Training in progress, step 1000
Browse files- all_results.json +13 -0
- eval_results.json +8 -0
- model.safetensors +1 -1
- runs/Apr15_20-26-16_bb4dc3bf47df/events.out.tfevents.1713213656.bb4dc3bf47df.3134.1 +3 -0
- runs/Apr15_20-45-42_bb4dc3bf47df/events.out.tfevents.1713213947.bb4dc3bf47df.3134.2 +3 -0
- runs/Apr15_21-08-26_bb4dc3bf47df/events.out.tfevents.1713215312.bb4dc3bf47df.3134.3 +3 -0
- runs/Apr15_21-08-26_bb4dc3bf47df/events.out.tfevents.1713215398.bb4dc3bf47df.3134.4 +3 -0
- runs/Apr15_21-11-15_bb4dc3bf47df/events.out.tfevents.1713215477.bb4dc3bf47df.15284.0 +3 -0
- train_results.json +8 -0
- trainer_state.json +2753 -0
- training_args.bin +1 -1
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_accuracy": 0.01789587852494577,
|
| 4 |
+
"eval_loss": 11.182476997375488,
|
| 5 |
+
"eval_runtime": 225.8898,
|
| 6 |
+
"eval_samples_per_second": 24.485,
|
| 7 |
+
"eval_steps_per_second": 8.163,
|
| 8 |
+
"total_flos": 9.121574551619174e+17,
|
| 9 |
+
"train_loss": 1.2556480791153037,
|
| 10 |
+
"train_runtime": 633.8251,
|
| 11 |
+
"train_samples_per_second": 18.444,
|
| 12 |
+
"train_steps_per_second": 6.148
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_accuracy": 0.01789587852494577,
|
| 4 |
+
"eval_loss": 11.182476997375488,
|
| 5 |
+
"eval_runtime": 225.8898,
|
| 6 |
+
"eval_samples_per_second": 24.485,
|
| 7 |
+
"eval_steps_per_second": 8.163
|
| 8 |
+
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 345602128
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:243e8e02e85b213cc91b4a9bff471b4d391253c5ef739ec0f847f2f01cbf9c5d
|
| 3 |
size 345602128
|
runs/Apr15_20-26-16_bb4dc3bf47df/events.out.tfevents.1713213656.bb4dc3bf47df.3134.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84e094e5ffe44a549e8ba11a8f5d74eb2996508ce0f5b51d4d9af61a51b285a0
|
| 3 |
+
size 411
|
runs/Apr15_20-45-42_bb4dc3bf47df/events.out.tfevents.1713213947.bb4dc3bf47df.3134.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cc26250face8bbd2e85548c8e4fbf1dadde4ea887542c473989e31f5d0d53f95
|
| 3 |
+
size 20289
|
runs/Apr15_21-08-26_bb4dc3bf47df/events.out.tfevents.1713215312.bb4dc3bf47df.3134.3
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e8e0cb0c1943d51bd27caaea2d96b028164c6516385bbe5bff1816ed911e26d5
|
| 3 |
+
size 7414
|
runs/Apr15_21-08-26_bb4dc3bf47df/events.out.tfevents.1713215398.bb4dc3bf47df.3134.4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ceaf5b5e5bf4d7853b7c7063f10d4cbf00af0f2043da2cfc8739e76e91993af
|
| 3 |
+
size 5551
|
runs/Apr15_21-11-15_bb4dc3bf47df/events.out.tfevents.1713215477.bb4dc3bf47df.15284.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7f1095ba3cfba6c5ad68bceda9a7983bc72ff0faf6385f9dbd61a77cbd40479b
|
| 3 |
+
size 25982
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 9.121574551619174e+17,
|
| 4 |
+
"train_loss": 1.2556480791153037,
|
| 5 |
+
"train_runtime": 633.8251,
|
| 6 |
+
"train_samples_per_second": 18.444,
|
| 7 |
+
"train_steps_per_second": 6.148
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2753 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 3897,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"grad_norm": 45.099327087402344,
|
| 14 |
+
"learning_rate": 0.00019948678470618425,
|
| 15 |
+
"loss": 2.0895,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01,
|
| 20 |
+
"grad_norm": 34.96247100830078,
|
| 21 |
+
"learning_rate": 0.0001989735694123685,
|
| 22 |
+
"loss": 2.3707,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.01,
|
| 27 |
+
"grad_norm": 16.490638732910156,
|
| 28 |
+
"learning_rate": 0.00019846035411855275,
|
| 29 |
+
"loss": 1.3701,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.01,
|
| 34 |
+
"grad_norm": 0.11159859597682953,
|
| 35 |
+
"learning_rate": 0.000197947138824737,
|
| 36 |
+
"loss": 1.0582,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.01,
|
| 41 |
+
"grad_norm": 0.12606161832809448,
|
| 42 |
+
"learning_rate": 0.00019743392353092123,
|
| 43 |
+
"loss": 1.1387,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.02,
|
| 48 |
+
"grad_norm": 4.384986400604248,
|
| 49 |
+
"learning_rate": 0.00019692070823710547,
|
| 50 |
+
"loss": 2.2113,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02,
|
| 55 |
+
"grad_norm": 0.7131040692329407,
|
| 56 |
+
"learning_rate": 0.00019640749294328974,
|
| 57 |
+
"loss": 0.6345,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.02,
|
| 62 |
+
"grad_norm": 5.308964252471924,
|
| 63 |
+
"learning_rate": 0.00019589427764947397,
|
| 64 |
+
"loss": 1.0728,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.02,
|
| 69 |
+
"grad_norm": 0.04876876622438431,
|
| 70 |
+
"learning_rate": 0.0001953810623556582,
|
| 71 |
+
"loss": 1.6609,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.03,
|
| 76 |
+
"grad_norm": 8.89411735534668,
|
| 77 |
+
"learning_rate": 0.00019486784706184245,
|
| 78 |
+
"loss": 1.8662,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.03,
|
| 83 |
+
"grad_norm": 0.05204994976520538,
|
| 84 |
+
"learning_rate": 0.00019435463176802672,
|
| 85 |
+
"loss": 0.5977,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.03,
|
| 90 |
+
"grad_norm": 34.89448165893555,
|
| 91 |
+
"learning_rate": 0.00019384141647421096,
|
| 92 |
+
"loss": 2.2815,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.03,
|
| 97 |
+
"grad_norm": 15.786956787109375,
|
| 98 |
+
"learning_rate": 0.0001933282011803952,
|
| 99 |
+
"loss": 1.6381,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.04,
|
| 104 |
+
"grad_norm": 29.863065719604492,
|
| 105 |
+
"learning_rate": 0.00019281498588657943,
|
| 106 |
+
"loss": 0.7649,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.04,
|
| 111 |
+
"grad_norm": 16.698570251464844,
|
| 112 |
+
"learning_rate": 0.00019230177059276367,
|
| 113 |
+
"loss": 1.1147,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.04,
|
| 118 |
+
"grad_norm": 34.869384765625,
|
| 119 |
+
"learning_rate": 0.00019178855529894794,
|
| 120 |
+
"loss": 1.4108,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.04,
|
| 125 |
+
"grad_norm": 0.3193562924861908,
|
| 126 |
+
"learning_rate": 0.00019127534000513218,
|
| 127 |
+
"loss": 1.0166,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.05,
|
| 132 |
+
"grad_norm": 23.633769989013672,
|
| 133 |
+
"learning_rate": 0.00019076212471131642,
|
| 134 |
+
"loss": 1.0537,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.05,
|
| 139 |
+
"grad_norm": 0.03772430121898651,
|
| 140 |
+
"learning_rate": 0.00019024890941750065,
|
| 141 |
+
"loss": 0.6254,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.05,
|
| 146 |
+
"grad_norm": 2.246229648590088,
|
| 147 |
+
"learning_rate": 0.0001897356941236849,
|
| 148 |
+
"loss": 2.0714,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.05,
|
| 153 |
+
"grad_norm": 2.4526054859161377,
|
| 154 |
+
"learning_rate": 0.00018922247882986913,
|
| 155 |
+
"loss": 1.0681,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.06,
|
| 160 |
+
"grad_norm": 1.3229526281356812,
|
| 161 |
+
"learning_rate": 0.00018870926353605337,
|
| 162 |
+
"loss": 0.6116,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.06,
|
| 167 |
+
"grad_norm": 0.042341507971286774,
|
| 168 |
+
"learning_rate": 0.0001881960482422376,
|
| 169 |
+
"loss": 1.4669,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.06,
|
| 174 |
+
"grad_norm": 35.0599479675293,
|
| 175 |
+
"learning_rate": 0.00018768283294842188,
|
| 176 |
+
"loss": 0.6897,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.06,
|
| 181 |
+
"grad_norm": 0.21465550363063812,
|
| 182 |
+
"learning_rate": 0.00018716961765460611,
|
| 183 |
+
"loss": 1.8696,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.07,
|
| 188 |
+
"grad_norm": 34.034175872802734,
|
| 189 |
+
"learning_rate": 0.00018665640236079035,
|
| 190 |
+
"loss": 1.5576,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.07,
|
| 195 |
+
"grad_norm": 34.895172119140625,
|
| 196 |
+
"learning_rate": 0.0001861431870669746,
|
| 197 |
+
"loss": 1.6542,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.07,
|
| 202 |
+
"grad_norm": 0.30317962169647217,
|
| 203 |
+
"learning_rate": 0.00018562997177315883,
|
| 204 |
+
"loss": 0.9893,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.07,
|
| 209 |
+
"grad_norm": 31.76571273803711,
|
| 210 |
+
"learning_rate": 0.0001851167564793431,
|
| 211 |
+
"loss": 2.43,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.08,
|
| 216 |
+
"grad_norm": 3.9265520572662354,
|
| 217 |
+
"learning_rate": 0.00018460354118552733,
|
| 218 |
+
"loss": 0.6246,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.08,
|
| 223 |
+
"grad_norm": 2.8710033893585205,
|
| 224 |
+
"learning_rate": 0.00018409032589171157,
|
| 225 |
+
"loss": 1.2509,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.08,
|
| 230 |
+
"grad_norm": 28.637104034423828,
|
| 231 |
+
"learning_rate": 0.0001835771105978958,
|
| 232 |
+
"loss": 1.4902,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.08,
|
| 237 |
+
"grad_norm": 4.821115970611572,
|
| 238 |
+
"learning_rate": 0.00018306389530408008,
|
| 239 |
+
"loss": 0.7693,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.09,
|
| 244 |
+
"grad_norm": 0.466738224029541,
|
| 245 |
+
"learning_rate": 0.00018255068001026432,
|
| 246 |
+
"loss": 1.9563,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.09,
|
| 251 |
+
"grad_norm": 34.78805160522461,
|
| 252 |
+
"learning_rate": 0.00018203746471644856,
|
| 253 |
+
"loss": 1.4862,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.09,
|
| 258 |
+
"grad_norm": 5.344199180603027,
|
| 259 |
+
"learning_rate": 0.0001815242494226328,
|
| 260 |
+
"loss": 1.1602,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.09,
|
| 265 |
+
"grad_norm": 0.4355538785457611,
|
| 266 |
+
"learning_rate": 0.00018101103412881703,
|
| 267 |
+
"loss": 1.2546,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.1,
|
| 272 |
+
"grad_norm": 0.7162923216819763,
|
| 273 |
+
"learning_rate": 0.0001804978188350013,
|
| 274 |
+
"loss": 3.0478,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.1,
|
| 279 |
+
"grad_norm": 0.09367480874061584,
|
| 280 |
+
"learning_rate": 0.00017998460354118554,
|
| 281 |
+
"loss": 1.0921,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.1,
|
| 286 |
+
"grad_norm": 0.20555052161216736,
|
| 287 |
+
"learning_rate": 0.00017947138824736978,
|
| 288 |
+
"loss": 0.0523,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.11,
|
| 293 |
+
"grad_norm": 0.08092090487480164,
|
| 294 |
+
"learning_rate": 0.00017895817295355402,
|
| 295 |
+
"loss": 1.5604,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.11,
|
| 300 |
+
"grad_norm": 0.32685595750808716,
|
| 301 |
+
"learning_rate": 0.00017844495765973828,
|
| 302 |
+
"loss": 0.5272,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.11,
|
| 307 |
+
"grad_norm": 0.5422254800796509,
|
| 308 |
+
"learning_rate": 0.00017793174236592252,
|
| 309 |
+
"loss": 3.4396,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.11,
|
| 314 |
+
"grad_norm": 35.279212951660156,
|
| 315 |
+
"learning_rate": 0.00017741852707210676,
|
| 316 |
+
"loss": 0.8554,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.12,
|
| 321 |
+
"grad_norm": 26.898597717285156,
|
| 322 |
+
"learning_rate": 0.000176905311778291,
|
| 323 |
+
"loss": 0.6776,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.12,
|
| 328 |
+
"grad_norm": 0.39926862716674805,
|
| 329 |
+
"learning_rate": 0.00017639209648447526,
|
| 330 |
+
"loss": 2.0551,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.12,
|
| 335 |
+
"grad_norm": 0.39992421865463257,
|
| 336 |
+
"learning_rate": 0.0001758788811906595,
|
| 337 |
+
"loss": 0.6316,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.12,
|
| 342 |
+
"grad_norm": 0.15786024928092957,
|
| 343 |
+
"learning_rate": 0.00017536566589684374,
|
| 344 |
+
"loss": 2.5885,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.13,
|
| 349 |
+
"grad_norm": 35.12531280517578,
|
| 350 |
+
"learning_rate": 0.00017485245060302798,
|
| 351 |
+
"loss": 1.5028,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.13,
|
| 356 |
+
"grad_norm": 1.712053656578064,
|
| 357 |
+
"learning_rate": 0.00017433923530921222,
|
| 358 |
+
"loss": 0.7324,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.13,
|
| 363 |
+
"grad_norm": 28.602834701538086,
|
| 364 |
+
"learning_rate": 0.00017382602001539648,
|
| 365 |
+
"loss": 2.7516,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.13,
|
| 370 |
+
"grad_norm": 8.03342342376709,
|
| 371 |
+
"learning_rate": 0.00017331280472158072,
|
| 372 |
+
"loss": 0.3827,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.14,
|
| 377 |
+
"grad_norm": 33.861934661865234,
|
| 378 |
+
"learning_rate": 0.00017279958942776496,
|
| 379 |
+
"loss": 1.1201,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.14,
|
| 384 |
+
"grad_norm": 0.0627857968211174,
|
| 385 |
+
"learning_rate": 0.0001722863741339492,
|
| 386 |
+
"loss": 0.6096,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.14,
|
| 391 |
+
"grad_norm": 34.3585319519043,
|
| 392 |
+
"learning_rate": 0.00017177315884013344,
|
| 393 |
+
"loss": 1.4903,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.14,
|
| 398 |
+
"grad_norm": 33.19193649291992,
|
| 399 |
+
"learning_rate": 0.00017125994354631768,
|
| 400 |
+
"loss": 1.4301,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.15,
|
| 405 |
+
"grad_norm": 34.12609100341797,
|
| 406 |
+
"learning_rate": 0.00017074672825250194,
|
| 407 |
+
"loss": 1.1228,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.15,
|
| 412 |
+
"grad_norm": 0.07128182798624039,
|
| 413 |
+
"learning_rate": 0.00017023351295868618,
|
| 414 |
+
"loss": 0.8195,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.15,
|
| 419 |
+
"grad_norm": 34.57550048828125,
|
| 420 |
+
"learning_rate": 0.00016972029766487042,
|
| 421 |
+
"loss": 1.9291,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.15,
|
| 426 |
+
"grad_norm": 60.265777587890625,
|
| 427 |
+
"learning_rate": 0.00016920708237105466,
|
| 428 |
+
"loss": 2.2235,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.16,
|
| 433 |
+
"grad_norm": 33.86371612548828,
|
| 434 |
+
"learning_rate": 0.0001686938670772389,
|
| 435 |
+
"loss": 1.101,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.16,
|
| 440 |
+
"grad_norm": 1.018540620803833,
|
| 441 |
+
"learning_rate": 0.00016818065178342314,
|
| 442 |
+
"loss": 1.0946,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.16,
|
| 447 |
+
"grad_norm": 0.05546221882104874,
|
| 448 |
+
"learning_rate": 0.00016766743648960738,
|
| 449 |
+
"loss": 1.3699,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.16,
|
| 454 |
+
"grad_norm": 0.4712983965873718,
|
| 455 |
+
"learning_rate": 0.00016715422119579164,
|
| 456 |
+
"loss": 2.366,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.17,
|
| 461 |
+
"grad_norm": 0.1462753266096115,
|
| 462 |
+
"learning_rate": 0.00016664100590197588,
|
| 463 |
+
"loss": 1.1087,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.17,
|
| 468 |
+
"grad_norm": 0.0955154225230217,
|
| 469 |
+
"learning_rate": 0.00016612779060816012,
|
| 470 |
+
"loss": 0.0171,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.17,
|
| 475 |
+
"grad_norm": 35.19109344482422,
|
| 476 |
+
"learning_rate": 0.00016561457531434436,
|
| 477 |
+
"loss": 1.6907,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.17,
|
| 482 |
+
"grad_norm": 0.16912657022476196,
|
| 483 |
+
"learning_rate": 0.00016510136002052862,
|
| 484 |
+
"loss": 0.0055,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.18,
|
| 489 |
+
"grad_norm": 34.529685974121094,
|
| 490 |
+
"learning_rate": 0.00016458814472671286,
|
| 491 |
+
"loss": 2.8261,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.18,
|
| 496 |
+
"grad_norm": 13.726387023925781,
|
| 497 |
+
"learning_rate": 0.0001640749294328971,
|
| 498 |
+
"loss": 1.7874,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.18,
|
| 503 |
+
"grad_norm": 29.94759750366211,
|
| 504 |
+
"learning_rate": 0.00016356171413908134,
|
| 505 |
+
"loss": 1.3416,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.18,
|
| 510 |
+
"grad_norm": 0.5144930481910706,
|
| 511 |
+
"learning_rate": 0.0001630484988452656,
|
| 512 |
+
"loss": 0.7381,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.19,
|
| 517 |
+
"grad_norm": 0.4536816477775574,
|
| 518 |
+
"learning_rate": 0.00016253528355144985,
|
| 519 |
+
"loss": 1.7744,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.19,
|
| 524 |
+
"grad_norm": 34.2490119934082,
|
| 525 |
+
"learning_rate": 0.00016202206825763408,
|
| 526 |
+
"loss": 1.6362,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.19,
|
| 531 |
+
"grad_norm": 0.7601956725120544,
|
| 532 |
+
"learning_rate": 0.00016150885296381832,
|
| 533 |
+
"loss": 1.0899,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.2,
|
| 538 |
+
"grad_norm": 0.16101330518722534,
|
| 539 |
+
"learning_rate": 0.00016099563767000256,
|
| 540 |
+
"loss": 1.4228,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.2,
|
| 545 |
+
"grad_norm": 2.7051339149475098,
|
| 546 |
+
"learning_rate": 0.00016048242237618683,
|
| 547 |
+
"loss": 1.2515,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.2,
|
| 552 |
+
"grad_norm": 34.69645690917969,
|
| 553 |
+
"learning_rate": 0.00015996920708237107,
|
| 554 |
+
"loss": 1.1568,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.2,
|
| 559 |
+
"grad_norm": 0.5005641579627991,
|
| 560 |
+
"learning_rate": 0.0001594559917885553,
|
| 561 |
+
"loss": 1.8697,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.21,
|
| 566 |
+
"grad_norm": 0.9808595776557922,
|
| 567 |
+
"learning_rate": 0.00015894277649473954,
|
| 568 |
+
"loss": 1.867,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.21,
|
| 573 |
+
"grad_norm": 34.12071228027344,
|
| 574 |
+
"learning_rate": 0.0001584295612009238,
|
| 575 |
+
"loss": 1.2028,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.21,
|
| 580 |
+
"grad_norm": 1.4215242862701416,
|
| 581 |
+
"learning_rate": 0.00015791634590710805,
|
| 582 |
+
"loss": 1.5028,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.21,
|
| 587 |
+
"grad_norm": 32.7039794921875,
|
| 588 |
+
"learning_rate": 0.0001574031306132923,
|
| 589 |
+
"loss": 1.851,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.22,
|
| 594 |
+
"grad_norm": 11.424134254455566,
|
| 595 |
+
"learning_rate": 0.00015688991531947653,
|
| 596 |
+
"loss": 1.1995,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.22,
|
| 601 |
+
"grad_norm": 18.550819396972656,
|
| 602 |
+
"learning_rate": 0.0001563767000256608,
|
| 603 |
+
"loss": 0.5795,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.22,
|
| 608 |
+
"grad_norm": 25.513269424438477,
|
| 609 |
+
"learning_rate": 0.00015586348473184503,
|
| 610 |
+
"loss": 0.6537,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.22,
|
| 615 |
+
"grad_norm": 0.043114595115184784,
|
| 616 |
+
"learning_rate": 0.00015535026943802927,
|
| 617 |
+
"loss": 1.4115,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.23,
|
| 622 |
+
"grad_norm": 0.06105157732963562,
|
| 623 |
+
"learning_rate": 0.0001548370541442135,
|
| 624 |
+
"loss": 0.5029,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.23,
|
| 629 |
+
"grad_norm": 6.027977466583252,
|
| 630 |
+
"learning_rate": 0.00015432383885039775,
|
| 631 |
+
"loss": 1.1377,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.23,
|
| 636 |
+
"grad_norm": 9.911833763122559,
|
| 637 |
+
"learning_rate": 0.000153810623556582,
|
| 638 |
+
"loss": 0.6785,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.23,
|
| 643 |
+
"grad_norm": 0.08647739142179489,
|
| 644 |
+
"learning_rate": 0.00015329740826276625,
|
| 645 |
+
"loss": 0.3428,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.24,
|
| 650 |
+
"grad_norm": 0.06115918606519699,
|
| 651 |
+
"learning_rate": 0.0001527841929689505,
|
| 652 |
+
"loss": 0.5422,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.24,
|
| 657 |
+
"grad_norm": 33.05713653564453,
|
| 658 |
+
"learning_rate": 0.00015227097767513473,
|
| 659 |
+
"loss": 0.9532,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.24,
|
| 664 |
+
"grad_norm": 2.755101442337036,
|
| 665 |
+
"learning_rate": 0.00015175776238131897,
|
| 666 |
+
"loss": 0.8933,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.24,
|
| 671 |
+
"grad_norm": 1.09266197681427,
|
| 672 |
+
"learning_rate": 0.0001512445470875032,
|
| 673 |
+
"loss": 1.3081,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.25,
|
| 678 |
+
"grad_norm": 1.8350268602371216,
|
| 679 |
+
"learning_rate": 0.00015073133179368745,
|
| 680 |
+
"loss": 1.6349,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.25,
|
| 685 |
+
"grad_norm": 28.465572357177734,
|
| 686 |
+
"learning_rate": 0.00015021811649987168,
|
| 687 |
+
"loss": 1.4432,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.25,
|
| 692 |
+
"grad_norm": 24.18708038330078,
|
| 693 |
+
"learning_rate": 0.00014970490120605595,
|
| 694 |
+
"loss": 0.8608,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.25,
|
| 699 |
+
"grad_norm": 0.06248585879802704,
|
| 700 |
+
"learning_rate": 0.0001491916859122402,
|
| 701 |
+
"loss": 0.7754,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.26,
|
| 706 |
+
"grad_norm": 0.02755439281463623,
|
| 707 |
+
"learning_rate": 0.00014867847061842443,
|
| 708 |
+
"loss": 0.3644,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.26,
|
| 713 |
+
"grad_norm": 0.03344785049557686,
|
| 714 |
+
"learning_rate": 0.00014816525532460867,
|
| 715 |
+
"loss": 2.7486,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.26,
|
| 720 |
+
"grad_norm": 4.339983940124512,
|
| 721 |
+
"learning_rate": 0.0001476520400307929,
|
| 722 |
+
"loss": 0.8171,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.26,
|
| 727 |
+
"grad_norm": 4.660573959350586,
|
| 728 |
+
"learning_rate": 0.00014713882473697717,
|
| 729 |
+
"loss": 2.1821,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.27,
|
| 734 |
+
"grad_norm": 21.810422897338867,
|
| 735 |
+
"learning_rate": 0.0001466256094431614,
|
| 736 |
+
"loss": 1.0611,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.27,
|
| 741 |
+
"grad_norm": 33.206607818603516,
|
| 742 |
+
"learning_rate": 0.00014611239414934565,
|
| 743 |
+
"loss": 1.5671,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.27,
|
| 748 |
+
"grad_norm": 0.678447425365448,
|
| 749 |
+
"learning_rate": 0.0001455991788555299,
|
| 750 |
+
"loss": 1.0065,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.27,
|
| 755 |
+
"grad_norm": 32.59206771850586,
|
| 756 |
+
"learning_rate": 0.00014508596356171415,
|
| 757 |
+
"loss": 1.8422,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.28,
|
| 762 |
+
"grad_norm": 28.002819061279297,
|
| 763 |
+
"learning_rate": 0.0001445727482678984,
|
| 764 |
+
"loss": 1.1739,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.28,
|
| 769 |
+
"grad_norm": 4.193221092224121,
|
| 770 |
+
"learning_rate": 0.00014405953297408263,
|
| 771 |
+
"loss": 0.431,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.28,
|
| 776 |
+
"grad_norm": 34.987388610839844,
|
| 777 |
+
"learning_rate": 0.00014354631768026687,
|
| 778 |
+
"loss": 1.1974,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.28,
|
| 783 |
+
"grad_norm": 0.1439794898033142,
|
| 784 |
+
"learning_rate": 0.0001430331023864511,
|
| 785 |
+
"loss": 2.1824,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.29,
|
| 790 |
+
"grad_norm": 0.07825344055891037,
|
| 791 |
+
"learning_rate": 0.00014251988709263537,
|
| 792 |
+
"loss": 1.103,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.29,
|
| 797 |
+
"grad_norm": 34.14964294433594,
|
| 798 |
+
"learning_rate": 0.0001420066717988196,
|
| 799 |
+
"loss": 2.1521,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.29,
|
| 804 |
+
"grad_norm": 1.839199185371399,
|
| 805 |
+
"learning_rate": 0.00014149345650500385,
|
| 806 |
+
"loss": 0.0536,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.3,
|
| 811 |
+
"grad_norm": 34.573585510253906,
|
| 812 |
+
"learning_rate": 0.0001409802412111881,
|
| 813 |
+
"loss": 1.5661,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.3,
|
| 818 |
+
"grad_norm": 0.41036948561668396,
|
| 819 |
+
"learning_rate": 0.00014046702591737236,
|
| 820 |
+
"loss": 1.1817,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.3,
|
| 825 |
+
"grad_norm": 0.08495642244815826,
|
| 826 |
+
"learning_rate": 0.0001399538106235566,
|
| 827 |
+
"loss": 1.5704,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.3,
|
| 832 |
+
"grad_norm": 3.0547990798950195,
|
| 833 |
+
"learning_rate": 0.00013944059532974083,
|
| 834 |
+
"loss": 1.5402,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.31,
|
| 839 |
+
"grad_norm": 7.8949713706970215,
|
| 840 |
+
"learning_rate": 0.00013892738003592507,
|
| 841 |
+
"loss": 0.4595,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.31,
|
| 846 |
+
"grad_norm": 24.375001907348633,
|
| 847 |
+
"learning_rate": 0.00013841416474210934,
|
| 848 |
+
"loss": 0.7511,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.31,
|
| 853 |
+
"grad_norm": 12.163422584533691,
|
| 854 |
+
"learning_rate": 0.00013790094944829358,
|
| 855 |
+
"loss": 0.5726,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.31,
|
| 860 |
+
"grad_norm": 34.362632751464844,
|
| 861 |
+
"learning_rate": 0.00013738773415447782,
|
| 862 |
+
"loss": 1.3018,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.32,
|
| 867 |
+
"grad_norm": 21.605195999145508,
|
| 868 |
+
"learning_rate": 0.00013687451886066205,
|
| 869 |
+
"loss": 1.2047,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.32,
|
| 874 |
+
"grad_norm": 4.356873035430908,
|
| 875 |
+
"learning_rate": 0.0001363613035668463,
|
| 876 |
+
"loss": 1.3787,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.32,
|
| 881 |
+
"grad_norm": 12.701129913330078,
|
| 882 |
+
"learning_rate": 0.00013584808827303056,
|
| 883 |
+
"loss": 1.0334,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.32,
|
| 888 |
+
"grad_norm": 7.2957892417907715,
|
| 889 |
+
"learning_rate": 0.0001353348729792148,
|
| 890 |
+
"loss": 0.2939,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.33,
|
| 895 |
+
"grad_norm": 0.25156155228614807,
|
| 896 |
+
"learning_rate": 0.00013482165768539904,
|
| 897 |
+
"loss": 0.7811,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.33,
|
| 902 |
+
"grad_norm": 0.8677279353141785,
|
| 903 |
+
"learning_rate": 0.00013430844239158328,
|
| 904 |
+
"loss": 1.5475,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.33,
|
| 909 |
+
"grad_norm": 0.13093720376491547,
|
| 910 |
+
"learning_rate": 0.00013379522709776754,
|
| 911 |
+
"loss": 0.7902,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.33,
|
| 916 |
+
"grad_norm": 0.2733534276485443,
|
| 917 |
+
"learning_rate": 0.00013328201180395178,
|
| 918 |
+
"loss": 0.9059,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.34,
|
| 923 |
+
"grad_norm": 34.48268508911133,
|
| 924 |
+
"learning_rate": 0.00013276879651013602,
|
| 925 |
+
"loss": 1.5932,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.34,
|
| 930 |
+
"grad_norm": 33.39665603637695,
|
| 931 |
+
"learning_rate": 0.00013225558121632026,
|
| 932 |
+
"loss": 2.4436,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.34,
|
| 937 |
+
"grad_norm": 0.6858634948730469,
|
| 938 |
+
"learning_rate": 0.0001317423659225045,
|
| 939 |
+
"loss": 0.8113,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.34,
|
| 944 |
+
"grad_norm": 32.40642166137695,
|
| 945 |
+
"learning_rate": 0.00013122915062868873,
|
| 946 |
+
"loss": 2.1392,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.35,
|
| 951 |
+
"grad_norm": 0.05664900690317154,
|
| 952 |
+
"learning_rate": 0.00013071593533487297,
|
| 953 |
+
"loss": 0.8724,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.35,
|
| 958 |
+
"grad_norm": 30.9477481842041,
|
| 959 |
+
"learning_rate": 0.0001302027200410572,
|
| 960 |
+
"loss": 1.2746,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.35,
|
| 965 |
+
"grad_norm": 2.9625468254089355,
|
| 966 |
+
"learning_rate": 0.00012968950474724145,
|
| 967 |
+
"loss": 2.0471,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.35,
|
| 972 |
+
"grad_norm": 6.3701629638671875,
|
| 973 |
+
"learning_rate": 0.00012917628945342572,
|
| 974 |
+
"loss": 0.9569,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.36,
|
| 979 |
+
"grad_norm": 5.0901007652282715,
|
| 980 |
+
"learning_rate": 0.00012866307415960996,
|
| 981 |
+
"loss": 0.8391,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.36,
|
| 986 |
+
"grad_norm": 2.827472448348999,
|
| 987 |
+
"learning_rate": 0.0001281498588657942,
|
| 988 |
+
"loss": 0.5071,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.36,
|
| 993 |
+
"grad_norm": 0.24257418513298035,
|
| 994 |
+
"learning_rate": 0.00012763664357197843,
|
| 995 |
+
"loss": 0.0258,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.36,
|
| 1000 |
+
"grad_norm": 34.39240264892578,
|
| 1001 |
+
"learning_rate": 0.0001271234282781627,
|
| 1002 |
+
"loss": 1.6826,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.37,
|
| 1007 |
+
"grad_norm": 0.06999631226062775,
|
| 1008 |
+
"learning_rate": 0.00012661021298434694,
|
| 1009 |
+
"loss": 2.854,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.37,
|
| 1014 |
+
"grad_norm": 0.7731397151947021,
|
| 1015 |
+
"learning_rate": 0.00012609699769053118,
|
| 1016 |
+
"loss": 1.5991,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.37,
|
| 1021 |
+
"grad_norm": 0.7090578675270081,
|
| 1022 |
+
"learning_rate": 0.00012558378239671542,
|
| 1023 |
+
"loss": 1.1829,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.37,
|
| 1028 |
+
"grad_norm": 33.13046646118164,
|
| 1029 |
+
"learning_rate": 0.00012507056710289968,
|
| 1030 |
+
"loss": 1.4736,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.38,
|
| 1035 |
+
"grad_norm": 0.6540558338165283,
|
| 1036 |
+
"learning_rate": 0.00012455735180908392,
|
| 1037 |
+
"loss": 1.6983,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.38,
|
| 1042 |
+
"grad_norm": 1.6605318784713745,
|
| 1043 |
+
"learning_rate": 0.00012404413651526816,
|
| 1044 |
+
"loss": 1.5005,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.38,
|
| 1049 |
+
"grad_norm": 31.07722282409668,
|
| 1050 |
+
"learning_rate": 0.0001235309212214524,
|
| 1051 |
+
"loss": 1.4853,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.38,
|
| 1056 |
+
"grad_norm": 1.135640025138855,
|
| 1057 |
+
"learning_rate": 0.00012301770592763664,
|
| 1058 |
+
"loss": 0.9949,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.39,
|
| 1063 |
+
"grad_norm": 0.14291101694107056,
|
| 1064 |
+
"learning_rate": 0.0001225044906338209,
|
| 1065 |
+
"loss": 1.7112,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.39,
|
| 1070 |
+
"grad_norm": 0.07842687517404556,
|
| 1071 |
+
"learning_rate": 0.00012199127534000514,
|
| 1072 |
+
"loss": 0.677,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.39,
|
| 1077 |
+
"grad_norm": 0.6385894417762756,
|
| 1078 |
+
"learning_rate": 0.00012147806004618938,
|
| 1079 |
+
"loss": 1.0463,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.4,
|
| 1084 |
+
"grad_norm": 34.2146110534668,
|
| 1085 |
+
"learning_rate": 0.00012096484475237362,
|
| 1086 |
+
"loss": 2.0908,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.4,
|
| 1091 |
+
"grad_norm": 0.15750457346439362,
|
| 1092 |
+
"learning_rate": 0.00012045162945855788,
|
| 1093 |
+
"loss": 1.6537,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.4,
|
| 1098 |
+
"grad_norm": 0.15273036062717438,
|
| 1099 |
+
"learning_rate": 0.00011993841416474212,
|
| 1100 |
+
"loss": 1.8717,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.4,
|
| 1105 |
+
"grad_norm": 4.405086517333984,
|
| 1106 |
+
"learning_rate": 0.00011942519887092636,
|
| 1107 |
+
"loss": 0.3942,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.41,
|
| 1112 |
+
"grad_norm": 2.592953681945801,
|
| 1113 |
+
"learning_rate": 0.0001189119835771106,
|
| 1114 |
+
"loss": 0.3217,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.41,
|
| 1119 |
+
"grad_norm": 0.5546024441719055,
|
| 1120 |
+
"learning_rate": 0.00011839876828329485,
|
| 1121 |
+
"loss": 1.0285,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.41,
|
| 1126 |
+
"grad_norm": 0.09479659050703049,
|
| 1127 |
+
"learning_rate": 0.00011788555298947909,
|
| 1128 |
+
"loss": 1.7519,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.41,
|
| 1133 |
+
"grad_norm": 0.09771596640348434,
|
| 1134 |
+
"learning_rate": 0.00011737233769566333,
|
| 1135 |
+
"loss": 1.2341,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.42,
|
| 1140 |
+
"grad_norm": 33.37510681152344,
|
| 1141 |
+
"learning_rate": 0.00011685912240184758,
|
| 1142 |
+
"loss": 3.3626,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.42,
|
| 1147 |
+
"grad_norm": 32.650489807128906,
|
| 1148 |
+
"learning_rate": 0.00011634590710803182,
|
| 1149 |
+
"loss": 0.9869,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.42,
|
| 1154 |
+
"grad_norm": 32.52039337158203,
|
| 1155 |
+
"learning_rate": 0.00011583269181421607,
|
| 1156 |
+
"loss": 1.8117,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.42,
|
| 1161 |
+
"grad_norm": 15.11254596710205,
|
| 1162 |
+
"learning_rate": 0.00011531947652040031,
|
| 1163 |
+
"loss": 1.288,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.43,
|
| 1168 |
+
"grad_norm": 3.6993765830993652,
|
| 1169 |
+
"learning_rate": 0.00011480626122658455,
|
| 1170 |
+
"loss": 0.863,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.43,
|
| 1175 |
+
"grad_norm": 34.809791564941406,
|
| 1176 |
+
"learning_rate": 0.00011429304593276879,
|
| 1177 |
+
"loss": 0.6192,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.43,
|
| 1182 |
+
"grad_norm": 0.10077141970396042,
|
| 1183 |
+
"learning_rate": 0.00011377983063895306,
|
| 1184 |
+
"loss": 0.571,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.43,
|
| 1189 |
+
"grad_norm": 0.07107020914554596,
|
| 1190 |
+
"learning_rate": 0.0001132666153451373,
|
| 1191 |
+
"loss": 0.996,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.44,
|
| 1196 |
+
"grad_norm": 33.04214096069336,
|
| 1197 |
+
"learning_rate": 0.00011275340005132153,
|
| 1198 |
+
"loss": 2.6481,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.44,
|
| 1203 |
+
"grad_norm": 32.640541076660156,
|
| 1204 |
+
"learning_rate": 0.00011224018475750577,
|
| 1205 |
+
"loss": 2.1778,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.44,
|
| 1210 |
+
"grad_norm": 1.7503925561904907,
|
| 1211 |
+
"learning_rate": 0.00011172696946369004,
|
| 1212 |
+
"loss": 1.5489,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.44,
|
| 1217 |
+
"grad_norm": 0.348849892616272,
|
| 1218 |
+
"learning_rate": 0.00011121375416987428,
|
| 1219 |
+
"loss": 1.1267,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.45,
|
| 1224 |
+
"grad_norm": 0.8353062272071838,
|
| 1225 |
+
"learning_rate": 0.00011070053887605852,
|
| 1226 |
+
"loss": 0.0233,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.45,
|
| 1231 |
+
"grad_norm": 0.4115784466266632,
|
| 1232 |
+
"learning_rate": 0.00011018732358224275,
|
| 1233 |
+
"loss": 1.3895,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.45,
|
| 1238 |
+
"grad_norm": 0.09716140478849411,
|
| 1239 |
+
"learning_rate": 0.00010967410828842699,
|
| 1240 |
+
"loss": 0.9971,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.45,
|
| 1245 |
+
"grad_norm": 0.07956521958112717,
|
| 1246 |
+
"learning_rate": 0.00010916089299461126,
|
| 1247 |
+
"loss": 0.4633,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.46,
|
| 1252 |
+
"grad_norm": 33.34837341308594,
|
| 1253 |
+
"learning_rate": 0.0001086476777007955,
|
| 1254 |
+
"loss": 0.5945,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.46,
|
| 1259 |
+
"grad_norm": 33.278682708740234,
|
| 1260 |
+
"learning_rate": 0.00010813446240697974,
|
| 1261 |
+
"loss": 1.0137,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.46,
|
| 1266 |
+
"grad_norm": 32.34352111816406,
|
| 1267 |
+
"learning_rate": 0.00010762124711316398,
|
| 1268 |
+
"loss": 2.238,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.46,
|
| 1273 |
+
"grad_norm": 0.8423387408256531,
|
| 1274 |
+
"learning_rate": 0.00010710803181934823,
|
| 1275 |
+
"loss": 0.0119,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.47,
|
| 1280 |
+
"grad_norm": 0.43691009283065796,
|
| 1281 |
+
"learning_rate": 0.00010659481652553247,
|
| 1282 |
+
"loss": 0.7785,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.47,
|
| 1287 |
+
"grad_norm": 0.27386584877967834,
|
| 1288 |
+
"learning_rate": 0.0001060816012317167,
|
| 1289 |
+
"loss": 1.4121,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.47,
|
| 1294 |
+
"grad_norm": 0.035622984170913696,
|
| 1295 |
+
"learning_rate": 0.00010556838593790094,
|
| 1296 |
+
"loss": 1.8601,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.47,
|
| 1301 |
+
"grad_norm": 0.6981528401374817,
|
| 1302 |
+
"learning_rate": 0.00010505517064408518,
|
| 1303 |
+
"loss": 1.2077,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.48,
|
| 1308 |
+
"grad_norm": 32.1064567565918,
|
| 1309 |
+
"learning_rate": 0.00010454195535026945,
|
| 1310 |
+
"loss": 1.5198,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.48,
|
| 1315 |
+
"grad_norm": 30.88555908203125,
|
| 1316 |
+
"learning_rate": 0.00010402874005645369,
|
| 1317 |
+
"loss": 2.0069,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.48,
|
| 1322 |
+
"grad_norm": 0.03494291752576828,
|
| 1323 |
+
"learning_rate": 0.00010351552476263793,
|
| 1324 |
+
"loss": 0.3124,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.48,
|
| 1329 |
+
"grad_norm": 1.9848475456237793,
|
| 1330 |
+
"learning_rate": 0.00010300230946882216,
|
| 1331 |
+
"loss": 0.8303,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.49,
|
| 1336 |
+
"grad_norm": 31.325864791870117,
|
| 1337 |
+
"learning_rate": 0.00010248909417500643,
|
| 1338 |
+
"loss": 0.9595,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.49,
|
| 1343 |
+
"grad_norm": 0.025403378531336784,
|
| 1344 |
+
"learning_rate": 0.00010197587888119067,
|
| 1345 |
+
"loss": 0.362,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.49,
|
| 1350 |
+
"grad_norm": 1.1182492971420288,
|
| 1351 |
+
"learning_rate": 0.00010146266358737491,
|
| 1352 |
+
"loss": 1.4183,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.5,
|
| 1357 |
+
"grad_norm": 0.0197270717471838,
|
| 1358 |
+
"learning_rate": 0.00010094944829355915,
|
| 1359 |
+
"loss": 0.9907,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.5,
|
| 1364 |
+
"grad_norm": 1.5864602327346802,
|
| 1365 |
+
"learning_rate": 0.00010043623299974341,
|
| 1366 |
+
"loss": 0.9328,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.5,
|
| 1371 |
+
"grad_norm": 0.019860761240124702,
|
| 1372 |
+
"learning_rate": 9.992301770592765e-05,
|
| 1373 |
+
"loss": 1.2027,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.5,
|
| 1378 |
+
"grad_norm": 0.27393853664398193,
|
| 1379 |
+
"learning_rate": 9.940980241211189e-05,
|
| 1380 |
+
"loss": 1.2231,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.51,
|
| 1385 |
+
"grad_norm": 0.06520062685012817,
|
| 1386 |
+
"learning_rate": 9.889658711829613e-05,
|
| 1387 |
+
"loss": 1.1252,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.51,
|
| 1392 |
+
"grad_norm": 0.10603094100952148,
|
| 1393 |
+
"learning_rate": 9.838337182448038e-05,
|
| 1394 |
+
"loss": 0.5362,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.51,
|
| 1399 |
+
"grad_norm": 0.21443594992160797,
|
| 1400 |
+
"learning_rate": 9.787015653066462e-05,
|
| 1401 |
+
"loss": 1.6563,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.51,
|
| 1406 |
+
"grad_norm": 0.15189635753631592,
|
| 1407 |
+
"learning_rate": 9.735694123684886e-05,
|
| 1408 |
+
"loss": 1.5657,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.52,
|
| 1413 |
+
"grad_norm": 32.36518478393555,
|
| 1414 |
+
"learning_rate": 9.68437259430331e-05,
|
| 1415 |
+
"loss": 2.1702,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.52,
|
| 1420 |
+
"grad_norm": 0.2264794558286667,
|
| 1421 |
+
"learning_rate": 9.633051064921735e-05,
|
| 1422 |
+
"loss": 1.2663,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.52,
|
| 1427 |
+
"grad_norm": 30.200082778930664,
|
| 1428 |
+
"learning_rate": 9.581729535540159e-05,
|
| 1429 |
+
"loss": 2.2151,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.52,
|
| 1434 |
+
"grad_norm": 3.5600388050079346,
|
| 1435 |
+
"learning_rate": 9.530408006158584e-05,
|
| 1436 |
+
"loss": 0.7643,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.53,
|
| 1441 |
+
"grad_norm": 0.1124127209186554,
|
| 1442 |
+
"learning_rate": 9.479086476777008e-05,
|
| 1443 |
+
"loss": 0.679,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.53,
|
| 1448 |
+
"grad_norm": 28.294071197509766,
|
| 1449 |
+
"learning_rate": 9.427764947395433e-05,
|
| 1450 |
+
"loss": 1.5172,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.53,
|
| 1455 |
+
"grad_norm": 26.79751205444336,
|
| 1456 |
+
"learning_rate": 9.376443418013857e-05,
|
| 1457 |
+
"loss": 0.4192,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.53,
|
| 1462 |
+
"grad_norm": 0.1510336697101593,
|
| 1463 |
+
"learning_rate": 9.325121888632282e-05,
|
| 1464 |
+
"loss": 2.1576,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.54,
|
| 1469 |
+
"grad_norm": 0.32501623034477234,
|
| 1470 |
+
"learning_rate": 9.273800359250706e-05,
|
| 1471 |
+
"loss": 1.1192,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.54,
|
| 1476 |
+
"grad_norm": 26.4614315032959,
|
| 1477 |
+
"learning_rate": 9.22247882986913e-05,
|
| 1478 |
+
"loss": 0.7766,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.54,
|
| 1483 |
+
"grad_norm": 0.20047006011009216,
|
| 1484 |
+
"learning_rate": 9.171157300487555e-05,
|
| 1485 |
+
"loss": 0.6836,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.54,
|
| 1490 |
+
"grad_norm": 10.277036666870117,
|
| 1491 |
+
"learning_rate": 9.119835771105979e-05,
|
| 1492 |
+
"loss": 0.4951,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.55,
|
| 1497 |
+
"grad_norm": 9.455358505249023,
|
| 1498 |
+
"learning_rate": 9.068514241724404e-05,
|
| 1499 |
+
"loss": 0.2494,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.55,
|
| 1504 |
+
"grad_norm": 5.371951103210449,
|
| 1505 |
+
"learning_rate": 9.017192712342828e-05,
|
| 1506 |
+
"loss": 0.7672,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.55,
|
| 1511 |
+
"grad_norm": 4.935911655426025,
|
| 1512 |
+
"learning_rate": 8.965871182961254e-05,
|
| 1513 |
+
"loss": 0.7815,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.55,
|
| 1518 |
+
"grad_norm": 0.05824290215969086,
|
| 1519 |
+
"learning_rate": 8.914549653579677e-05,
|
| 1520 |
+
"loss": 0.4563,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.56,
|
| 1525 |
+
"grad_norm": 29.326370239257812,
|
| 1526 |
+
"learning_rate": 8.863228124198101e-05,
|
| 1527 |
+
"loss": 0.9615,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.56,
|
| 1532 |
+
"grad_norm": 30.599393844604492,
|
| 1533 |
+
"learning_rate": 8.811906594816526e-05,
|
| 1534 |
+
"loss": 0.617,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.56,
|
| 1539 |
+
"grad_norm": 0.6331340074539185,
|
| 1540 |
+
"learning_rate": 8.76058506543495e-05,
|
| 1541 |
+
"loss": 1.0741,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.56,
|
| 1546 |
+
"grad_norm": 34.11085891723633,
|
| 1547 |
+
"learning_rate": 8.709263536053374e-05,
|
| 1548 |
+
"loss": 1.9807,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.57,
|
| 1553 |
+
"grad_norm": 0.4941723942756653,
|
| 1554 |
+
"learning_rate": 8.6579420066718e-05,
|
| 1555 |
+
"loss": 1.6712,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.57,
|
| 1560 |
+
"grad_norm": 0.8858903646469116,
|
| 1561 |
+
"learning_rate": 8.606620477290223e-05,
|
| 1562 |
+
"loss": 1.1382,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.57,
|
| 1567 |
+
"grad_norm": 0.4389359951019287,
|
| 1568 |
+
"learning_rate": 8.555298947908647e-05,
|
| 1569 |
+
"loss": 0.0166,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.57,
|
| 1574 |
+
"grad_norm": 0.27877870202064514,
|
| 1575 |
+
"learning_rate": 8.503977418527072e-05,
|
| 1576 |
+
"loss": 1.8697,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.58,
|
| 1581 |
+
"grad_norm": 0.0968392938375473,
|
| 1582 |
+
"learning_rate": 8.452655889145496e-05,
|
| 1583 |
+
"loss": 0.9307,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.58,
|
| 1588 |
+
"grad_norm": 0.07415905594825745,
|
| 1589 |
+
"learning_rate": 8.401334359763922e-05,
|
| 1590 |
+
"loss": 1.3474,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.58,
|
| 1595 |
+
"grad_norm": 32.0526008605957,
|
| 1596 |
+
"learning_rate": 8.350012830382345e-05,
|
| 1597 |
+
"loss": 1.9036,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.59,
|
| 1602 |
+
"grad_norm": 31.857463836669922,
|
| 1603 |
+
"learning_rate": 8.29869130100077e-05,
|
| 1604 |
+
"loss": 1.7843,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.59,
|
| 1609 |
+
"grad_norm": 0.16152064502239227,
|
| 1610 |
+
"learning_rate": 8.247369771619195e-05,
|
| 1611 |
+
"loss": 1.138,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.59,
|
| 1616 |
+
"grad_norm": 31.4908504486084,
|
| 1617 |
+
"learning_rate": 8.19604824223762e-05,
|
| 1618 |
+
"loss": 1.348,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.59,
|
| 1623 |
+
"grad_norm": 0.8940775990486145,
|
| 1624 |
+
"learning_rate": 8.144726712856044e-05,
|
| 1625 |
+
"loss": 1.4317,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.6,
|
| 1630 |
+
"grad_norm": 1.3364328145980835,
|
| 1631 |
+
"learning_rate": 8.093405183474469e-05,
|
| 1632 |
+
"loss": 1.3774,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.6,
|
| 1637 |
+
"grad_norm": 32.3911018371582,
|
| 1638 |
+
"learning_rate": 8.042083654092893e-05,
|
| 1639 |
+
"loss": 1.8806,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.6,
|
| 1644 |
+
"grad_norm": 0.0908791646361351,
|
| 1645 |
+
"learning_rate": 7.990762124711317e-05,
|
| 1646 |
+
"loss": 0.3828,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.6,
|
| 1651 |
+
"grad_norm": 0.7057339549064636,
|
| 1652 |
+
"learning_rate": 7.939440595329742e-05,
|
| 1653 |
+
"loss": 1.8282,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.61,
|
| 1658 |
+
"grad_norm": 0.9124455451965332,
|
| 1659 |
+
"learning_rate": 7.888119065948166e-05,
|
| 1660 |
+
"loss": 0.8194,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.61,
|
| 1665 |
+
"grad_norm": 0.6995695233345032,
|
| 1666 |
+
"learning_rate": 7.83679753656659e-05,
|
| 1667 |
+
"loss": 2.4169,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.61,
|
| 1672 |
+
"grad_norm": 30.961750030517578,
|
| 1673 |
+
"learning_rate": 7.785476007185013e-05,
|
| 1674 |
+
"loss": 1.0782,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.61,
|
| 1679 |
+
"grad_norm": 30.930931091308594,
|
| 1680 |
+
"learning_rate": 7.734154477803439e-05,
|
| 1681 |
+
"loss": 1.773,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.62,
|
| 1686 |
+
"grad_norm": 2.427703857421875,
|
| 1687 |
+
"learning_rate": 7.682832948421863e-05,
|
| 1688 |
+
"loss": 1.1838,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.62,
|
| 1693 |
+
"grad_norm": 28.99689292907715,
|
| 1694 |
+
"learning_rate": 7.631511419040288e-05,
|
| 1695 |
+
"loss": 0.9657,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.62,
|
| 1700 |
+
"grad_norm": 0.0632254034280777,
|
| 1701 |
+
"learning_rate": 7.580189889658712e-05,
|
| 1702 |
+
"loss": 0.9273,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.62,
|
| 1707 |
+
"grad_norm": 0.06876543164253235,
|
| 1708 |
+
"learning_rate": 7.528868360277137e-05,
|
| 1709 |
+
"loss": 0.6686,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.63,
|
| 1714 |
+
"grad_norm": 0.059902504086494446,
|
| 1715 |
+
"learning_rate": 7.477546830895561e-05,
|
| 1716 |
+
"loss": 0.6242,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.63,
|
| 1721 |
+
"grad_norm": 30.04041290283203,
|
| 1722 |
+
"learning_rate": 7.426225301513986e-05,
|
| 1723 |
+
"loss": 2.3915,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.63,
|
| 1728 |
+
"grad_norm": 2.442535400390625,
|
| 1729 |
+
"learning_rate": 7.37490377213241e-05,
|
| 1730 |
+
"loss": 0.5622,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.63,
|
| 1735 |
+
"grad_norm": 0.0483098030090332,
|
| 1736 |
+
"learning_rate": 7.323582242750834e-05,
|
| 1737 |
+
"loss": 1.5706,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.64,
|
| 1742 |
+
"grad_norm": 0.048585113137960434,
|
| 1743 |
+
"learning_rate": 7.272260713369259e-05,
|
| 1744 |
+
"loss": 0.8918,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.64,
|
| 1749 |
+
"grad_norm": 29.45044708251953,
|
| 1750 |
+
"learning_rate": 7.220939183987683e-05,
|
| 1751 |
+
"loss": 1.4446,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.64,
|
| 1756 |
+
"grad_norm": 0.03927851840853691,
|
| 1757 |
+
"learning_rate": 7.169617654606108e-05,
|
| 1758 |
+
"loss": 0.6336,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.64,
|
| 1763 |
+
"grad_norm": 5.8573174476623535,
|
| 1764 |
+
"learning_rate": 7.118296125224532e-05,
|
| 1765 |
+
"loss": 0.5349,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.65,
|
| 1770 |
+
"grad_norm": 0.04964911565184593,
|
| 1771 |
+
"learning_rate": 7.066974595842957e-05,
|
| 1772 |
+
"loss": 1.471,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.65,
|
| 1777 |
+
"grad_norm": 1.5833439826965332,
|
| 1778 |
+
"learning_rate": 7.015653066461381e-05,
|
| 1779 |
+
"loss": 0.3127,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.65,
|
| 1784 |
+
"grad_norm": 0.6209262609481812,
|
| 1785 |
+
"learning_rate": 6.964331537079806e-05,
|
| 1786 |
+
"loss": 0.3781,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.65,
|
| 1791 |
+
"grad_norm": 34.23354721069336,
|
| 1792 |
+
"learning_rate": 6.91301000769823e-05,
|
| 1793 |
+
"loss": 1.8894,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.66,
|
| 1798 |
+
"grad_norm": 0.40863272547721863,
|
| 1799 |
+
"learning_rate": 6.861688478316654e-05,
|
| 1800 |
+
"loss": 0.8985,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.66,
|
| 1805 |
+
"grad_norm": 0.26213815808296204,
|
| 1806 |
+
"learning_rate": 6.810366948935078e-05,
|
| 1807 |
+
"loss": 1.038,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.66,
|
| 1812 |
+
"grad_norm": 32.263370513916016,
|
| 1813 |
+
"learning_rate": 6.759045419553503e-05,
|
| 1814 |
+
"loss": 2.2679,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.66,
|
| 1819 |
+
"grad_norm": 34.00870895385742,
|
| 1820 |
+
"learning_rate": 6.707723890171927e-05,
|
| 1821 |
+
"loss": 2.075,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.67,
|
| 1826 |
+
"grad_norm": 0.4284890294075012,
|
| 1827 |
+
"learning_rate": 6.656402360790351e-05,
|
| 1828 |
+
"loss": 0.7597,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.67,
|
| 1833 |
+
"grad_norm": 0.3407612144947052,
|
| 1834 |
+
"learning_rate": 6.605080831408776e-05,
|
| 1835 |
+
"loss": 0.4383,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.67,
|
| 1840 |
+
"grad_norm": 0.467129111289978,
|
| 1841 |
+
"learning_rate": 6.5537593020272e-05,
|
| 1842 |
+
"loss": 1.2687,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.67,
|
| 1847 |
+
"grad_norm": 0.13389898836612701,
|
| 1848 |
+
"learning_rate": 6.502437772645625e-05,
|
| 1849 |
+
"loss": 2.1633,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.68,
|
| 1854 |
+
"grad_norm": 0.13675318658351898,
|
| 1855 |
+
"learning_rate": 6.451116243264049e-05,
|
| 1856 |
+
"loss": 0.7836,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.68,
|
| 1861 |
+
"grad_norm": 0.5350820422172546,
|
| 1862 |
+
"learning_rate": 6.399794713882474e-05,
|
| 1863 |
+
"loss": 0.7649,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.68,
|
| 1868 |
+
"grad_norm": 0.48448631167411804,
|
| 1869 |
+
"learning_rate": 6.348473184500898e-05,
|
| 1870 |
+
"loss": 1.38,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.69,
|
| 1875 |
+
"grad_norm": 0.5778862237930298,
|
| 1876 |
+
"learning_rate": 6.297151655119323e-05,
|
| 1877 |
+
"loss": 2.2181,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.69,
|
| 1882 |
+
"grad_norm": 1.1682463884353638,
|
| 1883 |
+
"learning_rate": 6.245830125737747e-05,
|
| 1884 |
+
"loss": 1.0876,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.69,
|
| 1889 |
+
"grad_norm": 31.393110275268555,
|
| 1890 |
+
"learning_rate": 6.194508596356173e-05,
|
| 1891 |
+
"loss": 1.1128,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.69,
|
| 1896 |
+
"grad_norm": 0.6889828443527222,
|
| 1897 |
+
"learning_rate": 6.143187066974596e-05,
|
| 1898 |
+
"loss": 0.7252,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.7,
|
| 1903 |
+
"grad_norm": 0.5311009883880615,
|
| 1904 |
+
"learning_rate": 6.0918655375930204e-05,
|
| 1905 |
+
"loss": 0.3959,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.7,
|
| 1910 |
+
"grad_norm": 0.43320441246032715,
|
| 1911 |
+
"learning_rate": 6.040544008211445e-05,
|
| 1912 |
+
"loss": 1.2466,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.7,
|
| 1917 |
+
"grad_norm": 0.06079118326306343,
|
| 1918 |
+
"learning_rate": 5.989222478829869e-05,
|
| 1919 |
+
"loss": 2.0125,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.7,
|
| 1924 |
+
"grad_norm": 0.042944520711898804,
|
| 1925 |
+
"learning_rate": 5.937900949448294e-05,
|
| 1926 |
+
"loss": 0.779,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.71,
|
| 1931 |
+
"grad_norm": 0.9245113730430603,
|
| 1932 |
+
"learning_rate": 5.886579420066718e-05,
|
| 1933 |
+
"loss": 0.7976,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.71,
|
| 1938 |
+
"grad_norm": 0.039791032671928406,
|
| 1939 |
+
"learning_rate": 5.835257890685143e-05,
|
| 1940 |
+
"loss": 1.4881,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.71,
|
| 1945 |
+
"grad_norm": 0.6234956979751587,
|
| 1946 |
+
"learning_rate": 5.783936361303567e-05,
|
| 1947 |
+
"loss": 0.6912,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.71,
|
| 1952 |
+
"grad_norm": 0.9020193219184875,
|
| 1953 |
+
"learning_rate": 5.7326148319219916e-05,
|
| 1954 |
+
"loss": 0.7522,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.72,
|
| 1959 |
+
"grad_norm": 0.034982066601514816,
|
| 1960 |
+
"learning_rate": 5.681293302540416e-05,
|
| 1961 |
+
"loss": 1.2431,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.72,
|
| 1966 |
+
"grad_norm": 0.04073556512594223,
|
| 1967 |
+
"learning_rate": 5.6299717731588407e-05,
|
| 1968 |
+
"loss": 1.1053,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.72,
|
| 1973 |
+
"grad_norm": 35.25714874267578,
|
| 1974 |
+
"learning_rate": 5.5786502437772645e-05,
|
| 1975 |
+
"loss": 1.4825,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.72,
|
| 1980 |
+
"grad_norm": 0.7252044081687927,
|
| 1981 |
+
"learning_rate": 5.52732871439569e-05,
|
| 1982 |
+
"loss": 2.1285,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.73,
|
| 1987 |
+
"grad_norm": 31.510704040527344,
|
| 1988 |
+
"learning_rate": 5.4760071850141136e-05,
|
| 1989 |
+
"loss": 0.7914,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.73,
|
| 1994 |
+
"grad_norm": 0.44585466384887695,
|
| 1995 |
+
"learning_rate": 5.4246856556325375e-05,
|
| 1996 |
+
"loss": 1.2094,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.73,
|
| 2001 |
+
"grad_norm": 0.2099919468164444,
|
| 2002 |
+
"learning_rate": 5.373364126250963e-05,
|
| 2003 |
+
"loss": 2.496,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.73,
|
| 2008 |
+
"grad_norm": 30.501766204833984,
|
| 2009 |
+
"learning_rate": 5.3220425968693866e-05,
|
| 2010 |
+
"loss": 1.0997,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.74,
|
| 2015 |
+
"grad_norm": 0.36132165789604187,
|
| 2016 |
+
"learning_rate": 5.270721067487812e-05,
|
| 2017 |
+
"loss": 1.4378,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.74,
|
| 2022 |
+
"grad_norm": 1.6164754629135132,
|
| 2023 |
+
"learning_rate": 5.219399538106236e-05,
|
| 2024 |
+
"loss": 1.1191,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.74,
|
| 2029 |
+
"grad_norm": 0.9825451374053955,
|
| 2030 |
+
"learning_rate": 5.16807800872466e-05,
|
| 2031 |
+
"loss": 0.861,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.74,
|
| 2036 |
+
"grad_norm": 0.28137707710266113,
|
| 2037 |
+
"learning_rate": 5.116756479343084e-05,
|
| 2038 |
+
"loss": 0.4387,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.75,
|
| 2043 |
+
"grad_norm": 0.2143826186656952,
|
| 2044 |
+
"learning_rate": 5.0654349499615094e-05,
|
| 2045 |
+
"loss": 0.9561,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.75,
|
| 2050 |
+
"grad_norm": 0.25483042001724243,
|
| 2051 |
+
"learning_rate": 5.014113420579933e-05,
|
| 2052 |
+
"loss": 1.4271,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.75,
|
| 2057 |
+
"grad_norm": 0.36998990178108215,
|
| 2058 |
+
"learning_rate": 4.962791891198358e-05,
|
| 2059 |
+
"loss": 0.8621,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.75,
|
| 2064 |
+
"grad_norm": 0.08702105283737183,
|
| 2065 |
+
"learning_rate": 4.9114703618167824e-05,
|
| 2066 |
+
"loss": 0.4751,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.76,
|
| 2071 |
+
"grad_norm": 0.455074667930603,
|
| 2072 |
+
"learning_rate": 4.860148832435207e-05,
|
| 2073 |
+
"loss": 2.7804,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.76,
|
| 2078 |
+
"grad_norm": 0.41690900921821594,
|
| 2079 |
+
"learning_rate": 4.8088273030536315e-05,
|
| 2080 |
+
"loss": 0.444,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.76,
|
| 2085 |
+
"grad_norm": 0.0824534222483635,
|
| 2086 |
+
"learning_rate": 4.757505773672056e-05,
|
| 2087 |
+
"loss": 2.2328,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.76,
|
| 2092 |
+
"grad_norm": 0.7097424268722534,
|
| 2093 |
+
"learning_rate": 4.70618424429048e-05,
|
| 2094 |
+
"loss": 1.232,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.77,
|
| 2099 |
+
"grad_norm": 32.93928527832031,
|
| 2100 |
+
"learning_rate": 4.6548627149089045e-05,
|
| 2101 |
+
"loss": 2.1723,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.77,
|
| 2106 |
+
"grad_norm": 1.5564465522766113,
|
| 2107 |
+
"learning_rate": 4.603541185527329e-05,
|
| 2108 |
+
"loss": 1.3908,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.77,
|
| 2113 |
+
"grad_norm": 1.1192642450332642,
|
| 2114 |
+
"learning_rate": 4.5522196561457536e-05,
|
| 2115 |
+
"loss": 0.2986,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.77,
|
| 2120 |
+
"grad_norm": 0.09087738394737244,
|
| 2121 |
+
"learning_rate": 4.5008981267641774e-05,
|
| 2122 |
+
"loss": 1.3239,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.78,
|
| 2127 |
+
"grad_norm": 31.66700553894043,
|
| 2128 |
+
"learning_rate": 4.449576597382602e-05,
|
| 2129 |
+
"loss": 1.9121,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.78,
|
| 2134 |
+
"grad_norm": 0.08067402243614197,
|
| 2135 |
+
"learning_rate": 4.3982550680010266e-05,
|
| 2136 |
+
"loss": 1.4794,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.78,
|
| 2141 |
+
"grad_norm": 31.138818740844727,
|
| 2142 |
+
"learning_rate": 4.346933538619451e-05,
|
| 2143 |
+
"loss": 1.3591,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.79,
|
| 2148 |
+
"grad_norm": 30.739635467529297,
|
| 2149 |
+
"learning_rate": 4.2956120092378757e-05,
|
| 2150 |
+
"loss": 2.1444,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.79,
|
| 2155 |
+
"grad_norm": 29.64994239807129,
|
| 2156 |
+
"learning_rate": 4.2442904798563e-05,
|
| 2157 |
+
"loss": 1.1349,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.79,
|
| 2162 |
+
"grad_norm": 0.06835217773914337,
|
| 2163 |
+
"learning_rate": 4.192968950474724e-05,
|
| 2164 |
+
"loss": 1.1995,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.79,
|
| 2169 |
+
"grad_norm": 0.07994609326124191,
|
| 2170 |
+
"learning_rate": 4.1416474210931486e-05,
|
| 2171 |
+
"loss": 1.106,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.8,
|
| 2176 |
+
"grad_norm": 0.077377088367939,
|
| 2177 |
+
"learning_rate": 4.090325891711573e-05,
|
| 2178 |
+
"loss": 0.5515,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.8,
|
| 2183 |
+
"grad_norm": 0.07298921793699265,
|
| 2184 |
+
"learning_rate": 4.039004362329998e-05,
|
| 2185 |
+
"loss": 1.3587,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.8,
|
| 2190 |
+
"grad_norm": 2.3832900524139404,
|
| 2191 |
+
"learning_rate": 3.987682832948422e-05,
|
| 2192 |
+
"loss": 0.5702,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.8,
|
| 2197 |
+
"grad_norm": 1.3254412412643433,
|
| 2198 |
+
"learning_rate": 3.936361303566847e-05,
|
| 2199 |
+
"loss": 0.6837,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.81,
|
| 2204 |
+
"grad_norm": 31.57373046875,
|
| 2205 |
+
"learning_rate": 3.885039774185271e-05,
|
| 2206 |
+
"loss": 1.4529,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.81,
|
| 2211 |
+
"grad_norm": 31.78739356994629,
|
| 2212 |
+
"learning_rate": 3.833718244803695e-05,
|
| 2213 |
+
"loss": 1.1592,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.81,
|
| 2218 |
+
"grad_norm": 0.885814368724823,
|
| 2219 |
+
"learning_rate": 3.78239671542212e-05,
|
| 2220 |
+
"loss": 2.0692,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.81,
|
| 2225 |
+
"grad_norm": 0.11006342619657516,
|
| 2226 |
+
"learning_rate": 3.7310751860405444e-05,
|
| 2227 |
+
"loss": 1.3088,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.82,
|
| 2232 |
+
"grad_norm": 0.7606900930404663,
|
| 2233 |
+
"learning_rate": 3.679753656658968e-05,
|
| 2234 |
+
"loss": 0.7503,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.82,
|
| 2239 |
+
"grad_norm": 34.01801681518555,
|
| 2240 |
+
"learning_rate": 3.628432127277393e-05,
|
| 2241 |
+
"loss": 2.3927,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.82,
|
| 2246 |
+
"grad_norm": 31.359045028686523,
|
| 2247 |
+
"learning_rate": 3.5771105978958174e-05,
|
| 2248 |
+
"loss": 1.7152,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.82,
|
| 2253 |
+
"grad_norm": 0.4896736741065979,
|
| 2254 |
+
"learning_rate": 3.525789068514242e-05,
|
| 2255 |
+
"loss": 1.1745,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.83,
|
| 2260 |
+
"grad_norm": 33.61710739135742,
|
| 2261 |
+
"learning_rate": 3.4744675391326665e-05,
|
| 2262 |
+
"loss": 2.6372,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.83,
|
| 2267 |
+
"grad_norm": 0.2002750039100647,
|
| 2268 |
+
"learning_rate": 3.423146009751091e-05,
|
| 2269 |
+
"loss": 0.3958,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.83,
|
| 2274 |
+
"grad_norm": 0.5248939394950867,
|
| 2275 |
+
"learning_rate": 3.3718244803695156e-05,
|
| 2276 |
+
"loss": 1.1428,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.83,
|
| 2281 |
+
"grad_norm": 0.20893482863903046,
|
| 2282 |
+
"learning_rate": 3.32050295098794e-05,
|
| 2283 |
+
"loss": 1.1695,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.84,
|
| 2288 |
+
"grad_norm": 31.339996337890625,
|
| 2289 |
+
"learning_rate": 3.269181421606364e-05,
|
| 2290 |
+
"loss": 1.9207,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.84,
|
| 2295 |
+
"grad_norm": 0.20348432660102844,
|
| 2296 |
+
"learning_rate": 3.2178598922247886e-05,
|
| 2297 |
+
"loss": 1.587,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.84,
|
| 2302 |
+
"grad_norm": 31.108985900878906,
|
| 2303 |
+
"learning_rate": 3.1665383628432124e-05,
|
| 2304 |
+
"loss": 1.3677,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.84,
|
| 2309 |
+
"grad_norm": 29.706878662109375,
|
| 2310 |
+
"learning_rate": 3.115216833461637e-05,
|
| 2311 |
+
"loss": 2.0426,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.85,
|
| 2316 |
+
"grad_norm": 1.0215935707092285,
|
| 2317 |
+
"learning_rate": 3.0638953040800616e-05,
|
| 2318 |
+
"loss": 1.7041,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.85,
|
| 2323 |
+
"grad_norm": 0.23811456561088562,
|
| 2324 |
+
"learning_rate": 3.012573774698486e-05,
|
| 2325 |
+
"loss": 1.2077,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.85,
|
| 2330 |
+
"grad_norm": 0.21963796019554138,
|
| 2331 |
+
"learning_rate": 2.9612522453169107e-05,
|
| 2332 |
+
"loss": 1.1672,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.85,
|
| 2337 |
+
"grad_norm": 1.149311900138855,
|
| 2338 |
+
"learning_rate": 2.9099307159353352e-05,
|
| 2339 |
+
"loss": 1.6651,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.86,
|
| 2344 |
+
"grad_norm": 1.1534388065338135,
|
| 2345 |
+
"learning_rate": 2.8586091865537594e-05,
|
| 2346 |
+
"loss": 1.2245,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.86,
|
| 2351 |
+
"grad_norm": 1.3270301818847656,
|
| 2352 |
+
"learning_rate": 2.807287657172184e-05,
|
| 2353 |
+
"loss": 0.6268,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.86,
|
| 2358 |
+
"grad_norm": 0.2713325619697571,
|
| 2359 |
+
"learning_rate": 2.7559661277906085e-05,
|
| 2360 |
+
"loss": 1.6962,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.86,
|
| 2365 |
+
"grad_norm": 0.25903043150901794,
|
| 2366 |
+
"learning_rate": 2.704644598409033e-05,
|
| 2367 |
+
"loss": 0.4256,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.87,
|
| 2372 |
+
"grad_norm": 0.456488698720932,
|
| 2373 |
+
"learning_rate": 2.6533230690274573e-05,
|
| 2374 |
+
"loss": 0.7452,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.87,
|
| 2379 |
+
"grad_norm": 30.572643280029297,
|
| 2380 |
+
"learning_rate": 2.6020015396458812e-05,
|
| 2381 |
+
"loss": 1.9218,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.87,
|
| 2386 |
+
"grad_norm": 31.96410369873047,
|
| 2387 |
+
"learning_rate": 2.5506800102643057e-05,
|
| 2388 |
+
"loss": 2.2211,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.88,
|
| 2393 |
+
"grad_norm": 0.5312764644622803,
|
| 2394 |
+
"learning_rate": 2.4993584808827306e-05,
|
| 2395 |
+
"loss": 0.6983,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.88,
|
| 2400 |
+
"grad_norm": 0.23434561491012573,
|
| 2401 |
+
"learning_rate": 2.448036951501155e-05,
|
| 2402 |
+
"loss": 0.7466,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.88,
|
| 2407 |
+
"grad_norm": 36.59482192993164,
|
| 2408 |
+
"learning_rate": 2.3967154221195794e-05,
|
| 2409 |
+
"loss": 1.6694,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.88,
|
| 2414 |
+
"grad_norm": 32.04636001586914,
|
| 2415 |
+
"learning_rate": 2.3453938927380036e-05,
|
| 2416 |
+
"loss": 1.6496,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.89,
|
| 2421 |
+
"grad_norm": 0.2819555699825287,
|
| 2422 |
+
"learning_rate": 2.294072363356428e-05,
|
| 2423 |
+
"loss": 1.3424,
|
| 2424 |
+
"step": 3450
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.89,
|
| 2428 |
+
"grad_norm": 0.46713873744010925,
|
| 2429 |
+
"learning_rate": 2.2427508339748527e-05,
|
| 2430 |
+
"loss": 0.0149,
|
| 2431 |
+
"step": 3460
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.89,
|
| 2435 |
+
"grad_norm": 0.5181002020835876,
|
| 2436 |
+
"learning_rate": 2.1914293045932773e-05,
|
| 2437 |
+
"loss": 0.7906,
|
| 2438 |
+
"step": 3470
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.89,
|
| 2442 |
+
"grad_norm": 0.20827196538448334,
|
| 2443 |
+
"learning_rate": 2.140107775211701e-05,
|
| 2444 |
+
"loss": 0.4505,
|
| 2445 |
+
"step": 3480
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.9,
|
| 2449 |
+
"grad_norm": 32.157432556152344,
|
| 2450 |
+
"learning_rate": 2.0887862458301257e-05,
|
| 2451 |
+
"loss": 2.5997,
|
| 2452 |
+
"step": 3490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.9,
|
| 2456 |
+
"grad_norm": 0.1746426224708557,
|
| 2457 |
+
"learning_rate": 2.0374647164485502e-05,
|
| 2458 |
+
"loss": 1.6267,
|
| 2459 |
+
"step": 3500
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.9,
|
| 2463 |
+
"grad_norm": 31.728593826293945,
|
| 2464 |
+
"learning_rate": 1.9861431870669748e-05,
|
| 2465 |
+
"loss": 0.8132,
|
| 2466 |
+
"step": 3510
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.9,
|
| 2470 |
+
"grad_norm": 0.8428208231925964,
|
| 2471 |
+
"learning_rate": 1.9348216576853994e-05,
|
| 2472 |
+
"loss": 1.4408,
|
| 2473 |
+
"step": 3520
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.91,
|
| 2477 |
+
"grad_norm": 0.6710448861122131,
|
| 2478 |
+
"learning_rate": 1.8835001283038236e-05,
|
| 2479 |
+
"loss": 0.7472,
|
| 2480 |
+
"step": 3530
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.91,
|
| 2484 |
+
"grad_norm": 0.15114788711071014,
|
| 2485 |
+
"learning_rate": 1.8321785989222478e-05,
|
| 2486 |
+
"loss": 0.3549,
|
| 2487 |
+
"step": 3540
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.91,
|
| 2491 |
+
"grad_norm": 0.48726460337638855,
|
| 2492 |
+
"learning_rate": 1.7808570695406723e-05,
|
| 2493 |
+
"loss": 1.0424,
|
| 2494 |
+
"step": 3550
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.91,
|
| 2498 |
+
"grad_norm": 31.90473747253418,
|
| 2499 |
+
"learning_rate": 1.729535540159097e-05,
|
| 2500 |
+
"loss": 1.6014,
|
| 2501 |
+
"step": 3560
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.92,
|
| 2505 |
+
"grad_norm": 0.11974696815013885,
|
| 2506 |
+
"learning_rate": 1.678214010777521e-05,
|
| 2507 |
+
"loss": 1.367,
|
| 2508 |
+
"step": 3570
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.92,
|
| 2512 |
+
"grad_norm": 31.7264461517334,
|
| 2513 |
+
"learning_rate": 1.6268924813959457e-05,
|
| 2514 |
+
"loss": 0.8339,
|
| 2515 |
+
"step": 3580
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.92,
|
| 2519 |
+
"grad_norm": 0.9428460001945496,
|
| 2520 |
+
"learning_rate": 1.5755709520143702e-05,
|
| 2521 |
+
"loss": 1.2831,
|
| 2522 |
+
"step": 3590
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.92,
|
| 2526 |
+
"grad_norm": 0.5526345372200012,
|
| 2527 |
+
"learning_rate": 1.5242494226327944e-05,
|
| 2528 |
+
"loss": 0.7811,
|
| 2529 |
+
"step": 3600
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.93,
|
| 2533 |
+
"grad_norm": 0.4995884299278259,
|
| 2534 |
+
"learning_rate": 1.4729278932512188e-05,
|
| 2535 |
+
"loss": 0.454,
|
| 2536 |
+
"step": 3610
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.93,
|
| 2540 |
+
"grad_norm": 32.1862907409668,
|
| 2541 |
+
"learning_rate": 1.4216063638696434e-05,
|
| 2542 |
+
"loss": 1.7057,
|
| 2543 |
+
"step": 3620
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.93,
|
| 2547 |
+
"grad_norm": 30.95826530456543,
|
| 2548 |
+
"learning_rate": 1.3702848344880677e-05,
|
| 2549 |
+
"loss": 1.6207,
|
| 2550 |
+
"step": 3630
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.93,
|
| 2554 |
+
"grad_norm": 0.10005848109722137,
|
| 2555 |
+
"learning_rate": 1.3189633051064923e-05,
|
| 2556 |
+
"loss": 0.8695,
|
| 2557 |
+
"step": 3640
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.94,
|
| 2561 |
+
"grad_norm": 31.5998477935791,
|
| 2562 |
+
"learning_rate": 1.2676417757249169e-05,
|
| 2563 |
+
"loss": 1.9645,
|
| 2564 |
+
"step": 3650
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.94,
|
| 2568 |
+
"grad_norm": 0.797754168510437,
|
| 2569 |
+
"learning_rate": 1.216320246343341e-05,
|
| 2570 |
+
"loss": 1.1224,
|
| 2571 |
+
"step": 3660
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.94,
|
| 2575 |
+
"grad_norm": 0.10280079394578934,
|
| 2576 |
+
"learning_rate": 1.1649987169617656e-05,
|
| 2577 |
+
"loss": 0.8117,
|
| 2578 |
+
"step": 3670
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.94,
|
| 2582 |
+
"grad_norm": 31.812061309814453,
|
| 2583 |
+
"learning_rate": 1.1136771875801898e-05,
|
| 2584 |
+
"loss": 0.5172,
|
| 2585 |
+
"step": 3680
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.95,
|
| 2589 |
+
"grad_norm": 0.08260977268218994,
|
| 2590 |
+
"learning_rate": 1.0623556581986144e-05,
|
| 2591 |
+
"loss": 1.6816,
|
| 2592 |
+
"step": 3690
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.95,
|
| 2596 |
+
"grad_norm": 0.6431272029876709,
|
| 2597 |
+
"learning_rate": 1.0110341288170388e-05,
|
| 2598 |
+
"loss": 2.2201,
|
| 2599 |
+
"step": 3700
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.95,
|
| 2603 |
+
"grad_norm": 0.5408867597579956,
|
| 2604 |
+
"learning_rate": 9.597125994354632e-06,
|
| 2605 |
+
"loss": 0.9963,
|
| 2606 |
+
"step": 3710
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.95,
|
| 2610 |
+
"grad_norm": 0.6773453950881958,
|
| 2611 |
+
"learning_rate": 9.083910700538877e-06,
|
| 2612 |
+
"loss": 2.3436,
|
| 2613 |
+
"step": 3720
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.96,
|
| 2617 |
+
"grad_norm": 33.85165786743164,
|
| 2618 |
+
"learning_rate": 8.570695406723121e-06,
|
| 2619 |
+
"loss": 1.229,
|
| 2620 |
+
"step": 3730
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.96,
|
| 2624 |
+
"grad_norm": 0.09623867273330688,
|
| 2625 |
+
"learning_rate": 8.057480112907365e-06,
|
| 2626 |
+
"loss": 0.9852,
|
| 2627 |
+
"step": 3740
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.96,
|
| 2631 |
+
"grad_norm": 0.6665917038917542,
|
| 2632 |
+
"learning_rate": 7.5442648190916095e-06,
|
| 2633 |
+
"loss": 0.9811,
|
| 2634 |
+
"step": 3750
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.96,
|
| 2638 |
+
"grad_norm": 0.1049143448472023,
|
| 2639 |
+
"learning_rate": 7.031049525275854e-06,
|
| 2640 |
+
"loss": 1.6639,
|
| 2641 |
+
"step": 3760
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.97,
|
| 2645 |
+
"grad_norm": 1.028926968574524,
|
| 2646 |
+
"learning_rate": 6.517834231460097e-06,
|
| 2647 |
+
"loss": 0.6745,
|
| 2648 |
+
"step": 3770
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.97,
|
| 2652 |
+
"grad_norm": 0.5345853567123413,
|
| 2653 |
+
"learning_rate": 6.004618937644342e-06,
|
| 2654 |
+
"loss": 1.9351,
|
| 2655 |
+
"step": 3780
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.97,
|
| 2659 |
+
"grad_norm": 0.09490037709474564,
|
| 2660 |
+
"learning_rate": 5.4914036438285865e-06,
|
| 2661 |
+
"loss": 1.0257,
|
| 2662 |
+
"step": 3790
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.98,
|
| 2666 |
+
"grad_norm": 0.7808061838150024,
|
| 2667 |
+
"learning_rate": 4.978188350012831e-06,
|
| 2668 |
+
"loss": 1.6728,
|
| 2669 |
+
"step": 3800
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.98,
|
| 2673 |
+
"grad_norm": 0.4944613575935364,
|
| 2674 |
+
"learning_rate": 4.464973056197075e-06,
|
| 2675 |
+
"loss": 1.6118,
|
| 2676 |
+
"step": 3810
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.98,
|
| 2680 |
+
"grad_norm": 1.0724005699157715,
|
| 2681 |
+
"learning_rate": 3.951757762381319e-06,
|
| 2682 |
+
"loss": 1.352,
|
| 2683 |
+
"step": 3820
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.98,
|
| 2687 |
+
"grad_norm": 0.8290554285049438,
|
| 2688 |
+
"learning_rate": 3.4385424685655636e-06,
|
| 2689 |
+
"loss": 2.4513,
|
| 2690 |
+
"step": 3830
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.99,
|
| 2694 |
+
"grad_norm": 0.532361626625061,
|
| 2695 |
+
"learning_rate": 2.9253271747498074e-06,
|
| 2696 |
+
"loss": 0.412,
|
| 2697 |
+
"step": 3840
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.99,
|
| 2701 |
+
"grad_norm": 0.1319778561592102,
|
| 2702 |
+
"learning_rate": 2.412111880934052e-06,
|
| 2703 |
+
"loss": 0.7912,
|
| 2704 |
+
"step": 3850
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.99,
|
| 2708 |
+
"grad_norm": 0.6048825979232788,
|
| 2709 |
+
"learning_rate": 1.8988965871182964e-06,
|
| 2710 |
+
"loss": 0.9813,
|
| 2711 |
+
"step": 3860
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.99,
|
| 2715 |
+
"grad_norm": 0.4283379018306732,
|
| 2716 |
+
"learning_rate": 1.3856812933025406e-06,
|
| 2717 |
+
"loss": 0.7374,
|
| 2718 |
+
"step": 3870
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 1.0,
|
| 2722 |
+
"grad_norm": 0.7439431548118591,
|
| 2723 |
+
"learning_rate": 8.724659994867848e-07,
|
| 2724 |
+
"loss": 0.3349,
|
| 2725 |
+
"step": 3880
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 1.0,
|
| 2729 |
+
"grad_norm": 0.49157199263572693,
|
| 2730 |
+
"learning_rate": 3.5925070567102904e-07,
|
| 2731 |
+
"loss": 1.4183,
|
| 2732 |
+
"step": 3890
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 1.0,
|
| 2736 |
+
"step": 3897,
|
| 2737 |
+
"total_flos": 9.121574551619174e+17,
|
| 2738 |
+
"train_loss": 1.2556480791153037,
|
| 2739 |
+
"train_runtime": 633.8251,
|
| 2740 |
+
"train_samples_per_second": 18.444,
|
| 2741 |
+
"train_steps_per_second": 6.148
|
| 2742 |
+
}
|
| 2743 |
+
],
|
| 2744 |
+
"logging_steps": 10,
|
| 2745 |
+
"max_steps": 3897,
|
| 2746 |
+
"num_input_tokens_seen": 0,
|
| 2747 |
+
"num_train_epochs": 1,
|
| 2748 |
+
"save_steps": 1000,
|
| 2749 |
+
"total_flos": 9.121574551619174e+17,
|
| 2750 |
+
"train_batch_size": 3,
|
| 2751 |
+
"trial_name": null,
|
| 2752 |
+
"trial_params": null
|
| 2753 |
+
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4920
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d74d1eabcbb9a8fa35f017225dd6959cab8db2e1365e4b2efe0be055d9fbb23f
|
| 3 |
size 4920
|