1751	360000	3.6444	0.7173
3.8948	15.5967	370000	3.6362	0.7186
3.8875	16.0182	380000	3.6316	0.7193
3.8842	16.4397	390000	3.6280	0.7199
3.8773	16.8613	400000	3.6256	0.7203
3.8766	17.2828	410000	3.6190	0.7214
3.8712	17.7043	420000	3.6155	0.7220
3.8638	18.1259	430000	3.6134	0.7225
3.8588	18.5474	440000	3.6109	0.7228
3.8586	18.9689	450000	3.6057	0.7237
3.8534	19.3905	460000	3.6018	0.7241
3.8508	19.8120	470000	3.5988	0.7247
3.8443	20.2335	480000	3.5953	0.7252
3.8393	20.6551	490000	3.5921	0.7257
3.8366	21.0766	500000	3.5886	0.7260
3.8324	21.4981	510000	3.5857	0.7266
3.8286	21.9197	520000	3.5843	0.7269
3.8317	22.3412	530000	3.5818	0.7273
3.8243	22.7627	540000	3.5811	0.7276
3.825	23.1843	550000	3.5771	0.7279
3.8227	23.6058	560000	3.5752	0.7284
3.8173	24.0273	570000	3.5741	0.7286
3.8174	24.4488	580000	3.5706	0.7291
3.8141	24.8704	590000	3.5708	0.7289

Framework versions

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