gokulsrinivasagan/distilbert_base_lda_train_book

distilbert_base_train_book

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

• Loss: 1.2047
• Accuracy: 0.7295

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: 160
• eval_batch_size: 160
• seed: 10
• 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
• lr_scheduler_warmup_steps: 10000
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
5.6039	0.7025	10000	5.4506	0.1653
4.4684	1.4051	20000	3.7450	0.3849
2.3547	2.1076	30000	2.0441	0.5926
2.0785	2.8102	40000	1.7986	0.6309
1.938	3.5127	50000	1.6650	0.6520
1.8476	4.2153	60000	1.5862	0.6648
1.7905	4.9178	70000	1.5281	0.6746
1.74	5.6203	80000	1.4845	0.6815
1.7042	6.3229	90000	1.4543	0.6868
1.6725	7.0254	100000	1.4226	0.6917
1.6516	7.7280	110000	1.4016	0.6957
1.6269	8.4305	120000	1.3791	0.6991
1.6032	9.1331	130000	1.3647	0.7016
1.5903	9.8356	140000	1.3465	0.7051
1.5759	10.5381	150000	1.3326	0.7074
1.5641	11.2407	160000	1.3235	0.7090
1.5487	11.9432	170000	1.3103	0.7110
1.5384	12.6458	180000	1.2964	0.7133
1.527	13.3483	190000	1.2920	0.7144
1.5186	14.0509	200000	1.2808	0.7160
1.5086	14.7534	210000	1.2729	0.7174
1.4991	15.4560	220000	1.2637	0.7191
1.4936	16.1585	230000	1.2589	0.7198
1.4843	16.8610	240000	1.2534	0.7209
1.4763	17.5636	250000	1.2467	0.7219
1.4701	18.2661	260000	1.2408	0.7230
1.4668	18.9687	270000	1.2353	0.7240
1.458	19.6712	280000	1.2307	0.7249
1.4547	20.3738	290000	1.2251	0.7258
1.4466	21.0763	300000	1.2207	0.7266
1.4446	21.7788	310000	1.2153	0.7275
1.4375	22.4814	320000	1.2119	0.7281
1.4343	23.1839	330000	1.2086	0.7286
1.4325	23.8865	340000	1.2057	0.7293
1.4294	24.5890	350000	1.2024	0.7297

Framework versions

• Transformers 4.46.1
• Pytorch 2.2.0+cu121
• Datasets 3.1.0
• Tokenizers 0.20.1