Training in progress, step 7800, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1000555808
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eb1d3fbeeaecaab609152e0153a01088cbd5c9ccb619aed7ff28fca6691379d0
|
| 3 |
size 1000555808
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1318473087
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae4aa17d380b71d4eed886f6e0f14d94362ff7b5c90e24cd161246c0252c9fd1
|
| 3 |
size 1318473087
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1b92b3721101d47b0835d502a18707c17451e53a5fc7f24dfcdbc3f097658d48
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -52508,6 +52508,2106 @@
|
|
| 52508 |
"learning_rate": 0.000517970140921568,
|
| 52509 |
"loss": 2.1059,
|
| 52510 |
"step": 7500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 52511 |
}
|
| 52512 |
],
|
| 52513 |
"logging_steps": 1,
|
|
@@ -52527,7 +54627,7 @@
|
|
| 52527 |
"attributes": {}
|
| 52528 |
}
|
| 52529 |
},
|
| 52530 |
-
"total_flos":
|
| 52531 |
"train_batch_size": 8,
|
| 52532 |
"trial_name": null,
|
| 52533 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.43333333333333335,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 7800,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 52508 |
"learning_rate": 0.000517970140921568,
|
| 52509 |
"loss": 2.1059,
|
| 52510 |
"step": 7500
|
| 52511 |
+
},
|
| 52512 |
+
{
|
| 52513 |
+
"epoch": 0.4167222222222222,
|
| 52514 |
+
"grad_norm": 0.1044921875,
|
| 52515 |
+
"learning_rate": 0.0005179020697643617,
|
| 52516 |
+
"loss": 2.1059,
|
| 52517 |
+
"step": 7501
|
| 52518 |
+
},
|
| 52519 |
+
{
|
| 52520 |
+
"epoch": 0.4167777777777778,
|
| 52521 |
+
"grad_norm": 0.10498046875,
|
| 52522 |
+
"learning_rate": 0.0005178339948675653,
|
| 52523 |
+
"loss": 2.0724,
|
| 52524 |
+
"step": 7502
|
| 52525 |
+
},
|
| 52526 |
+
{
|
| 52527 |
+
"epoch": 0.41683333333333333,
|
| 52528 |
+
"grad_norm": 0.09814453125,
|
| 52529 |
+
"learning_rate": 0.0005177659162333379,
|
| 52530 |
+
"loss": 2.0396,
|
| 52531 |
+
"step": 7503
|
| 52532 |
+
},
|
| 52533 |
+
{
|
| 52534 |
+
"epoch": 0.41688888888888886,
|
| 52535 |
+
"grad_norm": 0.09765625,
|
| 52536 |
+
"learning_rate": 0.0005176978338638387,
|
| 52537 |
+
"loss": 2.0582,
|
| 52538 |
+
"step": 7504
|
| 52539 |
+
},
|
| 52540 |
+
{
|
| 52541 |
+
"epoch": 0.41694444444444445,
|
| 52542 |
+
"grad_norm": 0.11083984375,
|
| 52543 |
+
"learning_rate": 0.0005176297477612273,
|
| 52544 |
+
"loss": 2.0626,
|
| 52545 |
+
"step": 7505
|
| 52546 |
+
},
|
| 52547 |
+
{
|
| 52548 |
+
"epoch": 0.417,
|
| 52549 |
+
"grad_norm": 0.10009765625,
|
| 52550 |
+
"learning_rate": 0.0005175616579276632,
|
| 52551 |
+
"loss": 2.0836,
|
| 52552 |
+
"step": 7506
|
| 52553 |
+
},
|
| 52554 |
+
{
|
| 52555 |
+
"epoch": 0.41705555555555557,
|
| 52556 |
+
"grad_norm": 0.11181640625,
|
| 52557 |
+
"learning_rate": 0.000517493564365306,
|
| 52558 |
+
"loss": 2.1313,
|
| 52559 |
+
"step": 7507
|
| 52560 |
+
},
|
| 52561 |
+
{
|
| 52562 |
+
"epoch": 0.4171111111111111,
|
| 52563 |
+
"grad_norm": 0.10302734375,
|
| 52564 |
+
"learning_rate": 0.0005174254670763154,
|
| 52565 |
+
"loss": 2.07,
|
| 52566 |
+
"step": 7508
|
| 52567 |
+
},
|
| 52568 |
+
{
|
| 52569 |
+
"epoch": 0.4171666666666667,
|
| 52570 |
+
"grad_norm": 0.1025390625,
|
| 52571 |
+
"learning_rate": 0.0005173573660628514,
|
| 52572 |
+
"loss": 2.1438,
|
| 52573 |
+
"step": 7509
|
| 52574 |
+
},
|
| 52575 |
+
{
|
| 52576 |
+
"epoch": 0.4172222222222222,
|
| 52577 |
+
"grad_norm": 0.0986328125,
|
| 52578 |
+
"learning_rate": 0.000517289261327074,
|
| 52579 |
+
"loss": 2.1405,
|
| 52580 |
+
"step": 7510
|
| 52581 |
+
},
|
| 52582 |
+
{
|
| 52583 |
+
"epoch": 0.4172777777777778,
|
| 52584 |
+
"grad_norm": 0.10546875,
|
| 52585 |
+
"learning_rate": 0.0005172211528711435,
|
| 52586 |
+
"loss": 2.13,
|
| 52587 |
+
"step": 7511
|
| 52588 |
+
},
|
| 52589 |
+
{
|
| 52590 |
+
"epoch": 0.41733333333333333,
|
| 52591 |
+
"grad_norm": 0.10107421875,
|
| 52592 |
+
"learning_rate": 0.0005171530406972196,
|
| 52593 |
+
"loss": 2.1344,
|
| 52594 |
+
"step": 7512
|
| 52595 |
+
},
|
| 52596 |
+
{
|
| 52597 |
+
"epoch": 0.41738888888888886,
|
| 52598 |
+
"grad_norm": 0.10302734375,
|
| 52599 |
+
"learning_rate": 0.0005170849248074634,
|
| 52600 |
+
"loss": 2.1153,
|
| 52601 |
+
"step": 7513
|
| 52602 |
+
},
|
| 52603 |
+
{
|
| 52604 |
+
"epoch": 0.41744444444444445,
|
| 52605 |
+
"grad_norm": 0.0986328125,
|
| 52606 |
+
"learning_rate": 0.0005170168052040349,
|
| 52607 |
+
"loss": 2.0954,
|
| 52608 |
+
"step": 7514
|
| 52609 |
+
},
|
| 52610 |
+
{
|
| 52611 |
+
"epoch": 0.4175,
|
| 52612 |
+
"grad_norm": 0.10400390625,
|
| 52613 |
+
"learning_rate": 0.0005169486818890947,
|
| 52614 |
+
"loss": 2.0643,
|
| 52615 |
+
"step": 7515
|
| 52616 |
+
},
|
| 52617 |
+
{
|
| 52618 |
+
"epoch": 0.41755555555555557,
|
| 52619 |
+
"grad_norm": 0.10107421875,
|
| 52620 |
+
"learning_rate": 0.0005168805548648038,
|
| 52621 |
+
"loss": 2.0501,
|
| 52622 |
+
"step": 7516
|
| 52623 |
+
},
|
| 52624 |
+
{
|
| 52625 |
+
"epoch": 0.4176111111111111,
|
| 52626 |
+
"grad_norm": 0.10009765625,
|
| 52627 |
+
"learning_rate": 0.0005168124241333228,
|
| 52628 |
+
"loss": 2.1143,
|
| 52629 |
+
"step": 7517
|
| 52630 |
+
},
|
| 52631 |
+
{
|
| 52632 |
+
"epoch": 0.4176666666666667,
|
| 52633 |
+
"grad_norm": 0.09765625,
|
| 52634 |
+
"learning_rate": 0.0005167442896968128,
|
| 52635 |
+
"loss": 2.093,
|
| 52636 |
+
"step": 7518
|
| 52637 |
+
},
|
| 52638 |
+
{
|
| 52639 |
+
"epoch": 0.4177222222222222,
|
| 52640 |
+
"grad_norm": 0.099609375,
|
| 52641 |
+
"learning_rate": 0.0005166761515574347,
|
| 52642 |
+
"loss": 2.0664,
|
| 52643 |
+
"step": 7519
|
| 52644 |
+
},
|
| 52645 |
+
{
|
| 52646 |
+
"epoch": 0.4177777777777778,
|
| 52647 |
+
"grad_norm": 0.10546875,
|
| 52648 |
+
"learning_rate": 0.0005166080097173499,
|
| 52649 |
+
"loss": 2.1043,
|
| 52650 |
+
"step": 7520
|
| 52651 |
+
},
|
| 52652 |
+
{
|
| 52653 |
+
"epoch": 0.41783333333333333,
|
| 52654 |
+
"grad_norm": 0.1015625,
|
| 52655 |
+
"learning_rate": 0.0005165398641787196,
|
| 52656 |
+
"loss": 2.1381,
|
| 52657 |
+
"step": 7521
|
| 52658 |
+
},
|
| 52659 |
+
{
|
| 52660 |
+
"epoch": 0.41788888888888887,
|
| 52661 |
+
"grad_norm": 0.1005859375,
|
| 52662 |
+
"learning_rate": 0.000516471714943705,
|
| 52663 |
+
"loss": 2.1213,
|
| 52664 |
+
"step": 7522
|
| 52665 |
+
},
|
| 52666 |
+
{
|
| 52667 |
+
"epoch": 0.41794444444444445,
|
| 52668 |
+
"grad_norm": 0.10009765625,
|
| 52669 |
+
"learning_rate": 0.0005164035620144683,
|
| 52670 |
+
"loss": 2.1077,
|
| 52671 |
+
"step": 7523
|
| 52672 |
+
},
|
| 52673 |
+
{
|
| 52674 |
+
"epoch": 0.418,
|
| 52675 |
+
"grad_norm": 0.103515625,
|
| 52676 |
+
"learning_rate": 0.0005163354053931703,
|
| 52677 |
+
"loss": 2.1451,
|
| 52678 |
+
"step": 7524
|
| 52679 |
+
},
|
| 52680 |
+
{
|
| 52681 |
+
"epoch": 0.41805555555555557,
|
| 52682 |
+
"grad_norm": 0.099609375,
|
| 52683 |
+
"learning_rate": 0.0005162672450819735,
|
| 52684 |
+
"loss": 2.024,
|
| 52685 |
+
"step": 7525
|
| 52686 |
+
},
|
| 52687 |
+
{
|
| 52688 |
+
"epoch": 0.4181111111111111,
|
| 52689 |
+
"grad_norm": 0.10107421875,
|
| 52690 |
+
"learning_rate": 0.0005161990810830393,
|
| 52691 |
+
"loss": 2.0873,
|
| 52692 |
+
"step": 7526
|
| 52693 |
+
},
|
| 52694 |
+
{
|
| 52695 |
+
"epoch": 0.4181666666666667,
|
| 52696 |
+
"grad_norm": 0.1015625,
|
| 52697 |
+
"learning_rate": 0.00051613091339853,
|
| 52698 |
+
"loss": 2.0956,
|
| 52699 |
+
"step": 7527
|
| 52700 |
+
},
|
| 52701 |
+
{
|
| 52702 |
+
"epoch": 0.4182222222222222,
|
| 52703 |
+
"grad_norm": 0.0966796875,
|
| 52704 |
+
"learning_rate": 0.0005160627420306076,
|
| 52705 |
+
"loss": 2.086,
|
| 52706 |
+
"step": 7528
|
| 52707 |
+
},
|
| 52708 |
+
{
|
| 52709 |
+
"epoch": 0.4182777777777778,
|
| 52710 |
+
"grad_norm": 0.09765625,
|
| 52711 |
+
"learning_rate": 0.0005159945669814345,
|
| 52712 |
+
"loss": 2.0757,
|
| 52713 |
+
"step": 7529
|
| 52714 |
+
},
|
| 52715 |
+
{
|
| 52716 |
+
"epoch": 0.41833333333333333,
|
| 52717 |
+
"grad_norm": 0.09814453125,
|
| 52718 |
+
"learning_rate": 0.0005159263882531729,
|
| 52719 |
+
"loss": 2.0952,
|
| 52720 |
+
"step": 7530
|
| 52721 |
+
},
|
| 52722 |
+
{
|
| 52723 |
+
"epoch": 0.41838888888888887,
|
| 52724 |
+
"grad_norm": 0.10400390625,
|
| 52725 |
+
"learning_rate": 0.0005158582058479851,
|
| 52726 |
+
"loss": 2.1616,
|
| 52727 |
+
"step": 7531
|
| 52728 |
+
},
|
| 52729 |
+
{
|
| 52730 |
+
"epoch": 0.41844444444444445,
|
| 52731 |
+
"grad_norm": 0.0986328125,
|
| 52732 |
+
"learning_rate": 0.0005157900197680341,
|
| 52733 |
+
"loss": 2.0258,
|
| 52734 |
+
"step": 7532
|
| 52735 |
+
},
|
| 52736 |
+
{
|
| 52737 |
+
"epoch": 0.4185,
|
| 52738 |
+
"grad_norm": 0.11083984375,
|
| 52739 |
+
"learning_rate": 0.0005157218300154825,
|
| 52740 |
+
"loss": 2.0548,
|
| 52741 |
+
"step": 7533
|
| 52742 |
+
},
|
| 52743 |
+
{
|
| 52744 |
+
"epoch": 0.41855555555555557,
|
| 52745 |
+
"grad_norm": 0.10107421875,
|
| 52746 |
+
"learning_rate": 0.0005156536365924928,
|
| 52747 |
+
"loss": 2.1145,
|
| 52748 |
+
"step": 7534
|
| 52749 |
+
},
|
| 52750 |
+
{
|
| 52751 |
+
"epoch": 0.4186111111111111,
|
| 52752 |
+
"grad_norm": 0.1025390625,
|
| 52753 |
+
"learning_rate": 0.0005155854395012282,
|
| 52754 |
+
"loss": 2.1498,
|
| 52755 |
+
"step": 7535
|
| 52756 |
+
},
|
| 52757 |
+
{
|
| 52758 |
+
"epoch": 0.4186666666666667,
|
| 52759 |
+
"grad_norm": 0.1044921875,
|
| 52760 |
+
"learning_rate": 0.000515517238743852,
|
| 52761 |
+
"loss": 2.1397,
|
| 52762 |
+
"step": 7536
|
| 52763 |
+
},
|
| 52764 |
+
{
|
| 52765 |
+
"epoch": 0.4187222222222222,
|
| 52766 |
+
"grad_norm": 0.10205078125,
|
| 52767 |
+
"learning_rate": 0.000515449034322527,
|
| 52768 |
+
"loss": 2.0906,
|
| 52769 |
+
"step": 7537
|
| 52770 |
+
},
|
| 52771 |
+
{
|
| 52772 |
+
"epoch": 0.4187777777777778,
|
| 52773 |
+
"grad_norm": 0.1025390625,
|
| 52774 |
+
"learning_rate": 0.0005153808262394165,
|
| 52775 |
+
"loss": 2.0436,
|
| 52776 |
+
"step": 7538
|
| 52777 |
+
},
|
| 52778 |
+
{
|
| 52779 |
+
"epoch": 0.41883333333333334,
|
| 52780 |
+
"grad_norm": 0.10205078125,
|
| 52781 |
+
"learning_rate": 0.0005153126144966842,
|
| 52782 |
+
"loss": 2.0026,
|
| 52783 |
+
"step": 7539
|
| 52784 |
+
},
|
| 52785 |
+
{
|
| 52786 |
+
"epoch": 0.41888888888888887,
|
| 52787 |
+
"grad_norm": 0.109375,
|
| 52788 |
+
"learning_rate": 0.0005152443990964933,
|
| 52789 |
+
"loss": 2.098,
|
| 52790 |
+
"step": 7540
|
| 52791 |
+
},
|
| 52792 |
+
{
|
| 52793 |
+
"epoch": 0.41894444444444445,
|
| 52794 |
+
"grad_norm": 0.1044921875,
|
| 52795 |
+
"learning_rate": 0.0005151761800410077,
|
| 52796 |
+
"loss": 2.1142,
|
| 52797 |
+
"step": 7541
|
| 52798 |
+
},
|
| 52799 |
+
{
|
| 52800 |
+
"epoch": 0.419,
|
| 52801 |
+
"grad_norm": 0.1015625,
|
| 52802 |
+
"learning_rate": 0.0005151079573323909,
|
| 52803 |
+
"loss": 2.1134,
|
| 52804 |
+
"step": 7542
|
| 52805 |
+
},
|
| 52806 |
+
{
|
| 52807 |
+
"epoch": 0.41905555555555557,
|
| 52808 |
+
"grad_norm": 0.09912109375,
|
| 52809 |
+
"learning_rate": 0.0005150397309728069,
|
| 52810 |
+
"loss": 2.0982,
|
| 52811 |
+
"step": 7543
|
| 52812 |
+
},
|
| 52813 |
+
{
|
| 52814 |
+
"epoch": 0.4191111111111111,
|
| 52815 |
+
"grad_norm": 0.10302734375,
|
| 52816 |
+
"learning_rate": 0.0005149715009644199,
|
| 52817 |
+
"loss": 2.0963,
|
| 52818 |
+
"step": 7544
|
| 52819 |
+
},
|
| 52820 |
+
{
|
| 52821 |
+
"epoch": 0.4191666666666667,
|
| 52822 |
+
"grad_norm": 0.09814453125,
|
| 52823 |
+
"learning_rate": 0.0005149032673093936,
|
| 52824 |
+
"loss": 2.1023,
|
| 52825 |
+
"step": 7545
|
| 52826 |
+
},
|
| 52827 |
+
{
|
| 52828 |
+
"epoch": 0.4192222222222222,
|
| 52829 |
+
"grad_norm": 0.10302734375,
|
| 52830 |
+
"learning_rate": 0.0005148350300098926,
|
| 52831 |
+
"loss": 2.0013,
|
| 52832 |
+
"step": 7546
|
| 52833 |
+
},
|
| 52834 |
+
{
|
| 52835 |
+
"epoch": 0.4192777777777778,
|
| 52836 |
+
"grad_norm": 0.10107421875,
|
| 52837 |
+
"learning_rate": 0.000514766789068081,
|
| 52838 |
+
"loss": 2.0681,
|
| 52839 |
+
"step": 7547
|
| 52840 |
+
},
|
| 52841 |
+
{
|
| 52842 |
+
"epoch": 0.41933333333333334,
|
| 52843 |
+
"grad_norm": 0.10205078125,
|
| 52844 |
+
"learning_rate": 0.0005146985444861231,
|
| 52845 |
+
"loss": 2.0836,
|
| 52846 |
+
"step": 7548
|
| 52847 |
+
},
|
| 52848 |
+
{
|
| 52849 |
+
"epoch": 0.41938888888888887,
|
| 52850 |
+
"grad_norm": 0.0986328125,
|
| 52851 |
+
"learning_rate": 0.0005146302962661839,
|
| 52852 |
+
"loss": 2.034,
|
| 52853 |
+
"step": 7549
|
| 52854 |
+
},
|
| 52855 |
+
{
|
| 52856 |
+
"epoch": 0.41944444444444445,
|
| 52857 |
+
"grad_norm": 0.10107421875,
|
| 52858 |
+
"learning_rate": 0.0005145620444104278,
|
| 52859 |
+
"loss": 2.0919,
|
| 52860 |
+
"step": 7550
|
| 52861 |
+
},
|
| 52862 |
+
{
|
| 52863 |
+
"epoch": 0.4195,
|
| 52864 |
+
"grad_norm": 0.09814453125,
|
| 52865 |
+
"learning_rate": 0.0005144937889210196,
|
| 52866 |
+
"loss": 2.0758,
|
| 52867 |
+
"step": 7551
|
| 52868 |
+
},
|
| 52869 |
+
{
|
| 52870 |
+
"epoch": 0.41955555555555557,
|
| 52871 |
+
"grad_norm": 0.10107421875,
|
| 52872 |
+
"learning_rate": 0.0005144255298001244,
|
| 52873 |
+
"loss": 2.1298,
|
| 52874 |
+
"step": 7552
|
| 52875 |
+
},
|
| 52876 |
+
{
|
| 52877 |
+
"epoch": 0.4196111111111111,
|
| 52878 |
+
"grad_norm": 0.095703125,
|
| 52879 |
+
"learning_rate": 0.000514357267049907,
|
| 52880 |
+
"loss": 1.9709,
|
| 52881 |
+
"step": 7553
|
| 52882 |
+
},
|
| 52883 |
+
{
|
| 52884 |
+
"epoch": 0.4196666666666667,
|
| 52885 |
+
"grad_norm": 0.10400390625,
|
| 52886 |
+
"learning_rate": 0.0005142890006725327,
|
| 52887 |
+
"loss": 2.049,
|
| 52888 |
+
"step": 7554
|
| 52889 |
+
},
|
| 52890 |
+
{
|
| 52891 |
+
"epoch": 0.4197222222222222,
|
| 52892 |
+
"grad_norm": 0.099609375,
|
| 52893 |
+
"learning_rate": 0.0005142207306701666,
|
| 52894 |
+
"loss": 2.0663,
|
| 52895 |
+
"step": 7555
|
| 52896 |
+
},
|
| 52897 |
+
{
|
| 52898 |
+
"epoch": 0.4197777777777778,
|
| 52899 |
+
"grad_norm": 0.0986328125,
|
| 52900 |
+
"learning_rate": 0.0005141524570449742,
|
| 52901 |
+
"loss": 2.1148,
|
| 52902 |
+
"step": 7556
|
| 52903 |
+
},
|
| 52904 |
+
{
|
| 52905 |
+
"epoch": 0.41983333333333334,
|
| 52906 |
+
"grad_norm": 0.10107421875,
|
| 52907 |
+
"learning_rate": 0.000514084179799121,
|
| 52908 |
+
"loss": 2.1108,
|
| 52909 |
+
"step": 7557
|
| 52910 |
+
},
|
| 52911 |
+
{
|
| 52912 |
+
"epoch": 0.41988888888888887,
|
| 52913 |
+
"grad_norm": 0.1015625,
|
| 52914 |
+
"learning_rate": 0.0005140158989347724,
|
| 52915 |
+
"loss": 2.093,
|
| 52916 |
+
"step": 7558
|
| 52917 |
+
},
|
| 52918 |
+
{
|
| 52919 |
+
"epoch": 0.41994444444444445,
|
| 52920 |
+
"grad_norm": 0.0986328125,
|
| 52921 |
+
"learning_rate": 0.0005139476144540943,
|
| 52922 |
+
"loss": 2.068,
|
| 52923 |
+
"step": 7559
|
| 52924 |
+
},
|
| 52925 |
+
{
|
| 52926 |
+
"epoch": 0.42,
|
| 52927 |
+
"grad_norm": 0.1025390625,
|
| 52928 |
+
"learning_rate": 0.0005138793263592525,
|
| 52929 |
+
"loss": 2.0416,
|
| 52930 |
+
"step": 7560
|
| 52931 |
+
},
|
| 52932 |
+
{
|
| 52933 |
+
"epoch": 0.42005555555555557,
|
| 52934 |
+
"grad_norm": 0.09814453125,
|
| 52935 |
+
"learning_rate": 0.000513811034652413,
|
| 52936 |
+
"loss": 2.0865,
|
| 52937 |
+
"step": 7561
|
| 52938 |
+
},
|
| 52939 |
+
{
|
| 52940 |
+
"epoch": 0.4201111111111111,
|
| 52941 |
+
"grad_norm": 0.09716796875,
|
| 52942 |
+
"learning_rate": 0.0005137427393357419,
|
| 52943 |
+
"loss": 2.1221,
|
| 52944 |
+
"step": 7562
|
| 52945 |
+
},
|
| 52946 |
+
{
|
| 52947 |
+
"epoch": 0.4201666666666667,
|
| 52948 |
+
"grad_norm": 0.10205078125,
|
| 52949 |
+
"learning_rate": 0.000513674440411405,
|
| 52950 |
+
"loss": 2.129,
|
| 52951 |
+
"step": 7563
|
| 52952 |
+
},
|
| 52953 |
+
{
|
| 52954 |
+
"epoch": 0.4202222222222222,
|
| 52955 |
+
"grad_norm": 0.095703125,
|
| 52956 |
+
"learning_rate": 0.0005136061378815691,
|
| 52957 |
+
"loss": 2.0759,
|
| 52958 |
+
"step": 7564
|
| 52959 |
+
},
|
| 52960 |
+
{
|
| 52961 |
+
"epoch": 0.42027777777777775,
|
| 52962 |
+
"grad_norm": 0.103515625,
|
| 52963 |
+
"learning_rate": 0.0005135378317484003,
|
| 52964 |
+
"loss": 2.1162,
|
| 52965 |
+
"step": 7565
|
| 52966 |
+
},
|
| 52967 |
+
{
|
| 52968 |
+
"epoch": 0.42033333333333334,
|
| 52969 |
+
"grad_norm": 0.09619140625,
|
| 52970 |
+
"learning_rate": 0.0005134695220140651,
|
| 52971 |
+
"loss": 1.9682,
|
| 52972 |
+
"step": 7566
|
| 52973 |
+
},
|
| 52974 |
+
{
|
| 52975 |
+
"epoch": 0.42038888888888887,
|
| 52976 |
+
"grad_norm": 0.1044921875,
|
| 52977 |
+
"learning_rate": 0.0005134012086807305,
|
| 52978 |
+
"loss": 2.0835,
|
| 52979 |
+
"step": 7567
|
| 52980 |
+
},
|
| 52981 |
+
{
|
| 52982 |
+
"epoch": 0.42044444444444445,
|
| 52983 |
+
"grad_norm": 0.1005859375,
|
| 52984 |
+
"learning_rate": 0.0005133328917505626,
|
| 52985 |
+
"loss": 2.0865,
|
| 52986 |
+
"step": 7568
|
| 52987 |
+
},
|
| 52988 |
+
{
|
| 52989 |
+
"epoch": 0.4205,
|
| 52990 |
+
"grad_norm": 0.1015625,
|
| 52991 |
+
"learning_rate": 0.0005132645712257289,
|
| 52992 |
+
"loss": 2.1249,
|
| 52993 |
+
"step": 7569
|
| 52994 |
+
},
|
| 52995 |
+
{
|
| 52996 |
+
"epoch": 0.42055555555555557,
|
| 52997 |
+
"grad_norm": 0.0966796875,
|
| 52998 |
+
"learning_rate": 0.0005131962471083959,
|
| 52999 |
+
"loss": 2.1114,
|
| 53000 |
+
"step": 7570
|
| 53001 |
+
},
|
| 53002 |
+
{
|
| 53003 |
+
"epoch": 0.4206111111111111,
|
| 53004 |
+
"grad_norm": 0.10205078125,
|
| 53005 |
+
"learning_rate": 0.000513127919400731,
|
| 53006 |
+
"loss": 2.1129,
|
| 53007 |
+
"step": 7571
|
| 53008 |
+
},
|
| 53009 |
+
{
|
| 53010 |
+
"epoch": 0.4206666666666667,
|
| 53011 |
+
"grad_norm": 0.09765625,
|
| 53012 |
+
"learning_rate": 0.0005130595881049013,
|
| 53013 |
+
"loss": 1.9645,
|
| 53014 |
+
"step": 7572
|
| 53015 |
+
},
|
| 53016 |
+
{
|
| 53017 |
+
"epoch": 0.4207222222222222,
|
| 53018 |
+
"grad_norm": 0.09912109375,
|
| 53019 |
+
"learning_rate": 0.0005129912532230741,
|
| 53020 |
+
"loss": 2.0422,
|
| 53021 |
+
"step": 7573
|
| 53022 |
+
},
|
| 53023 |
+
{
|
| 53024 |
+
"epoch": 0.42077777777777775,
|
| 53025 |
+
"grad_norm": 0.0966796875,
|
| 53026 |
+
"learning_rate": 0.0005129229147574169,
|
| 53027 |
+
"loss": 2.0715,
|
| 53028 |
+
"step": 7574
|
| 53029 |
+
},
|
| 53030 |
+
{
|
| 53031 |
+
"epoch": 0.42083333333333334,
|
| 53032 |
+
"grad_norm": 0.09716796875,
|
| 53033 |
+
"learning_rate": 0.000512854572710097,
|
| 53034 |
+
"loss": 2.1004,
|
| 53035 |
+
"step": 7575
|
| 53036 |
+
},
|
| 53037 |
+
{
|
| 53038 |
+
"epoch": 0.42088888888888887,
|
| 53039 |
+
"grad_norm": 0.10107421875,
|
| 53040 |
+
"learning_rate": 0.0005127862270832823,
|
| 53041 |
+
"loss": 2.0762,
|
| 53042 |
+
"step": 7576
|
| 53043 |
+
},
|
| 53044 |
+
{
|
| 53045 |
+
"epoch": 0.42094444444444445,
|
| 53046 |
+
"grad_norm": 0.09716796875,
|
| 53047 |
+
"learning_rate": 0.0005127178778791406,
|
| 53048 |
+
"loss": 2.0759,
|
| 53049 |
+
"step": 7577
|
| 53050 |
+
},
|
| 53051 |
+
{
|
| 53052 |
+
"epoch": 0.421,
|
| 53053 |
+
"grad_norm": 0.09765625,
|
| 53054 |
+
"learning_rate": 0.0005126495250998396,
|
| 53055 |
+
"loss": 2.0664,
|
| 53056 |
+
"step": 7578
|
| 53057 |
+
},
|
| 53058 |
+
{
|
| 53059 |
+
"epoch": 0.42105555555555557,
|
| 53060 |
+
"grad_norm": 0.09912109375,
|
| 53061 |
+
"learning_rate": 0.0005125811687475474,
|
| 53062 |
+
"loss": 1.9866,
|
| 53063 |
+
"step": 7579
|
| 53064 |
+
},
|
| 53065 |
+
{
|
| 53066 |
+
"epoch": 0.4211111111111111,
|
| 53067 |
+
"grad_norm": 0.10302734375,
|
| 53068 |
+
"learning_rate": 0.0005125128088244321,
|
| 53069 |
+
"loss": 2.1185,
|
| 53070 |
+
"step": 7580
|
| 53071 |
+
},
|
| 53072 |
+
{
|
| 53073 |
+
"epoch": 0.4211666666666667,
|
| 53074 |
+
"grad_norm": 0.0986328125,
|
| 53075 |
+
"learning_rate": 0.000512444445332662,
|
| 53076 |
+
"loss": 2.0592,
|
| 53077 |
+
"step": 7581
|
| 53078 |
+
},
|
| 53079 |
+
{
|
| 53080 |
+
"epoch": 0.4212222222222222,
|
| 53081 |
+
"grad_norm": 0.1015625,
|
| 53082 |
+
"learning_rate": 0.0005123760782744053,
|
| 53083 |
+
"loss": 2.0888,
|
| 53084 |
+
"step": 7582
|
| 53085 |
+
},
|
| 53086 |
+
{
|
| 53087 |
+
"epoch": 0.42127777777777775,
|
| 53088 |
+
"grad_norm": 0.09521484375,
|
| 53089 |
+
"learning_rate": 0.0005123077076518306,
|
| 53090 |
+
"loss": 2.102,
|
| 53091 |
+
"step": 7583
|
| 53092 |
+
},
|
| 53093 |
+
{
|
| 53094 |
+
"epoch": 0.42133333333333334,
|
| 53095 |
+
"grad_norm": 0.09619140625,
|
| 53096 |
+
"learning_rate": 0.0005122393334671063,
|
| 53097 |
+
"loss": 2.0184,
|
| 53098 |
+
"step": 7584
|
| 53099 |
+
},
|
| 53100 |
+
{
|
| 53101 |
+
"epoch": 0.42138888888888887,
|
| 53102 |
+
"grad_norm": 0.103515625,
|
| 53103 |
+
"learning_rate": 0.0005121709557224011,
|
| 53104 |
+
"loss": 2.049,
|
| 53105 |
+
"step": 7585
|
| 53106 |
+
},
|
| 53107 |
+
{
|
| 53108 |
+
"epoch": 0.42144444444444445,
|
| 53109 |
+
"grad_norm": 0.0986328125,
|
| 53110 |
+
"learning_rate": 0.0005121025744198839,
|
| 53111 |
+
"loss": 2.0432,
|
| 53112 |
+
"step": 7586
|
| 53113 |
+
},
|
| 53114 |
+
{
|
| 53115 |
+
"epoch": 0.4215,
|
| 53116 |
+
"grad_norm": 0.1015625,
|
| 53117 |
+
"learning_rate": 0.0005120341895617236,
|
| 53118 |
+
"loss": 2.1472,
|
| 53119 |
+
"step": 7587
|
| 53120 |
+
},
|
| 53121 |
+
{
|
| 53122 |
+
"epoch": 0.42155555555555557,
|
| 53123 |
+
"grad_norm": 0.10302734375,
|
| 53124 |
+
"learning_rate": 0.000511965801150089,
|
| 53125 |
+
"loss": 2.0686,
|
| 53126 |
+
"step": 7588
|
| 53127 |
+
},
|
| 53128 |
+
{
|
| 53129 |
+
"epoch": 0.4216111111111111,
|
| 53130 |
+
"grad_norm": 0.09814453125,
|
| 53131 |
+
"learning_rate": 0.0005118974091871495,
|
| 53132 |
+
"loss": 2.1102,
|
| 53133 |
+
"step": 7589
|
| 53134 |
+
},
|
| 53135 |
+
{
|
| 53136 |
+
"epoch": 0.4216666666666667,
|
| 53137 |
+
"grad_norm": 0.1005859375,
|
| 53138 |
+
"learning_rate": 0.0005118290136750743,
|
| 53139 |
+
"loss": 2.0471,
|
| 53140 |
+
"step": 7590
|
| 53141 |
+
},
|
| 53142 |
+
{
|
| 53143 |
+
"epoch": 0.4217222222222222,
|
| 53144 |
+
"grad_norm": 0.10400390625,
|
| 53145 |
+
"learning_rate": 0.0005117606146160325,
|
| 53146 |
+
"loss": 2.0421,
|
| 53147 |
+
"step": 7591
|
| 53148 |
+
},
|
| 53149 |
+
{
|
| 53150 |
+
"epoch": 0.42177777777777775,
|
| 53151 |
+
"grad_norm": 0.10205078125,
|
| 53152 |
+
"learning_rate": 0.0005116922120121939,
|
| 53153 |
+
"loss": 2.049,
|
| 53154 |
+
"step": 7592
|
| 53155 |
+
},
|
| 53156 |
+
{
|
| 53157 |
+
"epoch": 0.42183333333333334,
|
| 53158 |
+
"grad_norm": 0.095703125,
|
| 53159 |
+
"learning_rate": 0.0005116238058657278,
|
| 53160 |
+
"loss": 2.066,
|
| 53161 |
+
"step": 7593
|
| 53162 |
+
},
|
| 53163 |
+
{
|
| 53164 |
+
"epoch": 0.42188888888888887,
|
| 53165 |
+
"grad_norm": 0.1005859375,
|
| 53166 |
+
"learning_rate": 0.0005115553961788041,
|
| 53167 |
+
"loss": 2.0956,
|
| 53168 |
+
"step": 7594
|
| 53169 |
+
},
|
| 53170 |
+
{
|
| 53171 |
+
"epoch": 0.42194444444444446,
|
| 53172 |
+
"grad_norm": 0.09814453125,
|
| 53173 |
+
"learning_rate": 0.0005114869829535925,
|
| 53174 |
+
"loss": 2.0665,
|
| 53175 |
+
"step": 7595
|
| 53176 |
+
},
|
| 53177 |
+
{
|
| 53178 |
+
"epoch": 0.422,
|
| 53179 |
+
"grad_norm": 0.095703125,
|
| 53180 |
+
"learning_rate": 0.0005114185661922628,
|
| 53181 |
+
"loss": 1.9808,
|
| 53182 |
+
"step": 7596
|
| 53183 |
+
},
|
| 53184 |
+
{
|
| 53185 |
+
"epoch": 0.4220555555555556,
|
| 53186 |
+
"grad_norm": 0.09716796875,
|
| 53187 |
+
"learning_rate": 0.0005113501458969854,
|
| 53188 |
+
"loss": 2.0726,
|
| 53189 |
+
"step": 7597
|
| 53190 |
+
},
|
| 53191 |
+
{
|
| 53192 |
+
"epoch": 0.4221111111111111,
|
| 53193 |
+
"grad_norm": 0.09521484375,
|
| 53194 |
+
"learning_rate": 0.0005112817220699299,
|
| 53195 |
+
"loss": 2.112,
|
| 53196 |
+
"step": 7598
|
| 53197 |
+
},
|
| 53198 |
+
{
|
| 53199 |
+
"epoch": 0.4221666666666667,
|
| 53200 |
+
"grad_norm": 0.09521484375,
|
| 53201 |
+
"learning_rate": 0.0005112132947132671,
|
| 53202 |
+
"loss": 2.0852,
|
| 53203 |
+
"step": 7599
|
| 53204 |
+
},
|
| 53205 |
+
{
|
| 53206 |
+
"epoch": 0.4222222222222222,
|
| 53207 |
+
"grad_norm": 0.09521484375,
|
| 53208 |
+
"learning_rate": 0.000511144863829167,
|
| 53209 |
+
"loss": 2.0538,
|
| 53210 |
+
"step": 7600
|
| 53211 |
+
},
|
| 53212 |
+
{
|
| 53213 |
+
"epoch": 0.42227777777777775,
|
| 53214 |
+
"grad_norm": 0.0927734375,
|
| 53215 |
+
"learning_rate": 0.0005110764294198001,
|
| 53216 |
+
"loss": 2.0285,
|
| 53217 |
+
"step": 7601
|
| 53218 |
+
},
|
| 53219 |
+
{
|
| 53220 |
+
"epoch": 0.42233333333333334,
|
| 53221 |
+
"grad_norm": 0.09228515625,
|
| 53222 |
+
"learning_rate": 0.0005110079914873372,
|
| 53223 |
+
"loss": 1.9614,
|
| 53224 |
+
"step": 7602
|
| 53225 |
+
},
|
| 53226 |
+
{
|
| 53227 |
+
"epoch": 0.42238888888888887,
|
| 53228 |
+
"grad_norm": 0.09619140625,
|
| 53229 |
+
"learning_rate": 0.0005109395500339488,
|
| 53230 |
+
"loss": 2.0708,
|
| 53231 |
+
"step": 7603
|
| 53232 |
+
},
|
| 53233 |
+
{
|
| 53234 |
+
"epoch": 0.42244444444444446,
|
| 53235 |
+
"grad_norm": 0.09423828125,
|
| 53236 |
+
"learning_rate": 0.0005108711050618059,
|
| 53237 |
+
"loss": 2.0298,
|
| 53238 |
+
"step": 7604
|
| 53239 |
+
},
|
| 53240 |
+
{
|
| 53241 |
+
"epoch": 0.4225,
|
| 53242 |
+
"grad_norm": 0.0947265625,
|
| 53243 |
+
"learning_rate": 0.0005108026565730792,
|
| 53244 |
+
"loss": 2.1128,
|
| 53245 |
+
"step": 7605
|
| 53246 |
+
},
|
| 53247 |
+
{
|
| 53248 |
+
"epoch": 0.4225555555555556,
|
| 53249 |
+
"grad_norm": 0.09765625,
|
| 53250 |
+
"learning_rate": 0.0005107342045699398,
|
| 53251 |
+
"loss": 2.0763,
|
| 53252 |
+
"step": 7606
|
| 53253 |
+
},
|
| 53254 |
+
{
|
| 53255 |
+
"epoch": 0.4226111111111111,
|
| 53256 |
+
"grad_norm": 0.09375,
|
| 53257 |
+
"learning_rate": 0.0005106657490545588,
|
| 53258 |
+
"loss": 1.9621,
|
| 53259 |
+
"step": 7607
|
| 53260 |
+
},
|
| 53261 |
+
{
|
| 53262 |
+
"epoch": 0.4226666666666667,
|
| 53263 |
+
"grad_norm": 0.09716796875,
|
| 53264 |
+
"learning_rate": 0.0005105972900291077,
|
| 53265 |
+
"loss": 2.0498,
|
| 53266 |
+
"step": 7608
|
| 53267 |
+
},
|
| 53268 |
+
{
|
| 53269 |
+
"epoch": 0.4227222222222222,
|
| 53270 |
+
"grad_norm": 0.0947265625,
|
| 53271 |
+
"learning_rate": 0.0005105288274957577,
|
| 53272 |
+
"loss": 2.0915,
|
| 53273 |
+
"step": 7609
|
| 53274 |
+
},
|
| 53275 |
+
{
|
| 53276 |
+
"epoch": 0.42277777777777775,
|
| 53277 |
+
"grad_norm": 0.09716796875,
|
| 53278 |
+
"learning_rate": 0.0005104603614566802,
|
| 53279 |
+
"loss": 2.1188,
|
| 53280 |
+
"step": 7610
|
| 53281 |
+
},
|
| 53282 |
+
{
|
| 53283 |
+
"epoch": 0.42283333333333334,
|
| 53284 |
+
"grad_norm": 0.09423828125,
|
| 53285 |
+
"learning_rate": 0.0005103918919140469,
|
| 53286 |
+
"loss": 2.0761,
|
| 53287 |
+
"step": 7611
|
| 53288 |
+
},
|
| 53289 |
+
{
|
| 53290 |
+
"epoch": 0.42288888888888887,
|
| 53291 |
+
"grad_norm": 0.0927734375,
|
| 53292 |
+
"learning_rate": 0.0005103234188700295,
|
| 53293 |
+
"loss": 1.9859,
|
| 53294 |
+
"step": 7612
|
| 53295 |
+
},
|
| 53296 |
+
{
|
| 53297 |
+
"epoch": 0.42294444444444446,
|
| 53298 |
+
"grad_norm": 0.10107421875,
|
| 53299 |
+
"learning_rate": 0.0005102549423267996,
|
| 53300 |
+
"loss": 2.0267,
|
| 53301 |
+
"step": 7613
|
| 53302 |
+
},
|
| 53303 |
+
{
|
| 53304 |
+
"epoch": 0.423,
|
| 53305 |
+
"grad_norm": 0.09716796875,
|
| 53306 |
+
"learning_rate": 0.0005101864622865295,
|
| 53307 |
+
"loss": 2.1066,
|
| 53308 |
+
"step": 7614
|
| 53309 |
+
},
|
| 53310 |
+
{
|
| 53311 |
+
"epoch": 0.4230555555555556,
|
| 53312 |
+
"grad_norm": 0.09716796875,
|
| 53313 |
+
"learning_rate": 0.0005101179787513911,
|
| 53314 |
+
"loss": 2.0789,
|
| 53315 |
+
"step": 7615
|
| 53316 |
+
},
|
| 53317 |
+
{
|
| 53318 |
+
"epoch": 0.4231111111111111,
|
| 53319 |
+
"grad_norm": 0.09912109375,
|
| 53320 |
+
"learning_rate": 0.0005100494917235562,
|
| 53321 |
+
"loss": 2.1107,
|
| 53322 |
+
"step": 7616
|
| 53323 |
+
},
|
| 53324 |
+
{
|
| 53325 |
+
"epoch": 0.4231666666666667,
|
| 53326 |
+
"grad_norm": 0.09375,
|
| 53327 |
+
"learning_rate": 0.0005099810012051977,
|
| 53328 |
+
"loss": 2.0383,
|
| 53329 |
+
"step": 7617
|
| 53330 |
+
},
|
| 53331 |
+
{
|
| 53332 |
+
"epoch": 0.4232222222222222,
|
| 53333 |
+
"grad_norm": 0.10693359375,
|
| 53334 |
+
"learning_rate": 0.0005099125071984874,
|
| 53335 |
+
"loss": 2.0663,
|
| 53336 |
+
"step": 7618
|
| 53337 |
+
},
|
| 53338 |
+
{
|
| 53339 |
+
"epoch": 0.42327777777777775,
|
| 53340 |
+
"grad_norm": 0.09619140625,
|
| 53341 |
+
"learning_rate": 0.0005098440097055978,
|
| 53342 |
+
"loss": 2.0692,
|
| 53343 |
+
"step": 7619
|
| 53344 |
+
},
|
| 53345 |
+
{
|
| 53346 |
+
"epoch": 0.42333333333333334,
|
| 53347 |
+
"grad_norm": 0.09765625,
|
| 53348 |
+
"learning_rate": 0.000509775508728702,
|
| 53349 |
+
"loss": 2.0812,
|
| 53350 |
+
"step": 7620
|
| 53351 |
+
},
|
| 53352 |
+
{
|
| 53353 |
+
"epoch": 0.42338888888888887,
|
| 53354 |
+
"grad_norm": 0.09521484375,
|
| 53355 |
+
"learning_rate": 0.0005097070042699722,
|
| 53356 |
+
"loss": 2.0297,
|
| 53357 |
+
"step": 7621
|
| 53358 |
+
},
|
| 53359 |
+
{
|
| 53360 |
+
"epoch": 0.42344444444444446,
|
| 53361 |
+
"grad_norm": 0.1015625,
|
| 53362 |
+
"learning_rate": 0.0005096384963315815,
|
| 53363 |
+
"loss": 2.0654,
|
| 53364 |
+
"step": 7622
|
| 53365 |
+
},
|
| 53366 |
+
{
|
| 53367 |
+
"epoch": 0.4235,
|
| 53368 |
+
"grad_norm": 0.09765625,
|
| 53369 |
+
"learning_rate": 0.0005095699849157026,
|
| 53370 |
+
"loss": 2.1006,
|
| 53371 |
+
"step": 7623
|
| 53372 |
+
},
|
| 53373 |
+
{
|
| 53374 |
+
"epoch": 0.4235555555555556,
|
| 53375 |
+
"grad_norm": 0.09716796875,
|
| 53376 |
+
"learning_rate": 0.0005095014700245086,
|
| 53377 |
+
"loss": 2.0812,
|
| 53378 |
+
"step": 7624
|
| 53379 |
+
},
|
| 53380 |
+
{
|
| 53381 |
+
"epoch": 0.4236111111111111,
|
| 53382 |
+
"grad_norm": 0.0947265625,
|
| 53383 |
+
"learning_rate": 0.0005094329516601727,
|
| 53384 |
+
"loss": 2.0113,
|
| 53385 |
+
"step": 7625
|
| 53386 |
+
},
|
| 53387 |
+
{
|
| 53388 |
+
"epoch": 0.4236666666666667,
|
| 53389 |
+
"grad_norm": 0.10107421875,
|
| 53390 |
+
"learning_rate": 0.000509364429824868,
|
| 53391 |
+
"loss": 2.0295,
|
| 53392 |
+
"step": 7626
|
| 53393 |
+
},
|
| 53394 |
+
{
|
| 53395 |
+
"epoch": 0.4237222222222222,
|
| 53396 |
+
"grad_norm": 0.0947265625,
|
| 53397 |
+
"learning_rate": 0.0005092959045207681,
|
| 53398 |
+
"loss": 2.045,
|
| 53399 |
+
"step": 7627
|
| 53400 |
+
},
|
| 53401 |
+
{
|
| 53402 |
+
"epoch": 0.42377777777777775,
|
| 53403 |
+
"grad_norm": 0.09619140625,
|
| 53404 |
+
"learning_rate": 0.0005092273757500463,
|
| 53405 |
+
"loss": 2.0756,
|
| 53406 |
+
"step": 7628
|
| 53407 |
+
},
|
| 53408 |
+
{
|
| 53409 |
+
"epoch": 0.42383333333333334,
|
| 53410 |
+
"grad_norm": 0.0947265625,
|
| 53411 |
+
"learning_rate": 0.0005091588435148761,
|
| 53412 |
+
"loss": 2.0314,
|
| 53413 |
+
"step": 7629
|
| 53414 |
+
},
|
| 53415 |
+
{
|
| 53416 |
+
"epoch": 0.42388888888888887,
|
| 53417 |
+
"grad_norm": 0.09375,
|
| 53418 |
+
"learning_rate": 0.0005090903078174315,
|
| 53419 |
+
"loss": 1.9947,
|
| 53420 |
+
"step": 7630
|
| 53421 |
+
},
|
| 53422 |
+
{
|
| 53423 |
+
"epoch": 0.42394444444444446,
|
| 53424 |
+
"grad_norm": 0.09521484375,
|
| 53425 |
+
"learning_rate": 0.0005090217686598861,
|
| 53426 |
+
"loss": 2.0239,
|
| 53427 |
+
"step": 7631
|
| 53428 |
+
},
|
| 53429 |
+
{
|
| 53430 |
+
"epoch": 0.424,
|
| 53431 |
+
"grad_norm": 0.09423828125,
|
| 53432 |
+
"learning_rate": 0.0005089532260444136,
|
| 53433 |
+
"loss": 2.1017,
|
| 53434 |
+
"step": 7632
|
| 53435 |
+
},
|
| 53436 |
+
{
|
| 53437 |
+
"epoch": 0.4240555555555556,
|
| 53438 |
+
"grad_norm": 0.09619140625,
|
| 53439 |
+
"learning_rate": 0.0005088846799731885,
|
| 53440 |
+
"loss": 2.042,
|
| 53441 |
+
"step": 7633
|
| 53442 |
+
},
|
| 53443 |
+
{
|
| 53444 |
+
"epoch": 0.4241111111111111,
|
| 53445 |
+
"grad_norm": 0.09765625,
|
| 53446 |
+
"learning_rate": 0.0005088161304483846,
|
| 53447 |
+
"loss": 2.047,
|
| 53448 |
+
"step": 7634
|
| 53449 |
+
},
|
| 53450 |
+
{
|
| 53451 |
+
"epoch": 0.4241666666666667,
|
| 53452 |
+
"grad_norm": 0.09716796875,
|
| 53453 |
+
"learning_rate": 0.0005087475774721761,
|
| 53454 |
+
"loss": 2.0053,
|
| 53455 |
+
"step": 7635
|
| 53456 |
+
},
|
| 53457 |
+
{
|
| 53458 |
+
"epoch": 0.4242222222222222,
|
| 53459 |
+
"grad_norm": 0.1005859375,
|
| 53460 |
+
"learning_rate": 0.0005086790210467375,
|
| 53461 |
+
"loss": 2.0699,
|
| 53462 |
+
"step": 7636
|
| 53463 |
+
},
|
| 53464 |
+
{
|
| 53465 |
+
"epoch": 0.42427777777777775,
|
| 53466 |
+
"grad_norm": 0.09423828125,
|
| 53467 |
+
"learning_rate": 0.0005086104611742432,
|
| 53468 |
+
"loss": 2.0531,
|
| 53469 |
+
"step": 7637
|
| 53470 |
+
},
|
| 53471 |
+
{
|
| 53472 |
+
"epoch": 0.42433333333333334,
|
| 53473 |
+
"grad_norm": 0.09423828125,
|
| 53474 |
+
"learning_rate": 0.0005085418978568679,
|
| 53475 |
+
"loss": 2.0823,
|
| 53476 |
+
"step": 7638
|
| 53477 |
+
},
|
| 53478 |
+
{
|
| 53479 |
+
"epoch": 0.42438888888888887,
|
| 53480 |
+
"grad_norm": 0.0966796875,
|
| 53481 |
+
"learning_rate": 0.0005084733310967859,
|
| 53482 |
+
"loss": 2.0269,
|
| 53483 |
+
"step": 7639
|
| 53484 |
+
},
|
| 53485 |
+
{
|
| 53486 |
+
"epoch": 0.42444444444444446,
|
| 53487 |
+
"grad_norm": 0.0986328125,
|
| 53488 |
+
"learning_rate": 0.0005084047608961724,
|
| 53489 |
+
"loss": 2.1033,
|
| 53490 |
+
"step": 7640
|
| 53491 |
+
},
|
| 53492 |
+
{
|
| 53493 |
+
"epoch": 0.4245,
|
| 53494 |
+
"grad_norm": 0.09521484375,
|
| 53495 |
+
"learning_rate": 0.0005083361872572021,
|
| 53496 |
+
"loss": 2.1325,
|
| 53497 |
+
"step": 7641
|
| 53498 |
+
},
|
| 53499 |
+
{
|
| 53500 |
+
"epoch": 0.4245555555555556,
|
| 53501 |
+
"grad_norm": 0.09912109375,
|
| 53502 |
+
"learning_rate": 0.00050826761018205,
|
| 53503 |
+
"loss": 2.0877,
|
| 53504 |
+
"step": 7642
|
| 53505 |
+
},
|
| 53506 |
+
{
|
| 53507 |
+
"epoch": 0.4246111111111111,
|
| 53508 |
+
"grad_norm": 0.09521484375,
|
| 53509 |
+
"learning_rate": 0.0005081990296728913,
|
| 53510 |
+
"loss": 2.1008,
|
| 53511 |
+
"step": 7643
|
| 53512 |
+
},
|
| 53513 |
+
{
|
| 53514 |
+
"epoch": 0.4246666666666667,
|
| 53515 |
+
"grad_norm": 0.09912109375,
|
| 53516 |
+
"learning_rate": 0.000508130445731901,
|
| 53517 |
+
"loss": 1.9853,
|
| 53518 |
+
"step": 7644
|
| 53519 |
+
},
|
| 53520 |
+
{
|
| 53521 |
+
"epoch": 0.4247222222222222,
|
| 53522 |
+
"grad_norm": 0.099609375,
|
| 53523 |
+
"learning_rate": 0.0005080618583612548,
|
| 53524 |
+
"loss": 2.0805,
|
| 53525 |
+
"step": 7645
|
| 53526 |
+
},
|
| 53527 |
+
{
|
| 53528 |
+
"epoch": 0.42477777777777775,
|
| 53529 |
+
"grad_norm": 0.09423828125,
|
| 53530 |
+
"learning_rate": 0.0005079932675631278,
|
| 53531 |
+
"loss": 2.0503,
|
| 53532 |
+
"step": 7646
|
| 53533 |
+
},
|
| 53534 |
+
{
|
| 53535 |
+
"epoch": 0.42483333333333334,
|
| 53536 |
+
"grad_norm": 0.09716796875,
|
| 53537 |
+
"learning_rate": 0.0005079246733396957,
|
| 53538 |
+
"loss": 2.0869,
|
| 53539 |
+
"step": 7647
|
| 53540 |
+
},
|
| 53541 |
+
{
|
| 53542 |
+
"epoch": 0.42488888888888887,
|
| 53543 |
+
"grad_norm": 0.09521484375,
|
| 53544 |
+
"learning_rate": 0.0005078560756931343,
|
| 53545 |
+
"loss": 2.0785,
|
| 53546 |
+
"step": 7648
|
| 53547 |
+
},
|
| 53548 |
+
{
|
| 53549 |
+
"epoch": 0.42494444444444446,
|
| 53550 |
+
"grad_norm": 0.09423828125,
|
| 53551 |
+
"learning_rate": 0.0005077874746256189,
|
| 53552 |
+
"loss": 2.0349,
|
| 53553 |
+
"step": 7649
|
| 53554 |
+
},
|
| 53555 |
+
{
|
| 53556 |
+
"epoch": 0.425,
|
| 53557 |
+
"grad_norm": 0.09814453125,
|
| 53558 |
+
"learning_rate": 0.0005077188701393258,
|
| 53559 |
+
"loss": 2.0101,
|
| 53560 |
+
"step": 7650
|
| 53561 |
+
},
|
| 53562 |
+
{
|
| 53563 |
+
"epoch": 0.4250555555555556,
|
| 53564 |
+
"grad_norm": 0.095703125,
|
| 53565 |
+
"learning_rate": 0.0005076502622364307,
|
| 53566 |
+
"loss": 2.0475,
|
| 53567 |
+
"step": 7651
|
| 53568 |
+
},
|
| 53569 |
+
{
|
| 53570 |
+
"epoch": 0.4251111111111111,
|
| 53571 |
+
"grad_norm": 0.09765625,
|
| 53572 |
+
"learning_rate": 0.0005075816509191101,
|
| 53573 |
+
"loss": 2.1006,
|
| 53574 |
+
"step": 7652
|
| 53575 |
+
},
|
| 53576 |
+
{
|
| 53577 |
+
"epoch": 0.4251666666666667,
|
| 53578 |
+
"grad_norm": 0.09619140625,
|
| 53579 |
+
"learning_rate": 0.0005075130361895397,
|
| 53580 |
+
"loss": 2.1172,
|
| 53581 |
+
"step": 7653
|
| 53582 |
+
},
|
| 53583 |
+
{
|
| 53584 |
+
"epoch": 0.4252222222222222,
|
| 53585 |
+
"grad_norm": 0.09375,
|
| 53586 |
+
"learning_rate": 0.0005074444180498961,
|
| 53587 |
+
"loss": 1.9908,
|
| 53588 |
+
"step": 7654
|
| 53589 |
+
},
|
| 53590 |
+
{
|
| 53591 |
+
"epoch": 0.42527777777777775,
|
| 53592 |
+
"grad_norm": 0.099609375,
|
| 53593 |
+
"learning_rate": 0.0005073757965023558,
|
| 53594 |
+
"loss": 2.0697,
|
| 53595 |
+
"step": 7655
|
| 53596 |
+
},
|
| 53597 |
+
{
|
| 53598 |
+
"epoch": 0.42533333333333334,
|
| 53599 |
+
"grad_norm": 0.09521484375,
|
| 53600 |
+
"learning_rate": 0.000507307171549095,
|
| 53601 |
+
"loss": 2.1164,
|
| 53602 |
+
"step": 7656
|
| 53603 |
+
},
|
| 53604 |
+
{
|
| 53605 |
+
"epoch": 0.42538888888888887,
|
| 53606 |
+
"grad_norm": 0.09619140625,
|
| 53607 |
+
"learning_rate": 0.0005072385431922908,
|
| 53608 |
+
"loss": 2.1024,
|
| 53609 |
+
"step": 7657
|
| 53610 |
+
},
|
| 53611 |
+
{
|
| 53612 |
+
"epoch": 0.42544444444444446,
|
| 53613 |
+
"grad_norm": 0.09765625,
|
| 53614 |
+
"learning_rate": 0.0005071699114341194,
|
| 53615 |
+
"loss": 2.0678,
|
| 53616 |
+
"step": 7658
|
| 53617 |
+
},
|
| 53618 |
+
{
|
| 53619 |
+
"epoch": 0.4255,
|
| 53620 |
+
"grad_norm": 0.09423828125,
|
| 53621 |
+
"learning_rate": 0.0005071012762767578,
|
| 53622 |
+
"loss": 2.0738,
|
| 53623 |
+
"step": 7659
|
| 53624 |
+
},
|
| 53625 |
+
{
|
| 53626 |
+
"epoch": 0.4255555555555556,
|
| 53627 |
+
"grad_norm": 0.0966796875,
|
| 53628 |
+
"learning_rate": 0.0005070326377223832,
|
| 53629 |
+
"loss": 2.07,
|
| 53630 |
+
"step": 7660
|
| 53631 |
+
},
|
| 53632 |
+
{
|
| 53633 |
+
"epoch": 0.4256111111111111,
|
| 53634 |
+
"grad_norm": 0.0966796875,
|
| 53635 |
+
"learning_rate": 0.0005069639957731725,
|
| 53636 |
+
"loss": 2.0173,
|
| 53637 |
+
"step": 7661
|
| 53638 |
+
},
|
| 53639 |
+
{
|
| 53640 |
+
"epoch": 0.4256666666666667,
|
| 53641 |
+
"grad_norm": 0.09521484375,
|
| 53642 |
+
"learning_rate": 0.0005068953504313027,
|
| 53643 |
+
"loss": 2.0319,
|
| 53644 |
+
"step": 7662
|
| 53645 |
+
},
|
| 53646 |
+
{
|
| 53647 |
+
"epoch": 0.4257222222222222,
|
| 53648 |
+
"grad_norm": 0.0966796875,
|
| 53649 |
+
"learning_rate": 0.0005068267016989515,
|
| 53650 |
+
"loss": 2.0446,
|
| 53651 |
+
"step": 7663
|
| 53652 |
+
},
|
| 53653 |
+
{
|
| 53654 |
+
"epoch": 0.42577777777777776,
|
| 53655 |
+
"grad_norm": 0.09619140625,
|
| 53656 |
+
"learning_rate": 0.000506758049578296,
|
| 53657 |
+
"loss": 2.0968,
|
| 53658 |
+
"step": 7664
|
| 53659 |
+
},
|
| 53660 |
+
{
|
| 53661 |
+
"epoch": 0.42583333333333334,
|
| 53662 |
+
"grad_norm": 0.09375,
|
| 53663 |
+
"learning_rate": 0.0005066893940715135,
|
| 53664 |
+
"loss": 2.0822,
|
| 53665 |
+
"step": 7665
|
| 53666 |
+
},
|
| 53667 |
+
{
|
| 53668 |
+
"epoch": 0.4258888888888889,
|
| 53669 |
+
"grad_norm": 0.09619140625,
|
| 53670 |
+
"learning_rate": 0.0005066207351807821,
|
| 53671 |
+
"loss": 2.0119,
|
| 53672 |
+
"step": 7666
|
| 53673 |
+
},
|
| 53674 |
+
{
|
| 53675 |
+
"epoch": 0.42594444444444446,
|
| 53676 |
+
"grad_norm": 0.09521484375,
|
| 53677 |
+
"learning_rate": 0.000506552072908279,
|
| 53678 |
+
"loss": 1.997,
|
| 53679 |
+
"step": 7667
|
| 53680 |
+
},
|
| 53681 |
+
{
|
| 53682 |
+
"epoch": 0.426,
|
| 53683 |
+
"grad_norm": 0.1025390625,
|
| 53684 |
+
"learning_rate": 0.0005064834072561825,
|
| 53685 |
+
"loss": 2.0509,
|
| 53686 |
+
"step": 7668
|
| 53687 |
+
},
|
| 53688 |
+
{
|
| 53689 |
+
"epoch": 0.4260555555555556,
|
| 53690 |
+
"grad_norm": 0.10009765625,
|
| 53691 |
+
"learning_rate": 0.0005064147382266701,
|
| 53692 |
+
"loss": 2.0613,
|
| 53693 |
+
"step": 7669
|
| 53694 |
+
},
|
| 53695 |
+
{
|
| 53696 |
+
"epoch": 0.4261111111111111,
|
| 53697 |
+
"grad_norm": 0.09765625,
|
| 53698 |
+
"learning_rate": 0.00050634606582192,
|
| 53699 |
+
"loss": 2.1096,
|
| 53700 |
+
"step": 7670
|
| 53701 |
+
},
|
| 53702 |
+
{
|
| 53703 |
+
"epoch": 0.4261666666666667,
|
| 53704 |
+
"grad_norm": 0.10400390625,
|
| 53705 |
+
"learning_rate": 0.0005062773900441104,
|
| 53706 |
+
"loss": 2.0428,
|
| 53707 |
+
"step": 7671
|
| 53708 |
+
},
|
| 53709 |
+
{
|
| 53710 |
+
"epoch": 0.4262222222222222,
|
| 53711 |
+
"grad_norm": 0.09716796875,
|
| 53712 |
+
"learning_rate": 0.0005062087108954193,
|
| 53713 |
+
"loss": 2.1075,
|
| 53714 |
+
"step": 7672
|
| 53715 |
+
},
|
| 53716 |
+
{
|
| 53717 |
+
"epoch": 0.42627777777777776,
|
| 53718 |
+
"grad_norm": 0.0947265625,
|
| 53719 |
+
"learning_rate": 0.0005061400283780255,
|
| 53720 |
+
"loss": 2.0474,
|
| 53721 |
+
"step": 7673
|
| 53722 |
+
},
|
| 53723 |
+
{
|
| 53724 |
+
"epoch": 0.42633333333333334,
|
| 53725 |
+
"grad_norm": 0.0986328125,
|
| 53726 |
+
"learning_rate": 0.000506071342494107,
|
| 53727 |
+
"loss": 2.0397,
|
| 53728 |
+
"step": 7674
|
| 53729 |
+
},
|
| 53730 |
+
{
|
| 53731 |
+
"epoch": 0.4263888888888889,
|
| 53732 |
+
"grad_norm": 0.09814453125,
|
| 53733 |
+
"learning_rate": 0.0005060026532458427,
|
| 53734 |
+
"loss": 2.0678,
|
| 53735 |
+
"step": 7675
|
| 53736 |
+
},
|
| 53737 |
+
{
|
| 53738 |
+
"epoch": 0.42644444444444446,
|
| 53739 |
+
"grad_norm": 0.095703125,
|
| 53740 |
+
"learning_rate": 0.0005059339606354109,
|
| 53741 |
+
"loss": 2.1112,
|
| 53742 |
+
"step": 7676
|
| 53743 |
+
},
|
| 53744 |
+
{
|
| 53745 |
+
"epoch": 0.4265,
|
| 53746 |
+
"grad_norm": 0.09814453125,
|
| 53747 |
+
"learning_rate": 0.0005058652646649906,
|
| 53748 |
+
"loss": 2.0548,
|
| 53749 |
+
"step": 7677
|
| 53750 |
+
},
|
| 53751 |
+
{
|
| 53752 |
+
"epoch": 0.4265555555555556,
|
| 53753 |
+
"grad_norm": 0.095703125,
|
| 53754 |
+
"learning_rate": 0.0005057965653367609,
|
| 53755 |
+
"loss": 2.108,
|
| 53756 |
+
"step": 7678
|
| 53757 |
+
},
|
| 53758 |
+
{
|
| 53759 |
+
"epoch": 0.4266111111111111,
|
| 53760 |
+
"grad_norm": 0.1005859375,
|
| 53761 |
+
"learning_rate": 0.0005057278626529004,
|
| 53762 |
+
"loss": 2.0976,
|
| 53763 |
+
"step": 7679
|
| 53764 |
+
},
|
| 53765 |
+
{
|
| 53766 |
+
"epoch": 0.4266666666666667,
|
| 53767 |
+
"grad_norm": 0.09521484375,
|
| 53768 |
+
"learning_rate": 0.0005056591566155883,
|
| 53769 |
+
"loss": 2.0214,
|
| 53770 |
+
"step": 7680
|
| 53771 |
+
},
|
| 53772 |
+
{
|
| 53773 |
+
"epoch": 0.4267222222222222,
|
| 53774 |
+
"grad_norm": 0.09521484375,
|
| 53775 |
+
"learning_rate": 0.0005055904472270039,
|
| 53776 |
+
"loss": 1.9936,
|
| 53777 |
+
"step": 7681
|
| 53778 |
+
},
|
| 53779 |
+
{
|
| 53780 |
+
"epoch": 0.42677777777777776,
|
| 53781 |
+
"grad_norm": 0.09423828125,
|
| 53782 |
+
"learning_rate": 0.0005055217344893266,
|
| 53783 |
+
"loss": 2.04,
|
| 53784 |
+
"step": 7682
|
| 53785 |
+
},
|
| 53786 |
+
{
|
| 53787 |
+
"epoch": 0.42683333333333334,
|
| 53788 |
+
"grad_norm": 0.0966796875,
|
| 53789 |
+
"learning_rate": 0.0005054530184047355,
|
| 53790 |
+
"loss": 2.0601,
|
| 53791 |
+
"step": 7683
|
| 53792 |
+
},
|
| 53793 |
+
{
|
| 53794 |
+
"epoch": 0.4268888888888889,
|
| 53795 |
+
"grad_norm": 0.095703125,
|
| 53796 |
+
"learning_rate": 0.0005053842989754104,
|
| 53797 |
+
"loss": 2.0246,
|
| 53798 |
+
"step": 7684
|
| 53799 |
+
},
|
| 53800 |
+
{
|
| 53801 |
+
"epoch": 0.42694444444444446,
|
| 53802 |
+
"grad_norm": 0.09228515625,
|
| 53803 |
+
"learning_rate": 0.0005053155762035308,
|
| 53804 |
+
"loss": 2.0611,
|
| 53805 |
+
"step": 7685
|
| 53806 |
+
},
|
| 53807 |
+
{
|
| 53808 |
+
"epoch": 0.427,
|
| 53809 |
+
"grad_norm": 0.09521484375,
|
| 53810 |
+
"learning_rate": 0.0005052468500912766,
|
| 53811 |
+
"loss": 2.1271,
|
| 53812 |
+
"step": 7686
|
| 53813 |
+
},
|
| 53814 |
+
{
|
| 53815 |
+
"epoch": 0.4270555555555556,
|
| 53816 |
+
"grad_norm": 0.09814453125,
|
| 53817 |
+
"learning_rate": 0.0005051781206408272,
|
| 53818 |
+
"loss": 2.1037,
|
| 53819 |
+
"step": 7687
|
| 53820 |
+
},
|
| 53821 |
+
{
|
| 53822 |
+
"epoch": 0.4271111111111111,
|
| 53823 |
+
"grad_norm": 0.0947265625,
|
| 53824 |
+
"learning_rate": 0.000505109387854363,
|
| 53825 |
+
"loss": 1.998,
|
| 53826 |
+
"step": 7688
|
| 53827 |
+
},
|
| 53828 |
+
{
|
| 53829 |
+
"epoch": 0.42716666666666664,
|
| 53830 |
+
"grad_norm": 0.0947265625,
|
| 53831 |
+
"learning_rate": 0.000505040651734064,
|
| 53832 |
+
"loss": 2.0144,
|
| 53833 |
+
"step": 7689
|
| 53834 |
+
},
|
| 53835 |
+
{
|
| 53836 |
+
"epoch": 0.4272222222222222,
|
| 53837 |
+
"grad_norm": 0.09521484375,
|
| 53838 |
+
"learning_rate": 0.00050497191228211,
|
| 53839 |
+
"loss": 2.0784,
|
| 53840 |
+
"step": 7690
|
| 53841 |
+
},
|
| 53842 |
+
{
|
| 53843 |
+
"epoch": 0.42727777777777776,
|
| 53844 |
+
"grad_norm": 0.09423828125,
|
| 53845 |
+
"learning_rate": 0.0005049031695006818,
|
| 53846 |
+
"loss": 2.0365,
|
| 53847 |
+
"step": 7691
|
| 53848 |
+
},
|
| 53849 |
+
{
|
| 53850 |
+
"epoch": 0.42733333333333334,
|
| 53851 |
+
"grad_norm": 0.0966796875,
|
| 53852 |
+
"learning_rate": 0.0005048344233919593,
|
| 53853 |
+
"loss": 2.0774,
|
| 53854 |
+
"step": 7692
|
| 53855 |
+
},
|
| 53856 |
+
{
|
| 53857 |
+
"epoch": 0.4273888888888889,
|
| 53858 |
+
"grad_norm": 0.09326171875,
|
| 53859 |
+
"learning_rate": 0.0005047656739581232,
|
| 53860 |
+
"loss": 2.0516,
|
| 53861 |
+
"step": 7693
|
| 53862 |
+
},
|
| 53863 |
+
{
|
| 53864 |
+
"epoch": 0.42744444444444446,
|
| 53865 |
+
"grad_norm": 0.095703125,
|
| 53866 |
+
"learning_rate": 0.000504696921201354,
|
| 53867 |
+
"loss": 2.0702,
|
| 53868 |
+
"step": 7694
|
| 53869 |
+
},
|
| 53870 |
+
{
|
| 53871 |
+
"epoch": 0.4275,
|
| 53872 |
+
"grad_norm": 0.09375,
|
| 53873 |
+
"learning_rate": 0.0005046281651238324,
|
| 53874 |
+
"loss": 2.0111,
|
| 53875 |
+
"step": 7695
|
| 53876 |
+
},
|
| 53877 |
+
{
|
| 53878 |
+
"epoch": 0.4275555555555556,
|
| 53879 |
+
"grad_norm": 0.09619140625,
|
| 53880 |
+
"learning_rate": 0.0005045594057277393,
|
| 53881 |
+
"loss": 2.0375,
|
| 53882 |
+
"step": 7696
|
| 53883 |
+
},
|
| 53884 |
+
{
|
| 53885 |
+
"epoch": 0.4276111111111111,
|
| 53886 |
+
"grad_norm": 0.0947265625,
|
| 53887 |
+
"learning_rate": 0.0005044906430152554,
|
| 53888 |
+
"loss": 2.0448,
|
| 53889 |
+
"step": 7697
|
| 53890 |
+
},
|
| 53891 |
+
{
|
| 53892 |
+
"epoch": 0.42766666666666664,
|
| 53893 |
+
"grad_norm": 0.09326171875,
|
| 53894 |
+
"learning_rate": 0.0005044218769885618,
|
| 53895 |
+
"loss": 2.0596,
|
| 53896 |
+
"step": 7698
|
| 53897 |
+
},
|
| 53898 |
+
{
|
| 53899 |
+
"epoch": 0.4277222222222222,
|
| 53900 |
+
"grad_norm": 0.09375,
|
| 53901 |
+
"learning_rate": 0.0005043531076498397,
|
| 53902 |
+
"loss": 2.0695,
|
| 53903 |
+
"step": 7699
|
| 53904 |
+
},
|
| 53905 |
+
{
|
| 53906 |
+
"epoch": 0.42777777777777776,
|
| 53907 |
+
"grad_norm": 0.0986328125,
|
| 53908 |
+
"learning_rate": 0.00050428433500127,
|
| 53909 |
+
"loss": 2.0738,
|
| 53910 |
+
"step": 7700
|
| 53911 |
+
},
|
| 53912 |
+
{
|
| 53913 |
+
"epoch": 0.42783333333333334,
|
| 53914 |
+
"grad_norm": 0.09326171875,
|
| 53915 |
+
"learning_rate": 0.0005042155590450345,
|
| 53916 |
+
"loss": 2.0416,
|
| 53917 |
+
"step": 7701
|
| 53918 |
+
},
|
| 53919 |
+
{
|
| 53920 |
+
"epoch": 0.4278888888888889,
|
| 53921 |
+
"grad_norm": 0.09375,
|
| 53922 |
+
"learning_rate": 0.0005041467797833142,
|
| 53923 |
+
"loss": 2.0729,
|
| 53924 |
+
"step": 7702
|
| 53925 |
+
},
|
| 53926 |
+
{
|
| 53927 |
+
"epoch": 0.42794444444444446,
|
| 53928 |
+
"grad_norm": 0.09423828125,
|
| 53929 |
+
"learning_rate": 0.0005040779972182907,
|
| 53930 |
+
"loss": 2.0806,
|
| 53931 |
+
"step": 7703
|
| 53932 |
+
},
|
| 53933 |
+
{
|
| 53934 |
+
"epoch": 0.428,
|
| 53935 |
+
"grad_norm": 0.0927734375,
|
| 53936 |
+
"learning_rate": 0.0005040092113521459,
|
| 53937 |
+
"loss": 2.0554,
|
| 53938 |
+
"step": 7704
|
| 53939 |
+
},
|
| 53940 |
+
{
|
| 53941 |
+
"epoch": 0.4280555555555556,
|
| 53942 |
+
"grad_norm": 0.09228515625,
|
| 53943 |
+
"learning_rate": 0.0005039404221870612,
|
| 53944 |
+
"loss": 2.0516,
|
| 53945 |
+
"step": 7705
|
| 53946 |
+
},
|
| 53947 |
+
{
|
| 53948 |
+
"epoch": 0.4281111111111111,
|
| 53949 |
+
"grad_norm": 0.09423828125,
|
| 53950 |
+
"learning_rate": 0.0005038716297252186,
|
| 53951 |
+
"loss": 2.0099,
|
| 53952 |
+
"step": 7706
|
| 53953 |
+
},
|
| 53954 |
+
{
|
| 53955 |
+
"epoch": 0.42816666666666664,
|
| 53956 |
+
"grad_norm": 0.09375,
|
| 53957 |
+
"learning_rate": 0.0005038028339688001,
|
| 53958 |
+
"loss": 2.0334,
|
| 53959 |
+
"step": 7707
|
| 53960 |
+
},
|
| 53961 |
+
{
|
| 53962 |
+
"epoch": 0.4282222222222222,
|
| 53963 |
+
"grad_norm": 0.095703125,
|
| 53964 |
+
"learning_rate": 0.0005037340349199875,
|
| 53965 |
+
"loss": 2.0626,
|
| 53966 |
+
"step": 7708
|
| 53967 |
+
},
|
| 53968 |
+
{
|
| 53969 |
+
"epoch": 0.42827777777777776,
|
| 53970 |
+
"grad_norm": 0.0927734375,
|
| 53971 |
+
"learning_rate": 0.0005036652325809632,
|
| 53972 |
+
"loss": 2.0602,
|
| 53973 |
+
"step": 7709
|
| 53974 |
+
},
|
| 53975 |
+
{
|
| 53976 |
+
"epoch": 0.42833333333333334,
|
| 53977 |
+
"grad_norm": 0.09521484375,
|
| 53978 |
+
"learning_rate": 0.0005035964269539093,
|
| 53979 |
+
"loss": 2.1144,
|
| 53980 |
+
"step": 7710
|
| 53981 |
+
},
|
| 53982 |
+
{
|
| 53983 |
+
"epoch": 0.4283888888888889,
|
| 53984 |
+
"grad_norm": 0.09326171875,
|
| 53985 |
+
"learning_rate": 0.0005035276180410083,
|
| 53986 |
+
"loss": 2.0858,
|
| 53987 |
+
"step": 7711
|
| 53988 |
+
},
|
| 53989 |
+
{
|
| 53990 |
+
"epoch": 0.42844444444444446,
|
| 53991 |
+
"grad_norm": 0.09521484375,
|
| 53992 |
+
"learning_rate": 0.0005034588058444426,
|
| 53993 |
+
"loss": 2.0949,
|
| 53994 |
+
"step": 7712
|
| 53995 |
+
},
|
| 53996 |
+
{
|
| 53997 |
+
"epoch": 0.4285,
|
| 53998 |
+
"grad_norm": 0.09521484375,
|
| 53999 |
+
"learning_rate": 0.0005033899903663947,
|
| 54000 |
+
"loss": 2.003,
|
| 54001 |
+
"step": 7713
|
| 54002 |
+
},
|
| 54003 |
+
{
|
| 54004 |
+
"epoch": 0.4285555555555556,
|
| 54005 |
+
"grad_norm": 0.09423828125,
|
| 54006 |
+
"learning_rate": 0.0005033211716090475,
|
| 54007 |
+
"loss": 2.0823,
|
| 54008 |
+
"step": 7714
|
| 54009 |
+
},
|
| 54010 |
+
{
|
| 54011 |
+
"epoch": 0.4286111111111111,
|
| 54012 |
+
"grad_norm": 0.09326171875,
|
| 54013 |
+
"learning_rate": 0.0005032523495745835,
|
| 54014 |
+
"loss": 2.0538,
|
| 54015 |
+
"step": 7715
|
| 54016 |
+
},
|
| 54017 |
+
{
|
| 54018 |
+
"epoch": 0.42866666666666664,
|
| 54019 |
+
"grad_norm": 0.0947265625,
|
| 54020 |
+
"learning_rate": 0.0005031835242651857,
|
| 54021 |
+
"loss": 2.0344,
|
| 54022 |
+
"step": 7716
|
| 54023 |
+
},
|
| 54024 |
+
{
|
| 54025 |
+
"epoch": 0.4287222222222222,
|
| 54026 |
+
"grad_norm": 0.09326171875,
|
| 54027 |
+
"learning_rate": 0.0005031146956830373,
|
| 54028 |
+
"loss": 2.013,
|
| 54029 |
+
"step": 7717
|
| 54030 |
+
},
|
| 54031 |
+
{
|
| 54032 |
+
"epoch": 0.42877777777777776,
|
| 54033 |
+
"grad_norm": 0.09423828125,
|
| 54034 |
+
"learning_rate": 0.0005030458638303207,
|
| 54035 |
+
"loss": 2.0275,
|
| 54036 |
+
"step": 7718
|
| 54037 |
+
},
|
| 54038 |
+
{
|
| 54039 |
+
"epoch": 0.42883333333333334,
|
| 54040 |
+
"grad_norm": 0.09375,
|
| 54041 |
+
"learning_rate": 0.00050297702870922,
|
| 54042 |
+
"loss": 2.0518,
|
| 54043 |
+
"step": 7719
|
| 54044 |
+
},
|
| 54045 |
+
{
|
| 54046 |
+
"epoch": 0.4288888888888889,
|
| 54047 |
+
"grad_norm": 0.09326171875,
|
| 54048 |
+
"learning_rate": 0.0005029081903219179,
|
| 54049 |
+
"loss": 2.0505,
|
| 54050 |
+
"step": 7720
|
| 54051 |
+
},
|
| 54052 |
+
{
|
| 54053 |
+
"epoch": 0.42894444444444446,
|
| 54054 |
+
"grad_norm": 0.0927734375,
|
| 54055 |
+
"learning_rate": 0.000502839348670598,
|
| 54056 |
+
"loss": 2.0341,
|
| 54057 |
+
"step": 7721
|
| 54058 |
+
},
|
| 54059 |
+
{
|
| 54060 |
+
"epoch": 0.429,
|
| 54061 |
+
"grad_norm": 0.09619140625,
|
| 54062 |
+
"learning_rate": 0.0005027705037574437,
|
| 54063 |
+
"loss": 2.0285,
|
| 54064 |
+
"step": 7722
|
| 54065 |
+
},
|
| 54066 |
+
{
|
| 54067 |
+
"epoch": 0.4290555555555556,
|
| 54068 |
+
"grad_norm": 0.0947265625,
|
| 54069 |
+
"learning_rate": 0.0005027016555846386,
|
| 54070 |
+
"loss": 2.0274,
|
| 54071 |
+
"step": 7723
|
| 54072 |
+
},
|
| 54073 |
+
{
|
| 54074 |
+
"epoch": 0.4291111111111111,
|
| 54075 |
+
"grad_norm": 0.0966796875,
|
| 54076 |
+
"learning_rate": 0.0005026328041543666,
|
| 54077 |
+
"loss": 2.0541,
|
| 54078 |
+
"step": 7724
|
| 54079 |
+
},
|
| 54080 |
+
{
|
| 54081 |
+
"epoch": 0.42916666666666664,
|
| 54082 |
+
"grad_norm": 0.0947265625,
|
| 54083 |
+
"learning_rate": 0.0005025639494688113,
|
| 54084 |
+
"loss": 2.1112,
|
| 54085 |
+
"step": 7725
|
| 54086 |
+
},
|
| 54087 |
+
{
|
| 54088 |
+
"epoch": 0.4292222222222222,
|
| 54089 |
+
"grad_norm": 0.09228515625,
|
| 54090 |
+
"learning_rate": 0.0005024950915301569,
|
| 54091 |
+
"loss": 2.0711,
|
| 54092 |
+
"step": 7726
|
| 54093 |
+
},
|
| 54094 |
+
{
|
| 54095 |
+
"epoch": 0.42927777777777776,
|
| 54096 |
+
"grad_norm": 0.09521484375,
|
| 54097 |
+
"learning_rate": 0.000502426230340587,
|
| 54098 |
+
"loss": 2.0835,
|
| 54099 |
+
"step": 7727
|
| 54100 |
+
},
|
| 54101 |
+
{
|
| 54102 |
+
"epoch": 0.42933333333333334,
|
| 54103 |
+
"grad_norm": 0.091796875,
|
| 54104 |
+
"learning_rate": 0.0005023573659022862,
|
| 54105 |
+
"loss": 2.0074,
|
| 54106 |
+
"step": 7728
|
| 54107 |
+
},
|
| 54108 |
+
{
|
| 54109 |
+
"epoch": 0.4293888888888889,
|
| 54110 |
+
"grad_norm": 0.091796875,
|
| 54111 |
+
"learning_rate": 0.0005022884982174384,
|
| 54112 |
+
"loss": 2.0123,
|
| 54113 |
+
"step": 7729
|
| 54114 |
+
},
|
| 54115 |
+
{
|
| 54116 |
+
"epoch": 0.42944444444444446,
|
| 54117 |
+
"grad_norm": 0.0927734375,
|
| 54118 |
+
"learning_rate": 0.0005022196272882279,
|
| 54119 |
+
"loss": 2.0383,
|
| 54120 |
+
"step": 7730
|
| 54121 |
+
},
|
| 54122 |
+
{
|
| 54123 |
+
"epoch": 0.4295,
|
| 54124 |
+
"grad_norm": 0.09228515625,
|
| 54125 |
+
"learning_rate": 0.0005021507531168393,
|
| 54126 |
+
"loss": 2.0588,
|
| 54127 |
+
"step": 7731
|
| 54128 |
+
},
|
| 54129 |
+
{
|
| 54130 |
+
"epoch": 0.4295555555555556,
|
| 54131 |
+
"grad_norm": 0.09326171875,
|
| 54132 |
+
"learning_rate": 0.0005020818757054571,
|
| 54133 |
+
"loss": 2.0616,
|
| 54134 |
+
"step": 7732
|
| 54135 |
+
},
|
| 54136 |
+
{
|
| 54137 |
+
"epoch": 0.4296111111111111,
|
| 54138 |
+
"grad_norm": 0.09423828125,
|
| 54139 |
+
"learning_rate": 0.0005020129950562658,
|
| 54140 |
+
"loss": 2.0823,
|
| 54141 |
+
"step": 7733
|
| 54142 |
+
},
|
| 54143 |
+
{
|
| 54144 |
+
"epoch": 0.42966666666666664,
|
| 54145 |
+
"grad_norm": 0.09375,
|
| 54146 |
+
"learning_rate": 0.0005019441111714503,
|
| 54147 |
+
"loss": 2.1112,
|
| 54148 |
+
"step": 7734
|
| 54149 |
+
},
|
| 54150 |
+
{
|
| 54151 |
+
"epoch": 0.4297222222222222,
|
| 54152 |
+
"grad_norm": 0.0947265625,
|
| 54153 |
+
"learning_rate": 0.0005018752240531955,
|
| 54154 |
+
"loss": 2.0731,
|
| 54155 |
+
"step": 7735
|
| 54156 |
+
},
|
| 54157 |
+
{
|
| 54158 |
+
"epoch": 0.42977777777777776,
|
| 54159 |
+
"grad_norm": 0.0927734375,
|
| 54160 |
+
"learning_rate": 0.0005018063337036863,
|
| 54161 |
+
"loss": 2.1218,
|
| 54162 |
+
"step": 7736
|
| 54163 |
+
},
|
| 54164 |
+
{
|
| 54165 |
+
"epoch": 0.42983333333333335,
|
| 54166 |
+
"grad_norm": 0.09375,
|
| 54167 |
+
"learning_rate": 0.0005017374401251075,
|
| 54168 |
+
"loss": 2.0987,
|
| 54169 |
+
"step": 7737
|
| 54170 |
+
},
|
| 54171 |
+
{
|
| 54172 |
+
"epoch": 0.4298888888888889,
|
| 54173 |
+
"grad_norm": 0.091796875,
|
| 54174 |
+
"learning_rate": 0.0005016685433196445,
|
| 54175 |
+
"loss": 2.0159,
|
| 54176 |
+
"step": 7738
|
| 54177 |
+
},
|
| 54178 |
+
{
|
| 54179 |
+
"epoch": 0.42994444444444446,
|
| 54180 |
+
"grad_norm": 0.091796875,
|
| 54181 |
+
"learning_rate": 0.0005015996432894825,
|
| 54182 |
+
"loss": 2.0296,
|
| 54183 |
+
"step": 7739
|
| 54184 |
+
},
|
| 54185 |
+
{
|
| 54186 |
+
"epoch": 0.43,
|
| 54187 |
+
"grad_norm": 0.09228515625,
|
| 54188 |
+
"learning_rate": 0.0005015307400368068,
|
| 54189 |
+
"loss": 2.0461,
|
| 54190 |
+
"step": 7740
|
| 54191 |
+
},
|
| 54192 |
+
{
|
| 54193 |
+
"epoch": 0.4300555555555556,
|
| 54194 |
+
"grad_norm": 0.09423828125,
|
| 54195 |
+
"learning_rate": 0.000501461833563803,
|
| 54196 |
+
"loss": 2.0542,
|
| 54197 |
+
"step": 7741
|
| 54198 |
+
},
|
| 54199 |
+
{
|
| 54200 |
+
"epoch": 0.4301111111111111,
|
| 54201 |
+
"grad_norm": 0.09228515625,
|
| 54202 |
+
"learning_rate": 0.0005013929238726565,
|
| 54203 |
+
"loss": 2.0473,
|
| 54204 |
+
"step": 7742
|
| 54205 |
+
},
|
| 54206 |
+
{
|
| 54207 |
+
"epoch": 0.43016666666666664,
|
| 54208 |
+
"grad_norm": 0.09326171875,
|
| 54209 |
+
"learning_rate": 0.000501324010965553,
|
| 54210 |
+
"loss": 2.0629,
|
| 54211 |
+
"step": 7743
|
| 54212 |
+
},
|
| 54213 |
+
{
|
| 54214 |
+
"epoch": 0.43022222222222223,
|
| 54215 |
+
"grad_norm": 0.09423828125,
|
| 54216 |
+
"learning_rate": 0.0005012550948446784,
|
| 54217 |
+
"loss": 2.0127,
|
| 54218 |
+
"step": 7744
|
| 54219 |
+
},
|
| 54220 |
+
{
|
| 54221 |
+
"epoch": 0.43027777777777776,
|
| 54222 |
+
"grad_norm": 0.09228515625,
|
| 54223 |
+
"learning_rate": 0.0005011861755122185,
|
| 54224 |
+
"loss": 2.0507,
|
| 54225 |
+
"step": 7745
|
| 54226 |
+
},
|
| 54227 |
+
{
|
| 54228 |
+
"epoch": 0.43033333333333335,
|
| 54229 |
+
"grad_norm": 0.0947265625,
|
| 54230 |
+
"learning_rate": 0.000501117252970359,
|
| 54231 |
+
"loss": 2.0902,
|
| 54232 |
+
"step": 7746
|
| 54233 |
+
},
|
| 54234 |
+
{
|
| 54235 |
+
"epoch": 0.4303888888888889,
|
| 54236 |
+
"grad_norm": 0.09130859375,
|
| 54237 |
+
"learning_rate": 0.0005010483272212863,
|
| 54238 |
+
"loss": 2.0839,
|
| 54239 |
+
"step": 7747
|
| 54240 |
+
},
|
| 54241 |
+
{
|
| 54242 |
+
"epoch": 0.43044444444444446,
|
| 54243 |
+
"grad_norm": 0.0927734375,
|
| 54244 |
+
"learning_rate": 0.0005009793982671866,
|
| 54245 |
+
"loss": 2.0393,
|
| 54246 |
+
"step": 7748
|
| 54247 |
+
},
|
| 54248 |
+
{
|
| 54249 |
+
"epoch": 0.4305,
|
| 54250 |
+
"grad_norm": 0.09326171875,
|
| 54251 |
+
"learning_rate": 0.0005009104661102459,
|
| 54252 |
+
"loss": 2.0691,
|
| 54253 |
+
"step": 7749
|
| 54254 |
+
},
|
| 54255 |
+
{
|
| 54256 |
+
"epoch": 0.4305555555555556,
|
| 54257 |
+
"grad_norm": 0.09326171875,
|
| 54258 |
+
"learning_rate": 0.0005008415307526507,
|
| 54259 |
+
"loss": 2.0803,
|
| 54260 |
+
"step": 7750
|
| 54261 |
+
},
|
| 54262 |
+
{
|
| 54263 |
+
"epoch": 0.4306111111111111,
|
| 54264 |
+
"grad_norm": 0.08984375,
|
| 54265 |
+
"learning_rate": 0.0005007725921965875,
|
| 54266 |
+
"loss": 2.0248,
|
| 54267 |
+
"step": 7751
|
| 54268 |
+
},
|
| 54269 |
+
{
|
| 54270 |
+
"epoch": 0.43066666666666664,
|
| 54271 |
+
"grad_norm": 0.09521484375,
|
| 54272 |
+
"learning_rate": 0.0005007036504442429,
|
| 54273 |
+
"loss": 2.0764,
|
| 54274 |
+
"step": 7752
|
| 54275 |
+
},
|
| 54276 |
+
{
|
| 54277 |
+
"epoch": 0.43072222222222223,
|
| 54278 |
+
"grad_norm": 0.0927734375,
|
| 54279 |
+
"learning_rate": 0.0005006347054978034,
|
| 54280 |
+
"loss": 2.0493,
|
| 54281 |
+
"step": 7753
|
| 54282 |
+
},
|
| 54283 |
+
{
|
| 54284 |
+
"epoch": 0.43077777777777776,
|
| 54285 |
+
"grad_norm": 0.09326171875,
|
| 54286 |
+
"learning_rate": 0.0005005657573594561,
|
| 54287 |
+
"loss": 2.1023,
|
| 54288 |
+
"step": 7754
|
| 54289 |
+
},
|
| 54290 |
+
{
|
| 54291 |
+
"epoch": 0.43083333333333335,
|
| 54292 |
+
"grad_norm": 0.0927734375,
|
| 54293 |
+
"learning_rate": 0.0005004968060313876,
|
| 54294 |
+
"loss": 2.0341,
|
| 54295 |
+
"step": 7755
|
| 54296 |
+
},
|
| 54297 |
+
{
|
| 54298 |
+
"epoch": 0.4308888888888889,
|
| 54299 |
+
"grad_norm": 0.09375,
|
| 54300 |
+
"learning_rate": 0.000500427851515785,
|
| 54301 |
+
"loss": 2.0457,
|
| 54302 |
+
"step": 7756
|
| 54303 |
+
},
|
| 54304 |
+
{
|
| 54305 |
+
"epoch": 0.43094444444444446,
|
| 54306 |
+
"grad_norm": 0.091796875,
|
| 54307 |
+
"learning_rate": 0.0005003588938148354,
|
| 54308 |
+
"loss": 2.0505,
|
| 54309 |
+
"step": 7757
|
| 54310 |
+
},
|
| 54311 |
+
{
|
| 54312 |
+
"epoch": 0.431,
|
| 54313 |
+
"grad_norm": 0.09375,
|
| 54314 |
+
"learning_rate": 0.000500289932930726,
|
| 54315 |
+
"loss": 2.0553,
|
| 54316 |
+
"step": 7758
|
| 54317 |
+
},
|
| 54318 |
+
{
|
| 54319 |
+
"epoch": 0.4310555555555556,
|
| 54320 |
+
"grad_norm": 0.0947265625,
|
| 54321 |
+
"learning_rate": 0.0005002209688656439,
|
| 54322 |
+
"loss": 2.0145,
|
| 54323 |
+
"step": 7759
|
| 54324 |
+
},
|
| 54325 |
+
{
|
| 54326 |
+
"epoch": 0.4311111111111111,
|
| 54327 |
+
"grad_norm": 0.091796875,
|
| 54328 |
+
"learning_rate": 0.0005001520016217766,
|
| 54329 |
+
"loss": 2.0799,
|
| 54330 |
+
"step": 7760
|
| 54331 |
+
},
|
| 54332 |
+
{
|
| 54333 |
+
"epoch": 0.43116666666666664,
|
| 54334 |
+
"grad_norm": 0.0927734375,
|
| 54335 |
+
"learning_rate": 0.0005000830312013116,
|
| 54336 |
+
"loss": 2.0242,
|
| 54337 |
+
"step": 7761
|
| 54338 |
+
},
|
| 54339 |
+
{
|
| 54340 |
+
"epoch": 0.43122222222222223,
|
| 54341 |
+
"grad_norm": 0.09228515625,
|
| 54342 |
+
"learning_rate": 0.0005000140576064367,
|
| 54343 |
+
"loss": 2.0332,
|
| 54344 |
+
"step": 7762
|
| 54345 |
+
},
|
| 54346 |
+
{
|
| 54347 |
+
"epoch": 0.43127777777777776,
|
| 54348 |
+
"grad_norm": 0.0908203125,
|
| 54349 |
+
"learning_rate": 0.0004999450808393392,
|
| 54350 |
+
"loss": 2.025,
|
| 54351 |
+
"step": 7763
|
| 54352 |
+
},
|
| 54353 |
+
{
|
| 54354 |
+
"epoch": 0.43133333333333335,
|
| 54355 |
+
"grad_norm": 0.0927734375,
|
| 54356 |
+
"learning_rate": 0.0004998761009022072,
|
| 54357 |
+
"loss": 2.0442,
|
| 54358 |
+
"step": 7764
|
| 54359 |
+
},
|
| 54360 |
+
{
|
| 54361 |
+
"epoch": 0.4313888888888889,
|
| 54362 |
+
"grad_norm": 0.09228515625,
|
| 54363 |
+
"learning_rate": 0.0004998071177972284,
|
| 54364 |
+
"loss": 2.021,
|
| 54365 |
+
"step": 7765
|
| 54366 |
+
},
|
| 54367 |
+
{
|
| 54368 |
+
"epoch": 0.43144444444444446,
|
| 54369 |
+
"grad_norm": 0.091796875,
|
| 54370 |
+
"learning_rate": 0.0004997381315265907,
|
| 54371 |
+
"loss": 2.1047,
|
| 54372 |
+
"step": 7766
|
| 54373 |
+
},
|
| 54374 |
+
{
|
| 54375 |
+
"epoch": 0.4315,
|
| 54376 |
+
"grad_norm": 0.09228515625,
|
| 54377 |
+
"learning_rate": 0.0004996691420924826,
|
| 54378 |
+
"loss": 2.0425,
|
| 54379 |
+
"step": 7767
|
| 54380 |
+
},
|
| 54381 |
+
{
|
| 54382 |
+
"epoch": 0.4315555555555556,
|
| 54383 |
+
"grad_norm": 0.091796875,
|
| 54384 |
+
"learning_rate": 0.0004996001494970919,
|
| 54385 |
+
"loss": 2.0331,
|
| 54386 |
+
"step": 7768
|
| 54387 |
+
},
|
| 54388 |
+
{
|
| 54389 |
+
"epoch": 0.4316111111111111,
|
| 54390 |
+
"grad_norm": 0.08935546875,
|
| 54391 |
+
"learning_rate": 0.000499531153742607,
|
| 54392 |
+
"loss": 2.006,
|
| 54393 |
+
"step": 7769
|
| 54394 |
+
},
|
| 54395 |
+
{
|
| 54396 |
+
"epoch": 0.43166666666666664,
|
| 54397 |
+
"grad_norm": 0.09033203125,
|
| 54398 |
+
"learning_rate": 0.0004994621548312164,
|
| 54399 |
+
"loss": 2.0547,
|
| 54400 |
+
"step": 7770
|
| 54401 |
+
},
|
| 54402 |
+
{
|
| 54403 |
+
"epoch": 0.43172222222222223,
|
| 54404 |
+
"grad_norm": 0.09130859375,
|
| 54405 |
+
"learning_rate": 0.0004993931527651084,
|
| 54406 |
+
"loss": 2.0425,
|
| 54407 |
+
"step": 7771
|
| 54408 |
+
},
|
| 54409 |
+
{
|
| 54410 |
+
"epoch": 0.43177777777777776,
|
| 54411 |
+
"grad_norm": 0.09033203125,
|
| 54412 |
+
"learning_rate": 0.0004993241475464719,
|
| 54413 |
+
"loss": 2.0691,
|
| 54414 |
+
"step": 7772
|
| 54415 |
+
},
|
| 54416 |
+
{
|
| 54417 |
+
"epoch": 0.43183333333333335,
|
| 54418 |
+
"grad_norm": 0.0947265625,
|
| 54419 |
+
"learning_rate": 0.0004992551391774953,
|
| 54420 |
+
"loss": 2.0741,
|
| 54421 |
+
"step": 7773
|
| 54422 |
+
},
|
| 54423 |
+
{
|
| 54424 |
+
"epoch": 0.4318888888888889,
|
| 54425 |
+
"grad_norm": 0.09228515625,
|
| 54426 |
+
"learning_rate": 0.0004991861276603673,
|
| 54427 |
+
"loss": 2.0586,
|
| 54428 |
+
"step": 7774
|
| 54429 |
+
},
|
| 54430 |
+
{
|
| 54431 |
+
"epoch": 0.43194444444444446,
|
| 54432 |
+
"grad_norm": 0.09130859375,
|
| 54433 |
+
"learning_rate": 0.0004991171129972772,
|
| 54434 |
+
"loss": 2.0175,
|
| 54435 |
+
"step": 7775
|
| 54436 |
+
},
|
| 54437 |
+
{
|
| 54438 |
+
"epoch": 0.432,
|
| 54439 |
+
"grad_norm": 0.09423828125,
|
| 54440 |
+
"learning_rate": 0.0004990480951904135,
|
| 54441 |
+
"loss": 2.0959,
|
| 54442 |
+
"step": 7776
|
| 54443 |
+
},
|
| 54444 |
+
{
|
| 54445 |
+
"epoch": 0.4320555555555556,
|
| 54446 |
+
"grad_norm": 0.09375,
|
| 54447 |
+
"learning_rate": 0.0004989790742419657,
|
| 54448 |
+
"loss": 2.0855,
|
| 54449 |
+
"step": 7777
|
| 54450 |
+
},
|
| 54451 |
+
{
|
| 54452 |
+
"epoch": 0.4321111111111111,
|
| 54453 |
+
"grad_norm": 0.09375,
|
| 54454 |
+
"learning_rate": 0.000498910050154123,
|
| 54455 |
+
"loss": 2.0646,
|
| 54456 |
+
"step": 7778
|
| 54457 |
+
},
|
| 54458 |
+
{
|
| 54459 |
+
"epoch": 0.43216666666666664,
|
| 54460 |
+
"grad_norm": 0.09326171875,
|
| 54461 |
+
"learning_rate": 0.0004988410229290743,
|
| 54462 |
+
"loss": 2.0294,
|
| 54463 |
+
"step": 7779
|
| 54464 |
+
},
|
| 54465 |
+
{
|
| 54466 |
+
"epoch": 0.43222222222222223,
|
| 54467 |
+
"grad_norm": 0.09326171875,
|
| 54468 |
+
"learning_rate": 0.0004987719925690094,
|
| 54469 |
+
"loss": 2.034,
|
| 54470 |
+
"step": 7780
|
| 54471 |
+
},
|
| 54472 |
+
{
|
| 54473 |
+
"epoch": 0.43227777777777776,
|
| 54474 |
+
"grad_norm": 0.09375,
|
| 54475 |
+
"learning_rate": 0.0004987029590761174,
|
| 54476 |
+
"loss": 2.0997,
|
| 54477 |
+
"step": 7781
|
| 54478 |
+
},
|
| 54479 |
+
{
|
| 54480 |
+
"epoch": 0.43233333333333335,
|
| 54481 |
+
"grad_norm": 0.09375,
|
| 54482 |
+
"learning_rate": 0.0004986339224525883,
|
| 54483 |
+
"loss": 2.0777,
|
| 54484 |
+
"step": 7782
|
| 54485 |
+
},
|
| 54486 |
+
{
|
| 54487 |
+
"epoch": 0.4323888888888889,
|
| 54488 |
+
"grad_norm": 0.091796875,
|
| 54489 |
+
"learning_rate": 0.0004985648827006116,
|
| 54490 |
+
"loss": 1.9679,
|
| 54491 |
+
"step": 7783
|
| 54492 |
+
},
|
| 54493 |
+
{
|
| 54494 |
+
"epoch": 0.43244444444444446,
|
| 54495 |
+
"grad_norm": 0.09130859375,
|
| 54496 |
+
"learning_rate": 0.000498495839822377,
|
| 54497 |
+
"loss": 2.0812,
|
| 54498 |
+
"step": 7784
|
| 54499 |
+
},
|
| 54500 |
+
{
|
| 54501 |
+
"epoch": 0.4325,
|
| 54502 |
+
"grad_norm": 0.09228515625,
|
| 54503 |
+
"learning_rate": 0.0004984267938200743,
|
| 54504 |
+
"loss": 2.0561,
|
| 54505 |
+
"step": 7785
|
| 54506 |
+
},
|
| 54507 |
+
{
|
| 54508 |
+
"epoch": 0.4325555555555556,
|
| 54509 |
+
"grad_norm": 0.09130859375,
|
| 54510 |
+
"learning_rate": 0.0004983577446958938,
|
| 54511 |
+
"loss": 2.062,
|
| 54512 |
+
"step": 7786
|
| 54513 |
+
},
|
| 54514 |
+
{
|
| 54515 |
+
"epoch": 0.4326111111111111,
|
| 54516 |
+
"grad_norm": 0.09130859375,
|
| 54517 |
+
"learning_rate": 0.0004982886924520255,
|
| 54518 |
+
"loss": 2.0651,
|
| 54519 |
+
"step": 7787
|
| 54520 |
+
},
|
| 54521 |
+
{
|
| 54522 |
+
"epoch": 0.43266666666666664,
|
| 54523 |
+
"grad_norm": 0.09326171875,
|
| 54524 |
+
"learning_rate": 0.0004982196370906594,
|
| 54525 |
+
"loss": 2.0434,
|
| 54526 |
+
"step": 7788
|
| 54527 |
+
},
|
| 54528 |
+
{
|
| 54529 |
+
"epoch": 0.43272222222222223,
|
| 54530 |
+
"grad_norm": 0.0947265625,
|
| 54531 |
+
"learning_rate": 0.000498150578613986,
|
| 54532 |
+
"loss": 2.0778,
|
| 54533 |
+
"step": 7789
|
| 54534 |
+
},
|
| 54535 |
+
{
|
| 54536 |
+
"epoch": 0.43277777777777776,
|
| 54537 |
+
"grad_norm": 0.08837890625,
|
| 54538 |
+
"learning_rate": 0.0004980815170241954,
|
| 54539 |
+
"loss": 1.9802,
|
| 54540 |
+
"step": 7790
|
| 54541 |
+
},
|
| 54542 |
+
{
|
| 54543 |
+
"epoch": 0.43283333333333335,
|
| 54544 |
+
"grad_norm": 0.0966796875,
|
| 54545 |
+
"learning_rate": 0.0004980124523234783,
|
| 54546 |
+
"loss": 2.0992,
|
| 54547 |
+
"step": 7791
|
| 54548 |
+
},
|
| 54549 |
+
{
|
| 54550 |
+
"epoch": 0.4328888888888889,
|
| 54551 |
+
"grad_norm": 0.0947265625,
|
| 54552 |
+
"learning_rate": 0.0004979433845140253,
|
| 54553 |
+
"loss": 2.0836,
|
| 54554 |
+
"step": 7792
|
| 54555 |
+
},
|
| 54556 |
+
{
|
| 54557 |
+
"epoch": 0.43294444444444447,
|
| 54558 |
+
"grad_norm": 0.08984375,
|
| 54559 |
+
"learning_rate": 0.000497874313598027,
|
| 54560 |
+
"loss": 2.0073,
|
| 54561 |
+
"step": 7793
|
| 54562 |
+
},
|
| 54563 |
+
{
|
| 54564 |
+
"epoch": 0.433,
|
| 54565 |
+
"grad_norm": 0.09521484375,
|
| 54566 |
+
"learning_rate": 0.000497805239577674,
|
| 54567 |
+
"loss": 2.0751,
|
| 54568 |
+
"step": 7794
|
| 54569 |
+
},
|
| 54570 |
+
{
|
| 54571 |
+
"epoch": 0.4330555555555556,
|
| 54572 |
+
"grad_norm": 0.09228515625,
|
| 54573 |
+
"learning_rate": 0.0004977361624551576,
|
| 54574 |
+
"loss": 2.0634,
|
| 54575 |
+
"step": 7795
|
| 54576 |
+
},
|
| 54577 |
+
{
|
| 54578 |
+
"epoch": 0.4331111111111111,
|
| 54579 |
+
"grad_norm": 0.09375,
|
| 54580 |
+
"learning_rate": 0.0004976670822326685,
|
| 54581 |
+
"loss": 2.0677,
|
| 54582 |
+
"step": 7796
|
| 54583 |
+
},
|
| 54584 |
+
{
|
| 54585 |
+
"epoch": 0.43316666666666664,
|
| 54586 |
+
"grad_norm": 0.0908203125,
|
| 54587 |
+
"learning_rate": 0.0004975979989123977,
|
| 54588 |
+
"loss": 1.9718,
|
| 54589 |
+
"step": 7797
|
| 54590 |
+
},
|
| 54591 |
+
{
|
| 54592 |
+
"epoch": 0.43322222222222223,
|
| 54593 |
+
"grad_norm": 0.0927734375,
|
| 54594 |
+
"learning_rate": 0.0004975289124965366,
|
| 54595 |
+
"loss": 2.0914,
|
| 54596 |
+
"step": 7798
|
| 54597 |
+
},
|
| 54598 |
+
{
|
| 54599 |
+
"epoch": 0.43327777777777776,
|
| 54600 |
+
"grad_norm": 0.09423828125,
|
| 54601 |
+
"learning_rate": 0.0004974598229872761,
|
| 54602 |
+
"loss": 2.0648,
|
| 54603 |
+
"step": 7799
|
| 54604 |
+
},
|
| 54605 |
+
{
|
| 54606 |
+
"epoch": 0.43333333333333335,
|
| 54607 |
+
"grad_norm": 0.0927734375,
|
| 54608 |
+
"learning_rate": 0.000497390730386808,
|
| 54609 |
+
"loss": 2.0361,
|
| 54610 |
+
"step": 7800
|
| 54611 |
}
|
| 54612 |
],
|
| 54613 |
"logging_steps": 1,
|
|
|
|
| 54627 |
"attributes": {}
|
| 54628 |
}
|
| 54629 |
},
|
| 54630 |
+
"total_flos": 4.113968951072259e+18,
|
| 54631 |
"train_batch_size": 8,
|
| 54632 |
"trial_name": null,
|
| 54633 |
"trial_params": null
|