flanT5_large_Fact_U_T1

This model is a fine-tuned version of google/flan-t5-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.1337
• Accuracy: 0.7718
• Precision: 0.8116
• Recall: 0.7308
• F1 score: 0.7690

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: 0.0001
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Accuracy	F1 score	Precision	Recall	Validation Loss
1.221	0.3923	2500	0.6682	0.6659	0.6990	0.6357	1.1740
1.2301	0.7846	5000	0.6635	0.6977	0.6548	0.7466	1.5113
1.0764	1.1768	7500	0.6894	0.6741	0.7418	0.6176	1.2812
1.0245	1.5691	10000	0.7153	0.6676	0.8497	0.5498	1.2591
0.986	1.9614	12500	0.7259	0.6830	0.8567	0.5679	1.2615
0.8337	2.3537	15000	0.7271	0.6915	0.8387	0.5882	1.1350
0.807	2.7460	17500	0.74	0.7034	0.8647	0.5928	1.0071
0.7575	3.1382	20000	0.7353	0.6930	0.8729	0.5747	1.5670
0.5663	3.5305	22500	0.7435	0.7341	0.7963	0.6810	1.0824
0.6546	3.9228	25000	0.7424	0.7319	0.7973	0.6765	1.1824
0.4215	4.3151	27500	0.7435	0.7465	0.7679	0.7262	1.7775
0.4255	4.7074	30000	0.7635	0.7698	0.7796	0.7602	1.3931
0.3478	5.0996	32500	0.7635	0.7581	0.8098	0.7127	1.6014
0.2632	5.4919	35000	0.7447	0.7331	0.8032	0.6742	1.4911
0.2555	5.8842	37500	0.7588	0.7453	0.8264	0.6787	1.7558
0.2237	6.2765	40000	0.7588	0.7574	0.7940	0.7240	1.8132
0.1014	6.6688	42500	0.7647	0.7596	0.8103	0.7149	1.8028
0.15	7.0610	45000	0.7682	0.7733	0.7869	0.7602	1.7902
0.076	7.4533	47500	0.7706	0.7559	0.8459	0.6833	2.1883
0.1015	7.8456	50000	0.7694	0.7531	0.8494	0.6765	1.8640
0.0876	8.2379	52500	0.78	0.7823	0.8058	0.7602	2.0889
0.095	8.6302	55000	0.7859	0.7797	0.8385	0.7285	1.7835
0.0873	9.0224	57500	0.7718	0.7651	0.8229	0.7149	1.8784
0.0444	9.4147	60000	0.7706	0.7761	0.7879	0.7647	2.2505
0.0486	9.8070	62500	2.1337	0.7718	0.8116	0.7308	0.7690

Framework versions

• Transformers 4.44.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1