Merged-Int-praj

This model is a fine-tuned version of prajjwal1/bert-tiny on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1460
• Accuracy: 0.96
• F1: 0.9600

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.0	50	0.6933	0.5	0.3333
No log	0.01	100	0.6929	0.58	0.4900
No log	0.01	150	0.6937	0.5	0.3333
No log	0.01	200	0.6951	0.5	0.3333
No log	0.02	250	0.6902	0.52	0.5130
No log	0.02	300	0.6909	0.5	0.3333
No log	0.02	350	0.6795	0.56	0.4762
No log	0.03	400	0.6524	0.61	0.6010
No log	0.03	450	0.6139	0.71	0.7100
0.6779	0.03	500	0.5827	0.71	0.7033
0.6779	0.04	550	0.5732	0.71	0.7033
0.6779	0.04	600	0.5467	0.74	0.7396
0.6779	0.04	650	0.5174	0.8	0.7980
0.6779	0.05	700	0.5193	0.74	0.7399
0.6779	0.05	750	0.4905	0.8	0.7980
0.6779	0.05	800	0.4710	0.8	0.7980
0.6779	0.06	850	0.4523	0.83	0.8271
0.6779	0.06	900	0.4373	0.84	0.8368
0.6779	0.06	950	0.4214	0.84	0.8368
0.5615	0.07	1000	0.4086	0.84	0.8368
0.5615	0.07	1050	0.3803	0.84	0.8368
0.5615	0.07	1100	0.3476	0.9	0.8994
0.5615	0.08	1150	0.3218	0.91	0.9096
0.5615	0.08	1200	0.3028	0.91	0.9096
0.5615	0.08	1250	0.2851	0.92	0.9195
0.5615	0.09	1300	0.2737	0.92	0.9195
0.5615	0.09	1350	0.2637	0.91	0.9096
0.5615	0.09	1400	0.2560	0.92	0.9195
0.5615	0.1	1450	0.2426	0.92	0.9199
0.4267	0.1	1500	0.2390	0.89	0.8897
0.4267	0.1	1550	0.2320	0.92	0.9199
0.4267	0.11	1600	0.2239	0.93	0.9298
0.4267	0.11	1650	0.2159	0.94	0.9398
0.4267	0.11	1700	0.2156	0.93	0.9298
0.4267	0.12	1750	0.2079	0.93	0.9298
0.4267	0.12	1800	0.1938	0.93	0.9298
0.4267	0.12	1850	0.1909	0.93	0.9298
0.4267	0.13	1900	0.1923	0.93	0.9298
0.4267	0.13	1950	0.1893	0.94	0.9398
0.3491	0.13	2000	0.1633	0.96	0.9600
0.3491	0.14	2050	0.1662	0.95	0.9500
0.3491	0.14	2100	0.1494	0.96	0.9600
0.3491	0.14	2150	0.1606	0.95	0.9499
0.3491	0.15	2200	0.1595	0.96	0.9599
0.3491	0.15	2250	0.1460	0.96	0.9600

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.0
• Tokenizers 0.15.0