50pos_model

This model is a fine-tuned version of t5-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.4089
• Bleu: 0.0001
• Wer: 0.9459
• Rougel: 0.1326
• Gen Len: 18.9987

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: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 2
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Wer	Rougel	Gen Len
2.191	0.16	1000	1.3669	0.0001	0.9468	0.1309	18.999
1.5151	0.32	2000	1.2748	0.0001	0.9458	0.1327	18.9987
1.6193	0.48	3000	1.4301	0.0001	0.9457	0.133	18.9986
1.7458	0.64	4000	1.4091	0.0001	0.9459	0.1326	18.9986
1.7281	0.8	5000	1.4088	0.0001	0.9459	0.1326	18.9986
1.7255	0.96	6000	1.4089	0.0001	0.9459	0.1326	18.9986
1.7297	1.12	7000	1.4089	0.0001	0.9459	0.1326	18.9987
1.7197	1.28	8000	1.4089	0.0001	0.9459	0.1326	18.9987
1.7287	1.44	9000	1.4088	0.0001	0.9459	0.1326	18.9986
1.7253	1.6	10000	1.4088	0.0001	0.9459	0.1326	18.9987
1.7189	1.76	11000	1.4089	0.0001	0.9459	0.1326	18.9987
1.7257	1.92	12000	1.4089	0.0001	0.9459	0.1326	18.9987

Framework versions

• Transformers 4.37.1
• Pytorch 2.3.0.dev20240122+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1