my_awesome_arith_model

This model is a fine-tuned version of facebook/bart-large on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3147
• Rouge1: 0.9917
• Rouge2: 0.9123
• Rougel: 0.9594
• Rougelsum: 0.9593
• Gen Len: 8.9019

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: 32
• eval_batch_size: 32
• 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	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	31	0.5797	0.977	0.7972	0.9038	0.9035	8.6517
No log	2.0	62	0.6579	0.9792	0.8002	0.9055	0.9055	10.6333
No log	3.0	93	0.4883	0.982	0.8346	0.9181	0.9182	8.7263
No log	4.0	124	0.4298	0.9813	0.8255	0.9164	0.9163	8.8856
No log	5.0	155	0.4968	0.9849	0.8408	0.9228	0.9225	10.5873
No log	6.0	186	0.4186	0.9805	0.8441	0.9253	0.9253	10.1277
No log	7.0	217	0.3675	0.9837	0.8633	0.9371	0.9373	9.2308
No log	8.0	248	0.3372	0.9909	0.9004	0.955	0.9546	9.4566
No log	9.0	279	0.3526	0.991	0.8999	0.953	0.9529	8.7896
No log	10.0	310	0.3147	0.9917	0.9123	0.9594	0.9593	8.9019

Framework versions

• Transformers 4.34.1
• Pytorch 2.0.0+cu117
• Datasets 2.14.6
• Tokenizers 0.14.1