sentiment_pc_weightedLoss

This model is a fine-tuned version of ahmedrachid/FinancialBERT-Sentiment-Analysis on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6463
• Accuracy: 0.86
• F1: 0.8290

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.1739	50	0.6153	0.8087	0.7818
No log	0.3478	100	0.4938	0.8165	0.7843
No log	0.5217	150	0.4613	0.8339	0.8016
No log	0.6957	200	0.4918	0.7913	0.7619
No log	0.8696	250	0.4520	0.8283	0.7961
No log	1.0435	300	0.4821	0.8339	0.8054
No log	1.2174	350	0.4868	0.8639	0.8327
No log	1.3913	400	0.5093	0.8574	0.8259
No log	1.5652	450	0.4648	0.8474	0.8175
0.4528	1.7391	500	0.4556	0.8470	0.8151
0.4528	1.9130	550	0.4747	0.8361	0.8062
0.4528	2.0870	600	0.5520	0.8543	0.8234
0.4528	2.2609	650	0.6130	0.8652	0.8367
0.4528	2.4348	700	0.5657	0.8722	0.8415
0.4528	2.6087	750	0.5357	0.8339	0.8033
0.4528	2.7826	800	0.5729	0.8513	0.8233
0.4528	2.9565	850	0.5304	0.8522	0.8215
0.4528	3.1304	900	0.5982	0.8683	0.8375
0.4528	3.3043	950	0.5684	0.8513	0.8197
0.1978	3.4783	1000	0.6463	0.86	0.8290
0.1978	3.6522	1050	0.6566	0.8565	0.8262
0.1978	3.8261	1100	0.6497	0.8578	0.8282
0.1978	4.0	1150	0.6531	0.8591	0.8266

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1