beto-sentiment-analysis-finetuned-ner

This model is a fine-tuned version of finiteautomata/beto-sentiment-analysis on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.9250
• Precision: 0.5603
• Recall: 0.6436
• F1: 0.5991
• Accuracy: 0.9863

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: 8e-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: 24

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
1.4102	1.0	3	1.2732	0.0455	0.0198	0.0276	0.9723
0.7776	2.0	6	0.9025	0.1056	0.1485	0.1235	0.9663
0.6861	3.0	9	0.7874	0.1176	0.1980	0.1476	0.9694
0.2837	4.0	12	0.8528	0.1067	0.2376	0.1472	0.9534
0.3182	5.0	15	0.7798	0.2360	0.3762	0.2901	0.9729
0.1673	6.0	18	0.8645	0.1461	0.2574	0.1864	0.9604
0.2065	7.0	21	0.8130	0.2941	0.5446	0.3819	0.9765
0.0794	8.0	24	0.6841	0.4276	0.6139	0.5041	0.9822
0.0543	9.0	27	0.7113	0.4104	0.5446	0.4681	0.9815
0.0278	10.0	30	0.7865	0.4565	0.6238	0.5272	0.9833
0.0598	11.0	33	0.8356	0.4155	0.5842	0.4856	0.9824
0.0108	12.0	36	0.8104	0.4460	0.6139	0.5167	0.9826
0.0235	13.0	39	0.7986	0.5194	0.6634	0.5826	0.9844
0.0134	14.0	42	0.8175	0.6182	0.6733	0.6445	0.9865
0.0124	15.0	45	0.8575	0.6036	0.6634	0.6321	0.9875
0.0049	16.0	48	0.8822	0.6019	0.6436	0.6220	0.9871
0.0097	17.0	51	0.8696	0.5556	0.6436	0.5963	0.9862
0.0067	18.0	54	0.8728	0.5410	0.6535	0.5919	0.9859
0.0045	19.0	57	0.8807	0.5159	0.6436	0.5727	0.9848
0.004	20.0	60	0.8938	0.52	0.6436	0.5752	0.9851
0.0038	21.0	63	0.9108	0.5203	0.6337	0.5714	0.9852
0.004	22.0	66	0.9243	0.5702	0.6436	0.6047	0.9864
0.0106	23.0	69	0.9261	0.5702	0.6436	0.6047	0.9865
0.004	24.0	72	0.9250	0.5603	0.6436	0.5991	0.9863

Framework versions

• Transformers 4.22.2
• Pytorch 1.12.1+cu113
• Datasets 2.5.2
• Tokenizers 0.12.1