flan-t5-large-da-multiwoz2.1_fs0.2

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: 0.3159
• Accuracy: 45.1554
• Num: 3689
• Gen Len: 15.5213

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: 8
• eval_batch_size: 24
• seed: 1799
• 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	Validation Loss	Accuracy	Num	Gen Len
0.9653	0.28	400	0.4635	31.3166	3689	15.196
0.5071	0.57	800	0.4031	35.8289	3689	15.6546
0.4603	0.85	1200	0.3718	37.6313	3689	15.6511
0.4219	1.13	1600	0.3577	37.9333	3689	16.5319
0.3991	1.42	2000	0.3491	40.5462	3689	15.453
0.394	1.7	2400	0.3409	40.9333	3689	15.5137
0.3822	1.98	2800	0.3370	41.2932	3689	15.225
0.3625	2.26	3200	0.3327	42.1132	3689	16.0718
0.3577	2.55	3600	0.3329	42.1372	3689	15.9973
0.3644	2.83	4000	0.3303	42.2529	3689	15.6525
0.349	3.11	4400	0.3256	43.2025	3689	15.6601
0.3355	3.4	4800	0.3243	43.791	3689	15.5451
0.338	3.68	5200	0.3231	43.5073	3689	15.7411
0.3424	3.96	5600	0.3196	44.5281	3689	15.1307
0.3299	4.25	6000	0.3159	45.1554	3689	15.5213
0.328	4.53	6400	0.3188	43.4699	3689	15.3849
0.3204	4.81	6800	0.3159	44.7764	3689	15.8219
0.3166	5.1	7200	0.3165	45.0608	3689	15.8791

Framework versions

• Transformers 4.18.0
• Pytorch 1.10.0+cu111
• Datasets 2.5.1
• Tokenizers 0.12.1