flan-t5-xl-absa-multitask-rest

This model is a fine-tuned version of ybelkada/flan-t5-xl-sharded-bf16 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1127

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
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 12

Training results

Training Loss	Epoch	Step	Validation Loss
4.3549	0.32	200	3.5848
1.5908	0.63	400	0.5331
0.4981	0.95	600	0.3159
0.351	1.27	800	0.2457
0.2884	1.58	1000	0.2118
0.2592	1.9	1200	0.2000
0.2323	2.22	1400	0.1839
0.2107	2.53	1600	0.1704
0.2071	2.85	1800	0.1649
0.1944	3.16	2000	0.1634
0.1774	3.48	2200	0.1549
0.1796	3.8	2400	0.1505
0.1695	4.11	2600	0.1427
0.1569	4.43	2800	0.1403
0.1662	4.75	3000	0.1395
0.15	5.06	3200	0.1351
0.1448	5.38	3400	0.1283
0.1444	5.7	3600	0.1302
0.1506	6.01	3800	0.1237
0.1321	6.33	4000	0.1264
0.1318	6.65	4200	0.1269
0.1298	6.96	4400	0.1207
0.1273	7.28	4600	0.1224
0.123	7.59	4800	0.1209
0.1278	7.91	5000	0.1222
0.1236	8.23	5200	0.1165
0.1188	8.54	5400	0.1154
0.1181	8.86	5600	0.1173
0.1126	9.18	5800	0.1177
0.113	9.49	6000	0.1194
0.1086	9.81	6200	0.1148
0.1086	10.13	6400	0.1158
0.1118	10.44	6600	0.1145
0.105	10.76	6800	0.1125
0.1119	11.08	7000	0.1146
0.1007	11.39	7200	0.1123
0.114	11.71	7400	0.1127

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2