hbertv1-mini-wt-48-Massive-intent-emb-comp

This model is a fine-tuned version of gokuls/model_v1_complete_training_wt_init_48_mini_emb_comp on the massive dataset. It achieves the following results on the evaluation set:

• Loss: 0.7077
• Accuracy: 0.8411

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: 5e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 33
• distributed_type: multi-GPU
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.8713	1.0	180	1.8255	0.5903
1.4406	2.0	360	1.1089	0.7177
0.9491	3.0	540	0.8839	0.7727
0.7165	4.0	720	0.7622	0.8072
0.5574	5.0	900	0.7180	0.8121
0.4491	6.0	1080	0.7020	0.8224
0.3617	7.0	1260	0.6915	0.8244
0.291	8.0	1440	0.6727	0.8352
0.2355	9.0	1620	0.6822	0.8362
0.1915	10.0	1800	0.6960	0.8293
0.1569	11.0	1980	0.7021	0.8367
0.1296	12.0	2160	0.7077	0.8411
0.1087	13.0	2340	0.7080	0.8406
0.0931	14.0	2520	0.7152	0.8411
0.0839	15.0	2700	0.7203	0.8401

Framework versions

• Transformers 4.31.0
• Pytorch 1.14.0a0+410ce96
• Datasets 2.13.1
• Tokenizers 0.13.3