godel_restaurants

This model is a fine-tuned version of microsoft/GODEL-v1_1-base-seq2seq on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2869
• Bleu: 0.0565
• Rouge2: 0.0546
• Rougel: 0.2087

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: 1e-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: constant
• lr_scheduler_warmup_steps: 100
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Rouge2	Rougel
6.0069	1.0	14	5.6691	0.0	0.0073	0.0748
3.4734	2.0	28	3.0378	0.0	0.0073	0.0637
1.1689	3.0	42	0.6709	0.0	0.0065	0.0626
0.523	4.0	56	0.4075	0.0	0.0065	0.0637
0.4393	5.0	70	0.3495	0.0	0.0108	0.1135
0.1731	6.0	84	0.3100	0.0	0.0248	0.1499
0.1428	7.0	98	0.2962	0.0	0.0582	0.1752
0.1594	8.0	112	0.2931	0.0	0.0407	0.2034
0.2354	9.0	126	0.2896	0.0562	0.0546	0.21
0.2704	10.0	140	0.2869	0.0565	0.0546	0.2087

Framework versions

• Transformers 4.27.4
• Pytorch 1.13.1+cu116
• Datasets 2.11.0
• Tokenizers 0.13.2