Training in progress, step 4100, 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 1235573136
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:251b80dbafc03e7542fea3f5fdb21a7406af20ef74904296f4e428f94d593cd6
|
| 3 |
size 1235573136
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2471218763
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f262e6cdc90a567db3bb7f9eb2f117e8cbecc2bea5244f233a9be17c3ff3d180
|
| 3 |
size 2471218763
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6bbe17a5fe328cdb084ec10afd49d2cb7bb9035b6736206385c2c19bf9f66c41
|
| 3 |
size 14645
|
last-checkpoint/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8377531a6924fec4658b9153d3ae4c11dbd3e0f08b336f91ff60fef0ac414eeb
|
| 3 |
size 1383
|
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 sha256:76337cd86bed08edb589ad03a7c68cee9c51af6250c0d13ab74fd126165aa258
|
| 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":
|
| 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,
|
|
@@ -1128,6 +1128,1756 @@
|
|
| 1128 |
"learning_rate": 0.00029098797595190375,
|
| 1129 |
"loss": 3103321489408.0,
|
| 1130 |
"step": 1600
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1131 |
}
|
| 1132 |
],
|
| 1133 |
"logging_steps": 10,
|
|
@@ -1147,7 +2897,7 @@
|
|
| 1147 |
"attributes": {}
|
| 1148 |
}
|
| 1149 |
},
|
| 1150 |
-
"total_flos": 2.
|
| 1151 |
"train_batch_size": 2,
|
| 1152 |
"trial_name": null,
|
| 1153 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 24.0017,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 4100,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1128 |
"learning_rate": 0.00029098797595190375,
|
| 1129 |
"loss": 3103321489408.0,
|
| 1130 |
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.0002,
|
| 1134 |
+
"grad_norm": 0.0,
|
| 1135 |
+
"learning_rate": 0.0002909278557114228,
|
| 1136 |
+
"loss": 95.5510498046875,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.0004,
|
| 1141 |
+
"grad_norm": 0.0,
|
| 1142 |
+
"learning_rate": 0.00029086773547094184,
|
| 1143 |
+
"loss": 98.69436645507812,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.0006,
|
| 1148 |
+
"grad_norm": 0.0,
|
| 1149 |
+
"learning_rate": 0.0002908076152304609,
|
| 1150 |
+
"loss": 99.0325439453125,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.0008,
|
| 1155 |
+
"grad_norm": 0.0,
|
| 1156 |
+
"learning_rate": 0.00029074749498997994,
|
| 1157 |
+
"loss": 97.31904296875,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.001,
|
| 1162 |
+
"grad_norm": 0.0,
|
| 1163 |
+
"learning_rate": 0.000290687374749499,
|
| 1164 |
+
"loss": 98.40114135742188,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.0012,
|
| 1169 |
+
"grad_norm": 0.0,
|
| 1170 |
+
"learning_rate": 0.00029062725450901803,
|
| 1171 |
+
"loss": 99.31021728515626,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.0014,
|
| 1176 |
+
"grad_norm": 0.0,
|
| 1177 |
+
"learning_rate": 0.00029056713426853703,
|
| 1178 |
+
"loss": 99.4878662109375,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.0016,
|
| 1183 |
+
"grad_norm": 0.0,
|
| 1184 |
+
"learning_rate": 0.0002905070140280561,
|
| 1185 |
+
"loss": 99.3884521484375,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.0018,
|
| 1190 |
+
"grad_norm": 0.0,
|
| 1191 |
+
"learning_rate": 0.0002904468937875751,
|
| 1192 |
+
"loss": 99.46731567382812,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.002,
|
| 1197 |
+
"grad_norm": 0.0,
|
| 1198 |
+
"learning_rate": 0.00029038677354709417,
|
| 1199 |
+
"loss": 99.4464599609375,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.0001,
|
| 1204 |
+
"grad_norm": 0.0,
|
| 1205 |
+
"learning_rate": 0.00029032665330661317,
|
| 1206 |
+
"loss": 104.66427001953124,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.0003,
|
| 1211 |
+
"grad_norm": 0.0,
|
| 1212 |
+
"learning_rate": 0.0002902665330661322,
|
| 1213 |
+
"loss": 98.24152221679688,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.0005,
|
| 1218 |
+
"grad_norm": 0.0,
|
| 1219 |
+
"learning_rate": 0.0002902064128256513,
|
| 1220 |
+
"loss": 99.4651123046875,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.0007,
|
| 1225 |
+
"grad_norm": 0.0,
|
| 1226 |
+
"learning_rate": 0.0002901462925851703,
|
| 1227 |
+
"loss": 99.3810546875,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.0009,
|
| 1232 |
+
"grad_norm": 0.0,
|
| 1233 |
+
"learning_rate": 0.00029008617234468936,
|
| 1234 |
+
"loss": 96.20846557617188,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.0011,
|
| 1239 |
+
"grad_norm": 0.0,
|
| 1240 |
+
"learning_rate": 0.0002900260521042084,
|
| 1241 |
+
"loss": 99.44166259765625,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.0013,
|
| 1246 |
+
"grad_norm": 0.0,
|
| 1247 |
+
"learning_rate": 0.00028996593186372745,
|
| 1248 |
+
"loss": 99.43170776367188,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.0015,
|
| 1253 |
+
"grad_norm": 0.0,
|
| 1254 |
+
"learning_rate": 0.00028990581162324644,
|
| 1255 |
+
"loss": 99.37213134765625,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.0017,
|
| 1260 |
+
"grad_norm": 0.0,
|
| 1261 |
+
"learning_rate": 0.0002898456913827655,
|
| 1262 |
+
"loss": 99.51829223632812,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.0019,
|
| 1267 |
+
"grad_norm": 0.0,
|
| 1268 |
+
"learning_rate": 0.00028978557114228454,
|
| 1269 |
+
"loss": 99.27789306640625,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.0021,
|
| 1274 |
+
"grad_norm": 0.0,
|
| 1275 |
+
"learning_rate": 0.0002897254509018036,
|
| 1276 |
+
"loss": 99.44300537109375,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 3.0002,
|
| 1281 |
+
"grad_norm": 0.0,
|
| 1282 |
+
"learning_rate": 0.00028966533066132264,
|
| 1283 |
+
"loss": 103.58494873046875,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 3.0004,
|
| 1288 |
+
"grad_norm": 0.0,
|
| 1289 |
+
"learning_rate": 0.0002896052104208417,
|
| 1290 |
+
"loss": 99.35955810546875,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 3.0006,
|
| 1295 |
+
"grad_norm": 0.0,
|
| 1296 |
+
"learning_rate": 0.00028954509018036073,
|
| 1297 |
+
"loss": 99.25761108398437,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 3.0008,
|
| 1302 |
+
"grad_norm": 0.0,
|
| 1303 |
+
"learning_rate": 0.0002894849699398797,
|
| 1304 |
+
"loss": 97.34839477539063,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 3.001,
|
| 1309 |
+
"grad_norm": 0.0,
|
| 1310 |
+
"learning_rate": 0.00028942484969939877,
|
| 1311 |
+
"loss": 98.359326171875,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 3.0012,
|
| 1316 |
+
"grad_norm": 0.0,
|
| 1317 |
+
"learning_rate": 0.0002893647294589178,
|
| 1318 |
+
"loss": 99.25703735351563,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 3.0014,
|
| 1323 |
+
"grad_norm": 0.0,
|
| 1324 |
+
"learning_rate": 0.00028930460921843687,
|
| 1325 |
+
"loss": 99.42723388671875,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 3.0016,
|
| 1330 |
+
"grad_norm": 0.0,
|
| 1331 |
+
"learning_rate": 0.00028924448897795586,
|
| 1332 |
+
"loss": 99.32404174804688,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 3.0018,
|
| 1337 |
+
"grad_norm": 0.0,
|
| 1338 |
+
"learning_rate": 0.0002891843687374749,
|
| 1339 |
+
"loss": 99.3990966796875,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 3.002,
|
| 1344 |
+
"grad_norm": 0.0,
|
| 1345 |
+
"learning_rate": 0.00028912424849699396,
|
| 1346 |
+
"loss": 99.39078369140626,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 4.0001,
|
| 1351 |
+
"grad_norm": 0.0,
|
| 1352 |
+
"learning_rate": 0.000289064128256513,
|
| 1353 |
+
"loss": 104.5905517578125,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 4.0003,
|
| 1358 |
+
"grad_norm": 0.0,
|
| 1359 |
+
"learning_rate": 0.00028900400801603205,
|
| 1360 |
+
"loss": 98.17286987304688,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 4.0005,
|
| 1365 |
+
"grad_norm": 0.0,
|
| 1366 |
+
"learning_rate": 0.0002889438877755511,
|
| 1367 |
+
"loss": 99.40567016601562,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 4.0007,
|
| 1372 |
+
"grad_norm": 0.0,
|
| 1373 |
+
"learning_rate": 0.00028888376753507015,
|
| 1374 |
+
"loss": 99.32139892578125,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 4.0009,
|
| 1379 |
+
"grad_norm": 0.0,
|
| 1380 |
+
"learning_rate": 0.00028882364729458914,
|
| 1381 |
+
"loss": 96.15136108398437,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 4.0011,
|
| 1386 |
+
"grad_norm": 0.0,
|
| 1387 |
+
"learning_rate": 0.0002887635270541082,
|
| 1388 |
+
"loss": 99.3803466796875,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 4.0013,
|
| 1393 |
+
"grad_norm": 0.0,
|
| 1394 |
+
"learning_rate": 0.00028870340681362724,
|
| 1395 |
+
"loss": 99.37009887695312,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 4.0015,
|
| 1400 |
+
"grad_norm": 0.0,
|
| 1401 |
+
"learning_rate": 0.0002886432865731463,
|
| 1402 |
+
"loss": 99.310498046875,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 4.0017,
|
| 1407 |
+
"grad_norm": 0.0,
|
| 1408 |
+
"learning_rate": 0.0002885831663326653,
|
| 1409 |
+
"loss": 99.45986938476562,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 4.0019,
|
| 1414 |
+
"grad_norm": 0.0,
|
| 1415 |
+
"learning_rate": 0.0002885230460921843,
|
| 1416 |
+
"loss": 99.21234741210938,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 4.0021,
|
| 1421 |
+
"grad_norm": 0.0,
|
| 1422 |
+
"learning_rate": 0.00028846292585170337,
|
| 1423 |
+
"loss": 99.381591796875,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 5.0002,
|
| 1428 |
+
"grad_norm": 0.0,
|
| 1429 |
+
"learning_rate": 0.0002884028056112224,
|
| 1430 |
+
"loss": 103.5211669921875,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 5.0004,
|
| 1435 |
+
"grad_norm": 0.0,
|
| 1436 |
+
"learning_rate": 0.00028834268537074147,
|
| 1437 |
+
"loss": 99.30160522460938,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 5.0006,
|
| 1442 |
+
"grad_norm": 0.0,
|
| 1443 |
+
"learning_rate": 0.0002882825651302605,
|
| 1444 |
+
"loss": 99.195703125,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 5.0008,
|
| 1449 |
+
"grad_norm": 0.0,
|
| 1450 |
+
"learning_rate": 0.0002882224448897795,
|
| 1451 |
+
"loss": 97.29168090820312,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 5.001,
|
| 1456 |
+
"grad_norm": 0.0,
|
| 1457 |
+
"learning_rate": 0.00028816232464929856,
|
| 1458 |
+
"loss": 98.30964965820313,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 5.0012,
|
| 1463 |
+
"grad_norm": 0.0,
|
| 1464 |
+
"learning_rate": 0.0002881022044088176,
|
| 1465 |
+
"loss": 99.19823608398437,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 5.0014,
|
| 1470 |
+
"grad_norm": 0.0,
|
| 1471 |
+
"learning_rate": 0.00028804208416833665,
|
| 1472 |
+
"loss": 99.3675048828125,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 5.0016,
|
| 1477 |
+
"grad_norm": 0.0,
|
| 1478 |
+
"learning_rate": 0.0002879819639278557,
|
| 1479 |
+
"loss": 99.27457275390626,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 5.0018,
|
| 1484 |
+
"grad_norm": 0.0,
|
| 1485 |
+
"learning_rate": 0.0002879218436873747,
|
| 1486 |
+
"loss": 99.3427490234375,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 5.002,
|
| 1491 |
+
"grad_norm": 0.0,
|
| 1492 |
+
"learning_rate": 0.0002878617234468938,
|
| 1493 |
+
"loss": 99.33101806640624,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 6.0001,
|
| 1498 |
+
"grad_norm": 0.0,
|
| 1499 |
+
"learning_rate": 0.0002878016032064128,
|
| 1500 |
+
"loss": 104.52410888671875,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 6.0003,
|
| 1505 |
+
"grad_norm": 0.0,
|
| 1506 |
+
"learning_rate": 0.00028774148296593184,
|
| 1507 |
+
"loss": 98.11661376953126,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 6.0005,
|
| 1512 |
+
"grad_norm": 0.0,
|
| 1513 |
+
"learning_rate": 0.0002876813627254509,
|
| 1514 |
+
"loss": 99.35023193359375,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 6.0007,
|
| 1519 |
+
"grad_norm": 0.0,
|
| 1520 |
+
"learning_rate": 0.00028762124248496993,
|
| 1521 |
+
"loss": 99.27654418945312,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 6.0009,
|
| 1526 |
+
"grad_norm": 0.0,
|
| 1527 |
+
"learning_rate": 0.0002875611222444889,
|
| 1528 |
+
"loss": 96.09561767578126,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 6.0011,
|
| 1533 |
+
"grad_norm": 0.0,
|
| 1534 |
+
"learning_rate": 0.000287501002004008,
|
| 1535 |
+
"loss": 99.32379760742188,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 6.0013,
|
| 1540 |
+
"grad_norm": 0.0,
|
| 1541 |
+
"learning_rate": 0.000287440881763527,
|
| 1542 |
+
"loss": 99.31390991210938,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 6.0015,
|
| 1547 |
+
"grad_norm": 0.0,
|
| 1548 |
+
"learning_rate": 0.00028738076152304607,
|
| 1549 |
+
"loss": 99.25233764648438,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 6.0017,
|
| 1554 |
+
"grad_norm": 0.0,
|
| 1555 |
+
"learning_rate": 0.0002873206412825651,
|
| 1556 |
+
"loss": 99.400146484375,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 6.0019,
|
| 1561 |
+
"grad_norm": 0.0,
|
| 1562 |
+
"learning_rate": 0.00028726052104208416,
|
| 1563 |
+
"loss": 99.16478271484375,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 6.0021,
|
| 1568 |
+
"grad_norm": 0.0,
|
| 1569 |
+
"learning_rate": 0.0002872004008016032,
|
| 1570 |
+
"loss": 99.3350830078125,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 7.0002,
|
| 1575 |
+
"grad_norm": 0.0,
|
| 1576 |
+
"learning_rate": 0.0002871402805611222,
|
| 1577 |
+
"loss": 103.46038818359375,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 7.0004,
|
| 1582 |
+
"grad_norm": 0.0,
|
| 1583 |
+
"learning_rate": 0.00028708016032064125,
|
| 1584 |
+
"loss": 99.24920043945312,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 7.0006,
|
| 1589 |
+
"grad_norm": 0.0,
|
| 1590 |
+
"learning_rate": 0.0002870200400801603,
|
| 1591 |
+
"loss": 99.14209594726563,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 7.0008,
|
| 1596 |
+
"grad_norm": 0.0,
|
| 1597 |
+
"learning_rate": 0.00028695991983967935,
|
| 1598 |
+
"loss": 97.23970336914063,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 7.001,
|
| 1603 |
+
"grad_norm": 0.0,
|
| 1604 |
+
"learning_rate": 0.00028689979959919834,
|
| 1605 |
+
"loss": 98.25855712890625,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 7.0012,
|
| 1610 |
+
"grad_norm": 0.0,
|
| 1611 |
+
"learning_rate": 0.0002868396793587174,
|
| 1612 |
+
"loss": 99.14596557617188,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 7.0014,
|
| 1617 |
+
"grad_norm": 0.0,
|
| 1618 |
+
"learning_rate": 0.00028677955911823644,
|
| 1619 |
+
"loss": 99.317333984375,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 7.0016,
|
| 1624 |
+
"grad_norm": 0.0,
|
| 1625 |
+
"learning_rate": 0.0002867194388777555,
|
| 1626 |
+
"loss": 99.22235107421875,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 7.0018,
|
| 1631 |
+
"grad_norm": 0.0,
|
| 1632 |
+
"learning_rate": 0.00028665931863727453,
|
| 1633 |
+
"loss": 99.2902587890625,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 7.002,
|
| 1638 |
+
"grad_norm": 0.0,
|
| 1639 |
+
"learning_rate": 0.0002865991983967936,
|
| 1640 |
+
"loss": 99.27511596679688,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 8.0001,
|
| 1645 |
+
"grad_norm": 0.0,
|
| 1646 |
+
"learning_rate": 0.00028653907815631263,
|
| 1647 |
+
"loss": 104.4739990234375,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 8.0003,
|
| 1652 |
+
"grad_norm": 0.0,
|
| 1653 |
+
"learning_rate": 0.0002864789579158316,
|
| 1654 |
+
"loss": 98.07261352539062,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 8.0005,
|
| 1659 |
+
"grad_norm": 0.0,
|
| 1660 |
+
"learning_rate": 0.00028641883767535067,
|
| 1661 |
+
"loss": 99.2950439453125,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 8.0007,
|
| 1666 |
+
"grad_norm": 0.0,
|
| 1667 |
+
"learning_rate": 0.0002863587174348697,
|
| 1668 |
+
"loss": 99.22308959960938,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 8.0009,
|
| 1673 |
+
"grad_norm": 0.0,
|
| 1674 |
+
"learning_rate": 0.00028629859719438876,
|
| 1675 |
+
"loss": 96.060498046875,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 8.0011,
|
| 1680 |
+
"grad_norm": 0.0,
|
| 1681 |
+
"learning_rate": 0.00028623847695390776,
|
| 1682 |
+
"loss": 99.27498168945313,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 8.0013,
|
| 1687 |
+
"grad_norm": 0.0,
|
| 1688 |
+
"learning_rate": 0.0002861783567134268,
|
| 1689 |
+
"loss": 99.27508544921875,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 8.0015,
|
| 1694 |
+
"grad_norm": 0.0,
|
| 1695 |
+
"learning_rate": 0.0002861182364729459,
|
| 1696 |
+
"loss": 99.20576782226563,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 8.0017,
|
| 1701 |
+
"grad_norm": 0.0,
|
| 1702 |
+
"learning_rate": 0.0002860581162324649,
|
| 1703 |
+
"loss": 99.35106811523437,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 8.0019,
|
| 1708 |
+
"grad_norm": 0.0,
|
| 1709 |
+
"learning_rate": 0.00028599799599198395,
|
| 1710 |
+
"loss": 99.11838989257812,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 8.0021,
|
| 1715 |
+
"grad_norm": 0.0,
|
| 1716 |
+
"learning_rate": 0.000285937875751503,
|
| 1717 |
+
"loss": 99.290185546875,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 9.0002,
|
| 1722 |
+
"grad_norm": 0.0,
|
| 1723 |
+
"learning_rate": 0.00028587775551102204,
|
| 1724 |
+
"loss": 103.40933837890626,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 9.0004,
|
| 1729 |
+
"grad_norm": 0.0,
|
| 1730 |
+
"learning_rate": 0.00028581763527054104,
|
| 1731 |
+
"loss": 99.20323486328125,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 9.0006,
|
| 1736 |
+
"grad_norm": 0.0,
|
| 1737 |
+
"learning_rate": 0.0002857575150300601,
|
| 1738 |
+
"loss": 99.10476684570312,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 9.0008,
|
| 1743 |
+
"grad_norm": 0.0,
|
| 1744 |
+
"learning_rate": 0.00028569739478957913,
|
| 1745 |
+
"loss": 97.20264282226563,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 9.001,
|
| 1750 |
+
"grad_norm": 0.0,
|
| 1751 |
+
"learning_rate": 0.0002856372745490982,
|
| 1752 |
+
"loss": 98.21160278320312,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 9.0012,
|
| 1757 |
+
"grad_norm": 0.0,
|
| 1758 |
+
"learning_rate": 0.0002855771543086172,
|
| 1759 |
+
"loss": 99.10530395507813,
|
| 1760 |
+
"step": 2500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 9.0014,
|
| 1764 |
+
"grad_norm": 0.0,
|
| 1765 |
+
"learning_rate": 0.0002855170340681363,
|
| 1766 |
+
"loss": 99.27515258789063,
|
| 1767 |
+
"step": 2510
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 9.0016,
|
| 1771 |
+
"grad_norm": 0.0,
|
| 1772 |
+
"learning_rate": 0.0002854569138276553,
|
| 1773 |
+
"loss": 99.17865600585938,
|
| 1774 |
+
"step": 2520
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 9.0018,
|
| 1778 |
+
"grad_norm": 0.0,
|
| 1779 |
+
"learning_rate": 0.0002853967935871743,
|
| 1780 |
+
"loss": 99.24905395507812,
|
| 1781 |
+
"step": 2530
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 9.002,
|
| 1785 |
+
"grad_norm": 0.0,
|
| 1786 |
+
"learning_rate": 0.00028533667334669336,
|
| 1787 |
+
"loss": 99.2433837890625,
|
| 1788 |
+
"step": 2540
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 10.0001,
|
| 1792 |
+
"grad_norm": 0.0,
|
| 1793 |
+
"learning_rate": 0.0002852765531062124,
|
| 1794 |
+
"loss": 104.4260498046875,
|
| 1795 |
+
"step": 2550
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 10.0003,
|
| 1799 |
+
"grad_norm": 0.0,
|
| 1800 |
+
"learning_rate": 0.00028521643286573146,
|
| 1801 |
+
"loss": 98.03045043945312,
|
| 1802 |
+
"step": 2560
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 10.0005,
|
| 1806 |
+
"grad_norm": 0.0,
|
| 1807 |
+
"learning_rate": 0.00028515631262525045,
|
| 1808 |
+
"loss": 99.26333618164062,
|
| 1809 |
+
"step": 2570
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 10.0007,
|
| 1813 |
+
"grad_norm": 0.0,
|
| 1814 |
+
"learning_rate": 0.0002850961923847695,
|
| 1815 |
+
"loss": 99.18395385742187,
|
| 1816 |
+
"step": 2580
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 10.0009,
|
| 1820 |
+
"grad_norm": 0.0,
|
| 1821 |
+
"learning_rate": 0.00028503607214428855,
|
| 1822 |
+
"loss": 96.02274169921876,
|
| 1823 |
+
"step": 2590
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 10.0011,
|
| 1827 |
+
"grad_norm": 0.0,
|
| 1828 |
+
"learning_rate": 0.0002849759519038076,
|
| 1829 |
+
"loss": 99.23676147460938,
|
| 1830 |
+
"step": 2600
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 10.0013,
|
| 1834 |
+
"grad_norm": 0.0,
|
| 1835 |
+
"learning_rate": 0.00028491583166332664,
|
| 1836 |
+
"loss": 99.226318359375,
|
| 1837 |
+
"step": 2610
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 10.0015,
|
| 1841 |
+
"grad_norm": 0.0,
|
| 1842 |
+
"learning_rate": 0.0002848557114228457,
|
| 1843 |
+
"loss": 99.171044921875,
|
| 1844 |
+
"step": 2620
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 10.0017,
|
| 1848 |
+
"grad_norm": 0.0,
|
| 1849 |
+
"learning_rate": 0.00028479559118236474,
|
| 1850 |
+
"loss": 99.31712036132812,
|
| 1851 |
+
"step": 2630
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 10.0019,
|
| 1855 |
+
"grad_norm": 0.0,
|
| 1856 |
+
"learning_rate": 0.00028473547094188373,
|
| 1857 |
+
"loss": 99.08142700195313,
|
| 1858 |
+
"step": 2640
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 10.0021,
|
| 1862 |
+
"grad_norm": 0.0,
|
| 1863 |
+
"learning_rate": 0.0002846753507014028,
|
| 1864 |
+
"loss": 99.24673461914062,
|
| 1865 |
+
"step": 2650
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 11.0002,
|
| 1869 |
+
"grad_norm": 0.0,
|
| 1870 |
+
"learning_rate": 0.00028461523046092183,
|
| 1871 |
+
"loss": 103.37125244140626,
|
| 1872 |
+
"step": 2660
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 11.0004,
|
| 1876 |
+
"grad_norm": 0.0,
|
| 1877 |
+
"learning_rate": 0.0002845551102204409,
|
| 1878 |
+
"loss": 99.16270141601562,
|
| 1879 |
+
"step": 2670
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 11.0006,
|
| 1883 |
+
"grad_norm": 0.0,
|
| 1884 |
+
"learning_rate": 0.00028449498997995987,
|
| 1885 |
+
"loss": 99.06387329101562,
|
| 1886 |
+
"step": 2680
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 11.0008,
|
| 1890 |
+
"grad_norm": 0.0,
|
| 1891 |
+
"learning_rate": 0.0002844348697394789,
|
| 1892 |
+
"loss": 97.17066040039063,
|
| 1893 |
+
"step": 2690
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 11.001,
|
| 1897 |
+
"grad_norm": 0.0,
|
| 1898 |
+
"learning_rate": 0.00028437474949899797,
|
| 1899 |
+
"loss": 98.17435302734376,
|
| 1900 |
+
"step": 2700
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 11.0012,
|
| 1904 |
+
"grad_norm": 0.0,
|
| 1905 |
+
"learning_rate": 0.000284314629258517,
|
| 1906 |
+
"loss": 99.06696166992188,
|
| 1907 |
+
"step": 2710
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 11.0014,
|
| 1911 |
+
"grad_norm": 0.0,
|
| 1912 |
+
"learning_rate": 0.00028425450901803606,
|
| 1913 |
+
"loss": 99.23809814453125,
|
| 1914 |
+
"step": 2720
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 11.0016,
|
| 1918 |
+
"grad_norm": 0.0,
|
| 1919 |
+
"learning_rate": 0.0002841943887775551,
|
| 1920 |
+
"loss": 99.14588623046875,
|
| 1921 |
+
"step": 2730
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 11.0018,
|
| 1925 |
+
"grad_norm": 0.0,
|
| 1926 |
+
"learning_rate": 0.0002841342685370741,
|
| 1927 |
+
"loss": 99.21581420898437,
|
| 1928 |
+
"step": 2740
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 11.002,
|
| 1932 |
+
"grad_norm": 0.0,
|
| 1933 |
+
"learning_rate": 0.00028407414829659315,
|
| 1934 |
+
"loss": 99.2086181640625,
|
| 1935 |
+
"step": 2750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 12.0001,
|
| 1939 |
+
"grad_norm": 0.0,
|
| 1940 |
+
"learning_rate": 0.0002840140280561122,
|
| 1941 |
+
"loss": 104.3935546875,
|
| 1942 |
+
"step": 2760
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 12.0003,
|
| 1946 |
+
"grad_norm": 0.0,
|
| 1947 |
+
"learning_rate": 0.00028395390781563125,
|
| 1948 |
+
"loss": 97.99188842773438,
|
| 1949 |
+
"step": 2770
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 12.0005,
|
| 1953 |
+
"grad_norm": 0.0,
|
| 1954 |
+
"learning_rate": 0.0002838937875751503,
|
| 1955 |
+
"loss": 99.22097778320312,
|
| 1956 |
+
"step": 2780
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 12.0007,
|
| 1960 |
+
"grad_norm": 0.0,
|
| 1961 |
+
"learning_rate": 0.0002838336673346693,
|
| 1962 |
+
"loss": 99.151953125,
|
| 1963 |
+
"step": 2790
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 12.0009,
|
| 1967 |
+
"grad_norm": 0.0,
|
| 1968 |
+
"learning_rate": 0.00028377354709418833,
|
| 1969 |
+
"loss": 95.98749389648438,
|
| 1970 |
+
"step": 2800
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 12.0011,
|
| 1974 |
+
"grad_norm": 0.0,
|
| 1975 |
+
"learning_rate": 0.0002837134268537074,
|
| 1976 |
+
"loss": 99.2070556640625,
|
| 1977 |
+
"step": 2810
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 12.0013,
|
| 1981 |
+
"grad_norm": 0.0,
|
| 1982 |
+
"learning_rate": 0.00028365330661322643,
|
| 1983 |
+
"loss": 99.2055419921875,
|
| 1984 |
+
"step": 2820
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 12.0015,
|
| 1988 |
+
"grad_norm": 0.0,
|
| 1989 |
+
"learning_rate": 0.0002835931863727455,
|
| 1990 |
+
"loss": 99.14091186523437,
|
| 1991 |
+
"step": 2830
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 12.0017,
|
| 1995 |
+
"grad_norm": 0.0,
|
| 1996 |
+
"learning_rate": 0.0002835330661322645,
|
| 1997 |
+
"loss": 99.283544921875,
|
| 1998 |
+
"step": 2840
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 12.0019,
|
| 2002 |
+
"grad_norm": 0.0,
|
| 2003 |
+
"learning_rate": 0.0002834729458917835,
|
| 2004 |
+
"loss": 99.05504150390625,
|
| 2005 |
+
"step": 2850
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 12.0021,
|
| 2009 |
+
"grad_norm": 0.0,
|
| 2010 |
+
"learning_rate": 0.00028341282565130257,
|
| 2011 |
+
"loss": 99.22215576171875,
|
| 2012 |
+
"step": 2860
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 13.0002,
|
| 2016 |
+
"grad_norm": 0.0,
|
| 2017 |
+
"learning_rate": 0.0002833527054108216,
|
| 2018 |
+
"loss": 103.33685302734375,
|
| 2019 |
+
"step": 2870
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 13.0004,
|
| 2023 |
+
"grad_norm": 0.0,
|
| 2024 |
+
"learning_rate": 0.00028329258517034066,
|
| 2025 |
+
"loss": 99.140380859375,
|
| 2026 |
+
"step": 2880
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 13.0006,
|
| 2030 |
+
"grad_norm": 0.0,
|
| 2031 |
+
"learning_rate": 0.00028323246492985965,
|
| 2032 |
+
"loss": 99.0400634765625,
|
| 2033 |
+
"step": 2890
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 13.0008,
|
| 2037 |
+
"grad_norm": 0.0,
|
| 2038 |
+
"learning_rate": 0.0002831723446893787,
|
| 2039 |
+
"loss": 97.14813842773438,
|
| 2040 |
+
"step": 2900
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 13.001,
|
| 2044 |
+
"grad_norm": 0.0,
|
| 2045 |
+
"learning_rate": 0.0002831122244488978,
|
| 2046 |
+
"loss": 98.15328979492188,
|
| 2047 |
+
"step": 2910
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 13.0012,
|
| 2051 |
+
"grad_norm": 0.0,
|
| 2052 |
+
"learning_rate": 0.0002830521042084168,
|
| 2053 |
+
"loss": 99.04591674804688,
|
| 2054 |
+
"step": 2920
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 13.0014,
|
| 2058 |
+
"grad_norm": 0.0,
|
| 2059 |
+
"learning_rate": 0.00028299198396793585,
|
| 2060 |
+
"loss": 99.20752563476563,
|
| 2061 |
+
"step": 2930
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 13.0016,
|
| 2065 |
+
"grad_norm": 0.0,
|
| 2066 |
+
"learning_rate": 0.0002829318637274549,
|
| 2067 |
+
"loss": 99.11998291015625,
|
| 2068 |
+
"step": 2940
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 13.0018,
|
| 2072 |
+
"grad_norm": 0.0,
|
| 2073 |
+
"learning_rate": 0.00028287174348697394,
|
| 2074 |
+
"loss": 99.19060668945312,
|
| 2075 |
+
"step": 2950
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 13.002,
|
| 2079 |
+
"grad_norm": 0.0,
|
| 2080 |
+
"learning_rate": 0.00028281162324649293,
|
| 2081 |
+
"loss": 99.18587036132813,
|
| 2082 |
+
"step": 2960
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 14.0001,
|
| 2086 |
+
"grad_norm": 0.0,
|
| 2087 |
+
"learning_rate": 0.000282751503006012,
|
| 2088 |
+
"loss": 104.3642578125,
|
| 2089 |
+
"step": 2970
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 14.0003,
|
| 2093 |
+
"grad_norm": 0.0,
|
| 2094 |
+
"learning_rate": 0.00028269138276553103,
|
| 2095 |
+
"loss": 97.97385864257812,
|
| 2096 |
+
"step": 2980
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 14.0005,
|
| 2100 |
+
"grad_norm": 0.0,
|
| 2101 |
+
"learning_rate": 0.0002826312625250501,
|
| 2102 |
+
"loss": 99.20270385742188,
|
| 2103 |
+
"step": 2990
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 14.0007,
|
| 2107 |
+
"grad_norm": 0.0,
|
| 2108 |
+
"learning_rate": 0.0002825711422845691,
|
| 2109 |
+
"loss": 99.12491455078126,
|
| 2110 |
+
"step": 3000
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 14.0009,
|
| 2114 |
+
"grad_norm": 0.0,
|
| 2115 |
+
"learning_rate": 0.00028251102204408817,
|
| 2116 |
+
"loss": 95.97457275390624,
|
| 2117 |
+
"step": 3010
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 14.0011,
|
| 2121 |
+
"grad_norm": 0.0,
|
| 2122 |
+
"learning_rate": 0.0002824509018036072,
|
| 2123 |
+
"loss": 99.18565673828125,
|
| 2124 |
+
"step": 3020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 14.0013,
|
| 2128 |
+
"grad_norm": 0.0,
|
| 2129 |
+
"learning_rate": 0.0002823907815631262,
|
| 2130 |
+
"loss": 99.17833251953125,
|
| 2131 |
+
"step": 3030
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 14.0015,
|
| 2135 |
+
"grad_norm": 0.0,
|
| 2136 |
+
"learning_rate": 0.00028233066132264526,
|
| 2137 |
+
"loss": 99.11763305664063,
|
| 2138 |
+
"step": 3040
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 14.0017,
|
| 2142 |
+
"grad_norm": 0.0,
|
| 2143 |
+
"learning_rate": 0.0002822705410821643,
|
| 2144 |
+
"loss": 99.26400146484374,
|
| 2145 |
+
"step": 3050
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 14.0019,
|
| 2149 |
+
"grad_norm": 0.0,
|
| 2150 |
+
"learning_rate": 0.00028221042084168336,
|
| 2151 |
+
"loss": 99.02830200195312,
|
| 2152 |
+
"step": 3060
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 14.0021,
|
| 2156 |
+
"grad_norm": 0.0,
|
| 2157 |
+
"learning_rate": 0.00028215030060120235,
|
| 2158 |
+
"loss": 99.20048217773437,
|
| 2159 |
+
"step": 3070
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 15.0002,
|
| 2163 |
+
"grad_norm": 0.0,
|
| 2164 |
+
"learning_rate": 0.0002820901803607214,
|
| 2165 |
+
"loss": 103.31695556640625,
|
| 2166 |
+
"step": 3080
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 15.0004,
|
| 2170 |
+
"grad_norm": 0.0,
|
| 2171 |
+
"learning_rate": 0.00028203006012024045,
|
| 2172 |
+
"loss": 99.11974487304687,
|
| 2173 |
+
"step": 3090
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 15.0006,
|
| 2177 |
+
"grad_norm": 0.0,
|
| 2178 |
+
"learning_rate": 0.0002819699398797595,
|
| 2179 |
+
"loss": 99.01674194335938,
|
| 2180 |
+
"step": 3100
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 15.0008,
|
| 2184 |
+
"grad_norm": 0.0,
|
| 2185 |
+
"learning_rate": 0.00028190981963927854,
|
| 2186 |
+
"loss": 97.12745361328125,
|
| 2187 |
+
"step": 3110
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 15.001,
|
| 2191 |
+
"grad_norm": 0.0,
|
| 2192 |
+
"learning_rate": 0.0002818496993987976,
|
| 2193 |
+
"loss": 98.1295654296875,
|
| 2194 |
+
"step": 3120
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 15.0012,
|
| 2198 |
+
"grad_norm": 0.0,
|
| 2199 |
+
"learning_rate": 0.00028178957915831664,
|
| 2200 |
+
"loss": 99.0256591796875,
|
| 2201 |
+
"step": 3130
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 15.0014,
|
| 2205 |
+
"grad_norm": 0.0,
|
| 2206 |
+
"learning_rate": 0.00028172945891783563,
|
| 2207 |
+
"loss": 99.1876220703125,
|
| 2208 |
+
"step": 3140
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 15.0016,
|
| 2212 |
+
"grad_norm": 0.0,
|
| 2213 |
+
"learning_rate": 0.0002816693386773547,
|
| 2214 |
+
"loss": 99.0940673828125,
|
| 2215 |
+
"step": 3150
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 15.0018,
|
| 2219 |
+
"grad_norm": 0.0,
|
| 2220 |
+
"learning_rate": 0.0002816092184368737,
|
| 2221 |
+
"loss": 99.16301879882812,
|
| 2222 |
+
"step": 3160
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 15.002,
|
| 2226 |
+
"grad_norm": 0.0,
|
| 2227 |
+
"learning_rate": 0.0002815490981963928,
|
| 2228 |
+
"loss": 99.16691284179687,
|
| 2229 |
+
"step": 3170
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 16.0001,
|
| 2233 |
+
"grad_norm": 0.0,
|
| 2234 |
+
"learning_rate": 0.00028148897795591177,
|
| 2235 |
+
"loss": 104.3411865234375,
|
| 2236 |
+
"step": 3180
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 16.0003,
|
| 2240 |
+
"grad_norm": 0.0,
|
| 2241 |
+
"learning_rate": 0.0002814288577154308,
|
| 2242 |
+
"loss": 97.95174560546874,
|
| 2243 |
+
"step": 3190
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 16.0005,
|
| 2247 |
+
"grad_norm": 0.0,
|
| 2248 |
+
"learning_rate": 0.0002813687374749499,
|
| 2249 |
+
"loss": 99.18150024414062,
|
| 2250 |
+
"step": 3200
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 16.0007,
|
| 2254 |
+
"grad_norm": 0.0,
|
| 2255 |
+
"learning_rate": 0.0002813086172344689,
|
| 2256 |
+
"loss": 99.10928344726562,
|
| 2257 |
+
"step": 3210
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 16.0009,
|
| 2261 |
+
"grad_norm": 0.0,
|
| 2262 |
+
"learning_rate": 0.00028124849699398796,
|
| 2263 |
+
"loss": 95.95228881835938,
|
| 2264 |
+
"step": 3220
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 16.0011,
|
| 2268 |
+
"grad_norm": 0.0,
|
| 2269 |
+
"learning_rate": 0.000281188376753507,
|
| 2270 |
+
"loss": 99.16500244140624,
|
| 2271 |
+
"step": 3230
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 16.0013,
|
| 2275 |
+
"grad_norm": 0.0,
|
| 2276 |
+
"learning_rate": 0.00028112825651302605,
|
| 2277 |
+
"loss": 99.16441040039062,
|
| 2278 |
+
"step": 3240
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 16.0015,
|
| 2282 |
+
"grad_norm": 0.0,
|
| 2283 |
+
"learning_rate": 0.00028106813627254505,
|
| 2284 |
+
"loss": 99.10928955078126,
|
| 2285 |
+
"step": 3250
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 16.0017,
|
| 2289 |
+
"grad_norm": 0.0,
|
| 2290 |
+
"learning_rate": 0.0002810080160320641,
|
| 2291 |
+
"loss": 99.24561157226563,
|
| 2292 |
+
"step": 3260
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 16.0019,
|
| 2296 |
+
"grad_norm": 0.0,
|
| 2297 |
+
"learning_rate": 0.00028094789579158314,
|
| 2298 |
+
"loss": 99.01414184570312,
|
| 2299 |
+
"step": 3270
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 16.0021,
|
| 2303 |
+
"grad_norm": 0.0,
|
| 2304 |
+
"learning_rate": 0.0002808877755511022,
|
| 2305 |
+
"loss": 99.175,
|
| 2306 |
+
"step": 3280
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 17.0002,
|
| 2310 |
+
"grad_norm": 0.0,
|
| 2311 |
+
"learning_rate": 0.0002808276553106212,
|
| 2312 |
+
"loss": 103.296435546875,
|
| 2313 |
+
"step": 3290
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 17.0004,
|
| 2317 |
+
"grad_norm": 0.0,
|
| 2318 |
+
"learning_rate": 0.0002807675350701403,
|
| 2319 |
+
"loss": 99.10460205078125,
|
| 2320 |
+
"step": 3300
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 17.0006,
|
| 2324 |
+
"grad_norm": 0.0,
|
| 2325 |
+
"learning_rate": 0.0002807074148296593,
|
| 2326 |
+
"loss": 99.00741577148438,
|
| 2327 |
+
"step": 3310
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 17.0008,
|
| 2331 |
+
"grad_norm": 0.0,
|
| 2332 |
+
"learning_rate": 0.0002806472945891783,
|
| 2333 |
+
"loss": 97.1107177734375,
|
| 2334 |
+
"step": 3320
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 17.001,
|
| 2338 |
+
"grad_norm": 0.0,
|
| 2339 |
+
"learning_rate": 0.0002805871743486974,
|
| 2340 |
+
"loss": 98.12017211914062,
|
| 2341 |
+
"step": 3330
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 17.0012,
|
| 2345 |
+
"grad_norm": 0.0,
|
| 2346 |
+
"learning_rate": 0.0002805270541082164,
|
| 2347 |
+
"loss": 99.01578979492187,
|
| 2348 |
+
"step": 3340
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 17.0014,
|
| 2352 |
+
"grad_norm": 0.0,
|
| 2353 |
+
"learning_rate": 0.00028046693386773547,
|
| 2354 |
+
"loss": 99.17923583984376,
|
| 2355 |
+
"step": 3350
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 17.0016,
|
| 2359 |
+
"grad_norm": 0.0,
|
| 2360 |
+
"learning_rate": 0.00028040681362725446,
|
| 2361 |
+
"loss": 99.08976440429687,
|
| 2362 |
+
"step": 3360
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 17.0018,
|
| 2366 |
+
"grad_norm": 0.0,
|
| 2367 |
+
"learning_rate": 0.0002803466933867735,
|
| 2368 |
+
"loss": 99.15631103515625,
|
| 2369 |
+
"step": 3370
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 17.002,
|
| 2373 |
+
"grad_norm": 0.0,
|
| 2374 |
+
"learning_rate": 0.00028028657314629256,
|
| 2375 |
+
"loss": 99.15965576171875,
|
| 2376 |
+
"step": 3380
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 18.0001,
|
| 2380 |
+
"grad_norm": 0.0,
|
| 2381 |
+
"learning_rate": 0.0002802264529058116,
|
| 2382 |
+
"loss": 104.33367919921875,
|
| 2383 |
+
"step": 3390
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 18.0003,
|
| 2387 |
+
"grad_norm": 0.0,
|
| 2388 |
+
"learning_rate": 0.00028016633266533065,
|
| 2389 |
+
"loss": 97.9498779296875,
|
| 2390 |
+
"step": 3400
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 18.0005,
|
| 2394 |
+
"grad_norm": 0.0,
|
| 2395 |
+
"learning_rate": 0.0002801062124248497,
|
| 2396 |
+
"loss": 99.17077026367187,
|
| 2397 |
+
"step": 3410
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 18.0007,
|
| 2401 |
+
"grad_norm": 0.0,
|
| 2402 |
+
"learning_rate": 0.0002800460921843687,
|
| 2403 |
+
"loss": 99.10169677734375,
|
| 2404 |
+
"step": 3420
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 18.0009,
|
| 2408 |
+
"grad_norm": 0.0,
|
| 2409 |
+
"learning_rate": 0.00027998597194388774,
|
| 2410 |
+
"loss": 95.9587158203125,
|
| 2411 |
+
"step": 3430
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 18.0011,
|
| 2415 |
+
"grad_norm": 0.0,
|
| 2416 |
+
"learning_rate": 0.0002799258517034068,
|
| 2417 |
+
"loss": 99.152587890625,
|
| 2418 |
+
"step": 3440
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 18.0013,
|
| 2422 |
+
"grad_norm": 0.0,
|
| 2423 |
+
"learning_rate": 0.00027986573146292584,
|
| 2424 |
+
"loss": 99.15311889648437,
|
| 2425 |
+
"step": 3450
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 18.0015,
|
| 2429 |
+
"grad_norm": 0.0,
|
| 2430 |
+
"learning_rate": 0.0002798056112224449,
|
| 2431 |
+
"loss": 99.09374389648437,
|
| 2432 |
+
"step": 3460
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 18.0017,
|
| 2436 |
+
"grad_norm": 0.0,
|
| 2437 |
+
"learning_rate": 0.0002797454909819639,
|
| 2438 |
+
"loss": 99.24362182617188,
|
| 2439 |
+
"step": 3470
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 18.0019,
|
| 2443 |
+
"grad_norm": 0.0,
|
| 2444 |
+
"learning_rate": 0.0002796853707414829,
|
| 2445 |
+
"loss": 99.01097412109375,
|
| 2446 |
+
"step": 3480
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 18.0021,
|
| 2450 |
+
"grad_norm": 0.0,
|
| 2451 |
+
"learning_rate": 0.000279625250501002,
|
| 2452 |
+
"loss": 99.16884155273438,
|
| 2453 |
+
"step": 3490
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 19.0002,
|
| 2457 |
+
"grad_norm": 0.0,
|
| 2458 |
+
"learning_rate": 0.000279565130260521,
|
| 2459 |
+
"loss": 103.30040283203125,
|
| 2460 |
+
"step": 3500
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 19.0004,
|
| 2464 |
+
"grad_norm": 0.0,
|
| 2465 |
+
"learning_rate": 0.00027950501002004007,
|
| 2466 |
+
"loss": 99.10455932617188,
|
| 2467 |
+
"step": 3510
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 19.0006,
|
| 2471 |
+
"grad_norm": 0.0,
|
| 2472 |
+
"learning_rate": 0.0002794448897795591,
|
| 2473 |
+
"loss": 98.99957275390625,
|
| 2474 |
+
"step": 3520
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 19.0008,
|
| 2478 |
+
"grad_norm": 0.0,
|
| 2479 |
+
"learning_rate": 0.0002793847695390781,
|
| 2480 |
+
"loss": 97.1145751953125,
|
| 2481 |
+
"step": 3530
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 19.001,
|
| 2485 |
+
"grad_norm": 0.0,
|
| 2486 |
+
"learning_rate": 0.00027932464929859716,
|
| 2487 |
+
"loss": 98.1151123046875,
|
| 2488 |
+
"step": 3540
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 19.0012,
|
| 2492 |
+
"grad_norm": 0.0,
|
| 2493 |
+
"learning_rate": 0.0002792645290581162,
|
| 2494 |
+
"loss": 99.0079833984375,
|
| 2495 |
+
"step": 3550
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 19.0014,
|
| 2499 |
+
"grad_norm": 0.0,
|
| 2500 |
+
"learning_rate": 0.00027920440881763525,
|
| 2501 |
+
"loss": 99.1750732421875,
|
| 2502 |
+
"step": 3560
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 19.0016,
|
| 2506 |
+
"grad_norm": 0.0,
|
| 2507 |
+
"learning_rate": 0.00027914428857715425,
|
| 2508 |
+
"loss": 99.08978271484375,
|
| 2509 |
+
"step": 3570
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 19.0018,
|
| 2513 |
+
"grad_norm": 0.0,
|
| 2514 |
+
"learning_rate": 0.0002790841683366733,
|
| 2515 |
+
"loss": 99.1601806640625,
|
| 2516 |
+
"step": 3580
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 19.002,
|
| 2520 |
+
"grad_norm": 0.0,
|
| 2521 |
+
"learning_rate": 0.0002790240480961924,
|
| 2522 |
+
"loss": 99.15270385742187,
|
| 2523 |
+
"step": 3590
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 20.0001,
|
| 2527 |
+
"grad_norm": 0.0,
|
| 2528 |
+
"learning_rate": 0.0002789639278557114,
|
| 2529 |
+
"loss": 104.329833984375,
|
| 2530 |
+
"step": 3600
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 20.0003,
|
| 2534 |
+
"grad_norm": 0.0,
|
| 2535 |
+
"learning_rate": 0.00027890380761523044,
|
| 2536 |
+
"loss": 97.94336547851563,
|
| 2537 |
+
"step": 3610
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 20.0005,
|
| 2541 |
+
"grad_norm": 0.0,
|
| 2542 |
+
"learning_rate": 0.0002788436873747495,
|
| 2543 |
+
"loss": 99.16564331054687,
|
| 2544 |
+
"step": 3620
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 20.0007,
|
| 2548 |
+
"grad_norm": 0.0,
|
| 2549 |
+
"learning_rate": 0.00027878356713426853,
|
| 2550 |
+
"loss": 99.0989990234375,
|
| 2551 |
+
"step": 3630
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 20.0009,
|
| 2555 |
+
"grad_norm": 0.0,
|
| 2556 |
+
"learning_rate": 0.00027872344689378753,
|
| 2557 |
+
"loss": 95.96222534179688,
|
| 2558 |
+
"step": 3640
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 20.0011,
|
| 2562 |
+
"grad_norm": 0.0,
|
| 2563 |
+
"learning_rate": 0.0002786633266533066,
|
| 2564 |
+
"loss": 99.14838256835938,
|
| 2565 |
+
"step": 3650
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 20.0013,
|
| 2569 |
+
"grad_norm": 0.0,
|
| 2570 |
+
"learning_rate": 0.0002786032064128256,
|
| 2571 |
+
"loss": 99.150390625,
|
| 2572 |
+
"step": 3660
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 20.0015,
|
| 2576 |
+
"grad_norm": 0.0,
|
| 2577 |
+
"learning_rate": 0.00027854308617234467,
|
| 2578 |
+
"loss": 99.0933837890625,
|
| 2579 |
+
"step": 3670
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 20.0017,
|
| 2583 |
+
"grad_norm": 0.0,
|
| 2584 |
+
"learning_rate": 0.00027848296593186366,
|
| 2585 |
+
"loss": 99.23428955078126,
|
| 2586 |
+
"step": 3680
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 20.0019,
|
| 2590 |
+
"grad_norm": 0.0,
|
| 2591 |
+
"learning_rate": 0.00027842284569138277,
|
| 2592 |
+
"loss": 99.016259765625,
|
| 2593 |
+
"step": 3690
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 20.0021,
|
| 2597 |
+
"grad_norm": 0.0,
|
| 2598 |
+
"learning_rate": 0.0002783627254509018,
|
| 2599 |
+
"loss": 99.17916870117188,
|
| 2600 |
+
"step": 3700
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 21.0002,
|
| 2604 |
+
"grad_norm": 0.0,
|
| 2605 |
+
"learning_rate": 0.0002783026052104208,
|
| 2606 |
+
"loss": 103.300244140625,
|
| 2607 |
+
"step": 3710
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 21.0004,
|
| 2611 |
+
"grad_norm": 0.0,
|
| 2612 |
+
"learning_rate": 0.00027824248496993985,
|
| 2613 |
+
"loss": 99.09691162109375,
|
| 2614 |
+
"step": 3720
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 21.0006,
|
| 2618 |
+
"grad_norm": 0.0,
|
| 2619 |
+
"learning_rate": 0.0002781823647294589,
|
| 2620 |
+
"loss": 99.0059326171875,
|
| 2621 |
+
"step": 3730
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 21.0008,
|
| 2625 |
+
"grad_norm": 0.0,
|
| 2626 |
+
"learning_rate": 0.00027812224448897795,
|
| 2627 |
+
"loss": 97.12512817382813,
|
| 2628 |
+
"step": 3740
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 21.001,
|
| 2632 |
+
"grad_norm": 0.0,
|
| 2633 |
+
"learning_rate": 0.00027806212424849694,
|
| 2634 |
+
"loss": 98.11283569335937,
|
| 2635 |
+
"step": 3750
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 21.0012,
|
| 2639 |
+
"grad_norm": 0.0,
|
| 2640 |
+
"learning_rate": 0.000278002004008016,
|
| 2641 |
+
"loss": 99.00809326171876,
|
| 2642 |
+
"step": 3760
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 21.0014,
|
| 2646 |
+
"grad_norm": 0.0,
|
| 2647 |
+
"learning_rate": 0.00027794188376753504,
|
| 2648 |
+
"loss": 99.1753662109375,
|
| 2649 |
+
"step": 3770
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 21.0016,
|
| 2653 |
+
"grad_norm": 0.0,
|
| 2654 |
+
"learning_rate": 0.0002778817635270541,
|
| 2655 |
+
"loss": 99.08587036132812,
|
| 2656 |
+
"step": 3780
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 21.0018,
|
| 2660 |
+
"grad_norm": 0.0,
|
| 2661 |
+
"learning_rate": 0.00027782164328657313,
|
| 2662 |
+
"loss": 99.15759887695313,
|
| 2663 |
+
"step": 3790
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 21.002,
|
| 2667 |
+
"grad_norm": 0.0,
|
| 2668 |
+
"learning_rate": 0.0002777615230460922,
|
| 2669 |
+
"loss": 99.1598388671875,
|
| 2670 |
+
"step": 3800
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 22.0001,
|
| 2674 |
+
"grad_norm": 0.0,
|
| 2675 |
+
"learning_rate": 0.00027770140280561123,
|
| 2676 |
+
"loss": 104.32696533203125,
|
| 2677 |
+
"step": 3810
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 22.0003,
|
| 2681 |
+
"grad_norm": 0.0,
|
| 2682 |
+
"learning_rate": 0.0002776412825651302,
|
| 2683 |
+
"loss": 97.95064697265624,
|
| 2684 |
+
"step": 3820
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 22.0005,
|
| 2688 |
+
"grad_norm": 0.0,
|
| 2689 |
+
"learning_rate": 0.00027758116232464927,
|
| 2690 |
+
"loss": 99.16984252929687,
|
| 2691 |
+
"step": 3830
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 22.0007,
|
| 2695 |
+
"grad_norm": 0.0,
|
| 2696 |
+
"learning_rate": 0.0002775210420841683,
|
| 2697 |
+
"loss": 99.10538940429687,
|
| 2698 |
+
"step": 3840
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 22.0009,
|
| 2702 |
+
"grad_norm": 0.0,
|
| 2703 |
+
"learning_rate": 0.00027746092184368737,
|
| 2704 |
+
"loss": 95.97281494140626,
|
| 2705 |
+
"step": 3850
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 22.0011,
|
| 2709 |
+
"grad_norm": 0.0,
|
| 2710 |
+
"learning_rate": 0.00027740080160320636,
|
| 2711 |
+
"loss": 99.15985717773438,
|
| 2712 |
+
"step": 3860
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 22.0013,
|
| 2716 |
+
"grad_norm": 0.0,
|
| 2717 |
+
"learning_rate": 0.0002773406813627254,
|
| 2718 |
+
"loss": 99.1621826171875,
|
| 2719 |
+
"step": 3870
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 22.0015,
|
| 2723 |
+
"grad_norm": 0.0,
|
| 2724 |
+
"learning_rate": 0.00027728056112224446,
|
| 2725 |
+
"loss": 99.09931030273438,
|
| 2726 |
+
"step": 3880
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 22.0017,
|
| 2730 |
+
"grad_norm": 0.0,
|
| 2731 |
+
"learning_rate": 0.0002772204408817635,
|
| 2732 |
+
"loss": 99.2378662109375,
|
| 2733 |
+
"step": 3890
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 22.0019,
|
| 2737 |
+
"grad_norm": 0.0,
|
| 2738 |
+
"learning_rate": 0.00027716032064128255,
|
| 2739 |
+
"loss": 99.01779174804688,
|
| 2740 |
+
"step": 3900
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 22.0021,
|
| 2744 |
+
"grad_norm": 0.0,
|
| 2745 |
+
"learning_rate": 0.0002771002004008016,
|
| 2746 |
+
"loss": 99.17398071289062,
|
| 2747 |
+
"step": 3910
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 23.0002,
|
| 2751 |
+
"grad_norm": 0.0,
|
| 2752 |
+
"learning_rate": 0.00027704008016032065,
|
| 2753 |
+
"loss": 103.30367431640624,
|
| 2754 |
+
"step": 3920
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 23.0004,
|
| 2758 |
+
"grad_norm": 0.0,
|
| 2759 |
+
"learning_rate": 0.00027697995991983964,
|
| 2760 |
+
"loss": 99.10546875,
|
| 2761 |
+
"step": 3930
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 23.0006,
|
| 2765 |
+
"grad_norm": 0.0,
|
| 2766 |
+
"learning_rate": 0.0002769198396793587,
|
| 2767 |
+
"loss": 99.01024780273437,
|
| 2768 |
+
"step": 3940
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 23.0008,
|
| 2772 |
+
"grad_norm": 0.0,
|
| 2773 |
+
"learning_rate": 0.00027685971943887773,
|
| 2774 |
+
"loss": 97.12568969726563,
|
| 2775 |
+
"step": 3950
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 23.001,
|
| 2779 |
+
"grad_norm": 0.0,
|
| 2780 |
+
"learning_rate": 0.0002767995991983968,
|
| 2781 |
+
"loss": 98.12539672851562,
|
| 2782 |
+
"step": 3960
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 23.0012,
|
| 2786 |
+
"grad_norm": 0.0,
|
| 2787 |
+
"learning_rate": 0.0002767394789579158,
|
| 2788 |
+
"loss": 99.0267578125,
|
| 2789 |
+
"step": 3970
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 23.0014,
|
| 2793 |
+
"grad_norm": 0.0,
|
| 2794 |
+
"learning_rate": 0.0002766793587174348,
|
| 2795 |
+
"loss": 99.17804565429688,
|
| 2796 |
+
"step": 3980
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 23.0016,
|
| 2800 |
+
"grad_norm": 0.0,
|
| 2801 |
+
"learning_rate": 0.00027661923847695387,
|
| 2802 |
+
"loss": 99.09422607421875,
|
| 2803 |
+
"step": 3990
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 23.0018,
|
| 2807 |
+
"grad_norm": 0.0,
|
| 2808 |
+
"learning_rate": 0.0002765591182364729,
|
| 2809 |
+
"loss": 99.1647705078125,
|
| 2810 |
+
"step": 4000
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 23.002,
|
| 2814 |
+
"grad_norm": 0.0,
|
| 2815 |
+
"learning_rate": 0.00027649899799599197,
|
| 2816 |
+
"loss": 99.16721801757812,
|
| 2817 |
+
"step": 4010
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 24.0001,
|
| 2821 |
+
"grad_norm": 0.0,
|
| 2822 |
+
"learning_rate": 0.000276438877755511,
|
| 2823 |
+
"loss": 104.34345703125,
|
| 2824 |
+
"step": 4020
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 24.0003,
|
| 2828 |
+
"grad_norm": 0.0,
|
| 2829 |
+
"learning_rate": 0.00027637875751503006,
|
| 2830 |
+
"loss": 97.96480712890624,
|
| 2831 |
+
"step": 4030
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 24.0005,
|
| 2835 |
+
"grad_norm": 0.0,
|
| 2836 |
+
"learning_rate": 0.00027631863727454906,
|
| 2837 |
+
"loss": 99.17822265625,
|
| 2838 |
+
"step": 4040
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 24.0007,
|
| 2842 |
+
"grad_norm": 0.0,
|
| 2843 |
+
"learning_rate": 0.0002762585170340681,
|
| 2844 |
+
"loss": 99.11380615234376,
|
| 2845 |
+
"step": 4050
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 24.0009,
|
| 2849 |
+
"grad_norm": 0.0,
|
| 2850 |
+
"learning_rate": 0.00027619839679358715,
|
| 2851 |
+
"loss": 95.99370727539062,
|
| 2852 |
+
"step": 4060
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 24.0011,
|
| 2856 |
+
"grad_norm": 0.0,
|
| 2857 |
+
"learning_rate": 0.0002761382765531062,
|
| 2858 |
+
"loss": 99.17216186523437,
|
| 2859 |
+
"step": 4070
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 24.0013,
|
| 2863 |
+
"grad_norm": 0.0,
|
| 2864 |
+
"learning_rate": 0.00027607815631262525,
|
| 2865 |
+
"loss": 99.17363891601562,
|
| 2866 |
+
"step": 4080
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 24.0015,
|
| 2870 |
+
"grad_norm": 0.0,
|
| 2871 |
+
"learning_rate": 0.0002760180360721443,
|
| 2872 |
+
"loss": 99.11543579101563,
|
| 2873 |
+
"step": 4090
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 24.0017,
|
| 2877 |
+
"grad_norm": 0.0,
|
| 2878 |
+
"learning_rate": 0.0002759579158316633,
|
| 2879 |
+
"loss": 99.25667724609374,
|
| 2880 |
+
"step": 4100
|
| 2881 |
}
|
| 2882 |
],
|
| 2883 |
"logging_steps": 10,
|
|
|
|
| 2897 |
"attributes": {}
|
| 2898 |
}
|
| 2899 |
},
|
| 2900 |
+
"total_flos": 2.832807993700608e+16,
|
| 2901 |
"train_batch_size": 2,
|
| 2902 |
"trial_name": null,
|
| 2903 |
"trial_params": null
|