distilroberta-finetuned-bloomberg-classifier

This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2983

Model description

DistilRoberta base finetuned on Bloomberg financial news.

Intended uses & limitations

Intended for analysis of financial articles and press releases in relation to sentiment of market

Training and evaluation data

Fine tuned on ~12,500 Bloomberg financial articles

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss
0.3714	1.0	313	0.3035
0.4374	2.0	626	0.2983
0.2918	3.0	939	0.5378
0.1115	4.0	1252	0.6534
0.0002	5.0	1565	0.7559

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.1+cpu
• Datasets 2.20.0
• Tokenizers 0.19.1