Update README.md
Browse files
README.md
CHANGED
|
@@ -56,14 +56,17 @@ The detail of Japanese data is summarized in the below table. The stockmark web
|
|
| 56 |
English data is sampled from [RedPajama-Data](https://github.com/togethercomputer/RedPajama-Data/tree/rp_v1).
|
| 57 |
|
| 58 |
## Training
|
|
|
|
| 59 |
- GPU: 48 nodes of a3 (8*H100) instances
|
| 60 |
- Training duration: about 7 weeks
|
| 61 |
- Container: [Pytorch NGC Container](https://catalog.ngc.nvidia.com/orgs/nvidia/containers/pytorch)
|
| 62 |
- Library: [Megatron-LM](https://github.com/NVIDIA/Megatron-LM)
|
| 63 |
|
| 64 |
## License
|
|
|
|
| 65 |
[MIT](https://opensource.org/licenses/MIT)
|
| 66 |
|
| 67 |
## Developed by
|
|
|
|
| 68 |
[Stockmark Inc.](https://stockmark.co.jp/)
|
| 69 |
|
|
|
|
| 56 |
English data is sampled from [RedPajama-Data](https://github.com/togethercomputer/RedPajama-Data/tree/rp_v1).
|
| 57 |
|
| 58 |
## Training
|
| 59 |
+
|
| 60 |
- GPU: 48 nodes of a3 (8*H100) instances
|
| 61 |
- Training duration: about 7 weeks
|
| 62 |
- Container: [Pytorch NGC Container](https://catalog.ngc.nvidia.com/orgs/nvidia/containers/pytorch)
|
| 63 |
- Library: [Megatron-LM](https://github.com/NVIDIA/Megatron-LM)
|
| 64 |
|
| 65 |
## License
|
| 66 |
+
|
| 67 |
[MIT](https://opensource.org/licenses/MIT)
|
| 68 |
|
| 69 |
## Developed by
|
| 70 |
+
|
| 71 |
[Stockmark Inc.](https://stockmark.co.jp/)
|
| 72 |
|