Training in progress, step 800, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2641039512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:701ca595ad1defd750a25409c943adf74f8781010426bdefa517964d000f7cb5
|
| 3 |
size 2641039512
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1322856248
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1aa61026d06844cb3156b00475a837e064d2f608ee32e985c65d38002d7cb243
|
| 3 |
size 1322856248
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 9785
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e1ba5e8415b30095e686cd73b6abdd2edac02ddfda122507f638f82ba2ed948a
|
| 3 |
size 9785
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1408,6 +1408,1406 @@
|
|
| 1408 |
"learning_rate": 0.0012,
|
| 1409 |
"loss": 2.2841665744781494,
|
| 1410 |
"step": 400
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1411 |
}
|
| 1412 |
],
|
| 1413 |
"logging_steps": 2,
|
|
@@ -1427,7 +2827,7 @@
|
|
| 1427 |
"attributes": {}
|
| 1428 |
}
|
| 1429 |
},
|
| 1430 |
-
"total_flos":
|
| 1431 |
"train_batch_size": 2,
|
| 1432 |
"trial_name": null,
|
| 1433 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.2,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 800,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1408 |
"learning_rate": 0.0012,
|
| 1409 |
"loss": 2.2841665744781494,
|
| 1410 |
"step": 400
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.1005,
|
| 1414 |
+
"grad_norm": 0.12054895609617233,
|
| 1415 |
+
"learning_rate": 0.0012,
|
| 1416 |
+
"loss": 2.326702356338501,
|
| 1417 |
+
"step": 402
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.101,
|
| 1421 |
+
"grad_norm": 0.12209802120923996,
|
| 1422 |
+
"learning_rate": 0.0012,
|
| 1423 |
+
"loss": 2.2534260749816895,
|
| 1424 |
+
"step": 404
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.1015,
|
| 1428 |
+
"grad_norm": 0.12105856090784073,
|
| 1429 |
+
"learning_rate": 0.0012,
|
| 1430 |
+
"loss": 2.2094016075134277,
|
| 1431 |
+
"step": 406
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.102,
|
| 1435 |
+
"grad_norm": 0.12396562099456787,
|
| 1436 |
+
"learning_rate": 0.0012,
|
| 1437 |
+
"loss": 2.3064889907836914,
|
| 1438 |
+
"step": 408
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 0.1025,
|
| 1442 |
+
"grad_norm": 0.12443626672029495,
|
| 1443 |
+
"learning_rate": 0.0012,
|
| 1444 |
+
"loss": 2.2636513710021973,
|
| 1445 |
+
"step": 410
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.103,
|
| 1449 |
+
"grad_norm": 0.12337794899940491,
|
| 1450 |
+
"learning_rate": 0.0012,
|
| 1451 |
+
"loss": 2.2778162956237793,
|
| 1452 |
+
"step": 412
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.1035,
|
| 1456 |
+
"grad_norm": 0.12057880312204361,
|
| 1457 |
+
"learning_rate": 0.0012,
|
| 1458 |
+
"loss": 2.0912723541259766,
|
| 1459 |
+
"step": 414
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.104,
|
| 1463 |
+
"grad_norm": 0.1243756115436554,
|
| 1464 |
+
"learning_rate": 0.0012,
|
| 1465 |
+
"loss": 2.2777326107025146,
|
| 1466 |
+
"step": 416
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.1045,
|
| 1470 |
+
"grad_norm": 0.12009448558092117,
|
| 1471 |
+
"learning_rate": 0.0012,
|
| 1472 |
+
"loss": 2.137455940246582,
|
| 1473 |
+
"step": 418
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.105,
|
| 1477 |
+
"grad_norm": 0.12110329419374466,
|
| 1478 |
+
"learning_rate": 0.0012,
|
| 1479 |
+
"loss": 2.1315882205963135,
|
| 1480 |
+
"step": 420
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.1055,
|
| 1484 |
+
"grad_norm": 0.12125954777002335,
|
| 1485 |
+
"learning_rate": 0.0012,
|
| 1486 |
+
"loss": 2.2602455615997314,
|
| 1487 |
+
"step": 422
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.106,
|
| 1491 |
+
"grad_norm": 0.12352994829416275,
|
| 1492 |
+
"learning_rate": 0.0012,
|
| 1493 |
+
"loss": 2.31404447555542,
|
| 1494 |
+
"step": 424
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.1065,
|
| 1498 |
+
"grad_norm": 0.12671133875846863,
|
| 1499 |
+
"learning_rate": 0.0012,
|
| 1500 |
+
"loss": 2.256730556488037,
|
| 1501 |
+
"step": 426
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 0.107,
|
| 1505 |
+
"grad_norm": 0.12330864369869232,
|
| 1506 |
+
"learning_rate": 0.0012,
|
| 1507 |
+
"loss": 2.2596917152404785,
|
| 1508 |
+
"step": 428
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.1075,
|
| 1512 |
+
"grad_norm": 0.12586729228496552,
|
| 1513 |
+
"learning_rate": 0.0012,
|
| 1514 |
+
"loss": 2.2614588737487793,
|
| 1515 |
+
"step": 430
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.108,
|
| 1519 |
+
"grad_norm": 0.12044867873191833,
|
| 1520 |
+
"learning_rate": 0.0012,
|
| 1521 |
+
"loss": 2.2217602729797363,
|
| 1522 |
+
"step": 432
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.1085,
|
| 1526 |
+
"grad_norm": 0.12189746648073196,
|
| 1527 |
+
"learning_rate": 0.0012,
|
| 1528 |
+
"loss": 2.298125982284546,
|
| 1529 |
+
"step": 434
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.109,
|
| 1533 |
+
"grad_norm": 0.11962965875864029,
|
| 1534 |
+
"learning_rate": 0.0012,
|
| 1535 |
+
"loss": 2.210402011871338,
|
| 1536 |
+
"step": 436
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.1095,
|
| 1540 |
+
"grad_norm": 0.12688636779785156,
|
| 1541 |
+
"learning_rate": 0.0012,
|
| 1542 |
+
"loss": 2.151930809020996,
|
| 1543 |
+
"step": 438
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 0.11,
|
| 1547 |
+
"grad_norm": 0.12192431092262268,
|
| 1548 |
+
"learning_rate": 0.0012,
|
| 1549 |
+
"loss": 2.2577176094055176,
|
| 1550 |
+
"step": 440
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.1105,
|
| 1554 |
+
"grad_norm": 0.11881116032600403,
|
| 1555 |
+
"learning_rate": 0.0012,
|
| 1556 |
+
"loss": 2.1138787269592285,
|
| 1557 |
+
"step": 442
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.111,
|
| 1561 |
+
"grad_norm": 0.12109513580799103,
|
| 1562 |
+
"learning_rate": 0.0012,
|
| 1563 |
+
"loss": 2.1739344596862793,
|
| 1564 |
+
"step": 444
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 0.1115,
|
| 1568 |
+
"grad_norm": 0.11925635486841202,
|
| 1569 |
+
"learning_rate": 0.0012,
|
| 1570 |
+
"loss": 2.2241151332855225,
|
| 1571 |
+
"step": 446
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.112,
|
| 1575 |
+
"grad_norm": 0.12064164876937866,
|
| 1576 |
+
"learning_rate": 0.0012,
|
| 1577 |
+
"loss": 2.214655876159668,
|
| 1578 |
+
"step": 448
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.1125,
|
| 1582 |
+
"grad_norm": 0.12177804112434387,
|
| 1583 |
+
"learning_rate": 0.0012,
|
| 1584 |
+
"loss": 2.304344654083252,
|
| 1585 |
+
"step": 450
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.113,
|
| 1589 |
+
"grad_norm": 0.12058129906654358,
|
| 1590 |
+
"learning_rate": 0.0012,
|
| 1591 |
+
"loss": 2.1164498329162598,
|
| 1592 |
+
"step": 452
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.1135,
|
| 1596 |
+
"grad_norm": 0.1220935806632042,
|
| 1597 |
+
"learning_rate": 0.0012,
|
| 1598 |
+
"loss": 2.2567191123962402,
|
| 1599 |
+
"step": 454
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.114,
|
| 1603 |
+
"grad_norm": 0.12614116072654724,
|
| 1604 |
+
"learning_rate": 0.0012,
|
| 1605 |
+
"loss": 2.3177077770233154,
|
| 1606 |
+
"step": 456
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 0.1145,
|
| 1610 |
+
"grad_norm": 0.12056887149810791,
|
| 1611 |
+
"learning_rate": 0.0012,
|
| 1612 |
+
"loss": 2.1604442596435547,
|
| 1613 |
+
"step": 458
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 0.115,
|
| 1617 |
+
"grad_norm": 0.1201133131980896,
|
| 1618 |
+
"learning_rate": 0.0012,
|
| 1619 |
+
"loss": 2.251906394958496,
|
| 1620 |
+
"step": 460
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 0.1155,
|
| 1624 |
+
"grad_norm": 0.12387340515851974,
|
| 1625 |
+
"learning_rate": 0.0012,
|
| 1626 |
+
"loss": 2.2524404525756836,
|
| 1627 |
+
"step": 462
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 0.116,
|
| 1631 |
+
"grad_norm": 0.12445537745952606,
|
| 1632 |
+
"learning_rate": 0.0012,
|
| 1633 |
+
"loss": 2.173084020614624,
|
| 1634 |
+
"step": 464
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.1165,
|
| 1638 |
+
"grad_norm": 0.12160599231719971,
|
| 1639 |
+
"learning_rate": 0.0012,
|
| 1640 |
+
"loss": 2.253653049468994,
|
| 1641 |
+
"step": 466
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.117,
|
| 1645 |
+
"grad_norm": 0.1241399273276329,
|
| 1646 |
+
"learning_rate": 0.0012,
|
| 1647 |
+
"loss": 2.1665000915527344,
|
| 1648 |
+
"step": 468
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.1175,
|
| 1652 |
+
"grad_norm": 0.1183992549777031,
|
| 1653 |
+
"learning_rate": 0.0012,
|
| 1654 |
+
"loss": 2.050638437271118,
|
| 1655 |
+
"step": 470
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.118,
|
| 1659 |
+
"grad_norm": 0.12072386592626572,
|
| 1660 |
+
"learning_rate": 0.0012,
|
| 1661 |
+
"loss": 2.1509904861450195,
|
| 1662 |
+
"step": 472
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.1185,
|
| 1666 |
+
"grad_norm": 0.12380450963973999,
|
| 1667 |
+
"learning_rate": 0.0012,
|
| 1668 |
+
"loss": 2.3313965797424316,
|
| 1669 |
+
"step": 474
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.119,
|
| 1673 |
+
"grad_norm": 0.12388438731431961,
|
| 1674 |
+
"learning_rate": 0.0012,
|
| 1675 |
+
"loss": 2.110642671585083,
|
| 1676 |
+
"step": 476
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.1195,
|
| 1680 |
+
"grad_norm": 0.123001329600811,
|
| 1681 |
+
"learning_rate": 0.0012,
|
| 1682 |
+
"loss": 2.225430965423584,
|
| 1683 |
+
"step": 478
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.12,
|
| 1687 |
+
"grad_norm": 0.12334371358156204,
|
| 1688 |
+
"learning_rate": 0.0012,
|
| 1689 |
+
"loss": 2.202939033508301,
|
| 1690 |
+
"step": 480
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.1205,
|
| 1694 |
+
"grad_norm": 0.11948074400424957,
|
| 1695 |
+
"learning_rate": 0.0012,
|
| 1696 |
+
"loss": 2.1103436946868896,
|
| 1697 |
+
"step": 482
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.121,
|
| 1701 |
+
"grad_norm": 0.1230388805270195,
|
| 1702 |
+
"learning_rate": 0.0012,
|
| 1703 |
+
"loss": 2.1396327018737793,
|
| 1704 |
+
"step": 484
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.1215,
|
| 1708 |
+
"grad_norm": 0.11782898753881454,
|
| 1709 |
+
"learning_rate": 0.0012,
|
| 1710 |
+
"loss": 2.1475939750671387,
|
| 1711 |
+
"step": 486
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.122,
|
| 1715 |
+
"grad_norm": 0.12395215034484863,
|
| 1716 |
+
"learning_rate": 0.0012,
|
| 1717 |
+
"loss": 2.3515262603759766,
|
| 1718 |
+
"step": 488
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.1225,
|
| 1722 |
+
"grad_norm": 0.12264861911535263,
|
| 1723 |
+
"learning_rate": 0.0012,
|
| 1724 |
+
"loss": 2.3518590927124023,
|
| 1725 |
+
"step": 490
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.123,
|
| 1729 |
+
"grad_norm": 0.12457684427499771,
|
| 1730 |
+
"learning_rate": 0.0012,
|
| 1731 |
+
"loss": 2.348440647125244,
|
| 1732 |
+
"step": 492
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.1235,
|
| 1736 |
+
"grad_norm": 0.12033570557832718,
|
| 1737 |
+
"learning_rate": 0.0012,
|
| 1738 |
+
"loss": 2.2577500343322754,
|
| 1739 |
+
"step": 494
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.124,
|
| 1743 |
+
"grad_norm": 0.12147720903158188,
|
| 1744 |
+
"learning_rate": 0.0012,
|
| 1745 |
+
"loss": 2.2284274101257324,
|
| 1746 |
+
"step": 496
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.1245,
|
| 1750 |
+
"grad_norm": 0.1266353577375412,
|
| 1751 |
+
"learning_rate": 0.0012,
|
| 1752 |
+
"loss": 2.2095959186553955,
|
| 1753 |
+
"step": 498
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.125,
|
| 1757 |
+
"grad_norm": 0.12260833382606506,
|
| 1758 |
+
"learning_rate": 0.0012,
|
| 1759 |
+
"loss": 2.2494609355926514,
|
| 1760 |
+
"step": 500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.1255,
|
| 1764 |
+
"grad_norm": 0.12166132777929306,
|
| 1765 |
+
"learning_rate": 0.0012,
|
| 1766 |
+
"loss": 2.227855682373047,
|
| 1767 |
+
"step": 502
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.126,
|
| 1771 |
+
"grad_norm": 0.12374623119831085,
|
| 1772 |
+
"learning_rate": 0.0012,
|
| 1773 |
+
"loss": 2.2455830574035645,
|
| 1774 |
+
"step": 504
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.1265,
|
| 1778 |
+
"grad_norm": 0.12395448237657547,
|
| 1779 |
+
"learning_rate": 0.0012,
|
| 1780 |
+
"loss": 2.155578136444092,
|
| 1781 |
+
"step": 506
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.127,
|
| 1785 |
+
"grad_norm": 0.12223073840141296,
|
| 1786 |
+
"learning_rate": 0.0012,
|
| 1787 |
+
"loss": 2.123234272003174,
|
| 1788 |
+
"step": 508
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.1275,
|
| 1792 |
+
"grad_norm": 0.12525120377540588,
|
| 1793 |
+
"learning_rate": 0.0012,
|
| 1794 |
+
"loss": 2.2657387256622314,
|
| 1795 |
+
"step": 510
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.128,
|
| 1799 |
+
"grad_norm": 0.12459076941013336,
|
| 1800 |
+
"learning_rate": 0.0012,
|
| 1801 |
+
"loss": 2.2407002449035645,
|
| 1802 |
+
"step": 512
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 0.1285,
|
| 1806 |
+
"grad_norm": 0.12786155939102173,
|
| 1807 |
+
"learning_rate": 0.0012,
|
| 1808 |
+
"loss": 2.2713661193847656,
|
| 1809 |
+
"step": 514
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 0.129,
|
| 1813 |
+
"grad_norm": 0.12424921989440918,
|
| 1814 |
+
"learning_rate": 0.0012,
|
| 1815 |
+
"loss": 2.277902603149414,
|
| 1816 |
+
"step": 516
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 0.1295,
|
| 1820 |
+
"grad_norm": 0.12083666026592255,
|
| 1821 |
+
"learning_rate": 0.0012,
|
| 1822 |
+
"loss": 2.229116201400757,
|
| 1823 |
+
"step": 518
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 0.13,
|
| 1827 |
+
"grad_norm": 0.12301281839609146,
|
| 1828 |
+
"learning_rate": 0.0012,
|
| 1829 |
+
"loss": 2.2580509185791016,
|
| 1830 |
+
"step": 520
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 0.1305,
|
| 1834 |
+
"grad_norm": 0.12253593653440475,
|
| 1835 |
+
"learning_rate": 0.0012,
|
| 1836 |
+
"loss": 2.060866355895996,
|
| 1837 |
+
"step": 522
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 0.131,
|
| 1841 |
+
"grad_norm": 0.12206588685512543,
|
| 1842 |
+
"learning_rate": 0.0012,
|
| 1843 |
+
"loss": 2.190617561340332,
|
| 1844 |
+
"step": 524
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 0.1315,
|
| 1848 |
+
"grad_norm": 0.11996202170848846,
|
| 1849 |
+
"learning_rate": 0.0012,
|
| 1850 |
+
"loss": 2.1966025829315186,
|
| 1851 |
+
"step": 526
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 0.132,
|
| 1855 |
+
"grad_norm": 0.12044856697320938,
|
| 1856 |
+
"learning_rate": 0.0012,
|
| 1857 |
+
"loss": 2.1235170364379883,
|
| 1858 |
+
"step": 528
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 0.1325,
|
| 1862 |
+
"grad_norm": 0.12207049876451492,
|
| 1863 |
+
"learning_rate": 0.0012,
|
| 1864 |
+
"loss": 2.2631895542144775,
|
| 1865 |
+
"step": 530
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 0.133,
|
| 1869 |
+
"grad_norm": 0.12151215225458145,
|
| 1870 |
+
"learning_rate": 0.0012,
|
| 1871 |
+
"loss": 2.146211624145508,
|
| 1872 |
+
"step": 532
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 0.1335,
|
| 1876 |
+
"grad_norm": 0.12725605070590973,
|
| 1877 |
+
"learning_rate": 0.0012,
|
| 1878 |
+
"loss": 2.3359932899475098,
|
| 1879 |
+
"step": 534
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 0.134,
|
| 1883 |
+
"grad_norm": 0.12424077838659286,
|
| 1884 |
+
"learning_rate": 0.0012,
|
| 1885 |
+
"loss": 2.205270528793335,
|
| 1886 |
+
"step": 536
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 0.1345,
|
| 1890 |
+
"grad_norm": 0.12291470915079117,
|
| 1891 |
+
"learning_rate": 0.0012,
|
| 1892 |
+
"loss": 2.1694424152374268,
|
| 1893 |
+
"step": 538
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 0.135,
|
| 1897 |
+
"grad_norm": 0.12399540841579437,
|
| 1898 |
+
"learning_rate": 0.0012,
|
| 1899 |
+
"loss": 2.1751201152801514,
|
| 1900 |
+
"step": 540
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 0.1355,
|
| 1904 |
+
"grad_norm": 0.12478026747703552,
|
| 1905 |
+
"learning_rate": 0.0012,
|
| 1906 |
+
"loss": 2.2821462154388428,
|
| 1907 |
+
"step": 542
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 0.136,
|
| 1911 |
+
"grad_norm": 0.12286561727523804,
|
| 1912 |
+
"learning_rate": 0.0012,
|
| 1913 |
+
"loss": 2.240694761276245,
|
| 1914 |
+
"step": 544
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 0.1365,
|
| 1918 |
+
"grad_norm": 0.11715179681777954,
|
| 1919 |
+
"learning_rate": 0.0012,
|
| 1920 |
+
"loss": 2.1502816677093506,
|
| 1921 |
+
"step": 546
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 0.137,
|
| 1925 |
+
"grad_norm": 0.1263730376958847,
|
| 1926 |
+
"learning_rate": 0.0012,
|
| 1927 |
+
"loss": 2.2521462440490723,
|
| 1928 |
+
"step": 548
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 0.1375,
|
| 1932 |
+
"grad_norm": 0.12617576122283936,
|
| 1933 |
+
"learning_rate": 0.0012,
|
| 1934 |
+
"loss": 2.1075775623321533,
|
| 1935 |
+
"step": 550
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 0.138,
|
| 1939 |
+
"grad_norm": 0.12463200837373734,
|
| 1940 |
+
"learning_rate": 0.0012,
|
| 1941 |
+
"loss": 2.2123312950134277,
|
| 1942 |
+
"step": 552
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 0.1385,
|
| 1946 |
+
"grad_norm": 0.12332479655742645,
|
| 1947 |
+
"learning_rate": 0.0012,
|
| 1948 |
+
"loss": 2.1256752014160156,
|
| 1949 |
+
"step": 554
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 0.139,
|
| 1953 |
+
"grad_norm": 0.12213516980409622,
|
| 1954 |
+
"learning_rate": 0.0012,
|
| 1955 |
+
"loss": 2.230741024017334,
|
| 1956 |
+
"step": 556
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 0.1395,
|
| 1960 |
+
"grad_norm": 0.12564736604690552,
|
| 1961 |
+
"learning_rate": 0.0012,
|
| 1962 |
+
"loss": 2.1856470108032227,
|
| 1963 |
+
"step": 558
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 0.14,
|
| 1967 |
+
"grad_norm": 0.1247808113694191,
|
| 1968 |
+
"learning_rate": 0.0012,
|
| 1969 |
+
"loss": 2.29241943359375,
|
| 1970 |
+
"step": 560
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 0.1405,
|
| 1974 |
+
"grad_norm": 0.12066401541233063,
|
| 1975 |
+
"learning_rate": 0.0012,
|
| 1976 |
+
"loss": 2.175884485244751,
|
| 1977 |
+
"step": 562
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 0.141,
|
| 1981 |
+
"grad_norm": 0.12430638819932938,
|
| 1982 |
+
"learning_rate": 0.0012,
|
| 1983 |
+
"loss": 2.3148648738861084,
|
| 1984 |
+
"step": 564
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 0.1415,
|
| 1988 |
+
"grad_norm": 0.12408965826034546,
|
| 1989 |
+
"learning_rate": 0.0012,
|
| 1990 |
+
"loss": 2.2461142539978027,
|
| 1991 |
+
"step": 566
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 0.142,
|
| 1995 |
+
"grad_norm": 0.1242135688662529,
|
| 1996 |
+
"learning_rate": 0.0012,
|
| 1997 |
+
"loss": 2.1376333236694336,
|
| 1998 |
+
"step": 568
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 0.1425,
|
| 2002 |
+
"grad_norm": 0.12672850489616394,
|
| 2003 |
+
"learning_rate": 0.0012,
|
| 2004 |
+
"loss": 2.148775100708008,
|
| 2005 |
+
"step": 570
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 0.143,
|
| 2009 |
+
"grad_norm": 0.12376829236745834,
|
| 2010 |
+
"learning_rate": 0.0012,
|
| 2011 |
+
"loss": 2.2954864501953125,
|
| 2012 |
+
"step": 572
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 0.1435,
|
| 2016 |
+
"grad_norm": 0.12005943804979324,
|
| 2017 |
+
"learning_rate": 0.0012,
|
| 2018 |
+
"loss": 2.115406036376953,
|
| 2019 |
+
"step": 574
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.144,
|
| 2023 |
+
"grad_norm": 0.1220192164182663,
|
| 2024 |
+
"learning_rate": 0.0012,
|
| 2025 |
+
"loss": 2.1036298274993896,
|
| 2026 |
+
"step": 576
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 0.1445,
|
| 2030 |
+
"grad_norm": 0.12172229588031769,
|
| 2031 |
+
"learning_rate": 0.0012,
|
| 2032 |
+
"loss": 2.1697356700897217,
|
| 2033 |
+
"step": 578
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 0.145,
|
| 2037 |
+
"grad_norm": 0.12480010837316513,
|
| 2038 |
+
"learning_rate": 0.0012,
|
| 2039 |
+
"loss": 2.2564525604248047,
|
| 2040 |
+
"step": 580
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 0.1455,
|
| 2044 |
+
"grad_norm": 0.12419505417346954,
|
| 2045 |
+
"learning_rate": 0.0012,
|
| 2046 |
+
"loss": 2.242037773132324,
|
| 2047 |
+
"step": 582
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 0.146,
|
| 2051 |
+
"grad_norm": 0.12008476257324219,
|
| 2052 |
+
"learning_rate": 0.0012,
|
| 2053 |
+
"loss": 2.149271011352539,
|
| 2054 |
+
"step": 584
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 0.1465,
|
| 2058 |
+
"grad_norm": 0.12122812122106552,
|
| 2059 |
+
"learning_rate": 0.0012,
|
| 2060 |
+
"loss": 2.2153196334838867,
|
| 2061 |
+
"step": 586
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 0.147,
|
| 2065 |
+
"grad_norm": 0.12544192373752594,
|
| 2066 |
+
"learning_rate": 0.0012,
|
| 2067 |
+
"loss": 2.3317863941192627,
|
| 2068 |
+
"step": 588
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 0.1475,
|
| 2072 |
+
"grad_norm": 0.1265304833650589,
|
| 2073 |
+
"learning_rate": 0.0012,
|
| 2074 |
+
"loss": 2.216907501220703,
|
| 2075 |
+
"step": 590
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 0.148,
|
| 2079 |
+
"grad_norm": 0.12547633051872253,
|
| 2080 |
+
"learning_rate": 0.0012,
|
| 2081 |
+
"loss": 2.1674623489379883,
|
| 2082 |
+
"step": 592
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 0.1485,
|
| 2086 |
+
"grad_norm": 0.1255500763654709,
|
| 2087 |
+
"learning_rate": 0.0012,
|
| 2088 |
+
"loss": 2.258897542953491,
|
| 2089 |
+
"step": 594
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 0.149,
|
| 2093 |
+
"grad_norm": 0.12620480358600616,
|
| 2094 |
+
"learning_rate": 0.0012,
|
| 2095 |
+
"loss": 2.1567251682281494,
|
| 2096 |
+
"step": 596
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.1495,
|
| 2100 |
+
"grad_norm": 0.12494870275259018,
|
| 2101 |
+
"learning_rate": 0.0012,
|
| 2102 |
+
"loss": 2.153189182281494,
|
| 2103 |
+
"step": 598
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 0.15,
|
| 2107 |
+
"grad_norm": 0.12307412177324295,
|
| 2108 |
+
"learning_rate": 0.0012,
|
| 2109 |
+
"loss": 2.1250929832458496,
|
| 2110 |
+
"step": 600
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 0.1505,
|
| 2114 |
+
"grad_norm": 0.12381011992692947,
|
| 2115 |
+
"learning_rate": 0.0012,
|
| 2116 |
+
"loss": 2.176429033279419,
|
| 2117 |
+
"step": 602
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 0.151,
|
| 2121 |
+
"grad_norm": 0.12082013487815857,
|
| 2122 |
+
"learning_rate": 0.0012,
|
| 2123 |
+
"loss": 2.184323787689209,
|
| 2124 |
+
"step": 604
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 0.1515,
|
| 2128 |
+
"grad_norm": 0.1267080008983612,
|
| 2129 |
+
"learning_rate": 0.0012,
|
| 2130 |
+
"loss": 2.1645710468292236,
|
| 2131 |
+
"step": 606
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 0.152,
|
| 2135 |
+
"grad_norm": 0.12078283727169037,
|
| 2136 |
+
"learning_rate": 0.0012,
|
| 2137 |
+
"loss": 2.2049546241760254,
|
| 2138 |
+
"step": 608
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 0.1525,
|
| 2142 |
+
"grad_norm": 0.12241696566343307,
|
| 2143 |
+
"learning_rate": 0.0012,
|
| 2144 |
+
"loss": 2.224576711654663,
|
| 2145 |
+
"step": 610
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 0.153,
|
| 2149 |
+
"grad_norm": 0.12433452904224396,
|
| 2150 |
+
"learning_rate": 0.0012,
|
| 2151 |
+
"loss": 2.225877523422241,
|
| 2152 |
+
"step": 612
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 0.1535,
|
| 2156 |
+
"grad_norm": 0.12083589285612106,
|
| 2157 |
+
"learning_rate": 0.0012,
|
| 2158 |
+
"loss": 2.0981287956237793,
|
| 2159 |
+
"step": 614
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 0.154,
|
| 2163 |
+
"grad_norm": 0.12109652161598206,
|
| 2164 |
+
"learning_rate": 0.0012,
|
| 2165 |
+
"loss": 2.0764999389648438,
|
| 2166 |
+
"step": 616
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 0.1545,
|
| 2170 |
+
"grad_norm": 0.12194561958312988,
|
| 2171 |
+
"learning_rate": 0.0012,
|
| 2172 |
+
"loss": 2.218468189239502,
|
| 2173 |
+
"step": 618
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 0.155,
|
| 2177 |
+
"grad_norm": 0.11935839802026749,
|
| 2178 |
+
"learning_rate": 0.0012,
|
| 2179 |
+
"loss": 2.1261422634124756,
|
| 2180 |
+
"step": 620
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 0.1555,
|
| 2184 |
+
"grad_norm": 0.1271117776632309,
|
| 2185 |
+
"learning_rate": 0.0012,
|
| 2186 |
+
"loss": 2.272076368331909,
|
| 2187 |
+
"step": 622
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 0.156,
|
| 2191 |
+
"grad_norm": 0.12194891273975372,
|
| 2192 |
+
"learning_rate": 0.0012,
|
| 2193 |
+
"loss": 2.199063777923584,
|
| 2194 |
+
"step": 624
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 0.1565,
|
| 2198 |
+
"grad_norm": 0.12229058146476746,
|
| 2199 |
+
"learning_rate": 0.0012,
|
| 2200 |
+
"loss": 2.124752998352051,
|
| 2201 |
+
"step": 626
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.157,
|
| 2205 |
+
"grad_norm": 0.12368135899305344,
|
| 2206 |
+
"learning_rate": 0.0012,
|
| 2207 |
+
"loss": 2.091749429702759,
|
| 2208 |
+
"step": 628
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.1575,
|
| 2212 |
+
"grad_norm": 0.12349650263786316,
|
| 2213 |
+
"learning_rate": 0.0012,
|
| 2214 |
+
"loss": 2.000277519226074,
|
| 2215 |
+
"step": 630
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.158,
|
| 2219 |
+
"grad_norm": 0.11999969929456711,
|
| 2220 |
+
"learning_rate": 0.0012,
|
| 2221 |
+
"loss": 2.1394622325897217,
|
| 2222 |
+
"step": 632
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.1585,
|
| 2226 |
+
"grad_norm": 0.12507598102092743,
|
| 2227 |
+
"learning_rate": 0.0012,
|
| 2228 |
+
"loss": 2.260275363922119,
|
| 2229 |
+
"step": 634
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.159,
|
| 2233 |
+
"grad_norm": 0.12678316235542297,
|
| 2234 |
+
"learning_rate": 0.0012,
|
| 2235 |
+
"loss": 2.2557196617126465,
|
| 2236 |
+
"step": 636
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 0.1595,
|
| 2240 |
+
"grad_norm": 0.11630699783563614,
|
| 2241 |
+
"learning_rate": 0.0012,
|
| 2242 |
+
"loss": 1.966517686843872,
|
| 2243 |
+
"step": 638
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.16,
|
| 2247 |
+
"grad_norm": 0.12710458040237427,
|
| 2248 |
+
"learning_rate": 0.0012,
|
| 2249 |
+
"loss": 2.217278480529785,
|
| 2250 |
+
"step": 640
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 0.1605,
|
| 2254 |
+
"grad_norm": 0.12581312656402588,
|
| 2255 |
+
"learning_rate": 0.0012,
|
| 2256 |
+
"loss": 2.2156848907470703,
|
| 2257 |
+
"step": 642
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.161,
|
| 2261 |
+
"grad_norm": 0.12341652810573578,
|
| 2262 |
+
"learning_rate": 0.0012,
|
| 2263 |
+
"loss": 2.111685276031494,
|
| 2264 |
+
"step": 644
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 0.1615,
|
| 2268 |
+
"grad_norm": 0.12299773842096329,
|
| 2269 |
+
"learning_rate": 0.0012,
|
| 2270 |
+
"loss": 2.181852340698242,
|
| 2271 |
+
"step": 646
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.162,
|
| 2275 |
+
"grad_norm": 0.1208917647600174,
|
| 2276 |
+
"learning_rate": 0.0012,
|
| 2277 |
+
"loss": 2.160839796066284,
|
| 2278 |
+
"step": 648
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 0.1625,
|
| 2282 |
+
"grad_norm": 0.1254134625196457,
|
| 2283 |
+
"learning_rate": 0.0012,
|
| 2284 |
+
"loss": 2.18388032913208,
|
| 2285 |
+
"step": 650
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 0.163,
|
| 2289 |
+
"grad_norm": 0.12103099375963211,
|
| 2290 |
+
"learning_rate": 0.0012,
|
| 2291 |
+
"loss": 2.1160151958465576,
|
| 2292 |
+
"step": 652
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 0.1635,
|
| 2296 |
+
"grad_norm": 0.12528404593467712,
|
| 2297 |
+
"learning_rate": 0.0012,
|
| 2298 |
+
"loss": 2.2673239707946777,
|
| 2299 |
+
"step": 654
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.164,
|
| 2303 |
+
"grad_norm": 0.12464321404695511,
|
| 2304 |
+
"learning_rate": 0.0012,
|
| 2305 |
+
"loss": 2.1309168338775635,
|
| 2306 |
+
"step": 656
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.1645,
|
| 2310 |
+
"grad_norm": 0.12006670981645584,
|
| 2311 |
+
"learning_rate": 0.0012,
|
| 2312 |
+
"loss": 2.0753886699676514,
|
| 2313 |
+
"step": 658
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.165,
|
| 2317 |
+
"grad_norm": 0.122077077627182,
|
| 2318 |
+
"learning_rate": 0.0012,
|
| 2319 |
+
"loss": 2.0909481048583984,
|
| 2320 |
+
"step": 660
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 0.1655,
|
| 2324 |
+
"grad_norm": 0.12410063296556473,
|
| 2325 |
+
"learning_rate": 0.0012,
|
| 2326 |
+
"loss": 2.2268733978271484,
|
| 2327 |
+
"step": 662
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 0.166,
|
| 2331 |
+
"grad_norm": 0.12509913742542267,
|
| 2332 |
+
"learning_rate": 0.0012,
|
| 2333 |
+
"loss": 2.1857738494873047,
|
| 2334 |
+
"step": 664
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 0.1665,
|
| 2338 |
+
"grad_norm": 0.12583136558532715,
|
| 2339 |
+
"learning_rate": 0.0012,
|
| 2340 |
+
"loss": 2.217729091644287,
|
| 2341 |
+
"step": 666
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.167,
|
| 2345 |
+
"grad_norm": 0.12470363080501556,
|
| 2346 |
+
"learning_rate": 0.0012,
|
| 2347 |
+
"loss": 2.1618192195892334,
|
| 2348 |
+
"step": 668
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 0.1675,
|
| 2352 |
+
"grad_norm": 0.12461432069540024,
|
| 2353 |
+
"learning_rate": 0.0012,
|
| 2354 |
+
"loss": 2.2985177040100098,
|
| 2355 |
+
"step": 670
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 0.168,
|
| 2359 |
+
"grad_norm": 0.12380319833755493,
|
| 2360 |
+
"learning_rate": 0.0012,
|
| 2361 |
+
"loss": 2.209359645843506,
|
| 2362 |
+
"step": 672
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 0.1685,
|
| 2366 |
+
"grad_norm": 0.12652839720249176,
|
| 2367 |
+
"learning_rate": 0.0012,
|
| 2368 |
+
"loss": 2.1950721740722656,
|
| 2369 |
+
"step": 674
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 0.169,
|
| 2373 |
+
"grad_norm": 0.12219655513763428,
|
| 2374 |
+
"learning_rate": 0.0012,
|
| 2375 |
+
"loss": 2.042815923690796,
|
| 2376 |
+
"step": 676
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 0.1695,
|
| 2380 |
+
"grad_norm": 0.12653478980064392,
|
| 2381 |
+
"learning_rate": 0.0012,
|
| 2382 |
+
"loss": 2.242166042327881,
|
| 2383 |
+
"step": 678
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 0.17,
|
| 2387 |
+
"grad_norm": 0.12441577017307281,
|
| 2388 |
+
"learning_rate": 0.0012,
|
| 2389 |
+
"loss": 2.1746740341186523,
|
| 2390 |
+
"step": 680
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 0.1705,
|
| 2394 |
+
"grad_norm": 0.12125367671251297,
|
| 2395 |
+
"learning_rate": 0.0012,
|
| 2396 |
+
"loss": 2.0081567764282227,
|
| 2397 |
+
"step": 682
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 0.171,
|
| 2401 |
+
"grad_norm": 0.12602663040161133,
|
| 2402 |
+
"learning_rate": 0.0012,
|
| 2403 |
+
"loss": 2.3285164833068848,
|
| 2404 |
+
"step": 684
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 0.1715,
|
| 2408 |
+
"grad_norm": 0.12515240907669067,
|
| 2409 |
+
"learning_rate": 0.0012,
|
| 2410 |
+
"loss": 2.0954160690307617,
|
| 2411 |
+
"step": 686
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 0.172,
|
| 2415 |
+
"grad_norm": 0.12215415388345718,
|
| 2416 |
+
"learning_rate": 0.0012,
|
| 2417 |
+
"loss": 2.1765799522399902,
|
| 2418 |
+
"step": 688
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 0.1725,
|
| 2422 |
+
"grad_norm": 0.12254252284765244,
|
| 2423 |
+
"learning_rate": 0.0012,
|
| 2424 |
+
"loss": 2.078477144241333,
|
| 2425 |
+
"step": 690
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 0.173,
|
| 2429 |
+
"grad_norm": 0.12743273377418518,
|
| 2430 |
+
"learning_rate": 0.0012,
|
| 2431 |
+
"loss": 2.262345790863037,
|
| 2432 |
+
"step": 692
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 0.1735,
|
| 2436 |
+
"grad_norm": 0.12541785836219788,
|
| 2437 |
+
"learning_rate": 0.0012,
|
| 2438 |
+
"loss": 2.200793504714966,
|
| 2439 |
+
"step": 694
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 0.174,
|
| 2443 |
+
"grad_norm": 0.12741735577583313,
|
| 2444 |
+
"learning_rate": 0.0012,
|
| 2445 |
+
"loss": 2.197943687438965,
|
| 2446 |
+
"step": 696
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 0.1745,
|
| 2450 |
+
"grad_norm": 0.12095748633146286,
|
| 2451 |
+
"learning_rate": 0.0012,
|
| 2452 |
+
"loss": 2.0293803215026855,
|
| 2453 |
+
"step": 698
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 0.175,
|
| 2457 |
+
"grad_norm": 0.11412248760461807,
|
| 2458 |
+
"learning_rate": 0.0012,
|
| 2459 |
+
"loss": 1.9167616367340088,
|
| 2460 |
+
"step": 700
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 0.1755,
|
| 2464 |
+
"grad_norm": 0.12195702642202377,
|
| 2465 |
+
"learning_rate": 0.0012,
|
| 2466 |
+
"loss": 2.086040496826172,
|
| 2467 |
+
"step": 702
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 0.176,
|
| 2471 |
+
"grad_norm": 0.12195327132940292,
|
| 2472 |
+
"learning_rate": 0.0012,
|
| 2473 |
+
"loss": 2.1532459259033203,
|
| 2474 |
+
"step": 704
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 0.1765,
|
| 2478 |
+
"grad_norm": 0.1243663877248764,
|
| 2479 |
+
"learning_rate": 0.0012,
|
| 2480 |
+
"loss": 2.200997829437256,
|
| 2481 |
+
"step": 706
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 0.177,
|
| 2485 |
+
"grad_norm": 0.11951598525047302,
|
| 2486 |
+
"learning_rate": 0.0012,
|
| 2487 |
+
"loss": 2.160093069076538,
|
| 2488 |
+
"step": 708
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 0.1775,
|
| 2492 |
+
"grad_norm": 0.1220458596944809,
|
| 2493 |
+
"learning_rate": 0.0012,
|
| 2494 |
+
"loss": 2.1600608825683594,
|
| 2495 |
+
"step": 710
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 0.178,
|
| 2499 |
+
"grad_norm": 0.12344282865524292,
|
| 2500 |
+
"learning_rate": 0.0012,
|
| 2501 |
+
"loss": 2.0668160915374756,
|
| 2502 |
+
"step": 712
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 0.1785,
|
| 2506 |
+
"grad_norm": 0.12584751844406128,
|
| 2507 |
+
"learning_rate": 0.0012,
|
| 2508 |
+
"loss": 2.249037504196167,
|
| 2509 |
+
"step": 714
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 0.179,
|
| 2513 |
+
"grad_norm": 0.12005162239074707,
|
| 2514 |
+
"learning_rate": 0.0012,
|
| 2515 |
+
"loss": 1.9391106367111206,
|
| 2516 |
+
"step": 716
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 0.1795,
|
| 2520 |
+
"grad_norm": 0.12259942293167114,
|
| 2521 |
+
"learning_rate": 0.0012,
|
| 2522 |
+
"loss": 2.0433120727539062,
|
| 2523 |
+
"step": 718
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 0.18,
|
| 2527 |
+
"grad_norm": 0.12099462002515793,
|
| 2528 |
+
"learning_rate": 0.0012,
|
| 2529 |
+
"loss": 2.034228801727295,
|
| 2530 |
+
"step": 720
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 0.1805,
|
| 2534 |
+
"grad_norm": 0.1258714497089386,
|
| 2535 |
+
"learning_rate": 0.0012,
|
| 2536 |
+
"loss": 2.2467880249023438,
|
| 2537 |
+
"step": 722
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 0.181,
|
| 2541 |
+
"grad_norm": 0.12345100194215775,
|
| 2542 |
+
"learning_rate": 0.0012,
|
| 2543 |
+
"loss": 2.1118154525756836,
|
| 2544 |
+
"step": 724
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 0.1815,
|
| 2548 |
+
"grad_norm": 0.12497245520353317,
|
| 2549 |
+
"learning_rate": 0.0012,
|
| 2550 |
+
"loss": 2.2561492919921875,
|
| 2551 |
+
"step": 726
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 0.182,
|
| 2555 |
+
"grad_norm": 0.12530677020549774,
|
| 2556 |
+
"learning_rate": 0.0012,
|
| 2557 |
+
"loss": 2.067718505859375,
|
| 2558 |
+
"step": 728
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 0.1825,
|
| 2562 |
+
"grad_norm": 0.12393652647733688,
|
| 2563 |
+
"learning_rate": 0.0012,
|
| 2564 |
+
"loss": 2.199213743209839,
|
| 2565 |
+
"step": 730
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 0.183,
|
| 2569 |
+
"grad_norm": 0.11859786510467529,
|
| 2570 |
+
"learning_rate": 0.0012,
|
| 2571 |
+
"loss": 1.8976398706436157,
|
| 2572 |
+
"step": 732
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 0.1835,
|
| 2576 |
+
"grad_norm": 0.12618829309940338,
|
| 2577 |
+
"learning_rate": 0.0012,
|
| 2578 |
+
"loss": 2.1067259311676025,
|
| 2579 |
+
"step": 734
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 0.184,
|
| 2583 |
+
"grad_norm": 0.12196256965398788,
|
| 2584 |
+
"learning_rate": 0.0012,
|
| 2585 |
+
"loss": 1.990835189819336,
|
| 2586 |
+
"step": 736
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 0.1845,
|
| 2590 |
+
"grad_norm": 0.12610235810279846,
|
| 2591 |
+
"learning_rate": 0.0012,
|
| 2592 |
+
"loss": 2.1363492012023926,
|
| 2593 |
+
"step": 738
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 0.185,
|
| 2597 |
+
"grad_norm": 0.13032737374305725,
|
| 2598 |
+
"learning_rate": 0.0012,
|
| 2599 |
+
"loss": 2.2302498817443848,
|
| 2600 |
+
"step": 740
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 0.1855,
|
| 2604 |
+
"grad_norm": 0.1223173514008522,
|
| 2605 |
+
"learning_rate": 0.0012,
|
| 2606 |
+
"loss": 2.102412223815918,
|
| 2607 |
+
"step": 742
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 0.186,
|
| 2611 |
+
"grad_norm": 0.12412013858556747,
|
| 2612 |
+
"learning_rate": 0.0012,
|
| 2613 |
+
"loss": 2.042909860610962,
|
| 2614 |
+
"step": 744
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 0.1865,
|
| 2618 |
+
"grad_norm": 0.12528584897518158,
|
| 2619 |
+
"learning_rate": 0.0012,
|
| 2620 |
+
"loss": 2.156968355178833,
|
| 2621 |
+
"step": 746
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 0.187,
|
| 2625 |
+
"grad_norm": 0.12142571061849594,
|
| 2626 |
+
"learning_rate": 0.0012,
|
| 2627 |
+
"loss": 2.0831336975097656,
|
| 2628 |
+
"step": 748
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 0.1875,
|
| 2632 |
+
"grad_norm": 0.12042216956615448,
|
| 2633 |
+
"learning_rate": 0.0012,
|
| 2634 |
+
"loss": 2.0399208068847656,
|
| 2635 |
+
"step": 750
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 0.188,
|
| 2639 |
+
"grad_norm": 0.1279473900794983,
|
| 2640 |
+
"learning_rate": 0.0012,
|
| 2641 |
+
"loss": 2.230341911315918,
|
| 2642 |
+
"step": 752
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 0.1885,
|
| 2646 |
+
"grad_norm": 0.1255028396844864,
|
| 2647 |
+
"learning_rate": 0.0012,
|
| 2648 |
+
"loss": 2.220186233520508,
|
| 2649 |
+
"step": 754
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 0.189,
|
| 2653 |
+
"grad_norm": 0.12432970106601715,
|
| 2654 |
+
"learning_rate": 0.0012,
|
| 2655 |
+
"loss": 2.112208843231201,
|
| 2656 |
+
"step": 756
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 0.1895,
|
| 2660 |
+
"grad_norm": 0.12302868068218231,
|
| 2661 |
+
"learning_rate": 0.0012,
|
| 2662 |
+
"loss": 2.1108741760253906,
|
| 2663 |
+
"step": 758
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 0.19,
|
| 2667 |
+
"grad_norm": 0.12439268827438354,
|
| 2668 |
+
"learning_rate": 0.0012,
|
| 2669 |
+
"loss": 2.093127489089966,
|
| 2670 |
+
"step": 760
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 0.1905,
|
| 2674 |
+
"grad_norm": 0.11993008106946945,
|
| 2675 |
+
"learning_rate": 0.0012,
|
| 2676 |
+
"loss": 1.958256721496582,
|
| 2677 |
+
"step": 762
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 0.191,
|
| 2681 |
+
"grad_norm": 0.1268967092037201,
|
| 2682 |
+
"learning_rate": 0.0012,
|
| 2683 |
+
"loss": 2.1421263217926025,
|
| 2684 |
+
"step": 764
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 0.1915,
|
| 2688 |
+
"grad_norm": 0.12185237556695938,
|
| 2689 |
+
"learning_rate": 0.0012,
|
| 2690 |
+
"loss": 2.036240816116333,
|
| 2691 |
+
"step": 766
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 0.192,
|
| 2695 |
+
"grad_norm": 0.1203518882393837,
|
| 2696 |
+
"learning_rate": 0.0012,
|
| 2697 |
+
"loss": 2.120025157928467,
|
| 2698 |
+
"step": 768
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 0.1925,
|
| 2702 |
+
"grad_norm": 0.12621000409126282,
|
| 2703 |
+
"learning_rate": 0.0012,
|
| 2704 |
+
"loss": 2.1707708835601807,
|
| 2705 |
+
"step": 770
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 0.193,
|
| 2709 |
+
"grad_norm": 0.12319474667310715,
|
| 2710 |
+
"learning_rate": 0.0012,
|
| 2711 |
+
"loss": 2.0842976570129395,
|
| 2712 |
+
"step": 772
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 0.1935,
|
| 2716 |
+
"grad_norm": 0.12575924396514893,
|
| 2717 |
+
"learning_rate": 0.0012,
|
| 2718 |
+
"loss": 2.174408197402954,
|
| 2719 |
+
"step": 774
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 0.194,
|
| 2723 |
+
"grad_norm": 0.12324059009552002,
|
| 2724 |
+
"learning_rate": 0.0012,
|
| 2725 |
+
"loss": 2.1565682888031006,
|
| 2726 |
+
"step": 776
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 0.1945,
|
| 2730 |
+
"grad_norm": 0.12696637213230133,
|
| 2731 |
+
"learning_rate": 0.0012,
|
| 2732 |
+
"loss": 2.242851734161377,
|
| 2733 |
+
"step": 778
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 0.195,
|
| 2737 |
+
"grad_norm": 0.12281235307455063,
|
| 2738 |
+
"learning_rate": 0.0012,
|
| 2739 |
+
"loss": 2.0886640548706055,
|
| 2740 |
+
"step": 780
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 0.1955,
|
| 2744 |
+
"grad_norm": 0.1271682232618332,
|
| 2745 |
+
"learning_rate": 0.0012,
|
| 2746 |
+
"loss": 2.181980609893799,
|
| 2747 |
+
"step": 782
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 0.196,
|
| 2751 |
+
"grad_norm": 0.12921912968158722,
|
| 2752 |
+
"learning_rate": 0.0012,
|
| 2753 |
+
"loss": 2.167280435562134,
|
| 2754 |
+
"step": 784
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.1965,
|
| 2758 |
+
"grad_norm": 0.1229398250579834,
|
| 2759 |
+
"learning_rate": 0.0012,
|
| 2760 |
+
"loss": 2.111304759979248,
|
| 2761 |
+
"step": 786
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.197,
|
| 2765 |
+
"grad_norm": 0.12247031927108765,
|
| 2766 |
+
"learning_rate": 0.0012,
|
| 2767 |
+
"loss": 2.1160202026367188,
|
| 2768 |
+
"step": 788
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 0.1975,
|
| 2772 |
+
"grad_norm": 0.12028486281633377,
|
| 2773 |
+
"learning_rate": 0.0012,
|
| 2774 |
+
"loss": 2.0347137451171875,
|
| 2775 |
+
"step": 790
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 0.198,
|
| 2779 |
+
"grad_norm": 0.12134865671396255,
|
| 2780 |
+
"learning_rate": 0.0012,
|
| 2781 |
+
"loss": 2.09946870803833,
|
| 2782 |
+
"step": 792
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 0.1985,
|
| 2786 |
+
"grad_norm": 0.12271501123905182,
|
| 2787 |
+
"learning_rate": 0.0012,
|
| 2788 |
+
"loss": 2.1185460090637207,
|
| 2789 |
+
"step": 794
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 0.199,
|
| 2793 |
+
"grad_norm": 0.12715184688568115,
|
| 2794 |
+
"learning_rate": 0.0012,
|
| 2795 |
+
"loss": 1.9932470321655273,
|
| 2796 |
+
"step": 796
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.1995,
|
| 2800 |
+
"grad_norm": 0.12658338248729706,
|
| 2801 |
+
"learning_rate": 0.0012,
|
| 2802 |
+
"loss": 2.1957900524139404,
|
| 2803 |
+
"step": 798
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 0.2,
|
| 2807 |
+
"grad_norm": 0.1263073831796646,
|
| 2808 |
+
"learning_rate": 0.0012,
|
| 2809 |
+
"loss": 2.262310028076172,
|
| 2810 |
+
"step": 800
|
| 2811 |
}
|
| 2812 |
],
|
| 2813 |
"logging_steps": 2,
|
|
|
|
| 2827 |
"attributes": {}
|
| 2828 |
}
|
| 2829 |
},
|
| 2830 |
+
"total_flos": 7.717577544727265e+17,
|
| 2831 |
"train_batch_size": 2,
|
| 2832 |
"trial_name": null,
|
| 2833 |
"trial_params": null
|