Beta
Explore
Marketplace
Neural Labs
Chat
Wallet
Docs
bert-base-cased-wikitext2 – AI Model by JoonJoon | AlphaNeural AI
You can deploy this model and start earning money today!
JoonJoon
/
bert-base-cased-wikitext2
like
0
transformers
pytorch
bert
fill-mask
generated_from_trainer
autotrain_compatible
endpoints_compatible
us
Views
No views yet
Model card
Files and Versions
Community
API
Deploy
bert-base-cased-wikitext2
This model is a fine-tuned version of
on the None dataset. It achieves the following results on the evaluation set:
Loss: 6.9846
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: 24
eval_batch_size: 24
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 3.0
Training results
Training Loss
Epoch
Step
Validation Loss
7.7422
1.0
782
7.1373
7.0302
2.0
1564
6.9972
6.9788
3.0
2346
7.0087
Framework versions
Transformers 4.11.3
Pytorch 1.12.0+cu102
Datasets 1.14.0
Tokenizers 0.10.3
