Update README.md
Browse files
README.md
CHANGED
|
@@ -20,7 +20,8 @@ tags:
|
|
| 20 |
- max-len: 2048
|
| 21 |
- lr: 1e-4 then 2e-5
|
| 22 |
- weight-decay: 0.01
|
| 23 |
-
- batch:
|
|
|
|
| 24 |
- bfloat16 precision
|
| 25 |
|
| 26 |
#### Loss
|
|
|
|
| 20 |
- max-len: 2048
|
| 21 |
- lr: 1e-4 then 2e-5
|
| 22 |
- weight-decay: 0.01
|
| 23 |
+
- batch-size-per-gpu: 8
|
| 24 |
+
- GPUs: 2 A100s 80gb
|
| 25 |
- bfloat16 precision
|
| 26 |
|
| 27 |
#### Loss
|