Beta
Explore
Marketplace
Neural Labs
Chat
Wallet
Docs
gpn_balanced_grass – Dataset by nberkowitz | AlphaNeural AI
You can deploy this model and start earning money today!
nberkowitz
/
gpn_balanced_grass
like
0
us
Views
No views yet
Model card
Files and Versions
Community
API
model
This model is a fine-tuned version of
on the nberkowitz/gpn_grass_balanced_v1 dataset. It achieves the following results on the evaluation set:
Loss: 1.0249
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: 64
eval_batch_size: 512
seed: 42
distributed_type: multi-GPU
gradient_accumulation_steps: 4
total_train_batch_size: 256
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
training_steps: 120000
Training results
Training Loss
Epoch
Step
Validation Loss
1.0777
2.01
60000
1.0535
1.0495
4.03
120000
1.0253
Framework versions
Transformers 4.30.0
Pytorch 2.4.1
Datasets 3.1.0
Tokenizers 0.13.3
