sentiment_pc_oversampler

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: 0.3909
• Accuracy: 0.9291
• F1: 0.9288

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	0.1134	50	0.5293	0.8154	0.8173
No log	0.2268	100	0.4512	0.8222	0.8224
No log	0.3401	150	0.4212	0.8356	0.8364
No log	0.4535	200	0.3978	0.8395	0.8400
No log	0.5669	250	0.3745	0.8631	0.8642
No log	0.6803	300	0.3593	0.8667	0.8675
No log	0.7937	350	0.3203	0.8821	0.8826
No log	0.9070	400	0.3130	0.8880	0.8889
No log	1.0204	450	0.3052	0.8903	0.8904
0.3514	1.1338	500	0.3216	0.8948	0.8954
0.3514	1.2472	550	0.3178	0.8979	0.8981
0.3514	1.3605	600	0.3366	0.8874	0.8877
0.3514	1.4739	650	0.3108	0.8951	0.8950
0.3514	1.5873	700	0.2551	0.9198	0.9200
0.3514	1.7007	750	0.3358	0.8911	0.8907
0.3514	1.8141	800	0.2812	0.9127	0.9125
0.3514	1.9274	850	0.2443	0.9240	0.9239
0.3514	2.0408	900	0.3059	0.9183	0.9182
0.3514	2.1542	950	0.3161	0.9155	0.9152
0.1587	2.2676	1000	0.2733	0.9237	0.9235
0.1587	2.3810	1050	0.3252	0.9141	0.9137
0.1587	2.4943	1100	0.3257	0.9141	0.9140
0.1587	2.6077	1150	0.2836	0.9254	0.9253
0.1587	2.7211	1200	0.3176	0.9166	0.9163
0.1587	2.8345	1250	0.3335	0.9232	0.9228
0.1587	2.9478	1300	0.3076	0.9257	0.9254
0.1587	3.0612	1350	0.3169	0.9269	0.9264
0.1587	3.1746	1400	0.3627	0.9240	0.9238
0.1587	3.2880	1450	0.4074	0.9127	0.9118
0.0731	3.4014	1500	0.3580	0.9251	0.9247
0.0731	3.5147	1550	0.3802	0.9240	0.9235
0.0731	3.6281	1600	0.3705	0.9257	0.9253
0.0731	3.7415	1650	0.3177	0.9362	0.9361
0.0731	3.8549	1700	0.3563	0.9314	0.9310
0.0731	3.9683	1750	0.4248	0.9158	0.9154
0.0731	4.0816	1800	0.3535	0.9314	0.9310
0.0731	4.1950	1850	0.3568	0.9308	0.9305
0.0731	4.3084	1900	0.4044	0.9266	0.9264
0.0731	4.4218	1950	0.3598	0.9331	0.9327
0.0358	4.5351	2000	0.3909	0.9291	0.9288
0.0358	4.6485	2050	0.3725	0.9325	0.9322
0.0358	4.7619	2100	0.3953	0.9305	0.9303
0.0358	4.8753	2150	0.3902	0.9305	0.9302
0.0358	4.9887	2200	0.3960	0.9286	0.9282

Framework versions

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