Beta
Explore
Marketplace
Neural Labs
Playground
Wallet
Docs
expert-arxiv – AI Model by Multi-Domain-Expert-Learning | AlphaNeural AI | AlphaNeural AI
You can deploy this model and start earning money today!
Multi-Domain-Expert-Learning
/
expert-arxiv
like
0
transformers
pytorch
gpt_neox
text-generation
generated_from_trainer
apache-2.0
autotrain_compatible
text-generation-inference
endpoints_compatible
us
Views
No views yet
Model card
Files and Versions
Community
API
Deploy
expert-arxiv
This model is a fine-tuned version of
EleutherAI/pythia-1b-deduped
on the None dataset. It achieves the following results on the evaluation set:
Loss: 1.8797
Accuracy: 0.5852
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: 0.0001
train_batch_size: 1
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 8
total_train_batch_size: 64
total_eval_batch_size: 64
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
training_steps: 1000
Training results
Training Loss
Epoch
Step
Validation Loss
Accuracy
1.8752
0.01
200
1.9087
0.5805
1.8809
0.01
400
1.9018
0.5815
1.9102
0.02
600
1.8933
0.5829
1.8764
0.02
800
1.8851
0.5843
1.8694
0.03
1000
1.8797
0.5852
Framework versions
Transformers 4.28.1
Pytorch 2.0.0+cu117
Datasets 2.11.0
Tokenizers 0.13.3
