End of training
Browse files- all_results.json +15 -0
- eval_results.json +10 -0
- logs/events.out.tfevents.1686635344.serv-9225.1554107.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +1768 -0
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.83,
|
| 3 |
+
"eval_accuracy": 0.3018414928644369,
|
| 4 |
+
"eval_loss": 326.3979187011719,
|
| 5 |
+
"eval_runtime": 4493.9318,
|
| 6 |
+
"eval_samples": 308356,
|
| 7 |
+
"eval_samples_per_second": 68.616,
|
| 8 |
+
"eval_steps_per_second": 1.906,
|
| 9 |
+
"perplexity": 5.659981303898174e+141,
|
| 10 |
+
"train_loss": 596.7565756253287,
|
| 11 |
+
"train_runtime": 197983.8753,
|
| 12 |
+
"train_samples": 5858758,
|
| 13 |
+
"train_samples_per_second": 147.96,
|
| 14 |
+
"train_steps_per_second": 4.11
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.83,
|
| 3 |
+
"eval_accuracy": 0.3018414928644369,
|
| 4 |
+
"eval_loss": 326.3979187011719,
|
| 5 |
+
"eval_runtime": 4493.9318,
|
| 6 |
+
"eval_samples": 308356,
|
| 7 |
+
"eval_samples_per_second": 68.616,
|
| 8 |
+
"eval_steps_per_second": 1.906,
|
| 9 |
+
"perplexity": 5.659981303898174e+141
|
| 10 |
+
}
|
logs/events.out.tfevents.1686635344.serv-9225.1554107.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0631f2311222934c746b5dd823d1cdb85ca6a5349b3654c963daa5c75dbda69f
|
| 3 |
+
size 369
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.83,
|
| 3 |
+
"train_loss": 596.7565756253287,
|
| 4 |
+
"train_runtime": 197983.8753,
|
| 5 |
+
"train_samples": 5858758,
|
| 6 |
+
"train_samples_per_second": 147.96,
|
| 7 |
+
"train_steps_per_second": 4.11
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1768 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 326.4412536621094,
|
| 3 |
+
"best_model_checkpoint": "bert_12_layer_model_v1_complete_training_new_48_KD/checkpoint-130000",
|
| 4 |
+
"epoch": 0.8326574251585311,
|
| 5 |
+
"global_step": 135510,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 5.000000000000001e-07,
|
| 13 |
+
"loss": 5299.6435,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 19 |
+
"loss": 4491.3755,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 1.5e-06,
|
| 25 |
+
"loss": 3930.7215,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 31 |
+
"loss": 3531.1685,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.02,
|
| 36 |
+
"learning_rate": 2.5e-06,
|
| 37 |
+
"loss": 3139.457,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.02,
|
| 42 |
+
"learning_rate": 3e-06,
|
| 43 |
+
"loss": 2777.934,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.02,
|
| 48 |
+
"learning_rate": 3.5e-06,
|
| 49 |
+
"loss": 2417.424,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.02,
|
| 54 |
+
"learning_rate": 4.000000000000001e-06,
|
| 55 |
+
"loss": 2093.8424,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.03,
|
| 60 |
+
"learning_rate": 4.5e-06,
|
| 61 |
+
"loss": 1831.8067,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.03,
|
| 66 |
+
"learning_rate": 5e-06,
|
| 67 |
+
"loss": 1621.8709,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.03,
|
| 72 |
+
"learning_rate": 5.500000000000001e-06,
|
| 73 |
+
"loss": 1464.8522,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.04,
|
| 78 |
+
"learning_rate": 6e-06,
|
| 79 |
+
"loss": 1331.6716,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.04,
|
| 84 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 85 |
+
"loss": 1233.671,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.04,
|
| 90 |
+
"learning_rate": 7e-06,
|
| 91 |
+
"loss": 1142.7462,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.05,
|
| 96 |
+
"learning_rate": 7.500000000000001e-06,
|
| 97 |
+
"loss": 1082.9845,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.05,
|
| 102 |
+
"learning_rate": 8.000000000000001e-06,
|
| 103 |
+
"loss": 1014.1482,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.05,
|
| 108 |
+
"learning_rate": 8.5e-06,
|
| 109 |
+
"loss": 964.7002,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.06,
|
| 114 |
+
"learning_rate": 9e-06,
|
| 115 |
+
"loss": 927.0119,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.06,
|
| 120 |
+
"learning_rate": 9.5e-06,
|
| 121 |
+
"loss": 880.1404,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.06,
|
| 126 |
+
"learning_rate": 1e-05,
|
| 127 |
+
"loss": 849.2694,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.06,
|
| 132 |
+
"eval_accuracy": 0.14347106203227555,
|
| 133 |
+
"eval_loss": 802.2138061523438,
|
| 134 |
+
"eval_runtime": 4494.5628,
|
| 135 |
+
"eval_samples_per_second": 68.606,
|
| 136 |
+
"eval_steps_per_second": 1.906,
|
| 137 |
+
"step": 10000
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.06,
|
| 141 |
+
"learning_rate": 9.993778927984871e-06,
|
| 142 |
+
"loss": 813.9781,
|
| 143 |
+
"step": 10500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.07,
|
| 147 |
+
"learning_rate": 9.987557855969741e-06,
|
| 148 |
+
"loss": 781.7992,
|
| 149 |
+
"step": 11000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.07,
|
| 153 |
+
"learning_rate": 9.981336783954611e-06,
|
| 154 |
+
"loss": 760.9813,
|
| 155 |
+
"step": 11500
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.07,
|
| 159 |
+
"learning_rate": 9.975115711939483e-06,
|
| 160 |
+
"loss": 744.9574,
|
| 161 |
+
"step": 12000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.08,
|
| 165 |
+
"learning_rate": 9.968894639924353e-06,
|
| 166 |
+
"loss": 725.8691,
|
| 167 |
+
"step": 12500
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.08,
|
| 171 |
+
"learning_rate": 9.962673567909223e-06,
|
| 172 |
+
"loss": 711.8804,
|
| 173 |
+
"step": 13000
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.08,
|
| 177 |
+
"learning_rate": 9.956452495894093e-06,
|
| 178 |
+
"loss": 699.3131,
|
| 179 |
+
"step": 13500
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.09,
|
| 183 |
+
"learning_rate": 9.950231423878963e-06,
|
| 184 |
+
"loss": 684.4532,
|
| 185 |
+
"step": 14000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.09,
|
| 189 |
+
"learning_rate": 9.944010351863835e-06,
|
| 190 |
+
"loss": 676.7353,
|
| 191 |
+
"step": 14500
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.09,
|
| 195 |
+
"learning_rate": 9.937789279848705e-06,
|
| 196 |
+
"loss": 667.6119,
|
| 197 |
+
"step": 15000
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.1,
|
| 201 |
+
"learning_rate": 9.931568207833575e-06,
|
| 202 |
+
"loss": 658.2263,
|
| 203 |
+
"step": 15500
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.1,
|
| 207 |
+
"learning_rate": 9.925347135818445e-06,
|
| 208 |
+
"loss": 652.0114,
|
| 209 |
+
"step": 16000
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.1,
|
| 213 |
+
"learning_rate": 9.919126063803315e-06,
|
| 214 |
+
"loss": 641.5294,
|
| 215 |
+
"step": 16500
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.1,
|
| 219 |
+
"learning_rate": 9.912904991788187e-06,
|
| 220 |
+
"loss": 635.9818,
|
| 221 |
+
"step": 17000
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.11,
|
| 225 |
+
"learning_rate": 9.906683919773055e-06,
|
| 226 |
+
"loss": 628.077,
|
| 227 |
+
"step": 17500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.11,
|
| 231 |
+
"learning_rate": 9.900462847757927e-06,
|
| 232 |
+
"loss": 623.7212,
|
| 233 |
+
"step": 18000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.11,
|
| 237 |
+
"learning_rate": 9.894241775742797e-06,
|
| 238 |
+
"loss": 618.7286,
|
| 239 |
+
"step": 18500
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.12,
|
| 243 |
+
"learning_rate": 9.888020703727668e-06,
|
| 244 |
+
"loss": 615.7981,
|
| 245 |
+
"step": 19000
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.12,
|
| 249 |
+
"learning_rate": 9.881799631712537e-06,
|
| 250 |
+
"loss": 608.6854,
|
| 251 |
+
"step": 19500
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.12,
|
| 255 |
+
"learning_rate": 9.875578559697407e-06,
|
| 256 |
+
"loss": 603.4255,
|
| 257 |
+
"step": 20000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.12,
|
| 261 |
+
"eval_accuracy": 0.14451028748778685,
|
| 262 |
+
"eval_loss": 597.5114135742188,
|
| 263 |
+
"eval_runtime": 4496.9192,
|
| 264 |
+
"eval_samples_per_second": 68.571,
|
| 265 |
+
"eval_steps_per_second": 1.905,
|
| 266 |
+
"step": 20000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.13,
|
| 270 |
+
"learning_rate": 9.869357487682279e-06,
|
| 271 |
+
"loss": 600.8989,
|
| 272 |
+
"step": 20500
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.13,
|
| 276 |
+
"learning_rate": 9.863136415667149e-06,
|
| 277 |
+
"loss": 595.9435,
|
| 278 |
+
"step": 21000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.13,
|
| 282 |
+
"learning_rate": 9.856915343652019e-06,
|
| 283 |
+
"loss": 591.4831,
|
| 284 |
+
"step": 21500
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.14,
|
| 288 |
+
"learning_rate": 9.850694271636889e-06,
|
| 289 |
+
"loss": 591.864,
|
| 290 |
+
"step": 22000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.14,
|
| 294 |
+
"learning_rate": 9.84447319962176e-06,
|
| 295 |
+
"loss": 583.3476,
|
| 296 |
+
"step": 22500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.14,
|
| 300 |
+
"learning_rate": 9.83825212760663e-06,
|
| 301 |
+
"loss": 583.3637,
|
| 302 |
+
"step": 23000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.14,
|
| 306 |
+
"learning_rate": 9.8320310555915e-06,
|
| 307 |
+
"loss": 581.8371,
|
| 308 |
+
"step": 23500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.15,
|
| 312 |
+
"learning_rate": 9.82580998357637e-06,
|
| 313 |
+
"loss": 577.9322,
|
| 314 |
+
"step": 24000
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.15,
|
| 318 |
+
"learning_rate": 9.81958891156124e-06,
|
| 319 |
+
"loss": 577.4843,
|
| 320 |
+
"step": 24500
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.15,
|
| 324 |
+
"learning_rate": 9.813367839546112e-06,
|
| 325 |
+
"loss": 573.0655,
|
| 326 |
+
"step": 25000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.16,
|
| 330 |
+
"learning_rate": 9.807146767530983e-06,
|
| 331 |
+
"loss": 568.9416,
|
| 332 |
+
"step": 25500
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.16,
|
| 336 |
+
"learning_rate": 9.800925695515851e-06,
|
| 337 |
+
"loss": 567.2173,
|
| 338 |
+
"step": 26000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.16,
|
| 342 |
+
"learning_rate": 9.794704623500723e-06,
|
| 343 |
+
"loss": 566.1221,
|
| 344 |
+
"step": 26500
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.17,
|
| 348 |
+
"learning_rate": 9.788483551485593e-06,
|
| 349 |
+
"loss": 562.4309,
|
| 350 |
+
"step": 27000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.17,
|
| 354 |
+
"learning_rate": 9.782262479470463e-06,
|
| 355 |
+
"loss": 561.3121,
|
| 356 |
+
"step": 27500
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.17,
|
| 360 |
+
"learning_rate": 9.776041407455333e-06,
|
| 361 |
+
"loss": 557.4592,
|
| 362 |
+
"step": 28000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.18,
|
| 366 |
+
"learning_rate": 9.769820335440205e-06,
|
| 367 |
+
"loss": 560.1322,
|
| 368 |
+
"step": 28500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.18,
|
| 372 |
+
"learning_rate": 9.763599263425075e-06,
|
| 373 |
+
"loss": 553.2739,
|
| 374 |
+
"step": 29000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.18,
|
| 378 |
+
"learning_rate": 9.757378191409945e-06,
|
| 379 |
+
"loss": 554.5388,
|
| 380 |
+
"step": 29500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.18,
|
| 384 |
+
"learning_rate": 9.751157119394815e-06,
|
| 385 |
+
"loss": 552.5588,
|
| 386 |
+
"step": 30000
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.18,
|
| 390 |
+
"eval_accuracy": 0.14544608101464893,
|
| 391 |
+
"eval_loss": 549.1309814453125,
|
| 392 |
+
"eval_runtime": 4500.0866,
|
| 393 |
+
"eval_samples_per_second": 68.522,
|
| 394 |
+
"eval_steps_per_second": 1.904,
|
| 395 |
+
"step": 30000
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.19,
|
| 399 |
+
"learning_rate": 9.744936047379685e-06,
|
| 400 |
+
"loss": 550.3801,
|
| 401 |
+
"step": 30500
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.19,
|
| 405 |
+
"learning_rate": 9.738714975364556e-06,
|
| 406 |
+
"loss": 546.1035,
|
| 407 |
+
"step": 31000
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.19,
|
| 411 |
+
"learning_rate": 9.732493903349427e-06,
|
| 412 |
+
"loss": 546.263,
|
| 413 |
+
"step": 31500
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.2,
|
| 417 |
+
"learning_rate": 9.726272831334297e-06,
|
| 418 |
+
"loss": 543.4885,
|
| 419 |
+
"step": 32000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.2,
|
| 423 |
+
"learning_rate": 9.720051759319167e-06,
|
| 424 |
+
"loss": 544.1279,
|
| 425 |
+
"step": 32500
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.2,
|
| 429 |
+
"learning_rate": 9.713830687304037e-06,
|
| 430 |
+
"loss": 540.8521,
|
| 431 |
+
"step": 33000
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.21,
|
| 435 |
+
"learning_rate": 9.707609615288908e-06,
|
| 436 |
+
"loss": 538.6831,
|
| 437 |
+
"step": 33500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.21,
|
| 441 |
+
"learning_rate": 9.701388543273777e-06,
|
| 442 |
+
"loss": 537.6094,
|
| 443 |
+
"step": 34000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.21,
|
| 447 |
+
"learning_rate": 9.695167471258648e-06,
|
| 448 |
+
"loss": 536.3544,
|
| 449 |
+
"step": 34500
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.22,
|
| 453 |
+
"learning_rate": 9.688946399243519e-06,
|
| 454 |
+
"loss": 538.46,
|
| 455 |
+
"step": 35000
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.22,
|
| 459 |
+
"learning_rate": 9.682725327228389e-06,
|
| 460 |
+
"loss": 535.3799,
|
| 461 |
+
"step": 35500
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.22,
|
| 465 |
+
"learning_rate": 9.676504255213259e-06,
|
| 466 |
+
"loss": 534.6113,
|
| 467 |
+
"step": 36000
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.22,
|
| 471 |
+
"learning_rate": 9.670283183198129e-06,
|
| 472 |
+
"loss": 532.2135,
|
| 473 |
+
"step": 36500
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.23,
|
| 477 |
+
"learning_rate": 9.664062111183e-06,
|
| 478 |
+
"loss": 533.0382,
|
| 479 |
+
"step": 37000
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.23,
|
| 483 |
+
"learning_rate": 9.65784103916787e-06,
|
| 484 |
+
"loss": 530.5372,
|
| 485 |
+
"step": 37500
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.23,
|
| 489 |
+
"learning_rate": 9.65161996715274e-06,
|
| 490 |
+
"loss": 528.8568,
|
| 491 |
+
"step": 38000
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.24,
|
| 495 |
+
"learning_rate": 9.64539889513761e-06,
|
| 496 |
+
"loss": 528.1091,
|
| 497 |
+
"step": 38500
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.24,
|
| 501 |
+
"learning_rate": 9.63917782312248e-06,
|
| 502 |
+
"loss": 526.5896,
|
| 503 |
+
"step": 39000
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.24,
|
| 507 |
+
"learning_rate": 9.632956751107352e-06,
|
| 508 |
+
"loss": 525.7636,
|
| 509 |
+
"step": 39500
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.25,
|
| 513 |
+
"learning_rate": 9.626735679092222e-06,
|
| 514 |
+
"loss": 525.5738,
|
| 515 |
+
"step": 40000
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.25,
|
| 519 |
+
"eval_accuracy": 0.14598640380258085,
|
| 520 |
+
"eval_loss": 523.0780639648438,
|
| 521 |
+
"eval_runtime": 4495.5628,
|
| 522 |
+
"eval_samples_per_second": 68.591,
|
| 523 |
+
"eval_steps_per_second": 1.905,
|
| 524 |
+
"step": 40000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.25,
|
| 528 |
+
"learning_rate": 9.620514607077092e-06,
|
| 529 |
+
"loss": 524.004,
|
| 530 |
+
"step": 40500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.25,
|
| 534 |
+
"learning_rate": 9.614293535061963e-06,
|
| 535 |
+
"loss": 523.962,
|
| 536 |
+
"step": 41000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.26,
|
| 540 |
+
"learning_rate": 9.608072463046833e-06,
|
| 541 |
+
"loss": 523.107,
|
| 542 |
+
"step": 41500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.26,
|
| 546 |
+
"learning_rate": 9.601851391031704e-06,
|
| 547 |
+
"loss": 522.7406,
|
| 548 |
+
"step": 42000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.26,
|
| 552 |
+
"learning_rate": 9.595630319016573e-06,
|
| 553 |
+
"loss": 519.7569,
|
| 554 |
+
"step": 42500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.26,
|
| 558 |
+
"learning_rate": 9.589409247001444e-06,
|
| 559 |
+
"loss": 519.8968,
|
| 560 |
+
"step": 43000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.27,
|
| 564 |
+
"learning_rate": 9.583188174986314e-06,
|
| 565 |
+
"loss": 518.1082,
|
| 566 |
+
"step": 43500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.27,
|
| 570 |
+
"learning_rate": 9.576967102971185e-06,
|
| 571 |
+
"loss": 516.9303,
|
| 572 |
+
"step": 44000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.27,
|
| 576 |
+
"learning_rate": 9.570746030956055e-06,
|
| 577 |
+
"loss": 516.3807,
|
| 578 |
+
"step": 44500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.28,
|
| 582 |
+
"learning_rate": 9.564524958940925e-06,
|
| 583 |
+
"loss": 515.6876,
|
| 584 |
+
"step": 45000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.28,
|
| 588 |
+
"learning_rate": 9.558303886925796e-06,
|
| 589 |
+
"loss": 515.7118,
|
| 590 |
+
"step": 45500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.28,
|
| 594 |
+
"learning_rate": 9.552082814910666e-06,
|
| 595 |
+
"loss": 513.0249,
|
| 596 |
+
"step": 46000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.29,
|
| 600 |
+
"learning_rate": 9.545861742895536e-06,
|
| 601 |
+
"loss": 513.6018,
|
| 602 |
+
"step": 46500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.29,
|
| 606 |
+
"learning_rate": 9.539640670880407e-06,
|
| 607 |
+
"loss": 513.4996,
|
| 608 |
+
"step": 47000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.29,
|
| 612 |
+
"learning_rate": 9.533419598865278e-06,
|
| 613 |
+
"loss": 513.2091,
|
| 614 |
+
"step": 47500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.29,
|
| 618 |
+
"learning_rate": 9.527198526850148e-06,
|
| 619 |
+
"loss": 510.1921,
|
| 620 |
+
"step": 48000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.3,
|
| 624 |
+
"learning_rate": 9.520977454835018e-06,
|
| 625 |
+
"loss": 511.1126,
|
| 626 |
+
"step": 48500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.3,
|
| 630 |
+
"learning_rate": 9.514756382819888e-06,
|
| 631 |
+
"loss": 510.3868,
|
| 632 |
+
"step": 49000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.3,
|
| 636 |
+
"learning_rate": 9.508535310804758e-06,
|
| 637 |
+
"loss": 510.4816,
|
| 638 |
+
"step": 49500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.31,
|
| 642 |
+
"learning_rate": 9.50231423878963e-06,
|
| 643 |
+
"loss": 508.5192,
|
| 644 |
+
"step": 50000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.31,
|
| 648 |
+
"eval_accuracy": 0.14626834935375188,
|
| 649 |
+
"eval_loss": 507.57720947265625,
|
| 650 |
+
"eval_runtime": 4507.4316,
|
| 651 |
+
"eval_samples_per_second": 68.411,
|
| 652 |
+
"eval_steps_per_second": 1.9,
|
| 653 |
+
"step": 50000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.31,
|
| 657 |
+
"learning_rate": 9.496093166774499e-06,
|
| 658 |
+
"loss": 507.3154,
|
| 659 |
+
"step": 50500
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 0.31,
|
| 663 |
+
"learning_rate": 9.489872094759369e-06,
|
| 664 |
+
"loss": 506.8317,
|
| 665 |
+
"step": 51000
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.32,
|
| 669 |
+
"learning_rate": 9.48365102274424e-06,
|
| 670 |
+
"loss": 505.1345,
|
| 671 |
+
"step": 51500
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 0.32,
|
| 675 |
+
"learning_rate": 9.47742995072911e-06,
|
| 676 |
+
"loss": 505.4357,
|
| 677 |
+
"step": 52000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.32,
|
| 681 |
+
"learning_rate": 9.47120887871398e-06,
|
| 682 |
+
"loss": 506.4242,
|
| 683 |
+
"step": 52500
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.33,
|
| 687 |
+
"learning_rate": 9.46498780669885e-06,
|
| 688 |
+
"loss": 503.4846,
|
| 689 |
+
"step": 53000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.33,
|
| 693 |
+
"learning_rate": 9.458766734683722e-06,
|
| 694 |
+
"loss": 505.3417,
|
| 695 |
+
"step": 53500
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.33,
|
| 699 |
+
"learning_rate": 9.452545662668592e-06,
|
| 700 |
+
"loss": 504.8216,
|
| 701 |
+
"step": 54000
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.33,
|
| 705 |
+
"learning_rate": 9.446324590653462e-06,
|
| 706 |
+
"loss": 503.261,
|
| 707 |
+
"step": 54500
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 0.34,
|
| 711 |
+
"learning_rate": 9.440103518638332e-06,
|
| 712 |
+
"loss": 501.3681,
|
| 713 |
+
"step": 55000
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.34,
|
| 717 |
+
"learning_rate": 9.433882446623202e-06,
|
| 718 |
+
"loss": 501.7554,
|
| 719 |
+
"step": 55500
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 0.34,
|
| 723 |
+
"learning_rate": 9.427661374608074e-06,
|
| 724 |
+
"loss": 500.2519,
|
| 725 |
+
"step": 56000
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.35,
|
| 729 |
+
"learning_rate": 9.421440302592944e-06,
|
| 730 |
+
"loss": 500.3549,
|
| 731 |
+
"step": 56500
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.35,
|
| 735 |
+
"learning_rate": 9.415219230577813e-06,
|
| 736 |
+
"loss": 500.4672,
|
| 737 |
+
"step": 57000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.35,
|
| 741 |
+
"learning_rate": 9.408998158562684e-06,
|
| 742 |
+
"loss": 499.2853,
|
| 743 |
+
"step": 57500
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 0.36,
|
| 747 |
+
"learning_rate": 9.402777086547554e-06,
|
| 748 |
+
"loss": 499.7778,
|
| 749 |
+
"step": 58000
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.36,
|
| 753 |
+
"learning_rate": 9.396556014532424e-06,
|
| 754 |
+
"loss": 498.6907,
|
| 755 |
+
"step": 58500
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 0.36,
|
| 759 |
+
"learning_rate": 9.390334942517294e-06,
|
| 760 |
+
"loss": 498.2245,
|
| 761 |
+
"step": 59000
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.37,
|
| 765 |
+
"learning_rate": 9.384113870502166e-06,
|
| 766 |
+
"loss": 497.6837,
|
| 767 |
+
"step": 59500
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.37,
|
| 771 |
+
"learning_rate": 9.377892798487036e-06,
|
| 772 |
+
"loss": 496.0482,
|
| 773 |
+
"step": 60000
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.37,
|
| 777 |
+
"eval_accuracy": 0.14565177575976343,
|
| 778 |
+
"eval_loss": 494.53851318359375,
|
| 779 |
+
"eval_runtime": 4501.2026,
|
| 780 |
+
"eval_samples_per_second": 68.505,
|
| 781 |
+
"eval_steps_per_second": 1.903,
|
| 782 |
+
"step": 60000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.37,
|
| 786 |
+
"learning_rate": 9.371671726471906e-06,
|
| 787 |
+
"loss": 497.4394,
|
| 788 |
+
"step": 60500
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.37,
|
| 792 |
+
"learning_rate": 9.365450654456776e-06,
|
| 793 |
+
"loss": 494.9281,
|
| 794 |
+
"step": 61000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.38,
|
| 798 |
+
"learning_rate": 9.359229582441646e-06,
|
| 799 |
+
"loss": 494.4248,
|
| 800 |
+
"step": 61500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.38,
|
| 804 |
+
"learning_rate": 9.353008510426518e-06,
|
| 805 |
+
"loss": 495.0921,
|
| 806 |
+
"step": 62000
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.38,
|
| 810 |
+
"learning_rate": 9.346787438411388e-06,
|
| 811 |
+
"loss": 492.6064,
|
| 812 |
+
"step": 62500
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.39,
|
| 816 |
+
"learning_rate": 9.340566366396258e-06,
|
| 817 |
+
"loss": 493.7769,
|
| 818 |
+
"step": 63000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.39,
|
| 822 |
+
"learning_rate": 9.334345294381128e-06,
|
| 823 |
+
"loss": 493.9243,
|
| 824 |
+
"step": 63500
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.39,
|
| 828 |
+
"learning_rate": 9.328124222365998e-06,
|
| 829 |
+
"loss": 492.37,
|
| 830 |
+
"step": 64000
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.4,
|
| 834 |
+
"learning_rate": 9.32190315035087e-06,
|
| 835 |
+
"loss": 492.0229,
|
| 836 |
+
"step": 64500
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.4,
|
| 840 |
+
"learning_rate": 9.31568207833574e-06,
|
| 841 |
+
"loss": 492.5365,
|
| 842 |
+
"step": 65000
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.4,
|
| 846 |
+
"learning_rate": 9.30946100632061e-06,
|
| 847 |
+
"loss": 491.0516,
|
| 848 |
+
"step": 65500
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.41,
|
| 852 |
+
"learning_rate": 9.30323993430548e-06,
|
| 853 |
+
"loss": 490.2487,
|
| 854 |
+
"step": 66000
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.41,
|
| 858 |
+
"learning_rate": 9.29701886229035e-06,
|
| 859 |
+
"loss": 489.6305,
|
| 860 |
+
"step": 66500
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.41,
|
| 864 |
+
"learning_rate": 9.29079779027522e-06,
|
| 865 |
+
"loss": 490.6653,
|
| 866 |
+
"step": 67000
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.41,
|
| 870 |
+
"learning_rate": 9.28457671826009e-06,
|
| 871 |
+
"loss": 487.7568,
|
| 872 |
+
"step": 67500
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.42,
|
| 876 |
+
"learning_rate": 9.278355646244962e-06,
|
| 877 |
+
"loss": 488.9199,
|
| 878 |
+
"step": 68000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.42,
|
| 882 |
+
"learning_rate": 9.272134574229832e-06,
|
| 883 |
+
"loss": 490.0546,
|
| 884 |
+
"step": 68500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.42,
|
| 888 |
+
"learning_rate": 9.265913502214702e-06,
|
| 889 |
+
"loss": 488.9438,
|
| 890 |
+
"step": 69000
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.43,
|
| 894 |
+
"learning_rate": 9.259692430199572e-06,
|
| 895 |
+
"loss": 487.9586,
|
| 896 |
+
"step": 69500
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.43,
|
| 900 |
+
"learning_rate": 9.253471358184442e-06,
|
| 901 |
+
"loss": 487.2105,
|
| 902 |
+
"step": 70000
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.43,
|
| 906 |
+
"eval_accuracy": 0.1464311024994026,
|
| 907 |
+
"eval_loss": 484.7273254394531,
|
| 908 |
+
"eval_runtime": 4500.233,
|
| 909 |
+
"eval_samples_per_second": 68.52,
|
| 910 |
+
"eval_steps_per_second": 1.903,
|
| 911 |
+
"step": 70000
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.43,
|
| 915 |
+
"learning_rate": 9.247250286169314e-06,
|
| 916 |
+
"loss": 487.3548,
|
| 917 |
+
"step": 70500
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 0.44,
|
| 921 |
+
"learning_rate": 9.241029214154184e-06,
|
| 922 |
+
"loss": 486.7473,
|
| 923 |
+
"step": 71000
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 0.44,
|
| 927 |
+
"learning_rate": 9.234808142139054e-06,
|
| 928 |
+
"loss": 484.9183,
|
| 929 |
+
"step": 71500
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.44,
|
| 933 |
+
"learning_rate": 9.228587070123924e-06,
|
| 934 |
+
"loss": 486.2899,
|
| 935 |
+
"step": 72000
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 0.45,
|
| 939 |
+
"learning_rate": 9.222365998108796e-06,
|
| 940 |
+
"loss": 485.0653,
|
| 941 |
+
"step": 72500
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.45,
|
| 945 |
+
"learning_rate": 9.216144926093666e-06,
|
| 946 |
+
"loss": 483.3236,
|
| 947 |
+
"step": 73000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.45,
|
| 951 |
+
"learning_rate": 9.209923854078534e-06,
|
| 952 |
+
"loss": 482.9156,
|
| 953 |
+
"step": 73500
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.45,
|
| 957 |
+
"learning_rate": 9.203702782063406e-06,
|
| 958 |
+
"loss": 483.0611,
|
| 959 |
+
"step": 74000
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 0.46,
|
| 963 |
+
"learning_rate": 9.197481710048276e-06,
|
| 964 |
+
"loss": 482.1246,
|
| 965 |
+
"step": 74500
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 0.46,
|
| 969 |
+
"learning_rate": 9.191260638033146e-06,
|
| 970 |
+
"loss": 482.9997,
|
| 971 |
+
"step": 75000
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.46,
|
| 975 |
+
"learning_rate": 9.185039566018016e-06,
|
| 976 |
+
"loss": 481.7453,
|
| 977 |
+
"step": 75500
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.47,
|
| 981 |
+
"learning_rate": 9.178818494002886e-06,
|
| 982 |
+
"loss": 479.9169,
|
| 983 |
+
"step": 76000
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.47,
|
| 987 |
+
"learning_rate": 9.172597421987758e-06,
|
| 988 |
+
"loss": 479.518,
|
| 989 |
+
"step": 76500
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.47,
|
| 993 |
+
"learning_rate": 9.166376349972628e-06,
|
| 994 |
+
"loss": 479.7603,
|
| 995 |
+
"step": 77000
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 0.48,
|
| 999 |
+
"learning_rate": 9.160155277957498e-06,
|
| 1000 |
+
"loss": 480.0271,
|
| 1001 |
+
"step": 77500
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.48,
|
| 1005 |
+
"learning_rate": 9.153934205942368e-06,
|
| 1006 |
+
"loss": 478.5475,
|
| 1007 |
+
"step": 78000
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 0.48,
|
| 1011 |
+
"learning_rate": 9.14771313392724e-06,
|
| 1012 |
+
"loss": 477.4423,
|
| 1013 |
+
"step": 78500
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.49,
|
| 1017 |
+
"learning_rate": 9.14149206191211e-06,
|
| 1018 |
+
"loss": 476.8239,
|
| 1019 |
+
"step": 79000
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 0.49,
|
| 1023 |
+
"learning_rate": 9.13527098989698e-06,
|
| 1024 |
+
"loss": 476.0955,
|
| 1025 |
+
"step": 79500
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.49,
|
| 1029 |
+
"learning_rate": 9.12904991788185e-06,
|
| 1030 |
+
"loss": 476.1281,
|
| 1031 |
+
"step": 80000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.49,
|
| 1035 |
+
"eval_accuracy": 0.14904986051099686,
|
| 1036 |
+
"eval_loss": 473.3443603515625,
|
| 1037 |
+
"eval_runtime": 4499.3007,
|
| 1038 |
+
"eval_samples_per_second": 68.534,
|
| 1039 |
+
"eval_steps_per_second": 1.904,
|
| 1040 |
+
"step": 80000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.49,
|
| 1044 |
+
"learning_rate": 9.12282884586672e-06,
|
| 1045 |
+
"loss": 473.7816,
|
| 1046 |
+
"step": 80500
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.5,
|
| 1050 |
+
"learning_rate": 9.116607773851592e-06,
|
| 1051 |
+
"loss": 474.6803,
|
| 1052 |
+
"step": 81000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.5,
|
| 1056 |
+
"learning_rate": 9.110386701836462e-06,
|
| 1057 |
+
"loss": 470.885,
|
| 1058 |
+
"step": 81500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.5,
|
| 1062 |
+
"learning_rate": 9.10416562982133e-06,
|
| 1063 |
+
"loss": 473.3452,
|
| 1064 |
+
"step": 82000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.51,
|
| 1068 |
+
"learning_rate": 9.097944557806202e-06,
|
| 1069 |
+
"loss": 470.5821,
|
| 1070 |
+
"step": 82500
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.51,
|
| 1074 |
+
"learning_rate": 9.091723485791072e-06,
|
| 1075 |
+
"loss": 469.3621,
|
| 1076 |
+
"step": 83000
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.51,
|
| 1080 |
+
"learning_rate": 9.085502413775942e-06,
|
| 1081 |
+
"loss": 470.4361,
|
| 1082 |
+
"step": 83500
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.52,
|
| 1086 |
+
"learning_rate": 9.079281341760812e-06,
|
| 1087 |
+
"loss": 470.2291,
|
| 1088 |
+
"step": 84000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.52,
|
| 1092 |
+
"learning_rate": 9.073060269745684e-06,
|
| 1093 |
+
"loss": 466.8307,
|
| 1094 |
+
"step": 84500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.52,
|
| 1098 |
+
"learning_rate": 9.066839197730554e-06,
|
| 1099 |
+
"loss": 467.9917,
|
| 1100 |
+
"step": 85000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.53,
|
| 1104 |
+
"learning_rate": 9.060618125715424e-06,
|
| 1105 |
+
"loss": 465.8809,
|
| 1106 |
+
"step": 85500
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.53,
|
| 1110 |
+
"learning_rate": 9.054397053700294e-06,
|
| 1111 |
+
"loss": 463.9713,
|
| 1112 |
+
"step": 86000
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.53,
|
| 1116 |
+
"learning_rate": 9.048175981685164e-06,
|
| 1117 |
+
"loss": 464.0281,
|
| 1118 |
+
"step": 86500
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.53,
|
| 1122 |
+
"learning_rate": 9.041954909670036e-06,
|
| 1123 |
+
"loss": 464.9262,
|
| 1124 |
+
"step": 87000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.54,
|
| 1128 |
+
"learning_rate": 9.035733837654906e-06,
|
| 1129 |
+
"loss": 460.958,
|
| 1130 |
+
"step": 87500
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.54,
|
| 1134 |
+
"learning_rate": 9.029512765639776e-06,
|
| 1135 |
+
"loss": 461.2699,
|
| 1136 |
+
"step": 88000
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.54,
|
| 1140 |
+
"learning_rate": 9.023291693624646e-06,
|
| 1141 |
+
"loss": 458.6046,
|
| 1142 |
+
"step": 88500
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 0.55,
|
| 1146 |
+
"learning_rate": 9.017070621609516e-06,
|
| 1147 |
+
"loss": 457.8228,
|
| 1148 |
+
"step": 89000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 0.55,
|
| 1152 |
+
"learning_rate": 9.010849549594388e-06,
|
| 1153 |
+
"loss": 455.9227,
|
| 1154 |
+
"step": 89500
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.55,
|
| 1158 |
+
"learning_rate": 9.004628477579256e-06,
|
| 1159 |
+
"loss": 456.0017,
|
| 1160 |
+
"step": 90000
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.55,
|
| 1164 |
+
"eval_accuracy": 0.16624086165401022,
|
| 1165 |
+
"eval_loss": 445.0464172363281,
|
| 1166 |
+
"eval_runtime": 4500.6588,
|
| 1167 |
+
"eval_samples_per_second": 68.514,
|
| 1168 |
+
"eval_steps_per_second": 1.903,
|
| 1169 |
+
"step": 90000
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 0.56,
|
| 1173 |
+
"learning_rate": 8.998407405564128e-06,
|
| 1174 |
+
"loss": 454.5954,
|
| 1175 |
+
"step": 90500
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.56,
|
| 1179 |
+
"learning_rate": 8.992186333548998e-06,
|
| 1180 |
+
"loss": 454.2384,
|
| 1181 |
+
"step": 91000
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 0.56,
|
| 1185 |
+
"learning_rate": 8.985965261533868e-06,
|
| 1186 |
+
"loss": 450.6002,
|
| 1187 |
+
"step": 91500
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 0.57,
|
| 1191 |
+
"learning_rate": 8.979744189518738e-06,
|
| 1192 |
+
"loss": 448.1904,
|
| 1193 |
+
"step": 92000
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.57,
|
| 1197 |
+
"learning_rate": 8.973523117503608e-06,
|
| 1198 |
+
"loss": 448.8409,
|
| 1199 |
+
"step": 92500
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.57,
|
| 1203 |
+
"learning_rate": 8.96730204548848e-06,
|
| 1204 |
+
"loss": 446.8822,
|
| 1205 |
+
"step": 93000
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 0.57,
|
| 1209 |
+
"learning_rate": 8.96108097347335e-06,
|
| 1210 |
+
"loss": 444.3815,
|
| 1211 |
+
"step": 93500
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 0.58,
|
| 1215 |
+
"learning_rate": 8.95485990145822e-06,
|
| 1216 |
+
"loss": 443.1149,
|
| 1217 |
+
"step": 94000
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.58,
|
| 1221 |
+
"learning_rate": 8.94863882944309e-06,
|
| 1222 |
+
"loss": 441.321,
|
| 1223 |
+
"step": 94500
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 0.58,
|
| 1227 |
+
"learning_rate": 8.94241775742796e-06,
|
| 1228 |
+
"loss": 437.646,
|
| 1229 |
+
"step": 95000
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 0.59,
|
| 1233 |
+
"learning_rate": 8.936196685412832e-06,
|
| 1234 |
+
"loss": 437.9417,
|
| 1235 |
+
"step": 95500
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.59,
|
| 1239 |
+
"learning_rate": 8.929975613397702e-06,
|
| 1240 |
+
"loss": 436.0795,
|
| 1241 |
+
"step": 96000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.59,
|
| 1245 |
+
"learning_rate": 8.923754541382572e-06,
|
| 1246 |
+
"loss": 435.8347,
|
| 1247 |
+
"step": 96500
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 0.6,
|
| 1251 |
+
"learning_rate": 8.917533469367442e-06,
|
| 1252 |
+
"loss": 431.6963,
|
| 1253 |
+
"step": 97000
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 0.6,
|
| 1257 |
+
"learning_rate": 8.911312397352314e-06,
|
| 1258 |
+
"loss": 431.1178,
|
| 1259 |
+
"step": 97500
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.6,
|
| 1263 |
+
"learning_rate": 8.905091325337182e-06,
|
| 1264 |
+
"loss": 429.1986,
|
| 1265 |
+
"step": 98000
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 0.61,
|
| 1269 |
+
"learning_rate": 8.898870253322052e-06,
|
| 1270 |
+
"loss": 424.9979,
|
| 1271 |
+
"step": 98500
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 0.61,
|
| 1275 |
+
"learning_rate": 8.892649181306924e-06,
|
| 1276 |
+
"loss": 426.5004,
|
| 1277 |
+
"step": 99000
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.61,
|
| 1281 |
+
"learning_rate": 8.886428109291794e-06,
|
| 1282 |
+
"loss": 424.2374,
|
| 1283 |
+
"step": 99500
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.61,
|
| 1287 |
+
"learning_rate": 8.880207037276664e-06,
|
| 1288 |
+
"loss": 421.6633,
|
| 1289 |
+
"step": 100000
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 0.61,
|
| 1293 |
+
"eval_accuracy": 0.20462580767987698,
|
| 1294 |
+
"eval_loss": 404.10711669921875,
|
| 1295 |
+
"eval_runtime": 4503.0526,
|
| 1296 |
+
"eval_samples_per_second": 68.477,
|
| 1297 |
+
"eval_steps_per_second": 1.902,
|
| 1298 |
+
"step": 100000
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.62,
|
| 1302 |
+
"learning_rate": 8.873985965261534e-06,
|
| 1303 |
+
"loss": 419.1612,
|
| 1304 |
+
"step": 100500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.62,
|
| 1308 |
+
"learning_rate": 8.867764893246404e-06,
|
| 1309 |
+
"loss": 416.2441,
|
| 1310 |
+
"step": 101000
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 0.62,
|
| 1314 |
+
"learning_rate": 8.861543821231276e-06,
|
| 1315 |
+
"loss": 414.9002,
|
| 1316 |
+
"step": 101500
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 0.63,
|
| 1320 |
+
"learning_rate": 8.855322749216146e-06,
|
| 1321 |
+
"loss": 412.1551,
|
| 1322 |
+
"step": 102000
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.63,
|
| 1326 |
+
"learning_rate": 8.849101677201016e-06,
|
| 1327 |
+
"loss": 410.3781,
|
| 1328 |
+
"step": 102500
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.63,
|
| 1332 |
+
"learning_rate": 8.842880605185886e-06,
|
| 1333 |
+
"loss": 408.377,
|
| 1334 |
+
"step": 103000
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 0.64,
|
| 1338 |
+
"learning_rate": 8.836659533170758e-06,
|
| 1339 |
+
"loss": 406.7924,
|
| 1340 |
+
"step": 103500
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.64,
|
| 1344 |
+
"learning_rate": 8.830438461155628e-06,
|
| 1345 |
+
"loss": 404.9551,
|
| 1346 |
+
"step": 104000
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.64,
|
| 1350 |
+
"learning_rate": 8.824217389140498e-06,
|
| 1351 |
+
"loss": 403.9014,
|
| 1352 |
+
"step": 104500
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 0.65,
|
| 1356 |
+
"learning_rate": 8.817996317125368e-06,
|
| 1357 |
+
"loss": 402.8424,
|
| 1358 |
+
"step": 105000
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 0.65,
|
| 1362 |
+
"learning_rate": 8.811775245110238e-06,
|
| 1363 |
+
"loss": 400.1352,
|
| 1364 |
+
"step": 105500
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.65,
|
| 1368 |
+
"learning_rate": 8.80555417309511e-06,
|
| 1369 |
+
"loss": 396.1973,
|
| 1370 |
+
"step": 106000
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.65,
|
| 1374 |
+
"learning_rate": 8.799333101079978e-06,
|
| 1375 |
+
"loss": 396.7836,
|
| 1376 |
+
"step": 106500
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 0.66,
|
| 1380 |
+
"learning_rate": 8.793112029064848e-06,
|
| 1381 |
+
"loss": 393.8781,
|
| 1382 |
+
"step": 107000
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.66,
|
| 1386 |
+
"learning_rate": 8.78689095704972e-06,
|
| 1387 |
+
"loss": 392.84,
|
| 1388 |
+
"step": 107500
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.66,
|
| 1392 |
+
"learning_rate": 8.78066988503459e-06,
|
| 1393 |
+
"loss": 389.1277,
|
| 1394 |
+
"step": 108000
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 0.67,
|
| 1398 |
+
"learning_rate": 8.77444881301946e-06,
|
| 1399 |
+
"loss": 388.1972,
|
| 1400 |
+
"step": 108500
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 0.67,
|
| 1404 |
+
"learning_rate": 8.76822774100433e-06,
|
| 1405 |
+
"loss": 387.9536,
|
| 1406 |
+
"step": 109000
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.67,
|
| 1410 |
+
"learning_rate": 8.762006668989202e-06,
|
| 1411 |
+
"loss": 387.2457,
|
| 1412 |
+
"step": 109500
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 0.68,
|
| 1416 |
+
"learning_rate": 8.755785596974072e-06,
|
| 1417 |
+
"loss": 382.6604,
|
| 1418 |
+
"step": 110000
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 0.68,
|
| 1422 |
+
"eval_accuracy": 0.24460605268573873,
|
| 1423 |
+
"eval_loss": 369.21484375,
|
| 1424 |
+
"eval_runtime": 4502.1476,
|
| 1425 |
+
"eval_samples_per_second": 68.491,
|
| 1426 |
+
"eval_steps_per_second": 1.903,
|
| 1427 |
+
"step": 110000
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.68,
|
| 1431 |
+
"learning_rate": 8.749564524958942e-06,
|
| 1432 |
+
"loss": 381.8682,
|
| 1433 |
+
"step": 110500
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 0.68,
|
| 1437 |
+
"learning_rate": 8.743343452943812e-06,
|
| 1438 |
+
"loss": 382.7058,
|
| 1439 |
+
"step": 111000
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 0.69,
|
| 1443 |
+
"learning_rate": 8.737122380928682e-06,
|
| 1444 |
+
"loss": 379.1437,
|
| 1445 |
+
"step": 111500
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.69,
|
| 1449 |
+
"learning_rate": 8.730901308913554e-06,
|
| 1450 |
+
"loss": 377.1325,
|
| 1451 |
+
"step": 112000
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.69,
|
| 1455 |
+
"learning_rate": 8.724680236898424e-06,
|
| 1456 |
+
"loss": 375.0838,
|
| 1457 |
+
"step": 112500
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 0.69,
|
| 1461 |
+
"learning_rate": 8.718459164883294e-06,
|
| 1462 |
+
"loss": 375.5783,
|
| 1463 |
+
"step": 113000
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 0.7,
|
| 1467 |
+
"learning_rate": 8.712238092868164e-06,
|
| 1468 |
+
"loss": 373.4636,
|
| 1469 |
+
"step": 113500
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 0.7,
|
| 1473 |
+
"learning_rate": 8.706017020853034e-06,
|
| 1474 |
+
"loss": 371.6336,
|
| 1475 |
+
"step": 114000
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 0.7,
|
| 1479 |
+
"learning_rate": 8.699795948837904e-06,
|
| 1480 |
+
"loss": 370.4187,
|
| 1481 |
+
"step": 114500
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 0.71,
|
| 1485 |
+
"learning_rate": 8.693574876822774e-06,
|
| 1486 |
+
"loss": 370.2962,
|
| 1487 |
+
"step": 115000
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.71,
|
| 1491 |
+
"learning_rate": 8.687353804807646e-06,
|
| 1492 |
+
"loss": 369.4474,
|
| 1493 |
+
"step": 115500
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.71,
|
| 1497 |
+
"learning_rate": 8.681132732792516e-06,
|
| 1498 |
+
"loss": 367.2834,
|
| 1499 |
+
"step": 116000
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 0.72,
|
| 1503 |
+
"learning_rate": 8.674911660777386e-06,
|
| 1504 |
+
"loss": 365.3446,
|
| 1505 |
+
"step": 116500
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 0.72,
|
| 1509 |
+
"learning_rate": 8.668690588762256e-06,
|
| 1510 |
+
"loss": 365.8585,
|
| 1511 |
+
"step": 117000
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 0.72,
|
| 1515 |
+
"learning_rate": 8.662469516747126e-06,
|
| 1516 |
+
"loss": 363.0709,
|
| 1517 |
+
"step": 117500
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 0.73,
|
| 1521 |
+
"learning_rate": 8.656248444731998e-06,
|
| 1522 |
+
"loss": 362.2264,
|
| 1523 |
+
"step": 118000
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 0.73,
|
| 1527 |
+
"learning_rate": 8.650027372716868e-06,
|
| 1528 |
+
"loss": 361.6181,
|
| 1529 |
+
"step": 118500
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.73,
|
| 1533 |
+
"learning_rate": 8.643806300701738e-06,
|
| 1534 |
+
"loss": 359.4933,
|
| 1535 |
+
"step": 119000
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.73,
|
| 1539 |
+
"learning_rate": 8.637585228686608e-06,
|
| 1540 |
+
"loss": 358.7548,
|
| 1541 |
+
"step": 119500
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 0.74,
|
| 1545 |
+
"learning_rate": 8.631364156671478e-06,
|
| 1546 |
+
"loss": 358.6727,
|
| 1547 |
+
"step": 120000
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 0.74,
|
| 1551 |
+
"eval_accuracy": 0.2775794485775731,
|
| 1552 |
+
"eval_loss": 341.1114196777344,
|
| 1553 |
+
"eval_runtime": 4501.5791,
|
| 1554 |
+
"eval_samples_per_second": 68.5,
|
| 1555 |
+
"eval_steps_per_second": 1.903,
|
| 1556 |
+
"step": 120000
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.74,
|
| 1560 |
+
"learning_rate": 8.62514308465635e-06,
|
| 1561 |
+
"loss": 356.8832,
|
| 1562 |
+
"step": 120500
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 0.74,
|
| 1566 |
+
"learning_rate": 8.618922012641218e-06,
|
| 1567 |
+
"loss": 357.4298,
|
| 1568 |
+
"step": 121000
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 0.75,
|
| 1572 |
+
"learning_rate": 8.61270094062609e-06,
|
| 1573 |
+
"loss": 354.2308,
|
| 1574 |
+
"step": 121500
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 0.75,
|
| 1578 |
+
"learning_rate": 8.60647986861096e-06,
|
| 1579 |
+
"loss": 353.5963,
|
| 1580 |
+
"step": 122000
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 0.75,
|
| 1584 |
+
"learning_rate": 8.600258796595831e-06,
|
| 1585 |
+
"loss": 352.4074,
|
| 1586 |
+
"step": 122500
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.76,
|
| 1590 |
+
"learning_rate": 8.5940377245807e-06,
|
| 1591 |
+
"loss": 351.2607,
|
| 1592 |
+
"step": 123000
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.76,
|
| 1596 |
+
"learning_rate": 8.58781665256557e-06,
|
| 1597 |
+
"loss": 350.996,
|
| 1598 |
+
"step": 123500
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.76,
|
| 1602 |
+
"learning_rate": 8.581595580550442e-06,
|
| 1603 |
+
"loss": 349.7997,
|
| 1604 |
+
"step": 124000
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 0.77,
|
| 1608 |
+
"learning_rate": 8.575374508535312e-06,
|
| 1609 |
+
"loss": 348.3944,
|
| 1610 |
+
"step": 124500
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 0.77,
|
| 1614 |
+
"learning_rate": 8.569153436520182e-06,
|
| 1615 |
+
"loss": 347.8057,
|
| 1616 |
+
"step": 125000
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.77,
|
| 1620 |
+
"learning_rate": 8.562932364505052e-06,
|
| 1621 |
+
"loss": 347.0378,
|
| 1622 |
+
"step": 125500
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 0.77,
|
| 1626 |
+
"learning_rate": 8.556711292489922e-06,
|
| 1627 |
+
"loss": 347.7662,
|
| 1628 |
+
"step": 126000
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 0.78,
|
| 1632 |
+
"learning_rate": 8.550490220474794e-06,
|
| 1633 |
+
"loss": 344.9208,
|
| 1634 |
+
"step": 126500
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.78,
|
| 1638 |
+
"learning_rate": 8.544269148459664e-06,
|
| 1639 |
+
"loss": 345.5751,
|
| 1640 |
+
"step": 127000
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.78,
|
| 1644 |
+
"learning_rate": 8.538048076444534e-06,
|
| 1645 |
+
"loss": 344.0301,
|
| 1646 |
+
"step": 127500
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 0.79,
|
| 1650 |
+
"learning_rate": 8.531827004429404e-06,
|
| 1651 |
+
"loss": 344.0015,
|
| 1652 |
+
"step": 128000
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 0.79,
|
| 1656 |
+
"learning_rate": 8.525605932414275e-06,
|
| 1657 |
+
"loss": 343.2538,
|
| 1658 |
+
"step": 128500
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 0.79,
|
| 1662 |
+
"learning_rate": 8.519384860399145e-06,
|
| 1663 |
+
"loss": 341.9897,
|
| 1664 |
+
"step": 129000
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 0.8,
|
| 1668 |
+
"learning_rate": 8.513163788384014e-06,
|
| 1669 |
+
"loss": 339.5061,
|
| 1670 |
+
"step": 129500
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 0.8,
|
| 1674 |
+
"learning_rate": 8.506942716368886e-06,
|
| 1675 |
+
"loss": 339.9395,
|
| 1676 |
+
"step": 130000
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.8,
|
| 1680 |
+
"eval_accuracy": 0.30175581659567297,
|
| 1681 |
+
"eval_loss": 326.4412536621094,
|
| 1682 |
+
"eval_runtime": 4500.3594,
|
| 1683 |
+
"eval_samples_per_second": 68.518,
|
| 1684 |
+
"eval_steps_per_second": 1.903,
|
| 1685 |
+
"step": 130000
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 0.8,
|
| 1689 |
+
"learning_rate": 8.500721644353756e-06,
|
| 1690 |
+
"loss": 339.0469,
|
| 1691 |
+
"step": 130500
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 0.8,
|
| 1695 |
+
"learning_rate": 8.494500572338626e-06,
|
| 1696 |
+
"loss": 337.8205,
|
| 1697 |
+
"step": 131000
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.81,
|
| 1701 |
+
"learning_rate": 8.488279500323496e-06,
|
| 1702 |
+
"loss": 337.9211,
|
| 1703 |
+
"step": 131500
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.81,
|
| 1707 |
+
"learning_rate": 8.482058428308367e-06,
|
| 1708 |
+
"loss": 337.4457,
|
| 1709 |
+
"step": 132000
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 0.81,
|
| 1713 |
+
"learning_rate": 8.475837356293238e-06,
|
| 1714 |
+
"loss": 335.1968,
|
| 1715 |
+
"step": 132500
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 0.82,
|
| 1719 |
+
"learning_rate": 8.469616284278108e-06,
|
| 1720 |
+
"loss": 333.5669,
|
| 1721 |
+
"step": 133000
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 0.82,
|
| 1725 |
+
"learning_rate": 8.463395212262978e-06,
|
| 1726 |
+
"loss": 333.9828,
|
| 1727 |
+
"step": 133500
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 0.82,
|
| 1731 |
+
"learning_rate": 8.457174140247848e-06,
|
| 1732 |
+
"loss": 332.9984,
|
| 1733 |
+
"step": 134000
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.83,
|
| 1737 |
+
"learning_rate": 8.45095306823272e-06,
|
| 1738 |
+
"loss": 332.9492,
|
| 1739 |
+
"step": 134500
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.83,
|
| 1743 |
+
"learning_rate": 8.44473199621759e-06,
|
| 1744 |
+
"loss": 332.6251,
|
| 1745 |
+
"step": 135000
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.83,
|
| 1749 |
+
"learning_rate": 8.43851092420246e-06,
|
| 1750 |
+
"loss": 332.3873,
|
| 1751 |
+
"step": 135500
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 0.83,
|
| 1755 |
+
"step": 135510,
|
| 1756 |
+
"total_flos": 1.445396136369062e+18,
|
| 1757 |
+
"train_loss": 596.7565756253287,
|
| 1758 |
+
"train_runtime": 197983.8753,
|
| 1759 |
+
"train_samples_per_second": 147.96,
|
| 1760 |
+
"train_steps_per_second": 4.11
|
| 1761 |
+
}
|
| 1762 |
+
],
|
| 1763 |
+
"max_steps": 813720,
|
| 1764 |
+
"num_train_epochs": 5,
|
| 1765 |
+
"total_flos": 1.445396136369062e+18,
|
| 1766 |
+
"trial_name": null,
|
| 1767 |
+
"trial_params": null
|
| 1768 |
+
}
|