Modelpublic

mr7371/distilbert-financial-phrasebank-allagree_LoRA_r1

sourceHugging Faceapache-2.0updated 10mo agoView on Hugging Face

0likes13downloads

Model Card

distilbert-financial-phrasebank-allagreeLoRAr1

This model is a fine-tuned version of mr7371/distilbert-financial-phrasebank-allagree_headonly on an unknown dataset. It achieves the following results on the evaluation set:

—Train Loss: 0.0628
—Train Accuracy: 0.9773
—Validation Loss: 0.1312
—Validation Accuracy: 0.9529
—Epoch: 20

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:

—optimizer: {'name': 'Adam', 'weightdecay': None, 'clipnorm': None, 'globalclipnorm': None, 'clipvalue': None, 'useema': False, 'emamomentum': 0.99, 'emaoverwritefrequency': None, 'jitcompile': True, 'islegacyoptimizer': False, 'learningrate': {'module': None, 'classname': 'CustomSchedule', 'config': {'dmodel': 768, 'warmupsteps': 250, 'maxrate': 0.0001}, 'registeredname': 'CustomSchedule'}, 'beta1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
—training_precision: float32

Training results

Train Loss	Train Accuracy	Validation Loss	Validation Accuracy	Epoch
0.2642	0.8883	0.2903	0.8500	0
0.2668	0.8857	0.2872	0.8500	1
0.2583	0.8990	0.2817	0.8588	2
0.2506	0.8958	0.2686	0.8559	3
0.2312	0.8958	0.2533	0.8794	4
0.2103	0.9085	0.2208	0.9029	5
0.1802	0.9287	0.1921	0.9147	6
0.1655	0.9306	0.1714	0.9294	7
0.1451	0.9426	0.1591	0.9353	8
0.1308	0.9463	0.1601	0.9324	9
0.1032	0.9596	0.1524	0.9412	10
0.1084	0.9621	0.1493	0.9471	11
0.1029	0.9615	0.1439	0.9471	12
0.0861	0.9665	0.1538	0.9441	13
0.0863	0.9716	0.1392	0.9471	14
0.0839	0.9691	0.1381	0.9529	15
0.0732	0.9747	0.1335	0.9529	16
0.0756	0.9729	0.1317	0.9500	17
0.0622	0.9747	0.1273	0.9529	18
0.0651	0.9785	0.1327	0.9559	19
0.0628	0.9773	0.1312	0.9529	20

Framework versions

—Transformers 4.57.2
—TensorFlow 2.19.0
—Datasets 3.6.0
—Tokenizers 0.22.1