Modelpublic

Kudod/Cognitive-Distortions-Llama-3.1-8B-v2

sourceHugging Facellama3.1updated 9mo agoView on Hugging Face

0likes7downloads

Model Card

Cognitive-Distortions-Llama-3.1-8B-v2

This model is a fine-tuned version of meta-llama/Llama-3.1-8B on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 1.8643
—Precision: 0.5529
—Recall: 0.5456
—F1: 0.5411

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: 5e-05
—trainbatchsize: 64
—evalbatchsize: 64
—seed: 42
—gradientaccumulationsteps: 4
—totaltrainbatch_size: 256
—optimizer: Use adamwtorchfused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
—lrschedulertype: linear
—lrschedulerwarmup_steps: 500
—num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1
No log	1.0	13	4.5195	0.0535	0.0779	0.0616
No log	2.0	26	4.2070	0.0786	0.0765	0.0695
No log	3.0	39	3.7695	0.0808	0.0912	0.0835
No log	4.0	52	3.2695	0.1445	0.1382	0.1357
No log	5.0	65	2.9609	0.2160	0.2353	0.1991
No log	6.0	78	2.5957	0.3343	0.3176	0.2974
No log	7.0	91	2.2344	0.3784	0.3632	0.3659
No log	8.0	104	2.2188	0.4388	0.4368	0.4316
No log	9.0	117	2.1445	0.4599	0.4485	0.4472
No log	10.0	130	2.1133	0.4971	0.475	0.4754
No log	11.0	143	1.9170	0.5066	0.5029	0.5023
No log	12.0	156	1.9746	0.5370	0.5191	0.5218
No log	13.0	169	1.8906	0.5481	0.5309	0.5331
No log	14.0	182	1.8984	0.5269	0.5294	0.5245
No log	15.0	195	1.8643	0.5529	0.5456	0.5411

Framework versions

—PEFT 0.18.0
—Transformers 4.57.3
—Pytorch 2.9.1+cu128
—Datasets 4.4.1
—Tokenizers 0.22.1