Training in progress, step 1148, 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 2934397032
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84df729e7e8fb5c2f1ea59d0a63abf4d13e2f12303386bd44c2ae490bfce5814
|
| 3 |
size 2934397032
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3631584265
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7a351fd4eb2f58e6f68c02c82f53a6bac6eebe34ffa3c3e5fe21a54813d142f
|
| 3 |
size 3631584265
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6bae4afb6fbbfacc38bc3d60430c52c058dd6c71ae8fa94c8dcf8854eb4fc176
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 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,
|
|
@@ -2017,6 +2017,2015 @@
|
|
| 2017 |
"learning_rate": 5e-05,
|
| 2018 |
"loss": 2.0299763679504395,
|
| 2019 |
"step": 574
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2020 |
}
|
| 2021 |
],
|
| 2022 |
"logging_steps": 2,
|
|
@@ -2036,7 +4045,7 @@
|
|
| 2036 |
"attributes": {}
|
| 2037 |
}
|
| 2038 |
},
|
| 2039 |
-
"total_flos":
|
| 2040 |
"train_batch_size": 1,
|
| 2041 |
"trial_name": null,
|
| 2042 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.40029987621388846,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 1148,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 2017 |
"learning_rate": 5e-05,
|
| 2018 |
"loss": 2.0299763679504395,
|
| 2019 |
"step": 574
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.20084732465087085,
|
| 2023 |
+
"grad_norm": 0.18850381672382355,
|
| 2024 |
+
"learning_rate": 5e-05,
|
| 2025 |
+
"loss": 2.131117105484009,
|
| 2026 |
+
"step": 576
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 0.2015447111947975,
|
| 2030 |
+
"grad_norm": 0.27086082100868225,
|
| 2031 |
+
"learning_rate": 5e-05,
|
| 2032 |
+
"loss": 2.1912379264831543,
|
| 2033 |
+
"step": 578
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 0.20224209773872412,
|
| 2037 |
+
"grad_norm": 0.27545830607414246,
|
| 2038 |
+
"learning_rate": 5e-05,
|
| 2039 |
+
"loss": 2.19915771484375,
|
| 2040 |
+
"step": 580
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 0.20293948428265077,
|
| 2044 |
+
"grad_norm": 0.2752111256122589,
|
| 2045 |
+
"learning_rate": 5e-05,
|
| 2046 |
+
"loss": 2.267606258392334,
|
| 2047 |
+
"step": 582
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 0.2036368708265774,
|
| 2051 |
+
"grad_norm": 0.319011926651001,
|
| 2052 |
+
"learning_rate": 5e-05,
|
| 2053 |
+
"loss": 2.2916200160980225,
|
| 2054 |
+
"step": 584
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 0.20433425737050404,
|
| 2058 |
+
"grad_norm": 0.20431022346019745,
|
| 2059 |
+
"learning_rate": 5e-05,
|
| 2060 |
+
"loss": 2.2031617164611816,
|
| 2061 |
+
"step": 586
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 0.20503164391443068,
|
| 2065 |
+
"grad_norm": 0.2223842591047287,
|
| 2066 |
+
"learning_rate": 5e-05,
|
| 2067 |
+
"loss": 2.196582317352295,
|
| 2068 |
+
"step": 588
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 0.2057290304583573,
|
| 2072 |
+
"grad_norm": 0.3272736370563507,
|
| 2073 |
+
"learning_rate": 5e-05,
|
| 2074 |
+
"loss": 2.1547532081604004,
|
| 2075 |
+
"step": 590
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 0.20642641700228395,
|
| 2079 |
+
"grad_norm": 0.18153901398181915,
|
| 2080 |
+
"learning_rate": 5e-05,
|
| 2081 |
+
"loss": 2.2748570442199707,
|
| 2082 |
+
"step": 592
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 0.20712380354621057,
|
| 2086 |
+
"grad_norm": 0.20026224851608276,
|
| 2087 |
+
"learning_rate": 5e-05,
|
| 2088 |
+
"loss": 2.185814619064331,
|
| 2089 |
+
"step": 594
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 0.20782119009013722,
|
| 2093 |
+
"grad_norm": 0.2095404714345932,
|
| 2094 |
+
"learning_rate": 5e-05,
|
| 2095 |
+
"loss": 2.116464138031006,
|
| 2096 |
+
"step": 596
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.20851857663406384,
|
| 2100 |
+
"grad_norm": 0.2254505455493927,
|
| 2101 |
+
"learning_rate": 5e-05,
|
| 2102 |
+
"loss": 2.290566921234131,
|
| 2103 |
+
"step": 598
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 0.2092159631779905,
|
| 2107 |
+
"grad_norm": 0.22954094409942627,
|
| 2108 |
+
"learning_rate": 5e-05,
|
| 2109 |
+
"loss": 2.1632447242736816,
|
| 2110 |
+
"step": 600
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 0.2099133497219171,
|
| 2114 |
+
"grad_norm": 0.17814919352531433,
|
| 2115 |
+
"learning_rate": 5e-05,
|
| 2116 |
+
"loss": 2.1754729747772217,
|
| 2117 |
+
"step": 602
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 0.21061073626584376,
|
| 2121 |
+
"grad_norm": 0.31458622217178345,
|
| 2122 |
+
"learning_rate": 5e-05,
|
| 2123 |
+
"loss": 2.4768474102020264,
|
| 2124 |
+
"step": 604
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 0.21130812280977038,
|
| 2128 |
+
"grad_norm": 0.16490019857883453,
|
| 2129 |
+
"learning_rate": 5e-05,
|
| 2130 |
+
"loss": 2.1874232292175293,
|
| 2131 |
+
"step": 606
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 0.21200550935369702,
|
| 2135 |
+
"grad_norm": 0.2257475107908249,
|
| 2136 |
+
"learning_rate": 5e-05,
|
| 2137 |
+
"loss": 2.2535409927368164,
|
| 2138 |
+
"step": 608
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 0.21270289589762365,
|
| 2142 |
+
"grad_norm": 0.2660852074623108,
|
| 2143 |
+
"learning_rate": 5e-05,
|
| 2144 |
+
"loss": 2.061890125274658,
|
| 2145 |
+
"step": 610
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 0.2134002824415503,
|
| 2149 |
+
"grad_norm": 0.2156171202659607,
|
| 2150 |
+
"learning_rate": 5e-05,
|
| 2151 |
+
"loss": 2.2851383686065674,
|
| 2152 |
+
"step": 612
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 0.2140976689854769,
|
| 2156 |
+
"grad_norm": 0.17925716936588287,
|
| 2157 |
+
"learning_rate": 5e-05,
|
| 2158 |
+
"loss": 2.082984447479248,
|
| 2159 |
+
"step": 614
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 0.21479505552940356,
|
| 2163 |
+
"grad_norm": 0.3436058759689331,
|
| 2164 |
+
"learning_rate": 5e-05,
|
| 2165 |
+
"loss": 2.1233651638031006,
|
| 2166 |
+
"step": 616
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 0.21549244207333018,
|
| 2170 |
+
"grad_norm": 0.22150059044361115,
|
| 2171 |
+
"learning_rate": 5e-05,
|
| 2172 |
+
"loss": 2.023638963699341,
|
| 2173 |
+
"step": 618
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 0.21618982861725683,
|
| 2177 |
+
"grad_norm": 0.19689801335334778,
|
| 2178 |
+
"learning_rate": 5e-05,
|
| 2179 |
+
"loss": 2.082303762435913,
|
| 2180 |
+
"step": 620
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 0.21688721516118348,
|
| 2184 |
+
"grad_norm": 0.1721586138010025,
|
| 2185 |
+
"learning_rate": 5e-05,
|
| 2186 |
+
"loss": 2.1851935386657715,
|
| 2187 |
+
"step": 622
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 0.2175846017051101,
|
| 2191 |
+
"grad_norm": 0.21074773371219635,
|
| 2192 |
+
"learning_rate": 5e-05,
|
| 2193 |
+
"loss": 2.125735282897949,
|
| 2194 |
+
"step": 624
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 0.21828198824903675,
|
| 2198 |
+
"grad_norm": 0.2333507388830185,
|
| 2199 |
+
"learning_rate": 5e-05,
|
| 2200 |
+
"loss": 2.28065824508667,
|
| 2201 |
+
"step": 626
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.21897937479296337,
|
| 2205 |
+
"grad_norm": 0.20217208564281464,
|
| 2206 |
+
"learning_rate": 5e-05,
|
| 2207 |
+
"loss": 2.1805195808410645,
|
| 2208 |
+
"step": 628
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.21967676133689001,
|
| 2212 |
+
"grad_norm": 0.21829625964164734,
|
| 2213 |
+
"learning_rate": 5e-05,
|
| 2214 |
+
"loss": 1.9504680633544922,
|
| 2215 |
+
"step": 630
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.22037414788081663,
|
| 2219 |
+
"grad_norm": 0.19575423002243042,
|
| 2220 |
+
"learning_rate": 5e-05,
|
| 2221 |
+
"loss": 2.1475088596343994,
|
| 2222 |
+
"step": 632
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.22107153442474328,
|
| 2226 |
+
"grad_norm": 0.23382477462291718,
|
| 2227 |
+
"learning_rate": 5e-05,
|
| 2228 |
+
"loss": 2.1880345344543457,
|
| 2229 |
+
"step": 634
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.2217689209686699,
|
| 2233 |
+
"grad_norm": 0.2851170599460602,
|
| 2234 |
+
"learning_rate": 5e-05,
|
| 2235 |
+
"loss": 2.126394033432007,
|
| 2236 |
+
"step": 636
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 0.22246630751259655,
|
| 2240 |
+
"grad_norm": 0.2504940927028656,
|
| 2241 |
+
"learning_rate": 5e-05,
|
| 2242 |
+
"loss": 2.028315544128418,
|
| 2243 |
+
"step": 638
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.22316369405652317,
|
| 2247 |
+
"grad_norm": 0.1850201040506363,
|
| 2248 |
+
"learning_rate": 5e-05,
|
| 2249 |
+
"loss": 2.234729766845703,
|
| 2250 |
+
"step": 640
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 0.22386108060044982,
|
| 2254 |
+
"grad_norm": 0.15266118943691254,
|
| 2255 |
+
"learning_rate": 5e-05,
|
| 2256 |
+
"loss": 2.186972141265869,
|
| 2257 |
+
"step": 642
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.22455846714437644,
|
| 2261 |
+
"grad_norm": 0.2082221508026123,
|
| 2262 |
+
"learning_rate": 5e-05,
|
| 2263 |
+
"loss": 2.3294734954833984,
|
| 2264 |
+
"step": 644
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 0.2252558536883031,
|
| 2268 |
+
"grad_norm": 0.2343757152557373,
|
| 2269 |
+
"learning_rate": 5e-05,
|
| 2270 |
+
"loss": 2.3024473190307617,
|
| 2271 |
+
"step": 646
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.2259532402322297,
|
| 2275 |
+
"grad_norm": 0.27744945883750916,
|
| 2276 |
+
"learning_rate": 5e-05,
|
| 2277 |
+
"loss": 2.151794910430908,
|
| 2278 |
+
"step": 648
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 0.22665062677615636,
|
| 2282 |
+
"grad_norm": 0.19519643485546112,
|
| 2283 |
+
"learning_rate": 5e-05,
|
| 2284 |
+
"loss": 2.2723190784454346,
|
| 2285 |
+
"step": 650
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 0.22734801332008298,
|
| 2289 |
+
"grad_norm": 0.21738861501216888,
|
| 2290 |
+
"learning_rate": 5e-05,
|
| 2291 |
+
"loss": 2.1555187702178955,
|
| 2292 |
+
"step": 652
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 0.22804539986400962,
|
| 2296 |
+
"grad_norm": 0.2636173367500305,
|
| 2297 |
+
"learning_rate": 5e-05,
|
| 2298 |
+
"loss": 2.214590072631836,
|
| 2299 |
+
"step": 654
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.22874278640793627,
|
| 2303 |
+
"grad_norm": 0.21495908498764038,
|
| 2304 |
+
"learning_rate": 5e-05,
|
| 2305 |
+
"loss": 2.2480390071868896,
|
| 2306 |
+
"step": 656
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.2294401729518629,
|
| 2310 |
+
"grad_norm": 0.2296120971441269,
|
| 2311 |
+
"learning_rate": 5e-05,
|
| 2312 |
+
"loss": 2.1904211044311523,
|
| 2313 |
+
"step": 658
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.23013755949578954,
|
| 2317 |
+
"grad_norm": 0.30581578612327576,
|
| 2318 |
+
"learning_rate": 5e-05,
|
| 2319 |
+
"loss": 2.1074323654174805,
|
| 2320 |
+
"step": 660
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 0.23083494603971616,
|
| 2324 |
+
"grad_norm": 0.19764243066310883,
|
| 2325 |
+
"learning_rate": 5e-05,
|
| 2326 |
+
"loss": 2.223814010620117,
|
| 2327 |
+
"step": 662
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 0.2315323325836428,
|
| 2331 |
+
"grad_norm": 0.21603018045425415,
|
| 2332 |
+
"learning_rate": 5e-05,
|
| 2333 |
+
"loss": 2.140528440475464,
|
| 2334 |
+
"step": 664
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 0.23222971912756943,
|
| 2338 |
+
"grad_norm": 0.1800299435853958,
|
| 2339 |
+
"learning_rate": 5e-05,
|
| 2340 |
+
"loss": 2.1906323432922363,
|
| 2341 |
+
"step": 666
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.23292710567149608,
|
| 2345 |
+
"grad_norm": 0.24724136292934418,
|
| 2346 |
+
"learning_rate": 5e-05,
|
| 2347 |
+
"loss": 2.120924949645996,
|
| 2348 |
+
"step": 668
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 0.2336244922154227,
|
| 2352 |
+
"grad_norm": 0.2484823316335678,
|
| 2353 |
+
"learning_rate": 5e-05,
|
| 2354 |
+
"loss": 2.010047435760498,
|
| 2355 |
+
"step": 670
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 0.23432187875934934,
|
| 2359 |
+
"grad_norm": 0.19774702191352844,
|
| 2360 |
+
"learning_rate": 5e-05,
|
| 2361 |
+
"loss": 2.1234774589538574,
|
| 2362 |
+
"step": 672
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 0.23501926530327596,
|
| 2366 |
+
"grad_norm": 0.2553282678127289,
|
| 2367 |
+
"learning_rate": 5e-05,
|
| 2368 |
+
"loss": 2.1027908325195312,
|
| 2369 |
+
"step": 674
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 0.2357166518472026,
|
| 2373 |
+
"grad_norm": 0.20291560888290405,
|
| 2374 |
+
"learning_rate": 5e-05,
|
| 2375 |
+
"loss": 2.1387977600097656,
|
| 2376 |
+
"step": 676
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 0.23641403839112923,
|
| 2380 |
+
"grad_norm": 0.34688642621040344,
|
| 2381 |
+
"learning_rate": 5e-05,
|
| 2382 |
+
"loss": 2.1452786922454834,
|
| 2383 |
+
"step": 678
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 0.23711142493505588,
|
| 2387 |
+
"grad_norm": 0.2841176688671112,
|
| 2388 |
+
"learning_rate": 5e-05,
|
| 2389 |
+
"loss": 1.9979159832000732,
|
| 2390 |
+
"step": 680
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 0.2378088114789825,
|
| 2394 |
+
"grad_norm": 0.21012328565120697,
|
| 2395 |
+
"learning_rate": 5e-05,
|
| 2396 |
+
"loss": 2.271960735321045,
|
| 2397 |
+
"step": 682
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 0.23850619802290915,
|
| 2401 |
+
"grad_norm": 0.161057710647583,
|
| 2402 |
+
"learning_rate": 5e-05,
|
| 2403 |
+
"loss": 2.0357096195220947,
|
| 2404 |
+
"step": 684
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 0.23920358456683577,
|
| 2408 |
+
"grad_norm": 0.2356063425540924,
|
| 2409 |
+
"learning_rate": 5e-05,
|
| 2410 |
+
"loss": 2.1391124725341797,
|
| 2411 |
+
"step": 686
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 0.23990097111076242,
|
| 2415 |
+
"grad_norm": 0.21011163294315338,
|
| 2416 |
+
"learning_rate": 5e-05,
|
| 2417 |
+
"loss": 2.1928865909576416,
|
| 2418 |
+
"step": 688
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 0.24059835765468907,
|
| 2422 |
+
"grad_norm": 0.2434798926115036,
|
| 2423 |
+
"learning_rate": 5e-05,
|
| 2424 |
+
"loss": 2.017970323562622,
|
| 2425 |
+
"step": 690
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 0.24129574419861569,
|
| 2429 |
+
"grad_norm": 0.24937564134597778,
|
| 2430 |
+
"learning_rate": 5e-05,
|
| 2431 |
+
"loss": 2.058408260345459,
|
| 2432 |
+
"step": 692
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 0.24199313074254233,
|
| 2436 |
+
"grad_norm": 0.2534058094024658,
|
| 2437 |
+
"learning_rate": 5e-05,
|
| 2438 |
+
"loss": 2.002953052520752,
|
| 2439 |
+
"step": 694
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 0.24269051728646895,
|
| 2443 |
+
"grad_norm": 0.35094359517097473,
|
| 2444 |
+
"learning_rate": 5e-05,
|
| 2445 |
+
"loss": 2.2217190265655518,
|
| 2446 |
+
"step": 696
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 0.2433879038303956,
|
| 2450 |
+
"grad_norm": 0.19374975562095642,
|
| 2451 |
+
"learning_rate": 5e-05,
|
| 2452 |
+
"loss": 2.1322073936462402,
|
| 2453 |
+
"step": 698
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 0.24408529037432222,
|
| 2457 |
+
"grad_norm": 0.4570079445838928,
|
| 2458 |
+
"learning_rate": 5e-05,
|
| 2459 |
+
"loss": 2.106457233428955,
|
| 2460 |
+
"step": 700
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 0.24478267691824887,
|
| 2464 |
+
"grad_norm": 0.2247321903705597,
|
| 2465 |
+
"learning_rate": 5e-05,
|
| 2466 |
+
"loss": 2.0081517696380615,
|
| 2467 |
+
"step": 702
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 0.2454800634621755,
|
| 2471 |
+
"grad_norm": 0.21547703444957733,
|
| 2472 |
+
"learning_rate": 5e-05,
|
| 2473 |
+
"loss": 2.137990713119507,
|
| 2474 |
+
"step": 704
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 0.24617745000610214,
|
| 2478 |
+
"grad_norm": 0.253587007522583,
|
| 2479 |
+
"learning_rate": 5e-05,
|
| 2480 |
+
"loss": 2.056151866912842,
|
| 2481 |
+
"step": 706
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 0.24687483655002876,
|
| 2485 |
+
"grad_norm": 0.3514871895313263,
|
| 2486 |
+
"learning_rate": 5e-05,
|
| 2487 |
+
"loss": 2.067183494567871,
|
| 2488 |
+
"step": 708
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 0.2475722230939554,
|
| 2492 |
+
"grad_norm": 0.18376290798187256,
|
| 2493 |
+
"learning_rate": 5e-05,
|
| 2494 |
+
"loss": 2.1726388931274414,
|
| 2495 |
+
"step": 710
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 0.24826960963788203,
|
| 2499 |
+
"grad_norm": 0.2481544315814972,
|
| 2500 |
+
"learning_rate": 5e-05,
|
| 2501 |
+
"loss": 2.035372257232666,
|
| 2502 |
+
"step": 712
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 0.24896699618180868,
|
| 2506 |
+
"grad_norm": 0.23347629606723785,
|
| 2507 |
+
"learning_rate": 5e-05,
|
| 2508 |
+
"loss": 2.2434043884277344,
|
| 2509 |
+
"step": 714
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 0.2496643827257353,
|
| 2513 |
+
"grad_norm": 0.2897323668003082,
|
| 2514 |
+
"learning_rate": 5e-05,
|
| 2515 |
+
"loss": 2.0160436630249023,
|
| 2516 |
+
"step": 716
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 0.2503617692696619,
|
| 2520 |
+
"grad_norm": 0.24364791810512543,
|
| 2521 |
+
"learning_rate": 5e-05,
|
| 2522 |
+
"loss": 2.09956955909729,
|
| 2523 |
+
"step": 718
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 0.25105915581358856,
|
| 2527 |
+
"grad_norm": 0.24172277748584747,
|
| 2528 |
+
"learning_rate": 5e-05,
|
| 2529 |
+
"loss": 2.1436986923217773,
|
| 2530 |
+
"step": 720
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 0.2517565423575152,
|
| 2534 |
+
"grad_norm": 0.24258746206760406,
|
| 2535 |
+
"learning_rate": 5e-05,
|
| 2536 |
+
"loss": 2.0807886123657227,
|
| 2537 |
+
"step": 722
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 0.25245392890144186,
|
| 2541 |
+
"grad_norm": 0.20534907281398773,
|
| 2542 |
+
"learning_rate": 5e-05,
|
| 2543 |
+
"loss": 1.9933953285217285,
|
| 2544 |
+
"step": 724
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 0.2531513154453685,
|
| 2548 |
+
"grad_norm": 0.19531792402267456,
|
| 2549 |
+
"learning_rate": 5e-05,
|
| 2550 |
+
"loss": 2.0712335109710693,
|
| 2551 |
+
"step": 726
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 0.2538487019892951,
|
| 2555 |
+
"grad_norm": 0.2692209780216217,
|
| 2556 |
+
"learning_rate": 5e-05,
|
| 2557 |
+
"loss": 1.8906567096710205,
|
| 2558 |
+
"step": 728
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 0.25454608853322175,
|
| 2562 |
+
"grad_norm": 0.28854358196258545,
|
| 2563 |
+
"learning_rate": 5e-05,
|
| 2564 |
+
"loss": 2.0448617935180664,
|
| 2565 |
+
"step": 730
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 0.2552434750771484,
|
| 2569 |
+
"grad_norm": 0.21004259586334229,
|
| 2570 |
+
"learning_rate": 5e-05,
|
| 2571 |
+
"loss": 2.061755418777466,
|
| 2572 |
+
"step": 732
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 0.25594086162107504,
|
| 2576 |
+
"grad_norm": 0.1910502016544342,
|
| 2577 |
+
"learning_rate": 5e-05,
|
| 2578 |
+
"loss": 2.150097131729126,
|
| 2579 |
+
"step": 734
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 0.25663824816500164,
|
| 2583 |
+
"grad_norm": 0.24876752495765686,
|
| 2584 |
+
"learning_rate": 5e-05,
|
| 2585 |
+
"loss": 2.1157419681549072,
|
| 2586 |
+
"step": 736
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 0.2573356347089283,
|
| 2590 |
+
"grad_norm": 0.22561496496200562,
|
| 2591 |
+
"learning_rate": 5e-05,
|
| 2592 |
+
"loss": 2.087238311767578,
|
| 2593 |
+
"step": 738
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 0.25803302125285493,
|
| 2597 |
+
"grad_norm": 0.2667919993400574,
|
| 2598 |
+
"learning_rate": 5e-05,
|
| 2599 |
+
"loss": 2.117800712585449,
|
| 2600 |
+
"step": 740
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 0.2587304077967816,
|
| 2604 |
+
"grad_norm": 0.33913540840148926,
|
| 2605 |
+
"learning_rate": 5e-05,
|
| 2606 |
+
"loss": 2.0394740104675293,
|
| 2607 |
+
"step": 742
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 0.2594277943407082,
|
| 2611 |
+
"grad_norm": 0.1690003126859665,
|
| 2612 |
+
"learning_rate": 5e-05,
|
| 2613 |
+
"loss": 2.0725057125091553,
|
| 2614 |
+
"step": 744
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 0.2601251808846348,
|
| 2618 |
+
"grad_norm": 0.2350507229566574,
|
| 2619 |
+
"learning_rate": 5e-05,
|
| 2620 |
+
"loss": 2.180880546569824,
|
| 2621 |
+
"step": 746
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 0.26082256742856147,
|
| 2625 |
+
"grad_norm": 0.26825854182243347,
|
| 2626 |
+
"learning_rate": 5e-05,
|
| 2627 |
+
"loss": 2.1213607788085938,
|
| 2628 |
+
"step": 748
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 0.2615199539724881,
|
| 2632 |
+
"grad_norm": 0.1949106603860855,
|
| 2633 |
+
"learning_rate": 5e-05,
|
| 2634 |
+
"loss": 2.0407943725585938,
|
| 2635 |
+
"step": 750
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 0.2622173405164147,
|
| 2639 |
+
"grad_norm": 0.3467162549495697,
|
| 2640 |
+
"learning_rate": 5e-05,
|
| 2641 |
+
"loss": 1.9864469766616821,
|
| 2642 |
+
"step": 752
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 0.26291472706034136,
|
| 2646 |
+
"grad_norm": 0.20342646539211273,
|
| 2647 |
+
"learning_rate": 5e-05,
|
| 2648 |
+
"loss": 2.171985387802124,
|
| 2649 |
+
"step": 754
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 0.263612113604268,
|
| 2653 |
+
"grad_norm": 0.209679514169693,
|
| 2654 |
+
"learning_rate": 5e-05,
|
| 2655 |
+
"loss": 2.0921454429626465,
|
| 2656 |
+
"step": 756
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 0.26430950014819465,
|
| 2660 |
+
"grad_norm": 0.237525075674057,
|
| 2661 |
+
"learning_rate": 5e-05,
|
| 2662 |
+
"loss": 2.1939406394958496,
|
| 2663 |
+
"step": 758
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 0.2650068866921213,
|
| 2667 |
+
"grad_norm": 0.2145080417394638,
|
| 2668 |
+
"learning_rate": 5e-05,
|
| 2669 |
+
"loss": 2.132781505584717,
|
| 2670 |
+
"step": 760
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 0.2657042732360479,
|
| 2674 |
+
"grad_norm": 0.2128620594739914,
|
| 2675 |
+
"learning_rate": 5e-05,
|
| 2676 |
+
"loss": 2.053044080734253,
|
| 2677 |
+
"step": 762
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 0.26640165977997454,
|
| 2681 |
+
"grad_norm": 0.213362455368042,
|
| 2682 |
+
"learning_rate": 5e-05,
|
| 2683 |
+
"loss": 2.026813507080078,
|
| 2684 |
+
"step": 764
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 0.2670990463239012,
|
| 2688 |
+
"grad_norm": 0.28608018159866333,
|
| 2689 |
+
"learning_rate": 5e-05,
|
| 2690 |
+
"loss": 2.0595083236694336,
|
| 2691 |
+
"step": 766
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 0.26779643286782784,
|
| 2695 |
+
"grad_norm": 0.2690437138080597,
|
| 2696 |
+
"learning_rate": 5e-05,
|
| 2697 |
+
"loss": 1.9456102848052979,
|
| 2698 |
+
"step": 768
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 0.26849381941175443,
|
| 2702 |
+
"grad_norm": 0.16010865569114685,
|
| 2703 |
+
"learning_rate": 5e-05,
|
| 2704 |
+
"loss": 2.2088522911071777,
|
| 2705 |
+
"step": 770
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 0.2691912059556811,
|
| 2709 |
+
"grad_norm": 0.2786443531513214,
|
| 2710 |
+
"learning_rate": 5e-05,
|
| 2711 |
+
"loss": 2.180513381958008,
|
| 2712 |
+
"step": 772
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 0.2698885924996077,
|
| 2716 |
+
"grad_norm": 0.1664266735315323,
|
| 2717 |
+
"learning_rate": 5e-05,
|
| 2718 |
+
"loss": 2.1334125995635986,
|
| 2719 |
+
"step": 774
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 0.2705859790435344,
|
| 2723 |
+
"grad_norm": 0.21176394820213318,
|
| 2724 |
+
"learning_rate": 5e-05,
|
| 2725 |
+
"loss": 2.072683334350586,
|
| 2726 |
+
"step": 776
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 0.27128336558746097,
|
| 2730 |
+
"grad_norm": 0.2954282760620117,
|
| 2731 |
+
"learning_rate": 5e-05,
|
| 2732 |
+
"loss": 1.9445130825042725,
|
| 2733 |
+
"step": 778
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 0.2719807521313876,
|
| 2737 |
+
"grad_norm": 0.3181959092617035,
|
| 2738 |
+
"learning_rate": 5e-05,
|
| 2739 |
+
"loss": 2.138401508331299,
|
| 2740 |
+
"step": 780
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 0.27267813867531426,
|
| 2744 |
+
"grad_norm": 0.34124651551246643,
|
| 2745 |
+
"learning_rate": 5e-05,
|
| 2746 |
+
"loss": 2.053771495819092,
|
| 2747 |
+
"step": 782
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 0.2733755252192409,
|
| 2751 |
+
"grad_norm": 0.18755890429019928,
|
| 2752 |
+
"learning_rate": 5e-05,
|
| 2753 |
+
"loss": 2.2385504245758057,
|
| 2754 |
+
"step": 784
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.2740729117631675,
|
| 2758 |
+
"grad_norm": 0.23205752670764923,
|
| 2759 |
+
"learning_rate": 5e-05,
|
| 2760 |
+
"loss": 1.8543893098831177,
|
| 2761 |
+
"step": 786
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.27477029830709415,
|
| 2765 |
+
"grad_norm": 0.2637777626514435,
|
| 2766 |
+
"learning_rate": 5e-05,
|
| 2767 |
+
"loss": 2.216790199279785,
|
| 2768 |
+
"step": 788
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 0.2754676848510208,
|
| 2772 |
+
"grad_norm": 0.23640374839305878,
|
| 2773 |
+
"learning_rate": 5e-05,
|
| 2774 |
+
"loss": 1.9534659385681152,
|
| 2775 |
+
"step": 790
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 0.27616507139494745,
|
| 2779 |
+
"grad_norm": 0.3152559697628021,
|
| 2780 |
+
"learning_rate": 5e-05,
|
| 2781 |
+
"loss": 1.9674100875854492,
|
| 2782 |
+
"step": 792
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 0.2768624579388741,
|
| 2786 |
+
"grad_norm": 0.24841763079166412,
|
| 2787 |
+
"learning_rate": 5e-05,
|
| 2788 |
+
"loss": 1.9673519134521484,
|
| 2789 |
+
"step": 794
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 0.2775598444828007,
|
| 2793 |
+
"grad_norm": 0.2557215392589569,
|
| 2794 |
+
"learning_rate": 5e-05,
|
| 2795 |
+
"loss": 2.1061301231384277,
|
| 2796 |
+
"step": 796
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.27825723102672734,
|
| 2800 |
+
"grad_norm": 0.26791617274284363,
|
| 2801 |
+
"learning_rate": 5e-05,
|
| 2802 |
+
"loss": 1.980527639389038,
|
| 2803 |
+
"step": 798
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 0.278954617570654,
|
| 2807 |
+
"grad_norm": 0.18597114086151123,
|
| 2808 |
+
"learning_rate": 5e-05,
|
| 2809 |
+
"loss": 2.1571919918060303,
|
| 2810 |
+
"step": 800
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 0.27965200411458063,
|
| 2814 |
+
"grad_norm": 0.24104788899421692,
|
| 2815 |
+
"learning_rate": 5e-05,
|
| 2816 |
+
"loss": 2.052887439727783,
|
| 2817 |
+
"step": 802
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 0.2803493906585072,
|
| 2821 |
+
"grad_norm": 0.2629118263721466,
|
| 2822 |
+
"learning_rate": 5e-05,
|
| 2823 |
+
"loss": 2.1480886936187744,
|
| 2824 |
+
"step": 804
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 0.2810467772024339,
|
| 2828 |
+
"grad_norm": 0.1891702115535736,
|
| 2829 |
+
"learning_rate": 5e-05,
|
| 2830 |
+
"loss": 1.8945896625518799,
|
| 2831 |
+
"step": 806
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 0.2817441637463605,
|
| 2835 |
+
"grad_norm": 0.23194870352745056,
|
| 2836 |
+
"learning_rate": 5e-05,
|
| 2837 |
+
"loss": 2.0242464542388916,
|
| 2838 |
+
"step": 808
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 0.28244155029028717,
|
| 2842 |
+
"grad_norm": 0.22819095849990845,
|
| 2843 |
+
"learning_rate": 5e-05,
|
| 2844 |
+
"loss": 1.9482749700546265,
|
| 2845 |
+
"step": 810
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 0.28313893683421376,
|
| 2849 |
+
"grad_norm": 0.20101813971996307,
|
| 2850 |
+
"learning_rate": 5e-05,
|
| 2851 |
+
"loss": 2.1300370693206787,
|
| 2852 |
+
"step": 812
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 0.2838363233781404,
|
| 2856 |
+
"grad_norm": 0.371175080537796,
|
| 2857 |
+
"learning_rate": 5e-05,
|
| 2858 |
+
"loss": 2.1477580070495605,
|
| 2859 |
+
"step": 814
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 0.28453370992206706,
|
| 2863 |
+
"grad_norm": 0.30335918068885803,
|
| 2864 |
+
"learning_rate": 5e-05,
|
| 2865 |
+
"loss": 1.829369068145752,
|
| 2866 |
+
"step": 816
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 0.2852310964659937,
|
| 2870 |
+
"grad_norm": 0.2618269920349121,
|
| 2871 |
+
"learning_rate": 5e-05,
|
| 2872 |
+
"loss": 1.8751198053359985,
|
| 2873 |
+
"step": 818
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 0.2859284830099203,
|
| 2877 |
+
"grad_norm": 0.2387172430753708,
|
| 2878 |
+
"learning_rate": 5e-05,
|
| 2879 |
+
"loss": 1.9326030015945435,
|
| 2880 |
+
"step": 820
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 0.28662586955384695,
|
| 2884 |
+
"grad_norm": 0.24157649278640747,
|
| 2885 |
+
"learning_rate": 5e-05,
|
| 2886 |
+
"loss": 1.9281904697418213,
|
| 2887 |
+
"step": 822
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 0.2873232560977736,
|
| 2891 |
+
"grad_norm": 0.2316661924123764,
|
| 2892 |
+
"learning_rate": 5e-05,
|
| 2893 |
+
"loss": 2.1554112434387207,
|
| 2894 |
+
"step": 824
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 0.28802064264170024,
|
| 2898 |
+
"grad_norm": 0.3204042613506317,
|
| 2899 |
+
"learning_rate": 5e-05,
|
| 2900 |
+
"loss": 1.948148488998413,
|
| 2901 |
+
"step": 826
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 0.2887180291856269,
|
| 2905 |
+
"grad_norm": 0.20631511509418488,
|
| 2906 |
+
"learning_rate": 5e-05,
|
| 2907 |
+
"loss": 2.2995004653930664,
|
| 2908 |
+
"step": 828
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 0.2894154157295535,
|
| 2912 |
+
"grad_norm": 0.28268465399742126,
|
| 2913 |
+
"learning_rate": 5e-05,
|
| 2914 |
+
"loss": 2.0452728271484375,
|
| 2915 |
+
"step": 830
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 0.29011280227348013,
|
| 2919 |
+
"grad_norm": 0.2298143059015274,
|
| 2920 |
+
"learning_rate": 5e-05,
|
| 2921 |
+
"loss": 2.0092597007751465,
|
| 2922 |
+
"step": 832
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 0.2908101888174068,
|
| 2926 |
+
"grad_norm": 0.184764564037323,
|
| 2927 |
+
"learning_rate": 5e-05,
|
| 2928 |
+
"loss": 1.9803777933120728,
|
| 2929 |
+
"step": 834
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 0.2915075753613334,
|
| 2933 |
+
"grad_norm": 0.21297743916511536,
|
| 2934 |
+
"learning_rate": 5e-05,
|
| 2935 |
+
"loss": 2.1882874965667725,
|
| 2936 |
+
"step": 836
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 0.29220496190526,
|
| 2940 |
+
"grad_norm": 0.26148441433906555,
|
| 2941 |
+
"learning_rate": 5e-05,
|
| 2942 |
+
"loss": 1.9838454723358154,
|
| 2943 |
+
"step": 838
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 0.29290234844918667,
|
| 2947 |
+
"grad_norm": 0.25849202275276184,
|
| 2948 |
+
"learning_rate": 5e-05,
|
| 2949 |
+
"loss": 2.170018196105957,
|
| 2950 |
+
"step": 840
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 0.2935997349931133,
|
| 2954 |
+
"grad_norm": 0.19042696058750153,
|
| 2955 |
+
"learning_rate": 5e-05,
|
| 2956 |
+
"loss": 2.167074203491211,
|
| 2957 |
+
"step": 842
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 0.29429712153703996,
|
| 2961 |
+
"grad_norm": 0.21104653179645538,
|
| 2962 |
+
"learning_rate": 5e-05,
|
| 2963 |
+
"loss": 2.199953079223633,
|
| 2964 |
+
"step": 844
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 0.29499450808096656,
|
| 2968 |
+
"grad_norm": 0.2772143483161926,
|
| 2969 |
+
"learning_rate": 5e-05,
|
| 2970 |
+
"loss": 2.1958885192871094,
|
| 2971 |
+
"step": 846
|
| 2972 |
+
},
|
| 2973 |
+
{
|
| 2974 |
+
"epoch": 0.2956918946248932,
|
| 2975 |
+
"grad_norm": 0.3362143337726593,
|
| 2976 |
+
"learning_rate": 5e-05,
|
| 2977 |
+
"loss": 1.926414966583252,
|
| 2978 |
+
"step": 848
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"epoch": 0.29638928116881985,
|
| 2982 |
+
"grad_norm": 0.4233688712120056,
|
| 2983 |
+
"learning_rate": 5e-05,
|
| 2984 |
+
"loss": 1.7717297077178955,
|
| 2985 |
+
"step": 850
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 0.2970866677127465,
|
| 2989 |
+
"grad_norm": 0.33860331773757935,
|
| 2990 |
+
"learning_rate": 5e-05,
|
| 2991 |
+
"loss": 1.8843942880630493,
|
| 2992 |
+
"step": 852
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 0.2977840542566731,
|
| 2996 |
+
"grad_norm": 0.3185754418373108,
|
| 2997 |
+
"learning_rate": 5e-05,
|
| 2998 |
+
"loss": 2.1160013675689697,
|
| 2999 |
+
"step": 854
|
| 3000 |
+
},
|
| 3001 |
+
{
|
| 3002 |
+
"epoch": 0.29848144080059974,
|
| 3003 |
+
"grad_norm": 0.20009522140026093,
|
| 3004 |
+
"learning_rate": 5e-05,
|
| 3005 |
+
"loss": 1.9877500534057617,
|
| 3006 |
+
"step": 856
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 0.2991788273445264,
|
| 3010 |
+
"grad_norm": 0.2403642237186432,
|
| 3011 |
+
"learning_rate": 5e-05,
|
| 3012 |
+
"loss": 2.1462478637695312,
|
| 3013 |
+
"step": 858
|
| 3014 |
+
},
|
| 3015 |
+
{
|
| 3016 |
+
"epoch": 0.29987621388845304,
|
| 3017 |
+
"grad_norm": 0.21490423381328583,
|
| 3018 |
+
"learning_rate": 5e-05,
|
| 3019 |
+
"loss": 2.0985560417175293,
|
| 3020 |
+
"step": 860
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 0.3005736004323797,
|
| 3024 |
+
"grad_norm": 0.2281097024679184,
|
| 3025 |
+
"learning_rate": 5e-05,
|
| 3026 |
+
"loss": 1.953871488571167,
|
| 3027 |
+
"step": 862
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 0.3012709869763063,
|
| 3031 |
+
"grad_norm": 0.22519268095493317,
|
| 3032 |
+
"learning_rate": 5e-05,
|
| 3033 |
+
"loss": 1.9938616752624512,
|
| 3034 |
+
"step": 864
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 0.3019683735202329,
|
| 3038 |
+
"grad_norm": 0.2882140576839447,
|
| 3039 |
+
"learning_rate": 5e-05,
|
| 3040 |
+
"loss": 2.0265111923217773,
|
| 3041 |
+
"step": 866
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 0.3026657600641596,
|
| 3045 |
+
"grad_norm": 0.26522985100746155,
|
| 3046 |
+
"learning_rate": 5e-05,
|
| 3047 |
+
"loss": 1.971495270729065,
|
| 3048 |
+
"step": 868
|
| 3049 |
+
},
|
| 3050 |
+
{
|
| 3051 |
+
"epoch": 0.3033631466080862,
|
| 3052 |
+
"grad_norm": 0.2441837340593338,
|
| 3053 |
+
"learning_rate": 5e-05,
|
| 3054 |
+
"loss": 1.90657639503479,
|
| 3055 |
+
"step": 870
|
| 3056 |
+
},
|
| 3057 |
+
{
|
| 3058 |
+
"epoch": 0.3040605331520128,
|
| 3059 |
+
"grad_norm": 0.33309900760650635,
|
| 3060 |
+
"learning_rate": 5e-05,
|
| 3061 |
+
"loss": 2.001520872116089,
|
| 3062 |
+
"step": 872
|
| 3063 |
+
},
|
| 3064 |
+
{
|
| 3065 |
+
"epoch": 0.30475791969593946,
|
| 3066 |
+
"grad_norm": 0.2041439563035965,
|
| 3067 |
+
"learning_rate": 5e-05,
|
| 3068 |
+
"loss": 1.968753695487976,
|
| 3069 |
+
"step": 874
|
| 3070 |
+
},
|
| 3071 |
+
{
|
| 3072 |
+
"epoch": 0.3054553062398661,
|
| 3073 |
+
"grad_norm": 0.17259293794631958,
|
| 3074 |
+
"learning_rate": 5e-05,
|
| 3075 |
+
"loss": 1.9523271322250366,
|
| 3076 |
+
"step": 876
|
| 3077 |
+
},
|
| 3078 |
+
{
|
| 3079 |
+
"epoch": 0.30615269278379276,
|
| 3080 |
+
"grad_norm": 0.1741812825202942,
|
| 3081 |
+
"learning_rate": 5e-05,
|
| 3082 |
+
"loss": 2.1121654510498047,
|
| 3083 |
+
"step": 878
|
| 3084 |
+
},
|
| 3085 |
+
{
|
| 3086 |
+
"epoch": 0.30685007932771935,
|
| 3087 |
+
"grad_norm": 0.23253613710403442,
|
| 3088 |
+
"learning_rate": 5e-05,
|
| 3089 |
+
"loss": 2.213024616241455,
|
| 3090 |
+
"step": 880
|
| 3091 |
+
},
|
| 3092 |
+
{
|
| 3093 |
+
"epoch": 0.307547465871646,
|
| 3094 |
+
"grad_norm": 0.2259003072977066,
|
| 3095 |
+
"learning_rate": 5e-05,
|
| 3096 |
+
"loss": 2.094529151916504,
|
| 3097 |
+
"step": 882
|
| 3098 |
+
},
|
| 3099 |
+
{
|
| 3100 |
+
"epoch": 0.30824485241557265,
|
| 3101 |
+
"grad_norm": 0.22867822647094727,
|
| 3102 |
+
"learning_rate": 5e-05,
|
| 3103 |
+
"loss": 2.0620737075805664,
|
| 3104 |
+
"step": 884
|
| 3105 |
+
},
|
| 3106 |
+
{
|
| 3107 |
+
"epoch": 0.3089422389594993,
|
| 3108 |
+
"grad_norm": 0.17488710582256317,
|
| 3109 |
+
"learning_rate": 5e-05,
|
| 3110 |
+
"loss": 1.985865592956543,
|
| 3111 |
+
"step": 886
|
| 3112 |
+
},
|
| 3113 |
+
{
|
| 3114 |
+
"epoch": 0.3096396255034259,
|
| 3115 |
+
"grad_norm": 0.28305572271347046,
|
| 3116 |
+
"learning_rate": 5e-05,
|
| 3117 |
+
"loss": 2.090651273727417,
|
| 3118 |
+
"step": 888
|
| 3119 |
+
},
|
| 3120 |
+
{
|
| 3121 |
+
"epoch": 0.31033701204735253,
|
| 3122 |
+
"grad_norm": 0.258344829082489,
|
| 3123 |
+
"learning_rate": 5e-05,
|
| 3124 |
+
"loss": 1.9839575290679932,
|
| 3125 |
+
"step": 890
|
| 3126 |
+
},
|
| 3127 |
+
{
|
| 3128 |
+
"epoch": 0.3110343985912792,
|
| 3129 |
+
"grad_norm": 0.2704688310623169,
|
| 3130 |
+
"learning_rate": 5e-05,
|
| 3131 |
+
"loss": 1.9838356971740723,
|
| 3132 |
+
"step": 892
|
| 3133 |
+
},
|
| 3134 |
+
{
|
| 3135 |
+
"epoch": 0.31173178513520583,
|
| 3136 |
+
"grad_norm": 0.21489857137203217,
|
| 3137 |
+
"learning_rate": 5e-05,
|
| 3138 |
+
"loss": 2.0196244716644287,
|
| 3139 |
+
"step": 894
|
| 3140 |
+
},
|
| 3141 |
+
{
|
| 3142 |
+
"epoch": 0.3124291716791325,
|
| 3143 |
+
"grad_norm": 0.42650023102760315,
|
| 3144 |
+
"learning_rate": 5e-05,
|
| 3145 |
+
"loss": 2.060972213745117,
|
| 3146 |
+
"step": 896
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 0.31312655822305907,
|
| 3150 |
+
"grad_norm": 0.17513501644134521,
|
| 3151 |
+
"learning_rate": 5e-05,
|
| 3152 |
+
"loss": 2.140475273132324,
|
| 3153 |
+
"step": 898
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 0.3138239447669857,
|
| 3157 |
+
"grad_norm": 0.2620815336704254,
|
| 3158 |
+
"learning_rate": 5e-05,
|
| 3159 |
+
"loss": 2.194441318511963,
|
| 3160 |
+
"step": 900
|
| 3161 |
+
},
|
| 3162 |
+
{
|
| 3163 |
+
"epoch": 0.31452133131091237,
|
| 3164 |
+
"grad_norm": 0.2540329396724701,
|
| 3165 |
+
"learning_rate": 5e-05,
|
| 3166 |
+
"loss": 1.9437588453292847,
|
| 3167 |
+
"step": 902
|
| 3168 |
+
},
|
| 3169 |
+
{
|
| 3170 |
+
"epoch": 0.315218717854839,
|
| 3171 |
+
"grad_norm": 0.19750094413757324,
|
| 3172 |
+
"learning_rate": 5e-05,
|
| 3173 |
+
"loss": 2.0057075023651123,
|
| 3174 |
+
"step": 904
|
| 3175 |
+
},
|
| 3176 |
+
{
|
| 3177 |
+
"epoch": 0.3159161043987656,
|
| 3178 |
+
"grad_norm": 0.23293620347976685,
|
| 3179 |
+
"learning_rate": 5e-05,
|
| 3180 |
+
"loss": 1.8982203006744385,
|
| 3181 |
+
"step": 906
|
| 3182 |
+
},
|
| 3183 |
+
{
|
| 3184 |
+
"epoch": 0.31661349094269226,
|
| 3185 |
+
"grad_norm": 0.2062148004770279,
|
| 3186 |
+
"learning_rate": 5e-05,
|
| 3187 |
+
"loss": 2.0358047485351562,
|
| 3188 |
+
"step": 908
|
| 3189 |
+
},
|
| 3190 |
+
{
|
| 3191 |
+
"epoch": 0.3173108774866189,
|
| 3192 |
+
"grad_norm": 0.26289623975753784,
|
| 3193 |
+
"learning_rate": 5e-05,
|
| 3194 |
+
"loss": 1.795961618423462,
|
| 3195 |
+
"step": 910
|
| 3196 |
+
},
|
| 3197 |
+
{
|
| 3198 |
+
"epoch": 0.31800826403054555,
|
| 3199 |
+
"grad_norm": 0.2925187349319458,
|
| 3200 |
+
"learning_rate": 5e-05,
|
| 3201 |
+
"loss": 1.9860413074493408,
|
| 3202 |
+
"step": 912
|
| 3203 |
+
},
|
| 3204 |
+
{
|
| 3205 |
+
"epoch": 0.31870565057447214,
|
| 3206 |
+
"grad_norm": 0.2811063826084137,
|
| 3207 |
+
"learning_rate": 5e-05,
|
| 3208 |
+
"loss": 1.8662142753601074,
|
| 3209 |
+
"step": 914
|
| 3210 |
+
},
|
| 3211 |
+
{
|
| 3212 |
+
"epoch": 0.3194030371183988,
|
| 3213 |
+
"grad_norm": 0.22942379117012024,
|
| 3214 |
+
"learning_rate": 5e-05,
|
| 3215 |
+
"loss": 2.0050673484802246,
|
| 3216 |
+
"step": 916
|
| 3217 |
+
},
|
| 3218 |
+
{
|
| 3219 |
+
"epoch": 0.32010042366232544,
|
| 3220 |
+
"grad_norm": 0.16354793310165405,
|
| 3221 |
+
"learning_rate": 5e-05,
|
| 3222 |
+
"loss": 2.1945548057556152,
|
| 3223 |
+
"step": 918
|
| 3224 |
+
},
|
| 3225 |
+
{
|
| 3226 |
+
"epoch": 0.3207978102062521,
|
| 3227 |
+
"grad_norm": 0.23471316695213318,
|
| 3228 |
+
"learning_rate": 5e-05,
|
| 3229 |
+
"loss": 2.2042596340179443,
|
| 3230 |
+
"step": 920
|
| 3231 |
+
},
|
| 3232 |
+
{
|
| 3233 |
+
"epoch": 0.3214951967501787,
|
| 3234 |
+
"grad_norm": 0.2259097695350647,
|
| 3235 |
+
"learning_rate": 5e-05,
|
| 3236 |
+
"loss": 2.16628360748291,
|
| 3237 |
+
"step": 922
|
| 3238 |
+
},
|
| 3239 |
+
{
|
| 3240 |
+
"epoch": 0.32219258329410533,
|
| 3241 |
+
"grad_norm": 0.2739982008934021,
|
| 3242 |
+
"learning_rate": 5e-05,
|
| 3243 |
+
"loss": 2.055208206176758,
|
| 3244 |
+
"step": 924
|
| 3245 |
+
},
|
| 3246 |
+
{
|
| 3247 |
+
"epoch": 0.322889969838032,
|
| 3248 |
+
"grad_norm": 0.3012983798980713,
|
| 3249 |
+
"learning_rate": 5e-05,
|
| 3250 |
+
"loss": 2.2179489135742188,
|
| 3251 |
+
"step": 926
|
| 3252 |
+
},
|
| 3253 |
+
{
|
| 3254 |
+
"epoch": 0.3235873563819586,
|
| 3255 |
+
"grad_norm": 0.23193614184856415,
|
| 3256 |
+
"learning_rate": 5e-05,
|
| 3257 |
+
"loss": 2.104721784591675,
|
| 3258 |
+
"step": 928
|
| 3259 |
+
},
|
| 3260 |
+
{
|
| 3261 |
+
"epoch": 0.32428474292588527,
|
| 3262 |
+
"grad_norm": 0.19617155194282532,
|
| 3263 |
+
"learning_rate": 5e-05,
|
| 3264 |
+
"loss": 1.9976686239242554,
|
| 3265 |
+
"step": 930
|
| 3266 |
+
},
|
| 3267 |
+
{
|
| 3268 |
+
"epoch": 0.32498212946981186,
|
| 3269 |
+
"grad_norm": 0.22112447023391724,
|
| 3270 |
+
"learning_rate": 5e-05,
|
| 3271 |
+
"loss": 2.2046685218811035,
|
| 3272 |
+
"step": 932
|
| 3273 |
+
},
|
| 3274 |
+
{
|
| 3275 |
+
"epoch": 0.3256795160137385,
|
| 3276 |
+
"grad_norm": 0.19225233793258667,
|
| 3277 |
+
"learning_rate": 5e-05,
|
| 3278 |
+
"loss": 1.9811292886734009,
|
| 3279 |
+
"step": 934
|
| 3280 |
+
},
|
| 3281 |
+
{
|
| 3282 |
+
"epoch": 0.32637690255766516,
|
| 3283 |
+
"grad_norm": 0.16594399511814117,
|
| 3284 |
+
"learning_rate": 5e-05,
|
| 3285 |
+
"loss": 2.0854458808898926,
|
| 3286 |
+
"step": 936
|
| 3287 |
+
},
|
| 3288 |
+
{
|
| 3289 |
+
"epoch": 0.3270742891015918,
|
| 3290 |
+
"grad_norm": 0.22566556930541992,
|
| 3291 |
+
"learning_rate": 5e-05,
|
| 3292 |
+
"loss": 2.0371317863464355,
|
| 3293 |
+
"step": 938
|
| 3294 |
+
},
|
| 3295 |
+
{
|
| 3296 |
+
"epoch": 0.3277716756455184,
|
| 3297 |
+
"grad_norm": 0.2621348798274994,
|
| 3298 |
+
"learning_rate": 5e-05,
|
| 3299 |
+
"loss": 2.2924721240997314,
|
| 3300 |
+
"step": 940
|
| 3301 |
+
},
|
| 3302 |
+
{
|
| 3303 |
+
"epoch": 0.32846906218944505,
|
| 3304 |
+
"grad_norm": 0.3321608006954193,
|
| 3305 |
+
"learning_rate": 5e-05,
|
| 3306 |
+
"loss": 1.9144505262374878,
|
| 3307 |
+
"step": 942
|
| 3308 |
+
},
|
| 3309 |
+
{
|
| 3310 |
+
"epoch": 0.3291664487333717,
|
| 3311 |
+
"grad_norm": 0.21470032632350922,
|
| 3312 |
+
"learning_rate": 5e-05,
|
| 3313 |
+
"loss": 2.0083987712860107,
|
| 3314 |
+
"step": 944
|
| 3315 |
+
},
|
| 3316 |
+
{
|
| 3317 |
+
"epoch": 0.32986383527729835,
|
| 3318 |
+
"grad_norm": 0.17992348968982697,
|
| 3319 |
+
"learning_rate": 5e-05,
|
| 3320 |
+
"loss": 2.096499443054199,
|
| 3321 |
+
"step": 946
|
| 3322 |
+
},
|
| 3323 |
+
{
|
| 3324 |
+
"epoch": 0.33056122182122494,
|
| 3325 |
+
"grad_norm": 0.191554456949234,
|
| 3326 |
+
"learning_rate": 5e-05,
|
| 3327 |
+
"loss": 1.956010103225708,
|
| 3328 |
+
"step": 948
|
| 3329 |
+
},
|
| 3330 |
+
{
|
| 3331 |
+
"epoch": 0.3312586083651516,
|
| 3332 |
+
"grad_norm": 0.2133859097957611,
|
| 3333 |
+
"learning_rate": 5e-05,
|
| 3334 |
+
"loss": 1.908299446105957,
|
| 3335 |
+
"step": 950
|
| 3336 |
+
},
|
| 3337 |
+
{
|
| 3338 |
+
"epoch": 0.33195599490907823,
|
| 3339 |
+
"grad_norm": 0.22485746443271637,
|
| 3340 |
+
"learning_rate": 5e-05,
|
| 3341 |
+
"loss": 1.7729823589324951,
|
| 3342 |
+
"step": 952
|
| 3343 |
+
},
|
| 3344 |
+
{
|
| 3345 |
+
"epoch": 0.3326533814530049,
|
| 3346 |
+
"grad_norm": 0.186967670917511,
|
| 3347 |
+
"learning_rate": 5e-05,
|
| 3348 |
+
"loss": 1.9951062202453613,
|
| 3349 |
+
"step": 954
|
| 3350 |
+
},
|
| 3351 |
+
{
|
| 3352 |
+
"epoch": 0.3333507679969315,
|
| 3353 |
+
"grad_norm": 0.26257404685020447,
|
| 3354 |
+
"learning_rate": 5e-05,
|
| 3355 |
+
"loss": 2.0230295658111572,
|
| 3356 |
+
"step": 956
|
| 3357 |
+
},
|
| 3358 |
+
{
|
| 3359 |
+
"epoch": 0.3340481545408581,
|
| 3360 |
+
"grad_norm": 0.2617090046405792,
|
| 3361 |
+
"learning_rate": 5e-05,
|
| 3362 |
+
"loss": 1.9331642389297485,
|
| 3363 |
+
"step": 958
|
| 3364 |
+
},
|
| 3365 |
+
{
|
| 3366 |
+
"epoch": 0.33474554108478477,
|
| 3367 |
+
"grad_norm": 0.20375195145606995,
|
| 3368 |
+
"learning_rate": 5e-05,
|
| 3369 |
+
"loss": 1.9368494749069214,
|
| 3370 |
+
"step": 960
|
| 3371 |
+
},
|
| 3372 |
+
{
|
| 3373 |
+
"epoch": 0.3354429276287114,
|
| 3374 |
+
"grad_norm": 0.1918850988149643,
|
| 3375 |
+
"learning_rate": 5e-05,
|
| 3376 |
+
"loss": 2.1231679916381836,
|
| 3377 |
+
"step": 962
|
| 3378 |
+
},
|
| 3379 |
+
{
|
| 3380 |
+
"epoch": 0.33614031417263807,
|
| 3381 |
+
"grad_norm": 0.26361358165740967,
|
| 3382 |
+
"learning_rate": 5e-05,
|
| 3383 |
+
"loss": 2.1133885383605957,
|
| 3384 |
+
"step": 964
|
| 3385 |
+
},
|
| 3386 |
+
{
|
| 3387 |
+
"epoch": 0.33683770071656466,
|
| 3388 |
+
"grad_norm": 0.19686871767044067,
|
| 3389 |
+
"learning_rate": 5e-05,
|
| 3390 |
+
"loss": 2.07720685005188,
|
| 3391 |
+
"step": 966
|
| 3392 |
+
},
|
| 3393 |
+
{
|
| 3394 |
+
"epoch": 0.3375350872604913,
|
| 3395 |
+
"grad_norm": 0.32117438316345215,
|
| 3396 |
+
"learning_rate": 5e-05,
|
| 3397 |
+
"loss": 2.0643956661224365,
|
| 3398 |
+
"step": 968
|
| 3399 |
+
},
|
| 3400 |
+
{
|
| 3401 |
+
"epoch": 0.33823247380441795,
|
| 3402 |
+
"grad_norm": 0.2792845368385315,
|
| 3403 |
+
"learning_rate": 5e-05,
|
| 3404 |
+
"loss": 2.0637881755828857,
|
| 3405 |
+
"step": 970
|
| 3406 |
+
},
|
| 3407 |
+
{
|
| 3408 |
+
"epoch": 0.3389298603483446,
|
| 3409 |
+
"grad_norm": 0.21377836167812347,
|
| 3410 |
+
"learning_rate": 5e-05,
|
| 3411 |
+
"loss": 2.150491714477539,
|
| 3412 |
+
"step": 972
|
| 3413 |
+
},
|
| 3414 |
+
{
|
| 3415 |
+
"epoch": 0.3396272468922712,
|
| 3416 |
+
"grad_norm": 0.22636762261390686,
|
| 3417 |
+
"learning_rate": 5e-05,
|
| 3418 |
+
"loss": 1.9907886981964111,
|
| 3419 |
+
"step": 974
|
| 3420 |
+
},
|
| 3421 |
+
{
|
| 3422 |
+
"epoch": 0.34032463343619784,
|
| 3423 |
+
"grad_norm": 0.2062670737504959,
|
| 3424 |
+
"learning_rate": 5e-05,
|
| 3425 |
+
"loss": 1.9640690088272095,
|
| 3426 |
+
"step": 976
|
| 3427 |
+
},
|
| 3428 |
+
{
|
| 3429 |
+
"epoch": 0.3410220199801245,
|
| 3430 |
+
"grad_norm": 0.2347593754529953,
|
| 3431 |
+
"learning_rate": 5e-05,
|
| 3432 |
+
"loss": 2.0291075706481934,
|
| 3433 |
+
"step": 978
|
| 3434 |
+
},
|
| 3435 |
+
{
|
| 3436 |
+
"epoch": 0.34171940652405114,
|
| 3437 |
+
"grad_norm": 0.22571015357971191,
|
| 3438 |
+
"learning_rate": 5e-05,
|
| 3439 |
+
"loss": 2.0671143531799316,
|
| 3440 |
+
"step": 980
|
| 3441 |
+
},
|
| 3442 |
+
{
|
| 3443 |
+
"epoch": 0.34241679306797773,
|
| 3444 |
+
"grad_norm": 0.2145722508430481,
|
| 3445 |
+
"learning_rate": 5e-05,
|
| 3446 |
+
"loss": 2.1719367504119873,
|
| 3447 |
+
"step": 982
|
| 3448 |
+
},
|
| 3449 |
+
{
|
| 3450 |
+
"epoch": 0.3431141796119044,
|
| 3451 |
+
"grad_norm": 0.25082701444625854,
|
| 3452 |
+
"learning_rate": 5e-05,
|
| 3453 |
+
"loss": 2.0923759937286377,
|
| 3454 |
+
"step": 984
|
| 3455 |
+
},
|
| 3456 |
+
{
|
| 3457 |
+
"epoch": 0.343811566155831,
|
| 3458 |
+
"grad_norm": 0.22755537927150726,
|
| 3459 |
+
"learning_rate": 5e-05,
|
| 3460 |
+
"loss": 1.9666999578475952,
|
| 3461 |
+
"step": 986
|
| 3462 |
+
},
|
| 3463 |
+
{
|
| 3464 |
+
"epoch": 0.3445089526997577,
|
| 3465 |
+
"grad_norm": 0.2901584804058075,
|
| 3466 |
+
"learning_rate": 5e-05,
|
| 3467 |
+
"loss": 1.8081202507019043,
|
| 3468 |
+
"step": 988
|
| 3469 |
+
},
|
| 3470 |
+
{
|
| 3471 |
+
"epoch": 0.34520633924368427,
|
| 3472 |
+
"grad_norm": 0.23690833151340485,
|
| 3473 |
+
"learning_rate": 5e-05,
|
| 3474 |
+
"loss": 2.040344476699829,
|
| 3475 |
+
"step": 990
|
| 3476 |
+
},
|
| 3477 |
+
{
|
| 3478 |
+
"epoch": 0.3459037257876109,
|
| 3479 |
+
"grad_norm": 0.21881553530693054,
|
| 3480 |
+
"learning_rate": 5e-05,
|
| 3481 |
+
"loss": 1.7107083797454834,
|
| 3482 |
+
"step": 992
|
| 3483 |
+
},
|
| 3484 |
+
{
|
| 3485 |
+
"epoch": 0.34660111233153756,
|
| 3486 |
+
"grad_norm": 0.26182490587234497,
|
| 3487 |
+
"learning_rate": 5e-05,
|
| 3488 |
+
"loss": 2.064363479614258,
|
| 3489 |
+
"step": 994
|
| 3490 |
+
},
|
| 3491 |
+
{
|
| 3492 |
+
"epoch": 0.3472984988754642,
|
| 3493 |
+
"grad_norm": 0.2080112099647522,
|
| 3494 |
+
"learning_rate": 5e-05,
|
| 3495 |
+
"loss": 2.166213274002075,
|
| 3496 |
+
"step": 996
|
| 3497 |
+
},
|
| 3498 |
+
{
|
| 3499 |
+
"epoch": 0.34799588541939086,
|
| 3500 |
+
"grad_norm": 0.2145555168390274,
|
| 3501 |
+
"learning_rate": 5e-05,
|
| 3502 |
+
"loss": 1.9562971591949463,
|
| 3503 |
+
"step": 998
|
| 3504 |
+
},
|
| 3505 |
+
{
|
| 3506 |
+
"epoch": 0.34869327196331745,
|
| 3507 |
+
"grad_norm": 0.21929831802845,
|
| 3508 |
+
"learning_rate": 5e-05,
|
| 3509 |
+
"loss": 2.0631940364837646,
|
| 3510 |
+
"step": 1000
|
| 3511 |
+
},
|
| 3512 |
+
{
|
| 3513 |
+
"epoch": 0.3493906585072441,
|
| 3514 |
+
"grad_norm": 0.2061082273721695,
|
| 3515 |
+
"learning_rate": 5e-05,
|
| 3516 |
+
"loss": 1.9821693897247314,
|
| 3517 |
+
"step": 1002
|
| 3518 |
+
},
|
| 3519 |
+
{
|
| 3520 |
+
"epoch": 0.35008804505117075,
|
| 3521 |
+
"grad_norm": 0.21569553017616272,
|
| 3522 |
+
"learning_rate": 5e-05,
|
| 3523 |
+
"loss": 2.1045546531677246,
|
| 3524 |
+
"step": 1004
|
| 3525 |
+
},
|
| 3526 |
+
{
|
| 3527 |
+
"epoch": 0.3507854315950974,
|
| 3528 |
+
"grad_norm": 0.20195575058460236,
|
| 3529 |
+
"learning_rate": 5e-05,
|
| 3530 |
+
"loss": 2.023355484008789,
|
| 3531 |
+
"step": 1006
|
| 3532 |
+
},
|
| 3533 |
+
{
|
| 3534 |
+
"epoch": 0.351482818139024,
|
| 3535 |
+
"grad_norm": 0.21589083969593048,
|
| 3536 |
+
"learning_rate": 5e-05,
|
| 3537 |
+
"loss": 1.9698195457458496,
|
| 3538 |
+
"step": 1008
|
| 3539 |
+
},
|
| 3540 |
+
{
|
| 3541 |
+
"epoch": 0.35218020468295064,
|
| 3542 |
+
"grad_norm": 0.35657018423080444,
|
| 3543 |
+
"learning_rate": 5e-05,
|
| 3544 |
+
"loss": 2.0477797985076904,
|
| 3545 |
+
"step": 1010
|
| 3546 |
+
},
|
| 3547 |
+
{
|
| 3548 |
+
"epoch": 0.3528775912268773,
|
| 3549 |
+
"grad_norm": 0.2049272358417511,
|
| 3550 |
+
"learning_rate": 5e-05,
|
| 3551 |
+
"loss": 2.010043144226074,
|
| 3552 |
+
"step": 1012
|
| 3553 |
+
},
|
| 3554 |
+
{
|
| 3555 |
+
"epoch": 0.35357497777080393,
|
| 3556 |
+
"grad_norm": 0.1918814480304718,
|
| 3557 |
+
"learning_rate": 5e-05,
|
| 3558 |
+
"loss": 1.9827858209609985,
|
| 3559 |
+
"step": 1014
|
| 3560 |
+
},
|
| 3561 |
+
{
|
| 3562 |
+
"epoch": 0.3542723643147305,
|
| 3563 |
+
"grad_norm": 0.22987155616283417,
|
| 3564 |
+
"learning_rate": 5e-05,
|
| 3565 |
+
"loss": 1.987635612487793,
|
| 3566 |
+
"step": 1016
|
| 3567 |
+
},
|
| 3568 |
+
{
|
| 3569 |
+
"epoch": 0.3549697508586572,
|
| 3570 |
+
"grad_norm": 0.24195624887943268,
|
| 3571 |
+
"learning_rate": 5e-05,
|
| 3572 |
+
"loss": 1.9928193092346191,
|
| 3573 |
+
"step": 1018
|
| 3574 |
+
},
|
| 3575 |
+
{
|
| 3576 |
+
"epoch": 0.3556671374025838,
|
| 3577 |
+
"grad_norm": 0.27684086561203003,
|
| 3578 |
+
"learning_rate": 5e-05,
|
| 3579 |
+
"loss": 1.9530699253082275,
|
| 3580 |
+
"step": 1020
|
| 3581 |
+
},
|
| 3582 |
+
{
|
| 3583 |
+
"epoch": 0.35636452394651047,
|
| 3584 |
+
"grad_norm": 0.21984905004501343,
|
| 3585 |
+
"learning_rate": 5e-05,
|
| 3586 |
+
"loss": 2.0691213607788086,
|
| 3587 |
+
"step": 1022
|
| 3588 |
+
},
|
| 3589 |
+
{
|
| 3590 |
+
"epoch": 0.35706191049043706,
|
| 3591 |
+
"grad_norm": 0.249387726187706,
|
| 3592 |
+
"learning_rate": 5e-05,
|
| 3593 |
+
"loss": 2.183511734008789,
|
| 3594 |
+
"step": 1024
|
| 3595 |
+
},
|
| 3596 |
+
{
|
| 3597 |
+
"epoch": 0.3577592970343637,
|
| 3598 |
+
"grad_norm": 0.23080800473690033,
|
| 3599 |
+
"learning_rate": 5e-05,
|
| 3600 |
+
"loss": 1.930666446685791,
|
| 3601 |
+
"step": 1026
|
| 3602 |
+
},
|
| 3603 |
+
{
|
| 3604 |
+
"epoch": 0.35845668357829036,
|
| 3605 |
+
"grad_norm": 0.23000819981098175,
|
| 3606 |
+
"learning_rate": 5e-05,
|
| 3607 |
+
"loss": 2.092984199523926,
|
| 3608 |
+
"step": 1028
|
| 3609 |
+
},
|
| 3610 |
+
{
|
| 3611 |
+
"epoch": 0.359154070122217,
|
| 3612 |
+
"grad_norm": 0.1848149597644806,
|
| 3613 |
+
"learning_rate": 5e-05,
|
| 3614 |
+
"loss": 2.0390803813934326,
|
| 3615 |
+
"step": 1030
|
| 3616 |
+
},
|
| 3617 |
+
{
|
| 3618 |
+
"epoch": 0.35985145666614365,
|
| 3619 |
+
"grad_norm": 0.27325889468193054,
|
| 3620 |
+
"learning_rate": 5e-05,
|
| 3621 |
+
"loss": 1.9925591945648193,
|
| 3622 |
+
"step": 1032
|
| 3623 |
+
},
|
| 3624 |
+
{
|
| 3625 |
+
"epoch": 0.36054884321007025,
|
| 3626 |
+
"grad_norm": 0.2851124405860901,
|
| 3627 |
+
"learning_rate": 5e-05,
|
| 3628 |
+
"loss": 2.2212448120117188,
|
| 3629 |
+
"step": 1034
|
| 3630 |
+
},
|
| 3631 |
+
{
|
| 3632 |
+
"epoch": 0.3612462297539969,
|
| 3633 |
+
"grad_norm": 0.357220321893692,
|
| 3634 |
+
"learning_rate": 5e-05,
|
| 3635 |
+
"loss": 1.5241745710372925,
|
| 3636 |
+
"step": 1036
|
| 3637 |
+
},
|
| 3638 |
+
{
|
| 3639 |
+
"epoch": 0.36194361629792354,
|
| 3640 |
+
"grad_norm": 0.28451353311538696,
|
| 3641 |
+
"learning_rate": 5e-05,
|
| 3642 |
+
"loss": 2.13199782371521,
|
| 3643 |
+
"step": 1038
|
| 3644 |
+
},
|
| 3645 |
+
{
|
| 3646 |
+
"epoch": 0.3626410028418502,
|
| 3647 |
+
"grad_norm": 0.17099227011203766,
|
| 3648 |
+
"learning_rate": 5e-05,
|
| 3649 |
+
"loss": 1.9190986156463623,
|
| 3650 |
+
"step": 1040
|
| 3651 |
+
},
|
| 3652 |
+
{
|
| 3653 |
+
"epoch": 0.3633383893857768,
|
| 3654 |
+
"grad_norm": 0.27135729789733887,
|
| 3655 |
+
"learning_rate": 5e-05,
|
| 3656 |
+
"loss": 2.1251871585845947,
|
| 3657 |
+
"step": 1042
|
| 3658 |
+
},
|
| 3659 |
+
{
|
| 3660 |
+
"epoch": 0.36403577592970343,
|
| 3661 |
+
"grad_norm": 0.28680723905563354,
|
| 3662 |
+
"learning_rate": 5e-05,
|
| 3663 |
+
"loss": 1.6807236671447754,
|
| 3664 |
+
"step": 1044
|
| 3665 |
+
},
|
| 3666 |
+
{
|
| 3667 |
+
"epoch": 0.3647331624736301,
|
| 3668 |
+
"grad_norm": 0.2310262769460678,
|
| 3669 |
+
"learning_rate": 5e-05,
|
| 3670 |
+
"loss": 2.03737735748291,
|
| 3671 |
+
"step": 1046
|
| 3672 |
+
},
|
| 3673 |
+
{
|
| 3674 |
+
"epoch": 0.3654305490175567,
|
| 3675 |
+
"grad_norm": 0.19282707571983337,
|
| 3676 |
+
"learning_rate": 5e-05,
|
| 3677 |
+
"loss": 2.1107892990112305,
|
| 3678 |
+
"step": 1048
|
| 3679 |
+
},
|
| 3680 |
+
{
|
| 3681 |
+
"epoch": 0.3661279355614833,
|
| 3682 |
+
"grad_norm": 0.22267811000347137,
|
| 3683 |
+
"learning_rate": 5e-05,
|
| 3684 |
+
"loss": 1.990861415863037,
|
| 3685 |
+
"step": 1050
|
| 3686 |
+
},
|
| 3687 |
+
{
|
| 3688 |
+
"epoch": 0.36682532210540997,
|
| 3689 |
+
"grad_norm": 0.23680366575717926,
|
| 3690 |
+
"learning_rate": 5e-05,
|
| 3691 |
+
"loss": 1.9701367616653442,
|
| 3692 |
+
"step": 1052
|
| 3693 |
+
},
|
| 3694 |
+
{
|
| 3695 |
+
"epoch": 0.3675227086493366,
|
| 3696 |
+
"grad_norm": 0.2254926711320877,
|
| 3697 |
+
"learning_rate": 5e-05,
|
| 3698 |
+
"loss": 1.8134162425994873,
|
| 3699 |
+
"step": 1054
|
| 3700 |
+
},
|
| 3701 |
+
{
|
| 3702 |
+
"epoch": 0.36822009519326326,
|
| 3703 |
+
"grad_norm": 0.22644150257110596,
|
| 3704 |
+
"learning_rate": 5e-05,
|
| 3705 |
+
"loss": 1.911505937576294,
|
| 3706 |
+
"step": 1056
|
| 3707 |
+
},
|
| 3708 |
+
{
|
| 3709 |
+
"epoch": 0.36891748173718986,
|
| 3710 |
+
"grad_norm": 0.1897391378879547,
|
| 3711 |
+
"learning_rate": 5e-05,
|
| 3712 |
+
"loss": 2.0960769653320312,
|
| 3713 |
+
"step": 1058
|
| 3714 |
+
},
|
| 3715 |
+
{
|
| 3716 |
+
"epoch": 0.3696148682811165,
|
| 3717 |
+
"grad_norm": 0.26435860991477966,
|
| 3718 |
+
"learning_rate": 5e-05,
|
| 3719 |
+
"loss": 2.0274524688720703,
|
| 3720 |
+
"step": 1060
|
| 3721 |
+
},
|
| 3722 |
+
{
|
| 3723 |
+
"epoch": 0.37031225482504315,
|
| 3724 |
+
"grad_norm": 0.33470162749290466,
|
| 3725 |
+
"learning_rate": 5e-05,
|
| 3726 |
+
"loss": 1.8201732635498047,
|
| 3727 |
+
"step": 1062
|
| 3728 |
+
},
|
| 3729 |
+
{
|
| 3730 |
+
"epoch": 0.3710096413689698,
|
| 3731 |
+
"grad_norm": 0.22211971879005432,
|
| 3732 |
+
"learning_rate": 5e-05,
|
| 3733 |
+
"loss": 1.9741380214691162,
|
| 3734 |
+
"step": 1064
|
| 3735 |
+
},
|
| 3736 |
+
{
|
| 3737 |
+
"epoch": 0.37170702791289645,
|
| 3738 |
+
"grad_norm": 0.38349875807762146,
|
| 3739 |
+
"learning_rate": 5e-05,
|
| 3740 |
+
"loss": 2.1487531661987305,
|
| 3741 |
+
"step": 1066
|
| 3742 |
+
},
|
| 3743 |
+
{
|
| 3744 |
+
"epoch": 0.37240441445682304,
|
| 3745 |
+
"grad_norm": 0.2047680765390396,
|
| 3746 |
+
"learning_rate": 5e-05,
|
| 3747 |
+
"loss": 2.2522268295288086,
|
| 3748 |
+
"step": 1068
|
| 3749 |
+
},
|
| 3750 |
+
{
|
| 3751 |
+
"epoch": 0.3731018010007497,
|
| 3752 |
+
"grad_norm": 0.2490031123161316,
|
| 3753 |
+
"learning_rate": 5e-05,
|
| 3754 |
+
"loss": 2.051534652709961,
|
| 3755 |
+
"step": 1070
|
| 3756 |
+
},
|
| 3757 |
+
{
|
| 3758 |
+
"epoch": 0.37379918754467634,
|
| 3759 |
+
"grad_norm": 0.22337035834789276,
|
| 3760 |
+
"learning_rate": 5e-05,
|
| 3761 |
+
"loss": 2.099733829498291,
|
| 3762 |
+
"step": 1072
|
| 3763 |
+
},
|
| 3764 |
+
{
|
| 3765 |
+
"epoch": 0.374496574088603,
|
| 3766 |
+
"grad_norm": 0.22020231187343597,
|
| 3767 |
+
"learning_rate": 5e-05,
|
| 3768 |
+
"loss": 2.078166961669922,
|
| 3769 |
+
"step": 1074
|
| 3770 |
+
},
|
| 3771 |
+
{
|
| 3772 |
+
"epoch": 0.3751939606325296,
|
| 3773 |
+
"grad_norm": 0.26717209815979004,
|
| 3774 |
+
"learning_rate": 5e-05,
|
| 3775 |
+
"loss": 1.8090195655822754,
|
| 3776 |
+
"step": 1076
|
| 3777 |
+
},
|
| 3778 |
+
{
|
| 3779 |
+
"epoch": 0.3758913471764562,
|
| 3780 |
+
"grad_norm": 0.1886700689792633,
|
| 3781 |
+
"learning_rate": 5e-05,
|
| 3782 |
+
"loss": 2.0888776779174805,
|
| 3783 |
+
"step": 1078
|
| 3784 |
+
},
|
| 3785 |
+
{
|
| 3786 |
+
"epoch": 0.3765887337203829,
|
| 3787 |
+
"grad_norm": 0.2425251603126526,
|
| 3788 |
+
"learning_rate": 5e-05,
|
| 3789 |
+
"loss": 1.9623768329620361,
|
| 3790 |
+
"step": 1080
|
| 3791 |
+
},
|
| 3792 |
+
{
|
| 3793 |
+
"epoch": 0.3772861202643095,
|
| 3794 |
+
"grad_norm": 0.22117263078689575,
|
| 3795 |
+
"learning_rate": 5e-05,
|
| 3796 |
+
"loss": 2.085573196411133,
|
| 3797 |
+
"step": 1082
|
| 3798 |
+
},
|
| 3799 |
+
{
|
| 3800 |
+
"epoch": 0.3779835068082361,
|
| 3801 |
+
"grad_norm": 0.2838881015777588,
|
| 3802 |
+
"learning_rate": 5e-05,
|
| 3803 |
+
"loss": 2.0469043254852295,
|
| 3804 |
+
"step": 1084
|
| 3805 |
+
},
|
| 3806 |
+
{
|
| 3807 |
+
"epoch": 0.37868089335216276,
|
| 3808 |
+
"grad_norm": 0.1819688081741333,
|
| 3809 |
+
"learning_rate": 5e-05,
|
| 3810 |
+
"loss": 1.5593383312225342,
|
| 3811 |
+
"step": 1086
|
| 3812 |
+
},
|
| 3813 |
+
{
|
| 3814 |
+
"epoch": 0.3793782798960894,
|
| 3815 |
+
"grad_norm": 0.2294865846633911,
|
| 3816 |
+
"learning_rate": 5e-05,
|
| 3817 |
+
"loss": 1.9427132606506348,
|
| 3818 |
+
"step": 1088
|
| 3819 |
+
},
|
| 3820 |
+
{
|
| 3821 |
+
"epoch": 0.38007566644001606,
|
| 3822 |
+
"grad_norm": 0.23989562690258026,
|
| 3823 |
+
"learning_rate": 5e-05,
|
| 3824 |
+
"loss": 1.9454470872879028,
|
| 3825 |
+
"step": 1090
|
| 3826 |
+
},
|
| 3827 |
+
{
|
| 3828 |
+
"epoch": 0.38077305298394265,
|
| 3829 |
+
"grad_norm": 0.31190624833106995,
|
| 3830 |
+
"learning_rate": 5e-05,
|
| 3831 |
+
"loss": 2.0495333671569824,
|
| 3832 |
+
"step": 1092
|
| 3833 |
+
},
|
| 3834 |
+
{
|
| 3835 |
+
"epoch": 0.3814704395278693,
|
| 3836 |
+
"grad_norm": 0.1790897399187088,
|
| 3837 |
+
"learning_rate": 5e-05,
|
| 3838 |
+
"loss": 2.115795850753784,
|
| 3839 |
+
"step": 1094
|
| 3840 |
+
},
|
| 3841 |
+
{
|
| 3842 |
+
"epoch": 0.38216782607179595,
|
| 3843 |
+
"grad_norm": 0.23868292570114136,
|
| 3844 |
+
"learning_rate": 5e-05,
|
| 3845 |
+
"loss": 2.1075124740600586,
|
| 3846 |
+
"step": 1096
|
| 3847 |
+
},
|
| 3848 |
+
{
|
| 3849 |
+
"epoch": 0.3828652126157226,
|
| 3850 |
+
"grad_norm": 0.2061663120985031,
|
| 3851 |
+
"learning_rate": 5e-05,
|
| 3852 |
+
"loss": 1.8393436670303345,
|
| 3853 |
+
"step": 1098
|
| 3854 |
+
},
|
| 3855 |
+
{
|
| 3856 |
+
"epoch": 0.3835625991596492,
|
| 3857 |
+
"grad_norm": 0.18991388380527496,
|
| 3858 |
+
"learning_rate": 5e-05,
|
| 3859 |
+
"loss": 2.1416378021240234,
|
| 3860 |
+
"step": 1100
|
| 3861 |
+
},
|
| 3862 |
+
{
|
| 3863 |
+
"epoch": 0.38425998570357583,
|
| 3864 |
+
"grad_norm": 0.21696053445339203,
|
| 3865 |
+
"learning_rate": 5e-05,
|
| 3866 |
+
"loss": 2.0470800399780273,
|
| 3867 |
+
"step": 1102
|
| 3868 |
+
},
|
| 3869 |
+
{
|
| 3870 |
+
"epoch": 0.3849573722475025,
|
| 3871 |
+
"grad_norm": 0.3593919277191162,
|
| 3872 |
+
"learning_rate": 5e-05,
|
| 3873 |
+
"loss": 1.9083117246627808,
|
| 3874 |
+
"step": 1104
|
| 3875 |
+
},
|
| 3876 |
+
{
|
| 3877 |
+
"epoch": 0.38565475879142913,
|
| 3878 |
+
"grad_norm": 0.17209303379058838,
|
| 3879 |
+
"learning_rate": 5e-05,
|
| 3880 |
+
"loss": 2.0877814292907715,
|
| 3881 |
+
"step": 1106
|
| 3882 |
+
},
|
| 3883 |
+
{
|
| 3884 |
+
"epoch": 0.3863521453353558,
|
| 3885 |
+
"grad_norm": 0.22013252973556519,
|
| 3886 |
+
"learning_rate": 5e-05,
|
| 3887 |
+
"loss": 2.1486687660217285,
|
| 3888 |
+
"step": 1108
|
| 3889 |
+
},
|
| 3890 |
+
{
|
| 3891 |
+
"epoch": 0.38704953187928237,
|
| 3892 |
+
"grad_norm": 0.1675334870815277,
|
| 3893 |
+
"learning_rate": 5e-05,
|
| 3894 |
+
"loss": 2.065140724182129,
|
| 3895 |
+
"step": 1110
|
| 3896 |
+
},
|
| 3897 |
+
{
|
| 3898 |
+
"epoch": 0.387746918423209,
|
| 3899 |
+
"grad_norm": 0.2227640151977539,
|
| 3900 |
+
"learning_rate": 5e-05,
|
| 3901 |
+
"loss": 1.9250330924987793,
|
| 3902 |
+
"step": 1112
|
| 3903 |
+
},
|
| 3904 |
+
{
|
| 3905 |
+
"epoch": 0.38844430496713567,
|
| 3906 |
+
"grad_norm": 0.21977634727954865,
|
| 3907 |
+
"learning_rate": 5e-05,
|
| 3908 |
+
"loss": 1.8304221630096436,
|
| 3909 |
+
"step": 1114
|
| 3910 |
+
},
|
| 3911 |
+
{
|
| 3912 |
+
"epoch": 0.3891416915110623,
|
| 3913 |
+
"grad_norm": 0.21973222494125366,
|
| 3914 |
+
"learning_rate": 5e-05,
|
| 3915 |
+
"loss": 1.9635357856750488,
|
| 3916 |
+
"step": 1116
|
| 3917 |
+
},
|
| 3918 |
+
{
|
| 3919 |
+
"epoch": 0.3898390780549889,
|
| 3920 |
+
"grad_norm": 0.2314382642507553,
|
| 3921 |
+
"learning_rate": 5e-05,
|
| 3922 |
+
"loss": 1.9372278451919556,
|
| 3923 |
+
"step": 1118
|
| 3924 |
+
},
|
| 3925 |
+
{
|
| 3926 |
+
"epoch": 0.39053646459891556,
|
| 3927 |
+
"grad_norm": 0.21442043781280518,
|
| 3928 |
+
"learning_rate": 5e-05,
|
| 3929 |
+
"loss": 1.7821249961853027,
|
| 3930 |
+
"step": 1120
|
| 3931 |
+
},
|
| 3932 |
+
{
|
| 3933 |
+
"epoch": 0.3912338511428422,
|
| 3934 |
+
"grad_norm": 0.22008800506591797,
|
| 3935 |
+
"learning_rate": 5e-05,
|
| 3936 |
+
"loss": 2.0237293243408203,
|
| 3937 |
+
"step": 1122
|
| 3938 |
+
},
|
| 3939 |
+
{
|
| 3940 |
+
"epoch": 0.39193123768676885,
|
| 3941 |
+
"grad_norm": 0.23807160556316376,
|
| 3942 |
+
"learning_rate": 5e-05,
|
| 3943 |
+
"loss": 2.054992198944092,
|
| 3944 |
+
"step": 1124
|
| 3945 |
+
},
|
| 3946 |
+
{
|
| 3947 |
+
"epoch": 0.39262862423069544,
|
| 3948 |
+
"grad_norm": 0.437844842672348,
|
| 3949 |
+
"learning_rate": 5e-05,
|
| 3950 |
+
"loss": 1.7409727573394775,
|
| 3951 |
+
"step": 1126
|
| 3952 |
+
},
|
| 3953 |
+
{
|
| 3954 |
+
"epoch": 0.3933260107746221,
|
| 3955 |
+
"grad_norm": 0.21472229063510895,
|
| 3956 |
+
"learning_rate": 5e-05,
|
| 3957 |
+
"loss": 2.0891048908233643,
|
| 3958 |
+
"step": 1128
|
| 3959 |
+
},
|
| 3960 |
+
{
|
| 3961 |
+
"epoch": 0.39402339731854874,
|
| 3962 |
+
"grad_norm": 0.38082173466682434,
|
| 3963 |
+
"learning_rate": 5e-05,
|
| 3964 |
+
"loss": 1.919767141342163,
|
| 3965 |
+
"step": 1130
|
| 3966 |
+
},
|
| 3967 |
+
{
|
| 3968 |
+
"epoch": 0.3947207838624754,
|
| 3969 |
+
"grad_norm": 0.3429301083087921,
|
| 3970 |
+
"learning_rate": 5e-05,
|
| 3971 |
+
"loss": 1.9058270454406738,
|
| 3972 |
+
"step": 1132
|
| 3973 |
+
},
|
| 3974 |
+
{
|
| 3975 |
+
"epoch": 0.395418170406402,
|
| 3976 |
+
"grad_norm": 0.22207891941070557,
|
| 3977 |
+
"learning_rate": 5e-05,
|
| 3978 |
+
"loss": 2.0372676849365234,
|
| 3979 |
+
"step": 1134
|
| 3980 |
+
},
|
| 3981 |
+
{
|
| 3982 |
+
"epoch": 0.39611555695032863,
|
| 3983 |
+
"grad_norm": 0.2745357155799866,
|
| 3984 |
+
"learning_rate": 5e-05,
|
| 3985 |
+
"loss": 1.9663827419281006,
|
| 3986 |
+
"step": 1136
|
| 3987 |
+
},
|
| 3988 |
+
{
|
| 3989 |
+
"epoch": 0.3968129434942553,
|
| 3990 |
+
"grad_norm": 0.3773990869522095,
|
| 3991 |
+
"learning_rate": 5e-05,
|
| 3992 |
+
"loss": 1.9896197319030762,
|
| 3993 |
+
"step": 1138
|
| 3994 |
+
},
|
| 3995 |
+
{
|
| 3996 |
+
"epoch": 0.3975103300381819,
|
| 3997 |
+
"grad_norm": 0.2900267243385315,
|
| 3998 |
+
"learning_rate": 5e-05,
|
| 3999 |
+
"loss": 1.7686326503753662,
|
| 4000 |
+
"step": 1140
|
| 4001 |
+
},
|
| 4002 |
+
{
|
| 4003 |
+
"epoch": 0.3982077165821086,
|
| 4004 |
+
"grad_norm": 0.22188392281532288,
|
| 4005 |
+
"learning_rate": 5e-05,
|
| 4006 |
+
"loss": 2.0737409591674805,
|
| 4007 |
+
"step": 1142
|
| 4008 |
+
},
|
| 4009 |
+
{
|
| 4010 |
+
"epoch": 0.39890510312603517,
|
| 4011 |
+
"grad_norm": 0.2207290530204773,
|
| 4012 |
+
"learning_rate": 5e-05,
|
| 4013 |
+
"loss": 2.016484260559082,
|
| 4014 |
+
"step": 1144
|
| 4015 |
+
},
|
| 4016 |
+
{
|
| 4017 |
+
"epoch": 0.3996024896699618,
|
| 4018 |
+
"grad_norm": 0.1925177276134491,
|
| 4019 |
+
"learning_rate": 5e-05,
|
| 4020 |
+
"loss": 1.7681307792663574,
|
| 4021 |
+
"step": 1146
|
| 4022 |
+
},
|
| 4023 |
+
{
|
| 4024 |
+
"epoch": 0.40029987621388846,
|
| 4025 |
+
"grad_norm": 0.19402268528938293,
|
| 4026 |
+
"learning_rate": 5e-05,
|
| 4027 |
+
"loss": 2.1917271614074707,
|
| 4028 |
+
"step": 1148
|
| 4029 |
}
|
| 4030 |
],
|
| 4031 |
"logging_steps": 2,
|
|
|
|
| 4045 |
"attributes": {}
|
| 4046 |
}
|
| 4047 |
},
|
| 4048 |
+
"total_flos": 2.0393064516707942e+17,
|
| 4049 |
"train_batch_size": 1,
|
| 4050 |
"trial_name": null,
|
| 4051 |
"trial_params": null
|