Ahmadzei's picture
added 3 more tables for large emb model
5fa1a76
yaml
{
"fp16": {
"enabled": "auto",
"loss_scale": 0,
"loss_scale_window": 1000,
"initial_scale_power": 16,
"hysteresis": 2,
"min_loss_scale": 1
}
}
For additional DeepSpeed fp16 training options, take a look at the FP16 Training Options reference.