adriansanz/test23042024_v2

stocks

This model is a fine-tuned version of projecte-aina/roberta-base-ca-v2-cased-te on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7277
• Accuracy: 0.8042
• Precision: 0.8164
• Recall: 0.8082
• F1: 0.8034
• Ratio: 0.5833

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: 4
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.06
• lr_scheduler_warmup_steps: 4
• num_epochs: 2
• label_smoothing_factor: 0.1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	Ratio
3.4817	0.1429	10	1.8123	0.55	0.5523	0.552	0.5497	0.5458
1.3575	0.2857	20	1.3976	0.5542	0.5525	0.5483	0.5421	0.3583
1.0748	0.4286	30	0.9633	0.5917	0.5908	0.5906	0.5906	0.4708
0.878	0.5714	40	0.8403	0.6583	0.6646	0.6616	0.6575	0.5708
0.8718	0.7143	50	0.7818	0.7125	0.7277	0.7174	0.7103	0.6083
0.7897	0.8571	60	0.7231	0.7833	0.7921	0.7868	0.7828	0.5708
0.6933	1.0	70	0.7606	0.7917	0.8026	0.7955	0.7910	0.5792
0.6953	1.1429	80	0.7457	0.7875	0.7939	0.7904	0.7872	0.5583
0.5491	1.2857	90	0.7324	0.7917	0.7939	0.7934	0.7917	0.5292
0.5106	1.4286	100	0.7595	0.8042	0.8093	0.8068	0.8040	0.55
0.7078	1.5714	110	0.7645	0.8042	0.8108	0.8071	0.8039	0.5583
0.7572	1.7143	120	0.7544	0.8	0.8111	0.8038	0.7993	0.5792
0.6879	1.8571	130	0.7307	0.8042	0.8164	0.8082	0.8034	0.5833
0.6025	2.0	140	0.7277	0.8042	0.8164	0.8082	0.8034	0.5833

Framework versions

• Transformers 4.40.0
• Pytorch 2.1.0+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1