llama-sentiment-classifier

This model is a fine-tuned version of meta-llama/Llama-3.1-8B-instruct on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3453
• Accuracy: 0.9677
• F1: 0.9735
• Precision: 0.9756
• Recall: 0.9713

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.5412	0.0783	100	0.5607	0.7675	0.7938	0.8627	0.7351
0.3145	0.1566	200	0.3603	0.9530	0.9608	0.975	0.9470
0.3636	0.2349	300	0.3565	0.9543	0.9621	0.9729	0.9514
0.3313	0.3132	400	0.3533	0.9597	0.9666	0.9753	0.9581
0.3633	0.3915	500	0.3516	0.9610	0.9681	0.9649	0.9713
0.3537	0.4699	600	0.3498	0.9624	0.9692	0.9650	0.9735
0.3381	0.5482	700	0.3483	0.9624	0.9691	0.9691	0.9691
0.3141	0.6265	800	0.3497	0.9637	0.9700	0.9754	0.9647
0.3661	0.7048	900	0.3472	0.9664	0.9724	0.9735	0.9713
0.3133	0.7831	1000	0.3476	0.9651	0.9714	0.9692	0.9735
0.3667	0.8614	1100	0.3517	0.9597	0.9666	0.9753	0.9581
0.3383	0.9397	1200	0.3463	0.9664	0.9726	0.9672	0.9779
0.3133	1.0180	1300	0.3455	0.9691	0.9746	0.9736	0.9757
0.3733	1.0963	1400	0.3570	0.9556	0.9630	0.9773	0.9492
0.3382	1.1746	1500	0.3516	0.9610	0.9677	0.9753	0.9603
0.3383	1.2529	1600	0.3464	0.9664	0.9724	0.9714	0.9735
0.3133	1.3312	1700	0.3461	0.9677	0.9736	0.9714	0.9757
0.3407	1.4096	1800	0.3482	0.9651	0.9712	0.9755	0.9669
0.3383	1.4879	1900	0.3560	0.9543	0.9618	0.9794	0.9448
0.3158	1.5662	2000	0.3466	0.9664	0.9723	0.9756	0.9691
0.3411	1.6445	2100	0.3452	0.9677	0.9735	0.9756	0.9713
0.3133	1.7228	2200	0.3451	0.9677	0.9735	0.9756	0.9713
0.3633	1.8011	2300	0.3439	0.9677	0.9735	0.9756	0.9713
0.3133	1.8794	2400	0.3459	0.9664	0.9723	0.9756	0.9691
0.3133	1.9577	2500	0.3463	0.9664	0.9723	0.9756	0.9691
0.3135	2.0360	2600	0.3447	0.9677	0.9735	0.9756	0.9713
0.3133	2.1143	2700	0.3436	0.9691	0.9746	0.9736	0.9757
0.3133	2.1926	2800	0.3461	0.9664	0.9723	0.9756	0.9691
0.3133	2.2709	2900	0.3498	0.9624	0.9689	0.9754	0.9625
0.3383	2.3493	3000	0.3473	0.9651	0.9712	0.9755	0.9669
0.3133	2.4276	3100	0.3461	0.9664	0.9723	0.9756	0.9691
0.3633	2.5059	3200	0.3457	0.9664	0.9723	0.9756	0.9691
0.3133	2.5842	3300	0.3457	0.9664	0.9723	0.9756	0.9691
0.3383	2.6625	3400	0.3451	0.9677	0.9735	0.9756	0.9713
0.3382	2.7408	3500	0.3452	0.9677	0.9735	0.9756	0.9713
0.3133	2.8191	3600	0.3452	0.9677	0.9735	0.9756	0.9713
0.3633	2.8974	3700	0.3454	0.9677	0.9735	0.9756	0.9713
0.3383	2.9757	3800	0.3453	0.9677	0.9735	0.9756	0.9713

Framework versions

• PEFT 0.13.2
• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.2
• Tokenizers 0.19.1