minoosh/finetuned_roberta-base_on_shEMO_transcripts

finetuned_roberta-base

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

• Loss: 1.2574
• Accuracy: 0.6033

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.48	1.0	75	1.4001	0.41
1.2847	2.0	150	1.1993	0.58
1.1522	3.0	225	1.0007	0.6333
0.9921	4.0	300	0.9189	0.66
0.9104	5.0	375	0.8855	0.69
0.8371	6.0	450	0.9431	0.6767
0.699	7.0	525	0.9500	0.6633
0.6872	8.0	600	0.9728	0.7033
0.5867	9.0	675	0.9939	0.6867
0.5323	10.0	750	1.1115	0.69
0.4066	11.0	825	1.2031	0.6667
0.3517	12.0	900	1.2193	0.65
0.3114	13.0	975	1.2281	0.67
0.3102	14.0	1050	1.2691	0.67
0.2681	15.0	1125	1.2818	0.6633

Framework versions

• Transformers 4.34.1
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1