Commit
•
96ed3d4
1
Parent(s):
6851ff6
Update README.md
Browse files
README.md
CHANGED
|
@@ -57,11 +57,11 @@ We have used Japanese corpus of total of about 220 billion tokens.
|
|
| 57 |
|CommonCrawl (snapshot: 2023-23, 2022-49, 2022-21, 2021-21)|112.9 billion|
|
| 58 |
|
| 59 |
|
| 60 |
-
##
|
| 61 |
-
- Library for distributed training: neuronx-nemo-megatron
|
| 62 |
-
- https://github.com/aws-neuron/neuronx-nemo-megatron
|
| 63 |
- Accelerator: AWS Trainium
|
| 64 |
- https://aws.amazon.com/machine-learning/trainium/
|
|
|
|
|
|
|
| 65 |
|
| 66 |
## License
|
| 67 |
[The MIT license](https://opensource.org/licenses/MIT)
|
|
|
|
| 57 |
|CommonCrawl (snapshot: 2023-23, 2022-49, 2022-21, 2021-21)|112.9 billion|
|
| 58 |
|
| 59 |
|
| 60 |
+
## Accelerator and Library
|
|
|
|
|
|
|
| 61 |
- Accelerator: AWS Trainium
|
| 62 |
- https://aws.amazon.com/machine-learning/trainium/
|
| 63 |
+
- Library for distributed training: neuronx-nemo-megatron
|
| 64 |
+
- https://github.com/aws-neuron/neuronx-nemo-megatron
|
| 65 |
|
| 66 |
## License
|
| 67 |
[The MIT license](https://opensource.org/licenses/MIT)
|