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

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: 0.3373
• Accuracy: 43.2245
• Num: 3689
• Gen Len: 15.3058

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
0.9683	0.56	400	0.4608	31.0524	3689	14.9279
0.5046	1.13	800	0.4028	35.7175	3689	15.1651
0.4488	1.69	1200	0.3803	36.4952	3689	16.3375
0.4267	2.25	1600	0.3613	38.4613	3689	15.2646
0.4003	2.81	2000	0.3538	39.8281	3689	15.5842
0.3862	3.38	2400	0.3497	40.0593	3689	15.2356
0.3729	3.94	2800	0.3433	40.857	3689	15.9675
0.3632	4.5	3200	0.3457	41.157	3689	15.8818
0.3534	5.06	3600	0.3367	42.9369	3689	15.7314
0.3432	5.63	4000	0.3358	41.9514	3689	15.7173
0.3395	6.19	4400	0.3373	43.2245	3689	15.3058
0.3345	6.75	4800	0.3351	42.4941	3689	14.8916
0.3266	7.31	5200	0.3360	42.9742	3689	15.7124
0.3233	7.88	5600	0.3327	43.1362	3689	15.9379

Framework versions

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