Beta
Explore
Marketplace
Neural Labs
Chat
Wallet
Docs
t5-small-finetuned-fi-to-en – AI Model by Sancha | AlphaNeural AI
You can deploy this model and start earning money today!
Sancha
/
t5-small-finetuned-fi-to-en
like
0
transformers
pytorch
tensorboard
t5
text2text-generation
generated_from_trainer
wmt19
apache-2.0
model-index
text-generation-inference
endpoints_compatible
us
Views
No views yet
Model card
Files and Versions
Community
API
Deploy
t5-small-finetuned-fi-to-en
This model is a fine-tuned version of
t5-small
on the wmt19 dataset. It achieves the following results on the evaluation set:
Loss: 3.5185
Bleu: 1.2541
Gen Len: 17.395
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: 16
eval_batch_size: 16
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
Gen Len
3.413
1.0
6250
3.5378
1.2291
17.4057
3.342
2.0
12500
3.5185
1.2541
17.395
Framework versions
Transformers 4.12.5
Pytorch 1.9.1
Datasets 1.16.1
Tokenizers 0.10.3
