metadata

license: apache-2.0
base_model: projecte-aina/roberta-base-ca-v2-cased-te
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: stocks
    results: []
pipeline_tag: zero-shot-classification

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: 0.7709
Accuracy: 0.7875
Precision: 0.5276
Recall: 0.5256
F1: 0.5261
Ratio: 0.5083

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
3.4508	0.1429	10	1.7714	0.5625	0.3772	0.3755	0.3758	0.5167
1.3131	0.2857	20	1.2703	0.6083	0.6073	0.6066	0.6066	0.4542
1.0393	0.4286	30	0.9366	0.6625	0.6623	0.6603	0.6604	0.4417
0.8129	0.5714	40	0.8434	0.7167	0.7179	0.7179	0.7167	0.5208
0.816	0.7143	50	0.9037	0.7042	0.7447	0.7122	0.6961	0.6833
0.7914	0.8571	60	0.7575	0.7583	0.7586	0.7569	0.7573	0.4542
0.7873	1.0	70	0.7795	0.75	0.7709	0.7555	0.7475	0.6208
0.6177	1.1429	80	0.7027	0.7917	0.7914	0.7910	0.7911	0.4708
0.5429	1.2857	90	0.7100	0.7917	0.7915	0.792	0.7915	0.4958
0.5314	1.4286	100	0.7451	0.7875	0.5276	0.5256	0.5261	0.5083
0.5945	1.5714	110	0.7605	0.8	0.5358	0.5324	0.5338	0.4542
0.661	1.7143	120	0.7722	0.7792	0.5215	0.5195	0.5204	0.4917
0.6144	1.8571	130	0.7688	0.7875	0.5273	0.5253	0.5260	0.5
0.5695	2.0	140	0.7709	0.7875	0.5276	0.5256	0.5261	0.5083

Framework versions

Transformers 4.40.0
Pytorch 2.1.0+cu121
Datasets 2.19.0
Tokenizers 0.19.1