Upload ds_stage1.json
Browse files- configs/ds_stage1.json +41 -0
configs/ds_stage1.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"zero_optimization": {
|
| 3 |
+
"stage": 1,
|
| 4 |
+
"allgather_partitions": true,
|
| 5 |
+
"allgather_bucket_size": 1e9,
|
| 6 |
+
"overlap_comm": true,
|
| 7 |
+
"reduce_scatter": true,
|
| 8 |
+
"reduce_bucket_size": 1e9,
|
| 9 |
+
"contiguous_gradients": true
|
| 10 |
+
},
|
| 11 |
+
"fp16": {
|
| 12 |
+
"enabled": "auto",
|
| 13 |
+
"auto_cast": true,
|
| 14 |
+
"loss_scale": 0,
|
| 15 |
+
"initial_scale_power": 32,
|
| 16 |
+
"loss_scale_window": 1000,
|
| 17 |
+
"hysteresis": 2,
|
| 18 |
+
"min_loss_scale": 1
|
| 19 |
+
},
|
| 20 |
+
"bf16": {
|
| 21 |
+
"enabled": "auto"
|
| 22 |
+
},
|
| 23 |
+
"optimizer": {
|
| 24 |
+
"type": "AdamW",
|
| 25 |
+
"params": {
|
| 26 |
+
"lr": "auto",
|
| 27 |
+
"betas": [
|
| 28 |
+
0.9,
|
| 29 |
+
0.999
|
| 30 |
+
],
|
| 31 |
+
"eps": 1e-8,
|
| 32 |
+
"weight_decay": "auto"
|
| 33 |
+
}
|
| 34 |
+
},
|
| 35 |
+
"gradient_accumulation_steps": "auto",
|
| 36 |
+
"gradient_clipping": "auto",
|
| 37 |
+
"steps_per_print": 2000,
|
| 38 |
+
"train_batch_size": "auto",
|
| 39 |
+
"train_micro_batch_size_per_gpu": "auto",
|
| 40 |
+
"wall_clock_breakdown": true
|
| 41 |
+
}
|