Training details edit
Browse files
README.md
CHANGED
|
@@ -145,7 +145,7 @@ The model was trained over approximately 26 hours using the STAM optimizer with
|
|
| 145 |
**Training Throughput:**
|
| 146 |
|
| 147 |
- Average step time: ~27 seconds
|
| 148 |
-
- Peak GPU memory usage: ~14.
|
| 149 |
- Total tokens processed: ~898M (input + target)
|
| 150 |
|
| 151 |
### Dataset Details
|
|
|
|
| 145 |
**Training Throughput:**
|
| 146 |
|
| 147 |
- Average step time: ~27 seconds
|
| 148 |
+
- Peak GPU memory usage: ~14.9 GB per GPU
|
| 149 |
- Total tokens processed: ~898M (input + target)
|
| 150 |
|
| 151 |
### Dataset Details
|