twitter-xlm-roberta-base-sentiment-finetunned-davincis-local

This model is a fine-tuned version of citizenlab/twitter-xlm-roberta-base-sentiment-finetunned on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5461
• Accuracy: 0.9302
• F1: 0.9301

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: 72
• eval_batch_size: 72
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.4006	1.0	41	0.3037	0.8779	0.8771
0.2165	2.0	82	0.2007	0.9205	0.9205
0.1311	3.0	123	0.2124	0.9244	0.9244
0.0839	4.0	164	0.2504	0.9341	0.9341
0.0525	5.0	205	0.3695	0.9147	0.9144
0.0392	6.0	246	0.3393	0.9244	0.9243
0.0282	7.0	287	0.4203	0.9244	0.9242
0.0205	8.0	328	0.3889	0.9302	0.9301
0.012	9.0	369	0.6586	0.9012	0.9006
0.0069	10.0	410	0.4873	0.9302	0.9301
0.005	11.0	451	0.6105	0.9089	0.9085
0.0082	12.0	492	0.4642	0.9302	0.9301
0.0022	13.0	533	0.3709	0.9516	0.9515
0.0088	14.0	574	0.5322	0.9283	0.9281
0.0067	15.0	615	0.6661	0.9128	0.9124
0.0015	16.0	656	0.5450	0.9283	0.9282
0.0006	17.0	697	0.5453	0.9302	0.9301
0.0002	18.0	738	0.5555	0.9302	0.9301
0.0018	19.0	779	0.5408	0.9302	0.9301
0.0022	20.0	820	0.5461	0.9302	0.9301

Framework versions

• Transformers 4.27.4
• Pytorch 1.13.1+cu116
• Datasets 2.11.0
• Tokenizers 0.13.2