HeavenlyJoe/flan-t5-large-eng-tgl-translation

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: 1.3378
• Bleu: 0.4953
• Gen Len: 19.0

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: 4e-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	Bleu	Gen Len
1.9527	0.44	25	1.5761	0.2146	19.0
1.8866	0.88	50	1.5303	0.293	19.0
1.8045	1.32	75	1.5092	0.2499	19.0
1.7596	1.75	100	1.4840	0.3498	19.0
1.7354	2.19	125	1.4628	0.3282	19.0
1.6866	2.63	150	1.4437	0.3205	19.0
1.6605	3.07	175	1.4275	0.3781	19.0
1.6157	3.51	200	1.4177	0.3805	19.0
1.6237	3.95	225	1.4007	0.398	19.0
1.5948	4.39	250	1.3954	0.4022	19.0
1.5555	4.82	275	1.3866	0.3854	19.0
1.5388	5.26	300	1.3761	0.4105	19.0
1.5448	5.7	325	1.3712	0.4339	19.0
1.5149	6.14	350	1.3635	0.4342	19.0
1.5104	6.58	375	1.3566	0.459	19.0
1.4955	7.02	400	1.3525	0.4888	19.0
1.467	7.46	425	1.3491	0.4723	19.0
1.4872	7.89	450	1.3440	0.491	19.0
1.4766	8.33	475	1.3423	0.5183	19.0
1.4553	8.77	500	1.3404	0.5026	19.0
1.464	9.21	525	1.3384	0.4979	19.0
1.454	9.65	550	1.3378	0.4953	19.0

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0