End of training
Browse files- all_results.json +13 -0
- eval_results.json +8 -0
- runs/Apr26_10-41-08_f5c94994a90d/events.out.tfevents.1714137706.f5c94994a90d.529.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +1778 -0
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 19.955456570155903,
|
| 3 |
+
"eval_accuracy": 0.8703819661865999,
|
| 4 |
+
"eval_loss": 0.5583016276359558,
|
| 5 |
+
"eval_runtime": 21.3673,
|
| 6 |
+
"eval_samples_per_second": 74.74,
|
| 7 |
+
"eval_steps_per_second": 2.34,
|
| 8 |
+
"total_flos": 2.928337686848471e+19,
|
| 9 |
+
"train_loss": 0.5139281852969102,
|
| 10 |
+
"train_runtime": 9596.9456,
|
| 11 |
+
"train_samples_per_second": 29.937,
|
| 12 |
+
"train_steps_per_second": 0.233
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 19.955456570155903,
|
| 3 |
+
"eval_accuracy": 0.8703819661865999,
|
| 4 |
+
"eval_loss": 0.5583016276359558,
|
| 5 |
+
"eval_runtime": 21.3673,
|
| 6 |
+
"eval_samples_per_second": 74.74,
|
| 7 |
+
"eval_steps_per_second": 2.34
|
| 8 |
+
}
|
runs/Apr26_10-41-08_f5c94994a90d/events.out.tfevents.1714137706.f5c94994a90d.529.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:307e6697ed09410f91c5616d3447ac63668d0c7f6d437e3fc2c1d1244120cbda
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 19.955456570155903,
|
| 3 |
+
"total_flos": 2.928337686848471e+19,
|
| 4 |
+
"train_loss": 0.5139281852969102,
|
| 5 |
+
"train_runtime": 9596.9456,
|
| 6 |
+
"train_samples_per_second": 29.937,
|
| 7 |
+
"train_steps_per_second": 0.233
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1778 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8703819661865999,
|
| 3 |
+
"best_model_checkpoint": "dinov2-base-finetuned-galaxy/checkpoint-2240",
|
| 4 |
+
"epoch": 19.955456570155903,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2240,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.08908685968819599,
|
| 13 |
+
"grad_norm": 41.429115295410156,
|
| 14 |
+
"learning_rate": 2.2321428571428573e-06,
|
| 15 |
+
"loss": 2.5179,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.17817371937639198,
|
| 20 |
+
"grad_norm": 39.681358337402344,
|
| 21 |
+
"learning_rate": 4.464285714285715e-06,
|
| 22 |
+
"loss": 2.0573,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.267260579064588,
|
| 27 |
+
"grad_norm": 43.90978240966797,
|
| 28 |
+
"learning_rate": 6.696428571428572e-06,
|
| 29 |
+
"loss": 1.5858,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.35634743875278396,
|
| 34 |
+
"grad_norm": 58.3934211730957,
|
| 35 |
+
"learning_rate": 8.92857142857143e-06,
|
| 36 |
+
"loss": 1.315,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.44543429844098,
|
| 41 |
+
"grad_norm": 50.093204498291016,
|
| 42 |
+
"learning_rate": 1.1160714285714287e-05,
|
| 43 |
+
"loss": 1.1625,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.534521158129176,
|
| 48 |
+
"grad_norm": 62.2159423828125,
|
| 49 |
+
"learning_rate": 1.3392857142857144e-05,
|
| 50 |
+
"loss": 1.068,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.623608017817372,
|
| 55 |
+
"grad_norm": 71.58743286132812,
|
| 56 |
+
"learning_rate": 1.5625e-05,
|
| 57 |
+
"loss": 0.9487,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.7126948775055679,
|
| 62 |
+
"grad_norm": 71.40558624267578,
|
| 63 |
+
"learning_rate": 1.785714285714286e-05,
|
| 64 |
+
"loss": 1.0095,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.8017817371937639,
|
| 69 |
+
"grad_norm": 100.2014389038086,
|
| 70 |
+
"learning_rate": 2.0089285714285717e-05,
|
| 71 |
+
"loss": 1.0315,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.89086859688196,
|
| 76 |
+
"grad_norm": 40.66311264038086,
|
| 77 |
+
"learning_rate": 2.2321428571428575e-05,
|
| 78 |
+
"loss": 0.8269,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.9799554565701559,
|
| 83 |
+
"grad_norm": 82.63133239746094,
|
| 84 |
+
"learning_rate": 2.455357142857143e-05,
|
| 85 |
+
"loss": 0.8589,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.9977728285077951,
|
| 90 |
+
"eval_accuracy": 0.7507827175954915,
|
| 91 |
+
"eval_loss": 0.7458012104034424,
|
| 92 |
+
"eval_runtime": 21.2216,
|
| 93 |
+
"eval_samples_per_second": 75.253,
|
| 94 |
+
"eval_steps_per_second": 2.356,
|
| 95 |
+
"step": 112
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.069042316258352,
|
| 99 |
+
"grad_norm": 38.83303451538086,
|
| 100 |
+
"learning_rate": 2.6785714285714288e-05,
|
| 101 |
+
"loss": 0.8078,
|
| 102 |
+
"step": 120
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.158129175946548,
|
| 106 |
+
"grad_norm": 68.20454406738281,
|
| 107 |
+
"learning_rate": 2.9017857142857146e-05,
|
| 108 |
+
"loss": 0.8021,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.247216035634744,
|
| 113 |
+
"grad_norm": 46.07291030883789,
|
| 114 |
+
"learning_rate": 3.125e-05,
|
| 115 |
+
"loss": 0.8225,
|
| 116 |
+
"step": 140
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.3363028953229399,
|
| 120 |
+
"grad_norm": 72.90870666503906,
|
| 121 |
+
"learning_rate": 3.348214285714286e-05,
|
| 122 |
+
"loss": 0.8584,
|
| 123 |
+
"step": 150
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.4253897550111359,
|
| 127 |
+
"grad_norm": 56.07601547241211,
|
| 128 |
+
"learning_rate": 3.571428571428572e-05,
|
| 129 |
+
"loss": 0.894,
|
| 130 |
+
"step": 160
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.5144766146993318,
|
| 134 |
+
"grad_norm": 39.8024787902832,
|
| 135 |
+
"learning_rate": 3.794642857142857e-05,
|
| 136 |
+
"loss": 0.9406,
|
| 137 |
+
"step": 170
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.6035634743875278,
|
| 141 |
+
"grad_norm": 59.87102127075195,
|
| 142 |
+
"learning_rate": 4.017857142857143e-05,
|
| 143 |
+
"loss": 0.8083,
|
| 144 |
+
"step": 180
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.692650334075724,
|
| 148 |
+
"grad_norm": 74.46855926513672,
|
| 149 |
+
"learning_rate": 4.2410714285714285e-05,
|
| 150 |
+
"loss": 0.8802,
|
| 151 |
+
"step": 190
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.7817371937639197,
|
| 155 |
+
"grad_norm": 30.22767448425293,
|
| 156 |
+
"learning_rate": 4.464285714285715e-05,
|
| 157 |
+
"loss": 0.8672,
|
| 158 |
+
"step": 200
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.8708240534521159,
|
| 162 |
+
"grad_norm": 32.281150817871094,
|
| 163 |
+
"learning_rate": 4.6875e-05,
|
| 164 |
+
"loss": 0.8663,
|
| 165 |
+
"step": 210
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.9599109131403119,
|
| 169 |
+
"grad_norm": 37.24480056762695,
|
| 170 |
+
"learning_rate": 4.910714285714286e-05,
|
| 171 |
+
"loss": 0.8465,
|
| 172 |
+
"step": 220
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.9955456570155903,
|
| 176 |
+
"eval_accuracy": 0.7864746399499061,
|
| 177 |
+
"eval_loss": 0.656895637512207,
|
| 178 |
+
"eval_runtime": 21.1029,
|
| 179 |
+
"eval_samples_per_second": 75.677,
|
| 180 |
+
"eval_steps_per_second": 2.369,
|
| 181 |
+
"step": 224
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.048997772828508,
|
| 185 |
+
"grad_norm": 40.045684814453125,
|
| 186 |
+
"learning_rate": 4.985119047619048e-05,
|
| 187 |
+
"loss": 0.7752,
|
| 188 |
+
"step": 230
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.138084632516704,
|
| 192 |
+
"grad_norm": 26.04545783996582,
|
| 193 |
+
"learning_rate": 4.960317460317461e-05,
|
| 194 |
+
"loss": 0.7703,
|
| 195 |
+
"step": 240
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.2271714922048997,
|
| 199 |
+
"grad_norm": 38.471458435058594,
|
| 200 |
+
"learning_rate": 4.9355158730158735e-05,
|
| 201 |
+
"loss": 0.8756,
|
| 202 |
+
"step": 250
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.316258351893096,
|
| 206 |
+
"grad_norm": 22.920825958251953,
|
| 207 |
+
"learning_rate": 4.910714285714286e-05,
|
| 208 |
+
"loss": 0.8719,
|
| 209 |
+
"step": 260
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.4053452115812917,
|
| 213 |
+
"grad_norm": 25.196147918701172,
|
| 214 |
+
"learning_rate": 4.8859126984126984e-05,
|
| 215 |
+
"loss": 0.8446,
|
| 216 |
+
"step": 270
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.494432071269488,
|
| 220 |
+
"grad_norm": 37.12013626098633,
|
| 221 |
+
"learning_rate": 4.8611111111111115e-05,
|
| 222 |
+
"loss": 0.7929,
|
| 223 |
+
"step": 280
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 2.5835189309576836,
|
| 227 |
+
"grad_norm": 40.25604248046875,
|
| 228 |
+
"learning_rate": 4.836309523809524e-05,
|
| 229 |
+
"loss": 0.7952,
|
| 230 |
+
"step": 290
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.6726057906458798,
|
| 234 |
+
"grad_norm": 33.80072021484375,
|
| 235 |
+
"learning_rate": 4.811507936507937e-05,
|
| 236 |
+
"loss": 0.7557,
|
| 237 |
+
"step": 300
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 2.7616926503340755,
|
| 241 |
+
"grad_norm": 18.884248733520508,
|
| 242 |
+
"learning_rate": 4.7867063492063496e-05,
|
| 243 |
+
"loss": 0.8124,
|
| 244 |
+
"step": 310
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 2.8507795100222717,
|
| 248 |
+
"grad_norm": 23.225616455078125,
|
| 249 |
+
"learning_rate": 4.761904761904762e-05,
|
| 250 |
+
"loss": 0.71,
|
| 251 |
+
"step": 320
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 2.939866369710468,
|
| 255 |
+
"grad_norm": 23.466827392578125,
|
| 256 |
+
"learning_rate": 4.7371031746031745e-05,
|
| 257 |
+
"loss": 0.732,
|
| 258 |
+
"step": 330
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.9933184855233854,
|
| 262 |
+
"eval_accuracy": 0.7595491546649968,
|
| 263 |
+
"eval_loss": 0.7020854353904724,
|
| 264 |
+
"eval_runtime": 21.064,
|
| 265 |
+
"eval_samples_per_second": 75.816,
|
| 266 |
+
"eval_steps_per_second": 2.374,
|
| 267 |
+
"step": 336
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 3.0289532293986636,
|
| 271 |
+
"grad_norm": 23.42888069152832,
|
| 272 |
+
"learning_rate": 4.7123015873015876e-05,
|
| 273 |
+
"loss": 0.7415,
|
| 274 |
+
"step": 340
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 3.11804008908686,
|
| 278 |
+
"grad_norm": 35.59011459350586,
|
| 279 |
+
"learning_rate": 4.6875e-05,
|
| 280 |
+
"loss": 0.7201,
|
| 281 |
+
"step": 350
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 3.2071269487750556,
|
| 285 |
+
"grad_norm": 39.206329345703125,
|
| 286 |
+
"learning_rate": 4.662698412698413e-05,
|
| 287 |
+
"loss": 0.7227,
|
| 288 |
+
"step": 360
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 3.2962138084632517,
|
| 292 |
+
"grad_norm": 25.05638885498047,
|
| 293 |
+
"learning_rate": 4.637896825396826e-05,
|
| 294 |
+
"loss": 0.7214,
|
| 295 |
+
"step": 370
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 3.3853006681514475,
|
| 299 |
+
"grad_norm": 19.9650936126709,
|
| 300 |
+
"learning_rate": 4.613095238095239e-05,
|
| 301 |
+
"loss": 0.7479,
|
| 302 |
+
"step": 380
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 3.4743875278396437,
|
| 306 |
+
"grad_norm": 30.844806671142578,
|
| 307 |
+
"learning_rate": 4.5882936507936506e-05,
|
| 308 |
+
"loss": 0.7454,
|
| 309 |
+
"step": 390
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 3.5634743875278394,
|
| 313 |
+
"grad_norm": 16.321706771850586,
|
| 314 |
+
"learning_rate": 4.563492063492064e-05,
|
| 315 |
+
"loss": 0.7032,
|
| 316 |
+
"step": 400
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 3.6525612472160356,
|
| 320 |
+
"grad_norm": 15.168939590454102,
|
| 321 |
+
"learning_rate": 4.538690476190476e-05,
|
| 322 |
+
"loss": 0.6498,
|
| 323 |
+
"step": 410
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 3.7416481069042318,
|
| 327 |
+
"grad_norm": 30.41251564025879,
|
| 328 |
+
"learning_rate": 4.5138888888888894e-05,
|
| 329 |
+
"loss": 0.7092,
|
| 330 |
+
"step": 420
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 3.8307349665924275,
|
| 334 |
+
"grad_norm": 22.254060745239258,
|
| 335 |
+
"learning_rate": 4.489087301587302e-05,
|
| 336 |
+
"loss": 0.7199,
|
| 337 |
+
"step": 430
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 3.9198218262806237,
|
| 341 |
+
"grad_norm": 17.9412841796875,
|
| 342 |
+
"learning_rate": 4.464285714285715e-05,
|
| 343 |
+
"loss": 0.7083,
|
| 344 |
+
"step": 440
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 4.0,
|
| 348 |
+
"eval_accuracy": 0.7883531621790858,
|
| 349 |
+
"eval_loss": 0.6093864440917969,
|
| 350 |
+
"eval_runtime": 21.2086,
|
| 351 |
+
"eval_samples_per_second": 75.3,
|
| 352 |
+
"eval_steps_per_second": 2.358,
|
| 353 |
+
"step": 449
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 4.008908685968819,
|
| 357 |
+
"grad_norm": 20.122547149658203,
|
| 358 |
+
"learning_rate": 4.439484126984127e-05,
|
| 359 |
+
"loss": 0.6959,
|
| 360 |
+
"step": 450
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 4.097995545657016,
|
| 364 |
+
"grad_norm": 20.436920166015625,
|
| 365 |
+
"learning_rate": 4.41468253968254e-05,
|
| 366 |
+
"loss": 0.5657,
|
| 367 |
+
"step": 460
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 4.187082405345212,
|
| 371 |
+
"grad_norm": 18.810733795166016,
|
| 372 |
+
"learning_rate": 4.3898809523809523e-05,
|
| 373 |
+
"loss": 0.645,
|
| 374 |
+
"step": 470
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 4.276169265033408,
|
| 378 |
+
"grad_norm": 19.18097496032715,
|
| 379 |
+
"learning_rate": 4.3650793650793655e-05,
|
| 380 |
+
"loss": 0.7224,
|
| 381 |
+
"step": 480
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 4.365256124721603,
|
| 385 |
+
"grad_norm": 14.759562492370605,
|
| 386 |
+
"learning_rate": 4.340277777777778e-05,
|
| 387 |
+
"loss": 0.6456,
|
| 388 |
+
"step": 490
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 4.4543429844097995,
|
| 392 |
+
"grad_norm": 26.570215225219727,
|
| 393 |
+
"learning_rate": 4.315476190476191e-05,
|
| 394 |
+
"loss": 0.7037,
|
| 395 |
+
"step": 500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 4.543429844097996,
|
| 399 |
+
"grad_norm": 22.715709686279297,
|
| 400 |
+
"learning_rate": 4.290674603174603e-05,
|
| 401 |
+
"loss": 0.7042,
|
| 402 |
+
"step": 510
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 4.632516703786192,
|
| 406 |
+
"grad_norm": 22.303823471069336,
|
| 407 |
+
"learning_rate": 4.265873015873016e-05,
|
| 408 |
+
"loss": 0.64,
|
| 409 |
+
"step": 520
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 4.721603563474387,
|
| 413 |
+
"grad_norm": 24.049057006835938,
|
| 414 |
+
"learning_rate": 4.2410714285714285e-05,
|
| 415 |
+
"loss": 0.6607,
|
| 416 |
+
"step": 530
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.810690423162583,
|
| 420 |
+
"grad_norm": 24.192787170410156,
|
| 421 |
+
"learning_rate": 4.2162698412698416e-05,
|
| 422 |
+
"loss": 0.6641,
|
| 423 |
+
"step": 540
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 4.8997772828507795,
|
| 427 |
+
"grad_norm": 18.463960647583008,
|
| 428 |
+
"learning_rate": 4.191468253968254e-05,
|
| 429 |
+
"loss": 0.6725,
|
| 430 |
+
"step": 550
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 4.988864142538976,
|
| 434 |
+
"grad_norm": 14.166847229003906,
|
| 435 |
+
"learning_rate": 4.166666666666667e-05,
|
| 436 |
+
"loss": 0.662,
|
| 437 |
+
"step": 560
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 4.997772828507795,
|
| 441 |
+
"eval_accuracy": 0.817783343769568,
|
| 442 |
+
"eval_loss": 0.5464382767677307,
|
| 443 |
+
"eval_runtime": 21.2366,
|
| 444 |
+
"eval_samples_per_second": 75.2,
|
| 445 |
+
"eval_steps_per_second": 2.354,
|
| 446 |
+
"step": 561
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 5.077951002227172,
|
| 450 |
+
"grad_norm": 17.516250610351562,
|
| 451 |
+
"learning_rate": 4.14186507936508e-05,
|
| 452 |
+
"loss": 0.5613,
|
| 453 |
+
"step": 570
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 5.167037861915367,
|
| 457 |
+
"grad_norm": 16.576993942260742,
|
| 458 |
+
"learning_rate": 4.117063492063492e-05,
|
| 459 |
+
"loss": 0.6792,
|
| 460 |
+
"step": 580
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 5.256124721603563,
|
| 464 |
+
"grad_norm": 21.692968368530273,
|
| 465 |
+
"learning_rate": 4.0922619047619046e-05,
|
| 466 |
+
"loss": 0.6623,
|
| 467 |
+
"step": 590
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 5.3452115812917596,
|
| 471 |
+
"grad_norm": 19.450206756591797,
|
| 472 |
+
"learning_rate": 4.067460317460318e-05,
|
| 473 |
+
"loss": 0.5972,
|
| 474 |
+
"step": 600
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 5.434298440979956,
|
| 478 |
+
"grad_norm": 15.104981422424316,
|
| 479 |
+
"learning_rate": 4.04265873015873e-05,
|
| 480 |
+
"loss": 0.5822,
|
| 481 |
+
"step": 610
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 5.523385300668151,
|
| 485 |
+
"grad_norm": 18.1156005859375,
|
| 486 |
+
"learning_rate": 4.017857142857143e-05,
|
| 487 |
+
"loss": 0.5892,
|
| 488 |
+
"step": 620
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 5.612472160356347,
|
| 492 |
+
"grad_norm": 17.83465576171875,
|
| 493 |
+
"learning_rate": 3.993055555555556e-05,
|
| 494 |
+
"loss": 0.5647,
|
| 495 |
+
"step": 630
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 5.701559020044543,
|
| 499 |
+
"grad_norm": 21.32401466369629,
|
| 500 |
+
"learning_rate": 3.968253968253968e-05,
|
| 501 |
+
"loss": 0.6736,
|
| 502 |
+
"step": 640
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 5.79064587973274,
|
| 506 |
+
"grad_norm": 20.872905731201172,
|
| 507 |
+
"learning_rate": 3.943452380952381e-05,
|
| 508 |
+
"loss": 0.6413,
|
| 509 |
+
"step": 650
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 5.879732739420936,
|
| 513 |
+
"grad_norm": 18.514385223388672,
|
| 514 |
+
"learning_rate": 3.918650793650794e-05,
|
| 515 |
+
"loss": 0.5748,
|
| 516 |
+
"step": 660
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 5.968819599109131,
|
| 520 |
+
"grad_norm": 15.119345664978027,
|
| 521 |
+
"learning_rate": 3.893849206349206e-05,
|
| 522 |
+
"loss": 0.6211,
|
| 523 |
+
"step": 670
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 5.99554565701559,
|
| 527 |
+
"eval_accuracy": 0.8134001252348153,
|
| 528 |
+
"eval_loss": 0.5666616559028625,
|
| 529 |
+
"eval_runtime": 21.3956,
|
| 530 |
+
"eval_samples_per_second": 74.642,
|
| 531 |
+
"eval_steps_per_second": 2.337,
|
| 532 |
+
"step": 673
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 6.057906458797327,
|
| 536 |
+
"grad_norm": 14.720823287963867,
|
| 537 |
+
"learning_rate": 3.8690476190476195e-05,
|
| 538 |
+
"loss": 0.5748,
|
| 539 |
+
"step": 680
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 6.146993318485523,
|
| 543 |
+
"grad_norm": 18.489421844482422,
|
| 544 |
+
"learning_rate": 3.844246031746032e-05,
|
| 545 |
+
"loss": 0.5077,
|
| 546 |
+
"step": 690
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 6.23608017817372,
|
| 550 |
+
"grad_norm": 20.35164451599121,
|
| 551 |
+
"learning_rate": 3.8194444444444444e-05,
|
| 552 |
+
"loss": 0.5337,
|
| 553 |
+
"step": 700
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 6.325167037861915,
|
| 557 |
+
"grad_norm": 19.843276977539062,
|
| 558 |
+
"learning_rate": 3.794642857142857e-05,
|
| 559 |
+
"loss": 0.5381,
|
| 560 |
+
"step": 710
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 6.414253897550111,
|
| 564 |
+
"grad_norm": 18.632654190063477,
|
| 565 |
+
"learning_rate": 3.76984126984127e-05,
|
| 566 |
+
"loss": 0.5889,
|
| 567 |
+
"step": 720
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 6.503340757238307,
|
| 571 |
+
"grad_norm": 17.311256408691406,
|
| 572 |
+
"learning_rate": 3.7450396825396824e-05,
|
| 573 |
+
"loss": 0.5198,
|
| 574 |
+
"step": 730
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 6.5924276169265035,
|
| 578 |
+
"grad_norm": 14.276023864746094,
|
| 579 |
+
"learning_rate": 3.7202380952380956e-05,
|
| 580 |
+
"loss": 0.5255,
|
| 581 |
+
"step": 740
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 6.6815144766147,
|
| 585 |
+
"grad_norm": 18.370847702026367,
|
| 586 |
+
"learning_rate": 3.695436507936508e-05,
|
| 587 |
+
"loss": 0.5429,
|
| 588 |
+
"step": 750
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 6.770601336302895,
|
| 592 |
+
"grad_norm": 15.048176765441895,
|
| 593 |
+
"learning_rate": 3.6706349206349205e-05,
|
| 594 |
+
"loss": 0.5933,
|
| 595 |
+
"step": 760
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 6.859688195991091,
|
| 599 |
+
"grad_norm": 23.03609275817871,
|
| 600 |
+
"learning_rate": 3.6458333333333336e-05,
|
| 601 |
+
"loss": 0.5956,
|
| 602 |
+
"step": 770
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 6.948775055679287,
|
| 606 |
+
"grad_norm": 16.367786407470703,
|
| 607 |
+
"learning_rate": 3.621031746031746e-05,
|
| 608 |
+
"loss": 0.5374,
|
| 609 |
+
"step": 780
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 6.993318485523385,
|
| 613 |
+
"eval_accuracy": 0.8296806512210394,
|
| 614 |
+
"eval_loss": 0.5134668946266174,
|
| 615 |
+
"eval_runtime": 21.1708,
|
| 616 |
+
"eval_samples_per_second": 75.434,
|
| 617 |
+
"eval_steps_per_second": 2.362,
|
| 618 |
+
"step": 785
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 7.0378619153674835,
|
| 622 |
+
"grad_norm": 14.078798294067383,
|
| 623 |
+
"learning_rate": 3.5962301587301586e-05,
|
| 624 |
+
"loss": 0.5302,
|
| 625 |
+
"step": 790
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 7.12694877505568,
|
| 629 |
+
"grad_norm": 16.1302433013916,
|
| 630 |
+
"learning_rate": 3.571428571428572e-05,
|
| 631 |
+
"loss": 0.5558,
|
| 632 |
+
"step": 800
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 7.216035634743875,
|
| 636 |
+
"grad_norm": 13.907476425170898,
|
| 637 |
+
"learning_rate": 3.546626984126984e-05,
|
| 638 |
+
"loss": 0.5582,
|
| 639 |
+
"step": 810
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 7.305122494432071,
|
| 643 |
+
"grad_norm": 17.780229568481445,
|
| 644 |
+
"learning_rate": 3.521825396825397e-05,
|
| 645 |
+
"loss": 0.5502,
|
| 646 |
+
"step": 820
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 7.394209354120267,
|
| 650 |
+
"grad_norm": 26.133169174194336,
|
| 651 |
+
"learning_rate": 3.49702380952381e-05,
|
| 652 |
+
"loss": 0.5396,
|
| 653 |
+
"step": 830
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 7.4832962138084635,
|
| 657 |
+
"grad_norm": 13.426424980163574,
|
| 658 |
+
"learning_rate": 3.472222222222222e-05,
|
| 659 |
+
"loss": 0.5536,
|
| 660 |
+
"step": 840
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 7.57238307349666,
|
| 664 |
+
"grad_norm": 24.7213077545166,
|
| 665 |
+
"learning_rate": 3.4474206349206354e-05,
|
| 666 |
+
"loss": 0.5081,
|
| 667 |
+
"step": 850
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 7.661469933184855,
|
| 671 |
+
"grad_norm": 20.476024627685547,
|
| 672 |
+
"learning_rate": 3.422619047619048e-05,
|
| 673 |
+
"loss": 0.5186,
|
| 674 |
+
"step": 860
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 7.750556792873051,
|
| 678 |
+
"grad_norm": 14.186267852783203,
|
| 679 |
+
"learning_rate": 3.397817460317461e-05,
|
| 680 |
+
"loss": 0.5517,
|
| 681 |
+
"step": 870
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 7.839643652561247,
|
| 685 |
+
"grad_norm": 17.4648494720459,
|
| 686 |
+
"learning_rate": 3.3730158730158734e-05,
|
| 687 |
+
"loss": 0.546,
|
| 688 |
+
"step": 880
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 7.928730512249444,
|
| 692 |
+
"grad_norm": 17.170608520507812,
|
| 693 |
+
"learning_rate": 3.348214285714286e-05,
|
| 694 |
+
"loss": 0.5147,
|
| 695 |
+
"step": 890
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 8.0,
|
| 699 |
+
"eval_accuracy": 0.828428303068253,
|
| 700 |
+
"eval_loss": 0.5184165835380554,
|
| 701 |
+
"eval_runtime": 21.3041,
|
| 702 |
+
"eval_samples_per_second": 74.962,
|
| 703 |
+
"eval_steps_per_second": 2.347,
|
| 704 |
+
"step": 898
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 8.017817371937639,
|
| 708 |
+
"grad_norm": 15.437606811523438,
|
| 709 |
+
"learning_rate": 3.3234126984126983e-05,
|
| 710 |
+
"loss": 0.5755,
|
| 711 |
+
"step": 900
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 8.106904231625835,
|
| 715 |
+
"grad_norm": 13.340350151062012,
|
| 716 |
+
"learning_rate": 3.2986111111111115e-05,
|
| 717 |
+
"loss": 0.4563,
|
| 718 |
+
"step": 910
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 8.195991091314031,
|
| 722 |
+
"grad_norm": 11.855537414550781,
|
| 723 |
+
"learning_rate": 3.273809523809524e-05,
|
| 724 |
+
"loss": 0.4695,
|
| 725 |
+
"step": 920
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 8.285077951002227,
|
| 729 |
+
"grad_norm": 19.414112091064453,
|
| 730 |
+
"learning_rate": 3.249007936507937e-05,
|
| 731 |
+
"loss": 0.4728,
|
| 732 |
+
"step": 930
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 8.374164810690424,
|
| 736 |
+
"grad_norm": 14.449013710021973,
|
| 737 |
+
"learning_rate": 3.2242063492063495e-05,
|
| 738 |
+
"loss": 0.5105,
|
| 739 |
+
"step": 940
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 8.46325167037862,
|
| 743 |
+
"grad_norm": 12.747797966003418,
|
| 744 |
+
"learning_rate": 3.199404761904762e-05,
|
| 745 |
+
"loss": 0.5009,
|
| 746 |
+
"step": 950
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 8.552338530066816,
|
| 750 |
+
"grad_norm": 17.288352966308594,
|
| 751 |
+
"learning_rate": 3.1746031746031745e-05,
|
| 752 |
+
"loss": 0.5235,
|
| 753 |
+
"step": 960
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 8.64142538975501,
|
| 757 |
+
"grad_norm": 10.824517250061035,
|
| 758 |
+
"learning_rate": 3.1498015873015876e-05,
|
| 759 |
+
"loss": 0.4365,
|
| 760 |
+
"step": 970
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 8.730512249443207,
|
| 764 |
+
"grad_norm": 11.664321899414062,
|
| 765 |
+
"learning_rate": 3.125e-05,
|
| 766 |
+
"loss": 0.4819,
|
| 767 |
+
"step": 980
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 8.819599109131403,
|
| 771 |
+
"grad_norm": 15.481782913208008,
|
| 772 |
+
"learning_rate": 3.100198412698413e-05,
|
| 773 |
+
"loss": 0.4624,
|
| 774 |
+
"step": 990
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 8.908685968819599,
|
| 778 |
+
"grad_norm": 15.801815032958984,
|
| 779 |
+
"learning_rate": 3.075396825396826e-05,
|
| 780 |
+
"loss": 0.4689,
|
| 781 |
+
"step": 1000
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 8.997772828507795,
|
| 785 |
+
"grad_norm": 12.902999877929688,
|
| 786 |
+
"learning_rate": 3.0505952380952385e-05,
|
| 787 |
+
"loss": 0.5101,
|
| 788 |
+
"step": 1010
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 8.997772828507795,
|
| 792 |
+
"eval_accuracy": 0.8090169067000627,
|
| 793 |
+
"eval_loss": 0.5775282382965088,
|
| 794 |
+
"eval_runtime": 21.2996,
|
| 795 |
+
"eval_samples_per_second": 74.978,
|
| 796 |
+
"eval_steps_per_second": 2.347,
|
| 797 |
+
"step": 1010
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 9.086859688195991,
|
| 801 |
+
"grad_norm": 14.9346342086792,
|
| 802 |
+
"learning_rate": 3.0257936507936506e-05,
|
| 803 |
+
"loss": 0.4874,
|
| 804 |
+
"step": 1020
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 9.175946547884188,
|
| 808 |
+
"grad_norm": 13.989501953125,
|
| 809 |
+
"learning_rate": 3.0009920634920634e-05,
|
| 810 |
+
"loss": 0.4677,
|
| 811 |
+
"step": 1030
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 9.265033407572384,
|
| 815 |
+
"grad_norm": 13.029789924621582,
|
| 816 |
+
"learning_rate": 2.9761904761904762e-05,
|
| 817 |
+
"loss": 0.4558,
|
| 818 |
+
"step": 1040
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 9.35412026726058,
|
| 822 |
+
"grad_norm": 13.129937171936035,
|
| 823 |
+
"learning_rate": 2.951388888888889e-05,
|
| 824 |
+
"loss": 0.477,
|
| 825 |
+
"step": 1050
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 9.443207126948774,
|
| 829 |
+
"grad_norm": 18.104257583618164,
|
| 830 |
+
"learning_rate": 2.9265873015873018e-05,
|
| 831 |
+
"loss": 0.4678,
|
| 832 |
+
"step": 1060
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 9.53229398663697,
|
| 836 |
+
"grad_norm": 14.129535675048828,
|
| 837 |
+
"learning_rate": 2.9017857142857146e-05,
|
| 838 |
+
"loss": 0.5203,
|
| 839 |
+
"step": 1070
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 9.621380846325167,
|
| 843 |
+
"grad_norm": 12.165337562561035,
|
| 844 |
+
"learning_rate": 2.876984126984127e-05,
|
| 845 |
+
"loss": 0.4555,
|
| 846 |
+
"step": 1080
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 9.710467706013363,
|
| 850 |
+
"grad_norm": 11.243370056152344,
|
| 851 |
+
"learning_rate": 2.8521825396825395e-05,
|
| 852 |
+
"loss": 0.4713,
|
| 853 |
+
"step": 1090
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 9.799554565701559,
|
| 857 |
+
"grad_norm": 13.38561725616455,
|
| 858 |
+
"learning_rate": 2.8273809523809523e-05,
|
| 859 |
+
"loss": 0.4461,
|
| 860 |
+
"step": 1100
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 9.888641425389755,
|
| 864 |
+
"grad_norm": 24.981054306030273,
|
| 865 |
+
"learning_rate": 2.802579365079365e-05,
|
| 866 |
+
"loss": 0.4867,
|
| 867 |
+
"step": 1110
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 9.977728285077951,
|
| 871 |
+
"grad_norm": 15.780611991882324,
|
| 872 |
+
"learning_rate": 2.777777777777778e-05,
|
| 873 |
+
"loss": 0.4791,
|
| 874 |
+
"step": 1120
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 9.99554565701559,
|
| 878 |
+
"eval_accuracy": 0.8278021289918598,
|
| 879 |
+
"eval_loss": 0.49649718403816223,
|
| 880 |
+
"eval_runtime": 21.0785,
|
| 881 |
+
"eval_samples_per_second": 75.765,
|
| 882 |
+
"eval_steps_per_second": 2.372,
|
| 883 |
+
"step": 1122
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 10.066815144766148,
|
| 887 |
+
"grad_norm": 20.694881439208984,
|
| 888 |
+
"learning_rate": 2.7529761904761907e-05,
|
| 889 |
+
"loss": 0.4615,
|
| 890 |
+
"step": 1130
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 10.155902004454344,
|
| 894 |
+
"grad_norm": 7.557748317718506,
|
| 895 |
+
"learning_rate": 2.7281746031746032e-05,
|
| 896 |
+
"loss": 0.5191,
|
| 897 |
+
"step": 1140
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 10.244988864142538,
|
| 901 |
+
"grad_norm": 21.552391052246094,
|
| 902 |
+
"learning_rate": 2.703373015873016e-05,
|
| 903 |
+
"loss": 0.4283,
|
| 904 |
+
"step": 1150
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 10.334075723830734,
|
| 908 |
+
"grad_norm": 16.425872802734375,
|
| 909 |
+
"learning_rate": 2.6785714285714288e-05,
|
| 910 |
+
"loss": 0.4424,
|
| 911 |
+
"step": 1160
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 10.42316258351893,
|
| 915 |
+
"grad_norm": 18.903982162475586,
|
| 916 |
+
"learning_rate": 2.6537698412698416e-05,
|
| 917 |
+
"loss": 0.4488,
|
| 918 |
+
"step": 1170
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 10.512249443207127,
|
| 922 |
+
"grad_norm": 20.344375610351562,
|
| 923 |
+
"learning_rate": 2.628968253968254e-05,
|
| 924 |
+
"loss": 0.4788,
|
| 925 |
+
"step": 1180
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 10.601336302895323,
|
| 929 |
+
"grad_norm": 9.902765274047852,
|
| 930 |
+
"learning_rate": 2.604166666666667e-05,
|
| 931 |
+
"loss": 0.4644,
|
| 932 |
+
"step": 1190
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 10.690423162583519,
|
| 936 |
+
"grad_norm": 14.26260757446289,
|
| 937 |
+
"learning_rate": 2.5793650793650796e-05,
|
| 938 |
+
"loss": 0.3828,
|
| 939 |
+
"step": 1200
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 10.779510022271715,
|
| 943 |
+
"grad_norm": 10.488040924072266,
|
| 944 |
+
"learning_rate": 2.554563492063492e-05,
|
| 945 |
+
"loss": 0.4348,
|
| 946 |
+
"step": 1210
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 10.868596881959911,
|
| 950 |
+
"grad_norm": 19.645410537719727,
|
| 951 |
+
"learning_rate": 2.529761904761905e-05,
|
| 952 |
+
"loss": 0.4537,
|
| 953 |
+
"step": 1220
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 10.957683741648108,
|
| 957 |
+
"grad_norm": 13.607442855834961,
|
| 958 |
+
"learning_rate": 2.5049603174603177e-05,
|
| 959 |
+
"loss": 0.4339,
|
| 960 |
+
"step": 1230
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 10.993318485523385,
|
| 964 |
+
"eval_accuracy": 0.8303068252974327,
|
| 965 |
+
"eval_loss": 0.5053770542144775,
|
| 966 |
+
"eval_runtime": 21.1844,
|
| 967 |
+
"eval_samples_per_second": 75.386,
|
| 968 |
+
"eval_steps_per_second": 2.36,
|
| 969 |
+
"step": 1234
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 11.046770601336302,
|
| 973 |
+
"grad_norm": 10.476160049438477,
|
| 974 |
+
"learning_rate": 2.4801587301587305e-05,
|
| 975 |
+
"loss": 0.3619,
|
| 976 |
+
"step": 1240
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 11.135857461024498,
|
| 980 |
+
"grad_norm": 12.624418258666992,
|
| 981 |
+
"learning_rate": 2.455357142857143e-05,
|
| 982 |
+
"loss": 0.3775,
|
| 983 |
+
"step": 1250
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 11.224944320712694,
|
| 987 |
+
"grad_norm": 14.461904525756836,
|
| 988 |
+
"learning_rate": 2.4305555555555558e-05,
|
| 989 |
+
"loss": 0.447,
|
| 990 |
+
"step": 1260
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 11.31403118040089,
|
| 994 |
+
"grad_norm": 12.849555969238281,
|
| 995 |
+
"learning_rate": 2.4057539682539686e-05,
|
| 996 |
+
"loss": 0.3806,
|
| 997 |
+
"step": 1270
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 11.403118040089087,
|
| 1001 |
+
"grad_norm": 13.973437309265137,
|
| 1002 |
+
"learning_rate": 2.380952380952381e-05,
|
| 1003 |
+
"loss": 0.4338,
|
| 1004 |
+
"step": 1280
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 11.492204899777283,
|
| 1008 |
+
"grad_norm": 9.459921836853027,
|
| 1009 |
+
"learning_rate": 2.3561507936507938e-05,
|
| 1010 |
+
"loss": 0.4273,
|
| 1011 |
+
"step": 1290
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 11.58129175946548,
|
| 1015 |
+
"grad_norm": 19.047231674194336,
|
| 1016 |
+
"learning_rate": 2.3313492063492066e-05,
|
| 1017 |
+
"loss": 0.4625,
|
| 1018 |
+
"step": 1300
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 11.670378619153675,
|
| 1022 |
+
"grad_norm": 14.69473934173584,
|
| 1023 |
+
"learning_rate": 2.3065476190476194e-05,
|
| 1024 |
+
"loss": 0.434,
|
| 1025 |
+
"step": 1310
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 11.759465478841872,
|
| 1029 |
+
"grad_norm": 12.16936206817627,
|
| 1030 |
+
"learning_rate": 2.281746031746032e-05,
|
| 1031 |
+
"loss": 0.4532,
|
| 1032 |
+
"step": 1320
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 11.848552338530066,
|
| 1036 |
+
"grad_norm": 11.761576652526855,
|
| 1037 |
+
"learning_rate": 2.2569444444444447e-05,
|
| 1038 |
+
"loss": 0.3755,
|
| 1039 |
+
"step": 1330
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 11.937639198218262,
|
| 1043 |
+
"grad_norm": 15.081666946411133,
|
| 1044 |
+
"learning_rate": 2.2321428571428575e-05,
|
| 1045 |
+
"loss": 0.4709,
|
| 1046 |
+
"step": 1340
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 12.0,
|
| 1050 |
+
"eval_accuracy": 0.8033813400125235,
|
| 1051 |
+
"eval_loss": 0.6274072527885437,
|
| 1052 |
+
"eval_runtime": 21.1787,
|
| 1053 |
+
"eval_samples_per_second": 75.406,
|
| 1054 |
+
"eval_steps_per_second": 2.361,
|
| 1055 |
+
"step": 1347
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 12.026726057906458,
|
| 1059 |
+
"grad_norm": 15.724143028259277,
|
| 1060 |
+
"learning_rate": 2.20734126984127e-05,
|
| 1061 |
+
"loss": 0.4501,
|
| 1062 |
+
"step": 1350
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 12.115812917594655,
|
| 1066 |
+
"grad_norm": 9.250158309936523,
|
| 1067 |
+
"learning_rate": 2.1825396825396827e-05,
|
| 1068 |
+
"loss": 0.3923,
|
| 1069 |
+
"step": 1360
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 12.20489977728285,
|
| 1073 |
+
"grad_norm": 17.66366195678711,
|
| 1074 |
+
"learning_rate": 2.1577380952380955e-05,
|
| 1075 |
+
"loss": 0.3874,
|
| 1076 |
+
"step": 1370
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 12.293986636971047,
|
| 1080 |
+
"grad_norm": 9.035892486572266,
|
| 1081 |
+
"learning_rate": 2.132936507936508e-05,
|
| 1082 |
+
"loss": 0.4155,
|
| 1083 |
+
"step": 1380
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 12.383073496659243,
|
| 1087 |
+
"grad_norm": 14.44294548034668,
|
| 1088 |
+
"learning_rate": 2.1081349206349208e-05,
|
| 1089 |
+
"loss": 0.4047,
|
| 1090 |
+
"step": 1390
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 12.47216035634744,
|
| 1094 |
+
"grad_norm": 12.312219619750977,
|
| 1095 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 1096 |
+
"loss": 0.4204,
|
| 1097 |
+
"step": 1400
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 12.561247216035635,
|
| 1101 |
+
"grad_norm": 11.931431770324707,
|
| 1102 |
+
"learning_rate": 2.058531746031746e-05,
|
| 1103 |
+
"loss": 0.4006,
|
| 1104 |
+
"step": 1410
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 12.65033407572383,
|
| 1108 |
+
"grad_norm": 9.212520599365234,
|
| 1109 |
+
"learning_rate": 2.033730158730159e-05,
|
| 1110 |
+
"loss": 0.3722,
|
| 1111 |
+
"step": 1420
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 12.739420935412026,
|
| 1115 |
+
"grad_norm": 13.458575248718262,
|
| 1116 |
+
"learning_rate": 2.0089285714285717e-05,
|
| 1117 |
+
"loss": 0.3816,
|
| 1118 |
+
"step": 1430
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 12.828507795100222,
|
| 1122 |
+
"grad_norm": 9.76841926574707,
|
| 1123 |
+
"learning_rate": 1.984126984126984e-05,
|
| 1124 |
+
"loss": 0.3839,
|
| 1125 |
+
"step": 1440
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 12.917594654788418,
|
| 1129 |
+
"grad_norm": 15.495878219604492,
|
| 1130 |
+
"learning_rate": 1.959325396825397e-05,
|
| 1131 |
+
"loss": 0.4092,
|
| 1132 |
+
"step": 1450
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 12.997772828507795,
|
| 1136 |
+
"eval_accuracy": 0.8484658735128365,
|
| 1137 |
+
"eval_loss": 0.47939950227737427,
|
| 1138 |
+
"eval_runtime": 21.0587,
|
| 1139 |
+
"eval_samples_per_second": 75.836,
|
| 1140 |
+
"eval_steps_per_second": 2.374,
|
| 1141 |
+
"step": 1459
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 13.006681514476615,
|
| 1145 |
+
"grad_norm": 10.429738998413086,
|
| 1146 |
+
"learning_rate": 1.9345238095238097e-05,
|
| 1147 |
+
"loss": 0.3871,
|
| 1148 |
+
"step": 1460
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 13.09576837416481,
|
| 1152 |
+
"grad_norm": 10.946513175964355,
|
| 1153 |
+
"learning_rate": 1.9097222222222222e-05,
|
| 1154 |
+
"loss": 0.348,
|
| 1155 |
+
"step": 1470
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 13.184855233853007,
|
| 1159 |
+
"grad_norm": 9.279400825500488,
|
| 1160 |
+
"learning_rate": 1.884920634920635e-05,
|
| 1161 |
+
"loss": 0.3221,
|
| 1162 |
+
"step": 1480
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 13.273942093541203,
|
| 1166 |
+
"grad_norm": 13.357484817504883,
|
| 1167 |
+
"learning_rate": 1.8601190476190478e-05,
|
| 1168 |
+
"loss": 0.3058,
|
| 1169 |
+
"step": 1490
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 13.3630289532294,
|
| 1173 |
+
"grad_norm": 13.243049621582031,
|
| 1174 |
+
"learning_rate": 1.8353174603174602e-05,
|
| 1175 |
+
"loss": 0.3596,
|
| 1176 |
+
"step": 1500
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 13.452115812917596,
|
| 1180 |
+
"grad_norm": 9.535561561584473,
|
| 1181 |
+
"learning_rate": 1.810515873015873e-05,
|
| 1182 |
+
"loss": 0.34,
|
| 1183 |
+
"step": 1510
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 13.54120267260579,
|
| 1187 |
+
"grad_norm": 11.322809219360352,
|
| 1188 |
+
"learning_rate": 1.785714285714286e-05,
|
| 1189 |
+
"loss": 0.3468,
|
| 1190 |
+
"step": 1520
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 13.630289532293986,
|
| 1194 |
+
"grad_norm": 13.22533130645752,
|
| 1195 |
+
"learning_rate": 1.7609126984126986e-05,
|
| 1196 |
+
"loss": 0.3749,
|
| 1197 |
+
"step": 1530
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 13.719376391982182,
|
| 1201 |
+
"grad_norm": 9.667810440063477,
|
| 1202 |
+
"learning_rate": 1.736111111111111e-05,
|
| 1203 |
+
"loss": 0.3601,
|
| 1204 |
+
"step": 1540
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 13.808463251670378,
|
| 1208 |
+
"grad_norm": 16.140737533569336,
|
| 1209 |
+
"learning_rate": 1.711309523809524e-05,
|
| 1210 |
+
"loss": 0.3634,
|
| 1211 |
+
"step": 1550
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 13.897550111358575,
|
| 1215 |
+
"grad_norm": 13.473885536193848,
|
| 1216 |
+
"learning_rate": 1.6865079365079367e-05,
|
| 1217 |
+
"loss": 0.3983,
|
| 1218 |
+
"step": 1560
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 13.98663697104677,
|
| 1222 |
+
"grad_norm": 13.371350288391113,
|
| 1223 |
+
"learning_rate": 1.6617063492063492e-05,
|
| 1224 |
+
"loss": 0.368,
|
| 1225 |
+
"step": 1570
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 13.99554565701559,
|
| 1229 |
+
"eval_accuracy": 0.8465873512836568,
|
| 1230 |
+
"eval_loss": 0.5031822919845581,
|
| 1231 |
+
"eval_runtime": 21.1723,
|
| 1232 |
+
"eval_samples_per_second": 75.429,
|
| 1233 |
+
"eval_steps_per_second": 2.362,
|
| 1234 |
+
"step": 1571
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 14.075723830734967,
|
| 1238 |
+
"grad_norm": 11.687047958374023,
|
| 1239 |
+
"learning_rate": 1.636904761904762e-05,
|
| 1240 |
+
"loss": 0.3465,
|
| 1241 |
+
"step": 1580
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 14.164810690423163,
|
| 1245 |
+
"grad_norm": 13.193872451782227,
|
| 1246 |
+
"learning_rate": 1.6121031746031748e-05,
|
| 1247 |
+
"loss": 0.3402,
|
| 1248 |
+
"step": 1590
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 14.25389755011136,
|
| 1252 |
+
"grad_norm": 15.279977798461914,
|
| 1253 |
+
"learning_rate": 1.5873015873015872e-05,
|
| 1254 |
+
"loss": 0.358,
|
| 1255 |
+
"step": 1600
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 14.342984409799554,
|
| 1259 |
+
"grad_norm": 10.373139381408691,
|
| 1260 |
+
"learning_rate": 1.5625e-05,
|
| 1261 |
+
"loss": 0.3423,
|
| 1262 |
+
"step": 1610
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 14.43207126948775,
|
| 1266 |
+
"grad_norm": 14.441068649291992,
|
| 1267 |
+
"learning_rate": 1.537698412698413e-05,
|
| 1268 |
+
"loss": 0.323,
|
| 1269 |
+
"step": 1620
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 14.521158129175946,
|
| 1273 |
+
"grad_norm": 13.793278694152832,
|
| 1274 |
+
"learning_rate": 1.5128968253968253e-05,
|
| 1275 |
+
"loss": 0.3443,
|
| 1276 |
+
"step": 1630
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 14.610244988864142,
|
| 1280 |
+
"grad_norm": 13.510929107666016,
|
| 1281 |
+
"learning_rate": 1.4880952380952381e-05,
|
| 1282 |
+
"loss": 0.3289,
|
| 1283 |
+
"step": 1640
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 14.699331848552339,
|
| 1287 |
+
"grad_norm": 14.0494384765625,
|
| 1288 |
+
"learning_rate": 1.4632936507936509e-05,
|
| 1289 |
+
"loss": 0.3266,
|
| 1290 |
+
"step": 1650
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 14.788418708240535,
|
| 1294 |
+
"grad_norm": 17.658891677856445,
|
| 1295 |
+
"learning_rate": 1.4384920634920635e-05,
|
| 1296 |
+
"loss": 0.343,
|
| 1297 |
+
"step": 1660
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 14.877505567928731,
|
| 1301 |
+
"grad_norm": 14.786270141601562,
|
| 1302 |
+
"learning_rate": 1.4136904761904762e-05,
|
| 1303 |
+
"loss": 0.3262,
|
| 1304 |
+
"step": 1670
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 14.966592427616927,
|
| 1308 |
+
"grad_norm": 11.525693893432617,
|
| 1309 |
+
"learning_rate": 1.388888888888889e-05,
|
| 1310 |
+
"loss": 0.3415,
|
| 1311 |
+
"step": 1680
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 14.993318485523385,
|
| 1315 |
+
"eval_accuracy": 0.8409517845961177,
|
| 1316 |
+
"eval_loss": 0.5257949829101562,
|
| 1317 |
+
"eval_runtime": 21.2165,
|
| 1318 |
+
"eval_samples_per_second": 75.272,
|
| 1319 |
+
"eval_steps_per_second": 2.357,
|
| 1320 |
+
"step": 1683
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 15.055679287305123,
|
| 1324 |
+
"grad_norm": 12.09720516204834,
|
| 1325 |
+
"learning_rate": 1.3640873015873016e-05,
|
| 1326 |
+
"loss": 0.276,
|
| 1327 |
+
"step": 1690
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 15.144766146993318,
|
| 1331 |
+
"grad_norm": 15.346843719482422,
|
| 1332 |
+
"learning_rate": 1.3392857142857144e-05,
|
| 1333 |
+
"loss": 0.3415,
|
| 1334 |
+
"step": 1700
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 15.233853006681514,
|
| 1338 |
+
"grad_norm": 10.313881874084473,
|
| 1339 |
+
"learning_rate": 1.314484126984127e-05,
|
| 1340 |
+
"loss": 0.2949,
|
| 1341 |
+
"step": 1710
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 15.32293986636971,
|
| 1345 |
+
"grad_norm": 11.033019065856934,
|
| 1346 |
+
"learning_rate": 1.2896825396825398e-05,
|
| 1347 |
+
"loss": 0.3653,
|
| 1348 |
+
"step": 1720
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 15.412026726057906,
|
| 1352 |
+
"grad_norm": 12.691286087036133,
|
| 1353 |
+
"learning_rate": 1.2648809523809524e-05,
|
| 1354 |
+
"loss": 0.3321,
|
| 1355 |
+
"step": 1730
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 15.501113585746102,
|
| 1359 |
+
"grad_norm": 8.7056884765625,
|
| 1360 |
+
"learning_rate": 1.2400793650793652e-05,
|
| 1361 |
+
"loss": 0.3351,
|
| 1362 |
+
"step": 1740
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 15.590200445434299,
|
| 1366 |
+
"grad_norm": 17.608694076538086,
|
| 1367 |
+
"learning_rate": 1.2152777777777779e-05,
|
| 1368 |
+
"loss": 0.2772,
|
| 1369 |
+
"step": 1750
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 15.679287305122495,
|
| 1373 |
+
"grad_norm": 12.325498580932617,
|
| 1374 |
+
"learning_rate": 1.1904761904761905e-05,
|
| 1375 |
+
"loss": 0.3475,
|
| 1376 |
+
"step": 1760
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 15.768374164810691,
|
| 1380 |
+
"grad_norm": 13.088592529296875,
|
| 1381 |
+
"learning_rate": 1.1656746031746033e-05,
|
| 1382 |
+
"loss": 0.3217,
|
| 1383 |
+
"step": 1770
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 15.857461024498887,
|
| 1387 |
+
"grad_norm": 10.744033813476562,
|
| 1388 |
+
"learning_rate": 1.140873015873016e-05,
|
| 1389 |
+
"loss": 0.3311,
|
| 1390 |
+
"step": 1780
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 15.946547884187082,
|
| 1394 |
+
"grad_norm": 12.45305347442627,
|
| 1395 |
+
"learning_rate": 1.1160714285714287e-05,
|
| 1396 |
+
"loss": 0.2963,
|
| 1397 |
+
"step": 1790
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 16.0,
|
| 1401 |
+
"eval_accuracy": 0.841577958672511,
|
| 1402 |
+
"eval_loss": 0.5382400155067444,
|
| 1403 |
+
"eval_runtime": 21.0587,
|
| 1404 |
+
"eval_samples_per_second": 75.836,
|
| 1405 |
+
"eval_steps_per_second": 2.374,
|
| 1406 |
+
"step": 1796
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 16.035634743875278,
|
| 1410 |
+
"grad_norm": 13.57647705078125,
|
| 1411 |
+
"learning_rate": 1.0912698412698414e-05,
|
| 1412 |
+
"loss": 0.3066,
|
| 1413 |
+
"step": 1800
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 16.124721603563476,
|
| 1417 |
+
"grad_norm": 13.313094139099121,
|
| 1418 |
+
"learning_rate": 1.066468253968254e-05,
|
| 1419 |
+
"loss": 0.2948,
|
| 1420 |
+
"step": 1810
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 16.21380846325167,
|
| 1424 |
+
"grad_norm": 10.375079154968262,
|
| 1425 |
+
"learning_rate": 1.0416666666666668e-05,
|
| 1426 |
+
"loss": 0.2852,
|
| 1427 |
+
"step": 1820
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 16.302895322939868,
|
| 1431 |
+
"grad_norm": 10.751969337463379,
|
| 1432 |
+
"learning_rate": 1.0168650793650794e-05,
|
| 1433 |
+
"loss": 0.3016,
|
| 1434 |
+
"step": 1830
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 16.391982182628063,
|
| 1438 |
+
"grad_norm": 9.340880393981934,
|
| 1439 |
+
"learning_rate": 9.92063492063492e-06,
|
| 1440 |
+
"loss": 0.2849,
|
| 1441 |
+
"step": 1840
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 16.481069042316257,
|
| 1445 |
+
"grad_norm": 11.936270713806152,
|
| 1446 |
+
"learning_rate": 9.672619047619049e-06,
|
| 1447 |
+
"loss": 0.2715,
|
| 1448 |
+
"step": 1850
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 16.570155902004455,
|
| 1452 |
+
"grad_norm": 12.299432754516602,
|
| 1453 |
+
"learning_rate": 9.424603174603175e-06,
|
| 1454 |
+
"loss": 0.2872,
|
| 1455 |
+
"step": 1860
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 16.65924276169265,
|
| 1459 |
+
"grad_norm": 11.702118873596191,
|
| 1460 |
+
"learning_rate": 9.176587301587301e-06,
|
| 1461 |
+
"loss": 0.2865,
|
| 1462 |
+
"step": 1870
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 16.748329621380847,
|
| 1466 |
+
"grad_norm": 9.740532875061035,
|
| 1467 |
+
"learning_rate": 8.92857142857143e-06,
|
| 1468 |
+
"loss": 0.2727,
|
| 1469 |
+
"step": 1880
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 16.83741648106904,
|
| 1473 |
+
"grad_norm": 12.387954711914062,
|
| 1474 |
+
"learning_rate": 8.680555555555556e-06,
|
| 1475 |
+
"loss": 0.3054,
|
| 1476 |
+
"step": 1890
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 16.92650334075724,
|
| 1480 |
+
"grad_norm": 17.06888198852539,
|
| 1481 |
+
"learning_rate": 8.432539682539684e-06,
|
| 1482 |
+
"loss": 0.2614,
|
| 1483 |
+
"step": 1900
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 16.997772828507795,
|
| 1487 |
+
"eval_accuracy": 0.8559799624295554,
|
| 1488 |
+
"eval_loss": 0.5104979872703552,
|
| 1489 |
+
"eval_runtime": 21.2594,
|
| 1490 |
+
"eval_samples_per_second": 75.12,
|
| 1491 |
+
"eval_steps_per_second": 2.352,
|
| 1492 |
+
"step": 1908
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 17.015590200445434,
|
| 1496 |
+
"grad_norm": 9.6270170211792,
|
| 1497 |
+
"learning_rate": 8.18452380952381e-06,
|
| 1498 |
+
"loss": 0.3009,
|
| 1499 |
+
"step": 1910
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 17.104677060133632,
|
| 1503 |
+
"grad_norm": 14.768975257873535,
|
| 1504 |
+
"learning_rate": 7.936507936507936e-06,
|
| 1505 |
+
"loss": 0.2854,
|
| 1506 |
+
"step": 1920
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 17.193763919821826,
|
| 1510 |
+
"grad_norm": 15.029583930969238,
|
| 1511 |
+
"learning_rate": 7.688492063492064e-06,
|
| 1512 |
+
"loss": 0.2663,
|
| 1513 |
+
"step": 1930
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 17.28285077951002,
|
| 1517 |
+
"grad_norm": 6.971687316894531,
|
| 1518 |
+
"learning_rate": 7.4404761904761905e-06,
|
| 1519 |
+
"loss": 0.2263,
|
| 1520 |
+
"step": 1940
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 17.37193763919822,
|
| 1524 |
+
"grad_norm": 12.997879981994629,
|
| 1525 |
+
"learning_rate": 7.192460317460318e-06,
|
| 1526 |
+
"loss": 0.2366,
|
| 1527 |
+
"step": 1950
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 17.461024498886413,
|
| 1531 |
+
"grad_norm": 20.721641540527344,
|
| 1532 |
+
"learning_rate": 6.944444444444445e-06,
|
| 1533 |
+
"loss": 0.2708,
|
| 1534 |
+
"step": 1960
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 17.55011135857461,
|
| 1538 |
+
"grad_norm": 8.18585205078125,
|
| 1539 |
+
"learning_rate": 6.696428571428572e-06,
|
| 1540 |
+
"loss": 0.2462,
|
| 1541 |
+
"step": 1970
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 17.639198218262806,
|
| 1545 |
+
"grad_norm": 9.622570991516113,
|
| 1546 |
+
"learning_rate": 6.448412698412699e-06,
|
| 1547 |
+
"loss": 0.282,
|
| 1548 |
+
"step": 1980
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 17.728285077951004,
|
| 1552 |
+
"grad_norm": 9.59653377532959,
|
| 1553 |
+
"learning_rate": 6.200396825396826e-06,
|
| 1554 |
+
"loss": 0.2607,
|
| 1555 |
+
"step": 1990
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 17.817371937639198,
|
| 1559 |
+
"grad_norm": 9.513813972473145,
|
| 1560 |
+
"learning_rate": 5.9523809523809525e-06,
|
| 1561 |
+
"loss": 0.2732,
|
| 1562 |
+
"step": 2000
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 17.906458797327396,
|
| 1566 |
+
"grad_norm": 11.091604232788086,
|
| 1567 |
+
"learning_rate": 5.70436507936508e-06,
|
| 1568 |
+
"loss": 0.2512,
|
| 1569 |
+
"step": 2010
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 17.99554565701559,
|
| 1573 |
+
"grad_norm": 11.867079734802246,
|
| 1574 |
+
"learning_rate": 5.456349206349207e-06,
|
| 1575 |
+
"loss": 0.2542,
|
| 1576 |
+
"step": 2020
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 17.99554565701559,
|
| 1580 |
+
"eval_accuracy": 0.8534752661239825,
|
| 1581 |
+
"eval_loss": 0.5331311821937561,
|
| 1582 |
+
"eval_runtime": 21.0758,
|
| 1583 |
+
"eval_samples_per_second": 75.774,
|
| 1584 |
+
"eval_steps_per_second": 2.372,
|
| 1585 |
+
"step": 2020
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 18.084632516703785,
|
| 1589 |
+
"grad_norm": 19.73405647277832,
|
| 1590 |
+
"learning_rate": 5.208333333333334e-06,
|
| 1591 |
+
"loss": 0.2383,
|
| 1592 |
+
"step": 2030
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 18.173719376391983,
|
| 1596 |
+
"grad_norm": 9.715604782104492,
|
| 1597 |
+
"learning_rate": 4.96031746031746e-06,
|
| 1598 |
+
"loss": 0.2284,
|
| 1599 |
+
"step": 2040
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 18.262806236080177,
|
| 1603 |
+
"grad_norm": 10.271077156066895,
|
| 1604 |
+
"learning_rate": 4.7123015873015875e-06,
|
| 1605 |
+
"loss": 0.2181,
|
| 1606 |
+
"step": 2050
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 18.351893095768375,
|
| 1610 |
+
"grad_norm": 14.901619911193848,
|
| 1611 |
+
"learning_rate": 4.464285714285715e-06,
|
| 1612 |
+
"loss": 0.2642,
|
| 1613 |
+
"step": 2060
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 18.44097995545657,
|
| 1617 |
+
"grad_norm": 8.47389030456543,
|
| 1618 |
+
"learning_rate": 4.216269841269842e-06,
|
| 1619 |
+
"loss": 0.2703,
|
| 1620 |
+
"step": 2070
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 18.530066815144767,
|
| 1624 |
+
"grad_norm": 18.4124755859375,
|
| 1625 |
+
"learning_rate": 3.968253968253968e-06,
|
| 1626 |
+
"loss": 0.2427,
|
| 1627 |
+
"step": 2080
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 18.619153674832962,
|
| 1631 |
+
"grad_norm": 6.289044380187988,
|
| 1632 |
+
"learning_rate": 3.7202380952380952e-06,
|
| 1633 |
+
"loss": 0.2784,
|
| 1634 |
+
"step": 2090
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 18.70824053452116,
|
| 1638 |
+
"grad_norm": 8.155685424804688,
|
| 1639 |
+
"learning_rate": 3.4722222222222224e-06,
|
| 1640 |
+
"loss": 0.2096,
|
| 1641 |
+
"step": 2100
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 18.797327394209354,
|
| 1645 |
+
"grad_norm": 10.127546310424805,
|
| 1646 |
+
"learning_rate": 3.2242063492063495e-06,
|
| 1647 |
+
"loss": 0.2247,
|
| 1648 |
+
"step": 2110
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 18.88641425389755,
|
| 1652 |
+
"grad_norm": 8.079633712768555,
|
| 1653 |
+
"learning_rate": 2.9761904761904763e-06,
|
| 1654 |
+
"loss": 0.2291,
|
| 1655 |
+
"step": 2120
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 18.975501113585747,
|
| 1659 |
+
"grad_norm": 11.977538108825684,
|
| 1660 |
+
"learning_rate": 2.7281746031746034e-06,
|
| 1661 |
+
"loss": 0.2436,
|
| 1662 |
+
"step": 2130
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 18.993318485523385,
|
| 1666 |
+
"eval_accuracy": 0.8609893550407013,
|
| 1667 |
+
"eval_loss": 0.5468289256095886,
|
| 1668 |
+
"eval_runtime": 21.1931,
|
| 1669 |
+
"eval_samples_per_second": 75.355,
|
| 1670 |
+
"eval_steps_per_second": 2.359,
|
| 1671 |
+
"step": 2132
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 19.06458797327394,
|
| 1675 |
+
"grad_norm": 11.12005615234375,
|
| 1676 |
+
"learning_rate": 2.48015873015873e-06,
|
| 1677 |
+
"loss": 0.2046,
|
| 1678 |
+
"step": 2140
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 19.15367483296214,
|
| 1682 |
+
"grad_norm": 8.387917518615723,
|
| 1683 |
+
"learning_rate": 2.2321428571428573e-06,
|
| 1684 |
+
"loss": 0.2178,
|
| 1685 |
+
"step": 2150
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 19.242761692650333,
|
| 1689 |
+
"grad_norm": 14.341036796569824,
|
| 1690 |
+
"learning_rate": 1.984126984126984e-06,
|
| 1691 |
+
"loss": 0.2194,
|
| 1692 |
+
"step": 2160
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 19.33184855233853,
|
| 1696 |
+
"grad_norm": 16.983957290649414,
|
| 1697 |
+
"learning_rate": 1.7361111111111112e-06,
|
| 1698 |
+
"loss": 0.2644,
|
| 1699 |
+
"step": 2170
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 19.420935412026726,
|
| 1703 |
+
"grad_norm": 6.247664928436279,
|
| 1704 |
+
"learning_rate": 1.4880952380952381e-06,
|
| 1705 |
+
"loss": 0.2198,
|
| 1706 |
+
"step": 2180
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 19.510022271714924,
|
| 1710 |
+
"grad_norm": 15.509531021118164,
|
| 1711 |
+
"learning_rate": 1.240079365079365e-06,
|
| 1712 |
+
"loss": 0.2524,
|
| 1713 |
+
"step": 2190
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 19.599109131403118,
|
| 1717 |
+
"grad_norm": 8.883134841918945,
|
| 1718 |
+
"learning_rate": 9.92063492063492e-07,
|
| 1719 |
+
"loss": 0.2226,
|
| 1720 |
+
"step": 2200
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 19.688195991091312,
|
| 1724 |
+
"grad_norm": 8.381498336791992,
|
| 1725 |
+
"learning_rate": 7.440476190476191e-07,
|
| 1726 |
+
"loss": 0.2067,
|
| 1727 |
+
"step": 2210
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 19.77728285077951,
|
| 1731 |
+
"grad_norm": 9.45773983001709,
|
| 1732 |
+
"learning_rate": 4.96031746031746e-07,
|
| 1733 |
+
"loss": 0.2116,
|
| 1734 |
+
"step": 2220
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 19.866369710467705,
|
| 1738 |
+
"grad_norm": 11.97305965423584,
|
| 1739 |
+
"learning_rate": 2.48015873015873e-07,
|
| 1740 |
+
"loss": 0.212,
|
| 1741 |
+
"step": 2230
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 19.955456570155903,
|
| 1745 |
+
"grad_norm": 7.919186592102051,
|
| 1746 |
+
"learning_rate": 0.0,
|
| 1747 |
+
"loss": 0.2275,
|
| 1748 |
+
"step": 2240
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 19.955456570155903,
|
| 1752 |
+
"eval_accuracy": 0.8703819661865999,
|
| 1753 |
+
"eval_loss": 0.5583016276359558,
|
| 1754 |
+
"eval_runtime": 21.2008,
|
| 1755 |
+
"eval_samples_per_second": 75.327,
|
| 1756 |
+
"eval_steps_per_second": 2.358,
|
| 1757 |
+
"step": 2240
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 19.955456570155903,
|
| 1761 |
+
"step": 2240,
|
| 1762 |
+
"total_flos": 2.928337686848471e+19,
|
| 1763 |
+
"train_loss": 0.5139281852969102,
|
| 1764 |
+
"train_runtime": 9596.9456,
|
| 1765 |
+
"train_samples_per_second": 29.937,
|
| 1766 |
+
"train_steps_per_second": 0.233
|
| 1767 |
+
}
|
| 1768 |
+
],
|
| 1769 |
+
"logging_steps": 10,
|
| 1770 |
+
"max_steps": 2240,
|
| 1771 |
+
"num_input_tokens_seen": 0,
|
| 1772 |
+
"num_train_epochs": 20,
|
| 1773 |
+
"save_steps": 500,
|
| 1774 |
+
"total_flos": 2.928337686848471e+19,
|
| 1775 |
+
"train_batch_size": 32,
|
| 1776 |
+
"trial_name": null,
|
| 1777 |
+
"trial_params": null
|
| 1778 |
+
}
|