Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism
Paper • 1909.08053 • Published • 5
Model Description
1.3 billion parameter Megatron version of FinBERT.
Evaluation results
Document classification
Results have been measured on 10% of the datasets
Ylilauta
| Model | Accuracy |
|---|---|
| Megatron-finbert-1.3B-cased | 82.64% |
| FinBERT-cased | 79.18% |
Yle
| Model | Accuracy |
|---|---|
| Megatron-finbert-1.3B-cased | 82.64% |
| FinBERT-cased | 79.18% |
More documentation coming soon...
- Downloads last month
- 7
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support