Training in progress, step 550, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 80792096
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7f2bba67c1c1484d2bc04c25f371989dddfa7218d9db23366bcfd17cb36894c8
|
| 3 |
size 80792096
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 41460084
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:507ef66950bd52137d14c378fc2f78b5ad9af0c9387506f9ca6699bcba5321d8
|
| 3 |
size 41460084
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b5d154d045ee189af4c648f80535098cfde6139351de9c4d32c890f904602cee
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:789cac547d76905ddc88036180d9f246f307a104c94da93e131a174052f790e8
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 275,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1948,6 +1948,1939 @@
|
|
| 1948 |
"eval_samples_per_second": 9.017,
|
| 1949 |
"eval_steps_per_second": 4.511,
|
| 1950 |
"step": 275
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1951 |
}
|
| 1952 |
],
|
| 1953 |
"logging_steps": 1,
|
|
@@ -1967,7 +3900,7 @@
|
|
| 1967 |
"attributes": {}
|
| 1968 |
}
|
| 1969 |
},
|
| 1970 |
-
"total_flos":
|
| 1971 |
"train_batch_size": 2,
|
| 1972 |
"trial_name": null,
|
| 1973 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.5006257822277848,
|
| 5 |
"eval_steps": 275,
|
| 6 |
+
"global_step": 550,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1948 |
"eval_samples_per_second": 9.017,
|
| 1949 |
"eval_steps_per_second": 4.511,
|
| 1950 |
"step": 275
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 0.251223119808852,
|
| 1954 |
+
"grad_norm": 0.39927324652671814,
|
| 1955 |
+
"learning_rate": 0.00017197404764772805,
|
| 1956 |
+
"loss": 2.1982,
|
| 1957 |
+
"step": 276
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 0.2521333485038116,
|
| 1961 |
+
"grad_norm": 0.4287830889225006,
|
| 1962 |
+
"learning_rate": 0.00017177347024911562,
|
| 1963 |
+
"loss": 2.2733,
|
| 1964 |
+
"step": 277
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 0.2530435771987712,
|
| 1968 |
+
"grad_norm": 0.3960012197494507,
|
| 1969 |
+
"learning_rate": 0.00017157229552971487,
|
| 1970 |
+
"loss": 2.1884,
|
| 1971 |
+
"step": 278
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 0.2539538058937308,
|
| 1975 |
+
"grad_norm": 0.40106138586997986,
|
| 1976 |
+
"learning_rate": 0.00017137052516376345,
|
| 1977 |
+
"loss": 2.1207,
|
| 1978 |
+
"step": 279
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 0.2548640345886904,
|
| 1982 |
+
"grad_norm": 0.4410867393016815,
|
| 1983 |
+
"learning_rate": 0.00017116816083045602,
|
| 1984 |
+
"loss": 2.3589,
|
| 1985 |
+
"step": 280
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 0.25577426328365,
|
| 1989 |
+
"grad_norm": 0.4092939794063568,
|
| 1990 |
+
"learning_rate": 0.0001709652042139306,
|
| 1991 |
+
"loss": 2.0842,
|
| 1992 |
+
"step": 281
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 0.25668449197860965,
|
| 1996 |
+
"grad_norm": 0.40820494294166565,
|
| 1997 |
+
"learning_rate": 0.0001707616570032542,
|
| 1998 |
+
"loss": 2.1658,
|
| 1999 |
+
"step": 282
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 0.25759472067356926,
|
| 2003 |
+
"grad_norm": 0.41664186120033264,
|
| 2004 |
+
"learning_rate": 0.00017055752089240907,
|
| 2005 |
+
"loss": 2.1389,
|
| 2006 |
+
"step": 283
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 0.25850494936852886,
|
| 2010 |
+
"grad_norm": 0.4125240445137024,
|
| 2011 |
+
"learning_rate": 0.00017035279758027832,
|
| 2012 |
+
"loss": 2.0615,
|
| 2013 |
+
"step": 284
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 0.25941517806348846,
|
| 2017 |
+
"grad_norm": 0.42702898383140564,
|
| 2018 |
+
"learning_rate": 0.00017014748877063214,
|
| 2019 |
+
"loss": 2.017,
|
| 2020 |
+
"step": 285
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 0.26032540675844806,
|
| 2024 |
+
"grad_norm": 0.44943541288375854,
|
| 2025 |
+
"learning_rate": 0.00016994159617211317,
|
| 2026 |
+
"loss": 2.1901,
|
| 2027 |
+
"step": 286
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 0.26123563545340767,
|
| 2031 |
+
"grad_norm": 0.4286860227584839,
|
| 2032 |
+
"learning_rate": 0.00016973512149822274,
|
| 2033 |
+
"loss": 2.0643,
|
| 2034 |
+
"step": 287
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 0.26214586414836727,
|
| 2038 |
+
"grad_norm": 0.44938111305236816,
|
| 2039 |
+
"learning_rate": 0.0001695280664673062,
|
| 2040 |
+
"loss": 2.1539,
|
| 2041 |
+
"step": 288
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 0.26305609284332687,
|
| 2045 |
+
"grad_norm": 0.4638296067714691,
|
| 2046 |
+
"learning_rate": 0.0001693204328025389,
|
| 2047 |
+
"loss": 2.291,
|
| 2048 |
+
"step": 289
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 0.2639663215382865,
|
| 2052 |
+
"grad_norm": 0.49295714497566223,
|
| 2053 |
+
"learning_rate": 0.00016911222223191182,
|
| 2054 |
+
"loss": 2.2538,
|
| 2055 |
+
"step": 290
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 0.2648765502332461,
|
| 2059 |
+
"grad_norm": 0.48185715079307556,
|
| 2060 |
+
"learning_rate": 0.00016890343648821697,
|
| 2061 |
+
"loss": 2.2792,
|
| 2062 |
+
"step": 291
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 0.26578677892820574,
|
| 2066 |
+
"grad_norm": 0.4750272035598755,
|
| 2067 |
+
"learning_rate": 0.0001686940773090333,
|
| 2068 |
+
"loss": 2.2774,
|
| 2069 |
+
"step": 292
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 0.26669700762316534,
|
| 2073 |
+
"grad_norm": 0.5073033571243286,
|
| 2074 |
+
"learning_rate": 0.00016848414643671195,
|
| 2075 |
+
"loss": 2.3261,
|
| 2076 |
+
"step": 293
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 0.26760723631812494,
|
| 2080 |
+
"grad_norm": 0.5343424081802368,
|
| 2081 |
+
"learning_rate": 0.00016827364561836187,
|
| 2082 |
+
"loss": 2.4097,
|
| 2083 |
+
"step": 294
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 0.26851746501308454,
|
| 2087 |
+
"grad_norm": 0.5311369895935059,
|
| 2088 |
+
"learning_rate": 0.00016806257660583534,
|
| 2089 |
+
"loss": 2.3821,
|
| 2090 |
+
"step": 295
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 0.26942769370804415,
|
| 2094 |
+
"grad_norm": 0.5551429986953735,
|
| 2095 |
+
"learning_rate": 0.00016785094115571322,
|
| 2096 |
+
"loss": 2.3795,
|
| 2097 |
+
"step": 296
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 0.27033792240300375,
|
| 2101 |
+
"grad_norm": 0.6279783248901367,
|
| 2102 |
+
"learning_rate": 0.0001676387410292906,
|
| 2103 |
+
"loss": 2.435,
|
| 2104 |
+
"step": 297
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 0.27124815109796335,
|
| 2108 |
+
"grad_norm": 0.7317250967025757,
|
| 2109 |
+
"learning_rate": 0.00016742597799256182,
|
| 2110 |
+
"loss": 2.6991,
|
| 2111 |
+
"step": 298
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 0.27215837979292296,
|
| 2115 |
+
"grad_norm": 0.8485302329063416,
|
| 2116 |
+
"learning_rate": 0.000167212653816206,
|
| 2117 |
+
"loss": 2.7005,
|
| 2118 |
+
"step": 299
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 0.27306860848788256,
|
| 2122 |
+
"grad_norm": 1.5959185361862183,
|
| 2123 |
+
"learning_rate": 0.00016699877027557226,
|
| 2124 |
+
"loss": 2.7536,
|
| 2125 |
+
"step": 300
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 0.2739788371828422,
|
| 2129 |
+
"grad_norm": 0.4755174219608307,
|
| 2130 |
+
"learning_rate": 0.00016678432915066488,
|
| 2131 |
+
"loss": 2.5907,
|
| 2132 |
+
"step": 301
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 0.2748890658778018,
|
| 2136 |
+
"grad_norm": 0.45389342308044434,
|
| 2137 |
+
"learning_rate": 0.00016656933222612854,
|
| 2138 |
+
"loss": 2.4622,
|
| 2139 |
+
"step": 302
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 0.2757992945727614,
|
| 2143 |
+
"grad_norm": 0.4949435591697693,
|
| 2144 |
+
"learning_rate": 0.00016635378129123342,
|
| 2145 |
+
"loss": 2.4185,
|
| 2146 |
+
"step": 303
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 0.276709523267721,
|
| 2150 |
+
"grad_norm": 0.4521631896495819,
|
| 2151 |
+
"learning_rate": 0.00016613767813986044,
|
| 2152 |
+
"loss": 2.4918,
|
| 2153 |
+
"step": 304
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 0.2776197519626806,
|
| 2157 |
+
"grad_norm": 0.4228963553905487,
|
| 2158 |
+
"learning_rate": 0.0001659210245704861,
|
| 2159 |
+
"loss": 2.4194,
|
| 2160 |
+
"step": 305
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 0.27852998065764023,
|
| 2164 |
+
"grad_norm": 0.4170341491699219,
|
| 2165 |
+
"learning_rate": 0.00016570382238616777,
|
| 2166 |
+
"loss": 2.4185,
|
| 2167 |
+
"step": 306
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"epoch": 0.27944020935259983,
|
| 2171 |
+
"grad_norm": 0.4123315215110779,
|
| 2172 |
+
"learning_rate": 0.00016548607339452853,
|
| 2173 |
+
"loss": 2.3737,
|
| 2174 |
+
"step": 307
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 0.28035043804755944,
|
| 2178 |
+
"grad_norm": 0.4320162832736969,
|
| 2179 |
+
"learning_rate": 0.00016526777940774204,
|
| 2180 |
+
"loss": 2.3317,
|
| 2181 |
+
"step": 308
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 0.28126066674251904,
|
| 2185 |
+
"grad_norm": 0.4118390381336212,
|
| 2186 |
+
"learning_rate": 0.00016504894224251778,
|
| 2187 |
+
"loss": 2.3786,
|
| 2188 |
+
"step": 309
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"epoch": 0.28217089543747864,
|
| 2192 |
+
"grad_norm": 0.39763331413269043,
|
| 2193 |
+
"learning_rate": 0.0001648295637200856,
|
| 2194 |
+
"loss": 2.2968,
|
| 2195 |
+
"step": 310
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 0.2830811241324383,
|
| 2199 |
+
"grad_norm": 0.4391527473926544,
|
| 2200 |
+
"learning_rate": 0.0001646096456661807,
|
| 2201 |
+
"loss": 2.3764,
|
| 2202 |
+
"step": 311
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 0.2839913528273979,
|
| 2206 |
+
"grad_norm": 0.43077877163887024,
|
| 2207 |
+
"learning_rate": 0.00016438918991102842,
|
| 2208 |
+
"loss": 2.2013,
|
| 2209 |
+
"step": 312
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 0.2849015815223575,
|
| 2213 |
+
"grad_norm": 0.43149155378341675,
|
| 2214 |
+
"learning_rate": 0.000164168198289329,
|
| 2215 |
+
"loss": 2.3097,
|
| 2216 |
+
"step": 313
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 0.2858118102173171,
|
| 2220 |
+
"grad_norm": 0.40134817361831665,
|
| 2221 |
+
"learning_rate": 0.00016394667264024246,
|
| 2222 |
+
"loss": 2.3306,
|
| 2223 |
+
"step": 314
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 0.2867220389122767,
|
| 2227 |
+
"grad_norm": 0.4056681990623474,
|
| 2228 |
+
"learning_rate": 0.00016372461480737297,
|
| 2229 |
+
"loss": 2.3146,
|
| 2230 |
+
"step": 315
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 0.2876322676072363,
|
| 2234 |
+
"grad_norm": 0.41738027334213257,
|
| 2235 |
+
"learning_rate": 0.00016350202663875386,
|
| 2236 |
+
"loss": 1.9997,
|
| 2237 |
+
"step": 316
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 0.2885424963021959,
|
| 2241 |
+
"grad_norm": 0.38182246685028076,
|
| 2242 |
+
"learning_rate": 0.00016327890998683192,
|
| 2243 |
+
"loss": 2.0466,
|
| 2244 |
+
"step": 317
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 0.2894527249971555,
|
| 2248 |
+
"grad_norm": 0.39759719371795654,
|
| 2249 |
+
"learning_rate": 0.00016305526670845226,
|
| 2250 |
+
"loss": 2.1788,
|
| 2251 |
+
"step": 318
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 0.2903629536921151,
|
| 2255 |
+
"grad_norm": 0.3982352614402771,
|
| 2256 |
+
"learning_rate": 0.0001628310986648427,
|
| 2257 |
+
"loss": 2.2115,
|
| 2258 |
+
"step": 319
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 0.2912731823870747,
|
| 2262 |
+
"grad_norm": 0.41679051518440247,
|
| 2263 |
+
"learning_rate": 0.0001626064077215983,
|
| 2264 |
+
"loss": 2.3036,
|
| 2265 |
+
"step": 320
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 0.2921834110820344,
|
| 2269 |
+
"grad_norm": 0.40436604619026184,
|
| 2270 |
+
"learning_rate": 0.00016238119574866588,
|
| 2271 |
+
"loss": 2.1493,
|
| 2272 |
+
"step": 321
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 0.293093639776994,
|
| 2276 |
+
"grad_norm": 0.4502476751804352,
|
| 2277 |
+
"learning_rate": 0.0001621554646203284,
|
| 2278 |
+
"loss": 1.8572,
|
| 2279 |
+
"step": 322
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 0.2940038684719536,
|
| 2283 |
+
"grad_norm": 0.44303473830223083,
|
| 2284 |
+
"learning_rate": 0.00016192921621518944,
|
| 2285 |
+
"loss": 2.1832,
|
| 2286 |
+
"step": 323
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 0.2949140971669132,
|
| 2290 |
+
"grad_norm": 0.4064692258834839,
|
| 2291 |
+
"learning_rate": 0.0001617024524161574,
|
| 2292 |
+
"loss": 2.2656,
|
| 2293 |
+
"step": 324
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 0.2958243258618728,
|
| 2297 |
+
"grad_norm": 0.4479392170906067,
|
| 2298 |
+
"learning_rate": 0.0001614751751104301,
|
| 2299 |
+
"loss": 2.2462,
|
| 2300 |
+
"step": 325
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 0.2967345545568324,
|
| 2304 |
+
"grad_norm": 0.4629363715648651,
|
| 2305 |
+
"learning_rate": 0.0001612473861894788,
|
| 2306 |
+
"loss": 1.9715,
|
| 2307 |
+
"step": 326
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 0.297644783251792,
|
| 2311 |
+
"grad_norm": 0.3991665542125702,
|
| 2312 |
+
"learning_rate": 0.00016101908754903268,
|
| 2313 |
+
"loss": 2.0642,
|
| 2314 |
+
"step": 327
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 0.2985550119467516,
|
| 2318 |
+
"grad_norm": 0.42503711581230164,
|
| 2319 |
+
"learning_rate": 0.00016079028108906282,
|
| 2320 |
+
"loss": 2.1403,
|
| 2321 |
+
"step": 328
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 0.2994652406417112,
|
| 2325 |
+
"grad_norm": 0.4499455392360687,
|
| 2326 |
+
"learning_rate": 0.00016056096871376667,
|
| 2327 |
+
"loss": 2.0534,
|
| 2328 |
+
"step": 329
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 0.30037546933667086,
|
| 2332 |
+
"grad_norm": 0.4549277424812317,
|
| 2333 |
+
"learning_rate": 0.00016033115233155202,
|
| 2334 |
+
"loss": 2.2083,
|
| 2335 |
+
"step": 330
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 0.30128569803163047,
|
| 2339 |
+
"grad_norm": 0.3974262773990631,
|
| 2340 |
+
"learning_rate": 0.0001601008338550211,
|
| 2341 |
+
"loss": 2.0156,
|
| 2342 |
+
"step": 331
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 0.30219592672659007,
|
| 2346 |
+
"grad_norm": 0.43566057085990906,
|
| 2347 |
+
"learning_rate": 0.00015987001520095478,
|
| 2348 |
+
"loss": 2.1801,
|
| 2349 |
+
"step": 332
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 0.3031061554215497,
|
| 2353 |
+
"grad_norm": 0.47677701711654663,
|
| 2354 |
+
"learning_rate": 0.00015963869829029658,
|
| 2355 |
+
"loss": 2.1415,
|
| 2356 |
+
"step": 333
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 0.3040163841165093,
|
| 2360 |
+
"grad_norm": 0.4603672921657562,
|
| 2361 |
+
"learning_rate": 0.00015940688504813662,
|
| 2362 |
+
"loss": 2.2967,
|
| 2363 |
+
"step": 334
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 0.3049266128114689,
|
| 2367 |
+
"grad_norm": 0.4428515136241913,
|
| 2368 |
+
"learning_rate": 0.00015917457740369565,
|
| 2369 |
+
"loss": 2.1447,
|
| 2370 |
+
"step": 335
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 0.3058368415064285,
|
| 2374 |
+
"grad_norm": 0.4379275441169739,
|
| 2375 |
+
"learning_rate": 0.000158941777290309,
|
| 2376 |
+
"loss": 2.0957,
|
| 2377 |
+
"step": 336
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 0.3067470702013881,
|
| 2381 |
+
"grad_norm": 0.4831966459751129,
|
| 2382 |
+
"learning_rate": 0.00015870848664541044,
|
| 2383 |
+
"loss": 2.2457,
|
| 2384 |
+
"step": 337
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 0.3076572988963477,
|
| 2388 |
+
"grad_norm": 0.45160865783691406,
|
| 2389 |
+
"learning_rate": 0.00015847470741051618,
|
| 2390 |
+
"loss": 2.1441,
|
| 2391 |
+
"step": 338
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 0.3085675275913073,
|
| 2395 |
+
"grad_norm": 0.44453370571136475,
|
| 2396 |
+
"learning_rate": 0.00015824044153120852,
|
| 2397 |
+
"loss": 2.1073,
|
| 2398 |
+
"step": 339
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 0.30947775628626695,
|
| 2402 |
+
"grad_norm": 0.49965375661849976,
|
| 2403 |
+
"learning_rate": 0.00015800569095711982,
|
| 2404 |
+
"loss": 2.1574,
|
| 2405 |
+
"step": 340
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 0.31038798498122655,
|
| 2409 |
+
"grad_norm": 0.48138341307640076,
|
| 2410 |
+
"learning_rate": 0.00015777045764191625,
|
| 2411 |
+
"loss": 2.0205,
|
| 2412 |
+
"step": 341
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 0.31129821367618615,
|
| 2416 |
+
"grad_norm": 0.5034924745559692,
|
| 2417 |
+
"learning_rate": 0.00015753474354328142,
|
| 2418 |
+
"loss": 2.2319,
|
| 2419 |
+
"step": 342
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 0.31220844237114576,
|
| 2423 |
+
"grad_norm": 0.5034711956977844,
|
| 2424 |
+
"learning_rate": 0.00015729855062290022,
|
| 2425 |
+
"loss": 2.4066,
|
| 2426 |
+
"step": 343
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 0.31311867106610536,
|
| 2430 |
+
"grad_norm": 0.5409703254699707,
|
| 2431 |
+
"learning_rate": 0.00015706188084644242,
|
| 2432 |
+
"loss": 2.2435,
|
| 2433 |
+
"step": 344
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 0.31402889976106496,
|
| 2437 |
+
"grad_norm": 0.544597327709198,
|
| 2438 |
+
"learning_rate": 0.00015682473618354635,
|
| 2439 |
+
"loss": 2.2625,
|
| 2440 |
+
"step": 345
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 0.31493912845602456,
|
| 2444 |
+
"grad_norm": 0.6114000082015991,
|
| 2445 |
+
"learning_rate": 0.0001565871186078025,
|
| 2446 |
+
"loss": 2.4302,
|
| 2447 |
+
"step": 346
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 0.31584935715098417,
|
| 2451 |
+
"grad_norm": 0.6364843845367432,
|
| 2452 |
+
"learning_rate": 0.00015634903009673705,
|
| 2453 |
+
"loss": 2.5153,
|
| 2454 |
+
"step": 347
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 0.31675958584594377,
|
| 2458 |
+
"grad_norm": 0.7510351538658142,
|
| 2459 |
+
"learning_rate": 0.00015611047263179548,
|
| 2460 |
+
"loss": 2.5605,
|
| 2461 |
+
"step": 348
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 0.31766981454090343,
|
| 2465 |
+
"grad_norm": 0.8501291275024414,
|
| 2466 |
+
"learning_rate": 0.000155871448198326,
|
| 2467 |
+
"loss": 2.6519,
|
| 2468 |
+
"step": 349
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 0.31858004323586303,
|
| 2472 |
+
"grad_norm": 1.7441632747650146,
|
| 2473 |
+
"learning_rate": 0.0001556319587855631,
|
| 2474 |
+
"loss": 2.7517,
|
| 2475 |
+
"step": 350
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 0.31949027193082263,
|
| 2479 |
+
"grad_norm": 0.5301811695098877,
|
| 2480 |
+
"learning_rate": 0.00015539200638661104,
|
| 2481 |
+
"loss": 2.6647,
|
| 2482 |
+
"step": 351
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 0.32040050062578224,
|
| 2486 |
+
"grad_norm": 0.5063616633415222,
|
| 2487 |
+
"learning_rate": 0.00015515159299842707,
|
| 2488 |
+
"loss": 2.4961,
|
| 2489 |
+
"step": 352
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 0.32131072932074184,
|
| 2493 |
+
"grad_norm": 0.4843781590461731,
|
| 2494 |
+
"learning_rate": 0.00015491072062180503,
|
| 2495 |
+
"loss": 2.496,
|
| 2496 |
+
"step": 353
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 0.32222095801570144,
|
| 2500 |
+
"grad_norm": 0.4524553716182709,
|
| 2501 |
+
"learning_rate": 0.00015466939126135856,
|
| 2502 |
+
"loss": 2.448,
|
| 2503 |
+
"step": 354
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 0.32313118671066104,
|
| 2507 |
+
"grad_norm": 0.43678200244903564,
|
| 2508 |
+
"learning_rate": 0.00015442760692550443,
|
| 2509 |
+
"loss": 2.2687,
|
| 2510 |
+
"step": 355
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 0.32404141540562065,
|
| 2514 |
+
"grad_norm": 0.4301970303058624,
|
| 2515 |
+
"learning_rate": 0.00015418536962644592,
|
| 2516 |
+
"loss": 2.4826,
|
| 2517 |
+
"step": 356
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 0.32495164410058025,
|
| 2521 |
+
"grad_norm": 0.42540326714515686,
|
| 2522 |
+
"learning_rate": 0.00015394268138015598,
|
| 2523 |
+
"loss": 2.4205,
|
| 2524 |
+
"step": 357
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 0.32586187279553985,
|
| 2528 |
+
"grad_norm": 0.4173906445503235,
|
| 2529 |
+
"learning_rate": 0.00015369954420636048,
|
| 2530 |
+
"loss": 2.394,
|
| 2531 |
+
"step": 358
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 0.3267721014904995,
|
| 2535 |
+
"grad_norm": 0.43184736371040344,
|
| 2536 |
+
"learning_rate": 0.00015345596012852138,
|
| 2537 |
+
"loss": 2.3504,
|
| 2538 |
+
"step": 359
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 0.3276823301854591,
|
| 2542 |
+
"grad_norm": 0.4002053141593933,
|
| 2543 |
+
"learning_rate": 0.00015321193117381996,
|
| 2544 |
+
"loss": 2.2951,
|
| 2545 |
+
"step": 360
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 0.3285925588804187,
|
| 2549 |
+
"grad_norm": 0.39067134261131287,
|
| 2550 |
+
"learning_rate": 0.00015296745937313987,
|
| 2551 |
+
"loss": 2.2768,
|
| 2552 |
+
"step": 361
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 0.3295027875753783,
|
| 2556 |
+
"grad_norm": 0.40051525831222534,
|
| 2557 |
+
"learning_rate": 0.00015272254676105025,
|
| 2558 |
+
"loss": 2.2235,
|
| 2559 |
+
"step": 362
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 0.3304130162703379,
|
| 2563 |
+
"grad_norm": 0.3954068422317505,
|
| 2564 |
+
"learning_rate": 0.00015247719537578883,
|
| 2565 |
+
"loss": 2.2502,
|
| 2566 |
+
"step": 363
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 0.3313232449652975,
|
| 2570 |
+
"grad_norm": 0.4123362600803375,
|
| 2571 |
+
"learning_rate": 0.00015223140725924495,
|
| 2572 |
+
"loss": 2.3309,
|
| 2573 |
+
"step": 364
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 0.33223347366025713,
|
| 2577 |
+
"grad_norm": 0.4138774871826172,
|
| 2578 |
+
"learning_rate": 0.00015198518445694255,
|
| 2579 |
+
"loss": 2.4107,
|
| 2580 |
+
"step": 365
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 0.33314370235521673,
|
| 2584 |
+
"grad_norm": 0.3983847498893738,
|
| 2585 |
+
"learning_rate": 0.0001517385290180231,
|
| 2586 |
+
"loss": 2.2718,
|
| 2587 |
+
"step": 366
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 0.33405393105017633,
|
| 2591 |
+
"grad_norm": 0.36962834000587463,
|
| 2592 |
+
"learning_rate": 0.00015149144299522873,
|
| 2593 |
+
"loss": 2.1744,
|
| 2594 |
+
"step": 367
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 0.334964159745136,
|
| 2598 |
+
"grad_norm": 0.37924104928970337,
|
| 2599 |
+
"learning_rate": 0.0001512439284448849,
|
| 2600 |
+
"loss": 2.1451,
|
| 2601 |
+
"step": 368
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 0.3358743884400956,
|
| 2605 |
+
"grad_norm": 0.39990487694740295,
|
| 2606 |
+
"learning_rate": 0.0001509959874268835,
|
| 2607 |
+
"loss": 2.2508,
|
| 2608 |
+
"step": 369
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 0.3367846171350552,
|
| 2612 |
+
"grad_norm": 0.3862214684486389,
|
| 2613 |
+
"learning_rate": 0.00015074762200466556,
|
| 2614 |
+
"loss": 2.1483,
|
| 2615 |
+
"step": 370
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 0.3376948458300148,
|
| 2619 |
+
"grad_norm": 0.4037676751613617,
|
| 2620 |
+
"learning_rate": 0.00015049883424520414,
|
| 2621 |
+
"loss": 2.2179,
|
| 2622 |
+
"step": 371
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 0.3386050745249744,
|
| 2626 |
+
"grad_norm": 0.40439948439598083,
|
| 2627 |
+
"learning_rate": 0.00015024962621898715,
|
| 2628 |
+
"loss": 2.2054,
|
| 2629 |
+
"step": 372
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 0.339515303219934,
|
| 2633 |
+
"grad_norm": 0.3871942460536957,
|
| 2634 |
+
"learning_rate": 0.00015000000000000001,
|
| 2635 |
+
"loss": 2.129,
|
| 2636 |
+
"step": 373
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 0.3404255319148936,
|
| 2640 |
+
"grad_norm": 0.4091387093067169,
|
| 2641 |
+
"learning_rate": 0.00014974995766570855,
|
| 2642 |
+
"loss": 2.1395,
|
| 2643 |
+
"step": 374
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 0.3413357606098532,
|
| 2647 |
+
"grad_norm": 0.4097527265548706,
|
| 2648 |
+
"learning_rate": 0.00014949950129704162,
|
| 2649 |
+
"loss": 2.1789,
|
| 2650 |
+
"step": 375
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 0.3422459893048128,
|
| 2654 |
+
"grad_norm": 0.4139934480190277,
|
| 2655 |
+
"learning_rate": 0.00014924863297837378,
|
| 2656 |
+
"loss": 2.0611,
|
| 2657 |
+
"step": 376
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 0.3431562179997724,
|
| 2661 |
+
"grad_norm": 0.4146927297115326,
|
| 2662 |
+
"learning_rate": 0.00014899735479750794,
|
| 2663 |
+
"loss": 2.2488,
|
| 2664 |
+
"step": 377
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 0.3440664466947321,
|
| 2668 |
+
"grad_norm": 0.4194958209991455,
|
| 2669 |
+
"learning_rate": 0.00014874566884565807,
|
| 2670 |
+
"loss": 2.0164,
|
| 2671 |
+
"step": 378
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 0.3449766753896917,
|
| 2675 |
+
"grad_norm": 0.41280898451805115,
|
| 2676 |
+
"learning_rate": 0.00014849357721743168,
|
| 2677 |
+
"loss": 2.1503,
|
| 2678 |
+
"step": 379
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 0.3458869040846513,
|
| 2682 |
+
"grad_norm": 0.4133208692073822,
|
| 2683 |
+
"learning_rate": 0.00014824108201081247,
|
| 2684 |
+
"loss": 2.0895,
|
| 2685 |
+
"step": 380
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 0.3467971327796109,
|
| 2689 |
+
"grad_norm": 0.41347819566726685,
|
| 2690 |
+
"learning_rate": 0.00014798818532714279,
|
| 2691 |
+
"loss": 2.0479,
|
| 2692 |
+
"step": 381
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 0.3477073614745705,
|
| 2696 |
+
"grad_norm": 0.43102580308914185,
|
| 2697 |
+
"learning_rate": 0.00014773488927110633,
|
| 2698 |
+
"loss": 2.1458,
|
| 2699 |
+
"step": 382
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 0.3486175901695301,
|
| 2703 |
+
"grad_norm": 0.41427451372146606,
|
| 2704 |
+
"learning_rate": 0.00014748119595071034,
|
| 2705 |
+
"loss": 1.9396,
|
| 2706 |
+
"step": 383
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 0.3495278188644897,
|
| 2710 |
+
"grad_norm": 0.46386152505874634,
|
| 2711 |
+
"learning_rate": 0.0001472271074772683,
|
| 2712 |
+
"loss": 2.2446,
|
| 2713 |
+
"step": 384
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 0.3504380475594493,
|
| 2717 |
+
"grad_norm": 0.4310764670372009,
|
| 2718 |
+
"learning_rate": 0.00014697262596538227,
|
| 2719 |
+
"loss": 2.2144,
|
| 2720 |
+
"step": 385
|
| 2721 |
+
},
|
| 2722 |
+
{
|
| 2723 |
+
"epoch": 0.3513482762544089,
|
| 2724 |
+
"grad_norm": 0.4956878423690796,
|
| 2725 |
+
"learning_rate": 0.00014671775353292525,
|
| 2726 |
+
"loss": 2.1875,
|
| 2727 |
+
"step": 386
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 0.35225850494936856,
|
| 2731 |
+
"grad_norm": 0.4793931543827057,
|
| 2732 |
+
"learning_rate": 0.00014646249230102366,
|
| 2733 |
+
"loss": 2.2733,
|
| 2734 |
+
"step": 387
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 0.35316873364432816,
|
| 2738 |
+
"grad_norm": 0.46217313408851624,
|
| 2739 |
+
"learning_rate": 0.00014620684439403962,
|
| 2740 |
+
"loss": 2.2812,
|
| 2741 |
+
"step": 388
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 0.35407896233928776,
|
| 2745 |
+
"grad_norm": 0.4721885323524475,
|
| 2746 |
+
"learning_rate": 0.00014595081193955324,
|
| 2747 |
+
"loss": 2.1223,
|
| 2748 |
+
"step": 389
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 0.35498919103424736,
|
| 2752 |
+
"grad_norm": 0.49550965428352356,
|
| 2753 |
+
"learning_rate": 0.000145694397068345,
|
| 2754 |
+
"loss": 2.156,
|
| 2755 |
+
"step": 390
|
| 2756 |
+
},
|
| 2757 |
+
{
|
| 2758 |
+
"epoch": 0.35589941972920697,
|
| 2759 |
+
"grad_norm": 0.5109139084815979,
|
| 2760 |
+
"learning_rate": 0.0001454376019143779,
|
| 2761 |
+
"loss": 2.1494,
|
| 2762 |
+
"step": 391
|
| 2763 |
+
},
|
| 2764 |
+
{
|
| 2765 |
+
"epoch": 0.35680964842416657,
|
| 2766 |
+
"grad_norm": 0.4725574553012848,
|
| 2767 |
+
"learning_rate": 0.00014518042861477986,
|
| 2768 |
+
"loss": 2.1793,
|
| 2769 |
+
"step": 392
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 0.3577198771191262,
|
| 2773 |
+
"grad_norm": 0.4739914536476135,
|
| 2774 |
+
"learning_rate": 0.00014492287930982576,
|
| 2775 |
+
"loss": 2.1763,
|
| 2776 |
+
"step": 393
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 0.3586301058140858,
|
| 2780 |
+
"grad_norm": 0.5420114994049072,
|
| 2781 |
+
"learning_rate": 0.00014466495614291977,
|
| 2782 |
+
"loss": 2.4521,
|
| 2783 |
+
"step": 394
|
| 2784 |
+
},
|
| 2785 |
+
{
|
| 2786 |
+
"epoch": 0.3595403345090454,
|
| 2787 |
+
"grad_norm": 0.5225427150726318,
|
| 2788 |
+
"learning_rate": 0.00014440666126057744,
|
| 2789 |
+
"loss": 2.372,
|
| 2790 |
+
"step": 395
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 0.360450563204005,
|
| 2794 |
+
"grad_norm": 0.5337964296340942,
|
| 2795 |
+
"learning_rate": 0.0001441479968124078,
|
| 2796 |
+
"loss": 2.397,
|
| 2797 |
+
"step": 396
|
| 2798 |
+
},
|
| 2799 |
+
{
|
| 2800 |
+
"epoch": 0.36136079189896464,
|
| 2801 |
+
"grad_norm": 0.5906230807304382,
|
| 2802 |
+
"learning_rate": 0.0001438889649510956,
|
| 2803 |
+
"loss": 2.506,
|
| 2804 |
+
"step": 397
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 0.36227102059392424,
|
| 2808 |
+
"grad_norm": 0.6578875780105591,
|
| 2809 |
+
"learning_rate": 0.00014362956783238324,
|
| 2810 |
+
"loss": 2.6408,
|
| 2811 |
+
"step": 398
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 0.36318124928888385,
|
| 2815 |
+
"grad_norm": 0.7982918620109558,
|
| 2816 |
+
"learning_rate": 0.00014336980761505297,
|
| 2817 |
+
"loss": 2.6612,
|
| 2818 |
+
"step": 399
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 0.36409147798384345,
|
| 2822 |
+
"grad_norm": 1.4390262365341187,
|
| 2823 |
+
"learning_rate": 0.00014310968646090883,
|
| 2824 |
+
"loss": 2.7073,
|
| 2825 |
+
"step": 400
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 0.36500170667880305,
|
| 2829 |
+
"grad_norm": 0.5260487198829651,
|
| 2830 |
+
"learning_rate": 0.00014284920653475866,
|
| 2831 |
+
"loss": 2.6269,
|
| 2832 |
+
"step": 401
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 0.36591193537376265,
|
| 2836 |
+
"grad_norm": 0.4492892026901245,
|
| 2837 |
+
"learning_rate": 0.00014258837000439618,
|
| 2838 |
+
"loss": 2.3863,
|
| 2839 |
+
"step": 402
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 0.36682216406872226,
|
| 2843 |
+
"grad_norm": 0.4619944095611572,
|
| 2844 |
+
"learning_rate": 0.0001423271790405828,
|
| 2845 |
+
"loss": 2.4595,
|
| 2846 |
+
"step": 403
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 0.36773239276368186,
|
| 2850 |
+
"grad_norm": 0.4437786638736725,
|
| 2851 |
+
"learning_rate": 0.00014206563581702964,
|
| 2852 |
+
"loss": 2.3674,
|
| 2853 |
+
"step": 404
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 0.36864262145864146,
|
| 2857 |
+
"grad_norm": 0.4789164364337921,
|
| 2858 |
+
"learning_rate": 0.0001418037425103795,
|
| 2859 |
+
"loss": 2.5203,
|
| 2860 |
+
"step": 405
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 0.36955285015360106,
|
| 2864 |
+
"grad_norm": 0.44783228635787964,
|
| 2865 |
+
"learning_rate": 0.00014154150130018866,
|
| 2866 |
+
"loss": 2.5183,
|
| 2867 |
+
"step": 406
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 0.3704630788485607,
|
| 2871 |
+
"grad_norm": 0.40067169070243835,
|
| 2872 |
+
"learning_rate": 0.00014127891436890868,
|
| 2873 |
+
"loss": 2.3846,
|
| 2874 |
+
"step": 407
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 0.3713733075435203,
|
| 2878 |
+
"grad_norm": 0.3978015184402466,
|
| 2879 |
+
"learning_rate": 0.0001410159839018684,
|
| 2880 |
+
"loss": 2.3146,
|
| 2881 |
+
"step": 408
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 0.37228353623847993,
|
| 2885 |
+
"grad_norm": 0.4096076190471649,
|
| 2886 |
+
"learning_rate": 0.0001407527120872557,
|
| 2887 |
+
"loss": 2.3617,
|
| 2888 |
+
"step": 409
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 0.37319376493343953,
|
| 2892 |
+
"grad_norm": 0.4160764217376709,
|
| 2893 |
+
"learning_rate": 0.00014048910111609915,
|
| 2894 |
+
"loss": 2.2909,
|
| 2895 |
+
"step": 410
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"epoch": 0.37410399362839913,
|
| 2899 |
+
"grad_norm": 0.3976461887359619,
|
| 2900 |
+
"learning_rate": 0.0001402251531822499,
|
| 2901 |
+
"loss": 2.3111,
|
| 2902 |
+
"step": 411
|
| 2903 |
+
},
|
| 2904 |
+
{
|
| 2905 |
+
"epoch": 0.37501422232335874,
|
| 2906 |
+
"grad_norm": 0.3890199065208435,
|
| 2907 |
+
"learning_rate": 0.00013996087048236358,
|
| 2908 |
+
"loss": 2.0969,
|
| 2909 |
+
"step": 412
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 0.37592445101831834,
|
| 2913 |
+
"grad_norm": 0.4157082140445709,
|
| 2914 |
+
"learning_rate": 0.00013969625521588158,
|
| 2915 |
+
"loss": 2.3205,
|
| 2916 |
+
"step": 413
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 0.37683467971327794,
|
| 2920 |
+
"grad_norm": 0.4103608727455139,
|
| 2921 |
+
"learning_rate": 0.00013943130958501317,
|
| 2922 |
+
"loss": 2.2622,
|
| 2923 |
+
"step": 414
|
| 2924 |
+
},
|
| 2925 |
+
{
|
| 2926 |
+
"epoch": 0.37774490840823755,
|
| 2927 |
+
"grad_norm": 0.40916207432746887,
|
| 2928 |
+
"learning_rate": 0.00013916603579471705,
|
| 2929 |
+
"loss": 2.3585,
|
| 2930 |
+
"step": 415
|
| 2931 |
+
},
|
| 2932 |
+
{
|
| 2933 |
+
"epoch": 0.3786551371031972,
|
| 2934 |
+
"grad_norm": 0.39642858505249023,
|
| 2935 |
+
"learning_rate": 0.00013890043605268283,
|
| 2936 |
+
"loss": 2.2196,
|
| 2937 |
+
"step": 416
|
| 2938 |
+
},
|
| 2939 |
+
{
|
| 2940 |
+
"epoch": 0.3795653657981568,
|
| 2941 |
+
"grad_norm": 0.3851282596588135,
|
| 2942 |
+
"learning_rate": 0.00013863451256931287,
|
| 2943 |
+
"loss": 2.0298,
|
| 2944 |
+
"step": 417
|
| 2945 |
+
},
|
| 2946 |
+
{
|
| 2947 |
+
"epoch": 0.3804755944931164,
|
| 2948 |
+
"grad_norm": 0.38890305161476135,
|
| 2949 |
+
"learning_rate": 0.00013836826755770384,
|
| 2950 |
+
"loss": 2.1601,
|
| 2951 |
+
"step": 418
|
| 2952 |
+
},
|
| 2953 |
+
{
|
| 2954 |
+
"epoch": 0.381385823188076,
|
| 2955 |
+
"grad_norm": 0.41382652521133423,
|
| 2956 |
+
"learning_rate": 0.00013810170323362816,
|
| 2957 |
+
"loss": 2.2656,
|
| 2958 |
+
"step": 419
|
| 2959 |
+
},
|
| 2960 |
+
{
|
| 2961 |
+
"epoch": 0.3822960518830356,
|
| 2962 |
+
"grad_norm": 0.3820722699165344,
|
| 2963 |
+
"learning_rate": 0.0001378348218155158,
|
| 2964 |
+
"loss": 2.0094,
|
| 2965 |
+
"step": 420
|
| 2966 |
+
},
|
| 2967 |
+
{
|
| 2968 |
+
"epoch": 0.3832062805779952,
|
| 2969 |
+
"grad_norm": 0.4150048494338989,
|
| 2970 |
+
"learning_rate": 0.00013756762552443553,
|
| 2971 |
+
"loss": 2.2529,
|
| 2972 |
+
"step": 421
|
| 2973 |
+
},
|
| 2974 |
+
{
|
| 2975 |
+
"epoch": 0.3841165092729548,
|
| 2976 |
+
"grad_norm": 0.452776700258255,
|
| 2977 |
+
"learning_rate": 0.00013730011658407676,
|
| 2978 |
+
"loss": 2.1972,
|
| 2979 |
+
"step": 422
|
| 2980 |
+
},
|
| 2981 |
+
{
|
| 2982 |
+
"epoch": 0.3850267379679144,
|
| 2983 |
+
"grad_norm": 0.4173040986061096,
|
| 2984 |
+
"learning_rate": 0.00013703229722073065,
|
| 2985 |
+
"loss": 2.1502,
|
| 2986 |
+
"step": 423
|
| 2987 |
+
},
|
| 2988 |
+
{
|
| 2989 |
+
"epoch": 0.385936966662874,
|
| 2990 |
+
"grad_norm": 0.4115488529205322,
|
| 2991 |
+
"learning_rate": 0.000136764169663272,
|
| 2992 |
+
"loss": 1.9828,
|
| 2993 |
+
"step": 424
|
| 2994 |
+
},
|
| 2995 |
+
{
|
| 2996 |
+
"epoch": 0.38684719535783363,
|
| 2997 |
+
"grad_norm": 0.4060666561126709,
|
| 2998 |
+
"learning_rate": 0.00013649573614314044,
|
| 2999 |
+
"loss": 2.267,
|
| 3000 |
+
"step": 425
|
| 3001 |
+
},
|
| 3002 |
+
{
|
| 3003 |
+
"epoch": 0.3877574240527933,
|
| 3004 |
+
"grad_norm": 0.4049409031867981,
|
| 3005 |
+
"learning_rate": 0.00013622699889432184,
|
| 3006 |
+
"loss": 2.2044,
|
| 3007 |
+
"step": 426
|
| 3008 |
+
},
|
| 3009 |
+
{
|
| 3010 |
+
"epoch": 0.3886676527477529,
|
| 3011 |
+
"grad_norm": 0.40970832109451294,
|
| 3012 |
+
"learning_rate": 0.00013595796015332984,
|
| 3013 |
+
"loss": 2.0984,
|
| 3014 |
+
"step": 427
|
| 3015 |
+
},
|
| 3016 |
+
{
|
| 3017 |
+
"epoch": 0.3895778814427125,
|
| 3018 |
+
"grad_norm": 0.4141111671924591,
|
| 3019 |
+
"learning_rate": 0.00013568862215918717,
|
| 3020 |
+
"loss": 2.109,
|
| 3021 |
+
"step": 428
|
| 3022 |
+
},
|
| 3023 |
+
{
|
| 3024 |
+
"epoch": 0.3904881101376721,
|
| 3025 |
+
"grad_norm": 0.43404263257980347,
|
| 3026 |
+
"learning_rate": 0.00013541898715340716,
|
| 3027 |
+
"loss": 2.1763,
|
| 3028 |
+
"step": 429
|
| 3029 |
+
},
|
| 3030 |
+
{
|
| 3031 |
+
"epoch": 0.3913983388326317,
|
| 3032 |
+
"grad_norm": 0.41949963569641113,
|
| 3033 |
+
"learning_rate": 0.00013514905737997473,
|
| 3034 |
+
"loss": 2.3086,
|
| 3035 |
+
"step": 430
|
| 3036 |
+
},
|
| 3037 |
+
{
|
| 3038 |
+
"epoch": 0.3923085675275913,
|
| 3039 |
+
"grad_norm": 0.41665390133857727,
|
| 3040 |
+
"learning_rate": 0.00013487883508532815,
|
| 3041 |
+
"loss": 2.0726,
|
| 3042 |
+
"step": 431
|
| 3043 |
+
},
|
| 3044 |
+
{
|
| 3045 |
+
"epoch": 0.3932187962225509,
|
| 3046 |
+
"grad_norm": 0.4305708110332489,
|
| 3047 |
+
"learning_rate": 0.00013460832251834011,
|
| 3048 |
+
"loss": 2.1975,
|
| 3049 |
+
"step": 432
|
| 3050 |
+
},
|
| 3051 |
+
{
|
| 3052 |
+
"epoch": 0.3941290249175105,
|
| 3053 |
+
"grad_norm": 0.44775405526161194,
|
| 3054 |
+
"learning_rate": 0.00013433752193029886,
|
| 3055 |
+
"loss": 2.1503,
|
| 3056 |
+
"step": 433
|
| 3057 |
+
},
|
| 3058 |
+
{
|
| 3059 |
+
"epoch": 0.3950392536124701,
|
| 3060 |
+
"grad_norm": 0.44451820850372314,
|
| 3061 |
+
"learning_rate": 0.0001340664355748899,
|
| 3062 |
+
"loss": 2.1004,
|
| 3063 |
+
"step": 434
|
| 3064 |
+
},
|
| 3065 |
+
{
|
| 3066 |
+
"epoch": 0.39594948230742977,
|
| 3067 |
+
"grad_norm": 0.44242945313453674,
|
| 3068 |
+
"learning_rate": 0.0001337950657081768,
|
| 3069 |
+
"loss": 2.1074,
|
| 3070 |
+
"step": 435
|
| 3071 |
+
},
|
| 3072 |
+
{
|
| 3073 |
+
"epoch": 0.39685971100238937,
|
| 3074 |
+
"grad_norm": 0.4649699926376343,
|
| 3075 |
+
"learning_rate": 0.00013352341458858265,
|
| 3076 |
+
"loss": 2.2468,
|
| 3077 |
+
"step": 436
|
| 3078 |
+
},
|
| 3079 |
+
{
|
| 3080 |
+
"epoch": 0.397769939697349,
|
| 3081 |
+
"grad_norm": 0.4718558192253113,
|
| 3082 |
+
"learning_rate": 0.00013325148447687125,
|
| 3083 |
+
"loss": 2.225,
|
| 3084 |
+
"step": 437
|
| 3085 |
+
},
|
| 3086 |
+
{
|
| 3087 |
+
"epoch": 0.3986801683923086,
|
| 3088 |
+
"grad_norm": 0.44748789072036743,
|
| 3089 |
+
"learning_rate": 0.0001329792776361282,
|
| 3090 |
+
"loss": 2.0243,
|
| 3091 |
+
"step": 438
|
| 3092 |
+
},
|
| 3093 |
+
{
|
| 3094 |
+
"epoch": 0.3995903970872682,
|
| 3095 |
+
"grad_norm": 0.4730619192123413,
|
| 3096 |
+
"learning_rate": 0.00013270679633174218,
|
| 3097 |
+
"loss": 2.0262,
|
| 3098 |
+
"step": 439
|
| 3099 |
+
},
|
| 3100 |
+
{
|
| 3101 |
+
"epoch": 0.4005006257822278,
|
| 3102 |
+
"grad_norm": 0.4742071032524109,
|
| 3103 |
+
"learning_rate": 0.00013243404283138597,
|
| 3104 |
+
"loss": 2.1171,
|
| 3105 |
+
"step": 440
|
| 3106 |
+
},
|
| 3107 |
+
{
|
| 3108 |
+
"epoch": 0.4014108544771874,
|
| 3109 |
+
"grad_norm": 0.4963454306125641,
|
| 3110 |
+
"learning_rate": 0.00013216101940499768,
|
| 3111 |
+
"loss": 2.051,
|
| 3112 |
+
"step": 441
|
| 3113 |
+
},
|
| 3114 |
+
{
|
| 3115 |
+
"epoch": 0.402321083172147,
|
| 3116 |
+
"grad_norm": 0.5127780437469482,
|
| 3117 |
+
"learning_rate": 0.00013188772832476188,
|
| 3118 |
+
"loss": 2.1664,
|
| 3119 |
+
"step": 442
|
| 3120 |
+
},
|
| 3121 |
+
{
|
| 3122 |
+
"epoch": 0.4032313118671066,
|
| 3123 |
+
"grad_norm": 0.5129209756851196,
|
| 3124 |
+
"learning_rate": 0.00013161417186509052,
|
| 3125 |
+
"loss": 2.2272,
|
| 3126 |
+
"step": 443
|
| 3127 |
+
},
|
| 3128 |
+
{
|
| 3129 |
+
"epoch": 0.4041415405620662,
|
| 3130 |
+
"grad_norm": 0.5068848133087158,
|
| 3131 |
+
"learning_rate": 0.00013134035230260427,
|
| 3132 |
+
"loss": 2.1007,
|
| 3133 |
+
"step": 444
|
| 3134 |
+
},
|
| 3135 |
+
{
|
| 3136 |
+
"epoch": 0.40505176925702585,
|
| 3137 |
+
"grad_norm": 0.5721228718757629,
|
| 3138 |
+
"learning_rate": 0.00013106627191611332,
|
| 3139 |
+
"loss": 2.255,
|
| 3140 |
+
"step": 445
|
| 3141 |
+
},
|
| 3142 |
+
{
|
| 3143 |
+
"epoch": 0.40596199795198545,
|
| 3144 |
+
"grad_norm": 0.6085918545722961,
|
| 3145 |
+
"learning_rate": 0.0001307919329865985,
|
| 3146 |
+
"loss": 2.456,
|
| 3147 |
+
"step": 446
|
| 3148 |
+
},
|
| 3149 |
+
{
|
| 3150 |
+
"epoch": 0.40687222664694506,
|
| 3151 |
+
"grad_norm": 0.6652196645736694,
|
| 3152 |
+
"learning_rate": 0.00013051733779719234,
|
| 3153 |
+
"loss": 2.5504,
|
| 3154 |
+
"step": 447
|
| 3155 |
+
},
|
| 3156 |
+
{
|
| 3157 |
+
"epoch": 0.40778245534190466,
|
| 3158 |
+
"grad_norm": 0.7234418392181396,
|
| 3159 |
+
"learning_rate": 0.00013024248863316012,
|
| 3160 |
+
"loss": 2.5796,
|
| 3161 |
+
"step": 448
|
| 3162 |
+
},
|
| 3163 |
+
{
|
| 3164 |
+
"epoch": 0.40869268403686426,
|
| 3165 |
+
"grad_norm": 0.8588744401931763,
|
| 3166 |
+
"learning_rate": 0.00012996738778188067,
|
| 3167 |
+
"loss": 2.5756,
|
| 3168 |
+
"step": 449
|
| 3169 |
+
},
|
| 3170 |
+
{
|
| 3171 |
+
"epoch": 0.40960291273182386,
|
| 3172 |
+
"grad_norm": 1.2627683877944946,
|
| 3173 |
+
"learning_rate": 0.0001296920375328275,
|
| 3174 |
+
"loss": 2.203,
|
| 3175 |
+
"step": 450
|
| 3176 |
+
},
|
| 3177 |
+
{
|
| 3178 |
+
"epoch": 0.41051314142678347,
|
| 3179 |
+
"grad_norm": 0.4838164746761322,
|
| 3180 |
+
"learning_rate": 0.00012941644017754964,
|
| 3181 |
+
"loss": 2.434,
|
| 3182 |
+
"step": 451
|
| 3183 |
+
},
|
| 3184 |
+
{
|
| 3185 |
+
"epoch": 0.41142337012174307,
|
| 3186 |
+
"grad_norm": 0.44005534052848816,
|
| 3187 |
+
"learning_rate": 0.00012914059800965268,
|
| 3188 |
+
"loss": 2.55,
|
| 3189 |
+
"step": 452
|
| 3190 |
+
},
|
| 3191 |
+
{
|
| 3192 |
+
"epoch": 0.4123335988167027,
|
| 3193 |
+
"grad_norm": 0.4343414604663849,
|
| 3194 |
+
"learning_rate": 0.0001288645133247795,
|
| 3195 |
+
"loss": 2.432,
|
| 3196 |
+
"step": 453
|
| 3197 |
+
},
|
| 3198 |
+
{
|
| 3199 |
+
"epoch": 0.41324382751166233,
|
| 3200 |
+
"grad_norm": 0.4588654339313507,
|
| 3201 |
+
"learning_rate": 0.00012858818842059145,
|
| 3202 |
+
"loss": 2.4434,
|
| 3203 |
+
"step": 454
|
| 3204 |
+
},
|
| 3205 |
+
{
|
| 3206 |
+
"epoch": 0.41415405620662193,
|
| 3207 |
+
"grad_norm": 0.4294244647026062,
|
| 3208 |
+
"learning_rate": 0.00012831162559674887,
|
| 3209 |
+
"loss": 2.4241,
|
| 3210 |
+
"step": 455
|
| 3211 |
+
},
|
| 3212 |
+
{
|
| 3213 |
+
"epoch": 0.41506428490158154,
|
| 3214 |
+
"grad_norm": 0.40034809708595276,
|
| 3215 |
+
"learning_rate": 0.0001280348271548923,
|
| 3216 |
+
"loss": 2.3191,
|
| 3217 |
+
"step": 456
|
| 3218 |
+
},
|
| 3219 |
+
{
|
| 3220 |
+
"epoch": 0.41597451359654114,
|
| 3221 |
+
"grad_norm": 0.40817153453826904,
|
| 3222 |
+
"learning_rate": 0.00012775779539862304,
|
| 3223 |
+
"loss": 2.589,
|
| 3224 |
+
"step": 457
|
| 3225 |
+
},
|
| 3226 |
+
{
|
| 3227 |
+
"epoch": 0.41688474229150074,
|
| 3228 |
+
"grad_norm": 0.40605810284614563,
|
| 3229 |
+
"learning_rate": 0.0001274805326334842,
|
| 3230 |
+
"loss": 2.3445,
|
| 3231 |
+
"step": 458
|
| 3232 |
+
},
|
| 3233 |
+
{
|
| 3234 |
+
"epoch": 0.41779497098646035,
|
| 3235 |
+
"grad_norm": 0.4386533200740814,
|
| 3236 |
+
"learning_rate": 0.00012720304116694138,
|
| 3237 |
+
"loss": 2.4002,
|
| 3238 |
+
"step": 459
|
| 3239 |
+
},
|
| 3240 |
+
{
|
| 3241 |
+
"epoch": 0.41870519968141995,
|
| 3242 |
+
"grad_norm": 0.40985172986984253,
|
| 3243 |
+
"learning_rate": 0.00012692532330836346,
|
| 3244 |
+
"loss": 2.3964,
|
| 3245 |
+
"step": 460
|
| 3246 |
+
},
|
| 3247 |
+
{
|
| 3248 |
+
"epoch": 0.41961542837637955,
|
| 3249 |
+
"grad_norm": 0.4220562279224396,
|
| 3250 |
+
"learning_rate": 0.00012664738136900348,
|
| 3251 |
+
"loss": 2.3145,
|
| 3252 |
+
"step": 461
|
| 3253 |
+
},
|
| 3254 |
+
{
|
| 3255 |
+
"epoch": 0.42052565707133915,
|
| 3256 |
+
"grad_norm": 0.4068267047405243,
|
| 3257 |
+
"learning_rate": 0.00012636921766197943,
|
| 3258 |
+
"loss": 2.3274,
|
| 3259 |
+
"step": 462
|
| 3260 |
+
},
|
| 3261 |
+
{
|
| 3262 |
+
"epoch": 0.42143588576629876,
|
| 3263 |
+
"grad_norm": 0.3973187208175659,
|
| 3264 |
+
"learning_rate": 0.0001260908345022547,
|
| 3265 |
+
"loss": 2.1801,
|
| 3266 |
+
"step": 463
|
| 3267 |
+
},
|
| 3268 |
+
{
|
| 3269 |
+
"epoch": 0.4223461144612584,
|
| 3270 |
+
"grad_norm": 0.432224303483963,
|
| 3271 |
+
"learning_rate": 0.00012581223420661913,
|
| 3272 |
+
"loss": 2.4079,
|
| 3273 |
+
"step": 464
|
| 3274 |
+
},
|
| 3275 |
+
{
|
| 3276 |
+
"epoch": 0.423256343156218,
|
| 3277 |
+
"grad_norm": 0.3939046859741211,
|
| 3278 |
+
"learning_rate": 0.00012553341909366978,
|
| 3279 |
+
"loss": 2.0749,
|
| 3280 |
+
"step": 465
|
| 3281 |
+
},
|
| 3282 |
+
{
|
| 3283 |
+
"epoch": 0.4241665718511776,
|
| 3284 |
+
"grad_norm": 0.36949658393859863,
|
| 3285 |
+
"learning_rate": 0.00012525439148379128,
|
| 3286 |
+
"loss": 2.1471,
|
| 3287 |
+
"step": 466
|
| 3288 |
+
},
|
| 3289 |
+
{
|
| 3290 |
+
"epoch": 0.4250768005461372,
|
| 3291 |
+
"grad_norm": 0.3828236758708954,
|
| 3292 |
+
"learning_rate": 0.00012497515369913685,
|
| 3293 |
+
"loss": 2.0466,
|
| 3294 |
+
"step": 467
|
| 3295 |
+
},
|
| 3296 |
+
{
|
| 3297 |
+
"epoch": 0.4259870292410968,
|
| 3298 |
+
"grad_norm": 0.3874993920326233,
|
| 3299 |
+
"learning_rate": 0.00012469570806360875,
|
| 3300 |
+
"loss": 2.1605,
|
| 3301 |
+
"step": 468
|
| 3302 |
+
},
|
| 3303 |
+
{
|
| 3304 |
+
"epoch": 0.42689725793605643,
|
| 3305 |
+
"grad_norm": 0.3854924738407135,
|
| 3306 |
+
"learning_rate": 0.00012441605690283915,
|
| 3307 |
+
"loss": 2.0584,
|
| 3308 |
+
"step": 469
|
| 3309 |
+
},
|
| 3310 |
+
{
|
| 3311 |
+
"epoch": 0.42780748663101603,
|
| 3312 |
+
"grad_norm": 0.40301740169525146,
|
| 3313 |
+
"learning_rate": 0.00012413620254417057,
|
| 3314 |
+
"loss": 2.1481,
|
| 3315 |
+
"step": 470
|
| 3316 |
+
},
|
| 3317 |
+
{
|
| 3318 |
+
"epoch": 0.42871771532597563,
|
| 3319 |
+
"grad_norm": 0.3891369104385376,
|
| 3320 |
+
"learning_rate": 0.00012385614731663666,
|
| 3321 |
+
"loss": 2.1968,
|
| 3322 |
+
"step": 471
|
| 3323 |
+
},
|
| 3324 |
+
{
|
| 3325 |
+
"epoch": 0.42962794402093524,
|
| 3326 |
+
"grad_norm": 0.4305795729160309,
|
| 3327 |
+
"learning_rate": 0.00012357589355094275,
|
| 3328 |
+
"loss": 2.0421,
|
| 3329 |
+
"step": 472
|
| 3330 |
+
},
|
| 3331 |
+
{
|
| 3332 |
+
"epoch": 0.4305381727158949,
|
| 3333 |
+
"grad_norm": 0.44661635160446167,
|
| 3334 |
+
"learning_rate": 0.0001232954435794464,
|
| 3335 |
+
"loss": 2.3347,
|
| 3336 |
+
"step": 473
|
| 3337 |
+
},
|
| 3338 |
+
{
|
| 3339 |
+
"epoch": 0.4314484014108545,
|
| 3340 |
+
"grad_norm": 0.3984116315841675,
|
| 3341 |
+
"learning_rate": 0.00012301479973613822,
|
| 3342 |
+
"loss": 2.1093,
|
| 3343 |
+
"step": 474
|
| 3344 |
+
},
|
| 3345 |
+
{
|
| 3346 |
+
"epoch": 0.4323586301058141,
|
| 3347 |
+
"grad_norm": 0.4153747856616974,
|
| 3348 |
+
"learning_rate": 0.00012273396435662212,
|
| 3349 |
+
"loss": 2.0698,
|
| 3350 |
+
"step": 475
|
| 3351 |
+
},
|
| 3352 |
+
{
|
| 3353 |
+
"epoch": 0.4332688588007737,
|
| 3354 |
+
"grad_norm": 0.4589189887046814,
|
| 3355 |
+
"learning_rate": 0.00012245293977809605,
|
| 3356 |
+
"loss": 2.1707,
|
| 3357 |
+
"step": 476
|
| 3358 |
+
},
|
| 3359 |
+
{
|
| 3360 |
+
"epoch": 0.4341790874957333,
|
| 3361 |
+
"grad_norm": 0.43936577439308167,
|
| 3362 |
+
"learning_rate": 0.0001221717283393326,
|
| 3363 |
+
"loss": 2.2608,
|
| 3364 |
+
"step": 477
|
| 3365 |
+
},
|
| 3366 |
+
{
|
| 3367 |
+
"epoch": 0.4350893161906929,
|
| 3368 |
+
"grad_norm": 0.4170132279396057,
|
| 3369 |
+
"learning_rate": 0.0001218903323806595,
|
| 3370 |
+
"loss": 2.0813,
|
| 3371 |
+
"step": 478
|
| 3372 |
+
},
|
| 3373 |
+
{
|
| 3374 |
+
"epoch": 0.4359995448856525,
|
| 3375 |
+
"grad_norm": 0.43124523758888245,
|
| 3376 |
+
"learning_rate": 0.00012160875424393996,
|
| 3377 |
+
"loss": 2.1674,
|
| 3378 |
+
"step": 479
|
| 3379 |
+
},
|
| 3380 |
+
{
|
| 3381 |
+
"epoch": 0.4369097735806121,
|
| 3382 |
+
"grad_norm": 0.4394627511501312,
|
| 3383 |
+
"learning_rate": 0.00012132699627255347,
|
| 3384 |
+
"loss": 2.1904,
|
| 3385 |
+
"step": 480
|
| 3386 |
+
},
|
| 3387 |
+
{
|
| 3388 |
+
"epoch": 0.4378200022755717,
|
| 3389 |
+
"grad_norm": 0.4404590427875519,
|
| 3390 |
+
"learning_rate": 0.00012104506081137608,
|
| 3391 |
+
"loss": 2.1313,
|
| 3392 |
+
"step": 481
|
| 3393 |
+
},
|
| 3394 |
+
{
|
| 3395 |
+
"epoch": 0.4387302309705313,
|
| 3396 |
+
"grad_norm": 0.4580220878124237,
|
| 3397 |
+
"learning_rate": 0.00012076295020676103,
|
| 3398 |
+
"loss": 2.16,
|
| 3399 |
+
"step": 482
|
| 3400 |
+
},
|
| 3401 |
+
{
|
| 3402 |
+
"epoch": 0.439640459665491,
|
| 3403 |
+
"grad_norm": 0.4533630311489105,
|
| 3404 |
+
"learning_rate": 0.00012048066680651908,
|
| 3405 |
+
"loss": 2.1153,
|
| 3406 |
+
"step": 483
|
| 3407 |
+
},
|
| 3408 |
+
{
|
| 3409 |
+
"epoch": 0.4405506883604506,
|
| 3410 |
+
"grad_norm": 0.47520536184310913,
|
| 3411 |
+
"learning_rate": 0.00012019821295989912,
|
| 3412 |
+
"loss": 2.2152,
|
| 3413 |
+
"step": 484
|
| 3414 |
+
},
|
| 3415 |
+
{
|
| 3416 |
+
"epoch": 0.4414609170554102,
|
| 3417 |
+
"grad_norm": 0.44196072220802307,
|
| 3418 |
+
"learning_rate": 0.00011991559101756852,
|
| 3419 |
+
"loss": 2.1375,
|
| 3420 |
+
"step": 485
|
| 3421 |
+
},
|
| 3422 |
+
{
|
| 3423 |
+
"epoch": 0.4423711457503698,
|
| 3424 |
+
"grad_norm": 0.43681493401527405,
|
| 3425 |
+
"learning_rate": 0.00011963280333159358,
|
| 3426 |
+
"loss": 2.0552,
|
| 3427 |
+
"step": 486
|
| 3428 |
+
},
|
| 3429 |
+
{
|
| 3430 |
+
"epoch": 0.4432813744453294,
|
| 3431 |
+
"grad_norm": 0.4537602961063385,
|
| 3432 |
+
"learning_rate": 0.00011934985225541998,
|
| 3433 |
+
"loss": 2.1473,
|
| 3434 |
+
"step": 487
|
| 3435 |
+
},
|
| 3436 |
+
{
|
| 3437 |
+
"epoch": 0.444191603140289,
|
| 3438 |
+
"grad_norm": 0.4935773015022278,
|
| 3439 |
+
"learning_rate": 0.00011906674014385318,
|
| 3440 |
+
"loss": 2.0623,
|
| 3441 |
+
"step": 488
|
| 3442 |
+
},
|
| 3443 |
+
{
|
| 3444 |
+
"epoch": 0.4451018318352486,
|
| 3445 |
+
"grad_norm": 0.4802737236022949,
|
| 3446 |
+
"learning_rate": 0.00011878346935303883,
|
| 3447 |
+
"loss": 2.2908,
|
| 3448 |
+
"step": 489
|
| 3449 |
+
},
|
| 3450 |
+
{
|
| 3451 |
+
"epoch": 0.4460120605302082,
|
| 3452 |
+
"grad_norm": 0.5020537376403809,
|
| 3453 |
+
"learning_rate": 0.00011850004224044315,
|
| 3454 |
+
"loss": 2.3101,
|
| 3455 |
+
"step": 490
|
| 3456 |
+
},
|
| 3457 |
+
{
|
| 3458 |
+
"epoch": 0.4469222892251678,
|
| 3459 |
+
"grad_norm": 0.5106056332588196,
|
| 3460 |
+
"learning_rate": 0.00011821646116483335,
|
| 3461 |
+
"loss": 2.2838,
|
| 3462 |
+
"step": 491
|
| 3463 |
+
},
|
| 3464 |
+
{
|
| 3465 |
+
"epoch": 0.44783251792012746,
|
| 3466 |
+
"grad_norm": 0.473910391330719,
|
| 3467 |
+
"learning_rate": 0.00011793272848625797,
|
| 3468 |
+
"loss": 2.0599,
|
| 3469 |
+
"step": 492
|
| 3470 |
+
},
|
| 3471 |
+
{
|
| 3472 |
+
"epoch": 0.44874274661508706,
|
| 3473 |
+
"grad_norm": 0.5086584091186523,
|
| 3474 |
+
"learning_rate": 0.0001176488465660271,
|
| 3475 |
+
"loss": 2.1578,
|
| 3476 |
+
"step": 493
|
| 3477 |
+
},
|
| 3478 |
+
{
|
| 3479 |
+
"epoch": 0.44965297531004667,
|
| 3480 |
+
"grad_norm": 0.5282394886016846,
|
| 3481 |
+
"learning_rate": 0.00011736481776669306,
|
| 3482 |
+
"loss": 2.2965,
|
| 3483 |
+
"step": 494
|
| 3484 |
+
},
|
| 3485 |
+
{
|
| 3486 |
+
"epoch": 0.45056320400500627,
|
| 3487 |
+
"grad_norm": 0.5987780094146729,
|
| 3488 |
+
"learning_rate": 0.00011708064445203042,
|
| 3489 |
+
"loss": 2.3542,
|
| 3490 |
+
"step": 495
|
| 3491 |
+
},
|
| 3492 |
+
{
|
| 3493 |
+
"epoch": 0.45147343269996587,
|
| 3494 |
+
"grad_norm": 0.5943189859390259,
|
| 3495 |
+
"learning_rate": 0.00011679632898701649,
|
| 3496 |
+
"loss": 2.4294,
|
| 3497 |
+
"step": 496
|
| 3498 |
+
},
|
| 3499 |
+
{
|
| 3500 |
+
"epoch": 0.4523836613949255,
|
| 3501 |
+
"grad_norm": 0.6443737149238586,
|
| 3502 |
+
"learning_rate": 0.0001165118737378116,
|
| 3503 |
+
"loss": 2.605,
|
| 3504 |
+
"step": 497
|
| 3505 |
+
},
|
| 3506 |
+
{
|
| 3507 |
+
"epoch": 0.4532938900898851,
|
| 3508 |
+
"grad_norm": 0.7082577347755432,
|
| 3509 |
+
"learning_rate": 0.00011622728107173946,
|
| 3510 |
+
"loss": 2.4254,
|
| 3511 |
+
"step": 498
|
| 3512 |
+
},
|
| 3513 |
+
{
|
| 3514 |
+
"epoch": 0.4542041187848447,
|
| 3515 |
+
"grad_norm": 0.8503845930099487,
|
| 3516 |
+
"learning_rate": 0.00011594255335726724,
|
| 3517 |
+
"loss": 2.5187,
|
| 3518 |
+
"step": 499
|
| 3519 |
+
},
|
| 3520 |
+
{
|
| 3521 |
+
"epoch": 0.4551143474798043,
|
| 3522 |
+
"grad_norm": 1.6775977611541748,
|
| 3523 |
+
"learning_rate": 0.00011565769296398618,
|
| 3524 |
+
"loss": 2.6669,
|
| 3525 |
+
"step": 500
|
| 3526 |
+
},
|
| 3527 |
+
{
|
| 3528 |
+
"epoch": 0.4560245761747639,
|
| 3529 |
+
"grad_norm": 0.45572495460510254,
|
| 3530 |
+
"learning_rate": 0.00011537270226259169,
|
| 3531 |
+
"loss": 2.5806,
|
| 3532 |
+
"step": 501
|
| 3533 |
+
},
|
| 3534 |
+
{
|
| 3535 |
+
"epoch": 0.45693480486972354,
|
| 3536 |
+
"grad_norm": 0.45138293504714966,
|
| 3537 |
+
"learning_rate": 0.00011508758362486358,
|
| 3538 |
+
"loss": 2.3935,
|
| 3539 |
+
"step": 502
|
| 3540 |
+
},
|
| 3541 |
+
{
|
| 3542 |
+
"epoch": 0.45784503356468315,
|
| 3543 |
+
"grad_norm": 0.4548013210296631,
|
| 3544 |
+
"learning_rate": 0.00011480233942364645,
|
| 3545 |
+
"loss": 2.321,
|
| 3546 |
+
"step": 503
|
| 3547 |
+
},
|
| 3548 |
+
{
|
| 3549 |
+
"epoch": 0.45875526225964275,
|
| 3550 |
+
"grad_norm": 0.434442400932312,
|
| 3551 |
+
"learning_rate": 0.00011451697203282982,
|
| 3552 |
+
"loss": 2.375,
|
| 3553 |
+
"step": 504
|
| 3554 |
+
},
|
| 3555 |
+
{
|
| 3556 |
+
"epoch": 0.45966549095460235,
|
| 3557 |
+
"grad_norm": 0.4139295816421509,
|
| 3558 |
+
"learning_rate": 0.00011423148382732853,
|
| 3559 |
+
"loss": 2.3997,
|
| 3560 |
+
"step": 505
|
| 3561 |
+
},
|
| 3562 |
+
{
|
| 3563 |
+
"epoch": 0.46057571964956195,
|
| 3564 |
+
"grad_norm": 0.46020230650901794,
|
| 3565 |
+
"learning_rate": 0.00011394587718306275,
|
| 3566 |
+
"loss": 2.5745,
|
| 3567 |
+
"step": 506
|
| 3568 |
+
},
|
| 3569 |
+
{
|
| 3570 |
+
"epoch": 0.46148594834452156,
|
| 3571 |
+
"grad_norm": 0.4194343090057373,
|
| 3572 |
+
"learning_rate": 0.00011366015447693837,
|
| 3573 |
+
"loss": 2.2597,
|
| 3574 |
+
"step": 507
|
| 3575 |
+
},
|
| 3576 |
+
{
|
| 3577 |
+
"epoch": 0.46239617703948116,
|
| 3578 |
+
"grad_norm": 0.43983832001686096,
|
| 3579 |
+
"learning_rate": 0.0001133743180868273,
|
| 3580 |
+
"loss": 2.3511,
|
| 3581 |
+
"step": 508
|
| 3582 |
+
},
|
| 3583 |
+
{
|
| 3584 |
+
"epoch": 0.46330640573444076,
|
| 3585 |
+
"grad_norm": 0.41047292947769165,
|
| 3586 |
+
"learning_rate": 0.00011308837039154739,
|
| 3587 |
+
"loss": 2.2614,
|
| 3588 |
+
"step": 509
|
| 3589 |
+
},
|
| 3590 |
+
{
|
| 3591 |
+
"epoch": 0.46421663442940037,
|
| 3592 |
+
"grad_norm": 0.4110110104084015,
|
| 3593 |
+
"learning_rate": 0.0001128023137708429,
|
| 3594 |
+
"loss": 2.2719,
|
| 3595 |
+
"step": 510
|
| 3596 |
+
},
|
| 3597 |
+
{
|
| 3598 |
+
"epoch": 0.46512686312435997,
|
| 3599 |
+
"grad_norm": 0.41848358511924744,
|
| 3600 |
+
"learning_rate": 0.0001125161506053646,
|
| 3601 |
+
"loss": 2.3872,
|
| 3602 |
+
"step": 511
|
| 3603 |
+
},
|
| 3604 |
+
{
|
| 3605 |
+
"epoch": 0.4660370918193196,
|
| 3606 |
+
"grad_norm": 0.39852631092071533,
|
| 3607 |
+
"learning_rate": 0.00011222988327664997,
|
| 3608 |
+
"loss": 2.2001,
|
| 3609 |
+
"step": 512
|
| 3610 |
+
},
|
| 3611 |
+
{
|
| 3612 |
+
"epoch": 0.46694732051427923,
|
| 3613 |
+
"grad_norm": 0.4060978293418884,
|
| 3614 |
+
"learning_rate": 0.00011194351416710324,
|
| 3615 |
+
"loss": 2.2474,
|
| 3616 |
+
"step": 513
|
| 3617 |
+
},
|
| 3618 |
+
{
|
| 3619 |
+
"epoch": 0.46785754920923883,
|
| 3620 |
+
"grad_norm": 0.4010358452796936,
|
| 3621 |
+
"learning_rate": 0.00011165704565997593,
|
| 3622 |
+
"loss": 2.1262,
|
| 3623 |
+
"step": 514
|
| 3624 |
+
},
|
| 3625 |
+
{
|
| 3626 |
+
"epoch": 0.46876777790419843,
|
| 3627 |
+
"grad_norm": 0.4063378572463989,
|
| 3628 |
+
"learning_rate": 0.00011137048013934656,
|
| 3629 |
+
"loss": 2.1583,
|
| 3630 |
+
"step": 515
|
| 3631 |
+
},
|
| 3632 |
+
{
|
| 3633 |
+
"epoch": 0.46967800659915804,
|
| 3634 |
+
"grad_norm": 0.40287846326828003,
|
| 3635 |
+
"learning_rate": 0.00011108381999010111,
|
| 3636 |
+
"loss": 2.2351,
|
| 3637 |
+
"step": 516
|
| 3638 |
+
},
|
| 3639 |
+
{
|
| 3640 |
+
"epoch": 0.47058823529411764,
|
| 3641 |
+
"grad_norm": 0.3861018717288971,
|
| 3642 |
+
"learning_rate": 0.00011079706759791311,
|
| 3643 |
+
"loss": 2.195,
|
| 3644 |
+
"step": 517
|
| 3645 |
+
},
|
| 3646 |
+
{
|
| 3647 |
+
"epoch": 0.47149846398907724,
|
| 3648 |
+
"grad_norm": 0.38855546712875366,
|
| 3649 |
+
"learning_rate": 0.00011051022534922371,
|
| 3650 |
+
"loss": 2.1575,
|
| 3651 |
+
"step": 518
|
| 3652 |
+
},
|
| 3653 |
+
{
|
| 3654 |
+
"epoch": 0.47240869268403685,
|
| 3655 |
+
"grad_norm": 0.3941628038883209,
|
| 3656 |
+
"learning_rate": 0.00011022329563122191,
|
| 3657 |
+
"loss": 2.2324,
|
| 3658 |
+
"step": 519
|
| 3659 |
+
},
|
| 3660 |
+
{
|
| 3661 |
+
"epoch": 0.47331892137899645,
|
| 3662 |
+
"grad_norm": 0.40604814887046814,
|
| 3663 |
+
"learning_rate": 0.00010993628083182467,
|
| 3664 |
+
"loss": 2.1641,
|
| 3665 |
+
"step": 520
|
| 3666 |
+
},
|
| 3667 |
+
{
|
| 3668 |
+
"epoch": 0.4742291500739561,
|
| 3669 |
+
"grad_norm": 0.407815158367157,
|
| 3670 |
+
"learning_rate": 0.000109649183339657,
|
| 3671 |
+
"loss": 2.1648,
|
| 3672 |
+
"step": 521
|
| 3673 |
+
},
|
| 3674 |
+
{
|
| 3675 |
+
"epoch": 0.4751393787689157,
|
| 3676 |
+
"grad_norm": 0.400680810213089,
|
| 3677 |
+
"learning_rate": 0.00010936200554403209,
|
| 3678 |
+
"loss": 2.1939,
|
| 3679 |
+
"step": 522
|
| 3680 |
+
},
|
| 3681 |
+
{
|
| 3682 |
+
"epoch": 0.4760496074638753,
|
| 3683 |
+
"grad_norm": 0.416537344455719,
|
| 3684 |
+
"learning_rate": 0.00010907474983493144,
|
| 3685 |
+
"loss": 2.1694,
|
| 3686 |
+
"step": 523
|
| 3687 |
+
},
|
| 3688 |
+
{
|
| 3689 |
+
"epoch": 0.4769598361588349,
|
| 3690 |
+
"grad_norm": 0.4097869396209717,
|
| 3691 |
+
"learning_rate": 0.00010878741860298503,
|
| 3692 |
+
"loss": 2.1785,
|
| 3693 |
+
"step": 524
|
| 3694 |
+
},
|
| 3695 |
+
{
|
| 3696 |
+
"epoch": 0.4778700648537945,
|
| 3697 |
+
"grad_norm": 0.4243004024028778,
|
| 3698 |
+
"learning_rate": 0.00010850001423945126,
|
| 3699 |
+
"loss": 1.9963,
|
| 3700 |
+
"step": 525
|
| 3701 |
+
},
|
| 3702 |
+
{
|
| 3703 |
+
"epoch": 0.4787802935487541,
|
| 3704 |
+
"grad_norm": 0.41958731412887573,
|
| 3705 |
+
"learning_rate": 0.00010821253913619726,
|
| 3706 |
+
"loss": 2.1629,
|
| 3707 |
+
"step": 526
|
| 3708 |
+
},
|
| 3709 |
+
{
|
| 3710 |
+
"epoch": 0.4796905222437137,
|
| 3711 |
+
"grad_norm": 0.4177284240722656,
|
| 3712 |
+
"learning_rate": 0.00010792499568567884,
|
| 3713 |
+
"loss": 2.1276,
|
| 3714 |
+
"step": 527
|
| 3715 |
+
},
|
| 3716 |
+
{
|
| 3717 |
+
"epoch": 0.4806007509386733,
|
| 3718 |
+
"grad_norm": 0.41077664494514465,
|
| 3719 |
+
"learning_rate": 0.00010763738628092062,
|
| 3720 |
+
"loss": 2.0852,
|
| 3721 |
+
"step": 528
|
| 3722 |
+
},
|
| 3723 |
+
{
|
| 3724 |
+
"epoch": 0.48151097963363293,
|
| 3725 |
+
"grad_norm": 0.4098223149776459,
|
| 3726 |
+
"learning_rate": 0.00010734971331549603,
|
| 3727 |
+
"loss": 1.9977,
|
| 3728 |
+
"step": 529
|
| 3729 |
+
},
|
| 3730 |
+
{
|
| 3731 |
+
"epoch": 0.48242120832859253,
|
| 3732 |
+
"grad_norm": 0.42255935072898865,
|
| 3733 |
+
"learning_rate": 0.00010706197918350758,
|
| 3734 |
+
"loss": 1.9822,
|
| 3735 |
+
"step": 530
|
| 3736 |
+
},
|
| 3737 |
+
{
|
| 3738 |
+
"epoch": 0.4833314370235522,
|
| 3739 |
+
"grad_norm": 0.45597127079963684,
|
| 3740 |
+
"learning_rate": 0.0001067741862795668,
|
| 3741 |
+
"loss": 2.1072,
|
| 3742 |
+
"step": 531
|
| 3743 |
+
},
|
| 3744 |
+
{
|
| 3745 |
+
"epoch": 0.4842416657185118,
|
| 3746 |
+
"grad_norm": 0.4538208544254303,
|
| 3747 |
+
"learning_rate": 0.0001064863369987743,
|
| 3748 |
+
"loss": 2.41,
|
| 3749 |
+
"step": 532
|
| 3750 |
+
},
|
| 3751 |
+
{
|
| 3752 |
+
"epoch": 0.4851518944134714,
|
| 3753 |
+
"grad_norm": 0.4586673676967621,
|
| 3754 |
+
"learning_rate": 0.00010619843373669993,
|
| 3755 |
+
"loss": 2.1736,
|
| 3756 |
+
"step": 533
|
| 3757 |
+
},
|
| 3758 |
+
{
|
| 3759 |
+
"epoch": 0.486062123108431,
|
| 3760 |
+
"grad_norm": 0.4433608055114746,
|
| 3761 |
+
"learning_rate": 0.00010591047888936274,
|
| 3762 |
+
"loss": 2.1324,
|
| 3763 |
+
"step": 534
|
| 3764 |
+
},
|
| 3765 |
+
{
|
| 3766 |
+
"epoch": 0.4869723518033906,
|
| 3767 |
+
"grad_norm": 0.4421234428882599,
|
| 3768 |
+
"learning_rate": 0.00010562247485321115,
|
| 3769 |
+
"loss": 2.0689,
|
| 3770 |
+
"step": 535
|
| 3771 |
+
},
|
| 3772 |
+
{
|
| 3773 |
+
"epoch": 0.4878825804983502,
|
| 3774 |
+
"grad_norm": 0.46843069791793823,
|
| 3775 |
+
"learning_rate": 0.00010533442402510284,
|
| 3776 |
+
"loss": 2.2252,
|
| 3777 |
+
"step": 536
|
| 3778 |
+
},
|
| 3779 |
+
{
|
| 3780 |
+
"epoch": 0.4887928091933098,
|
| 3781 |
+
"grad_norm": 0.4747142493724823,
|
| 3782 |
+
"learning_rate": 0.00010504632880228498,
|
| 3783 |
+
"loss": 2.2503,
|
| 3784 |
+
"step": 537
|
| 3785 |
+
},
|
| 3786 |
+
{
|
| 3787 |
+
"epoch": 0.4897030378882694,
|
| 3788 |
+
"grad_norm": 0.46643224358558655,
|
| 3789 |
+
"learning_rate": 0.00010475819158237425,
|
| 3790 |
+
"loss": 2.2628,
|
| 3791 |
+
"step": 538
|
| 3792 |
+
},
|
| 3793 |
+
{
|
| 3794 |
+
"epoch": 0.490613266583229,
|
| 3795 |
+
"grad_norm": 0.47085490822792053,
|
| 3796 |
+
"learning_rate": 0.00010447001476333673,
|
| 3797 |
+
"loss": 2.0888,
|
| 3798 |
+
"step": 539
|
| 3799 |
+
},
|
| 3800 |
+
{
|
| 3801 |
+
"epoch": 0.49152349527818867,
|
| 3802 |
+
"grad_norm": 0.5102598071098328,
|
| 3803 |
+
"learning_rate": 0.00010418180074346815,
|
| 3804 |
+
"loss": 2.2736,
|
| 3805 |
+
"step": 540
|
| 3806 |
+
},
|
| 3807 |
+
{
|
| 3808 |
+
"epoch": 0.4924337239731483,
|
| 3809 |
+
"grad_norm": 0.49878573417663574,
|
| 3810 |
+
"learning_rate": 0.00010389355192137377,
|
| 3811 |
+
"loss": 2.1107,
|
| 3812 |
+
"step": 541
|
| 3813 |
+
},
|
| 3814 |
+
{
|
| 3815 |
+
"epoch": 0.4933439526681079,
|
| 3816 |
+
"grad_norm": 0.5236616134643555,
|
| 3817 |
+
"learning_rate": 0.00010360527069594859,
|
| 3818 |
+
"loss": 2.4099,
|
| 3819 |
+
"step": 542
|
| 3820 |
+
},
|
| 3821 |
+
{
|
| 3822 |
+
"epoch": 0.4942541813630675,
|
| 3823 |
+
"grad_norm": 0.49875032901763916,
|
| 3824 |
+
"learning_rate": 0.00010331695946635708,
|
| 3825 |
+
"loss": 2.1381,
|
| 3826 |
+
"step": 543
|
| 3827 |
+
},
|
| 3828 |
+
{
|
| 3829 |
+
"epoch": 0.4951644100580271,
|
| 3830 |
+
"grad_norm": 0.5333012938499451,
|
| 3831 |
+
"learning_rate": 0.00010302862063201367,
|
| 3832 |
+
"loss": 2.2274,
|
| 3833 |
+
"step": 544
|
| 3834 |
+
},
|
| 3835 |
+
{
|
| 3836 |
+
"epoch": 0.4960746387529867,
|
| 3837 |
+
"grad_norm": 0.5504993200302124,
|
| 3838 |
+
"learning_rate": 0.00010274025659256232,
|
| 3839 |
+
"loss": 2.2348,
|
| 3840 |
+
"step": 545
|
| 3841 |
+
},
|
| 3842 |
+
{
|
| 3843 |
+
"epoch": 0.4969848674479463,
|
| 3844 |
+
"grad_norm": 0.5924202799797058,
|
| 3845 |
+
"learning_rate": 0.00010245186974785685,
|
| 3846 |
+
"loss": 2.3686,
|
| 3847 |
+
"step": 546
|
| 3848 |
+
},
|
| 3849 |
+
{
|
| 3850 |
+
"epoch": 0.4978950961429059,
|
| 3851 |
+
"grad_norm": 0.6003567576408386,
|
| 3852 |
+
"learning_rate": 0.00010216346249794087,
|
| 3853 |
+
"loss": 2.3336,
|
| 3854 |
+
"step": 547
|
| 3855 |
+
},
|
| 3856 |
+
{
|
| 3857 |
+
"epoch": 0.4988053248378655,
|
| 3858 |
+
"grad_norm": 0.6700019836425781,
|
| 3859 |
+
"learning_rate": 0.00010187503724302776,
|
| 3860 |
+
"loss": 2.4446,
|
| 3861 |
+
"step": 548
|
| 3862 |
+
},
|
| 3863 |
+
{
|
| 3864 |
+
"epoch": 0.4997155535328251,
|
| 3865 |
+
"grad_norm": 0.8171781897544861,
|
| 3866 |
+
"learning_rate": 0.00010158659638348081,
|
| 3867 |
+
"loss": 2.4278,
|
| 3868 |
+
"step": 549
|
| 3869 |
+
},
|
| 3870 |
+
{
|
| 3871 |
+
"epoch": 0.5006257822277848,
|
| 3872 |
+
"grad_norm": 1.4212020635604858,
|
| 3873 |
+
"learning_rate": 0.0001012981423197931,
|
| 3874 |
+
"loss": 2.6229,
|
| 3875 |
+
"step": 550
|
| 3876 |
+
},
|
| 3877 |
+
{
|
| 3878 |
+
"epoch": 0.5006257822277848,
|
| 3879 |
+
"eval_loss": 2.2479705810546875,
|
| 3880 |
+
"eval_runtime": 205.3622,
|
| 3881 |
+
"eval_samples_per_second": 9.013,
|
| 3882 |
+
"eval_steps_per_second": 4.509,
|
| 3883 |
+
"step": 550
|
| 3884 |
}
|
| 3885 |
],
|
| 3886 |
"logging_steps": 1,
|
|
|
|
| 3900 |
"attributes": {}
|
| 3901 |
}
|
| 3902 |
},
|
| 3903 |
+
"total_flos": 1.0074178982447677e+18,
|
| 3904 |
"train_batch_size": 2,
|
| 3905 |
"trial_name": null,
|
| 3906 |
"trial_params": null
|