Training in progress, step 975, 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 17640136
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2fcb142089042c9fd3481a70a7df0e6f391d749780a0b1522b3c3a7031043996
|
| 3 |
size 17640136
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 9569204
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e149bc8a2d73999c20111adf159a4120f1329cfa81026b882d1ae9bcb18b519e
|
| 3 |
size 9569204
|
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:aa45536a258cdce6fd2a5fd9b3c167b41832ddeee0d25c2b58a64004e45653ab
|
| 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 sha256:9117aba07e0435be53a45cbfc1aca6d938bf07f10ec7ee995185ddb3faaecea3
|
| 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":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -5131,6 +5131,1707 @@
|
|
| 5131 |
"learning_rate": 1.470196091938691e-05,
|
| 5132 |
"loss": 0.6484,
|
| 5133 |
"step": 732
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5134 |
}
|
| 5135 |
],
|
| 5136 |
"logging_steps": 1,
|
|
@@ -5145,12 +6846,12 @@
|
|
| 5145 |
"should_evaluate": false,
|
| 5146 |
"should_log": false,
|
| 5147 |
"should_save": true,
|
| 5148 |
-
"should_training_stop":
|
| 5149 |
},
|
| 5150 |
"attributes": {}
|
| 5151 |
}
|
| 5152 |
},
|
| 5153 |
-
"total_flos":
|
| 5154 |
"train_batch_size": 4,
|
| 5155 |
"trial_name": null,
|
| 5156 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 975,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 5131 |
"learning_rate": 1.470196091938691e-05,
|
| 5132 |
"loss": 0.6484,
|
| 5133 |
"step": 732
|
| 5134 |
+
},
|
| 5135 |
+
{
|
| 5136 |
+
"epoch": 0.7517948717948718,
|
| 5137 |
+
"grad_norm": 6.877423286437988,
|
| 5138 |
+
"learning_rate": 1.4587453677213348e-05,
|
| 5139 |
+
"loss": 1.0808,
|
| 5140 |
+
"step": 733
|
| 5141 |
+
},
|
| 5142 |
+
{
|
| 5143 |
+
"epoch": 0.7528205128205128,
|
| 5144 |
+
"grad_norm": 4.961154937744141,
|
| 5145 |
+
"learning_rate": 1.4473317895893772e-05,
|
| 5146 |
+
"loss": 0.7292,
|
| 5147 |
+
"step": 734
|
| 5148 |
+
},
|
| 5149 |
+
{
|
| 5150 |
+
"epoch": 0.7538461538461538,
|
| 5151 |
+
"grad_norm": 5.3160810470581055,
|
| 5152 |
+
"learning_rate": 1.4359554772658552e-05,
|
| 5153 |
+
"loss": 0.7776,
|
| 5154 |
+
"step": 735
|
| 5155 |
+
},
|
| 5156 |
+
{
|
| 5157 |
+
"epoch": 0.7548717948717949,
|
| 5158 |
+
"grad_norm": 5.270435333251953,
|
| 5159 |
+
"learning_rate": 1.4246165500828972e-05,
|
| 5160 |
+
"loss": 0.8105,
|
| 5161 |
+
"step": 736
|
| 5162 |
+
},
|
| 5163 |
+
{
|
| 5164 |
+
"epoch": 0.7558974358974359,
|
| 5165 |
+
"grad_norm": 5.729541301727295,
|
| 5166 |
+
"learning_rate": 1.4133151269804873e-05,
|
| 5167 |
+
"loss": 1.291,
|
| 5168 |
+
"step": 737
|
| 5169 |
+
},
|
| 5170 |
+
{
|
| 5171 |
+
"epoch": 0.7569230769230769,
|
| 5172 |
+
"grad_norm": 3.9068515300750732,
|
| 5173 |
+
"learning_rate": 1.402051326505207e-05,
|
| 5174 |
+
"loss": 0.5514,
|
| 5175 |
+
"step": 738
|
| 5176 |
+
},
|
| 5177 |
+
{
|
| 5178 |
+
"epoch": 0.757948717948718,
|
| 5179 |
+
"grad_norm": 4.7010111808776855,
|
| 5180 |
+
"learning_rate": 1.3908252668089899e-05,
|
| 5181 |
+
"loss": 0.5938,
|
| 5182 |
+
"step": 739
|
| 5183 |
+
},
|
| 5184 |
+
{
|
| 5185 |
+
"epoch": 0.7589743589743589,
|
| 5186 |
+
"grad_norm": 6.777288436889648,
|
| 5187 |
+
"learning_rate": 1.3796370656478935e-05,
|
| 5188 |
+
"loss": 1.1077,
|
| 5189 |
+
"step": 740
|
| 5190 |
+
},
|
| 5191 |
+
{
|
| 5192 |
+
"epoch": 0.76,
|
| 5193 |
+
"grad_norm": 5.429808139801025,
|
| 5194 |
+
"learning_rate": 1.368486840380851e-05,
|
| 5195 |
+
"loss": 0.7201,
|
| 5196 |
+
"step": 741
|
| 5197 |
+
},
|
| 5198 |
+
{
|
| 5199 |
+
"epoch": 0.7610256410256411,
|
| 5200 |
+
"grad_norm": 5.458662033081055,
|
| 5201 |
+
"learning_rate": 1.3573747079684518e-05,
|
| 5202 |
+
"loss": 0.7914,
|
| 5203 |
+
"step": 742
|
| 5204 |
+
},
|
| 5205 |
+
{
|
| 5206 |
+
"epoch": 0.762051282051282,
|
| 5207 |
+
"grad_norm": 4.463566303253174,
|
| 5208 |
+
"learning_rate": 1.3463007849717036e-05,
|
| 5209 |
+
"loss": 0.5937,
|
| 5210 |
+
"step": 743
|
| 5211 |
+
},
|
| 5212 |
+
{
|
| 5213 |
+
"epoch": 0.7630769230769231,
|
| 5214 |
+
"grad_norm": 6.832617282867432,
|
| 5215 |
+
"learning_rate": 1.3352651875508204e-05,
|
| 5216 |
+
"loss": 1.0022,
|
| 5217 |
+
"step": 744
|
| 5218 |
+
},
|
| 5219 |
+
{
|
| 5220 |
+
"epoch": 0.764102564102564,
|
| 5221 |
+
"grad_norm": 7.7007293701171875,
|
| 5222 |
+
"learning_rate": 1.3242680314639993e-05,
|
| 5223 |
+
"loss": 1.1953,
|
| 5224 |
+
"step": 745
|
| 5225 |
+
},
|
| 5226 |
+
{
|
| 5227 |
+
"epoch": 0.7651282051282051,
|
| 5228 |
+
"grad_norm": 6.916568756103516,
|
| 5229 |
+
"learning_rate": 1.3133094320661998e-05,
|
| 5230 |
+
"loss": 0.8883,
|
| 5231 |
+
"step": 746
|
| 5232 |
+
},
|
| 5233 |
+
{
|
| 5234 |
+
"epoch": 0.7661538461538462,
|
| 5235 |
+
"grad_norm": 7.466612815856934,
|
| 5236 |
+
"learning_rate": 1.3023895043079476e-05,
|
| 5237 |
+
"loss": 0.9905,
|
| 5238 |
+
"step": 747
|
| 5239 |
+
},
|
| 5240 |
+
{
|
| 5241 |
+
"epoch": 0.7671794871794871,
|
| 5242 |
+
"grad_norm": 8.147319793701172,
|
| 5243 |
+
"learning_rate": 1.2915083627341128e-05,
|
| 5244 |
+
"loss": 1.0014,
|
| 5245 |
+
"step": 748
|
| 5246 |
+
},
|
| 5247 |
+
{
|
| 5248 |
+
"epoch": 0.7682051282051282,
|
| 5249 |
+
"grad_norm": 9.48543643951416,
|
| 5250 |
+
"learning_rate": 1.2806661214827287e-05,
|
| 5251 |
+
"loss": 0.9548,
|
| 5252 |
+
"step": 749
|
| 5253 |
+
},
|
| 5254 |
+
{
|
| 5255 |
+
"epoch": 0.7692307692307693,
|
| 5256 |
+
"grad_norm": 18.878480911254883,
|
| 5257 |
+
"learning_rate": 1.2698628942837699e-05,
|
| 5258 |
+
"loss": 3.1681,
|
| 5259 |
+
"step": 750
|
| 5260 |
+
},
|
| 5261 |
+
{
|
| 5262 |
+
"epoch": 0.7702564102564102,
|
| 5263 |
+
"grad_norm": 3.2501041889190674,
|
| 5264 |
+
"learning_rate": 1.2590987944579807e-05,
|
| 5265 |
+
"loss": 1.0004,
|
| 5266 |
+
"step": 751
|
| 5267 |
+
},
|
| 5268 |
+
{
|
| 5269 |
+
"epoch": 0.7712820512820513,
|
| 5270 |
+
"grad_norm": 3.6627116203308105,
|
| 5271 |
+
"learning_rate": 1.2483739349156726e-05,
|
| 5272 |
+
"loss": 0.9882,
|
| 5273 |
+
"step": 752
|
| 5274 |
+
},
|
| 5275 |
+
{
|
| 5276 |
+
"epoch": 0.7723076923076924,
|
| 5277 |
+
"grad_norm": 3.411827802658081,
|
| 5278 |
+
"learning_rate": 1.2376884281555484e-05,
|
| 5279 |
+
"loss": 0.8563,
|
| 5280 |
+
"step": 753
|
| 5281 |
+
},
|
| 5282 |
+
{
|
| 5283 |
+
"epoch": 0.7733333333333333,
|
| 5284 |
+
"grad_norm": 4.175601482391357,
|
| 5285 |
+
"learning_rate": 1.2270423862635189e-05,
|
| 5286 |
+
"loss": 1.2118,
|
| 5287 |
+
"step": 754
|
| 5288 |
+
},
|
| 5289 |
+
{
|
| 5290 |
+
"epoch": 0.7743589743589744,
|
| 5291 |
+
"grad_norm": 3.686321973800659,
|
| 5292 |
+
"learning_rate": 1.2164359209115234e-05,
|
| 5293 |
+
"loss": 0.7145,
|
| 5294 |
+
"step": 755
|
| 5295 |
+
},
|
| 5296 |
+
{
|
| 5297 |
+
"epoch": 0.7753846153846153,
|
| 5298 |
+
"grad_norm": 3.6827125549316406,
|
| 5299 |
+
"learning_rate": 1.2058691433563674e-05,
|
| 5300 |
+
"loss": 0.741,
|
| 5301 |
+
"step": 756
|
| 5302 |
+
},
|
| 5303 |
+
{
|
| 5304 |
+
"epoch": 0.7764102564102564,
|
| 5305 |
+
"grad_norm": 7.265366554260254,
|
| 5306 |
+
"learning_rate": 1.1953421644385443e-05,
|
| 5307 |
+
"loss": 0.9565,
|
| 5308 |
+
"step": 757
|
| 5309 |
+
},
|
| 5310 |
+
{
|
| 5311 |
+
"epoch": 0.7774358974358975,
|
| 5312 |
+
"grad_norm": 3.1191012859344482,
|
| 5313 |
+
"learning_rate": 1.1848550945810848e-05,
|
| 5314 |
+
"loss": 0.7104,
|
| 5315 |
+
"step": 758
|
| 5316 |
+
},
|
| 5317 |
+
{
|
| 5318 |
+
"epoch": 0.7784615384615384,
|
| 5319 |
+
"grad_norm": 3.883070468902588,
|
| 5320 |
+
"learning_rate": 1.1744080437883859e-05,
|
| 5321 |
+
"loss": 0.911,
|
| 5322 |
+
"step": 759
|
| 5323 |
+
},
|
| 5324 |
+
{
|
| 5325 |
+
"epoch": 0.7794871794871795,
|
| 5326 |
+
"grad_norm": 3.3209662437438965,
|
| 5327 |
+
"learning_rate": 1.1640011216450691e-05,
|
| 5328 |
+
"loss": 0.7606,
|
| 5329 |
+
"step": 760
|
| 5330 |
+
},
|
| 5331 |
+
{
|
| 5332 |
+
"epoch": 0.7805128205128206,
|
| 5333 |
+
"grad_norm": 4.259766578674316,
|
| 5334 |
+
"learning_rate": 1.1536344373148244e-05,
|
| 5335 |
+
"loss": 0.8399,
|
| 5336 |
+
"step": 761
|
| 5337 |
+
},
|
| 5338 |
+
{
|
| 5339 |
+
"epoch": 0.7815384615384615,
|
| 5340 |
+
"grad_norm": 3.5853612422943115,
|
| 5341 |
+
"learning_rate": 1.1433080995392614e-05,
|
| 5342 |
+
"loss": 0.6282,
|
| 5343 |
+
"step": 762
|
| 5344 |
+
},
|
| 5345 |
+
{
|
| 5346 |
+
"epoch": 0.7825641025641026,
|
| 5347 |
+
"grad_norm": 5.121779918670654,
|
| 5348 |
+
"learning_rate": 1.133022216636781e-05,
|
| 5349 |
+
"loss": 0.9597,
|
| 5350 |
+
"step": 763
|
| 5351 |
+
},
|
| 5352 |
+
{
|
| 5353 |
+
"epoch": 0.7835897435897435,
|
| 5354 |
+
"grad_norm": 3.7153100967407227,
|
| 5355 |
+
"learning_rate": 1.1227768965014246e-05,
|
| 5356 |
+
"loss": 0.9798,
|
| 5357 |
+
"step": 764
|
| 5358 |
+
},
|
| 5359 |
+
{
|
| 5360 |
+
"epoch": 0.7846153846153846,
|
| 5361 |
+
"grad_norm": 4.351454734802246,
|
| 5362 |
+
"learning_rate": 1.1125722466017547e-05,
|
| 5363 |
+
"loss": 0.84,
|
| 5364 |
+
"step": 765
|
| 5365 |
+
},
|
| 5366 |
+
{
|
| 5367 |
+
"epoch": 0.7856410256410257,
|
| 5368 |
+
"grad_norm": 4.489700794219971,
|
| 5369 |
+
"learning_rate": 1.102408373979717e-05,
|
| 5370 |
+
"loss": 0.9916,
|
| 5371 |
+
"step": 766
|
| 5372 |
+
},
|
| 5373 |
+
{
|
| 5374 |
+
"epoch": 0.7866666666666666,
|
| 5375 |
+
"grad_norm": 3.369802951812744,
|
| 5376 |
+
"learning_rate": 1.092285385249528e-05,
|
| 5377 |
+
"loss": 0.541,
|
| 5378 |
+
"step": 767
|
| 5379 |
+
},
|
| 5380 |
+
{
|
| 5381 |
+
"epoch": 0.7876923076923077,
|
| 5382 |
+
"grad_norm": 4.338639736175537,
|
| 5383 |
+
"learning_rate": 1.0822033865965503e-05,
|
| 5384 |
+
"loss": 0.8406,
|
| 5385 |
+
"step": 768
|
| 5386 |
+
},
|
| 5387 |
+
{
|
| 5388 |
+
"epoch": 0.7887179487179488,
|
| 5389 |
+
"grad_norm": 3.770918607711792,
|
| 5390 |
+
"learning_rate": 1.0721624837761768e-05,
|
| 5391 |
+
"loss": 0.7247,
|
| 5392 |
+
"step": 769
|
| 5393 |
+
},
|
| 5394 |
+
{
|
| 5395 |
+
"epoch": 0.7897435897435897,
|
| 5396 |
+
"grad_norm": 4.7859039306640625,
|
| 5397 |
+
"learning_rate": 1.0621627821127289e-05,
|
| 5398 |
+
"loss": 0.7578,
|
| 5399 |
+
"step": 770
|
| 5400 |
+
},
|
| 5401 |
+
{
|
| 5402 |
+
"epoch": 0.7907692307692308,
|
| 5403 |
+
"grad_norm": 4.7032623291015625,
|
| 5404 |
+
"learning_rate": 1.0522043864983428e-05,
|
| 5405 |
+
"loss": 0.8842,
|
| 5406 |
+
"step": 771
|
| 5407 |
+
},
|
| 5408 |
+
{
|
| 5409 |
+
"epoch": 0.7917948717948718,
|
| 5410 |
+
"grad_norm": 5.078568458557129,
|
| 5411 |
+
"learning_rate": 1.0422874013918793e-05,
|
| 5412 |
+
"loss": 0.9599,
|
| 5413 |
+
"step": 772
|
| 5414 |
+
},
|
| 5415 |
+
{
|
| 5416 |
+
"epoch": 0.7928205128205128,
|
| 5417 |
+
"grad_norm": 4.564446449279785,
|
| 5418 |
+
"learning_rate": 1.0324119308178166e-05,
|
| 5419 |
+
"loss": 0.9349,
|
| 5420 |
+
"step": 773
|
| 5421 |
+
},
|
| 5422 |
+
{
|
| 5423 |
+
"epoch": 0.7938461538461539,
|
| 5424 |
+
"grad_norm": 4.562546253204346,
|
| 5425 |
+
"learning_rate": 1.022578078365169e-05,
|
| 5426 |
+
"loss": 0.7547,
|
| 5427 |
+
"step": 774
|
| 5428 |
+
},
|
| 5429 |
+
{
|
| 5430 |
+
"epoch": 0.7948717948717948,
|
| 5431 |
+
"grad_norm": 4.574657917022705,
|
| 5432 |
+
"learning_rate": 1.012785947186397e-05,
|
| 5433 |
+
"loss": 0.6049,
|
| 5434 |
+
"step": 775
|
| 5435 |
+
},
|
| 5436 |
+
{
|
| 5437 |
+
"epoch": 0.7958974358974359,
|
| 5438 |
+
"grad_norm": 4.5475053787231445,
|
| 5439 |
+
"learning_rate": 1.0030356399963204e-05,
|
| 5440 |
+
"loss": 0.6777,
|
| 5441 |
+
"step": 776
|
| 5442 |
+
},
|
| 5443 |
+
{
|
| 5444 |
+
"epoch": 0.796923076923077,
|
| 5445 |
+
"grad_norm": 3.587956190109253,
|
| 5446 |
+
"learning_rate": 9.933272590710507e-06,
|
| 5447 |
+
"loss": 0.5728,
|
| 5448 |
+
"step": 777
|
| 5449 |
+
},
|
| 5450 |
+
{
|
| 5451 |
+
"epoch": 0.7979487179487179,
|
| 5452 |
+
"grad_norm": 3.741227388381958,
|
| 5453 |
+
"learning_rate": 9.836609062469066e-06,
|
| 5454 |
+
"loss": 0.5207,
|
| 5455 |
+
"step": 778
|
| 5456 |
+
},
|
| 5457 |
+
{
|
| 5458 |
+
"epoch": 0.798974358974359,
|
| 5459 |
+
"grad_norm": 6.142617702484131,
|
| 5460 |
+
"learning_rate": 9.740366829193587e-06,
|
| 5461 |
+
"loss": 1.1337,
|
| 5462 |
+
"step": 779
|
| 5463 |
+
},
|
| 5464 |
+
{
|
| 5465 |
+
"epoch": 0.8,
|
| 5466 |
+
"grad_norm": 4.094137191772461,
|
| 5467 |
+
"learning_rate": 9.644546900419533e-06,
|
| 5468 |
+
"loss": 0.6916,
|
| 5469 |
+
"step": 780
|
| 5470 |
+
},
|
| 5471 |
+
{
|
| 5472 |
+
"epoch": 0.801025641025641,
|
| 5473 |
+
"grad_norm": 5.459488391876221,
|
| 5474 |
+
"learning_rate": 9.549150281252633e-06,
|
| 5475 |
+
"loss": 0.9617,
|
| 5476 |
+
"step": 781
|
| 5477 |
+
},
|
| 5478 |
+
{
|
| 5479 |
+
"epoch": 0.8020512820512821,
|
| 5480 |
+
"grad_norm": 5.044382095336914,
|
| 5481 |
+
"learning_rate": 9.454177972358259e-06,
|
| 5482 |
+
"loss": 0.9123,
|
| 5483 |
+
"step": 782
|
| 5484 |
+
},
|
| 5485 |
+
{
|
| 5486 |
+
"epoch": 0.803076923076923,
|
| 5487 |
+
"grad_norm": 4.838907718658447,
|
| 5488 |
+
"learning_rate": 9.35963096995101e-06,
|
| 5489 |
+
"loss": 0.8702,
|
| 5490 |
+
"step": 783
|
| 5491 |
+
},
|
| 5492 |
+
{
|
| 5493 |
+
"epoch": 0.8041025641025641,
|
| 5494 |
+
"grad_norm": 5.395598411560059,
|
| 5495 |
+
"learning_rate": 9.265510265784189e-06,
|
| 5496 |
+
"loss": 1.2312,
|
| 5497 |
+
"step": 784
|
| 5498 |
+
},
|
| 5499 |
+
{
|
| 5500 |
+
"epoch": 0.8051282051282052,
|
| 5501 |
+
"grad_norm": 3.8278088569641113,
|
| 5502 |
+
"learning_rate": 9.171816847139448e-06,
|
| 5503 |
+
"loss": 0.5905,
|
| 5504 |
+
"step": 785
|
| 5505 |
+
},
|
| 5506 |
+
{
|
| 5507 |
+
"epoch": 0.8061538461538461,
|
| 5508 |
+
"grad_norm": 4.812705039978027,
|
| 5509 |
+
"learning_rate": 9.078551696816434e-06,
|
| 5510 |
+
"loss": 0.7691,
|
| 5511 |
+
"step": 786
|
| 5512 |
+
},
|
| 5513 |
+
{
|
| 5514 |
+
"epoch": 0.8071794871794872,
|
| 5515 |
+
"grad_norm": 5.471896648406982,
|
| 5516 |
+
"learning_rate": 8.985715793122407e-06,
|
| 5517 |
+
"loss": 0.9679,
|
| 5518 |
+
"step": 787
|
| 5519 |
+
},
|
| 5520 |
+
{
|
| 5521 |
+
"epoch": 0.8082051282051282,
|
| 5522 |
+
"grad_norm": 5.208366394042969,
|
| 5523 |
+
"learning_rate": 8.893310109862102e-06,
|
| 5524 |
+
"loss": 1.0003,
|
| 5525 |
+
"step": 788
|
| 5526 |
+
},
|
| 5527 |
+
{
|
| 5528 |
+
"epoch": 0.8092307692307692,
|
| 5529 |
+
"grad_norm": 5.83384370803833,
|
| 5530 |
+
"learning_rate": 8.801335616327377e-06,
|
| 5531 |
+
"loss": 0.853,
|
| 5532 |
+
"step": 789
|
| 5533 |
+
},
|
| 5534 |
+
{
|
| 5535 |
+
"epoch": 0.8102564102564103,
|
| 5536 |
+
"grad_norm": 4.475183963775635,
|
| 5537 |
+
"learning_rate": 8.70979327728718e-06,
|
| 5538 |
+
"loss": 0.6102,
|
| 5539 |
+
"step": 790
|
| 5540 |
+
},
|
| 5541 |
+
{
|
| 5542 |
+
"epoch": 0.8112820512820513,
|
| 5543 |
+
"grad_norm": 4.50262975692749,
|
| 5544 |
+
"learning_rate": 8.618684052977305e-06,
|
| 5545 |
+
"loss": 0.6673,
|
| 5546 |
+
"step": 791
|
| 5547 |
+
},
|
| 5548 |
+
{
|
| 5549 |
+
"epoch": 0.8123076923076923,
|
| 5550 |
+
"grad_norm": 6.1138691902160645,
|
| 5551 |
+
"learning_rate": 8.528008899090412e-06,
|
| 5552 |
+
"loss": 1.1774,
|
| 5553 |
+
"step": 792
|
| 5554 |
+
},
|
| 5555 |
+
{
|
| 5556 |
+
"epoch": 0.8133333333333334,
|
| 5557 |
+
"grad_norm": 5.287784099578857,
|
| 5558 |
+
"learning_rate": 8.437768766765974e-06,
|
| 5559 |
+
"loss": 0.8959,
|
| 5560 |
+
"step": 793
|
| 5561 |
+
},
|
| 5562 |
+
{
|
| 5563 |
+
"epoch": 0.8143589743589743,
|
| 5564 |
+
"grad_norm": 4.6912994384765625,
|
| 5565 |
+
"learning_rate": 8.347964602580243e-06,
|
| 5566 |
+
"loss": 0.5866,
|
| 5567 |
+
"step": 794
|
| 5568 |
+
},
|
| 5569 |
+
{
|
| 5570 |
+
"epoch": 0.8153846153846154,
|
| 5571 |
+
"grad_norm": 4.410092353820801,
|
| 5572 |
+
"learning_rate": 8.25859734853645e-06,
|
| 5573 |
+
"loss": 0.7006,
|
| 5574 |
+
"step": 795
|
| 5575 |
+
},
|
| 5576 |
+
{
|
| 5577 |
+
"epoch": 0.8164102564102564,
|
| 5578 |
+
"grad_norm": 6.904149532318115,
|
| 5579 |
+
"learning_rate": 8.169667942054759e-06,
|
| 5580 |
+
"loss": 1.2351,
|
| 5581 |
+
"step": 796
|
| 5582 |
+
},
|
| 5583 |
+
{
|
| 5584 |
+
"epoch": 0.8174358974358974,
|
| 5585 |
+
"grad_norm": 6.1469550132751465,
|
| 5586 |
+
"learning_rate": 8.081177315962602e-06,
|
| 5587 |
+
"loss": 0.9114,
|
| 5588 |
+
"step": 797
|
| 5589 |
+
},
|
| 5590 |
+
{
|
| 5591 |
+
"epoch": 0.8184615384615385,
|
| 5592 |
+
"grad_norm": 7.1000895500183105,
|
| 5593 |
+
"learning_rate": 7.99312639848474e-06,
|
| 5594 |
+
"loss": 1.0631,
|
| 5595 |
+
"step": 798
|
| 5596 |
+
},
|
| 5597 |
+
{
|
| 5598 |
+
"epoch": 0.8194871794871795,
|
| 5599 |
+
"grad_norm": 7.745519638061523,
|
| 5600 |
+
"learning_rate": 7.905516113233653e-06,
|
| 5601 |
+
"loss": 0.9995,
|
| 5602 |
+
"step": 799
|
| 5603 |
+
},
|
| 5604 |
+
{
|
| 5605 |
+
"epoch": 0.8205128205128205,
|
| 5606 |
+
"grad_norm": 14.297471046447754,
|
| 5607 |
+
"learning_rate": 7.81834737919978e-06,
|
| 5608 |
+
"loss": 2.0584,
|
| 5609 |
+
"step": 800
|
| 5610 |
+
},
|
| 5611 |
+
{
|
| 5612 |
+
"epoch": 0.8215384615384616,
|
| 5613 |
+
"grad_norm": 3.660001754760742,
|
| 5614 |
+
"learning_rate": 7.731621110741872e-06,
|
| 5615 |
+
"loss": 0.9644,
|
| 5616 |
+
"step": 801
|
| 5617 |
+
},
|
| 5618 |
+
{
|
| 5619 |
+
"epoch": 0.8225641025641026,
|
| 5620 |
+
"grad_norm": 3.4968271255493164,
|
| 5621 |
+
"learning_rate": 7.645338217577475e-06,
|
| 5622 |
+
"loss": 0.9172,
|
| 5623 |
+
"step": 802
|
| 5624 |
+
},
|
| 5625 |
+
{
|
| 5626 |
+
"epoch": 0.8235897435897436,
|
| 5627 |
+
"grad_norm": 3.162583351135254,
|
| 5628 |
+
"learning_rate": 7.559499604773279e-06,
|
| 5629 |
+
"loss": 0.7705,
|
| 5630 |
+
"step": 803
|
| 5631 |
+
},
|
| 5632 |
+
{
|
| 5633 |
+
"epoch": 0.8246153846153846,
|
| 5634 |
+
"grad_norm": 3.755932331085205,
|
| 5635 |
+
"learning_rate": 7.474106172735746e-06,
|
| 5636 |
+
"loss": 0.8697,
|
| 5637 |
+
"step": 804
|
| 5638 |
+
},
|
| 5639 |
+
{
|
| 5640 |
+
"epoch": 0.8256410256410256,
|
| 5641 |
+
"grad_norm": 3.186575412750244,
|
| 5642 |
+
"learning_rate": 7.389158817201542e-06,
|
| 5643 |
+
"loss": 0.903,
|
| 5644 |
+
"step": 805
|
| 5645 |
+
},
|
| 5646 |
+
{
|
| 5647 |
+
"epoch": 0.8266666666666667,
|
| 5648 |
+
"grad_norm": 3.6311652660369873,
|
| 5649 |
+
"learning_rate": 7.304658429228245e-06,
|
| 5650 |
+
"loss": 1.001,
|
| 5651 |
+
"step": 806
|
| 5652 |
+
},
|
| 5653 |
+
{
|
| 5654 |
+
"epoch": 0.8276923076923077,
|
| 5655 |
+
"grad_norm": 3.348775625228882,
|
| 5656 |
+
"learning_rate": 7.220605895184945e-06,
|
| 5657 |
+
"loss": 0.75,
|
| 5658 |
+
"step": 807
|
| 5659 |
+
},
|
| 5660 |
+
{
|
| 5661 |
+
"epoch": 0.8287179487179487,
|
| 5662 |
+
"grad_norm": 3.6370596885681152,
|
| 5663 |
+
"learning_rate": 7.13700209674294e-06,
|
| 5664 |
+
"loss": 0.9634,
|
| 5665 |
+
"step": 808
|
| 5666 |
+
},
|
| 5667 |
+
{
|
| 5668 |
+
"epoch": 0.8297435897435897,
|
| 5669 |
+
"grad_norm": 4.622466564178467,
|
| 5670 |
+
"learning_rate": 7.053847910866513e-06,
|
| 5671 |
+
"loss": 1.2344,
|
| 5672 |
+
"step": 809
|
| 5673 |
+
},
|
| 5674 |
+
{
|
| 5675 |
+
"epoch": 0.8307692307692308,
|
| 5676 |
+
"grad_norm": 3.5539474487304688,
|
| 5677 |
+
"learning_rate": 6.9711442098037375e-06,
|
| 5678 |
+
"loss": 0.8593,
|
| 5679 |
+
"step": 810
|
| 5680 |
+
},
|
| 5681 |
+
{
|
| 5682 |
+
"epoch": 0.8317948717948718,
|
| 5683 |
+
"grad_norm": 3.8597571849823,
|
| 5684 |
+
"learning_rate": 6.888891861077301e-06,
|
| 5685 |
+
"loss": 0.7546,
|
| 5686 |
+
"step": 811
|
| 5687 |
+
},
|
| 5688 |
+
{
|
| 5689 |
+
"epoch": 0.8328205128205128,
|
| 5690 |
+
"grad_norm": 3.2743446826934814,
|
| 5691 |
+
"learning_rate": 6.807091727475412e-06,
|
| 5692 |
+
"loss": 0.6182,
|
| 5693 |
+
"step": 812
|
| 5694 |
+
},
|
| 5695 |
+
{
|
| 5696 |
+
"epoch": 0.8338461538461538,
|
| 5697 |
+
"grad_norm": 5.070436477661133,
|
| 5698 |
+
"learning_rate": 6.725744667042777e-06,
|
| 5699 |
+
"loss": 1.3895,
|
| 5700 |
+
"step": 813
|
| 5701 |
+
},
|
| 5702 |
+
{
|
| 5703 |
+
"epoch": 0.8348717948717949,
|
| 5704 |
+
"grad_norm": 3.76391339302063,
|
| 5705 |
+
"learning_rate": 6.644851533071556e-06,
|
| 5706 |
+
"loss": 0.6863,
|
| 5707 |
+
"step": 814
|
| 5708 |
+
},
|
| 5709 |
+
{
|
| 5710 |
+
"epoch": 0.8358974358974359,
|
| 5711 |
+
"grad_norm": 3.7185099124908447,
|
| 5712 |
+
"learning_rate": 6.564413174092443e-06,
|
| 5713 |
+
"loss": 0.6729,
|
| 5714 |
+
"step": 815
|
| 5715 |
+
},
|
| 5716 |
+
{
|
| 5717 |
+
"epoch": 0.8369230769230769,
|
| 5718 |
+
"grad_norm": 3.9108810424804688,
|
| 5719 |
+
"learning_rate": 6.4844304338657844e-06,
|
| 5720 |
+
"loss": 0.8355,
|
| 5721 |
+
"step": 816
|
| 5722 |
+
},
|
| 5723 |
+
{
|
| 5724 |
+
"epoch": 0.837948717948718,
|
| 5725 |
+
"grad_norm": 4.387764930725098,
|
| 5726 |
+
"learning_rate": 6.4049041513726485e-06,
|
| 5727 |
+
"loss": 0.9891,
|
| 5728 |
+
"step": 817
|
| 5729 |
+
},
|
| 5730 |
+
{
|
| 5731 |
+
"epoch": 0.838974358974359,
|
| 5732 |
+
"grad_norm": 4.212962627410889,
|
| 5733 |
+
"learning_rate": 6.3258351608061315e-06,
|
| 5734 |
+
"loss": 1.0,
|
| 5735 |
+
"step": 818
|
| 5736 |
+
},
|
| 5737 |
+
{
|
| 5738 |
+
"epoch": 0.84,
|
| 5739 |
+
"grad_norm": 3.4375810623168945,
|
| 5740 |
+
"learning_rate": 6.247224291562509e-06,
|
| 5741 |
+
"loss": 0.6408,
|
| 5742 |
+
"step": 819
|
| 5743 |
+
},
|
| 5744 |
+
{
|
| 5745 |
+
"epoch": 0.841025641025641,
|
| 5746 |
+
"grad_norm": 4.018990993499756,
|
| 5747 |
+
"learning_rate": 6.16907236823262e-06,
|
| 5748 |
+
"loss": 0.7336,
|
| 5749 |
+
"step": 820
|
| 5750 |
+
},
|
| 5751 |
+
{
|
| 5752 |
+
"epoch": 0.8420512820512821,
|
| 5753 |
+
"grad_norm": 4.63192892074585,
|
| 5754 |
+
"learning_rate": 6.091380210593145e-06,
|
| 5755 |
+
"loss": 0.9198,
|
| 5756 |
+
"step": 821
|
| 5757 |
+
},
|
| 5758 |
+
{
|
| 5759 |
+
"epoch": 0.8430769230769231,
|
| 5760 |
+
"grad_norm": 3.9990158081054688,
|
| 5761 |
+
"learning_rate": 6.014148633598055e-06,
|
| 5762 |
+
"loss": 0.852,
|
| 5763 |
+
"step": 822
|
| 5764 |
+
},
|
| 5765 |
+
{
|
| 5766 |
+
"epoch": 0.8441025641025641,
|
| 5767 |
+
"grad_norm": 4.2810163497924805,
|
| 5768 |
+
"learning_rate": 5.937378447370067e-06,
|
| 5769 |
+
"loss": 0.7765,
|
| 5770 |
+
"step": 823
|
| 5771 |
+
},
|
| 5772 |
+
{
|
| 5773 |
+
"epoch": 0.8451282051282051,
|
| 5774 |
+
"grad_norm": 4.975213527679443,
|
| 5775 |
+
"learning_rate": 5.86107045719208e-06,
|
| 5776 |
+
"loss": 1.2429,
|
| 5777 |
+
"step": 824
|
| 5778 |
+
},
|
| 5779 |
+
{
|
| 5780 |
+
"epoch": 0.8461538461538461,
|
| 5781 |
+
"grad_norm": 4.42708158493042,
|
| 5782 |
+
"learning_rate": 5.785225463498828e-06,
|
| 5783 |
+
"loss": 0.8026,
|
| 5784 |
+
"step": 825
|
| 5785 |
+
},
|
| 5786 |
+
{
|
| 5787 |
+
"epoch": 0.8471794871794872,
|
| 5788 |
+
"grad_norm": 4.280889987945557,
|
| 5789 |
+
"learning_rate": 5.70984426186838e-06,
|
| 5790 |
+
"loss": 0.7082,
|
| 5791 |
+
"step": 826
|
| 5792 |
+
},
|
| 5793 |
+
{
|
| 5794 |
+
"epoch": 0.8482051282051282,
|
| 5795 |
+
"grad_norm": 6.7891645431518555,
|
| 5796 |
+
"learning_rate": 5.634927643013899e-06,
|
| 5797 |
+
"loss": 0.7442,
|
| 5798 |
+
"step": 827
|
| 5799 |
+
},
|
| 5800 |
+
{
|
| 5801 |
+
"epoch": 0.8492307692307692,
|
| 5802 |
+
"grad_norm": 4.235758304595947,
|
| 5803 |
+
"learning_rate": 5.560476392775238e-06,
|
| 5804 |
+
"loss": 0.8627,
|
| 5805 |
+
"step": 828
|
| 5806 |
+
},
|
| 5807 |
+
{
|
| 5808 |
+
"epoch": 0.8502564102564103,
|
| 5809 |
+
"grad_norm": 4.850736141204834,
|
| 5810 |
+
"learning_rate": 5.486491292110796e-06,
|
| 5811 |
+
"loss": 0.8016,
|
| 5812 |
+
"step": 829
|
| 5813 |
+
},
|
| 5814 |
+
{
|
| 5815 |
+
"epoch": 0.8512820512820513,
|
| 5816 |
+
"grad_norm": 3.9992334842681885,
|
| 5817 |
+
"learning_rate": 5.412973117089287e-06,
|
| 5818 |
+
"loss": 0.7891,
|
| 5819 |
+
"step": 830
|
| 5820 |
+
},
|
| 5821 |
+
{
|
| 5822 |
+
"epoch": 0.8523076923076923,
|
| 5823 |
+
"grad_norm": 4.047679901123047,
|
| 5824 |
+
"learning_rate": 5.339922638881545e-06,
|
| 5825 |
+
"loss": 0.5642,
|
| 5826 |
+
"step": 831
|
| 5827 |
+
},
|
| 5828 |
+
{
|
| 5829 |
+
"epoch": 0.8533333333333334,
|
| 5830 |
+
"grad_norm": 5.204737663269043,
|
| 5831 |
+
"learning_rate": 5.267340623752553e-06,
|
| 5832 |
+
"loss": 1.0036,
|
| 5833 |
+
"step": 832
|
| 5834 |
+
},
|
| 5835 |
+
{
|
| 5836 |
+
"epoch": 0.8543589743589743,
|
| 5837 |
+
"grad_norm": 5.493986129760742,
|
| 5838 |
+
"learning_rate": 5.195227833053274e-06,
|
| 5839 |
+
"loss": 0.9189,
|
| 5840 |
+
"step": 833
|
| 5841 |
+
},
|
| 5842 |
+
{
|
| 5843 |
+
"epoch": 0.8553846153846154,
|
| 5844 |
+
"grad_norm": 6.99483060836792,
|
| 5845 |
+
"learning_rate": 5.123585023212785e-06,
|
| 5846 |
+
"loss": 1.2142,
|
| 5847 |
+
"step": 834
|
| 5848 |
+
},
|
| 5849 |
+
{
|
| 5850 |
+
"epoch": 0.8564102564102564,
|
| 5851 |
+
"grad_norm": 4.1271491050720215,
|
| 5852 |
+
"learning_rate": 5.05241294573024e-06,
|
| 5853 |
+
"loss": 0.5715,
|
| 5854 |
+
"step": 835
|
| 5855 |
+
},
|
| 5856 |
+
{
|
| 5857 |
+
"epoch": 0.8574358974358974,
|
| 5858 |
+
"grad_norm": 5.13178014755249,
|
| 5859 |
+
"learning_rate": 4.981712347167061e-06,
|
| 5860 |
+
"loss": 0.9571,
|
| 5861 |
+
"step": 836
|
| 5862 |
+
},
|
| 5863 |
+
{
|
| 5864 |
+
"epoch": 0.8584615384615385,
|
| 5865 |
+
"grad_norm": 5.234194755554199,
|
| 5866 |
+
"learning_rate": 4.9114839691390854e-06,
|
| 5867 |
+
"loss": 0.9082,
|
| 5868 |
+
"step": 837
|
| 5869 |
+
},
|
| 5870 |
+
{
|
| 5871 |
+
"epoch": 0.8594871794871795,
|
| 5872 |
+
"grad_norm": 5.754642963409424,
|
| 5873 |
+
"learning_rate": 4.841728548308744e-06,
|
| 5874 |
+
"loss": 0.9004,
|
| 5875 |
+
"step": 838
|
| 5876 |
+
},
|
| 5877 |
+
{
|
| 5878 |
+
"epoch": 0.8605128205128205,
|
| 5879 |
+
"grad_norm": 4.62493371963501,
|
| 5880 |
+
"learning_rate": 4.772446816377407e-06,
|
| 5881 |
+
"loss": 0.6618,
|
| 5882 |
+
"step": 839
|
| 5883 |
+
},
|
| 5884 |
+
{
|
| 5885 |
+
"epoch": 0.8615384615384616,
|
| 5886 |
+
"grad_norm": 6.897546768188477,
|
| 5887 |
+
"learning_rate": 4.703639500077656e-06,
|
| 5888 |
+
"loss": 1.1324,
|
| 5889 |
+
"step": 840
|
| 5890 |
+
},
|
| 5891 |
+
{
|
| 5892 |
+
"epoch": 0.8625641025641025,
|
| 5893 |
+
"grad_norm": 4.677278995513916,
|
| 5894 |
+
"learning_rate": 4.635307321165688e-06,
|
| 5895 |
+
"loss": 0.6254,
|
| 5896 |
+
"step": 841
|
| 5897 |
+
},
|
| 5898 |
+
{
|
| 5899 |
+
"epoch": 0.8635897435897436,
|
| 5900 |
+
"grad_norm": 6.221714973449707,
|
| 5901 |
+
"learning_rate": 4.567450996413713e-06,
|
| 5902 |
+
"loss": 1.146,
|
| 5903 |
+
"step": 842
|
| 5904 |
+
},
|
| 5905 |
+
{
|
| 5906 |
+
"epoch": 0.8646153846153846,
|
| 5907 |
+
"grad_norm": 5.52736759185791,
|
| 5908 |
+
"learning_rate": 4.500071237602482e-06,
|
| 5909 |
+
"loss": 0.8631,
|
| 5910 |
+
"step": 843
|
| 5911 |
+
},
|
| 5912 |
+
{
|
| 5913 |
+
"epoch": 0.8656410256410256,
|
| 5914 |
+
"grad_norm": 5.572301387786865,
|
| 5915 |
+
"learning_rate": 4.433168751513761e-06,
|
| 5916 |
+
"loss": 0.8625,
|
| 5917 |
+
"step": 844
|
| 5918 |
+
},
|
| 5919 |
+
{
|
| 5920 |
+
"epoch": 0.8666666666666667,
|
| 5921 |
+
"grad_norm": 6.426190376281738,
|
| 5922 |
+
"learning_rate": 4.366744239922998e-06,
|
| 5923 |
+
"loss": 0.8229,
|
| 5924 |
+
"step": 845
|
| 5925 |
+
},
|
| 5926 |
+
{
|
| 5927 |
+
"epoch": 0.8676923076923077,
|
| 5928 |
+
"grad_norm": 7.4777679443359375,
|
| 5929 |
+
"learning_rate": 4.30079839959186e-06,
|
| 5930 |
+
"loss": 0.9241,
|
| 5931 |
+
"step": 846
|
| 5932 |
+
},
|
| 5933 |
+
{
|
| 5934 |
+
"epoch": 0.8687179487179487,
|
| 5935 |
+
"grad_norm": 6.525635719299316,
|
| 5936 |
+
"learning_rate": 4.235331922261026e-06,
|
| 5937 |
+
"loss": 0.9205,
|
| 5938 |
+
"step": 847
|
| 5939 |
+
},
|
| 5940 |
+
{
|
| 5941 |
+
"epoch": 0.8697435897435898,
|
| 5942 |
+
"grad_norm": 7.924921035766602,
|
| 5943 |
+
"learning_rate": 4.170345494642863e-06,
|
| 5944 |
+
"loss": 1.4281,
|
| 5945 |
+
"step": 848
|
| 5946 |
+
},
|
| 5947 |
+
{
|
| 5948 |
+
"epoch": 0.8707692307692307,
|
| 5949 |
+
"grad_norm": 8.663949966430664,
|
| 5950 |
+
"learning_rate": 4.10583979841424e-06,
|
| 5951 |
+
"loss": 1.4309,
|
| 5952 |
+
"step": 849
|
| 5953 |
+
},
|
| 5954 |
+
{
|
| 5955 |
+
"epoch": 0.8717948717948718,
|
| 5956 |
+
"grad_norm": 14.360984802246094,
|
| 5957 |
+
"learning_rate": 4.041815510209396e-06,
|
| 5958 |
+
"loss": 1.9381,
|
| 5959 |
+
"step": 850
|
| 5960 |
+
},
|
| 5961 |
+
{
|
| 5962 |
+
"epoch": 0.8728205128205129,
|
| 5963 |
+
"grad_norm": 3.2258362770080566,
|
| 5964 |
+
"learning_rate": 3.9782733016128005e-06,
|
| 5965 |
+
"loss": 0.9545,
|
| 5966 |
+
"step": 851
|
| 5967 |
+
},
|
| 5968 |
+
{
|
| 5969 |
+
"epoch": 0.8738461538461538,
|
| 5970 |
+
"grad_norm": 3.155390739440918,
|
| 5971 |
+
"learning_rate": 3.915213839152176e-06,
|
| 5972 |
+
"loss": 1.0386,
|
| 5973 |
+
"step": 852
|
| 5974 |
+
},
|
| 5975 |
+
{
|
| 5976 |
+
"epoch": 0.8748717948717949,
|
| 5977 |
+
"grad_norm": 3.149043560028076,
|
| 5978 |
+
"learning_rate": 3.852637784291424e-06,
|
| 5979 |
+
"loss": 0.9249,
|
| 5980 |
+
"step": 853
|
| 5981 |
+
},
|
| 5982 |
+
{
|
| 5983 |
+
"epoch": 0.8758974358974358,
|
| 5984 |
+
"grad_norm": 3.604149341583252,
|
| 5985 |
+
"learning_rate": 3.790545793423761e-06,
|
| 5986 |
+
"loss": 1.0529,
|
| 5987 |
+
"step": 854
|
| 5988 |
+
},
|
| 5989 |
+
{
|
| 5990 |
+
"epoch": 0.8769230769230769,
|
| 5991 |
+
"grad_norm": 3.0421512126922607,
|
| 5992 |
+
"learning_rate": 3.728938517864794e-06,
|
| 5993 |
+
"loss": 0.7461,
|
| 5994 |
+
"step": 855
|
| 5995 |
+
},
|
| 5996 |
+
{
|
| 5997 |
+
"epoch": 0.877948717948718,
|
| 5998 |
+
"grad_norm": 3.300569534301758,
|
| 5999 |
+
"learning_rate": 3.6678166038456807e-06,
|
| 6000 |
+
"loss": 0.7742,
|
| 6001 |
+
"step": 856
|
| 6002 |
+
},
|
| 6003 |
+
{
|
| 6004 |
+
"epoch": 0.8789743589743589,
|
| 6005 |
+
"grad_norm": 3.0753672122955322,
|
| 6006 |
+
"learning_rate": 3.60718069250639e-06,
|
| 6007 |
+
"loss": 0.6109,
|
| 6008 |
+
"step": 857
|
| 6009 |
+
},
|
| 6010 |
+
{
|
| 6011 |
+
"epoch": 0.88,
|
| 6012 |
+
"grad_norm": 4.064855098724365,
|
| 6013 |
+
"learning_rate": 3.5470314198889185e-06,
|
| 6014 |
+
"loss": 0.9794,
|
| 6015 |
+
"step": 858
|
| 6016 |
+
},
|
| 6017 |
+
{
|
| 6018 |
+
"epoch": 0.8810256410256411,
|
| 6019 |
+
"grad_norm": 3.6606357097625732,
|
| 6020 |
+
"learning_rate": 3.4873694169306913e-06,
|
| 6021 |
+
"loss": 0.7496,
|
| 6022 |
+
"step": 859
|
| 6023 |
+
},
|
| 6024 |
+
{
|
| 6025 |
+
"epoch": 0.882051282051282,
|
| 6026 |
+
"grad_norm": 4.065293788909912,
|
| 6027 |
+
"learning_rate": 3.4281953094578877e-06,
|
| 6028 |
+
"loss": 0.904,
|
| 6029 |
+
"step": 860
|
| 6030 |
+
},
|
| 6031 |
+
{
|
| 6032 |
+
"epoch": 0.8830769230769231,
|
| 6033 |
+
"grad_norm": 3.2183470726013184,
|
| 6034 |
+
"learning_rate": 3.369509718178887e-06,
|
| 6035 |
+
"loss": 0.5392,
|
| 6036 |
+
"step": 861
|
| 6037 |
+
},
|
| 6038 |
+
{
|
| 6039 |
+
"epoch": 0.884102564102564,
|
| 6040 |
+
"grad_norm": 4.25131368637085,
|
| 6041 |
+
"learning_rate": 3.3113132586777784e-06,
|
| 6042 |
+
"loss": 0.6751,
|
| 6043 |
+
"step": 862
|
| 6044 |
+
},
|
| 6045 |
+
{
|
| 6046 |
+
"epoch": 0.8851282051282051,
|
| 6047 |
+
"grad_norm": 4.080262184143066,
|
| 6048 |
+
"learning_rate": 3.253606541407872e-06,
|
| 6049 |
+
"loss": 1.0854,
|
| 6050 |
+
"step": 863
|
| 6051 |
+
},
|
| 6052 |
+
{
|
| 6053 |
+
"epoch": 0.8861538461538462,
|
| 6054 |
+
"grad_norm": 5.6910247802734375,
|
| 6055 |
+
"learning_rate": 3.196390171685343e-06,
|
| 6056 |
+
"loss": 1.1699,
|
| 6057 |
+
"step": 864
|
| 6058 |
+
},
|
| 6059 |
+
{
|
| 6060 |
+
"epoch": 0.8871794871794871,
|
| 6061 |
+
"grad_norm": 3.5453498363494873,
|
| 6062 |
+
"learning_rate": 3.1396647496828247e-06,
|
| 6063 |
+
"loss": 0.8176,
|
| 6064 |
+
"step": 865
|
| 6065 |
+
},
|
| 6066 |
+
{
|
| 6067 |
+
"epoch": 0.8882051282051282,
|
| 6068 |
+
"grad_norm": 4.133404731750488,
|
| 6069 |
+
"learning_rate": 3.0834308704231484e-06,
|
| 6070 |
+
"loss": 0.7646,
|
| 6071 |
+
"step": 866
|
| 6072 |
+
},
|
| 6073 |
+
{
|
| 6074 |
+
"epoch": 0.8892307692307693,
|
| 6075 |
+
"grad_norm": 3.6567392349243164,
|
| 6076 |
+
"learning_rate": 3.027689123773109e-06,
|
| 6077 |
+
"loss": 0.6477,
|
| 6078 |
+
"step": 867
|
| 6079 |
+
},
|
| 6080 |
+
{
|
| 6081 |
+
"epoch": 0.8902564102564102,
|
| 6082 |
+
"grad_norm": 5.468822956085205,
|
| 6083 |
+
"learning_rate": 2.9724400944372398e-06,
|
| 6084 |
+
"loss": 0.8332,
|
| 6085 |
+
"step": 868
|
| 6086 |
+
},
|
| 6087 |
+
{
|
| 6088 |
+
"epoch": 0.8912820512820513,
|
| 6089 |
+
"grad_norm": 4.6036882400512695,
|
| 6090 |
+
"learning_rate": 2.917684361951728e-06,
|
| 6091 |
+
"loss": 0.8671,
|
| 6092 |
+
"step": 869
|
| 6093 |
+
},
|
| 6094 |
+
{
|
| 6095 |
+
"epoch": 0.8923076923076924,
|
| 6096 |
+
"grad_norm": 3.711745500564575,
|
| 6097 |
+
"learning_rate": 2.8634225006782865e-06,
|
| 6098 |
+
"loss": 0.6129,
|
| 6099 |
+
"step": 870
|
| 6100 |
+
},
|
| 6101 |
+
{
|
| 6102 |
+
"epoch": 0.8933333333333333,
|
| 6103 |
+
"grad_norm": 4.098292827606201,
|
| 6104 |
+
"learning_rate": 2.8096550797981792e-06,
|
| 6105 |
+
"loss": 0.625,
|
| 6106 |
+
"step": 871
|
| 6107 |
+
},
|
| 6108 |
+
{
|
| 6109 |
+
"epoch": 0.8943589743589744,
|
| 6110 |
+
"grad_norm": 4.045825958251953,
|
| 6111 |
+
"learning_rate": 2.7563826633061896e-06,
|
| 6112 |
+
"loss": 0.5502,
|
| 6113 |
+
"step": 872
|
| 6114 |
+
},
|
| 6115 |
+
{
|
| 6116 |
+
"epoch": 0.8953846153846153,
|
| 6117 |
+
"grad_norm": 3.8534200191497803,
|
| 6118 |
+
"learning_rate": 2.703605810004772e-06,
|
| 6119 |
+
"loss": 0.7319,
|
| 6120 |
+
"step": 873
|
| 6121 |
+
},
|
| 6122 |
+
{
|
| 6123 |
+
"epoch": 0.8964102564102564,
|
| 6124 |
+
"grad_norm": 3.8560359477996826,
|
| 6125 |
+
"learning_rate": 2.65132507349814e-06,
|
| 6126 |
+
"loss": 0.6883,
|
| 6127 |
+
"step": 874
|
| 6128 |
+
},
|
| 6129 |
+
{
|
| 6130 |
+
"epoch": 0.8974358974358975,
|
| 6131 |
+
"grad_norm": 5.30885124206543,
|
| 6132 |
+
"learning_rate": 2.5995410021864787e-06,
|
| 6133 |
+
"loss": 0.7475,
|
| 6134 |
+
"step": 875
|
| 6135 |
+
},
|
| 6136 |
+
{
|
| 6137 |
+
"epoch": 0.8984615384615384,
|
| 6138 |
+
"grad_norm": 5.563632488250732,
|
| 6139 |
+
"learning_rate": 2.548254139260192e-06,
|
| 6140 |
+
"loss": 1.1936,
|
| 6141 |
+
"step": 876
|
| 6142 |
+
},
|
| 6143 |
+
{
|
| 6144 |
+
"epoch": 0.8994871794871795,
|
| 6145 |
+
"grad_norm": 4.295792102813721,
|
| 6146 |
+
"learning_rate": 2.4974650226942064e-06,
|
| 6147 |
+
"loss": 0.7475,
|
| 6148 |
+
"step": 877
|
| 6149 |
+
},
|
| 6150 |
+
{
|
| 6151 |
+
"epoch": 0.9005128205128206,
|
| 6152 |
+
"grad_norm": 7.171024322509766,
|
| 6153 |
+
"learning_rate": 2.4471741852423237e-06,
|
| 6154 |
+
"loss": 1.4032,
|
| 6155 |
+
"step": 878
|
| 6156 |
+
},
|
| 6157 |
+
{
|
| 6158 |
+
"epoch": 0.9015384615384615,
|
| 6159 |
+
"grad_norm": 5.549228191375732,
|
| 6160 |
+
"learning_rate": 2.397382154431621e-06,
|
| 6161 |
+
"loss": 1.0653,
|
| 6162 |
+
"step": 879
|
| 6163 |
+
},
|
| 6164 |
+
{
|
| 6165 |
+
"epoch": 0.9025641025641026,
|
| 6166 |
+
"grad_norm": 4.5205979347229,
|
| 6167 |
+
"learning_rate": 2.3480894525569562e-06,
|
| 6168 |
+
"loss": 0.7206,
|
| 6169 |
+
"step": 880
|
| 6170 |
+
},
|
| 6171 |
+
{
|
| 6172 |
+
"epoch": 0.9035897435897436,
|
| 6173 |
+
"grad_norm": 5.302089214324951,
|
| 6174 |
+
"learning_rate": 2.2992965966754377e-06,
|
| 6175 |
+
"loss": 0.8298,
|
| 6176 |
+
"step": 881
|
| 6177 |
+
},
|
| 6178 |
+
{
|
| 6179 |
+
"epoch": 0.9046153846153846,
|
| 6180 |
+
"grad_norm": 4.690470218658447,
|
| 6181 |
+
"learning_rate": 2.251004098601034e-06,
|
| 6182 |
+
"loss": 0.653,
|
| 6183 |
+
"step": 882
|
| 6184 |
+
},
|
| 6185 |
+
{
|
| 6186 |
+
"epoch": 0.9056410256410257,
|
| 6187 |
+
"grad_norm": 4.091747760772705,
|
| 6188 |
+
"learning_rate": 2.2032124648992015e-06,
|
| 6189 |
+
"loss": 0.5173,
|
| 6190 |
+
"step": 883
|
| 6191 |
+
},
|
| 6192 |
+
{
|
| 6193 |
+
"epoch": 0.9066666666666666,
|
| 6194 |
+
"grad_norm": 4.637690544128418,
|
| 6195 |
+
"learning_rate": 2.1559221968815546e-06,
|
| 6196 |
+
"loss": 0.6316,
|
| 6197 |
+
"step": 884
|
| 6198 |
+
},
|
| 6199 |
+
{
|
| 6200 |
+
"epoch": 0.9076923076923077,
|
| 6201 |
+
"grad_norm": 4.933149814605713,
|
| 6202 |
+
"learning_rate": 2.1091337906006482e-06,
|
| 6203 |
+
"loss": 0.991,
|
| 6204 |
+
"step": 885
|
| 6205 |
+
},
|
| 6206 |
+
{
|
| 6207 |
+
"epoch": 0.9087179487179488,
|
| 6208 |
+
"grad_norm": 5.4409027099609375,
|
| 6209 |
+
"learning_rate": 2.062847736844703e-06,
|
| 6210 |
+
"loss": 0.9278,
|
| 6211 |
+
"step": 886
|
| 6212 |
+
},
|
| 6213 |
+
{
|
| 6214 |
+
"epoch": 0.9097435897435897,
|
| 6215 |
+
"grad_norm": 5.3039398193359375,
|
| 6216 |
+
"learning_rate": 2.017064521132533e-06,
|
| 6217 |
+
"loss": 0.7831,
|
| 6218 |
+
"step": 887
|
| 6219 |
+
},
|
| 6220 |
+
{
|
| 6221 |
+
"epoch": 0.9107692307692308,
|
| 6222 |
+
"grad_norm": 5.107433319091797,
|
| 6223 |
+
"learning_rate": 1.9717846237084005e-06,
|
| 6224 |
+
"loss": 0.8805,
|
| 6225 |
+
"step": 888
|
| 6226 |
+
},
|
| 6227 |
+
{
|
| 6228 |
+
"epoch": 0.9117948717948718,
|
| 6229 |
+
"grad_norm": 5.635267734527588,
|
| 6230 |
+
"learning_rate": 1.9270085195370045e-06,
|
| 6231 |
+
"loss": 0.8542,
|
| 6232 |
+
"step": 889
|
| 6233 |
+
},
|
| 6234 |
+
{
|
| 6235 |
+
"epoch": 0.9128205128205128,
|
| 6236 |
+
"grad_norm": 6.010116100311279,
|
| 6237 |
+
"learning_rate": 1.8827366782984913e-06,
|
| 6238 |
+
"loss": 1.1149,
|
| 6239 |
+
"step": 890
|
| 6240 |
+
},
|
| 6241 |
+
{
|
| 6242 |
+
"epoch": 0.9138461538461539,
|
| 6243 |
+
"grad_norm": 4.707157135009766,
|
| 6244 |
+
"learning_rate": 1.8389695643835247e-06,
|
| 6245 |
+
"loss": 0.6094,
|
| 6246 |
+
"step": 891
|
| 6247 |
+
},
|
| 6248 |
+
{
|
| 6249 |
+
"epoch": 0.9148717948717948,
|
| 6250 |
+
"grad_norm": 5.779000759124756,
|
| 6251 |
+
"learning_rate": 1.7957076368884274e-06,
|
| 6252 |
+
"loss": 0.9566,
|
| 6253 |
+
"step": 892
|
| 6254 |
+
},
|
| 6255 |
+
{
|
| 6256 |
+
"epoch": 0.9158974358974359,
|
| 6257 |
+
"grad_norm": 5.43092679977417,
|
| 6258 |
+
"learning_rate": 1.7529513496103323e-06,
|
| 6259 |
+
"loss": 0.7754,
|
| 6260 |
+
"step": 893
|
| 6261 |
+
},
|
| 6262 |
+
{
|
| 6263 |
+
"epoch": 0.916923076923077,
|
| 6264 |
+
"grad_norm": 6.918627738952637,
|
| 6265 |
+
"learning_rate": 1.7107011510424764e-06,
|
| 6266 |
+
"loss": 1.151,
|
| 6267 |
+
"step": 894
|
| 6268 |
+
},
|
| 6269 |
+
{
|
| 6270 |
+
"epoch": 0.9179487179487179,
|
| 6271 |
+
"grad_norm": 6.21636962890625,
|
| 6272 |
+
"learning_rate": 1.6689574843694433e-06,
|
| 6273 |
+
"loss": 0.7324,
|
| 6274 |
+
"step": 895
|
| 6275 |
+
},
|
| 6276 |
+
{
|
| 6277 |
+
"epoch": 0.918974358974359,
|
| 6278 |
+
"grad_norm": 8.857195854187012,
|
| 6279 |
+
"learning_rate": 1.6277207874625444e-06,
|
| 6280 |
+
"loss": 0.9653,
|
| 6281 |
+
"step": 896
|
| 6282 |
+
},
|
| 6283 |
+
{
|
| 6284 |
+
"epoch": 0.92,
|
| 6285 |
+
"grad_norm": 7.489193916320801,
|
| 6286 |
+
"learning_rate": 1.5869914928752117e-06,
|
| 6287 |
+
"loss": 0.9201,
|
| 6288 |
+
"step": 897
|
| 6289 |
+
},
|
| 6290 |
+
{
|
| 6291 |
+
"epoch": 0.921025641025641,
|
| 6292 |
+
"grad_norm": 6.698476791381836,
|
| 6293 |
+
"learning_rate": 1.546770027838479e-06,
|
| 6294 |
+
"loss": 1.0077,
|
| 6295 |
+
"step": 898
|
| 6296 |
+
},
|
| 6297 |
+
{
|
| 6298 |
+
"epoch": 0.9220512820512821,
|
| 6299 |
+
"grad_norm": 9.213371276855469,
|
| 6300 |
+
"learning_rate": 1.507056814256491e-06,
|
| 6301 |
+
"loss": 1.2657,
|
| 6302 |
+
"step": 899
|
| 6303 |
+
},
|
| 6304 |
+
{
|
| 6305 |
+
"epoch": 0.9230769230769231,
|
| 6306 |
+
"grad_norm": 15.109607696533203,
|
| 6307 |
+
"learning_rate": 1.4678522687020413e-06,
|
| 6308 |
+
"loss": 2.0015,
|
| 6309 |
+
"step": 900
|
| 6310 |
+
},
|
| 6311 |
+
{
|
| 6312 |
+
"epoch": 0.9241025641025641,
|
| 6313 |
+
"grad_norm": 3.072422981262207,
|
| 6314 |
+
"learning_rate": 1.4291568024122848e-06,
|
| 6315 |
+
"loss": 1.0559,
|
| 6316 |
+
"step": 901
|
| 6317 |
+
},
|
| 6318 |
+
{
|
| 6319 |
+
"epoch": 0.9251282051282051,
|
| 6320 |
+
"grad_norm": 2.742314577102661,
|
| 6321 |
+
"learning_rate": 1.390970821284343e-06,
|
| 6322 |
+
"loss": 0.7734,
|
| 6323 |
+
"step": 902
|
| 6324 |
+
},
|
| 6325 |
+
{
|
| 6326 |
+
"epoch": 0.9261538461538461,
|
| 6327 |
+
"grad_norm": 3.3717610836029053,
|
| 6328 |
+
"learning_rate": 1.3532947258710904e-06,
|
| 6329 |
+
"loss": 0.9725,
|
| 6330 |
+
"step": 903
|
| 6331 |
+
},
|
| 6332 |
+
{
|
| 6333 |
+
"epoch": 0.9271794871794872,
|
| 6334 |
+
"grad_norm": 3.4587204456329346,
|
| 6335 |
+
"learning_rate": 1.3161289113769404e-06,
|
| 6336 |
+
"loss": 0.827,
|
| 6337 |
+
"step": 904
|
| 6338 |
+
},
|
| 6339 |
+
{
|
| 6340 |
+
"epoch": 0.9282051282051282,
|
| 6341 |
+
"grad_norm": 6.684597969055176,
|
| 6342 |
+
"learning_rate": 1.2794737676536994e-06,
|
| 6343 |
+
"loss": 0.9701,
|
| 6344 |
+
"step": 905
|
| 6345 |
+
},
|
| 6346 |
+
{
|
| 6347 |
+
"epoch": 0.9292307692307692,
|
| 6348 |
+
"grad_norm": 3.692589521408081,
|
| 6349 |
+
"learning_rate": 1.2433296791964754e-06,
|
| 6350 |
+
"loss": 0.9335,
|
| 6351 |
+
"step": 906
|
| 6352 |
+
},
|
| 6353 |
+
{
|
| 6354 |
+
"epoch": 0.9302564102564103,
|
| 6355 |
+
"grad_norm": 4.69821834564209,
|
| 6356 |
+
"learning_rate": 1.207697025139659e-06,
|
| 6357 |
+
"loss": 0.9612,
|
| 6358 |
+
"step": 907
|
| 6359 |
+
},
|
| 6360 |
+
{
|
| 6361 |
+
"epoch": 0.9312820512820513,
|
| 6362 |
+
"grad_norm": 3.526686906814575,
|
| 6363 |
+
"learning_rate": 1.1725761792529377e-06,
|
| 6364 |
+
"loss": 0.6213,
|
| 6365 |
+
"step": 908
|
| 6366 |
+
},
|
| 6367 |
+
{
|
| 6368 |
+
"epoch": 0.9323076923076923,
|
| 6369 |
+
"grad_norm": 4.1468939781188965,
|
| 6370 |
+
"learning_rate": 1.1379675099373488e-06,
|
| 6371 |
+
"loss": 0.8209,
|
| 6372 |
+
"step": 909
|
| 6373 |
+
},
|
| 6374 |
+
{
|
| 6375 |
+
"epoch": 0.9333333333333333,
|
| 6376 |
+
"grad_norm": 3.930953025817871,
|
| 6377 |
+
"learning_rate": 1.1038713802214717e-06,
|
| 6378 |
+
"loss": 0.9604,
|
| 6379 |
+
"step": 910
|
| 6380 |
+
},
|
| 6381 |
+
{
|
| 6382 |
+
"epoch": 0.9343589743589743,
|
| 6383 |
+
"grad_norm": 3.896559000015259,
|
| 6384 |
+
"learning_rate": 1.0702881477575588e-06,
|
| 6385 |
+
"loss": 0.8926,
|
| 6386 |
+
"step": 911
|
| 6387 |
+
},
|
| 6388 |
+
{
|
| 6389 |
+
"epoch": 0.9353846153846154,
|
| 6390 |
+
"grad_norm": 3.752502679824829,
|
| 6391 |
+
"learning_rate": 1.0372181648178436e-06,
|
| 6392 |
+
"loss": 0.9382,
|
| 6393 |
+
"step": 912
|
| 6394 |
+
},
|
| 6395 |
+
{
|
| 6396 |
+
"epoch": 0.9364102564102564,
|
| 6397 |
+
"grad_norm": 3.334625244140625,
|
| 6398 |
+
"learning_rate": 1.004661778290783e-06,
|
| 6399 |
+
"loss": 0.6633,
|
| 6400 |
+
"step": 913
|
| 6401 |
+
},
|
| 6402 |
+
{
|
| 6403 |
+
"epoch": 0.9374358974358974,
|
| 6404 |
+
"grad_norm": 3.897460460662842,
|
| 6405 |
+
"learning_rate": 9.726193296774766e-07,
|
| 6406 |
+
"loss": 0.7738,
|
| 6407 |
+
"step": 914
|
| 6408 |
+
},
|
| 6409 |
+
{
|
| 6410 |
+
"epoch": 0.9384615384615385,
|
| 6411 |
+
"grad_norm": 4.598691940307617,
|
| 6412 |
+
"learning_rate": 9.410911550880475e-07,
|
| 6413 |
+
"loss": 0.8331,
|
| 6414 |
+
"step": 915
|
| 6415 |
+
},
|
| 6416 |
+
{
|
| 6417 |
+
"epoch": 0.9394871794871795,
|
| 6418 |
+
"grad_norm": 4.020224571228027,
|
| 6419 |
+
"learning_rate": 9.100775852381226e-07,
|
| 6420 |
+
"loss": 0.9878,
|
| 6421 |
+
"step": 916
|
| 6422 |
+
},
|
| 6423 |
+
{
|
| 6424 |
+
"epoch": 0.9405128205128205,
|
| 6425 |
+
"grad_norm": 4.421807289123535,
|
| 6426 |
+
"learning_rate": 8.795789454453862e-07,
|
| 6427 |
+
"loss": 1.1013,
|
| 6428 |
+
"step": 917
|
| 6429 |
+
},
|
| 6430 |
+
{
|
| 6431 |
+
"epoch": 0.9415384615384615,
|
| 6432 |
+
"grad_norm": 4.398540019989014,
|
| 6433 |
+
"learning_rate": 8.495955556261204e-07,
|
| 6434 |
+
"loss": 0.8937,
|
| 6435 |
+
"step": 918
|
| 6436 |
+
},
|
| 6437 |
+
{
|
| 6438 |
+
"epoch": 0.9425641025641026,
|
| 6439 |
+
"grad_norm": 6.260739803314209,
|
| 6440 |
+
"learning_rate": 8.201277302919086e-07,
|
| 6441 |
+
"loss": 1.1365,
|
| 6442 |
+
"step": 919
|
| 6443 |
+
},
|
| 6444 |
+
{
|
| 6445 |
+
"epoch": 0.9435897435897436,
|
| 6446 |
+
"grad_norm": 3.387582778930664,
|
| 6447 |
+
"learning_rate": 7.911757785462881e-07,
|
| 6448 |
+
"loss": 0.5612,
|
| 6449 |
+
"step": 920
|
| 6450 |
+
},
|
| 6451 |
+
{
|
| 6452 |
+
"epoch": 0.9446153846153846,
|
| 6453 |
+
"grad_norm": 4.812023162841797,
|
| 6454 |
+
"learning_rate": 7.627400040815413e-07,
|
| 6455 |
+
"loss": 1.0781,
|
| 6456 |
+
"step": 921
|
| 6457 |
+
},
|
| 6458 |
+
{
|
| 6459 |
+
"epoch": 0.9456410256410256,
|
| 6460 |
+
"grad_norm": 4.107918739318848,
|
| 6461 |
+
"learning_rate": 7.348207051754818e-07,
|
| 6462 |
+
"loss": 0.7609,
|
| 6463 |
+
"step": 922
|
| 6464 |
+
},
|
| 6465 |
+
{
|
| 6466 |
+
"epoch": 0.9466666666666667,
|
| 6467 |
+
"grad_norm": 4.248846530914307,
|
| 6468 |
+
"learning_rate": 7.074181746883401e-07,
|
| 6469 |
+
"loss": 0.7369,
|
| 6470 |
+
"step": 923
|
| 6471 |
+
},
|
| 6472 |
+
{
|
| 6473 |
+
"epoch": 0.9476923076923077,
|
| 6474 |
+
"grad_norm": 4.63174295425415,
|
| 6475 |
+
"learning_rate": 6.805327000596995e-07,
|
| 6476 |
+
"loss": 0.6237,
|
| 6477 |
+
"step": 924
|
| 6478 |
+
},
|
| 6479 |
+
{
|
| 6480 |
+
"epoch": 0.9487179487179487,
|
| 6481 |
+
"grad_norm": 3.942258596420288,
|
| 6482 |
+
"learning_rate": 6.54164563305465e-07,
|
| 6483 |
+
"loss": 0.7415,
|
| 6484 |
+
"step": 925
|
| 6485 |
+
},
|
| 6486 |
+
{
|
| 6487 |
+
"epoch": 0.9497435897435897,
|
| 6488 |
+
"grad_norm": 5.168544292449951,
|
| 6489 |
+
"learning_rate": 6.283140410149213e-07,
|
| 6490 |
+
"loss": 1.0296,
|
| 6491 |
+
"step": 926
|
| 6492 |
+
},
|
| 6493 |
+
{
|
| 6494 |
+
"epoch": 0.9507692307692308,
|
| 6495 |
+
"grad_norm": 4.351825714111328,
|
| 6496 |
+
"learning_rate": 6.029814043478022e-07,
|
| 6497 |
+
"loss": 0.9396,
|
| 6498 |
+
"step": 927
|
| 6499 |
+
},
|
| 6500 |
+
{
|
| 6501 |
+
"epoch": 0.9517948717948718,
|
| 6502 |
+
"grad_norm": 5.255843162536621,
|
| 6503 |
+
"learning_rate": 5.781669190314809e-07,
|
| 6504 |
+
"loss": 0.6457,
|
| 6505 |
+
"step": 928
|
| 6506 |
+
},
|
| 6507 |
+
{
|
| 6508 |
+
"epoch": 0.9528205128205128,
|
| 6509 |
+
"grad_norm": 5.010144233703613,
|
| 6510 |
+
"learning_rate": 5.538708453581787e-07,
|
| 6511 |
+
"loss": 0.8027,
|
| 6512 |
+
"step": 929
|
| 6513 |
+
},
|
| 6514 |
+
{
|
| 6515 |
+
"epoch": 0.9538461538461539,
|
| 6516 |
+
"grad_norm": 5.048897743225098,
|
| 6517 |
+
"learning_rate": 5.300934381821998e-07,
|
| 6518 |
+
"loss": 0.7858,
|
| 6519 |
+
"step": 930
|
| 6520 |
+
},
|
| 6521 |
+
{
|
| 6522 |
+
"epoch": 0.9548717948717949,
|
| 6523 |
+
"grad_norm": 5.082890510559082,
|
| 6524 |
+
"learning_rate": 5.068349469173006e-07,
|
| 6525 |
+
"loss": 0.8434,
|
| 6526 |
+
"step": 931
|
| 6527 |
+
},
|
| 6528 |
+
{
|
| 6529 |
+
"epoch": 0.9558974358974359,
|
| 6530 |
+
"grad_norm": 4.683363437652588,
|
| 6531 |
+
"learning_rate": 4.840956155340415e-07,
|
| 6532 |
+
"loss": 0.846,
|
| 6533 |
+
"step": 932
|
| 6534 |
+
},
|
| 6535 |
+
{
|
| 6536 |
+
"epoch": 0.9569230769230769,
|
| 6537 |
+
"grad_norm": 4.555981636047363,
|
| 6538 |
+
"learning_rate": 4.6187568255726124e-07,
|
| 6539 |
+
"loss": 0.7578,
|
| 6540 |
+
"step": 933
|
| 6541 |
+
},
|
| 6542 |
+
{
|
| 6543 |
+
"epoch": 0.9579487179487179,
|
| 6544 |
+
"grad_norm": 4.5115814208984375,
|
| 6545 |
+
"learning_rate": 4.4017538106352894e-07,
|
| 6546 |
+
"loss": 0.654,
|
| 6547 |
+
"step": 934
|
| 6548 |
+
},
|
| 6549 |
+
{
|
| 6550 |
+
"epoch": 0.958974358974359,
|
| 6551 |
+
"grad_norm": 5.074052810668945,
|
| 6552 |
+
"learning_rate": 4.189949386787462e-07,
|
| 6553 |
+
"loss": 1.1065,
|
| 6554 |
+
"step": 935
|
| 6555 |
+
},
|
| 6556 |
+
{
|
| 6557 |
+
"epoch": 0.96,
|
| 6558 |
+
"grad_norm": 5.363509654998779,
|
| 6559 |
+
"learning_rate": 3.9833457757572636e-07,
|
| 6560 |
+
"loss": 0.7843,
|
| 6561 |
+
"step": 936
|
| 6562 |
+
},
|
| 6563 |
+
{
|
| 6564 |
+
"epoch": 0.961025641025641,
|
| 6565 |
+
"grad_norm": 6.79396915435791,
|
| 6566 |
+
"learning_rate": 3.7819451447189126e-07,
|
| 6567 |
+
"loss": 0.7907,
|
| 6568 |
+
"step": 937
|
| 6569 |
+
},
|
| 6570 |
+
{
|
| 6571 |
+
"epoch": 0.9620512820512821,
|
| 6572 |
+
"grad_norm": 4.234402656555176,
|
| 6573 |
+
"learning_rate": 3.5857496062695617e-07,
|
| 6574 |
+
"loss": 0.621,
|
| 6575 |
+
"step": 938
|
| 6576 |
+
},
|
| 6577 |
+
{
|
| 6578 |
+
"epoch": 0.963076923076923,
|
| 6579 |
+
"grad_norm": 4.6178131103515625,
|
| 6580 |
+
"learning_rate": 3.394761218407705e-07,
|
| 6581 |
+
"loss": 0.716,
|
| 6582 |
+
"step": 939
|
| 6583 |
+
},
|
| 6584 |
+
{
|
| 6585 |
+
"epoch": 0.9641025641025641,
|
| 6586 |
+
"grad_norm": 5.502908706665039,
|
| 6587 |
+
"learning_rate": 3.208981984511195e-07,
|
| 6588 |
+
"loss": 1.1312,
|
| 6589 |
+
"step": 940
|
| 6590 |
+
},
|
| 6591 |
+
{
|
| 6592 |
+
"epoch": 0.9651282051282051,
|
| 6593 |
+
"grad_norm": 4.802907943725586,
|
| 6594 |
+
"learning_rate": 3.0284138533160924e-07,
|
| 6595 |
+
"loss": 0.6067,
|
| 6596 |
+
"step": 941
|
| 6597 |
+
},
|
| 6598 |
+
{
|
| 6599 |
+
"epoch": 0.9661538461538461,
|
| 6600 |
+
"grad_norm": 4.86731481552124,
|
| 6601 |
+
"learning_rate": 2.8530587188968504e-07,
|
| 6602 |
+
"loss": 0.4606,
|
| 6603 |
+
"step": 942
|
| 6604 |
+
},
|
| 6605 |
+
{
|
| 6606 |
+
"epoch": 0.9671794871794872,
|
| 6607 |
+
"grad_norm": 6.775837421417236,
|
| 6608 |
+
"learning_rate": 2.682918420645719e-07,
|
| 6609 |
+
"loss": 1.2505,
|
| 6610 |
+
"step": 943
|
| 6611 |
+
},
|
| 6612 |
+
{
|
| 6613 |
+
"epoch": 0.9682051282051282,
|
| 6614 |
+
"grad_norm": 5.894476413726807,
|
| 6615 |
+
"learning_rate": 2.5179947432540376e-07,
|
| 6616 |
+
"loss": 0.7912,
|
| 6617 |
+
"step": 944
|
| 6618 |
+
},
|
| 6619 |
+
{
|
| 6620 |
+
"epoch": 0.9692307692307692,
|
| 6621 |
+
"grad_norm": 5.020647048950195,
|
| 6622 |
+
"learning_rate": 2.3582894166930268e-07,
|
| 6623 |
+
"loss": 0.6911,
|
| 6624 |
+
"step": 945
|
| 6625 |
+
},
|
| 6626 |
+
{
|
| 6627 |
+
"epoch": 0.9702564102564103,
|
| 6628 |
+
"grad_norm": 6.510494232177734,
|
| 6629 |
+
"learning_rate": 2.2038041161960288e-07,
|
| 6630 |
+
"loss": 0.8434,
|
| 6631 |
+
"step": 946
|
| 6632 |
+
},
|
| 6633 |
+
{
|
| 6634 |
+
"epoch": 0.9712820512820513,
|
| 6635 |
+
"grad_norm": 8.360902786254883,
|
| 6636 |
+
"learning_rate": 2.0545404622407393e-07,
|
| 6637 |
+
"loss": 1.4077,
|
| 6638 |
+
"step": 947
|
| 6639 |
+
},
|
| 6640 |
+
{
|
| 6641 |
+
"epoch": 0.9723076923076923,
|
| 6642 |
+
"grad_norm": 8.382393836975098,
|
| 6643 |
+
"learning_rate": 1.91050002053228e-07,
|
| 6644 |
+
"loss": 1.0157,
|
| 6645 |
+
"step": 948
|
| 6646 |
+
},
|
| 6647 |
+
{
|
| 6648 |
+
"epoch": 0.9733333333333334,
|
| 6649 |
+
"grad_norm": 10.754633903503418,
|
| 6650 |
+
"learning_rate": 1.7716843019867646e-07,
|
| 6651 |
+
"loss": 1.6177,
|
| 6652 |
+
"step": 949
|
| 6653 |
+
},
|
| 6654 |
+
{
|
| 6655 |
+
"epoch": 0.9743589743589743,
|
| 6656 |
+
"grad_norm": 16.695589065551758,
|
| 6657 |
+
"learning_rate": 1.6380947627153143e-07,
|
| 6658 |
+
"loss": 2.4161,
|
| 6659 |
+
"step": 950
|
| 6660 |
+
},
|
| 6661 |
+
{
|
| 6662 |
+
"epoch": 0.9753846153846154,
|
| 6663 |
+
"grad_norm": 3.2838544845581055,
|
| 6664 |
+
"learning_rate": 1.509732804009012e-07,
|
| 6665 |
+
"loss": 1.1099,
|
| 6666 |
+
"step": 951
|
| 6667 |
+
},
|
| 6668 |
+
{
|
| 6669 |
+
"epoch": 0.9764102564102564,
|
| 6670 |
+
"grad_norm": 2.8751566410064697,
|
| 6671 |
+
"learning_rate": 1.386599772324082e-07,
|
| 6672 |
+
"loss": 0.8006,
|
| 6673 |
+
"step": 952
|
| 6674 |
+
},
|
| 6675 |
+
{
|
| 6676 |
+
"epoch": 0.9774358974358974,
|
| 6677 |
+
"grad_norm": 3.975473403930664,
|
| 6678 |
+
"learning_rate": 1.268696959267679e-07,
|
| 6679 |
+
"loss": 0.9313,
|
| 6680 |
+
"step": 953
|
| 6681 |
+
},
|
| 6682 |
+
{
|
| 6683 |
+
"epoch": 0.9784615384615385,
|
| 6684 |
+
"grad_norm": 4.9324235916137695,
|
| 6685 |
+
"learning_rate": 1.156025601584676e-07,
|
| 6686 |
+
"loss": 0.8868,
|
| 6687 |
+
"step": 954
|
| 6688 |
+
},
|
| 6689 |
+
{
|
| 6690 |
+
"epoch": 0.9794871794871794,
|
| 6691 |
+
"grad_norm": 3.668186664581299,
|
| 6692 |
+
"learning_rate": 1.0485868811441757e-07,
|
| 6693 |
+
"loss": 0.819,
|
| 6694 |
+
"step": 955
|
| 6695 |
+
},
|
| 6696 |
+
{
|
| 6697 |
+
"epoch": 0.9805128205128205,
|
| 6698 |
+
"grad_norm": 3.585177183151245,
|
| 6699 |
+
"learning_rate": 9.463819249275751e-08,
|
| 6700 |
+
"loss": 0.67,
|
| 6701 |
+
"step": 956
|
| 6702 |
+
},
|
| 6703 |
+
{
|
| 6704 |
+
"epoch": 0.9815384615384616,
|
| 6705 |
+
"grad_norm": 5.605104446411133,
|
| 6706 |
+
"learning_rate": 8.494118050164646e-08,
|
| 6707 |
+
"loss": 1.3032,
|
| 6708 |
+
"step": 957
|
| 6709 |
+
},
|
| 6710 |
+
{
|
| 6711 |
+
"epoch": 0.9825641025641025,
|
| 6712 |
+
"grad_norm": 5.455284118652344,
|
| 6713 |
+
"learning_rate": 7.576775385815249e-08,
|
| 6714 |
+
"loss": 0.7616,
|
| 6715 |
+
"step": 958
|
| 6716 |
+
},
|
| 6717 |
+
{
|
| 6718 |
+
"epoch": 0.9835897435897436,
|
| 6719 |
+
"grad_norm": 4.1384711265563965,
|
| 6720 |
+
"learning_rate": 6.711800878718144e-08,
|
| 6721 |
+
"loss": 0.8938,
|
| 6722 |
+
"step": 959
|
| 6723 |
+
},
|
| 6724 |
+
{
|
| 6725 |
+
"epoch": 0.9846153846153847,
|
| 6726 |
+
"grad_norm": 3.7249691486358643,
|
| 6727 |
+
"learning_rate": 5.899203602046655e-08,
|
| 6728 |
+
"loss": 0.6647,
|
| 6729 |
+
"step": 960
|
| 6730 |
+
},
|
| 6731 |
+
{
|
| 6732 |
+
"epoch": 0.9856410256410256,
|
| 6733 |
+
"grad_norm": 5.5328474044799805,
|
| 6734 |
+
"learning_rate": 5.1389920795613667e-08,
|
| 6735 |
+
"loss": 1.0767,
|
| 6736 |
+
"step": 961
|
| 6737 |
+
},
|
| 6738 |
+
{
|
| 6739 |
+
"epoch": 0.9866666666666667,
|
| 6740 |
+
"grad_norm": 4.143031120300293,
|
| 6741 |
+
"learning_rate": 4.4311742855218664e-08,
|
| 6742 |
+
"loss": 0.7569,
|
| 6743 |
+
"step": 962
|
| 6744 |
+
},
|
| 6745 |
+
{
|
| 6746 |
+
"epoch": 0.9876923076923076,
|
| 6747 |
+
"grad_norm": 5.039745807647705,
|
| 6748 |
+
"learning_rate": 3.775757644601807e-08,
|
| 6749 |
+
"loss": 1.0621,
|
| 6750 |
+
"step": 963
|
| 6751 |
+
},
|
| 6752 |
+
{
|
| 6753 |
+
"epoch": 0.9887179487179487,
|
| 6754 |
+
"grad_norm": 4.493295192718506,
|
| 6755 |
+
"learning_rate": 3.172749031811195e-08,
|
| 6756 |
+
"loss": 1.0456,
|
| 6757 |
+
"step": 964
|
| 6758 |
+
},
|
| 6759 |
+
{
|
| 6760 |
+
"epoch": 0.9897435897435898,
|
| 6761 |
+
"grad_norm": 3.9239814281463623,
|
| 6762 |
+
"learning_rate": 2.6221547724253337e-08,
|
| 6763 |
+
"loss": 0.6195,
|
| 6764 |
+
"step": 965
|
| 6765 |
+
},
|
| 6766 |
+
{
|
| 6767 |
+
"epoch": 0.9907692307692307,
|
| 6768 |
+
"grad_norm": 6.028420448303223,
|
| 6769 |
+
"learning_rate": 2.1239806419176556e-08,
|
| 6770 |
+
"loss": 0.6658,
|
| 6771 |
+
"step": 966
|
| 6772 |
+
},
|
| 6773 |
+
{
|
| 6774 |
+
"epoch": 0.9917948717948718,
|
| 6775 |
+
"grad_norm": 4.432763576507568,
|
| 6776 |
+
"learning_rate": 1.678231865899216e-08,
|
| 6777 |
+
"loss": 0.7334,
|
| 6778 |
+
"step": 967
|
| 6779 |
+
},
|
| 6780 |
+
{
|
| 6781 |
+
"epoch": 0.9928205128205129,
|
| 6782 |
+
"grad_norm": 3.8991336822509766,
|
| 6783 |
+
"learning_rate": 1.2849131200631802e-08,
|
| 6784 |
+
"loss": 0.7016,
|
| 6785 |
+
"step": 968
|
| 6786 |
+
},
|
| 6787 |
+
{
|
| 6788 |
+
"epoch": 0.9938461538461538,
|
| 6789 |
+
"grad_norm": 6.39743185043335,
|
| 6790 |
+
"learning_rate": 9.440285301370865e-09,
|
| 6791 |
+
"loss": 1.026,
|
| 6792 |
+
"step": 969
|
| 6793 |
+
},
|
| 6794 |
+
{
|
| 6795 |
+
"epoch": 0.9948717948717949,
|
| 6796 |
+
"grad_norm": 5.205996513366699,
|
| 6797 |
+
"learning_rate": 6.5558167183898955e-09,
|
| 6798 |
+
"loss": 1.2714,
|
| 6799 |
+
"step": 970
|
| 6800 |
+
},
|
| 6801 |
+
{
|
| 6802 |
+
"epoch": 0.9958974358974358,
|
| 6803 |
+
"grad_norm": 5.89625358581543,
|
| 6804 |
+
"learning_rate": 4.1957557084082445e-09,
|
| 6805 |
+
"loss": 0.9163,
|
| 6806 |
+
"step": 971
|
| 6807 |
+
},
|
| 6808 |
+
{
|
| 6809 |
+
"epoch": 0.9969230769230769,
|
| 6810 |
+
"grad_norm": 6.269636631011963,
|
| 6811 |
+
"learning_rate": 2.360127027339898e-09,
|
| 6812 |
+
"loss": 1.2324,
|
| 6813 |
+
"step": 972
|
| 6814 |
+
},
|
| 6815 |
+
{
|
| 6816 |
+
"epoch": 0.997948717948718,
|
| 6817 |
+
"grad_norm": 5.002432823181152,
|
| 6818 |
+
"learning_rate": 1.0489499300603278e-09,
|
| 6819 |
+
"loss": 0.8293,
|
| 6820 |
+
"step": 973
|
| 6821 |
+
},
|
| 6822 |
+
{
|
| 6823 |
+
"epoch": 0.9989743589743589,
|
| 6824 |
+
"grad_norm": 5.600581169128418,
|
| 6825 |
+
"learning_rate": 2.622381702066523e-10,
|
| 6826 |
+
"loss": 0.6417,
|
| 6827 |
+
"step": 974
|
| 6828 |
+
},
|
| 6829 |
+
{
|
| 6830 |
+
"epoch": 1.0,
|
| 6831 |
+
"grad_norm": 10.595677375793457,
|
| 6832 |
+
"learning_rate": 0.0,
|
| 6833 |
+
"loss": 1.405,
|
| 6834 |
+
"step": 975
|
| 6835 |
}
|
| 6836 |
],
|
| 6837 |
"logging_steps": 1,
|
|
|
|
| 6846 |
"should_evaluate": false,
|
| 6847 |
"should_log": false,
|
| 6848 |
"should_save": true,
|
| 6849 |
+
"should_training_stop": true
|
| 6850 |
},
|
| 6851 |
"attributes": {}
|
| 6852 |
}
|
| 6853 |
},
|
| 6854 |
+
"total_flos": 3.4722657929527296e+16,
|
| 6855 |
"train_batch_size": 4,
|
| 6856 |
"trial_name": null,
|
| 6857 |
"trial_params": null
|