Beta
Explore
Marketplace
Neural Labs
Chat
Wallet
Docs
san_nli_2409_1325 – AI Model by JMosig | AlphaNeural AI
You can deploy this model and start earning money today!
JMosig
/
san_nli_2409_1325
like
0
transformers
pytorch
tensorboard
bert
text-classification
generated_from_trainer
autotrain_compatible
endpoints_compatible
us
Views
No views yet
Model card
Files and Versions
Community
API
Deploy
san_nli_2409_1325
This model is a fine-tuned version of
svalabs/gbert-large-zeroshot-nli
on the None dataset. It achieves the following results on the evaluation set:
Loss: 0.3856
F1: 0.9219
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: 5e-05
train_batch_size: 16
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 64
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 4
Training results
Training Loss
Epoch
Step
Validation Loss
F1
No log
0.93
10
0.2410
0.9219
No log
1.93
20
0.5240
0.9149
No log
2.93
30
0.4756
0.9219
No log
3.93
40
0.3856
0.9219
Framework versions
Transformers 4.22.2
Pytorch 1.12.1+cu113
Datasets 2.5.1
Tokenizers 0.12.1
