slac-palate

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

• Loss: 1.9763
• Accuracy: 0.9463
• F1 Macro: 0.8328
• Precision Macro: 0.8463
• Recall Macro: 0.8206
• Total Tf: [1516, 86, 1516, 86]

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: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 212
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro	Precision Macro	Recall Macro	Total Tf
0.7327	1.0	213	0.6511	0.7609	0.6358	0.6324	0.8470	[1219, 383, 1219, 383]
0.593	2.0	426	0.4161	0.9051	0.7883	0.7389	0.8958	[1450, 152, 1450, 152]
0.371	3.0	639	0.4254	0.9413	0.8492	0.8085	0.9096	[1508, 94, 1508, 94]
0.3035	4.0	852	0.5906	0.9419	0.8372	0.8172	0.8610	[1509, 93, 1509, 93]
0.2537	5.0	1065	0.7122	0.9426	0.8403	0.8179	0.8675	[1510, 92, 1510, 92]
0.1256	6.0	1278	0.9292	0.9370	0.8232	0.8040	0.8460	[1501, 101, 1501, 101]
0.0757	7.0	1491	1.1844	0.9351	0.8164	0.7997	0.8358	[1498, 104, 1498, 104]
0.0609	8.0	1704	1.5595	0.9363	0.8113	0.8076	0.8151	[1500, 102, 1500, 102]
0.037	9.0	1917	1.6620	0.9438	0.8283	0.8346	0.8223	[1512, 90, 1512, 90]
0.0599	10.0	2130	1.8541	0.9426	0.8256	0.8297	0.8216	[1510, 92, 1510, 92]
0.0381	11.0	2343	1.9677	0.9469	0.8331	0.8504	0.8179	[1517, 85, 1517, 85]
0.0313	12.0	2556	2.0030	0.9463	0.8317	0.8476	0.8176	[1516, 86, 1516, 86]
0.0151	13.0	2769	1.8765	0.9426	0.8256	0.8297	0.8216	[1510, 92, 1510, 92]
0.0227	14.0	2982	1.8954	0.9432	0.8259	0.8332	0.8189	[1511, 91, 1511, 91]
0.0031	15.0	3195	1.9763	0.9463	0.8328	0.8463	0.8206	[1516, 86, 1516, 86]

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0
• Datasets 2.21.0
• Tokenizers 0.19.1