Training in progress, step 956, 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 2934397032
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:58959cc122ef45ef7e3281ae82d05b84b00be9bf4513c53b7413f367166734e8
|
| 3 |
size 2934397032
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1474334008
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:98c4c34aa637be6fb4c1285964ab8971571b63e53f81529b2ec831e6afd58587
|
| 3 |
size 1474334008
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 9785
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:152e8c4eec5f7df7c408a099d31d440534d55833d85518cda8dfda4a88017624
|
| 3 |
size 9785
|
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,
|
|
@@ -1681,6 +1681,1679 @@
|
|
| 1681 |
"learning_rate": 0.001,
|
| 1682 |
"loss": 1.9332520961761475,
|
| 1683 |
"step": 478
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1684 |
}
|
| 1685 |
],
|
| 1686 |
"logging_steps": 2,
|
|
@@ -1700,7 +3373,7 @@
|
|
| 1700 |
"attributes": {}
|
| 1701 |
}
|
| 1702 |
},
|
| 1703 |
-
"total_flos":
|
| 1704 |
"train_batch_size": 1,
|
| 1705 |
"trial_name": null,
|
| 1706 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.4000209215963178,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 956,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1681 |
"learning_rate": 0.001,
|
| 1682 |
"loss": 1.9332520961761475,
|
| 1683 |
"step": 478
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.20084732465087085,
|
| 1687 |
+
"grad_norm": 0.1926761120557785,
|
| 1688 |
+
"learning_rate": 0.001,
|
| 1689 |
+
"loss": 2.089540481567383,
|
| 1690 |
+
"step": 480
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.20168418850358283,
|
| 1694 |
+
"grad_norm": 0.19629117846488953,
|
| 1695 |
+
"learning_rate": 0.001,
|
| 1696 |
+
"loss": 2.1012043952941895,
|
| 1697 |
+
"step": 482
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.20252105235629478,
|
| 1701 |
+
"grad_norm": 0.18073342740535736,
|
| 1702 |
+
"learning_rate": 0.001,
|
| 1703 |
+
"loss": 2.1839537620544434,
|
| 1704 |
+
"step": 484
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.20335791620900676,
|
| 1708 |
+
"grad_norm": 0.18920746445655823,
|
| 1709 |
+
"learning_rate": 0.001,
|
| 1710 |
+
"loss": 2.149573802947998,
|
| 1711 |
+
"step": 486
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.2041947800617187,
|
| 1715 |
+
"grad_norm": 0.16358095407485962,
|
| 1716 |
+
"learning_rate": 0.001,
|
| 1717 |
+
"loss": 2.1348447799682617,
|
| 1718 |
+
"step": 488
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.20503164391443068,
|
| 1722 |
+
"grad_norm": 0.1463562697172165,
|
| 1723 |
+
"learning_rate": 0.001,
|
| 1724 |
+
"loss": 2.136003017425537,
|
| 1725 |
+
"step": 490
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.20586850776714263,
|
| 1729 |
+
"grad_norm": 0.15765909850597382,
|
| 1730 |
+
"learning_rate": 0.001,
|
| 1731 |
+
"loss": 2.055952310562134,
|
| 1732 |
+
"step": 492
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.20670537161985458,
|
| 1736 |
+
"grad_norm": 0.15273720026016235,
|
| 1737 |
+
"learning_rate": 0.001,
|
| 1738 |
+
"loss": 2.1596319675445557,
|
| 1739 |
+
"step": 494
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.20754223547256656,
|
| 1743 |
+
"grad_norm": 0.17214162647724152,
|
| 1744 |
+
"learning_rate": 0.001,
|
| 1745 |
+
"loss": 2.2181458473205566,
|
| 1746 |
+
"step": 496
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.2083790993252785,
|
| 1750 |
+
"grad_norm": 0.15340030193328857,
|
| 1751 |
+
"learning_rate": 0.001,
|
| 1752 |
+
"loss": 2.064846992492676,
|
| 1753 |
+
"step": 498
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.2092159631779905,
|
| 1757 |
+
"grad_norm": 0.2015158087015152,
|
| 1758 |
+
"learning_rate": 0.001,
|
| 1759 |
+
"loss": 2.139336109161377,
|
| 1760 |
+
"step": 500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.21005282703070244,
|
| 1764 |
+
"grad_norm": 0.18293273448944092,
|
| 1765 |
+
"learning_rate": 0.001,
|
| 1766 |
+
"loss": 2.13197660446167,
|
| 1767 |
+
"step": 502
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.21088969088341442,
|
| 1771 |
+
"grad_norm": 0.13791531324386597,
|
| 1772 |
+
"learning_rate": 0.001,
|
| 1773 |
+
"loss": 2.3071653842926025,
|
| 1774 |
+
"step": 504
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.21172655473612637,
|
| 1778 |
+
"grad_norm": 0.1721826046705246,
|
| 1779 |
+
"learning_rate": 0.001,
|
| 1780 |
+
"loss": 2.104956865310669,
|
| 1781 |
+
"step": 506
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.21256341858883832,
|
| 1785 |
+
"grad_norm": 0.18442171812057495,
|
| 1786 |
+
"learning_rate": 0.001,
|
| 1787 |
+
"loss": 2.0128395557403564,
|
| 1788 |
+
"step": 508
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.2134002824415503,
|
| 1792 |
+
"grad_norm": 0.14987221360206604,
|
| 1793 |
+
"learning_rate": 0.001,
|
| 1794 |
+
"loss": 2.206230878829956,
|
| 1795 |
+
"step": 510
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.21423714629426224,
|
| 1799 |
+
"grad_norm": 0.14149565994739532,
|
| 1800 |
+
"learning_rate": 0.001,
|
| 1801 |
+
"loss": 2.0205750465393066,
|
| 1802 |
+
"step": 512
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 0.21507401014697422,
|
| 1806 |
+
"grad_norm": 0.1332736611366272,
|
| 1807 |
+
"learning_rate": 0.001,
|
| 1808 |
+
"loss": 2.0022032260894775,
|
| 1809 |
+
"step": 514
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 0.21591087399968617,
|
| 1813 |
+
"grad_norm": 0.20089702308177948,
|
| 1814 |
+
"learning_rate": 0.001,
|
| 1815 |
+
"loss": 1.882354736328125,
|
| 1816 |
+
"step": 516
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 0.21674773785239815,
|
| 1820 |
+
"grad_norm": 0.1603361815214157,
|
| 1821 |
+
"learning_rate": 0.001,
|
| 1822 |
+
"loss": 2.14081072807312,
|
| 1823 |
+
"step": 518
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 0.2175846017051101,
|
| 1827 |
+
"grad_norm": 0.1853979080915451,
|
| 1828 |
+
"learning_rate": 0.001,
|
| 1829 |
+
"loss": 2.075469493865967,
|
| 1830 |
+
"step": 520
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 0.21842146555782208,
|
| 1834 |
+
"grad_norm": 0.20383664965629578,
|
| 1835 |
+
"learning_rate": 0.001,
|
| 1836 |
+
"loss": 2.151334524154663,
|
| 1837 |
+
"step": 522
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 0.21925832941053403,
|
| 1841 |
+
"grad_norm": 0.17710188031196594,
|
| 1842 |
+
"learning_rate": 0.001,
|
| 1843 |
+
"loss": 2.0214757919311523,
|
| 1844 |
+
"step": 524
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 0.22009519326324598,
|
| 1848 |
+
"grad_norm": 0.19047263264656067,
|
| 1849 |
+
"learning_rate": 0.001,
|
| 1850 |
+
"loss": 2.008554220199585,
|
| 1851 |
+
"step": 526
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 0.22093205711595795,
|
| 1855 |
+
"grad_norm": 0.2155023068189621,
|
| 1856 |
+
"learning_rate": 0.001,
|
| 1857 |
+
"loss": 2.1172423362731934,
|
| 1858 |
+
"step": 528
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 0.2217689209686699,
|
| 1862 |
+
"grad_norm": 0.22524666786193848,
|
| 1863 |
+
"learning_rate": 0.001,
|
| 1864 |
+
"loss": 2.026920795440674,
|
| 1865 |
+
"step": 530
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 0.22260578482138188,
|
| 1869 |
+
"grad_norm": 0.21426229178905487,
|
| 1870 |
+
"learning_rate": 0.001,
|
| 1871 |
+
"loss": 1.9445011615753174,
|
| 1872 |
+
"step": 532
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 0.22344264867409383,
|
| 1876 |
+
"grad_norm": 0.14955450594425201,
|
| 1877 |
+
"learning_rate": 0.001,
|
| 1878 |
+
"loss": 2.127941846847534,
|
| 1879 |
+
"step": 534
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 0.2242795125268058,
|
| 1883 |
+
"grad_norm": 0.26866406202316284,
|
| 1884 |
+
"learning_rate": 0.001,
|
| 1885 |
+
"loss": 2.163763999938965,
|
| 1886 |
+
"step": 536
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 0.22511637637951776,
|
| 1890 |
+
"grad_norm": 0.17179425060749054,
|
| 1891 |
+
"learning_rate": 0.001,
|
| 1892 |
+
"loss": 2.2329812049865723,
|
| 1893 |
+
"step": 538
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 0.2259532402322297,
|
| 1897 |
+
"grad_norm": 0.17996813356876373,
|
| 1898 |
+
"learning_rate": 0.001,
|
| 1899 |
+
"loss": 2.159874439239502,
|
| 1900 |
+
"step": 540
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 0.22679010408494168,
|
| 1904 |
+
"grad_norm": 0.1732066124677658,
|
| 1905 |
+
"learning_rate": 0.001,
|
| 1906 |
+
"loss": 2.158291816711426,
|
| 1907 |
+
"step": 542
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 0.22762696793765363,
|
| 1911 |
+
"grad_norm": 0.19226931035518646,
|
| 1912 |
+
"learning_rate": 0.001,
|
| 1913 |
+
"loss": 2.0978221893310547,
|
| 1914 |
+
"step": 544
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 0.2284638317903656,
|
| 1918 |
+
"grad_norm": 0.15011784434318542,
|
| 1919 |
+
"learning_rate": 0.001,
|
| 1920 |
+
"loss": 2.1564321517944336,
|
| 1921 |
+
"step": 546
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 0.22930069564307756,
|
| 1925 |
+
"grad_norm": 0.15603330731391907,
|
| 1926 |
+
"learning_rate": 0.001,
|
| 1927 |
+
"loss": 2.1928114891052246,
|
| 1928 |
+
"step": 548
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 0.23013755949578954,
|
| 1932 |
+
"grad_norm": 0.22185857594013214,
|
| 1933 |
+
"learning_rate": 0.001,
|
| 1934 |
+
"loss": 2.0062198638916016,
|
| 1935 |
+
"step": 550
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 0.2309744233485015,
|
| 1939 |
+
"grad_norm": 0.15651273727416992,
|
| 1940 |
+
"learning_rate": 0.001,
|
| 1941 |
+
"loss": 2.171813488006592,
|
| 1942 |
+
"step": 552
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 0.23181128720121344,
|
| 1946 |
+
"grad_norm": 0.2166646420955658,
|
| 1947 |
+
"learning_rate": 0.001,
|
| 1948 |
+
"loss": 2.009000778198242,
|
| 1949 |
+
"step": 554
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 0.23264815105392542,
|
| 1953 |
+
"grad_norm": 0.2018618881702423,
|
| 1954 |
+
"learning_rate": 0.001,
|
| 1955 |
+
"loss": 2.0537290573120117,
|
| 1956 |
+
"step": 556
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 0.23348501490663737,
|
| 1960 |
+
"grad_norm": 0.17080949246883392,
|
| 1961 |
+
"learning_rate": 0.001,
|
| 1962 |
+
"loss": 2.0336785316467285,
|
| 1963 |
+
"step": 558
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 0.23432187875934934,
|
| 1967 |
+
"grad_norm": 0.17523957788944244,
|
| 1968 |
+
"learning_rate": 0.001,
|
| 1969 |
+
"loss": 2.074345350265503,
|
| 1970 |
+
"step": 560
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 0.2351587426120613,
|
| 1974 |
+
"grad_norm": 0.1538330614566803,
|
| 1975 |
+
"learning_rate": 0.001,
|
| 1976 |
+
"loss": 2.040306329727173,
|
| 1977 |
+
"step": 562
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 0.23599560646477327,
|
| 1981 |
+
"grad_norm": 0.14634568989276886,
|
| 1982 |
+
"learning_rate": 0.001,
|
| 1983 |
+
"loss": 2.1243209838867188,
|
| 1984 |
+
"step": 564
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 0.23683247031748522,
|
| 1988 |
+
"grad_norm": 0.1562218964099884,
|
| 1989 |
+
"learning_rate": 0.001,
|
| 1990 |
+
"loss": 2.043984889984131,
|
| 1991 |
+
"step": 566
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 0.2376693341701972,
|
| 1995 |
+
"grad_norm": 0.22568483650684357,
|
| 1996 |
+
"learning_rate": 0.001,
|
| 1997 |
+
"loss": 2.0602188110351562,
|
| 1998 |
+
"step": 568
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 0.23850619802290915,
|
| 2002 |
+
"grad_norm": 0.14369411766529083,
|
| 2003 |
+
"learning_rate": 0.001,
|
| 2004 |
+
"loss": 2.0794129371643066,
|
| 2005 |
+
"step": 570
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 0.2393430618756211,
|
| 2009 |
+
"grad_norm": 0.1667439341545105,
|
| 2010 |
+
"learning_rate": 0.001,
|
| 2011 |
+
"loss": 2.101592540740967,
|
| 2012 |
+
"step": 572
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 0.24017992572833308,
|
| 2016 |
+
"grad_norm": 0.11787199974060059,
|
| 2017 |
+
"learning_rate": 0.001,
|
| 2018 |
+
"loss": 2.067819118499756,
|
| 2019 |
+
"step": 574
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.24101678958104503,
|
| 2023 |
+
"grad_norm": 0.15923157334327698,
|
| 2024 |
+
"learning_rate": 0.001,
|
| 2025 |
+
"loss": 2.0479063987731934,
|
| 2026 |
+
"step": 576
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 0.241853653433757,
|
| 2030 |
+
"grad_norm": 0.2317262589931488,
|
| 2031 |
+
"learning_rate": 0.001,
|
| 2032 |
+
"loss": 1.8864357471466064,
|
| 2033 |
+
"step": 578
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 0.24269051728646895,
|
| 2037 |
+
"grad_norm": 0.22314757108688354,
|
| 2038 |
+
"learning_rate": 0.001,
|
| 2039 |
+
"loss": 2.1488709449768066,
|
| 2040 |
+
"step": 580
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 0.24352738113918093,
|
| 2044 |
+
"grad_norm": 0.14596961438655853,
|
| 2045 |
+
"learning_rate": 0.001,
|
| 2046 |
+
"loss": 2.0879902839660645,
|
| 2047 |
+
"step": 582
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 0.24436424499189288,
|
| 2051 |
+
"grad_norm": 0.17858095467090607,
|
| 2052 |
+
"learning_rate": 0.001,
|
| 2053 |
+
"loss": 2.049140691757202,
|
| 2054 |
+
"step": 584
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 0.24520110884460483,
|
| 2058 |
+
"grad_norm": 0.26916590332984924,
|
| 2059 |
+
"learning_rate": 0.001,
|
| 2060 |
+
"loss": 2.07326602935791,
|
| 2061 |
+
"step": 586
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 0.2460379726973168,
|
| 2065 |
+
"grad_norm": 0.16663160920143127,
|
| 2066 |
+
"learning_rate": 0.001,
|
| 2067 |
+
"loss": 1.9899840354919434,
|
| 2068 |
+
"step": 588
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 0.24687483655002876,
|
| 2072 |
+
"grad_norm": 0.24591751396656036,
|
| 2073 |
+
"learning_rate": 0.001,
|
| 2074 |
+
"loss": 2.017153263092041,
|
| 2075 |
+
"step": 590
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 0.24771170040274074,
|
| 2079 |
+
"grad_norm": 0.1736748367547989,
|
| 2080 |
+
"learning_rate": 0.001,
|
| 2081 |
+
"loss": 2.1265370845794678,
|
| 2082 |
+
"step": 592
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 0.2485485642554527,
|
| 2086 |
+
"grad_norm": 0.16665787994861603,
|
| 2087 |
+
"learning_rate": 0.001,
|
| 2088 |
+
"loss": 2.1271588802337646,
|
| 2089 |
+
"step": 594
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 0.24938542810816466,
|
| 2093 |
+
"grad_norm": 0.17151160538196564,
|
| 2094 |
+
"learning_rate": 0.001,
|
| 2095 |
+
"loss": 2.1545591354370117,
|
| 2096 |
+
"step": 596
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.25022229196087664,
|
| 2100 |
+
"grad_norm": 0.23257796466350555,
|
| 2101 |
+
"learning_rate": 0.001,
|
| 2102 |
+
"loss": 1.9667472839355469,
|
| 2103 |
+
"step": 598
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 0.25105915581358856,
|
| 2107 |
+
"grad_norm": 0.20128676295280457,
|
| 2108 |
+
"learning_rate": 0.001,
|
| 2109 |
+
"loss": 2.1109392642974854,
|
| 2110 |
+
"step": 600
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 0.25189601966630054,
|
| 2114 |
+
"grad_norm": 0.22417353093624115,
|
| 2115 |
+
"learning_rate": 0.001,
|
| 2116 |
+
"loss": 1.9775571823120117,
|
| 2117 |
+
"step": 602
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 0.2527328835190125,
|
| 2121 |
+
"grad_norm": 0.20211748778820038,
|
| 2122 |
+
"learning_rate": 0.001,
|
| 2123 |
+
"loss": 2.028385877609253,
|
| 2124 |
+
"step": 604
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 0.25356974737172444,
|
| 2128 |
+
"grad_norm": 0.17851772904396057,
|
| 2129 |
+
"learning_rate": 0.001,
|
| 2130 |
+
"loss": 2.0029194355010986,
|
| 2131 |
+
"step": 606
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 0.2544066112244364,
|
| 2135 |
+
"grad_norm": 0.17671839892864227,
|
| 2136 |
+
"learning_rate": 0.001,
|
| 2137 |
+
"loss": 1.8936843872070312,
|
| 2138 |
+
"step": 608
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 0.2552434750771484,
|
| 2142 |
+
"grad_norm": 0.19212132692337036,
|
| 2143 |
+
"learning_rate": 0.001,
|
| 2144 |
+
"loss": 2.0227270126342773,
|
| 2145 |
+
"step": 610
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 0.2560803389298604,
|
| 2149 |
+
"grad_norm": 0.1454077512025833,
|
| 2150 |
+
"learning_rate": 0.001,
|
| 2151 |
+
"loss": 2.0801706314086914,
|
| 2152 |
+
"step": 612
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 0.2569172027825723,
|
| 2156 |
+
"grad_norm": 0.16919741034507751,
|
| 2157 |
+
"learning_rate": 0.001,
|
| 2158 |
+
"loss": 2.0349369049072266,
|
| 2159 |
+
"step": 614
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 0.2577540666352843,
|
| 2163 |
+
"grad_norm": 0.20194627344608307,
|
| 2164 |
+
"learning_rate": 0.001,
|
| 2165 |
+
"loss": 2.0809738636016846,
|
| 2166 |
+
"step": 616
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 0.25859093048799625,
|
| 2170 |
+
"grad_norm": 0.16600646078586578,
|
| 2171 |
+
"learning_rate": 0.001,
|
| 2172 |
+
"loss": 2.086526870727539,
|
| 2173 |
+
"step": 618
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 0.2594277943407082,
|
| 2177 |
+
"grad_norm": 0.1607326716184616,
|
| 2178 |
+
"learning_rate": 0.001,
|
| 2179 |
+
"loss": 2.010438919067383,
|
| 2180 |
+
"step": 620
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 0.26026465819342015,
|
| 2184 |
+
"grad_norm": 0.1957024782896042,
|
| 2185 |
+
"learning_rate": 0.001,
|
| 2186 |
+
"loss": 2.181964874267578,
|
| 2187 |
+
"step": 622
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 0.26110152204613213,
|
| 2191 |
+
"grad_norm": 0.1674378514289856,
|
| 2192 |
+
"learning_rate": 0.001,
|
| 2193 |
+
"loss": 1.9782319068908691,
|
| 2194 |
+
"step": 624
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 0.2619383858988441,
|
| 2198 |
+
"grad_norm": 0.15856251120567322,
|
| 2199 |
+
"learning_rate": 0.001,
|
| 2200 |
+
"loss": 1.9276220798492432,
|
| 2201 |
+
"step": 626
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.26277524975155603,
|
| 2205 |
+
"grad_norm": 0.17853853106498718,
|
| 2206 |
+
"learning_rate": 0.001,
|
| 2207 |
+
"loss": 2.1943626403808594,
|
| 2208 |
+
"step": 628
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.263612113604268,
|
| 2212 |
+
"grad_norm": 0.15641216933727264,
|
| 2213 |
+
"learning_rate": 0.001,
|
| 2214 |
+
"loss": 1.9760797023773193,
|
| 2215 |
+
"step": 630
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.26444897745698,
|
| 2219 |
+
"grad_norm": 0.1873551905155182,
|
| 2220 |
+
"learning_rate": 0.001,
|
| 2221 |
+
"loss": 2.114103078842163,
|
| 2222 |
+
"step": 632
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.2652858413096919,
|
| 2226 |
+
"grad_norm": 0.161077082157135,
|
| 2227 |
+
"learning_rate": 0.001,
|
| 2228 |
+
"loss": 2.1406664848327637,
|
| 2229 |
+
"step": 634
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.2661227051624039,
|
| 2233 |
+
"grad_norm": 0.1653231978416443,
|
| 2234 |
+
"learning_rate": 0.001,
|
| 2235 |
+
"loss": 1.916658878326416,
|
| 2236 |
+
"step": 636
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 0.26695956901511586,
|
| 2240 |
+
"grad_norm": 0.1952139437198639,
|
| 2241 |
+
"learning_rate": 0.001,
|
| 2242 |
+
"loss": 2.056473970413208,
|
| 2243 |
+
"step": 638
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.26779643286782784,
|
| 2247 |
+
"grad_norm": 0.20149773359298706,
|
| 2248 |
+
"learning_rate": 0.001,
|
| 2249 |
+
"loss": 1.865290641784668,
|
| 2250 |
+
"step": 640
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 0.26863329672053976,
|
| 2254 |
+
"grad_norm": 0.13062399625778198,
|
| 2255 |
+
"learning_rate": 0.001,
|
| 2256 |
+
"loss": 2.161708354949951,
|
| 2257 |
+
"step": 642
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.26947016057325174,
|
| 2261 |
+
"grad_norm": 0.1794498860836029,
|
| 2262 |
+
"learning_rate": 0.001,
|
| 2263 |
+
"loss": 2.113748788833618,
|
| 2264 |
+
"step": 644
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 0.2703070244259637,
|
| 2268 |
+
"grad_norm": 0.15839435160160065,
|
| 2269 |
+
"learning_rate": 0.001,
|
| 2270 |
+
"loss": 2.032397747039795,
|
| 2271 |
+
"step": 646
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.27114388827867564,
|
| 2275 |
+
"grad_norm": 0.17412367463111877,
|
| 2276 |
+
"learning_rate": 0.001,
|
| 2277 |
+
"loss": 1.953669548034668,
|
| 2278 |
+
"step": 648
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 0.2719807521313876,
|
| 2282 |
+
"grad_norm": 0.22392594814300537,
|
| 2283 |
+
"learning_rate": 0.001,
|
| 2284 |
+
"loss": 2.0704774856567383,
|
| 2285 |
+
"step": 650
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 0.2728176159840996,
|
| 2289 |
+
"grad_norm": 0.21520493924617767,
|
| 2290 |
+
"learning_rate": 0.001,
|
| 2291 |
+
"loss": 2.0462725162506104,
|
| 2292 |
+
"step": 652
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 0.27365447983681157,
|
| 2296 |
+
"grad_norm": 0.18177847564220428,
|
| 2297 |
+
"learning_rate": 0.001,
|
| 2298 |
+
"loss": 1.9804809093475342,
|
| 2299 |
+
"step": 654
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.2744913436895235,
|
| 2303 |
+
"grad_norm": 0.12944908440113068,
|
| 2304 |
+
"learning_rate": 0.001,
|
| 2305 |
+
"loss": 2.1314516067504883,
|
| 2306 |
+
"step": 656
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.27532820754223547,
|
| 2310 |
+
"grad_norm": 0.15082746744155884,
|
| 2311 |
+
"learning_rate": 0.001,
|
| 2312 |
+
"loss": 1.9282715320587158,
|
| 2313 |
+
"step": 658
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.27616507139494745,
|
| 2317 |
+
"grad_norm": 0.22981098294258118,
|
| 2318 |
+
"learning_rate": 0.001,
|
| 2319 |
+
"loss": 1.9650275707244873,
|
| 2320 |
+
"step": 660
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 0.27700193524765937,
|
| 2324 |
+
"grad_norm": 0.18900924921035767,
|
| 2325 |
+
"learning_rate": 0.001,
|
| 2326 |
+
"loss": 1.933212399482727,
|
| 2327 |
+
"step": 662
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 0.27783879910037135,
|
| 2331 |
+
"grad_norm": 0.14509092271327972,
|
| 2332 |
+
"learning_rate": 0.001,
|
| 2333 |
+
"loss": 2.0642037391662598,
|
| 2334 |
+
"step": 664
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 0.2786756629530833,
|
| 2338 |
+
"grad_norm": 0.13011862337589264,
|
| 2339 |
+
"learning_rate": 0.001,
|
| 2340 |
+
"loss": 1.9835015535354614,
|
| 2341 |
+
"step": 666
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.2795125268057953,
|
| 2345 |
+
"grad_norm": 0.1856154203414917,
|
| 2346 |
+
"learning_rate": 0.001,
|
| 2347 |
+
"loss": 2.056840419769287,
|
| 2348 |
+
"step": 668
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 0.2803493906585072,
|
| 2352 |
+
"grad_norm": 0.17407356202602386,
|
| 2353 |
+
"learning_rate": 0.001,
|
| 2354 |
+
"loss": 2.089240550994873,
|
| 2355 |
+
"step": 670
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 0.2811862545112192,
|
| 2359 |
+
"grad_norm": 0.15162459015846252,
|
| 2360 |
+
"learning_rate": 0.001,
|
| 2361 |
+
"loss": 1.8737280368804932,
|
| 2362 |
+
"step": 672
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 0.2820231183639312,
|
| 2366 |
+
"grad_norm": 0.181808203458786,
|
| 2367 |
+
"learning_rate": 0.001,
|
| 2368 |
+
"loss": 1.9263813495635986,
|
| 2369 |
+
"step": 674
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 0.2828599822166431,
|
| 2373 |
+
"grad_norm": 0.14570149779319763,
|
| 2374 |
+
"learning_rate": 0.001,
|
| 2375 |
+
"loss": 2.0938119888305664,
|
| 2376 |
+
"step": 676
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 0.2836968460693551,
|
| 2380 |
+
"grad_norm": 0.19376394152641296,
|
| 2381 |
+
"learning_rate": 0.001,
|
| 2382 |
+
"loss": 2.0293610095977783,
|
| 2383 |
+
"step": 678
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 0.28453370992206706,
|
| 2387 |
+
"grad_norm": 0.24984364211559296,
|
| 2388 |
+
"learning_rate": 0.001,
|
| 2389 |
+
"loss": 1.791812539100647,
|
| 2390 |
+
"step": 680
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 0.28537057377477903,
|
| 2394 |
+
"grad_norm": 0.18278868496418,
|
| 2395 |
+
"learning_rate": 0.001,
|
| 2396 |
+
"loss": 1.8272709846496582,
|
| 2397 |
+
"step": 682
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 0.28620743762749096,
|
| 2401 |
+
"grad_norm": 0.16873054206371307,
|
| 2402 |
+
"learning_rate": 0.001,
|
| 2403 |
+
"loss": 1.9225802421569824,
|
| 2404 |
+
"step": 684
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 0.28704430148020293,
|
| 2408 |
+
"grad_norm": 0.1797211617231369,
|
| 2409 |
+
"learning_rate": 0.001,
|
| 2410 |
+
"loss": 1.9720535278320312,
|
| 2411 |
+
"step": 686
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 0.2878811653329149,
|
| 2415 |
+
"grad_norm": 0.1909167766571045,
|
| 2416 |
+
"learning_rate": 0.001,
|
| 2417 |
+
"loss": 1.9139065742492676,
|
| 2418 |
+
"step": 688
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 0.2887180291856269,
|
| 2422 |
+
"grad_norm": 0.16065365076065063,
|
| 2423 |
+
"learning_rate": 0.001,
|
| 2424 |
+
"loss": 2.234241485595703,
|
| 2425 |
+
"step": 690
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 0.2895548930383388,
|
| 2429 |
+
"grad_norm": 0.17435690760612488,
|
| 2430 |
+
"learning_rate": 0.001,
|
| 2431 |
+
"loss": 2.0211985111236572,
|
| 2432 |
+
"step": 692
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 0.2903917568910508,
|
| 2436 |
+
"grad_norm": 0.1799331158399582,
|
| 2437 |
+
"learning_rate": 0.001,
|
| 2438 |
+
"loss": 1.8879200220108032,
|
| 2439 |
+
"step": 694
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 0.29122862074376277,
|
| 2443 |
+
"grad_norm": 0.16190172731876373,
|
| 2444 |
+
"learning_rate": 0.001,
|
| 2445 |
+
"loss": 2.0217933654785156,
|
| 2446 |
+
"step": 696
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 0.2920654845964747,
|
| 2450 |
+
"grad_norm": 0.15402548015117645,
|
| 2451 |
+
"learning_rate": 0.001,
|
| 2452 |
+
"loss": 2.0288939476013184,
|
| 2453 |
+
"step": 698
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 0.29290234844918667,
|
| 2457 |
+
"grad_norm": 0.2108118087053299,
|
| 2458 |
+
"learning_rate": 0.001,
|
| 2459 |
+
"loss": 2.0979666709899902,
|
| 2460 |
+
"step": 700
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 0.29373921230189864,
|
| 2464 |
+
"grad_norm": 0.18792341649532318,
|
| 2465 |
+
"learning_rate": 0.001,
|
| 2466 |
+
"loss": 2.1385817527770996,
|
| 2467 |
+
"step": 702
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 0.2945760761546106,
|
| 2471 |
+
"grad_norm": 0.19642780721187592,
|
| 2472 |
+
"learning_rate": 0.001,
|
| 2473 |
+
"loss": 2.1903281211853027,
|
| 2474 |
+
"step": 704
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 0.29541294000732254,
|
| 2478 |
+
"grad_norm": 0.15158486366271973,
|
| 2479 |
+
"learning_rate": 0.001,
|
| 2480 |
+
"loss": 1.959409236907959,
|
| 2481 |
+
"step": 706
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 0.2962498038600345,
|
| 2485 |
+
"grad_norm": 0.21778351068496704,
|
| 2486 |
+
"learning_rate": 0.001,
|
| 2487 |
+
"loss": 1.8310878276824951,
|
| 2488 |
+
"step": 708
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 0.2970866677127465,
|
| 2492 |
+
"grad_norm": 0.23621246218681335,
|
| 2493 |
+
"learning_rate": 0.001,
|
| 2494 |
+
"loss": 1.7779185771942139,
|
| 2495 |
+
"step": 710
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 0.2979235315654584,
|
| 2499 |
+
"grad_norm": 0.21618059277534485,
|
| 2500 |
+
"learning_rate": 0.001,
|
| 2501 |
+
"loss": 2.0669431686401367,
|
| 2502 |
+
"step": 712
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 0.2987603954181704,
|
| 2506 |
+
"grad_norm": 0.15448474884033203,
|
| 2507 |
+
"learning_rate": 0.001,
|
| 2508 |
+
"loss": 1.9449490308761597,
|
| 2509 |
+
"step": 714
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 0.2995972592708824,
|
| 2513 |
+
"grad_norm": 0.17804761230945587,
|
| 2514 |
+
"learning_rate": 0.001,
|
| 2515 |
+
"loss": 2.1330461502075195,
|
| 2516 |
+
"step": 716
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 0.30043412312359435,
|
| 2520 |
+
"grad_norm": 0.20111820101737976,
|
| 2521 |
+
"learning_rate": 0.001,
|
| 2522 |
+
"loss": 1.892564296722412,
|
| 2523 |
+
"step": 718
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 0.3012709869763063,
|
| 2527 |
+
"grad_norm": 0.15836556255817413,
|
| 2528 |
+
"learning_rate": 0.001,
|
| 2529 |
+
"loss": 1.9832303524017334,
|
| 2530 |
+
"step": 720
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 0.30210785082901825,
|
| 2534 |
+
"grad_norm": 0.2553568184375763,
|
| 2535 |
+
"learning_rate": 0.001,
|
| 2536 |
+
"loss": 2.0218451023101807,
|
| 2537 |
+
"step": 722
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 0.30294471468173023,
|
| 2541 |
+
"grad_norm": 0.2228379249572754,
|
| 2542 |
+
"learning_rate": 0.001,
|
| 2543 |
+
"loss": 1.815321445465088,
|
| 2544 |
+
"step": 724
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 0.30378157853444215,
|
| 2548 |
+
"grad_norm": 0.13462747633457184,
|
| 2549 |
+
"learning_rate": 0.001,
|
| 2550 |
+
"loss": 1.999401330947876,
|
| 2551 |
+
"step": 726
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 0.30461844238715413,
|
| 2555 |
+
"grad_norm": 0.15057668089866638,
|
| 2556 |
+
"learning_rate": 0.001,
|
| 2557 |
+
"loss": 1.8639626502990723,
|
| 2558 |
+
"step": 728
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 0.3054553062398661,
|
| 2562 |
+
"grad_norm": 0.12982071936130524,
|
| 2563 |
+
"learning_rate": 0.001,
|
| 2564 |
+
"loss": 1.8446266651153564,
|
| 2565 |
+
"step": 730
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 0.3062921700925781,
|
| 2569 |
+
"grad_norm": 0.1756584495306015,
|
| 2570 |
+
"learning_rate": 0.001,
|
| 2571 |
+
"loss": 2.070432662963867,
|
| 2572 |
+
"step": 732
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 0.30712903394529,
|
| 2576 |
+
"grad_norm": 0.20863129198551178,
|
| 2577 |
+
"learning_rate": 0.001,
|
| 2578 |
+
"loss": 2.210751533508301,
|
| 2579 |
+
"step": 734
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 0.307965897798002,
|
| 2583 |
+
"grad_norm": 0.16472575068473816,
|
| 2584 |
+
"learning_rate": 0.001,
|
| 2585 |
+
"loss": 1.9010889530181885,
|
| 2586 |
+
"step": 736
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 0.30880276165071396,
|
| 2590 |
+
"grad_norm": 0.1404307335615158,
|
| 2591 |
+
"learning_rate": 0.001,
|
| 2592 |
+
"loss": 2.031334638595581,
|
| 2593 |
+
"step": 738
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 0.3096396255034259,
|
| 2597 |
+
"grad_norm": 0.2110474854707718,
|
| 2598 |
+
"learning_rate": 0.001,
|
| 2599 |
+
"loss": 2.0073390007019043,
|
| 2600 |
+
"step": 740
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 0.31047648935613786,
|
| 2604 |
+
"grad_norm": 0.19802813231945038,
|
| 2605 |
+
"learning_rate": 0.001,
|
| 2606 |
+
"loss": 1.9414364099502563,
|
| 2607 |
+
"step": 742
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 0.31131335320884984,
|
| 2611 |
+
"grad_norm": 0.15902012586593628,
|
| 2612 |
+
"learning_rate": 0.001,
|
| 2613 |
+
"loss": 2.0136821269989014,
|
| 2614 |
+
"step": 744
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 0.3121502170615618,
|
| 2618 |
+
"grad_norm": 0.1590680330991745,
|
| 2619 |
+
"learning_rate": 0.001,
|
| 2620 |
+
"loss": 1.9856327772140503,
|
| 2621 |
+
"step": 746
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 0.31298708091427374,
|
| 2625 |
+
"grad_norm": 0.18464486300945282,
|
| 2626 |
+
"learning_rate": 0.001,
|
| 2627 |
+
"loss": 1.9156696796417236,
|
| 2628 |
+
"step": 748
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 0.3138239447669857,
|
| 2632 |
+
"grad_norm": 0.14798279106616974,
|
| 2633 |
+
"learning_rate": 0.001,
|
| 2634 |
+
"loss": 2.203310489654541,
|
| 2635 |
+
"step": 750
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 0.3146608086196977,
|
| 2639 |
+
"grad_norm": 0.15232564508914948,
|
| 2640 |
+
"learning_rate": 0.001,
|
| 2641 |
+
"loss": 1.9305493831634521,
|
| 2642 |
+
"step": 752
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 0.3154976724724096,
|
| 2646 |
+
"grad_norm": 0.17355622351169586,
|
| 2647 |
+
"learning_rate": 0.001,
|
| 2648 |
+
"loss": 1.9124706983566284,
|
| 2649 |
+
"step": 754
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 0.3163345363251216,
|
| 2653 |
+
"grad_norm": 0.17948277294635773,
|
| 2654 |
+
"learning_rate": 0.001,
|
| 2655 |
+
"loss": 1.9012525081634521,
|
| 2656 |
+
"step": 756
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 0.3171714001778336,
|
| 2660 |
+
"grad_norm": 0.2285899966955185,
|
| 2661 |
+
"learning_rate": 0.001,
|
| 2662 |
+
"loss": 1.7846107482910156,
|
| 2663 |
+
"step": 758
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 0.31800826403054555,
|
| 2667 |
+
"grad_norm": 0.22635407745838165,
|
| 2668 |
+
"learning_rate": 0.001,
|
| 2669 |
+
"loss": 1.95014488697052,
|
| 2670 |
+
"step": 760
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 0.3188451278832575,
|
| 2674 |
+
"grad_norm": 0.15514934062957764,
|
| 2675 |
+
"learning_rate": 0.001,
|
| 2676 |
+
"loss": 1.7931287288665771,
|
| 2677 |
+
"step": 762
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 0.31968199173596945,
|
| 2681 |
+
"grad_norm": 0.19126024842262268,
|
| 2682 |
+
"learning_rate": 0.001,
|
| 2683 |
+
"loss": 2.028442621231079,
|
| 2684 |
+
"step": 764
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 0.32051885558868143,
|
| 2688 |
+
"grad_norm": 0.17914006114006042,
|
| 2689 |
+
"learning_rate": 0.001,
|
| 2690 |
+
"loss": 2.107748031616211,
|
| 2691 |
+
"step": 766
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 0.3213557194413934,
|
| 2695 |
+
"grad_norm": 0.1412804126739502,
|
| 2696 |
+
"learning_rate": 0.001,
|
| 2697 |
+
"loss": 2.089745283126831,
|
| 2698 |
+
"step": 768
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 0.32219258329410533,
|
| 2702 |
+
"grad_norm": 0.2493792176246643,
|
| 2703 |
+
"learning_rate": 0.001,
|
| 2704 |
+
"loss": 2.0231454372406006,
|
| 2705 |
+
"step": 770
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 0.3230294471468173,
|
| 2709 |
+
"grad_norm": 0.17295268177986145,
|
| 2710 |
+
"learning_rate": 0.001,
|
| 2711 |
+
"loss": 2.1804652214050293,
|
| 2712 |
+
"step": 772
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 0.3238663109995293,
|
| 2716 |
+
"grad_norm": 0.1355276256799698,
|
| 2717 |
+
"learning_rate": 0.001,
|
| 2718 |
+
"loss": 1.9094395637512207,
|
| 2719 |
+
"step": 774
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 0.3247031748522412,
|
| 2723 |
+
"grad_norm": 0.21440669894218445,
|
| 2724 |
+
"learning_rate": 0.001,
|
| 2725 |
+
"loss": 2.178769111633301,
|
| 2726 |
+
"step": 776
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 0.3255400387049532,
|
| 2730 |
+
"grad_norm": 0.1524726301431656,
|
| 2731 |
+
"learning_rate": 0.001,
|
| 2732 |
+
"loss": 1.963311791419983,
|
| 2733 |
+
"step": 778
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 0.32637690255766516,
|
| 2737 |
+
"grad_norm": 0.14417000114917755,
|
| 2738 |
+
"learning_rate": 0.001,
|
| 2739 |
+
"loss": 1.98314368724823,
|
| 2740 |
+
"step": 780
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 0.32721376641037714,
|
| 2744 |
+
"grad_norm": 0.16447897255420685,
|
| 2745 |
+
"learning_rate": 0.001,
|
| 2746 |
+
"loss": 2.0183463096618652,
|
| 2747 |
+
"step": 782
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 0.32805063026308906,
|
| 2751 |
+
"grad_norm": 0.15285257995128632,
|
| 2752 |
+
"learning_rate": 0.001,
|
| 2753 |
+
"loss": 2.1279056072235107,
|
| 2754 |
+
"step": 784
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.32888749411580104,
|
| 2758 |
+
"grad_norm": 0.1796996146440506,
|
| 2759 |
+
"learning_rate": 0.001,
|
| 2760 |
+
"loss": 1.8257269859313965,
|
| 2761 |
+
"step": 786
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.329724357968513,
|
| 2765 |
+
"grad_norm": 0.19257581233978271,
|
| 2766 |
+
"learning_rate": 0.001,
|
| 2767 |
+
"loss": 2.0261998176574707,
|
| 2768 |
+
"step": 788
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 0.33056122182122494,
|
| 2772 |
+
"grad_norm": 0.16922463476657867,
|
| 2773 |
+
"learning_rate": 0.001,
|
| 2774 |
+
"loss": 1.9479306936264038,
|
| 2775 |
+
"step": 790
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 0.3313980856739369,
|
| 2779 |
+
"grad_norm": 0.15687814354896545,
|
| 2780 |
+
"learning_rate": 0.001,
|
| 2781 |
+
"loss": 1.752540111541748,
|
| 2782 |
+
"step": 792
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 0.3322349495266489,
|
| 2786 |
+
"grad_norm": 0.15943767130374908,
|
| 2787 |
+
"learning_rate": 0.001,
|
| 2788 |
+
"loss": 1.9209232330322266,
|
| 2789 |
+
"step": 794
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 0.33307181337936087,
|
| 2793 |
+
"grad_norm": 0.12818294763565063,
|
| 2794 |
+
"learning_rate": 0.001,
|
| 2795 |
+
"loss": 1.8480098247528076,
|
| 2796 |
+
"step": 796
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.3339086772320728,
|
| 2800 |
+
"grad_norm": 0.17472955584526062,
|
| 2801 |
+
"learning_rate": 0.001,
|
| 2802 |
+
"loss": 1.998063325881958,
|
| 2803 |
+
"step": 798
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 0.33474554108478477,
|
| 2807 |
+
"grad_norm": 0.19923201203346252,
|
| 2808 |
+
"learning_rate": 0.001,
|
| 2809 |
+
"loss": 1.8344371318817139,
|
| 2810 |
+
"step": 800
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 0.33558240493749675,
|
| 2814 |
+
"grad_norm": 0.15498565137386322,
|
| 2815 |
+
"learning_rate": 0.001,
|
| 2816 |
+
"loss": 2.0607728958129883,
|
| 2817 |
+
"step": 802
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 0.33641926879020867,
|
| 2821 |
+
"grad_norm": 0.11883684992790222,
|
| 2822 |
+
"learning_rate": 0.001,
|
| 2823 |
+
"loss": 2.0854551792144775,
|
| 2824 |
+
"step": 804
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 0.33725613264292065,
|
| 2828 |
+
"grad_norm": 0.19124728441238403,
|
| 2829 |
+
"learning_rate": 0.001,
|
| 2830 |
+
"loss": 2.0002872943878174,
|
| 2831 |
+
"step": 806
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 0.3380929964956326,
|
| 2835 |
+
"grad_norm": 0.14578793942928314,
|
| 2836 |
+
"learning_rate": 0.001,
|
| 2837 |
+
"loss": 1.9360275268554688,
|
| 2838 |
+
"step": 808
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 0.3389298603483446,
|
| 2842 |
+
"grad_norm": 0.1889970749616623,
|
| 2843 |
+
"learning_rate": 0.001,
|
| 2844 |
+
"loss": 2.085242509841919,
|
| 2845 |
+
"step": 810
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 0.3397667242010565,
|
| 2849 |
+
"grad_norm": 0.2232929915189743,
|
| 2850 |
+
"learning_rate": 0.001,
|
| 2851 |
+
"loss": 1.9264602661132812,
|
| 2852 |
+
"step": 812
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 0.3406035880537685,
|
| 2856 |
+
"grad_norm": 0.1422855406999588,
|
| 2857 |
+
"learning_rate": 0.001,
|
| 2858 |
+
"loss": 1.9356168508529663,
|
| 2859 |
+
"step": 814
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 0.3414404519064805,
|
| 2863 |
+
"grad_norm": 0.1817842572927475,
|
| 2864 |
+
"learning_rate": 0.001,
|
| 2865 |
+
"loss": 1.9961140155792236,
|
| 2866 |
+
"step": 816
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 0.3422773157591924,
|
| 2870 |
+
"grad_norm": 0.1512877196073532,
|
| 2871 |
+
"learning_rate": 0.001,
|
| 2872 |
+
"loss": 1.9533355236053467,
|
| 2873 |
+
"step": 818
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 0.3431141796119044,
|
| 2877 |
+
"grad_norm": 0.2053813636302948,
|
| 2878 |
+
"learning_rate": 0.001,
|
| 2879 |
+
"loss": 2.0831804275512695,
|
| 2880 |
+
"step": 820
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 0.34395104346461636,
|
| 2884 |
+
"grad_norm": 0.18717814981937408,
|
| 2885 |
+
"learning_rate": 0.001,
|
| 2886 |
+
"loss": 1.8606504201889038,
|
| 2887 |
+
"step": 822
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 0.34478790731732833,
|
| 2891 |
+
"grad_norm": 0.2104620635509491,
|
| 2892 |
+
"learning_rate": 0.001,
|
| 2893 |
+
"loss": 1.7364555597305298,
|
| 2894 |
+
"step": 824
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 0.34562477117004026,
|
| 2898 |
+
"grad_norm": 0.2242583930492401,
|
| 2899 |
+
"learning_rate": 0.001,
|
| 2900 |
+
"loss": 1.8400073051452637,
|
| 2901 |
+
"step": 826
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 0.34646163502275223,
|
| 2905 |
+
"grad_norm": 0.1672690212726593,
|
| 2906 |
+
"learning_rate": 0.001,
|
| 2907 |
+
"loss": 1.877899408340454,
|
| 2908 |
+
"step": 828
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 0.3472984988754642,
|
| 2912 |
+
"grad_norm": 0.1541302353143692,
|
| 2913 |
+
"learning_rate": 0.001,
|
| 2914 |
+
"loss": 2.074535608291626,
|
| 2915 |
+
"step": 830
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 0.34813536272817613,
|
| 2919 |
+
"grad_norm": 0.1765347421169281,
|
| 2920 |
+
"learning_rate": 0.001,
|
| 2921 |
+
"loss": 1.8991279602050781,
|
| 2922 |
+
"step": 832
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 0.3489722265808881,
|
| 2926 |
+
"grad_norm": 0.16678385436534882,
|
| 2927 |
+
"learning_rate": 0.001,
|
| 2928 |
+
"loss": 1.8577218055725098,
|
| 2929 |
+
"step": 834
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 0.3498090904336001,
|
| 2933 |
+
"grad_norm": 0.18591059744358063,
|
| 2934 |
+
"learning_rate": 0.001,
|
| 2935 |
+
"loss": 2.009002208709717,
|
| 2936 |
+
"step": 836
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 0.35064595428631207,
|
| 2940 |
+
"grad_norm": 0.1719459444284439,
|
| 2941 |
+
"learning_rate": 0.001,
|
| 2942 |
+
"loss": 2.012207508087158,
|
| 2943 |
+
"step": 838
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 0.351482818139024,
|
| 2947 |
+
"grad_norm": 0.20304518938064575,
|
| 2948 |
+
"learning_rate": 0.001,
|
| 2949 |
+
"loss": 1.855114221572876,
|
| 2950 |
+
"step": 840
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 0.35231968199173597,
|
| 2954 |
+
"grad_norm": 0.1599912792444229,
|
| 2955 |
+
"learning_rate": 0.001,
|
| 2956 |
+
"loss": 1.8909435272216797,
|
| 2957 |
+
"step": 842
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 0.35315654584444794,
|
| 2961 |
+
"grad_norm": 0.19980160892009735,
|
| 2962 |
+
"learning_rate": 0.001,
|
| 2963 |
+
"loss": 2.0490949153900146,
|
| 2964 |
+
"step": 844
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 0.35399340969715987,
|
| 2968 |
+
"grad_norm": 0.17447344958782196,
|
| 2969 |
+
"learning_rate": 0.001,
|
| 2970 |
+
"loss": 1.7521772384643555,
|
| 2971 |
+
"step": 846
|
| 2972 |
+
},
|
| 2973 |
+
{
|
| 2974 |
+
"epoch": 0.35483027354987184,
|
| 2975 |
+
"grad_norm": 0.18320021033287048,
|
| 2976 |
+
"learning_rate": 0.001,
|
| 2977 |
+
"loss": 1.9941444396972656,
|
| 2978 |
+
"step": 848
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"epoch": 0.3556671374025838,
|
| 2982 |
+
"grad_norm": 0.1856023073196411,
|
| 2983 |
+
"learning_rate": 0.001,
|
| 2984 |
+
"loss": 1.864702582359314,
|
| 2985 |
+
"step": 850
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 0.3565040012552958,
|
| 2989 |
+
"grad_norm": 0.18598812818527222,
|
| 2990 |
+
"learning_rate": 0.001,
|
| 2991 |
+
"loss": 2.048583507537842,
|
| 2992 |
+
"step": 852
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 0.3573408651080077,
|
| 2996 |
+
"grad_norm": 0.1361086517572403,
|
| 2997 |
+
"learning_rate": 0.001,
|
| 2998 |
+
"loss": 2.0076942443847656,
|
| 2999 |
+
"step": 854
|
| 3000 |
+
},
|
| 3001 |
+
{
|
| 3002 |
+
"epoch": 0.3581777289607197,
|
| 3003 |
+
"grad_norm": 0.16787421703338623,
|
| 3004 |
+
"learning_rate": 0.001,
|
| 3005 |
+
"loss": 1.9802899360656738,
|
| 3006 |
+
"step": 856
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 0.3590145928134317,
|
| 3010 |
+
"grad_norm": 0.14613385498523712,
|
| 3011 |
+
"learning_rate": 0.001,
|
| 3012 |
+
"loss": 1.9639887809753418,
|
| 3013 |
+
"step": 858
|
| 3014 |
+
},
|
| 3015 |
+
{
|
| 3016 |
+
"epoch": 0.35985145666614365,
|
| 3017 |
+
"grad_norm": 0.20427103340625763,
|
| 3018 |
+
"learning_rate": 0.001,
|
| 3019 |
+
"loss": 1.9706238508224487,
|
| 3020 |
+
"step": 860
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 0.3606883205188556,
|
| 3024 |
+
"grad_norm": 0.18077662587165833,
|
| 3025 |
+
"learning_rate": 0.001,
|
| 3026 |
+
"loss": 2.0072827339172363,
|
| 3027 |
+
"step": 862
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 0.36152518437156755,
|
| 3031 |
+
"grad_norm": 0.24144840240478516,
|
| 3032 |
+
"learning_rate": 0.001,
|
| 3033 |
+
"loss": 1.7195979356765747,
|
| 3034 |
+
"step": 864
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 0.36236204822427953,
|
| 3038 |
+
"grad_norm": 0.1936212033033371,
|
| 3039 |
+
"learning_rate": 0.001,
|
| 3040 |
+
"loss": 1.9409396648406982,
|
| 3041 |
+
"step": 866
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 0.36319891207699145,
|
| 3045 |
+
"grad_norm": 0.17916715145111084,
|
| 3046 |
+
"learning_rate": 0.001,
|
| 3047 |
+
"loss": 1.9611454010009766,
|
| 3048 |
+
"step": 868
|
| 3049 |
+
},
|
| 3050 |
+
{
|
| 3051 |
+
"epoch": 0.36403577592970343,
|
| 3052 |
+
"grad_norm": 0.17761600017547607,
|
| 3053 |
+
"learning_rate": 0.001,
|
| 3054 |
+
"loss": 1.7109124660491943,
|
| 3055 |
+
"step": 870
|
| 3056 |
+
},
|
| 3057 |
+
{
|
| 3058 |
+
"epoch": 0.3648726397824154,
|
| 3059 |
+
"grad_norm": 0.20071087777614594,
|
| 3060 |
+
"learning_rate": 0.001,
|
| 3061 |
+
"loss": 2.0035483837127686,
|
| 3062 |
+
"step": 872
|
| 3063 |
+
},
|
| 3064 |
+
{
|
| 3065 |
+
"epoch": 0.3657095036351274,
|
| 3066 |
+
"grad_norm": 0.13550806045532227,
|
| 3067 |
+
"learning_rate": 0.001,
|
| 3068 |
+
"loss": 2.0695624351501465,
|
| 3069 |
+
"step": 874
|
| 3070 |
+
},
|
| 3071 |
+
{
|
| 3072 |
+
"epoch": 0.3665463674878393,
|
| 3073 |
+
"grad_norm": 0.18393978476524353,
|
| 3074 |
+
"learning_rate": 0.001,
|
| 3075 |
+
"loss": 1.8894411325454712,
|
| 3076 |
+
"step": 876
|
| 3077 |
+
},
|
| 3078 |
+
{
|
| 3079 |
+
"epoch": 0.3673832313405513,
|
| 3080 |
+
"grad_norm": 0.15232665836811066,
|
| 3081 |
+
"learning_rate": 0.001,
|
| 3082 |
+
"loss": 1.8594024181365967,
|
| 3083 |
+
"step": 878
|
| 3084 |
+
},
|
| 3085 |
+
{
|
| 3086 |
+
"epoch": 0.36822009519326326,
|
| 3087 |
+
"grad_norm": 0.17846406996250153,
|
| 3088 |
+
"learning_rate": 0.001,
|
| 3089 |
+
"loss": 1.7387466430664062,
|
| 3090 |
+
"step": 880
|
| 3091 |
+
},
|
| 3092 |
+
{
|
| 3093 |
+
"epoch": 0.3690569590459752,
|
| 3094 |
+
"grad_norm": 0.16683928668498993,
|
| 3095 |
+
"learning_rate": 0.001,
|
| 3096 |
+
"loss": 2.0036778450012207,
|
| 3097 |
+
"step": 882
|
| 3098 |
+
},
|
| 3099 |
+
{
|
| 3100 |
+
"epoch": 0.36989382289868716,
|
| 3101 |
+
"grad_norm": 0.15304864943027496,
|
| 3102 |
+
"learning_rate": 0.001,
|
| 3103 |
+
"loss": 1.8767156600952148,
|
| 3104 |
+
"step": 884
|
| 3105 |
+
},
|
| 3106 |
+
{
|
| 3107 |
+
"epoch": 0.37073068675139914,
|
| 3108 |
+
"grad_norm": 0.1274169683456421,
|
| 3109 |
+
"learning_rate": 0.001,
|
| 3110 |
+
"loss": 1.8328161239624023,
|
| 3111 |
+
"step": 886
|
| 3112 |
+
},
|
| 3113 |
+
{
|
| 3114 |
+
"epoch": 0.3715675506041111,
|
| 3115 |
+
"grad_norm": 0.2863754630088806,
|
| 3116 |
+
"learning_rate": 0.001,
|
| 3117 |
+
"loss": 1.9944405555725098,
|
| 3118 |
+
"step": 888
|
| 3119 |
+
},
|
| 3120 |
+
{
|
| 3121 |
+
"epoch": 0.37240441445682304,
|
| 3122 |
+
"grad_norm": 0.1632489562034607,
|
| 3123 |
+
"learning_rate": 0.001,
|
| 3124 |
+
"loss": 2.201812505722046,
|
| 3125 |
+
"step": 890
|
| 3126 |
+
},
|
| 3127 |
+
{
|
| 3128 |
+
"epoch": 0.373241278309535,
|
| 3129 |
+
"grad_norm": 0.18280380964279175,
|
| 3130 |
+
"learning_rate": 0.001,
|
| 3131 |
+
"loss": 1.984431505203247,
|
| 3132 |
+
"step": 892
|
| 3133 |
+
},
|
| 3134 |
+
{
|
| 3135 |
+
"epoch": 0.374078142162247,
|
| 3136 |
+
"grad_norm": 0.2175862044095993,
|
| 3137 |
+
"learning_rate": 0.001,
|
| 3138 |
+
"loss": 2.0239691734313965,
|
| 3139 |
+
"step": 894
|
| 3140 |
+
},
|
| 3141 |
+
{
|
| 3142 |
+
"epoch": 0.3749150060149589,
|
| 3143 |
+
"grad_norm": 0.2042212039232254,
|
| 3144 |
+
"learning_rate": 0.001,
|
| 3145 |
+
"loss": 1.94392991065979,
|
| 3146 |
+
"step": 896
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 0.3757518698676709,
|
| 3150 |
+
"grad_norm": 0.19343379139900208,
|
| 3151 |
+
"learning_rate": 0.001,
|
| 3152 |
+
"loss": 1.8200492858886719,
|
| 3153 |
+
"step": 898
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 0.3765887337203829,
|
| 3157 |
+
"grad_norm": 0.15494228899478912,
|
| 3158 |
+
"learning_rate": 0.001,
|
| 3159 |
+
"loss": 1.9902455806732178,
|
| 3160 |
+
"step": 900
|
| 3161 |
+
},
|
| 3162 |
+
{
|
| 3163 |
+
"epoch": 0.37742559757309485,
|
| 3164 |
+
"grad_norm": 0.16390930116176605,
|
| 3165 |
+
"learning_rate": 0.001,
|
| 3166 |
+
"loss": 1.9770724773406982,
|
| 3167 |
+
"step": 902
|
| 3168 |
+
},
|
| 3169 |
+
{
|
| 3170 |
+
"epoch": 0.3782624614258068,
|
| 3171 |
+
"grad_norm": 0.1899786740541458,
|
| 3172 |
+
"learning_rate": 0.001,
|
| 3173 |
+
"loss": 1.7640061378479004,
|
| 3174 |
+
"step": 904
|
| 3175 |
+
},
|
| 3176 |
+
{
|
| 3177 |
+
"epoch": 0.37909932527851875,
|
| 3178 |
+
"grad_norm": 0.15270715951919556,
|
| 3179 |
+
"learning_rate": 0.001,
|
| 3180 |
+
"loss": 1.744924783706665,
|
| 3181 |
+
"step": 906
|
| 3182 |
+
},
|
| 3183 |
+
{
|
| 3184 |
+
"epoch": 0.37993618913123073,
|
| 3185 |
+
"grad_norm": 0.16899310052394867,
|
| 3186 |
+
"learning_rate": 0.001,
|
| 3187 |
+
"loss": 1.882214069366455,
|
| 3188 |
+
"step": 908
|
| 3189 |
+
},
|
| 3190 |
+
{
|
| 3191 |
+
"epoch": 0.38077305298394265,
|
| 3192 |
+
"grad_norm": 0.21054407954216003,
|
| 3193 |
+
"learning_rate": 0.001,
|
| 3194 |
+
"loss": 2.0300934314727783,
|
| 3195 |
+
"step": 910
|
| 3196 |
+
},
|
| 3197 |
+
{
|
| 3198 |
+
"epoch": 0.38160991683665463,
|
| 3199 |
+
"grad_norm": 0.14994964003562927,
|
| 3200 |
+
"learning_rate": 0.001,
|
| 3201 |
+
"loss": 2.0306694507598877,
|
| 3202 |
+
"step": 912
|
| 3203 |
+
},
|
| 3204 |
+
{
|
| 3205 |
+
"epoch": 0.3824467806893666,
|
| 3206 |
+
"grad_norm": 0.1923508197069168,
|
| 3207 |
+
"learning_rate": 0.001,
|
| 3208 |
+
"loss": 2.017732620239258,
|
| 3209 |
+
"step": 914
|
| 3210 |
+
},
|
| 3211 |
+
{
|
| 3212 |
+
"epoch": 0.3832836445420786,
|
| 3213 |
+
"grad_norm": 0.20616550743579865,
|
| 3214 |
+
"learning_rate": 0.001,
|
| 3215 |
+
"loss": 1.876326084136963,
|
| 3216 |
+
"step": 916
|
| 3217 |
+
},
|
| 3218 |
+
{
|
| 3219 |
+
"epoch": 0.3841205083947905,
|
| 3220 |
+
"grad_norm": 0.17768679559230804,
|
| 3221 |
+
"learning_rate": 0.001,
|
| 3222 |
+
"loss": 2.000495672225952,
|
| 3223 |
+
"step": 918
|
| 3224 |
+
},
|
| 3225 |
+
{
|
| 3226 |
+
"epoch": 0.3849573722475025,
|
| 3227 |
+
"grad_norm": 0.15679720044136047,
|
| 3228 |
+
"learning_rate": 0.001,
|
| 3229 |
+
"loss": 1.852207899093628,
|
| 3230 |
+
"step": 920
|
| 3231 |
+
},
|
| 3232 |
+
{
|
| 3233 |
+
"epoch": 0.38579423610021446,
|
| 3234 |
+
"grad_norm": 0.2106035351753235,
|
| 3235 |
+
"learning_rate": 0.001,
|
| 3236 |
+
"loss": 2.0443673133850098,
|
| 3237 |
+
"step": 922
|
| 3238 |
+
},
|
| 3239 |
+
{
|
| 3240 |
+
"epoch": 0.3866310999529264,
|
| 3241 |
+
"grad_norm": 0.15228495001792908,
|
| 3242 |
+
"learning_rate": 0.001,
|
| 3243 |
+
"loss": 2.0514485836029053,
|
| 3244 |
+
"step": 924
|
| 3245 |
+
},
|
| 3246 |
+
{
|
| 3247 |
+
"epoch": 0.38746796380563836,
|
| 3248 |
+
"grad_norm": 0.1557074636220932,
|
| 3249 |
+
"learning_rate": 0.001,
|
| 3250 |
+
"loss": 1.8469433784484863,
|
| 3251 |
+
"step": 926
|
| 3252 |
+
},
|
| 3253 |
+
{
|
| 3254 |
+
"epoch": 0.38830482765835034,
|
| 3255 |
+
"grad_norm": 0.17438408732414246,
|
| 3256 |
+
"learning_rate": 0.001,
|
| 3257 |
+
"loss": 1.872908353805542,
|
| 3258 |
+
"step": 928
|
| 3259 |
+
},
|
| 3260 |
+
{
|
| 3261 |
+
"epoch": 0.3891416915110623,
|
| 3262 |
+
"grad_norm": 0.17715111374855042,
|
| 3263 |
+
"learning_rate": 0.001,
|
| 3264 |
+
"loss": 1.8595507144927979,
|
| 3265 |
+
"step": 930
|
| 3266 |
+
},
|
| 3267 |
+
{
|
| 3268 |
+
"epoch": 0.38997855536377424,
|
| 3269 |
+
"grad_norm": 0.1674790382385254,
|
| 3270 |
+
"learning_rate": 0.001,
|
| 3271 |
+
"loss": 1.8077672719955444,
|
| 3272 |
+
"step": 932
|
| 3273 |
+
},
|
| 3274 |
+
{
|
| 3275 |
+
"epoch": 0.3908154192164862,
|
| 3276 |
+
"grad_norm": 0.16420049965381622,
|
| 3277 |
+
"learning_rate": 0.001,
|
| 3278 |
+
"loss": 1.7856109142303467,
|
| 3279 |
+
"step": 934
|
| 3280 |
+
},
|
| 3281 |
+
{
|
| 3282 |
+
"epoch": 0.3916522830691982,
|
| 3283 |
+
"grad_norm": 0.17187325656414032,
|
| 3284 |
+
"learning_rate": 0.001,
|
| 3285 |
+
"loss": 1.9843409061431885,
|
| 3286 |
+
"step": 936
|
| 3287 |
+
},
|
| 3288 |
+
{
|
| 3289 |
+
"epoch": 0.3924891469219101,
|
| 3290 |
+
"grad_norm": 0.17182452976703644,
|
| 3291 |
+
"learning_rate": 0.001,
|
| 3292 |
+
"loss": 1.8276844024658203,
|
| 3293 |
+
"step": 938
|
| 3294 |
+
},
|
| 3295 |
+
{
|
| 3296 |
+
"epoch": 0.3933260107746221,
|
| 3297 |
+
"grad_norm": 0.14653123915195465,
|
| 3298 |
+
"learning_rate": 0.001,
|
| 3299 |
+
"loss": 1.9275693893432617,
|
| 3300 |
+
"step": 940
|
| 3301 |
+
},
|
| 3302 |
+
{
|
| 3303 |
+
"epoch": 0.39416287462733407,
|
| 3304 |
+
"grad_norm": 0.18896397948265076,
|
| 3305 |
+
"learning_rate": 0.001,
|
| 3306 |
+
"loss": 1.8866583108901978,
|
| 3307 |
+
"step": 942
|
| 3308 |
+
},
|
| 3309 |
+
{
|
| 3310 |
+
"epoch": 0.39499973848004605,
|
| 3311 |
+
"grad_norm": 0.1829843521118164,
|
| 3312 |
+
"learning_rate": 0.001,
|
| 3313 |
+
"loss": 1.9205105304718018,
|
| 3314 |
+
"step": 944
|
| 3315 |
+
},
|
| 3316 |
+
{
|
| 3317 |
+
"epoch": 0.39583660233275797,
|
| 3318 |
+
"grad_norm": 0.1734960973262787,
|
| 3319 |
+
"learning_rate": 0.001,
|
| 3320 |
+
"loss": 1.876957893371582,
|
| 3321 |
+
"step": 946
|
| 3322 |
+
},
|
| 3323 |
+
{
|
| 3324 |
+
"epoch": 0.39667346618546995,
|
| 3325 |
+
"grad_norm": 0.21156294643878937,
|
| 3326 |
+
"learning_rate": 0.001,
|
| 3327 |
+
"loss": 1.9193332195281982,
|
| 3328 |
+
"step": 948
|
| 3329 |
+
},
|
| 3330 |
+
{
|
| 3331 |
+
"epoch": 0.3975103300381819,
|
| 3332 |
+
"grad_norm": 0.20835177600383759,
|
| 3333 |
+
"learning_rate": 0.001,
|
| 3334 |
+
"loss": 1.6833326816558838,
|
| 3335 |
+
"step": 950
|
| 3336 |
+
},
|
| 3337 |
+
{
|
| 3338 |
+
"epoch": 0.3983471938908939,
|
| 3339 |
+
"grad_norm": 0.14049121737480164,
|
| 3340 |
+
"learning_rate": 0.001,
|
| 3341 |
+
"loss": 1.945499300956726,
|
| 3342 |
+
"step": 952
|
| 3343 |
+
},
|
| 3344 |
+
{
|
| 3345 |
+
"epoch": 0.3991840577436058,
|
| 3346 |
+
"grad_norm": 0.18569767475128174,
|
| 3347 |
+
"learning_rate": 0.001,
|
| 3348 |
+
"loss": 1.7446472644805908,
|
| 3349 |
+
"step": 954
|
| 3350 |
+
},
|
| 3351 |
+
{
|
| 3352 |
+
"epoch": 0.4000209215963178,
|
| 3353 |
+
"grad_norm": 0.18301931023597717,
|
| 3354 |
+
"learning_rate": 0.001,
|
| 3355 |
+
"loss": 1.8743345737457275,
|
| 3356 |
+
"step": 956
|
| 3357 |
}
|
| 3358 |
],
|
| 3359 |
"logging_steps": 2,
|
|
|
|
| 3373 |
"attributes": {}
|
| 3374 |
}
|
| 3375 |
},
|
| 3376 |
+
"total_flos": 2.0375091667376538e+17,
|
| 3377 |
"train_batch_size": 1,
|
| 3378 |
"trial_name": null,
|
| 3379 |
"trial_params": null
|