Update README.md (#31)
Browse files- Update README.md (00ccb31b042b205bd8d2dbdd501003138af90ce4)
Co-authored-by: Niels Rogge <nielsr@users.noreply.huggingface.co>
README.md
CHANGED
|
@@ -63,8 +63,6 @@ For code examples, we refer to the [documentation](https://huggingface.co/docs/t
|
|
| 63 |
|
| 64 |
The memory requirements differ based on the precision one uses. One can use 4-bit inference using [Bitsandbytes](https://huggingface.co/blog/4bit-transformers-bitsandbytes), which greatly reduce the memory requirements.
|
| 65 |
|
| 66 |
-
Training requires 4 times the
|
| 67 |
-
|
| 68 |
| dtype | Largest Layer or Residual Group | Total Size | Training using Adam |
|
| 69 |
|-------------------|---------------------------------|------------|----------------------|
|
| 70 |
| float32 | 490.94 MB | 14.43 GB | 57.72 GB |
|
|
|
|
| 63 |
|
| 64 |
The memory requirements differ based on the precision one uses. One can use 4-bit inference using [Bitsandbytes](https://huggingface.co/blog/4bit-transformers-bitsandbytes), which greatly reduce the memory requirements.
|
| 65 |
|
|
|
|
|
|
|
| 66 |
| dtype | Largest Layer or Residual Group | Total Size | Training using Adam |
|
| 67 |
|-------------------|---------------------------------|------------|----------------------|
|
| 68 |
| float32 | 490.94 MB | 14.43 GB | 57.72 GB |
|