Training in progress, step 460, 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 sha256:192318a76dbdabc5493ecc666ef558e5cba0c805846db4e5de70a9d8fb1661f5
|
| 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
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1854159ce72c01c49c84990cca6c5a4c9293a5b32835fcf2a7b719770ed5797b
|
| 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 sha256:70db357d365c307295f0f171f822ed92fa64ae5a966b05696ee6a22bf77bdd7c
|
| 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:e0a1b967b0e0452d968af4c52dc7534380ad7a06675bc9b554ab04875453a50f
|
| 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": 230,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1625,6 +1625,1624 @@
|
|
| 1625 |
"eval_samples_per_second": 15.85,
|
| 1626 |
"eval_steps_per_second": 7.945,
|
| 1627 |
"step": 230
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1628 |
}
|
| 1629 |
],
|
| 1630 |
"logging_steps": 1,
|
|
@@ -1644,7 +3262,7 @@
|
|
| 1644 |
"attributes": {}
|
| 1645 |
}
|
| 1646 |
},
|
| 1647 |
-
"total_flos":
|
| 1648 |
"train_batch_size": 2,
|
| 1649 |
"trial_name": null,
|
| 1650 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.500544069640914,
|
| 5 |
"eval_steps": 230,
|
| 6 |
+
"global_step": 460,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1625 |
"eval_samples_per_second": 15.85,
|
| 1626 |
"eval_steps_per_second": 7.945,
|
| 1627 |
"step": 230
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 0.2513601741022851,
|
| 1631 |
+
"grad_norm": 0.14655253291130066,
|
| 1632 |
+
"learning_rate": 0.00017222145741734626,
|
| 1633 |
+
"loss": 0.0879,
|
| 1634 |
+
"step": 231
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.25244831338411317,
|
| 1638 |
+
"grad_norm": 0.17367680370807648,
|
| 1639 |
+
"learning_rate": 0.00017198197898712404,
|
| 1640 |
+
"loss": 0.1261,
|
| 1641 |
+
"step": 232
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.2535364526659412,
|
| 1645 |
+
"grad_norm": 0.14948749542236328,
|
| 1646 |
+
"learning_rate": 0.0001717416407610824,
|
| 1647 |
+
"loss": 0.0874,
|
| 1648 |
+
"step": 233
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.2546245919477693,
|
| 1652 |
+
"grad_norm": 0.19695116579532623,
|
| 1653 |
+
"learning_rate": 0.00017150044560996488,
|
| 1654 |
+
"loss": 0.1119,
|
| 1655 |
+
"step": 234
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.25571273122959737,
|
| 1659 |
+
"grad_norm": 0.2416209876537323,
|
| 1660 |
+
"learning_rate": 0.00017125839641475072,
|
| 1661 |
+
"loss": 0.1495,
|
| 1662 |
+
"step": 235
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.25680087051142547,
|
| 1666 |
+
"grad_norm": 0.23595106601715088,
|
| 1667 |
+
"learning_rate": 0.00017101549606662024,
|
| 1668 |
+
"loss": 0.092,
|
| 1669 |
+
"step": 236
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.2578890097932535,
|
| 1673 |
+
"grad_norm": 0.3377005457878113,
|
| 1674 |
+
"learning_rate": 0.00017077174746692056,
|
| 1675 |
+
"loss": 0.1537,
|
| 1676 |
+
"step": 237
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.2589771490750816,
|
| 1680 |
+
"grad_norm": 0.31011515855789185,
|
| 1681 |
+
"learning_rate": 0.00017052715352713075,
|
| 1682 |
+
"loss": 0.2351,
|
| 1683 |
+
"step": 238
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.2600652883569097,
|
| 1687 |
+
"grad_norm": 0.2296973615884781,
|
| 1688 |
+
"learning_rate": 0.00017028171716882714,
|
| 1689 |
+
"loss": 0.1034,
|
| 1690 |
+
"step": 239
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.2611534276387378,
|
| 1694 |
+
"grad_norm": 0.33184927701950073,
|
| 1695 |
+
"learning_rate": 0.00017003544132364846,
|
| 1696 |
+
"loss": 0.1518,
|
| 1697 |
+
"step": 240
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.2622415669205658,
|
| 1701 |
+
"grad_norm": 0.333794504404068,
|
| 1702 |
+
"learning_rate": 0.00016978832893326074,
|
| 1703 |
+
"loss": 0.1167,
|
| 1704 |
+
"step": 241
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.26332970620239393,
|
| 1708 |
+
"grad_norm": 0.33567357063293457,
|
| 1709 |
+
"learning_rate": 0.00016954038294932216,
|
| 1710 |
+
"loss": 0.1672,
|
| 1711 |
+
"step": 242
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.264417845484222,
|
| 1715 |
+
"grad_norm": 0.3648099899291992,
|
| 1716 |
+
"learning_rate": 0.0001692916063334479,
|
| 1717 |
+
"loss": 0.1562,
|
| 1718 |
+
"step": 243
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.26550598476605003,
|
| 1722 |
+
"grad_norm": 0.3762454092502594,
|
| 1723 |
+
"learning_rate": 0.0001690420020571747,
|
| 1724 |
+
"loss": 0.1495,
|
| 1725 |
+
"step": 244
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.26659412404787813,
|
| 1729 |
+
"grad_norm": 0.42424383759498596,
|
| 1730 |
+
"learning_rate": 0.00016879157310192535,
|
| 1731 |
+
"loss": 0.1763,
|
| 1732 |
+
"step": 245
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.2676822633297062,
|
| 1736 |
+
"grad_norm": 0.4968826472759247,
|
| 1737 |
+
"learning_rate": 0.00016854032245897308,
|
| 1738 |
+
"loss": 0.2473,
|
| 1739 |
+
"step": 246
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.2687704026115343,
|
| 1743 |
+
"grad_norm": 0.5231485366821289,
|
| 1744 |
+
"learning_rate": 0.00016828825312940592,
|
| 1745 |
+
"loss": 0.2924,
|
| 1746 |
+
"step": 247
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.26985854189336234,
|
| 1750 |
+
"grad_norm": 0.5466935634613037,
|
| 1751 |
+
"learning_rate": 0.00016803536812409075,
|
| 1752 |
+
"loss": 0.2519,
|
| 1753 |
+
"step": 248
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.27094668117519044,
|
| 1757 |
+
"grad_norm": 0.6696439981460571,
|
| 1758 |
+
"learning_rate": 0.00016778167046363734,
|
| 1759 |
+
"loss": 0.2106,
|
| 1760 |
+
"step": 249
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.2720348204570185,
|
| 1764 |
+
"grad_norm": 0.7066907286643982,
|
| 1765 |
+
"learning_rate": 0.00016752716317836229,
|
| 1766 |
+
"loss": 0.2733,
|
| 1767 |
+
"step": 250
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.2731229597388466,
|
| 1771 |
+
"grad_norm": 0.058309707790613174,
|
| 1772 |
+
"learning_rate": 0.00016727184930825288,
|
| 1773 |
+
"loss": 0.0459,
|
| 1774 |
+
"step": 251
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.27421109902067464,
|
| 1778 |
+
"grad_norm": 0.06278934329748154,
|
| 1779 |
+
"learning_rate": 0.00016701573190293077,
|
| 1780 |
+
"loss": 0.049,
|
| 1781 |
+
"step": 252
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.27529923830250275,
|
| 1785 |
+
"grad_norm": 0.07942797988653183,
|
| 1786 |
+
"learning_rate": 0.00016675881402161536,
|
| 1787 |
+
"loss": 0.0757,
|
| 1788 |
+
"step": 253
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.2763873775843308,
|
| 1792 |
+
"grad_norm": 0.0874176099896431,
|
| 1793 |
+
"learning_rate": 0.00016650109873308765,
|
| 1794 |
+
"loss": 0.0952,
|
| 1795 |
+
"step": 254
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.27747551686615884,
|
| 1799 |
+
"grad_norm": 0.0788157656788826,
|
| 1800 |
+
"learning_rate": 0.0001662425891156531,
|
| 1801 |
+
"loss": 0.0655,
|
| 1802 |
+
"step": 255
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 0.27856365614798695,
|
| 1806 |
+
"grad_norm": 0.08784733712673187,
|
| 1807 |
+
"learning_rate": 0.00016598328825710533,
|
| 1808 |
+
"loss": 0.0778,
|
| 1809 |
+
"step": 256
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 0.279651795429815,
|
| 1813 |
+
"grad_norm": 0.09089700132608414,
|
| 1814 |
+
"learning_rate": 0.00016572319925468892,
|
| 1815 |
+
"loss": 0.0767,
|
| 1816 |
+
"step": 257
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 0.2807399347116431,
|
| 1820 |
+
"grad_norm": 0.07957662642002106,
|
| 1821 |
+
"learning_rate": 0.0001654623252150624,
|
| 1822 |
+
"loss": 0.0623,
|
| 1823 |
+
"step": 258
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 0.28182807399347115,
|
| 1827 |
+
"grad_norm": 0.08320681005716324,
|
| 1828 |
+
"learning_rate": 0.00016520066925426144,
|
| 1829 |
+
"loss": 0.0812,
|
| 1830 |
+
"step": 259
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 0.28291621327529926,
|
| 1834 |
+
"grad_norm": 0.10143834352493286,
|
| 1835 |
+
"learning_rate": 0.00016493823449766136,
|
| 1836 |
+
"loss": 0.0953,
|
| 1837 |
+
"step": 260
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 0.2840043525571273,
|
| 1841 |
+
"grad_norm": 0.1192905604839325,
|
| 1842 |
+
"learning_rate": 0.00016467502407993992,
|
| 1843 |
+
"loss": 0.1163,
|
| 1844 |
+
"step": 261
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 0.2850924918389554,
|
| 1848 |
+
"grad_norm": 0.11428846418857574,
|
| 1849 |
+
"learning_rate": 0.0001644110411450398,
|
| 1850 |
+
"loss": 0.1028,
|
| 1851 |
+
"step": 262
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 0.28618063112078346,
|
| 1855 |
+
"grad_norm": 0.11233223229646683,
|
| 1856 |
+
"learning_rate": 0.00016414628884613107,
|
| 1857 |
+
"loss": 0.091,
|
| 1858 |
+
"step": 263
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 0.28726877040261156,
|
| 1862 |
+
"grad_norm": 0.10367966443300247,
|
| 1863 |
+
"learning_rate": 0.00016388077034557355,
|
| 1864 |
+
"loss": 0.0797,
|
| 1865 |
+
"step": 264
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 0.2883569096844396,
|
| 1869 |
+
"grad_norm": 0.11604032665491104,
|
| 1870 |
+
"learning_rate": 0.00016361448881487914,
|
| 1871 |
+
"loss": 0.0919,
|
| 1872 |
+
"step": 265
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 0.28944504896626766,
|
| 1876 |
+
"grad_norm": 0.10309276729822159,
|
| 1877 |
+
"learning_rate": 0.00016334744743467364,
|
| 1878 |
+
"loss": 0.1065,
|
| 1879 |
+
"step": 266
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 0.29053318824809576,
|
| 1883 |
+
"grad_norm": 0.11475658416748047,
|
| 1884 |
+
"learning_rate": 0.00016307964939465914,
|
| 1885 |
+
"loss": 0.0959,
|
| 1886 |
+
"step": 267
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 0.2916213275299238,
|
| 1890 |
+
"grad_norm": 0.1230575293302536,
|
| 1891 |
+
"learning_rate": 0.0001628110978935756,
|
| 1892 |
+
"loss": 0.1031,
|
| 1893 |
+
"step": 268
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 0.2927094668117519,
|
| 1897 |
+
"grad_norm": 0.1267620474100113,
|
| 1898 |
+
"learning_rate": 0.00016254179613916278,
|
| 1899 |
+
"loss": 0.1219,
|
| 1900 |
+
"step": 269
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 0.29379760609357997,
|
| 1904 |
+
"grad_norm": 0.1032036617398262,
|
| 1905 |
+
"learning_rate": 0.000162271747348122,
|
| 1906 |
+
"loss": 0.0792,
|
| 1907 |
+
"step": 270
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 0.29488574537540807,
|
| 1911 |
+
"grad_norm": 0.10867134481668472,
|
| 1912 |
+
"learning_rate": 0.00016200095474607753,
|
| 1913 |
+
"loss": 0.0964,
|
| 1914 |
+
"step": 271
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 0.2959738846572361,
|
| 1918 |
+
"grad_norm": 0.13934585452079773,
|
| 1919 |
+
"learning_rate": 0.0001617294215675382,
|
| 1920 |
+
"loss": 0.1493,
|
| 1921 |
+
"step": 272
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 0.2970620239390642,
|
| 1925 |
+
"grad_norm": 0.1254916936159134,
|
| 1926 |
+
"learning_rate": 0.0001614571510558588,
|
| 1927 |
+
"loss": 0.1035,
|
| 1928 |
+
"step": 273
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 0.2981501632208923,
|
| 1932 |
+
"grad_norm": 0.10226383805274963,
|
| 1933 |
+
"learning_rate": 0.0001611841464632011,
|
| 1934 |
+
"loss": 0.0777,
|
| 1935 |
+
"step": 274
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 0.2992383025027203,
|
| 1939 |
+
"grad_norm": 0.11369970440864563,
|
| 1940 |
+
"learning_rate": 0.0001609104110504954,
|
| 1941 |
+
"loss": 0.0823,
|
| 1942 |
+
"step": 275
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 0.3003264417845484,
|
| 1946 |
+
"grad_norm": 0.11098276078701019,
|
| 1947 |
+
"learning_rate": 0.00016063594808740113,
|
| 1948 |
+
"loss": 0.0976,
|
| 1949 |
+
"step": 276
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 0.3014145810663765,
|
| 1953 |
+
"grad_norm": 0.13366885483264923,
|
| 1954 |
+
"learning_rate": 0.00016036076085226814,
|
| 1955 |
+
"loss": 0.1378,
|
| 1956 |
+
"step": 277
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 0.3025027203482046,
|
| 1960 |
+
"grad_norm": 0.11494230479001999,
|
| 1961 |
+
"learning_rate": 0.00016008485263209742,
|
| 1962 |
+
"loss": 0.072,
|
| 1963 |
+
"step": 278
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 0.30359085963003263,
|
| 1967 |
+
"grad_norm": 0.11145862936973572,
|
| 1968 |
+
"learning_rate": 0.0001598082267225018,
|
| 1969 |
+
"loss": 0.066,
|
| 1970 |
+
"step": 279
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 0.30467899891186073,
|
| 1974 |
+
"grad_norm": 0.1483200490474701,
|
| 1975 |
+
"learning_rate": 0.0001595308864276666,
|
| 1976 |
+
"loss": 0.1023,
|
| 1977 |
+
"step": 280
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 0.3057671381936888,
|
| 1981 |
+
"grad_norm": 0.12836772203445435,
|
| 1982 |
+
"learning_rate": 0.0001592528350603103,
|
| 1983 |
+
"loss": 0.0682,
|
| 1984 |
+
"step": 281
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 0.3068552774755169,
|
| 1988 |
+
"grad_norm": 0.16118410229682922,
|
| 1989 |
+
"learning_rate": 0.00015897407594164467,
|
| 1990 |
+
"loss": 0.1025,
|
| 1991 |
+
"step": 282
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 0.30794341675734493,
|
| 1995 |
+
"grad_norm": 0.22559022903442383,
|
| 1996 |
+
"learning_rate": 0.0001586946124013354,
|
| 1997 |
+
"loss": 0.1228,
|
| 1998 |
+
"step": 283
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 0.30903155603917304,
|
| 2002 |
+
"grad_norm": 0.233434796333313,
|
| 2003 |
+
"learning_rate": 0.0001584144477774623,
|
| 2004 |
+
"loss": 0.1928,
|
| 2005 |
+
"step": 284
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 0.3101196953210011,
|
| 2009 |
+
"grad_norm": 0.21861650049686432,
|
| 2010 |
+
"learning_rate": 0.00015813358541647915,
|
| 2011 |
+
"loss": 0.1054,
|
| 2012 |
+
"step": 285
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 0.31120783460282914,
|
| 2016 |
+
"grad_norm": 0.2723356783390045,
|
| 2017 |
+
"learning_rate": 0.00015785202867317407,
|
| 2018 |
+
"loss": 0.1411,
|
| 2019 |
+
"step": 286
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.31229597388465724,
|
| 2023 |
+
"grad_norm": 0.3065739870071411,
|
| 2024 |
+
"learning_rate": 0.0001575697809106292,
|
| 2025 |
+
"loss": 0.1785,
|
| 2026 |
+
"step": 287
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 0.3133841131664853,
|
| 2030 |
+
"grad_norm": 0.2983495593070984,
|
| 2031 |
+
"learning_rate": 0.00015728684550018064,
|
| 2032 |
+
"loss": 0.1402,
|
| 2033 |
+
"step": 288
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 0.3144722524483134,
|
| 2037 |
+
"grad_norm": 0.3250825107097626,
|
| 2038 |
+
"learning_rate": 0.00015700322582137827,
|
| 2039 |
+
"loss": 0.1929,
|
| 2040 |
+
"step": 289
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 0.31556039173014144,
|
| 2044 |
+
"grad_norm": 0.35388973355293274,
|
| 2045 |
+
"learning_rate": 0.00015671892526194516,
|
| 2046 |
+
"loss": 0.1791,
|
| 2047 |
+
"step": 290
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 0.31664853101196955,
|
| 2051 |
+
"grad_norm": 0.32610148191452026,
|
| 2052 |
+
"learning_rate": 0.0001564339472177373,
|
| 2053 |
+
"loss": 0.1289,
|
| 2054 |
+
"step": 291
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 0.3177366702937976,
|
| 2058 |
+
"grad_norm": 0.4028049409389496,
|
| 2059 |
+
"learning_rate": 0.0001561482950927029,
|
| 2060 |
+
"loss": 0.2026,
|
| 2061 |
+
"step": 292
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 0.3188248095756257,
|
| 2065 |
+
"grad_norm": 0.2420492172241211,
|
| 2066 |
+
"learning_rate": 0.00015586197229884184,
|
| 2067 |
+
"loss": 0.098,
|
| 2068 |
+
"step": 293
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 0.31991294885745375,
|
| 2072 |
+
"grad_norm": 0.3512971103191376,
|
| 2073 |
+
"learning_rate": 0.00015557498225616487,
|
| 2074 |
+
"loss": 0.205,
|
| 2075 |
+
"step": 294
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 0.32100108813928185,
|
| 2079 |
+
"grad_norm": 0.39271989464759827,
|
| 2080 |
+
"learning_rate": 0.00015528732839265272,
|
| 2081 |
+
"loss": 0.1473,
|
| 2082 |
+
"step": 295
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 0.3220892274211099,
|
| 2086 |
+
"grad_norm": 0.3802226185798645,
|
| 2087 |
+
"learning_rate": 0.0001549990141442153,
|
| 2088 |
+
"loss": 0.136,
|
| 2089 |
+
"step": 296
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 0.32317736670293795,
|
| 2093 |
+
"grad_norm": 0.5737869739532471,
|
| 2094 |
+
"learning_rate": 0.00015471004295465035,
|
| 2095 |
+
"loss": 0.3053,
|
| 2096 |
+
"step": 297
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.32426550598476606,
|
| 2100 |
+
"grad_norm": 0.45224013924598694,
|
| 2101 |
+
"learning_rate": 0.00015442041827560274,
|
| 2102 |
+
"loss": 0.222,
|
| 2103 |
+
"step": 298
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 0.3253536452665941,
|
| 2107 |
+
"grad_norm": 0.522432267665863,
|
| 2108 |
+
"learning_rate": 0.00015413014356652286,
|
| 2109 |
+
"loss": 0.1809,
|
| 2110 |
+
"step": 299
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 0.3264417845484222,
|
| 2114 |
+
"grad_norm": 0.6229780316352844,
|
| 2115 |
+
"learning_rate": 0.00015383922229462549,
|
| 2116 |
+
"loss": 0.2081,
|
| 2117 |
+
"step": 300
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 0.32752992383025026,
|
| 2121 |
+
"grad_norm": 0.053111448884010315,
|
| 2122 |
+
"learning_rate": 0.00015354765793484834,
|
| 2123 |
+
"loss": 0.0414,
|
| 2124 |
+
"step": 301
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 0.32861806311207836,
|
| 2128 |
+
"grad_norm": 0.07464036345481873,
|
| 2129 |
+
"learning_rate": 0.0001532554539698105,
|
| 2130 |
+
"loss": 0.0639,
|
| 2131 |
+
"step": 302
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 0.3297062023939064,
|
| 2135 |
+
"grad_norm": 0.08635352551937103,
|
| 2136 |
+
"learning_rate": 0.00015296261388977108,
|
| 2137 |
+
"loss": 0.074,
|
| 2138 |
+
"step": 303
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 0.3307943416757345,
|
| 2142 |
+
"grad_norm": 0.0818236917257309,
|
| 2143 |
+
"learning_rate": 0.000152669141192587,
|
| 2144 |
+
"loss": 0.0843,
|
| 2145 |
+
"step": 304
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 0.33188248095756256,
|
| 2149 |
+
"grad_norm": 0.08959626406431198,
|
| 2150 |
+
"learning_rate": 0.00015237503938367186,
|
| 2151 |
+
"loss": 0.0752,
|
| 2152 |
+
"step": 305
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 0.33297062023939067,
|
| 2156 |
+
"grad_norm": 0.087018683552742,
|
| 2157 |
+
"learning_rate": 0.00015208031197595356,
|
| 2158 |
+
"loss": 0.074,
|
| 2159 |
+
"step": 306
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 0.3340587595212187,
|
| 2163 |
+
"grad_norm": 0.10946961492300034,
|
| 2164 |
+
"learning_rate": 0.00015178496248983254,
|
| 2165 |
+
"loss": 0.0907,
|
| 2166 |
+
"step": 307
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 0.33514689880304677,
|
| 2170 |
+
"grad_norm": 0.09914237260818481,
|
| 2171 |
+
"learning_rate": 0.00015148899445313981,
|
| 2172 |
+
"loss": 0.0939,
|
| 2173 |
+
"step": 308
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 0.33623503808487487,
|
| 2177 |
+
"grad_norm": 0.07641992717981339,
|
| 2178 |
+
"learning_rate": 0.00015119241140109467,
|
| 2179 |
+
"loss": 0.0587,
|
| 2180 |
+
"step": 309
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 0.3373231773667029,
|
| 2184 |
+
"grad_norm": 0.10857319831848145,
|
| 2185 |
+
"learning_rate": 0.00015089521687626243,
|
| 2186 |
+
"loss": 0.1052,
|
| 2187 |
+
"step": 310
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 0.338411316648531,
|
| 2191 |
+
"grad_norm": 0.0868939459323883,
|
| 2192 |
+
"learning_rate": 0.0001505974144285124,
|
| 2193 |
+
"loss": 0.0723,
|
| 2194 |
+
"step": 311
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 0.3394994559303591,
|
| 2198 |
+
"grad_norm": 0.11470666527748108,
|
| 2199 |
+
"learning_rate": 0.00015029900761497506,
|
| 2200 |
+
"loss": 0.0972,
|
| 2201 |
+
"step": 312
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.3405875952121872,
|
| 2205 |
+
"grad_norm": 0.09828225523233414,
|
| 2206 |
+
"learning_rate": 0.00015000000000000001,
|
| 2207 |
+
"loss": 0.0904,
|
| 2208 |
+
"step": 313
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.3416757344940152,
|
| 2212 |
+
"grad_norm": 0.09422045201063156,
|
| 2213 |
+
"learning_rate": 0.00014970039515511304,
|
| 2214 |
+
"loss": 0.0736,
|
| 2215 |
+
"step": 314
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.34276387377584333,
|
| 2219 |
+
"grad_norm": 0.09876245260238647,
|
| 2220 |
+
"learning_rate": 0.0001494001966589736,
|
| 2221 |
+
"loss": 0.0849,
|
| 2222 |
+
"step": 315
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.3438520130576714,
|
| 2226 |
+
"grad_norm": 0.1073005348443985,
|
| 2227 |
+
"learning_rate": 0.00014909940809733222,
|
| 2228 |
+
"loss": 0.0842,
|
| 2229 |
+
"step": 316
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.34494015233949943,
|
| 2233 |
+
"grad_norm": 0.11519600450992584,
|
| 2234 |
+
"learning_rate": 0.00014879803306298736,
|
| 2235 |
+
"loss": 0.0901,
|
| 2236 |
+
"step": 317
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 0.34602829162132753,
|
| 2240 |
+
"grad_norm": 0.10380937904119492,
|
| 2241 |
+
"learning_rate": 0.00014849607515574276,
|
| 2242 |
+
"loss": 0.0688,
|
| 2243 |
+
"step": 318
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.3471164309031556,
|
| 2247 |
+
"grad_norm": 0.10230353474617004,
|
| 2248 |
+
"learning_rate": 0.00014819353798236427,
|
| 2249 |
+
"loss": 0.064,
|
| 2250 |
+
"step": 319
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 0.3482045701849837,
|
| 2254 |
+
"grad_norm": 0.10846245288848877,
|
| 2255 |
+
"learning_rate": 0.00014789042515653687,
|
| 2256 |
+
"loss": 0.0815,
|
| 2257 |
+
"step": 320
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.34929270946681173,
|
| 2261 |
+
"grad_norm": 0.11520566791296005,
|
| 2262 |
+
"learning_rate": 0.00014758674029882152,
|
| 2263 |
+
"loss": 0.0846,
|
| 2264 |
+
"step": 321
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 0.35038084874863984,
|
| 2268 |
+
"grad_norm": 0.16834412515163422,
|
| 2269 |
+
"learning_rate": 0.00014728248703661182,
|
| 2270 |
+
"loss": 0.1249,
|
| 2271 |
+
"step": 322
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.3514689880304679,
|
| 2275 |
+
"grad_norm": 0.11053828150033951,
|
| 2276 |
+
"learning_rate": 0.00014697766900409074,
|
| 2277 |
+
"loss": 0.073,
|
| 2278 |
+
"step": 323
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 0.352557127312296,
|
| 2282 |
+
"grad_norm": 0.12219499796628952,
|
| 2283 |
+
"learning_rate": 0.0001466722898421873,
|
| 2284 |
+
"loss": 0.0943,
|
| 2285 |
+
"step": 324
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 0.35364526659412404,
|
| 2289 |
+
"grad_norm": 0.1294214427471161,
|
| 2290 |
+
"learning_rate": 0.00014636635319853275,
|
| 2291 |
+
"loss": 0.0761,
|
| 2292 |
+
"step": 325
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 0.35473340587595215,
|
| 2296 |
+
"grad_norm": 0.13043484091758728,
|
| 2297 |
+
"learning_rate": 0.00014605986272741748,
|
| 2298 |
+
"loss": 0.1065,
|
| 2299 |
+
"step": 326
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.3558215451577802,
|
| 2303 |
+
"grad_norm": 0.1078469529747963,
|
| 2304 |
+
"learning_rate": 0.00014575282208974702,
|
| 2305 |
+
"loss": 0.0718,
|
| 2306 |
+
"step": 327
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.35690968443960824,
|
| 2310 |
+
"grad_norm": 0.17083537578582764,
|
| 2311 |
+
"learning_rate": 0.00014544523495299842,
|
| 2312 |
+
"loss": 0.1035,
|
| 2313 |
+
"step": 328
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.35799782372143635,
|
| 2317 |
+
"grad_norm": 0.1370207518339157,
|
| 2318 |
+
"learning_rate": 0.00014513710499117647,
|
| 2319 |
+
"loss": 0.089,
|
| 2320 |
+
"step": 329
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 0.3590859630032644,
|
| 2324 |
+
"grad_norm": 0.1698474884033203,
|
| 2325 |
+
"learning_rate": 0.00014482843588476974,
|
| 2326 |
+
"loss": 0.1172,
|
| 2327 |
+
"step": 330
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 0.3601741022850925,
|
| 2331 |
+
"grad_norm": 0.1472265124320984,
|
| 2332 |
+
"learning_rate": 0.0001445192313207067,
|
| 2333 |
+
"loss": 0.0782,
|
| 2334 |
+
"step": 331
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 0.36126224156692055,
|
| 2338 |
+
"grad_norm": 0.153669074177742,
|
| 2339 |
+
"learning_rate": 0.00014420949499231172,
|
| 2340 |
+
"loss": 0.0844,
|
| 2341 |
+
"step": 332
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.36235038084874865,
|
| 2345 |
+
"grad_norm": 0.2612091600894928,
|
| 2346 |
+
"learning_rate": 0.00014389923059926062,
|
| 2347 |
+
"loss": 0.1256,
|
| 2348 |
+
"step": 333
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 0.3634385201305767,
|
| 2352 |
+
"grad_norm": 0.18867933750152588,
|
| 2353 |
+
"learning_rate": 0.00014358844184753712,
|
| 2354 |
+
"loss": 0.0822,
|
| 2355 |
+
"step": 334
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 0.3645266594124048,
|
| 2359 |
+
"grad_norm": 0.29405227303504944,
|
| 2360 |
+
"learning_rate": 0.0001432771324493879,
|
| 2361 |
+
"loss": 0.1562,
|
| 2362 |
+
"step": 335
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 0.36561479869423286,
|
| 2366 |
+
"grad_norm": 0.299411803483963,
|
| 2367 |
+
"learning_rate": 0.00014296530612327863,
|
| 2368 |
+
"loss": 0.151,
|
| 2369 |
+
"step": 336
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 0.36670293797606096,
|
| 2373 |
+
"grad_norm": 0.4020368754863739,
|
| 2374 |
+
"learning_rate": 0.00014265296659384956,
|
| 2375 |
+
"loss": 0.2564,
|
| 2376 |
+
"step": 337
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 0.367791077257889,
|
| 2380 |
+
"grad_norm": 0.27561965584754944,
|
| 2381 |
+
"learning_rate": 0.00014234011759187083,
|
| 2382 |
+
"loss": 0.1193,
|
| 2383 |
+
"step": 338
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 0.36887921653971706,
|
| 2387 |
+
"grad_norm": 0.36899837851524353,
|
| 2388 |
+
"learning_rate": 0.00014202676285419812,
|
| 2389 |
+
"loss": 0.1844,
|
| 2390 |
+
"step": 339
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 0.36996735582154516,
|
| 2394 |
+
"grad_norm": 0.3305605351924896,
|
| 2395 |
+
"learning_rate": 0.0001417129061237278,
|
| 2396 |
+
"loss": 0.0825,
|
| 2397 |
+
"step": 340
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 0.3710554951033732,
|
| 2401 |
+
"grad_norm": 0.45063266158103943,
|
| 2402 |
+
"learning_rate": 0.00014139855114935252,
|
| 2403 |
+
"loss": 0.2383,
|
| 2404 |
+
"step": 341
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 0.3721436343852013,
|
| 2408 |
+
"grad_norm": 0.319297730922699,
|
| 2409 |
+
"learning_rate": 0.0001410837016859161,
|
| 2410 |
+
"loss": 0.1054,
|
| 2411 |
+
"step": 342
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 0.37323177366702937,
|
| 2415 |
+
"grad_norm": 0.3711492121219635,
|
| 2416 |
+
"learning_rate": 0.00014076836149416887,
|
| 2417 |
+
"loss": 0.1265,
|
| 2418 |
+
"step": 343
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 0.37431991294885747,
|
| 2422 |
+
"grad_norm": 0.4362325966358185,
|
| 2423 |
+
"learning_rate": 0.0001404525343407228,
|
| 2424 |
+
"loss": 0.1818,
|
| 2425 |
+
"step": 344
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 0.3754080522306855,
|
| 2429 |
+
"grad_norm": 0.39839836955070496,
|
| 2430 |
+
"learning_rate": 0.00014013622399800627,
|
| 2431 |
+
"loss": 0.173,
|
| 2432 |
+
"step": 345
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 0.3764961915125136,
|
| 2436 |
+
"grad_norm": 0.4215060770511627,
|
| 2437 |
+
"learning_rate": 0.00013981943424421932,
|
| 2438 |
+
"loss": 0.1225,
|
| 2439 |
+
"step": 346
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 0.37758433079434167,
|
| 2443 |
+
"grad_norm": 0.4466668963432312,
|
| 2444 |
+
"learning_rate": 0.0001395021688632882,
|
| 2445 |
+
"loss": 0.1336,
|
| 2446 |
+
"step": 347
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 0.3786724700761698,
|
| 2450 |
+
"grad_norm": 0.490313321352005,
|
| 2451 |
+
"learning_rate": 0.00013918443164482046,
|
| 2452 |
+
"loss": 0.1364,
|
| 2453 |
+
"step": 348
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 0.3797606093579978,
|
| 2457 |
+
"grad_norm": 0.6865617036819458,
|
| 2458 |
+
"learning_rate": 0.00013886622638405952,
|
| 2459 |
+
"loss": 0.2864,
|
| 2460 |
+
"step": 349
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 0.3808487486398259,
|
| 2464 |
+
"grad_norm": 0.7716324925422668,
|
| 2465 |
+
"learning_rate": 0.0001385475568818394,
|
| 2466 |
+
"loss": 0.2994,
|
| 2467 |
+
"step": 350
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 0.381936887921654,
|
| 2471 |
+
"grad_norm": 0.04748038947582245,
|
| 2472 |
+
"learning_rate": 0.00013822842694453924,
|
| 2473 |
+
"loss": 0.0425,
|
| 2474 |
+
"step": 351
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 0.383025027203482,
|
| 2478 |
+
"grad_norm": 0.06222306191921234,
|
| 2479 |
+
"learning_rate": 0.00013790884038403795,
|
| 2480 |
+
"loss": 0.0519,
|
| 2481 |
+
"step": 352
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 0.38411316648531013,
|
| 2485 |
+
"grad_norm": 0.07450418174266815,
|
| 2486 |
+
"learning_rate": 0.0001375888010176686,
|
| 2487 |
+
"loss": 0.0629,
|
| 2488 |
+
"step": 353
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 0.3852013057671382,
|
| 2492 |
+
"grad_norm": 0.08733393251895905,
|
| 2493 |
+
"learning_rate": 0.00013726831266817278,
|
| 2494 |
+
"loss": 0.0701,
|
| 2495 |
+
"step": 354
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 0.3862894450489663,
|
| 2499 |
+
"grad_norm": 0.09622704982757568,
|
| 2500 |
+
"learning_rate": 0.00013694737916365517,
|
| 2501 |
+
"loss": 0.0909,
|
| 2502 |
+
"step": 355
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 0.38737758433079433,
|
| 2506 |
+
"grad_norm": 0.08062370121479034,
|
| 2507 |
+
"learning_rate": 0.00013662600433753745,
|
| 2508 |
+
"loss": 0.0722,
|
| 2509 |
+
"step": 356
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 0.38846572361262244,
|
| 2513 |
+
"grad_norm": 0.09811591356992722,
|
| 2514 |
+
"learning_rate": 0.00013630419202851284,
|
| 2515 |
+
"loss": 0.0894,
|
| 2516 |
+
"step": 357
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 0.3895538628944505,
|
| 2520 |
+
"grad_norm": 0.0917980894446373,
|
| 2521 |
+
"learning_rate": 0.0001359819460805001,
|
| 2522 |
+
"loss": 0.0816,
|
| 2523 |
+
"step": 358
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 0.3906420021762786,
|
| 2527 |
+
"grad_norm": 0.08292034268379211,
|
| 2528 |
+
"learning_rate": 0.0001356592703425976,
|
| 2529 |
+
"loss": 0.0832,
|
| 2530 |
+
"step": 359
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 0.39173014145810664,
|
| 2534 |
+
"grad_norm": 0.0940559059381485,
|
| 2535 |
+
"learning_rate": 0.00013533616866903735,
|
| 2536 |
+
"loss": 0.078,
|
| 2537 |
+
"step": 360
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 0.3928182807399347,
|
| 2541 |
+
"grad_norm": 0.09960496425628662,
|
| 2542 |
+
"learning_rate": 0.00013501264491913906,
|
| 2543 |
+
"loss": 0.0899,
|
| 2544 |
+
"step": 361
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 0.3939064200217628,
|
| 2548 |
+
"grad_norm": 0.1174091249704361,
|
| 2549 |
+
"learning_rate": 0.00013468870295726398,
|
| 2550 |
+
"loss": 0.0983,
|
| 2551 |
+
"step": 362
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 0.39499455930359084,
|
| 2555 |
+
"grad_norm": 0.1083730086684227,
|
| 2556 |
+
"learning_rate": 0.00013436434665276865,
|
| 2557 |
+
"loss": 0.1004,
|
| 2558 |
+
"step": 363
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 0.39608269858541895,
|
| 2562 |
+
"grad_norm": 0.09829343855381012,
|
| 2563 |
+
"learning_rate": 0.00013403957987995882,
|
| 2564 |
+
"loss": 0.0851,
|
| 2565 |
+
"step": 364
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 0.397170837867247,
|
| 2569 |
+
"grad_norm": 0.1172933354973793,
|
| 2570 |
+
"learning_rate": 0.00013371440651804313,
|
| 2571 |
+
"loss": 0.1033,
|
| 2572 |
+
"step": 365
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 0.3982589771490751,
|
| 2576 |
+
"grad_norm": 0.11004797369241714,
|
| 2577 |
+
"learning_rate": 0.00013338883045108674,
|
| 2578 |
+
"loss": 0.0852,
|
| 2579 |
+
"step": 366
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 0.39934711643090315,
|
| 2583 |
+
"grad_norm": 0.10466606914997101,
|
| 2584 |
+
"learning_rate": 0.00013306285556796495,
|
| 2585 |
+
"loss": 0.0893,
|
| 2586 |
+
"step": 367
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 0.40043525571273125,
|
| 2590 |
+
"grad_norm": 0.121376633644104,
|
| 2591 |
+
"learning_rate": 0.0001327364857623168,
|
| 2592 |
+
"loss": 0.1037,
|
| 2593 |
+
"step": 368
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 0.4015233949945593,
|
| 2597 |
+
"grad_norm": 0.10333437472581863,
|
| 2598 |
+
"learning_rate": 0.00013240972493249847,
|
| 2599 |
+
"loss": 0.0981,
|
| 2600 |
+
"step": 369
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 0.40261153427638735,
|
| 2604 |
+
"grad_norm": 0.1174560934305191,
|
| 2605 |
+
"learning_rate": 0.00013208257698153677,
|
| 2606 |
+
"loss": 0.1187,
|
| 2607 |
+
"step": 370
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 0.40369967355821545,
|
| 2611 |
+
"grad_norm": 0.09671124815940857,
|
| 2612 |
+
"learning_rate": 0.0001317550458170826,
|
| 2613 |
+
"loss": 0.0771,
|
| 2614 |
+
"step": 371
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 0.4047878128400435,
|
| 2618 |
+
"grad_norm": 0.11311496794223785,
|
| 2619 |
+
"learning_rate": 0.00013142713535136414,
|
| 2620 |
+
"loss": 0.0915,
|
| 2621 |
+
"step": 372
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 0.4058759521218716,
|
| 2625 |
+
"grad_norm": 0.11149045825004578,
|
| 2626 |
+
"learning_rate": 0.00013109884950114007,
|
| 2627 |
+
"loss": 0.078,
|
| 2628 |
+
"step": 373
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 0.40696409140369966,
|
| 2632 |
+
"grad_norm": 0.15049664676189423,
|
| 2633 |
+
"learning_rate": 0.00013077019218765305,
|
| 2634 |
+
"loss": 0.1008,
|
| 2635 |
+
"step": 374
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 0.40805223068552776,
|
| 2639 |
+
"grad_norm": 0.13566477596759796,
|
| 2640 |
+
"learning_rate": 0.0001304411673365826,
|
| 2641 |
+
"loss": 0.1116,
|
| 2642 |
+
"step": 375
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 0.4091403699673558,
|
| 2646 |
+
"grad_norm": 0.1317652463912964,
|
| 2647 |
+
"learning_rate": 0.00013011177887799845,
|
| 2648 |
+
"loss": 0.1068,
|
| 2649 |
+
"step": 376
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 0.4102285092491839,
|
| 2653 |
+
"grad_norm": 0.12117652595043182,
|
| 2654 |
+
"learning_rate": 0.00012978203074631334,
|
| 2655 |
+
"loss": 0.0926,
|
| 2656 |
+
"step": 377
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 0.41131664853101196,
|
| 2660 |
+
"grad_norm": 0.13246335089206696,
|
| 2661 |
+
"learning_rate": 0.00012945192688023624,
|
| 2662 |
+
"loss": 0.0867,
|
| 2663 |
+
"step": 378
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 0.41240478781284007,
|
| 2667 |
+
"grad_norm": 0.1427900642156601,
|
| 2668 |
+
"learning_rate": 0.00012912147122272523,
|
| 2669 |
+
"loss": 0.1054,
|
| 2670 |
+
"step": 379
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 0.4134929270946681,
|
| 2674 |
+
"grad_norm": 0.13975268602371216,
|
| 2675 |
+
"learning_rate": 0.0001287906677209403,
|
| 2676 |
+
"loss": 0.0993,
|
| 2677 |
+
"step": 380
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 0.41458106637649617,
|
| 2681 |
+
"grad_norm": 0.16829046607017517,
|
| 2682 |
+
"learning_rate": 0.0001284595203261965,
|
| 2683 |
+
"loss": 0.0986,
|
| 2684 |
+
"step": 381
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 0.41566920565832427,
|
| 2688 |
+
"grad_norm": 0.18288354575634003,
|
| 2689 |
+
"learning_rate": 0.00012812803299391628,
|
| 2690 |
+
"loss": 0.1164,
|
| 2691 |
+
"step": 382
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 0.4167573449401523,
|
| 2695 |
+
"grad_norm": 0.20097504556179047,
|
| 2696 |
+
"learning_rate": 0.00012779620968358273,
|
| 2697 |
+
"loss": 0.1273,
|
| 2698 |
+
"step": 383
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 0.4178454842219804,
|
| 2702 |
+
"grad_norm": 0.1646791398525238,
|
| 2703 |
+
"learning_rate": 0.00012746405435869198,
|
| 2704 |
+
"loss": 0.0833,
|
| 2705 |
+
"step": 384
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 0.41893362350380847,
|
| 2709 |
+
"grad_norm": 0.1997787058353424,
|
| 2710 |
+
"learning_rate": 0.0001271315709867059,
|
| 2711 |
+
"loss": 0.1495,
|
| 2712 |
+
"step": 385
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 0.4200217627856366,
|
| 2716 |
+
"grad_norm": 0.1489897519350052,
|
| 2717 |
+
"learning_rate": 0.00012679876353900482,
|
| 2718 |
+
"loss": 0.0756,
|
| 2719 |
+
"step": 386
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 0.4211099020674646,
|
| 2723 |
+
"grad_norm": 0.22502455115318298,
|
| 2724 |
+
"learning_rate": 0.00012646563599083996,
|
| 2725 |
+
"loss": 0.1427,
|
| 2726 |
+
"step": 387
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 0.42219804134929273,
|
| 2730 |
+
"grad_norm": 0.19359458982944489,
|
| 2731 |
+
"learning_rate": 0.00012613219232128608,
|
| 2732 |
+
"loss": 0.121,
|
| 2733 |
+
"step": 388
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 0.4232861806311208,
|
| 2737 |
+
"grad_norm": 0.244260773062706,
|
| 2738 |
+
"learning_rate": 0.0001257984365131938,
|
| 2739 |
+
"loss": 0.1469,
|
| 2740 |
+
"step": 389
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 0.4243743199129489,
|
| 2744 |
+
"grad_norm": 0.18485282361507416,
|
| 2745 |
+
"learning_rate": 0.00012546437255314222,
|
| 2746 |
+
"loss": 0.0892,
|
| 2747 |
+
"step": 390
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 0.42546245919477693,
|
| 2751 |
+
"grad_norm": 0.3717290461063385,
|
| 2752 |
+
"learning_rate": 0.00012513000443139112,
|
| 2753 |
+
"loss": 0.2099,
|
| 2754 |
+
"step": 391
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.426550598476605,
|
| 2758 |
+
"grad_norm": 0.28721094131469727,
|
| 2759 |
+
"learning_rate": 0.00012479533614183334,
|
| 2760 |
+
"loss": 0.1193,
|
| 2761 |
+
"step": 392
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.4276387377584331,
|
| 2765 |
+
"grad_norm": 0.2697299122810364,
|
| 2766 |
+
"learning_rate": 0.00012446037168194714,
|
| 2767 |
+
"loss": 0.0965,
|
| 2768 |
+
"step": 393
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 0.42872687704026113,
|
| 2772 |
+
"grad_norm": 0.32627496123313904,
|
| 2773 |
+
"learning_rate": 0.00012412511505274844,
|
| 2774 |
+
"loss": 0.1832,
|
| 2775 |
+
"step": 394
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 0.42981501632208924,
|
| 2779 |
+
"grad_norm": 0.37745073437690735,
|
| 2780 |
+
"learning_rate": 0.000123789570258743,
|
| 2781 |
+
"loss": 0.1572,
|
| 2782 |
+
"step": 395
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 0.4309031556039173,
|
| 2786 |
+
"grad_norm": 0.4901193082332611,
|
| 2787 |
+
"learning_rate": 0.00012345374130787854,
|
| 2788 |
+
"loss": 0.1873,
|
| 2789 |
+
"step": 396
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 0.4319912948857454,
|
| 2793 |
+
"grad_norm": 0.44663485884666443,
|
| 2794 |
+
"learning_rate": 0.000123117632211497,
|
| 2795 |
+
"loss": 0.3353,
|
| 2796 |
+
"step": 397
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.43307943416757344,
|
| 2800 |
+
"grad_norm": 0.34345391392707825,
|
| 2801 |
+
"learning_rate": 0.0001227812469842864,
|
| 2802 |
+
"loss": 0.1924,
|
| 2803 |
+
"step": 398
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 0.43416757344940154,
|
| 2807 |
+
"grad_norm": 0.5725805759429932,
|
| 2808 |
+
"learning_rate": 0.00012244458964423327,
|
| 2809 |
+
"loss": 0.2352,
|
| 2810 |
+
"step": 399
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 0.4352557127312296,
|
| 2814 |
+
"grad_norm": 0.5519152879714966,
|
| 2815 |
+
"learning_rate": 0.0001221076642125742,
|
| 2816 |
+
"loss": 0.167,
|
| 2817 |
+
"step": 400
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 0.4363438520130577,
|
| 2821 |
+
"grad_norm": 0.055198315531015396,
|
| 2822 |
+
"learning_rate": 0.00012177047471374807,
|
| 2823 |
+
"loss": 0.0472,
|
| 2824 |
+
"step": 401
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 0.43743199129488575,
|
| 2828 |
+
"grad_norm": 0.09932799637317657,
|
| 2829 |
+
"learning_rate": 0.0001214330251753481,
|
| 2830 |
+
"loss": 0.0783,
|
| 2831 |
+
"step": 402
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 0.4385201305767138,
|
| 2835 |
+
"grad_norm": 0.08226185292005539,
|
| 2836 |
+
"learning_rate": 0.00012109531962807332,
|
| 2837 |
+
"loss": 0.0656,
|
| 2838 |
+
"step": 403
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 0.4396082698585419,
|
| 2842 |
+
"grad_norm": 0.0858379453420639,
|
| 2843 |
+
"learning_rate": 0.0001207573621056809,
|
| 2844 |
+
"loss": 0.0741,
|
| 2845 |
+
"step": 404
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 0.44069640914036995,
|
| 2849 |
+
"grad_norm": 0.07838830351829529,
|
| 2850 |
+
"learning_rate": 0.00012041915664493761,
|
| 2851 |
+
"loss": 0.066,
|
| 2852 |
+
"step": 405
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 0.44178454842219805,
|
| 2856 |
+
"grad_norm": 0.08843716233968735,
|
| 2857 |
+
"learning_rate": 0.00012008070728557186,
|
| 2858 |
+
"loss": 0.0817,
|
| 2859 |
+
"step": 406
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 0.4428726877040261,
|
| 2863 |
+
"grad_norm": 0.09485173225402832,
|
| 2864 |
+
"learning_rate": 0.00011974201807022525,
|
| 2865 |
+
"loss": 0.0719,
|
| 2866 |
+
"step": 407
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 0.4439608269858542,
|
| 2870 |
+
"grad_norm": 0.12550269067287445,
|
| 2871 |
+
"learning_rate": 0.00011940309304440433,
|
| 2872 |
+
"loss": 0.1025,
|
| 2873 |
+
"step": 408
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 0.44504896626768226,
|
| 2877 |
+
"grad_norm": 0.10056477040052414,
|
| 2878 |
+
"learning_rate": 0.00011906393625643244,
|
| 2879 |
+
"loss": 0.0822,
|
| 2880 |
+
"step": 409
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 0.44613710554951036,
|
| 2884 |
+
"grad_norm": 0.08779609948396683,
|
| 2885 |
+
"learning_rate": 0.00011872455175740112,
|
| 2886 |
+
"loss": 0.0731,
|
| 2887 |
+
"step": 410
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 0.4472252448313384,
|
| 2891 |
+
"grad_norm": 0.08771763741970062,
|
| 2892 |
+
"learning_rate": 0.00011838494360112185,
|
| 2893 |
+
"loss": 0.0686,
|
| 2894 |
+
"step": 411
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 0.44831338411316646,
|
| 2898 |
+
"grad_norm": 0.09602241218090057,
|
| 2899 |
+
"learning_rate": 0.00011804511584407763,
|
| 2900 |
+
"loss": 0.0826,
|
| 2901 |
+
"step": 412
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 0.44940152339499456,
|
| 2905 |
+
"grad_norm": 0.10052221268415451,
|
| 2906 |
+
"learning_rate": 0.00011770507254537453,
|
| 2907 |
+
"loss": 0.0711,
|
| 2908 |
+
"step": 413
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 0.4504896626768226,
|
| 2912 |
+
"grad_norm": 0.08452215045690536,
|
| 2913 |
+
"learning_rate": 0.00011736481776669306,
|
| 2914 |
+
"loss": 0.061,
|
| 2915 |
+
"step": 414
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 0.4515778019586507,
|
| 2919 |
+
"grad_norm": 0.09362675249576569,
|
| 2920 |
+
"learning_rate": 0.00011702435557223987,
|
| 2921 |
+
"loss": 0.0852,
|
| 2922 |
+
"step": 415
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 0.45266594124047876,
|
| 2926 |
+
"grad_norm": 0.10676004737615585,
|
| 2927 |
+
"learning_rate": 0.00011668369002869912,
|
| 2928 |
+
"loss": 0.101,
|
| 2929 |
+
"step": 416
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 0.45375408052230687,
|
| 2933 |
+
"grad_norm": 0.10523767024278641,
|
| 2934 |
+
"learning_rate": 0.00011634282520518383,
|
| 2935 |
+
"loss": 0.0892,
|
| 2936 |
+
"step": 417
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 0.4548422198041349,
|
| 2940 |
+
"grad_norm": 0.09733587503433228,
|
| 2941 |
+
"learning_rate": 0.00011600176517318741,
|
| 2942 |
+
"loss": 0.0714,
|
| 2943 |
+
"step": 418
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 0.455930359085963,
|
| 2947 |
+
"grad_norm": 0.13115546107292175,
|
| 2948 |
+
"learning_rate": 0.00011566051400653486,
|
| 2949 |
+
"loss": 0.1079,
|
| 2950 |
+
"step": 419
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 0.45701849836779107,
|
| 2954 |
+
"grad_norm": 0.10536440461874008,
|
| 2955 |
+
"learning_rate": 0.00011531907578133429,
|
| 2956 |
+
"loss": 0.0807,
|
| 2957 |
+
"step": 420
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 0.4581066376496192,
|
| 2961 |
+
"grad_norm": 0.10071249306201935,
|
| 2962 |
+
"learning_rate": 0.00011497745457592816,
|
| 2963 |
+
"loss": 0.0647,
|
| 2964 |
+
"step": 421
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 0.4591947769314472,
|
| 2968 |
+
"grad_norm": 0.12494815140962601,
|
| 2969 |
+
"learning_rate": 0.00011463565447084445,
|
| 2970 |
+
"loss": 0.0969,
|
| 2971 |
+
"step": 422
|
| 2972 |
+
},
|
| 2973 |
+
{
|
| 2974 |
+
"epoch": 0.4602829162132753,
|
| 2975 |
+
"grad_norm": 0.10858377069234848,
|
| 2976 |
+
"learning_rate": 0.00011429367954874819,
|
| 2977 |
+
"loss": 0.0709,
|
| 2978 |
+
"step": 423
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"epoch": 0.4613710554951034,
|
| 2982 |
+
"grad_norm": 0.10477497428655624,
|
| 2983 |
+
"learning_rate": 0.00011395153389439233,
|
| 2984 |
+
"loss": 0.0875,
|
| 2985 |
+
"step": 424
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 0.4624591947769314,
|
| 2989 |
+
"grad_norm": 0.12716920673847198,
|
| 2990 |
+
"learning_rate": 0.00011360922159456928,
|
| 2991 |
+
"loss": 0.0864,
|
| 2992 |
+
"step": 425
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 0.46354733405875953,
|
| 2996 |
+
"grad_norm": 0.13803425431251526,
|
| 2997 |
+
"learning_rate": 0.00011326674673806195,
|
| 2998 |
+
"loss": 0.1028,
|
| 2999 |
+
"step": 426
|
| 3000 |
+
},
|
| 3001 |
+
{
|
| 3002 |
+
"epoch": 0.4646354733405876,
|
| 3003 |
+
"grad_norm": 0.1662827581167221,
|
| 3004 |
+
"learning_rate": 0.0001129241134155949,
|
| 3005 |
+
"loss": 0.1053,
|
| 3006 |
+
"step": 427
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 0.4657236126224157,
|
| 3010 |
+
"grad_norm": 0.13029906153678894,
|
| 3011 |
+
"learning_rate": 0.00011258132571978555,
|
| 3012 |
+
"loss": 0.0797,
|
| 3013 |
+
"step": 428
|
| 3014 |
+
},
|
| 3015 |
+
{
|
| 3016 |
+
"epoch": 0.46681175190424373,
|
| 3017 |
+
"grad_norm": 0.18869560956954956,
|
| 3018 |
+
"learning_rate": 0.00011223838774509514,
|
| 3019 |
+
"loss": 0.1292,
|
| 3020 |
+
"step": 429
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 0.46789989118607184,
|
| 3024 |
+
"grad_norm": 0.14279034733772278,
|
| 3025 |
+
"learning_rate": 0.00011189530358778005,
|
| 3026 |
+
"loss": 0.0951,
|
| 3027 |
+
"step": 430
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 0.4689880304678999,
|
| 3031 |
+
"grad_norm": 0.11428643018007278,
|
| 3032 |
+
"learning_rate": 0.00011155207734584263,
|
| 3033 |
+
"loss": 0.0664,
|
| 3034 |
+
"step": 431
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 0.470076169749728,
|
| 3038 |
+
"grad_norm": 0.20008297264575958,
|
| 3039 |
+
"learning_rate": 0.00011120871311898254,
|
| 3040 |
+
"loss": 0.1027,
|
| 3041 |
+
"step": 432
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 0.47116430903155604,
|
| 3045 |
+
"grad_norm": 0.22173888981342316,
|
| 3046 |
+
"learning_rate": 0.00011086521500854745,
|
| 3047 |
+
"loss": 0.1356,
|
| 3048 |
+
"step": 433
|
| 3049 |
+
},
|
| 3050 |
+
{
|
| 3051 |
+
"epoch": 0.4722524483133841,
|
| 3052 |
+
"grad_norm": 0.2382795661687851,
|
| 3053 |
+
"learning_rate": 0.00011052158711748434,
|
| 3054 |
+
"loss": 0.1516,
|
| 3055 |
+
"step": 434
|
| 3056 |
+
},
|
| 3057 |
+
{
|
| 3058 |
+
"epoch": 0.4733405875952122,
|
| 3059 |
+
"grad_norm": 0.2854343354701996,
|
| 3060 |
+
"learning_rate": 0.00011017783355029026,
|
| 3061 |
+
"loss": 0.1116,
|
| 3062 |
+
"step": 435
|
| 3063 |
+
},
|
| 3064 |
+
{
|
| 3065 |
+
"epoch": 0.47442872687704024,
|
| 3066 |
+
"grad_norm": 0.23063793778419495,
|
| 3067 |
+
"learning_rate": 0.00010983395841296348,
|
| 3068 |
+
"loss": 0.107,
|
| 3069 |
+
"step": 436
|
| 3070 |
+
},
|
| 3071 |
+
{
|
| 3072 |
+
"epoch": 0.47551686615886835,
|
| 3073 |
+
"grad_norm": 0.19402769207954407,
|
| 3074 |
+
"learning_rate": 0.00010948996581295436,
|
| 3075 |
+
"loss": 0.0883,
|
| 3076 |
+
"step": 437
|
| 3077 |
+
},
|
| 3078 |
+
{
|
| 3079 |
+
"epoch": 0.4766050054406964,
|
| 3080 |
+
"grad_norm": 0.2664678692817688,
|
| 3081 |
+
"learning_rate": 0.00010914585985911632,
|
| 3082 |
+
"loss": 0.1161,
|
| 3083 |
+
"step": 438
|
| 3084 |
+
},
|
| 3085 |
+
{
|
| 3086 |
+
"epoch": 0.4776931447225245,
|
| 3087 |
+
"grad_norm": 0.29061347246170044,
|
| 3088 |
+
"learning_rate": 0.00010880164466165674,
|
| 3089 |
+
"loss": 0.1833,
|
| 3090 |
+
"step": 439
|
| 3091 |
+
},
|
| 3092 |
+
{
|
| 3093 |
+
"epoch": 0.47878128400435255,
|
| 3094 |
+
"grad_norm": 0.33060985803604126,
|
| 3095 |
+
"learning_rate": 0.00010845732433208779,
|
| 3096 |
+
"loss": 0.1521,
|
| 3097 |
+
"step": 440
|
| 3098 |
+
},
|
| 3099 |
+
{
|
| 3100 |
+
"epoch": 0.47986942328618065,
|
| 3101 |
+
"grad_norm": 0.28285855054855347,
|
| 3102 |
+
"learning_rate": 0.00010811290298317755,
|
| 3103 |
+
"loss": 0.1248,
|
| 3104 |
+
"step": 441
|
| 3105 |
+
},
|
| 3106 |
+
{
|
| 3107 |
+
"epoch": 0.4809575625680087,
|
| 3108 |
+
"grad_norm": 0.49815383553504944,
|
| 3109 |
+
"learning_rate": 0.00010776838472890065,
|
| 3110 |
+
"loss": 0.2238,
|
| 3111 |
+
"step": 442
|
| 3112 |
+
},
|
| 3113 |
+
{
|
| 3114 |
+
"epoch": 0.4820457018498368,
|
| 3115 |
+
"grad_norm": 0.367214173078537,
|
| 3116 |
+
"learning_rate": 0.00010742377368438914,
|
| 3117 |
+
"loss": 0.2344,
|
| 3118 |
+
"step": 443
|
| 3119 |
+
},
|
| 3120 |
+
{
|
| 3121 |
+
"epoch": 0.48313384113166485,
|
| 3122 |
+
"grad_norm": 0.3444245159626007,
|
| 3123 |
+
"learning_rate": 0.00010707907396588361,
|
| 3124 |
+
"loss": 0.1254,
|
| 3125 |
+
"step": 444
|
| 3126 |
+
},
|
| 3127 |
+
{
|
| 3128 |
+
"epoch": 0.4842219804134929,
|
| 3129 |
+
"grad_norm": 0.31096217036247253,
|
| 3130 |
+
"learning_rate": 0.00010673428969068364,
|
| 3131 |
+
"loss": 0.1313,
|
| 3132 |
+
"step": 445
|
| 3133 |
+
},
|
| 3134 |
+
{
|
| 3135 |
+
"epoch": 0.485310119695321,
|
| 3136 |
+
"grad_norm": 0.5377318263053894,
|
| 3137 |
+
"learning_rate": 0.0001063894249770989,
|
| 3138 |
+
"loss": 0.2526,
|
| 3139 |
+
"step": 446
|
| 3140 |
+
},
|
| 3141 |
+
{
|
| 3142 |
+
"epoch": 0.48639825897714906,
|
| 3143 |
+
"grad_norm": 0.4121945798397064,
|
| 3144 |
+
"learning_rate": 0.00010604448394439983,
|
| 3145 |
+
"loss": 0.1663,
|
| 3146 |
+
"step": 447
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 0.48748639825897716,
|
| 3150 |
+
"grad_norm": 0.48366662859916687,
|
| 3151 |
+
"learning_rate": 0.00010569947071276847,
|
| 3152 |
+
"loss": 0.2457,
|
| 3153 |
+
"step": 448
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 0.4885745375408052,
|
| 3157 |
+
"grad_norm": 0.6081061959266663,
|
| 3158 |
+
"learning_rate": 0.0001053543894032493,
|
| 3159 |
+
"loss": 0.2678,
|
| 3160 |
+
"step": 449
|
| 3161 |
+
},
|
| 3162 |
+
{
|
| 3163 |
+
"epoch": 0.4896626768226333,
|
| 3164 |
+
"grad_norm": 1.0334888696670532,
|
| 3165 |
+
"learning_rate": 0.00010500924413769988,
|
| 3166 |
+
"loss": 0.3153,
|
| 3167 |
+
"step": 450
|
| 3168 |
+
},
|
| 3169 |
+
{
|
| 3170 |
+
"epoch": 0.49075081610446136,
|
| 3171 |
+
"grad_norm": 0.0451701320707798,
|
| 3172 |
+
"learning_rate": 0.00010466403903874176,
|
| 3173 |
+
"loss": 0.0455,
|
| 3174 |
+
"step": 451
|
| 3175 |
+
},
|
| 3176 |
+
{
|
| 3177 |
+
"epoch": 0.49183895538628947,
|
| 3178 |
+
"grad_norm": 0.05214075744152069,
|
| 3179 |
+
"learning_rate": 0.00010431877822971117,
|
| 3180 |
+
"loss": 0.0519,
|
| 3181 |
+
"step": 452
|
| 3182 |
+
},
|
| 3183 |
+
{
|
| 3184 |
+
"epoch": 0.4929270946681175,
|
| 3185 |
+
"grad_norm": 0.06553710252046585,
|
| 3186 |
+
"learning_rate": 0.00010397346583460971,
|
| 3187 |
+
"loss": 0.0557,
|
| 3188 |
+
"step": 453
|
| 3189 |
+
},
|
| 3190 |
+
{
|
| 3191 |
+
"epoch": 0.4940152339499456,
|
| 3192 |
+
"grad_norm": 0.06424305588006973,
|
| 3193 |
+
"learning_rate": 0.00010362810597805526,
|
| 3194 |
+
"loss": 0.0672,
|
| 3195 |
+
"step": 454
|
| 3196 |
+
},
|
| 3197 |
+
{
|
| 3198 |
+
"epoch": 0.49510337323177367,
|
| 3199 |
+
"grad_norm": 0.08962132036685944,
|
| 3200 |
+
"learning_rate": 0.00010328270278523256,
|
| 3201 |
+
"loss": 0.0957,
|
| 3202 |
+
"step": 455
|
| 3203 |
+
},
|
| 3204 |
+
{
|
| 3205 |
+
"epoch": 0.4961915125136017,
|
| 3206 |
+
"grad_norm": 0.08780992031097412,
|
| 3207 |
+
"learning_rate": 0.00010293726038184393,
|
| 3208 |
+
"loss": 0.0894,
|
| 3209 |
+
"step": 456
|
| 3210 |
+
},
|
| 3211 |
+
{
|
| 3212 |
+
"epoch": 0.4972796517954298,
|
| 3213 |
+
"grad_norm": 0.08018220961093903,
|
| 3214 |
+
"learning_rate": 0.00010259178289406011,
|
| 3215 |
+
"loss": 0.0663,
|
| 3216 |
+
"step": 457
|
| 3217 |
+
},
|
| 3218 |
+
{
|
| 3219 |
+
"epoch": 0.49836779107725787,
|
| 3220 |
+
"grad_norm": 0.07880765199661255,
|
| 3221 |
+
"learning_rate": 0.0001022462744484709,
|
| 3222 |
+
"loss": 0.0832,
|
| 3223 |
+
"step": 458
|
| 3224 |
+
},
|
| 3225 |
+
{
|
| 3226 |
+
"epoch": 0.499455930359086,
|
| 3227 |
+
"grad_norm": 0.07101229578256607,
|
| 3228 |
+
"learning_rate": 0.00010190073917203589,
|
| 3229 |
+
"loss": 0.054,
|
| 3230 |
+
"step": 459
|
| 3231 |
+
},
|
| 3232 |
+
{
|
| 3233 |
+
"epoch": 0.500544069640914,
|
| 3234 |
+
"grad_norm": 0.07901884615421295,
|
| 3235 |
+
"learning_rate": 0.0001015551811920351,
|
| 3236 |
+
"loss": 0.0611,
|
| 3237 |
+
"step": 460
|
| 3238 |
+
},
|
| 3239 |
+
{
|
| 3240 |
+
"epoch": 0.500544069640914,
|
| 3241 |
+
"eval_loss": 0.11412899941205978,
|
| 3242 |
+
"eval_runtime": 24.4441,
|
| 3243 |
+
"eval_samples_per_second": 15.832,
|
| 3244 |
+
"eval_steps_per_second": 7.936,
|
| 3245 |
+
"step": 460
|
| 3246 |
}
|
| 3247 |
],
|
| 3248 |
"logging_steps": 1,
|
|
|
|
| 3262 |
"attributes": {}
|
| 3263 |
}
|
| 3264 |
},
|
| 3265 |
+
"total_flos": 1.8463228908326093e+17,
|
| 3266 |
"train_batch_size": 2,
|
| 3267 |
"trial_name": null,
|
| 3268 |
"trial_params": null
|