Training in progress, step 3360, 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 2519936480
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e7231d9e756c08a904e99667434c9d8b7b5aa980d0a1c25051e5387770555d31
|
| 3 |
size 2519936480
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1265545196
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6a51e23e42a5d649730c02c59d357c365579703c91878d3fb113a04b58886f2f
|
| 3 |
size 1265545196
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 10233
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:741c79f6f0644e5e1a8c98e05c7861c22187fc96145ff60c21c8fa583f48422c
|
| 3 |
size 10233
|
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": 959,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -10112,6 +10112,1686 @@
|
|
| 10112 |
"learning_rate": 0.00047686079103083156,
|
| 10113 |
"loss": 1.5678623914718628,
|
| 10114 |
"step": 2880
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10115 |
}
|
| 10116 |
],
|
| 10117 |
"logging_steps": 2,
|
|
@@ -10131,7 +11811,7 @@
|
|
| 10131 |
"attributes": {}
|
| 10132 |
}
|
| 10133 |
},
|
| 10134 |
-
"total_flos":
|
| 10135 |
"train_batch_size": 8,
|
| 10136 |
"trial_name": null,
|
| 10137 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.701095461658842,
|
| 6 |
"eval_steps": 959,
|
| 7 |
+
"global_step": 3360,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 10112 |
"learning_rate": 0.00047686079103083156,
|
| 10113 |
"loss": 1.5678623914718628,
|
| 10114 |
"step": 2880
|
| 10115 |
+
},
|
| 10116 |
+
{
|
| 10117 |
+
"epoch": 0.6013562858633281,
|
| 10118 |
+
"grad_norm": 0.19277074933052063,
|
| 10119 |
+
"learning_rate": 0.00047636250389286827,
|
| 10120 |
+
"loss": 1.7933660745620728,
|
| 10121 |
+
"step": 2882
|
| 10122 |
+
},
|
| 10123 |
+
{
|
| 10124 |
+
"epoch": 0.601773604590506,
|
| 10125 |
+
"grad_norm": 0.17304041981697083,
|
| 10126 |
+
"learning_rate": 0.00047586421675490504,
|
| 10127 |
+
"loss": 1.6832016706466675,
|
| 10128 |
+
"step": 2884
|
| 10129 |
+
},
|
| 10130 |
+
{
|
| 10131 |
+
"epoch": 0.6021909233176839,
|
| 10132 |
+
"grad_norm": 0.17196182906627655,
|
| 10133 |
+
"learning_rate": 0.0004753659296169418,
|
| 10134 |
+
"loss": 1.5628772974014282,
|
| 10135 |
+
"step": 2886
|
| 10136 |
+
},
|
| 10137 |
+
{
|
| 10138 |
+
"epoch": 0.6026082420448617,
|
| 10139 |
+
"grad_norm": 0.18928901851177216,
|
| 10140 |
+
"learning_rate": 0.00047486764247897857,
|
| 10141 |
+
"loss": 1.625766396522522,
|
| 10142 |
+
"step": 2888
|
| 10143 |
+
},
|
| 10144 |
+
{
|
| 10145 |
+
"epoch": 0.6030255607720396,
|
| 10146 |
+
"grad_norm": 0.2047906070947647,
|
| 10147 |
+
"learning_rate": 0.0004743693553410153,
|
| 10148 |
+
"loss": 1.4966003894805908,
|
| 10149 |
+
"step": 2890
|
| 10150 |
+
},
|
| 10151 |
+
{
|
| 10152 |
+
"epoch": 0.6034428794992175,
|
| 10153 |
+
"grad_norm": 0.17776289582252502,
|
| 10154 |
+
"learning_rate": 0.00047387106820305204,
|
| 10155 |
+
"loss": 1.7307140827178955,
|
| 10156 |
+
"step": 2892
|
| 10157 |
+
},
|
| 10158 |
+
{
|
| 10159 |
+
"epoch": 0.6038601982263954,
|
| 10160 |
+
"grad_norm": 0.1764279305934906,
|
| 10161 |
+
"learning_rate": 0.0004733727810650888,
|
| 10162 |
+
"loss": 1.7942173480987549,
|
| 10163 |
+
"step": 2894
|
| 10164 |
+
},
|
| 10165 |
+
{
|
| 10166 |
+
"epoch": 0.6042775169535733,
|
| 10167 |
+
"grad_norm": 0.1908806711435318,
|
| 10168 |
+
"learning_rate": 0.00047287449392712557,
|
| 10169 |
+
"loss": 1.4785549640655518,
|
| 10170 |
+
"step": 2896
|
| 10171 |
+
},
|
| 10172 |
+
{
|
| 10173 |
+
"epoch": 0.6046948356807512,
|
| 10174 |
+
"grad_norm": 0.1910744160413742,
|
| 10175 |
+
"learning_rate": 0.0004723762067891623,
|
| 10176 |
+
"loss": 1.695818305015564,
|
| 10177 |
+
"step": 2898
|
| 10178 |
+
},
|
| 10179 |
+
{
|
| 10180 |
+
"epoch": 0.6051121544079291,
|
| 10181 |
+
"grad_norm": 0.1757185459136963,
|
| 10182 |
+
"learning_rate": 0.00047187791965119905,
|
| 10183 |
+
"loss": 1.6474026441574097,
|
| 10184 |
+
"step": 2900
|
| 10185 |
+
},
|
| 10186 |
+
{
|
| 10187 |
+
"epoch": 0.605529473135107,
|
| 10188 |
+
"grad_norm": 0.18998144567012787,
|
| 10189 |
+
"learning_rate": 0.0004713796325132358,
|
| 10190 |
+
"loss": 1.6418334245681763,
|
| 10191 |
+
"step": 2902
|
| 10192 |
+
},
|
| 10193 |
+
{
|
| 10194 |
+
"epoch": 0.6059467918622848,
|
| 10195 |
+
"grad_norm": 0.19940714538097382,
|
| 10196 |
+
"learning_rate": 0.0004708813453752726,
|
| 10197 |
+
"loss": 1.650315284729004,
|
| 10198 |
+
"step": 2904
|
| 10199 |
+
},
|
| 10200 |
+
{
|
| 10201 |
+
"epoch": 0.6063641105894627,
|
| 10202 |
+
"grad_norm": 0.19629891216754913,
|
| 10203 |
+
"learning_rate": 0.0004703830582373093,
|
| 10204 |
+
"loss": 1.6142354011535645,
|
| 10205 |
+
"step": 2906
|
| 10206 |
+
},
|
| 10207 |
+
{
|
| 10208 |
+
"epoch": 0.6067814293166406,
|
| 10209 |
+
"grad_norm": 0.19428136944770813,
|
| 10210 |
+
"learning_rate": 0.00046988477109934605,
|
| 10211 |
+
"loss": 1.5597343444824219,
|
| 10212 |
+
"step": 2908
|
| 10213 |
+
},
|
| 10214 |
+
{
|
| 10215 |
+
"epoch": 0.6071987480438185,
|
| 10216 |
+
"grad_norm": 0.1555328220129013,
|
| 10217 |
+
"learning_rate": 0.0004693864839613828,
|
| 10218 |
+
"loss": 1.6591336727142334,
|
| 10219 |
+
"step": 2910
|
| 10220 |
+
},
|
| 10221 |
+
{
|
| 10222 |
+
"epoch": 0.6076160667709963,
|
| 10223 |
+
"grad_norm": 0.17966602742671967,
|
| 10224 |
+
"learning_rate": 0.0004688881968234196,
|
| 10225 |
+
"loss": 1.544377088546753,
|
| 10226 |
+
"step": 2912
|
| 10227 |
+
},
|
| 10228 |
+
{
|
| 10229 |
+
"epoch": 0.6080333854981742,
|
| 10230 |
+
"grad_norm": 0.18002371490001678,
|
| 10231 |
+
"learning_rate": 0.0004683899096854563,
|
| 10232 |
+
"loss": 1.77950119972229,
|
| 10233 |
+
"step": 2914
|
| 10234 |
+
},
|
| 10235 |
+
{
|
| 10236 |
+
"epoch": 0.6084507042253521,
|
| 10237 |
+
"grad_norm": 0.1758643537759781,
|
| 10238 |
+
"learning_rate": 0.00046789162254749306,
|
| 10239 |
+
"loss": 1.751016616821289,
|
| 10240 |
+
"step": 2916
|
| 10241 |
+
},
|
| 10242 |
+
{
|
| 10243 |
+
"epoch": 0.60886802295253,
|
| 10244 |
+
"grad_norm": 0.1944553703069687,
|
| 10245 |
+
"learning_rate": 0.0004673933354095297,
|
| 10246 |
+
"loss": 1.6245421171188354,
|
| 10247 |
+
"step": 2918
|
| 10248 |
+
},
|
| 10249 |
+
{
|
| 10250 |
+
"epoch": 0.6092853416797078,
|
| 10251 |
+
"grad_norm": 0.16478294134140015,
|
| 10252 |
+
"learning_rate": 0.0004668950482715665,
|
| 10253 |
+
"loss": 1.6343562602996826,
|
| 10254 |
+
"step": 2920
|
| 10255 |
+
},
|
| 10256 |
+
{
|
| 10257 |
+
"epoch": 0.6097026604068858,
|
| 10258 |
+
"grad_norm": 0.18271125853061676,
|
| 10259 |
+
"learning_rate": 0.00046639676113360325,
|
| 10260 |
+
"loss": 1.5838673114776611,
|
| 10261 |
+
"step": 2922
|
| 10262 |
+
},
|
| 10263 |
+
{
|
| 10264 |
+
"epoch": 0.6101199791340637,
|
| 10265 |
+
"grad_norm": 0.1791974902153015,
|
| 10266 |
+
"learning_rate": 0.00046589847399563996,
|
| 10267 |
+
"loss": 1.5538192987442017,
|
| 10268 |
+
"step": 2924
|
| 10269 |
+
},
|
| 10270 |
+
{
|
| 10271 |
+
"epoch": 0.6105372978612416,
|
| 10272 |
+
"grad_norm": 0.1821575164794922,
|
| 10273 |
+
"learning_rate": 0.0004654001868576767,
|
| 10274 |
+
"loss": 1.7037311792373657,
|
| 10275 |
+
"step": 2926
|
| 10276 |
+
},
|
| 10277 |
+
{
|
| 10278 |
+
"epoch": 0.6109546165884194,
|
| 10279 |
+
"grad_norm": 0.17619025707244873,
|
| 10280 |
+
"learning_rate": 0.0004649018997197135,
|
| 10281 |
+
"loss": 1.686895728111267,
|
| 10282 |
+
"step": 2928
|
| 10283 |
+
},
|
| 10284 |
+
{
|
| 10285 |
+
"epoch": 0.6113719353155973,
|
| 10286 |
+
"grad_norm": 0.15952201187610626,
|
| 10287 |
+
"learning_rate": 0.00046440361258175025,
|
| 10288 |
+
"loss": 1.6833465099334717,
|
| 10289 |
+
"step": 2930
|
| 10290 |
+
},
|
| 10291 |
+
{
|
| 10292 |
+
"epoch": 0.6117892540427752,
|
| 10293 |
+
"grad_norm": 0.23567429184913635,
|
| 10294 |
+
"learning_rate": 0.00046390532544378696,
|
| 10295 |
+
"loss": 1.4512883424758911,
|
| 10296 |
+
"step": 2932
|
| 10297 |
+
},
|
| 10298 |
+
{
|
| 10299 |
+
"epoch": 0.612206572769953,
|
| 10300 |
+
"grad_norm": 0.19210046529769897,
|
| 10301 |
+
"learning_rate": 0.0004634070383058237,
|
| 10302 |
+
"loss": 1.7234846353530884,
|
| 10303 |
+
"step": 2934
|
| 10304 |
+
},
|
| 10305 |
+
{
|
| 10306 |
+
"epoch": 0.6126238914971309,
|
| 10307 |
+
"grad_norm": 0.1734471470117569,
|
| 10308 |
+
"learning_rate": 0.0004629087511678605,
|
| 10309 |
+
"loss": 1.7545769214630127,
|
| 10310 |
+
"step": 2936
|
| 10311 |
+
},
|
| 10312 |
+
{
|
| 10313 |
+
"epoch": 0.6130412102243088,
|
| 10314 |
+
"grad_norm": 0.18496385216712952,
|
| 10315 |
+
"learning_rate": 0.0004624104640298972,
|
| 10316 |
+
"loss": 1.572465181350708,
|
| 10317 |
+
"step": 2938
|
| 10318 |
+
},
|
| 10319 |
+
{
|
| 10320 |
+
"epoch": 0.6134585289514867,
|
| 10321 |
+
"grad_norm": 0.2221444547176361,
|
| 10322 |
+
"learning_rate": 0.00046191217689193397,
|
| 10323 |
+
"loss": 1.519800066947937,
|
| 10324 |
+
"step": 2940
|
| 10325 |
+
},
|
| 10326 |
+
{
|
| 10327 |
+
"epoch": 0.6138758476786645,
|
| 10328 |
+
"grad_norm": 0.24368079006671906,
|
| 10329 |
+
"learning_rate": 0.00046141388975397073,
|
| 10330 |
+
"loss": 1.6060869693756104,
|
| 10331 |
+
"step": 2942
|
| 10332 |
+
},
|
| 10333 |
+
{
|
| 10334 |
+
"epoch": 0.6142931664058424,
|
| 10335 |
+
"grad_norm": 0.16458842158317566,
|
| 10336 |
+
"learning_rate": 0.0004609156026160075,
|
| 10337 |
+
"loss": 1.6020828485488892,
|
| 10338 |
+
"step": 2944
|
| 10339 |
+
},
|
| 10340 |
+
{
|
| 10341 |
+
"epoch": 0.6147104851330203,
|
| 10342 |
+
"grad_norm": 0.1970018744468689,
|
| 10343 |
+
"learning_rate": 0.0004604173154780442,
|
| 10344 |
+
"loss": 1.6586238145828247,
|
| 10345 |
+
"step": 2946
|
| 10346 |
+
},
|
| 10347 |
+
{
|
| 10348 |
+
"epoch": 0.6151278038601983,
|
| 10349 |
+
"grad_norm": 0.1822272688150406,
|
| 10350 |
+
"learning_rate": 0.00045991902834008097,
|
| 10351 |
+
"loss": 1.720170497894287,
|
| 10352 |
+
"step": 2948
|
| 10353 |
+
},
|
| 10354 |
+
{
|
| 10355 |
+
"epoch": 0.6155451225873761,
|
| 10356 |
+
"grad_norm": 0.17016743123531342,
|
| 10357 |
+
"learning_rate": 0.00045942074120211774,
|
| 10358 |
+
"loss": 1.5761950016021729,
|
| 10359 |
+
"step": 2950
|
| 10360 |
+
},
|
| 10361 |
+
{
|
| 10362 |
+
"epoch": 0.615962441314554,
|
| 10363 |
+
"grad_norm": 0.2120160609483719,
|
| 10364 |
+
"learning_rate": 0.0004589224540641545,
|
| 10365 |
+
"loss": 1.6407349109649658,
|
| 10366 |
+
"step": 2952
|
| 10367 |
+
},
|
| 10368 |
+
{
|
| 10369 |
+
"epoch": 0.6163797600417319,
|
| 10370 |
+
"grad_norm": 0.16476835310459137,
|
| 10371 |
+
"learning_rate": 0.0004584241669261912,
|
| 10372 |
+
"loss": 1.7095156908035278,
|
| 10373 |
+
"step": 2954
|
| 10374 |
+
},
|
| 10375 |
+
{
|
| 10376 |
+
"epoch": 0.6167970787689098,
|
| 10377 |
+
"grad_norm": 0.1890491098165512,
|
| 10378 |
+
"learning_rate": 0.000457925879788228,
|
| 10379 |
+
"loss": 1.591587781906128,
|
| 10380 |
+
"step": 2956
|
| 10381 |
+
},
|
| 10382 |
+
{
|
| 10383 |
+
"epoch": 0.6172143974960876,
|
| 10384 |
+
"grad_norm": 0.18666179478168488,
|
| 10385 |
+
"learning_rate": 0.00045742759265026474,
|
| 10386 |
+
"loss": 1.5489561557769775,
|
| 10387 |
+
"step": 2958
|
| 10388 |
+
},
|
| 10389 |
+
{
|
| 10390 |
+
"epoch": 0.6176317162232655,
|
| 10391 |
+
"grad_norm": 0.2605527341365814,
|
| 10392 |
+
"learning_rate": 0.0004569293055123015,
|
| 10393 |
+
"loss": 1.5896910429000854,
|
| 10394 |
+
"step": 2960
|
| 10395 |
+
},
|
| 10396 |
+
{
|
| 10397 |
+
"epoch": 0.6180490349504434,
|
| 10398 |
+
"grad_norm": 0.16350412368774414,
|
| 10399 |
+
"learning_rate": 0.0004564310183743382,
|
| 10400 |
+
"loss": 1.6582103967666626,
|
| 10401 |
+
"step": 2962
|
| 10402 |
+
},
|
| 10403 |
+
{
|
| 10404 |
+
"epoch": 0.6184663536776213,
|
| 10405 |
+
"grad_norm": 0.18763825297355652,
|
| 10406 |
+
"learning_rate": 0.000455932731236375,
|
| 10407 |
+
"loss": 1.7946631908416748,
|
| 10408 |
+
"step": 2964
|
| 10409 |
+
},
|
| 10410 |
+
{
|
| 10411 |
+
"epoch": 0.6188836724047991,
|
| 10412 |
+
"grad_norm": 0.22416430711746216,
|
| 10413 |
+
"learning_rate": 0.00045543444409841175,
|
| 10414 |
+
"loss": 1.6434133052825928,
|
| 10415 |
+
"step": 2966
|
| 10416 |
+
},
|
| 10417 |
+
{
|
| 10418 |
+
"epoch": 0.619300991131977,
|
| 10419 |
+
"grad_norm": 0.17112663388252258,
|
| 10420 |
+
"learning_rate": 0.0004549361569604485,
|
| 10421 |
+
"loss": 1.5859663486480713,
|
| 10422 |
+
"step": 2968
|
| 10423 |
+
},
|
| 10424 |
+
{
|
| 10425 |
+
"epoch": 0.6197183098591549,
|
| 10426 |
+
"grad_norm": 0.22274653613567352,
|
| 10427 |
+
"learning_rate": 0.0004544378698224852,
|
| 10428 |
+
"loss": 1.6013418436050415,
|
| 10429 |
+
"step": 2970
|
| 10430 |
+
},
|
| 10431 |
+
{
|
| 10432 |
+
"epoch": 0.6201356285863329,
|
| 10433 |
+
"grad_norm": 0.19702008366584778,
|
| 10434 |
+
"learning_rate": 0.000453939582684522,
|
| 10435 |
+
"loss": 1.6691360473632812,
|
| 10436 |
+
"step": 2972
|
| 10437 |
+
},
|
| 10438 |
+
{
|
| 10439 |
+
"epoch": 0.6205529473135107,
|
| 10440 |
+
"grad_norm": 0.1698775291442871,
|
| 10441 |
+
"learning_rate": 0.00045344129554655875,
|
| 10442 |
+
"loss": 1.8140881061553955,
|
| 10443 |
+
"step": 2974
|
| 10444 |
+
},
|
| 10445 |
+
{
|
| 10446 |
+
"epoch": 0.6209702660406886,
|
| 10447 |
+
"grad_norm": 0.1693205088376999,
|
| 10448 |
+
"learning_rate": 0.0004529430084085955,
|
| 10449 |
+
"loss": 1.6525713205337524,
|
| 10450 |
+
"step": 2976
|
| 10451 |
+
},
|
| 10452 |
+
{
|
| 10453 |
+
"epoch": 0.6213875847678665,
|
| 10454 |
+
"grad_norm": 0.1770809441804886,
|
| 10455 |
+
"learning_rate": 0.00045244472127063223,
|
| 10456 |
+
"loss": 1.5939429998397827,
|
| 10457 |
+
"step": 2978
|
| 10458 |
+
},
|
| 10459 |
+
{
|
| 10460 |
+
"epoch": 0.6218049034950444,
|
| 10461 |
+
"grad_norm": 0.1748695820569992,
|
| 10462 |
+
"learning_rate": 0.000451946434132669,
|
| 10463 |
+
"loss": 1.5831851959228516,
|
| 10464 |
+
"step": 2980
|
| 10465 |
+
},
|
| 10466 |
+
{
|
| 10467 |
+
"epoch": 0.6222222222222222,
|
| 10468 |
+
"grad_norm": 0.19646704196929932,
|
| 10469 |
+
"learning_rate": 0.00045144814699470576,
|
| 10470 |
+
"loss": 1.6287367343902588,
|
| 10471 |
+
"step": 2982
|
| 10472 |
+
},
|
| 10473 |
+
{
|
| 10474 |
+
"epoch": 0.6226395409494001,
|
| 10475 |
+
"grad_norm": 0.1683228760957718,
|
| 10476 |
+
"learning_rate": 0.00045094985985674247,
|
| 10477 |
+
"loss": 1.6986901760101318,
|
| 10478 |
+
"step": 2984
|
| 10479 |
+
},
|
| 10480 |
+
{
|
| 10481 |
+
"epoch": 0.623056859676578,
|
| 10482 |
+
"grad_norm": 0.1828620731830597,
|
| 10483 |
+
"learning_rate": 0.00045045157271877924,
|
| 10484 |
+
"loss": 1.7912113666534424,
|
| 10485 |
+
"step": 2986
|
| 10486 |
+
},
|
| 10487 |
+
{
|
| 10488 |
+
"epoch": 0.6234741784037559,
|
| 10489 |
+
"grad_norm": 0.16508342325687408,
|
| 10490 |
+
"learning_rate": 0.000449953285580816,
|
| 10491 |
+
"loss": 1.5801665782928467,
|
| 10492 |
+
"step": 2988
|
| 10493 |
+
},
|
| 10494 |
+
{
|
| 10495 |
+
"epoch": 0.6238914971309337,
|
| 10496 |
+
"grad_norm": 0.15644466876983643,
|
| 10497 |
+
"learning_rate": 0.00044945499844285277,
|
| 10498 |
+
"loss": 1.6213196516036987,
|
| 10499 |
+
"step": 2990
|
| 10500 |
+
},
|
| 10501 |
+
{
|
| 10502 |
+
"epoch": 0.6243088158581116,
|
| 10503 |
+
"grad_norm": 0.16748785972595215,
|
| 10504 |
+
"learning_rate": 0.0004489567113048895,
|
| 10505 |
+
"loss": 1.6507726907730103,
|
| 10506 |
+
"step": 2992
|
| 10507 |
+
},
|
| 10508 |
+
{
|
| 10509 |
+
"epoch": 0.6247261345852895,
|
| 10510 |
+
"grad_norm": 0.18665555119514465,
|
| 10511 |
+
"learning_rate": 0.00044845842416692624,
|
| 10512 |
+
"loss": 1.7788985967636108,
|
| 10513 |
+
"step": 2994
|
| 10514 |
+
},
|
| 10515 |
+
{
|
| 10516 |
+
"epoch": 0.6251434533124673,
|
| 10517 |
+
"grad_norm": 0.21414385735988617,
|
| 10518 |
+
"learning_rate": 0.000447960137028963,
|
| 10519 |
+
"loss": 1.587815284729004,
|
| 10520 |
+
"step": 2996
|
| 10521 |
+
},
|
| 10522 |
+
{
|
| 10523 |
+
"epoch": 0.6255607720396453,
|
| 10524 |
+
"grad_norm": 0.2360076606273651,
|
| 10525 |
+
"learning_rate": 0.00044746184989099977,
|
| 10526 |
+
"loss": 1.6898614168167114,
|
| 10527 |
+
"step": 2998
|
| 10528 |
+
},
|
| 10529 |
+
{
|
| 10530 |
+
"epoch": 0.6259780907668232,
|
| 10531 |
+
"grad_norm": 0.19727914035320282,
|
| 10532 |
+
"learning_rate": 0.0004469635627530365,
|
| 10533 |
+
"loss": 1.5027776956558228,
|
| 10534 |
+
"step": 3000
|
| 10535 |
+
},
|
| 10536 |
+
{
|
| 10537 |
+
"epoch": 0.6263954094940011,
|
| 10538 |
+
"grad_norm": 0.16130492091178894,
|
| 10539 |
+
"learning_rate": 0.00044646527561507314,
|
| 10540 |
+
"loss": 1.6677963733673096,
|
| 10541 |
+
"step": 3002
|
| 10542 |
+
},
|
| 10543 |
+
{
|
| 10544 |
+
"epoch": 0.626812728221179,
|
| 10545 |
+
"grad_norm": 0.17396186292171478,
|
| 10546 |
+
"learning_rate": 0.0004459669884771099,
|
| 10547 |
+
"loss": 1.7318880558013916,
|
| 10548 |
+
"step": 3004
|
| 10549 |
+
},
|
| 10550 |
+
{
|
| 10551 |
+
"epoch": 0.6272300469483568,
|
| 10552 |
+
"grad_norm": 0.16842906177043915,
|
| 10553 |
+
"learning_rate": 0.00044546870133914667,
|
| 10554 |
+
"loss": 1.7320445775985718,
|
| 10555 |
+
"step": 3006
|
| 10556 |
+
},
|
| 10557 |
+
{
|
| 10558 |
+
"epoch": 0.6276473656755347,
|
| 10559 |
+
"grad_norm": 0.17343546450138092,
|
| 10560 |
+
"learning_rate": 0.00044497041420118343,
|
| 10561 |
+
"loss": 1.5647716522216797,
|
| 10562 |
+
"step": 3008
|
| 10563 |
+
},
|
| 10564 |
+
{
|
| 10565 |
+
"epoch": 0.6280646844027126,
|
| 10566 |
+
"grad_norm": 0.18864287436008453,
|
| 10567 |
+
"learning_rate": 0.00044447212706322014,
|
| 10568 |
+
"loss": 1.6539498567581177,
|
| 10569 |
+
"step": 3010
|
| 10570 |
+
},
|
| 10571 |
+
{
|
| 10572 |
+
"epoch": 0.6284820031298904,
|
| 10573 |
+
"grad_norm": 0.18099819123744965,
|
| 10574 |
+
"learning_rate": 0.0004439738399252569,
|
| 10575 |
+
"loss": 1.6597203016281128,
|
| 10576 |
+
"step": 3012
|
| 10577 |
+
},
|
| 10578 |
+
{
|
| 10579 |
+
"epoch": 0.6288993218570683,
|
| 10580 |
+
"grad_norm": 0.1965656727552414,
|
| 10581 |
+
"learning_rate": 0.0004434755527872937,
|
| 10582 |
+
"loss": 1.5832552909851074,
|
| 10583 |
+
"step": 3014
|
| 10584 |
+
},
|
| 10585 |
+
{
|
| 10586 |
+
"epoch": 0.6293166405842462,
|
| 10587 |
+
"grad_norm": 0.1879548877477646,
|
| 10588 |
+
"learning_rate": 0.00044297726564933044,
|
| 10589 |
+
"loss": 1.5848406553268433,
|
| 10590 |
+
"step": 3016
|
| 10591 |
+
},
|
| 10592 |
+
{
|
| 10593 |
+
"epoch": 0.6297339593114241,
|
| 10594 |
+
"grad_norm": 0.16935324668884277,
|
| 10595 |
+
"learning_rate": 0.00044247897851136715,
|
| 10596 |
+
"loss": 1.6081706285476685,
|
| 10597 |
+
"step": 3018
|
| 10598 |
+
},
|
| 10599 |
+
{
|
| 10600 |
+
"epoch": 0.6301512780386019,
|
| 10601 |
+
"grad_norm": 0.17128311097621918,
|
| 10602 |
+
"learning_rate": 0.0004419806913734039,
|
| 10603 |
+
"loss": 1.7153724431991577,
|
| 10604 |
+
"step": 3020
|
| 10605 |
+
},
|
| 10606 |
+
{
|
| 10607 |
+
"epoch": 0.6305685967657798,
|
| 10608 |
+
"grad_norm": 0.18790321052074432,
|
| 10609 |
+
"learning_rate": 0.0004414824042354407,
|
| 10610 |
+
"loss": 1.7321152687072754,
|
| 10611 |
+
"step": 3022
|
| 10612 |
+
},
|
| 10613 |
+
{
|
| 10614 |
+
"epoch": 0.6309859154929578,
|
| 10615 |
+
"grad_norm": 0.17196448147296906,
|
| 10616 |
+
"learning_rate": 0.00044098411709747744,
|
| 10617 |
+
"loss": 1.6450070142745972,
|
| 10618 |
+
"step": 3024
|
| 10619 |
+
},
|
| 10620 |
+
{
|
| 10621 |
+
"epoch": 0.6314032342201357,
|
| 10622 |
+
"grad_norm": 0.181528240442276,
|
| 10623 |
+
"learning_rate": 0.00044048582995951415,
|
| 10624 |
+
"loss": 1.6748415231704712,
|
| 10625 |
+
"step": 3026
|
| 10626 |
+
},
|
| 10627 |
+
{
|
| 10628 |
+
"epoch": 0.6318205529473135,
|
| 10629 |
+
"grad_norm": 0.18286748230457306,
|
| 10630 |
+
"learning_rate": 0.0004399875428215509,
|
| 10631 |
+
"loss": 1.7617762088775635,
|
| 10632 |
+
"step": 3028
|
| 10633 |
+
},
|
| 10634 |
+
{
|
| 10635 |
+
"epoch": 0.6322378716744914,
|
| 10636 |
+
"grad_norm": 0.16749003529548645,
|
| 10637 |
+
"learning_rate": 0.0004394892556835877,
|
| 10638 |
+
"loss": 1.6863735914230347,
|
| 10639 |
+
"step": 3030
|
| 10640 |
+
},
|
| 10641 |
+
{
|
| 10642 |
+
"epoch": 0.6326551904016693,
|
| 10643 |
+
"grad_norm": 0.22944116592407227,
|
| 10644 |
+
"learning_rate": 0.00043899096854562445,
|
| 10645 |
+
"loss": 1.5296680927276611,
|
| 10646 |
+
"step": 3032
|
| 10647 |
+
},
|
| 10648 |
+
{
|
| 10649 |
+
"epoch": 0.6330725091288472,
|
| 10650 |
+
"grad_norm": 0.1561669260263443,
|
| 10651 |
+
"learning_rate": 0.00043849268140766116,
|
| 10652 |
+
"loss": 1.5262645483016968,
|
| 10653 |
+
"step": 3034
|
| 10654 |
+
},
|
| 10655 |
+
{
|
| 10656 |
+
"epoch": 0.633489827856025,
|
| 10657 |
+
"grad_norm": 0.16601932048797607,
|
| 10658 |
+
"learning_rate": 0.0004379943942696979,
|
| 10659 |
+
"loss": 1.5379654169082642,
|
| 10660 |
+
"step": 3036
|
| 10661 |
+
},
|
| 10662 |
+
{
|
| 10663 |
+
"epoch": 0.6339071465832029,
|
| 10664 |
+
"grad_norm": 0.17442798614501953,
|
| 10665 |
+
"learning_rate": 0.0004374961071317347,
|
| 10666 |
+
"loss": 1.7634894847869873,
|
| 10667 |
+
"step": 3038
|
| 10668 |
+
},
|
| 10669 |
+
{
|
| 10670 |
+
"epoch": 0.6343244653103808,
|
| 10671 |
+
"grad_norm": 0.1849924623966217,
|
| 10672 |
+
"learning_rate": 0.00043699781999377146,
|
| 10673 |
+
"loss": 1.6289749145507812,
|
| 10674 |
+
"step": 3040
|
| 10675 |
+
},
|
| 10676 |
+
{
|
| 10677 |
+
"epoch": 0.6347417840375587,
|
| 10678 |
+
"grad_norm": 0.17964428663253784,
|
| 10679 |
+
"learning_rate": 0.00043649953285580817,
|
| 10680 |
+
"loss": 1.7756881713867188,
|
| 10681 |
+
"step": 3042
|
| 10682 |
+
},
|
| 10683 |
+
{
|
| 10684 |
+
"epoch": 0.6351591027647365,
|
| 10685 |
+
"grad_norm": 0.1700710654258728,
|
| 10686 |
+
"learning_rate": 0.00043600124571784493,
|
| 10687 |
+
"loss": 1.654688835144043,
|
| 10688 |
+
"step": 3044
|
| 10689 |
+
},
|
| 10690 |
+
{
|
| 10691 |
+
"epoch": 0.6355764214919144,
|
| 10692 |
+
"grad_norm": 0.18866358697414398,
|
| 10693 |
+
"learning_rate": 0.0004355029585798817,
|
| 10694 |
+
"loss": 1.6230889558792114,
|
| 10695 |
+
"step": 3046
|
| 10696 |
+
},
|
| 10697 |
+
{
|
| 10698 |
+
"epoch": 0.6359937402190924,
|
| 10699 |
+
"grad_norm": 0.23964236676692963,
|
| 10700 |
+
"learning_rate": 0.0004350046714419184,
|
| 10701 |
+
"loss": 1.4981260299682617,
|
| 10702 |
+
"step": 3048
|
| 10703 |
+
},
|
| 10704 |
+
{
|
| 10705 |
+
"epoch": 0.6364110589462703,
|
| 10706 |
+
"grad_norm": 0.1868370622396469,
|
| 10707 |
+
"learning_rate": 0.00043450638430395517,
|
| 10708 |
+
"loss": 1.6155753135681152,
|
| 10709 |
+
"step": 3050
|
| 10710 |
+
},
|
| 10711 |
+
{
|
| 10712 |
+
"epoch": 0.6368283776734481,
|
| 10713 |
+
"grad_norm": 0.17351564764976501,
|
| 10714 |
+
"learning_rate": 0.00043400809716599194,
|
| 10715 |
+
"loss": 1.6848652362823486,
|
| 10716 |
+
"step": 3052
|
| 10717 |
+
},
|
| 10718 |
+
{
|
| 10719 |
+
"epoch": 0.637245696400626,
|
| 10720 |
+
"grad_norm": 0.1712159514427185,
|
| 10721 |
+
"learning_rate": 0.0004335098100280287,
|
| 10722 |
+
"loss": 1.689025640487671,
|
| 10723 |
+
"step": 3054
|
| 10724 |
+
},
|
| 10725 |
+
{
|
| 10726 |
+
"epoch": 0.6376630151278039,
|
| 10727 |
+
"grad_norm": 0.2635113596916199,
|
| 10728 |
+
"learning_rate": 0.0004330115228900654,
|
| 10729 |
+
"loss": 1.5704872608184814,
|
| 10730 |
+
"step": 3056
|
| 10731 |
+
},
|
| 10732 |
+
{
|
| 10733 |
+
"epoch": 0.6380803338549818,
|
| 10734 |
+
"grad_norm": 0.20459075272083282,
|
| 10735 |
+
"learning_rate": 0.0004325132357521022,
|
| 10736 |
+
"loss": 1.628517985343933,
|
| 10737 |
+
"step": 3058
|
| 10738 |
+
},
|
| 10739 |
+
{
|
| 10740 |
+
"epoch": 0.6384976525821596,
|
| 10741 |
+
"grad_norm": 0.19031086564064026,
|
| 10742 |
+
"learning_rate": 0.00043201494861413894,
|
| 10743 |
+
"loss": 1.6251946687698364,
|
| 10744 |
+
"step": 3060
|
| 10745 |
+
},
|
| 10746 |
+
{
|
| 10747 |
+
"epoch": 0.6389149713093375,
|
| 10748 |
+
"grad_norm": 0.2103298157453537,
|
| 10749 |
+
"learning_rate": 0.0004315166614761757,
|
| 10750 |
+
"loss": 1.6923656463623047,
|
| 10751 |
+
"step": 3062
|
| 10752 |
+
},
|
| 10753 |
+
{
|
| 10754 |
+
"epoch": 0.6393322900365154,
|
| 10755 |
+
"grad_norm": 0.15932463109493256,
|
| 10756 |
+
"learning_rate": 0.0004310183743382124,
|
| 10757 |
+
"loss": 1.6129586696624756,
|
| 10758 |
+
"step": 3064
|
| 10759 |
+
},
|
| 10760 |
+
{
|
| 10761 |
+
"epoch": 0.6397496087636932,
|
| 10762 |
+
"grad_norm": 0.17592453956604004,
|
| 10763 |
+
"learning_rate": 0.0004305200872002492,
|
| 10764 |
+
"loss": 1.536402702331543,
|
| 10765 |
+
"step": 3066
|
| 10766 |
+
},
|
| 10767 |
+
{
|
| 10768 |
+
"epoch": 0.6401669274908711,
|
| 10769 |
+
"grad_norm": 0.16769124567508698,
|
| 10770 |
+
"learning_rate": 0.00043002180006228595,
|
| 10771 |
+
"loss": 1.7355247735977173,
|
| 10772 |
+
"step": 3068
|
| 10773 |
+
},
|
| 10774 |
+
{
|
| 10775 |
+
"epoch": 0.640584246218049,
|
| 10776 |
+
"grad_norm": 0.17105732858181,
|
| 10777 |
+
"learning_rate": 0.0004295235129243227,
|
| 10778 |
+
"loss": 1.634315013885498,
|
| 10779 |
+
"step": 3070
|
| 10780 |
+
},
|
| 10781 |
+
{
|
| 10782 |
+
"epoch": 0.6410015649452269,
|
| 10783 |
+
"grad_norm": 0.17621351778507233,
|
| 10784 |
+
"learning_rate": 0.0004290252257863594,
|
| 10785 |
+
"loss": 1.6633837223052979,
|
| 10786 |
+
"step": 3072
|
| 10787 |
+
},
|
| 10788 |
+
{
|
| 10789 |
+
"epoch": 0.6414188836724048,
|
| 10790 |
+
"grad_norm": 0.1565503627061844,
|
| 10791 |
+
"learning_rate": 0.0004285269386483962,
|
| 10792 |
+
"loss": 1.7071964740753174,
|
| 10793 |
+
"step": 3074
|
| 10794 |
+
},
|
| 10795 |
+
{
|
| 10796 |
+
"epoch": 0.6418362023995827,
|
| 10797 |
+
"grad_norm": 0.19788403809070587,
|
| 10798 |
+
"learning_rate": 0.00042802865151043295,
|
| 10799 |
+
"loss": 1.5612205266952515,
|
| 10800 |
+
"step": 3076
|
| 10801 |
+
},
|
| 10802 |
+
{
|
| 10803 |
+
"epoch": 0.6422535211267606,
|
| 10804 |
+
"grad_norm": 0.17223519086837769,
|
| 10805 |
+
"learning_rate": 0.0004275303643724697,
|
| 10806 |
+
"loss": 1.6041910648345947,
|
| 10807 |
+
"step": 3078
|
| 10808 |
+
},
|
| 10809 |
+
{
|
| 10810 |
+
"epoch": 0.6426708398539385,
|
| 10811 |
+
"grad_norm": 0.17532415688037872,
|
| 10812 |
+
"learning_rate": 0.00042703207723450643,
|
| 10813 |
+
"loss": 1.779238224029541,
|
| 10814 |
+
"step": 3080
|
| 10815 |
+
},
|
| 10816 |
+
{
|
| 10817 |
+
"epoch": 0.6430881585811163,
|
| 10818 |
+
"grad_norm": 0.21478240191936493,
|
| 10819 |
+
"learning_rate": 0.0004265337900965432,
|
| 10820 |
+
"loss": 1.5742828845977783,
|
| 10821 |
+
"step": 3082
|
| 10822 |
+
},
|
| 10823 |
+
{
|
| 10824 |
+
"epoch": 0.6435054773082942,
|
| 10825 |
+
"grad_norm": 0.20716965198516846,
|
| 10826 |
+
"learning_rate": 0.00042603550295857996,
|
| 10827 |
+
"loss": 1.6765398979187012,
|
| 10828 |
+
"step": 3084
|
| 10829 |
+
},
|
| 10830 |
+
{
|
| 10831 |
+
"epoch": 0.6439227960354721,
|
| 10832 |
+
"grad_norm": 0.22852440178394318,
|
| 10833 |
+
"learning_rate": 0.0004255372158206166,
|
| 10834 |
+
"loss": 1.6046693325042725,
|
| 10835 |
+
"step": 3086
|
| 10836 |
+
},
|
| 10837 |
+
{
|
| 10838 |
+
"epoch": 0.64434011476265,
|
| 10839 |
+
"grad_norm": 0.18507982790470123,
|
| 10840 |
+
"learning_rate": 0.0004250389286826534,
|
| 10841 |
+
"loss": 1.5526434183120728,
|
| 10842 |
+
"step": 3088
|
| 10843 |
+
},
|
| 10844 |
+
{
|
| 10845 |
+
"epoch": 0.6447574334898278,
|
| 10846 |
+
"grad_norm": 0.17970949411392212,
|
| 10847 |
+
"learning_rate": 0.0004245406415446901,
|
| 10848 |
+
"loss": 1.5689854621887207,
|
| 10849 |
+
"step": 3090
|
| 10850 |
+
},
|
| 10851 |
+
{
|
| 10852 |
+
"epoch": 0.6451747522170057,
|
| 10853 |
+
"grad_norm": 0.18663153052330017,
|
| 10854 |
+
"learning_rate": 0.00042404235440672686,
|
| 10855 |
+
"loss": 1.7312533855438232,
|
| 10856 |
+
"step": 3092
|
| 10857 |
+
},
|
| 10858 |
+
{
|
| 10859 |
+
"epoch": 0.6455920709441836,
|
| 10860 |
+
"grad_norm": 0.20818455517292023,
|
| 10861 |
+
"learning_rate": 0.0004235440672687636,
|
| 10862 |
+
"loss": 1.6105419397354126,
|
| 10863 |
+
"step": 3094
|
| 10864 |
+
},
|
| 10865 |
+
{
|
| 10866 |
+
"epoch": 0.6460093896713615,
|
| 10867 |
+
"grad_norm": 0.15261031687259674,
|
| 10868 |
+
"learning_rate": 0.0004230457801308004,
|
| 10869 |
+
"loss": 1.723541021347046,
|
| 10870 |
+
"step": 3096
|
| 10871 |
+
},
|
| 10872 |
+
{
|
| 10873 |
+
"epoch": 0.6464267083985393,
|
| 10874 |
+
"grad_norm": 0.19954752922058105,
|
| 10875 |
+
"learning_rate": 0.0004225474929928371,
|
| 10876 |
+
"loss": 1.619447946548462,
|
| 10877 |
+
"step": 3098
|
| 10878 |
+
},
|
| 10879 |
+
{
|
| 10880 |
+
"epoch": 0.6468440271257173,
|
| 10881 |
+
"grad_norm": 0.20010371506214142,
|
| 10882 |
+
"learning_rate": 0.00042204920585487386,
|
| 10883 |
+
"loss": 1.6042195558547974,
|
| 10884 |
+
"step": 3100
|
| 10885 |
+
},
|
| 10886 |
+
{
|
| 10887 |
+
"epoch": 0.6472613458528952,
|
| 10888 |
+
"grad_norm": 0.16776025295257568,
|
| 10889 |
+
"learning_rate": 0.0004215509187169106,
|
| 10890 |
+
"loss": 1.804923176765442,
|
| 10891 |
+
"step": 3102
|
| 10892 |
+
},
|
| 10893 |
+
{
|
| 10894 |
+
"epoch": 0.6476786645800731,
|
| 10895 |
+
"grad_norm": 0.19969533383846283,
|
| 10896 |
+
"learning_rate": 0.0004210526315789474,
|
| 10897 |
+
"loss": 1.5725339651107788,
|
| 10898 |
+
"step": 3104
|
| 10899 |
+
},
|
| 10900 |
+
{
|
| 10901 |
+
"epoch": 0.6480959833072509,
|
| 10902 |
+
"grad_norm": 0.18492990732192993,
|
| 10903 |
+
"learning_rate": 0.0004205543444409841,
|
| 10904 |
+
"loss": 1.6459952592849731,
|
| 10905 |
+
"step": 3106
|
| 10906 |
+
},
|
| 10907 |
+
{
|
| 10908 |
+
"epoch": 0.6485133020344288,
|
| 10909 |
+
"grad_norm": 0.19893598556518555,
|
| 10910 |
+
"learning_rate": 0.00042005605730302087,
|
| 10911 |
+
"loss": 1.7250248193740845,
|
| 10912 |
+
"step": 3108
|
| 10913 |
+
},
|
| 10914 |
+
{
|
| 10915 |
+
"epoch": 0.6489306207616067,
|
| 10916 |
+
"grad_norm": 0.19987525045871735,
|
| 10917 |
+
"learning_rate": 0.00041955777016505763,
|
| 10918 |
+
"loss": 1.687385082244873,
|
| 10919 |
+
"step": 3110
|
| 10920 |
+
},
|
| 10921 |
+
{
|
| 10922 |
+
"epoch": 0.6493479394887846,
|
| 10923 |
+
"grad_norm": 0.27105215191841125,
|
| 10924 |
+
"learning_rate": 0.00041905948302709434,
|
| 10925 |
+
"loss": 1.5936906337738037,
|
| 10926 |
+
"step": 3112
|
| 10927 |
+
},
|
| 10928 |
+
{
|
| 10929 |
+
"epoch": 0.6497652582159624,
|
| 10930 |
+
"grad_norm": 0.19652622938156128,
|
| 10931 |
+
"learning_rate": 0.0004185611958891311,
|
| 10932 |
+
"loss": 1.7589534521102905,
|
| 10933 |
+
"step": 3114
|
| 10934 |
+
},
|
| 10935 |
+
{
|
| 10936 |
+
"epoch": 0.6501825769431403,
|
| 10937 |
+
"grad_norm": 0.1824251115322113,
|
| 10938 |
+
"learning_rate": 0.00041806290875116787,
|
| 10939 |
+
"loss": 1.63213050365448,
|
| 10940 |
+
"step": 3116
|
| 10941 |
+
},
|
| 10942 |
+
{
|
| 10943 |
+
"epoch": 0.6505998956703182,
|
| 10944 |
+
"grad_norm": 0.18713854253292084,
|
| 10945 |
+
"learning_rate": 0.00041756462161320464,
|
| 10946 |
+
"loss": 1.6345391273498535,
|
| 10947 |
+
"step": 3118
|
| 10948 |
+
},
|
| 10949 |
+
{
|
| 10950 |
+
"epoch": 0.651017214397496,
|
| 10951 |
+
"grad_norm": 0.17366652190685272,
|
| 10952 |
+
"learning_rate": 0.00041706633447524135,
|
| 10953 |
+
"loss": 1.5746169090270996,
|
| 10954 |
+
"step": 3120
|
| 10955 |
+
},
|
| 10956 |
+
{
|
| 10957 |
+
"epoch": 0.6514345331246739,
|
| 10958 |
+
"grad_norm": 0.18607227504253387,
|
| 10959 |
+
"learning_rate": 0.0004165680473372781,
|
| 10960 |
+
"loss": 1.6285501718521118,
|
| 10961 |
+
"step": 3122
|
| 10962 |
+
},
|
| 10963 |
+
{
|
| 10964 |
+
"epoch": 0.6518518518518519,
|
| 10965 |
+
"grad_norm": 0.17764310538768768,
|
| 10966 |
+
"learning_rate": 0.0004160697601993149,
|
| 10967 |
+
"loss": 1.5877246856689453,
|
| 10968 |
+
"step": 3124
|
| 10969 |
+
},
|
| 10970 |
+
{
|
| 10971 |
+
"epoch": 0.6522691705790298,
|
| 10972 |
+
"grad_norm": 0.18224091827869415,
|
| 10973 |
+
"learning_rate": 0.00041557147306135164,
|
| 10974 |
+
"loss": 1.6738301515579224,
|
| 10975 |
+
"step": 3126
|
| 10976 |
+
},
|
| 10977 |
+
{
|
| 10978 |
+
"epoch": 0.6526864893062077,
|
| 10979 |
+
"grad_norm": 0.2037447690963745,
|
| 10980 |
+
"learning_rate": 0.00041507318592338835,
|
| 10981 |
+
"loss": 1.7492774724960327,
|
| 10982 |
+
"step": 3128
|
| 10983 |
+
},
|
| 10984 |
+
{
|
| 10985 |
+
"epoch": 0.6531038080333855,
|
| 10986 |
+
"grad_norm": 0.19194239377975464,
|
| 10987 |
+
"learning_rate": 0.0004145748987854251,
|
| 10988 |
+
"loss": 1.7368258237838745,
|
| 10989 |
+
"step": 3130
|
| 10990 |
+
},
|
| 10991 |
+
{
|
| 10992 |
+
"epoch": 0.6535211267605634,
|
| 10993 |
+
"grad_norm": 0.16388799250125885,
|
| 10994 |
+
"learning_rate": 0.0004140766116474619,
|
| 10995 |
+
"loss": 1.6800915002822876,
|
| 10996 |
+
"step": 3132
|
| 10997 |
+
},
|
| 10998 |
+
{
|
| 10999 |
+
"epoch": 0.6539384454877413,
|
| 11000 |
+
"grad_norm": 0.19569306075572968,
|
| 11001 |
+
"learning_rate": 0.00041357832450949865,
|
| 11002 |
+
"loss": 1.6806323528289795,
|
| 11003 |
+
"step": 3134
|
| 11004 |
+
},
|
| 11005 |
+
{
|
| 11006 |
+
"epoch": 0.6543557642149191,
|
| 11007 |
+
"grad_norm": 0.19107472896575928,
|
| 11008 |
+
"learning_rate": 0.00041308003737153536,
|
| 11009 |
+
"loss": 1.5883365869522095,
|
| 11010 |
+
"step": 3136
|
| 11011 |
+
},
|
| 11012 |
+
{
|
| 11013 |
+
"epoch": 0.654773082942097,
|
| 11014 |
+
"grad_norm": 0.17638075351715088,
|
| 11015 |
+
"learning_rate": 0.0004125817502335721,
|
| 11016 |
+
"loss": 1.636826992034912,
|
| 11017 |
+
"step": 3138
|
| 11018 |
+
},
|
| 11019 |
+
{
|
| 11020 |
+
"epoch": 0.6551904016692749,
|
| 11021 |
+
"grad_norm": 0.19983480870723724,
|
| 11022 |
+
"learning_rate": 0.0004120834630956089,
|
| 11023 |
+
"loss": 1.6066926717758179,
|
| 11024 |
+
"step": 3140
|
| 11025 |
+
},
|
| 11026 |
+
{
|
| 11027 |
+
"epoch": 0.6556077203964528,
|
| 11028 |
+
"grad_norm": 0.1928826868534088,
|
| 11029 |
+
"learning_rate": 0.00041158517595764565,
|
| 11030 |
+
"loss": 1.6404352188110352,
|
| 11031 |
+
"step": 3142
|
| 11032 |
+
},
|
| 11033 |
+
{
|
| 11034 |
+
"epoch": 0.6560250391236306,
|
| 11035 |
+
"grad_norm": 0.19987136125564575,
|
| 11036 |
+
"learning_rate": 0.00041108688881968237,
|
| 11037 |
+
"loss": 1.7169287204742432,
|
| 11038 |
+
"step": 3144
|
| 11039 |
+
},
|
| 11040 |
+
{
|
| 11041 |
+
"epoch": 0.6564423578508085,
|
| 11042 |
+
"grad_norm": 0.1764020472764969,
|
| 11043 |
+
"learning_rate": 0.00041058860168171913,
|
| 11044 |
+
"loss": 1.6418157815933228,
|
| 11045 |
+
"step": 3146
|
| 11046 |
+
},
|
| 11047 |
+
{
|
| 11048 |
+
"epoch": 0.6568596765779864,
|
| 11049 |
+
"grad_norm": 0.1640889197587967,
|
| 11050 |
+
"learning_rate": 0.0004100903145437559,
|
| 11051 |
+
"loss": 1.641084909439087,
|
| 11052 |
+
"step": 3148
|
| 11053 |
+
},
|
| 11054 |
+
{
|
| 11055 |
+
"epoch": 0.6572769953051644,
|
| 11056 |
+
"grad_norm": 0.1755753457546234,
|
| 11057 |
+
"learning_rate": 0.00040959202740579266,
|
| 11058 |
+
"loss": 1.692739725112915,
|
| 11059 |
+
"step": 3150
|
| 11060 |
+
},
|
| 11061 |
+
{
|
| 11062 |
+
"epoch": 0.6576943140323422,
|
| 11063 |
+
"grad_norm": 0.17704693973064423,
|
| 11064 |
+
"learning_rate": 0.00040909374026782937,
|
| 11065 |
+
"loss": 1.5750787258148193,
|
| 11066 |
+
"step": 3152
|
| 11067 |
+
},
|
| 11068 |
+
{
|
| 11069 |
+
"epoch": 0.6581116327595201,
|
| 11070 |
+
"grad_norm": 0.17037749290466309,
|
| 11071 |
+
"learning_rate": 0.00040859545312986614,
|
| 11072 |
+
"loss": 1.7550424337387085,
|
| 11073 |
+
"step": 3154
|
| 11074 |
+
},
|
| 11075 |
+
{
|
| 11076 |
+
"epoch": 0.658528951486698,
|
| 11077 |
+
"grad_norm": 0.19859790802001953,
|
| 11078 |
+
"learning_rate": 0.0004080971659919029,
|
| 11079 |
+
"loss": 1.6859033107757568,
|
| 11080 |
+
"step": 3156
|
| 11081 |
+
},
|
| 11082 |
+
{
|
| 11083 |
+
"epoch": 0.6589462702138759,
|
| 11084 |
+
"grad_norm": 0.18174374103546143,
|
| 11085 |
+
"learning_rate": 0.0004075988788539396,
|
| 11086 |
+
"loss": 1.6949917078018188,
|
| 11087 |
+
"step": 3158
|
| 11088 |
+
},
|
| 11089 |
+
{
|
| 11090 |
+
"epoch": 0.6593635889410537,
|
| 11091 |
+
"grad_norm": 0.18143464624881744,
|
| 11092 |
+
"learning_rate": 0.0004071005917159764,
|
| 11093 |
+
"loss": 1.66156005859375,
|
| 11094 |
+
"step": 3160
|
| 11095 |
+
},
|
| 11096 |
+
{
|
| 11097 |
+
"epoch": 0.6597809076682316,
|
| 11098 |
+
"grad_norm": 0.18153072893619537,
|
| 11099 |
+
"learning_rate": 0.00040660230457801314,
|
| 11100 |
+
"loss": 1.502579689025879,
|
| 11101 |
+
"step": 3162
|
| 11102 |
+
},
|
| 11103 |
+
{
|
| 11104 |
+
"epoch": 0.6601982263954095,
|
| 11105 |
+
"grad_norm": 0.19244757294654846,
|
| 11106 |
+
"learning_rate": 0.0004061040174400499,
|
| 11107 |
+
"loss": 1.6462054252624512,
|
| 11108 |
+
"step": 3164
|
| 11109 |
+
},
|
| 11110 |
+
{
|
| 11111 |
+
"epoch": 0.6606155451225874,
|
| 11112 |
+
"grad_norm": 0.18178164958953857,
|
| 11113 |
+
"learning_rate": 0.0004056057303020866,
|
| 11114 |
+
"loss": 1.523628830909729,
|
| 11115 |
+
"step": 3166
|
| 11116 |
+
},
|
| 11117 |
+
{
|
| 11118 |
+
"epoch": 0.6610328638497652,
|
| 11119 |
+
"grad_norm": 0.1802728772163391,
|
| 11120 |
+
"learning_rate": 0.00040510744316412333,
|
| 11121 |
+
"loss": 1.7697495222091675,
|
| 11122 |
+
"step": 3168
|
| 11123 |
+
},
|
| 11124 |
+
{
|
| 11125 |
+
"epoch": 0.6614501825769431,
|
| 11126 |
+
"grad_norm": 0.22055105865001678,
|
| 11127 |
+
"learning_rate": 0.00040460915602616004,
|
| 11128 |
+
"loss": 1.6208723783493042,
|
| 11129 |
+
"step": 3170
|
| 11130 |
+
},
|
| 11131 |
+
{
|
| 11132 |
+
"epoch": 0.661867501304121,
|
| 11133 |
+
"grad_norm": 0.22543752193450928,
|
| 11134 |
+
"learning_rate": 0.0004041108688881968,
|
| 11135 |
+
"loss": 1.6431186199188232,
|
| 11136 |
+
"step": 3172
|
| 11137 |
+
},
|
| 11138 |
+
{
|
| 11139 |
+
"epoch": 0.662284820031299,
|
| 11140 |
+
"grad_norm": 0.1864599883556366,
|
| 11141 |
+
"learning_rate": 0.00040361258175023357,
|
| 11142 |
+
"loss": 1.689392328262329,
|
| 11143 |
+
"step": 3174
|
| 11144 |
+
},
|
| 11145 |
+
{
|
| 11146 |
+
"epoch": 0.6627021387584768,
|
| 11147 |
+
"grad_norm": 0.20197515189647675,
|
| 11148 |
+
"learning_rate": 0.0004031142946122703,
|
| 11149 |
+
"loss": 1.6715501546859741,
|
| 11150 |
+
"step": 3176
|
| 11151 |
+
},
|
| 11152 |
+
{
|
| 11153 |
+
"epoch": 0.6631194574856547,
|
| 11154 |
+
"grad_norm": 0.18714426457881927,
|
| 11155 |
+
"learning_rate": 0.00040261600747430704,
|
| 11156 |
+
"loss": 1.7265963554382324,
|
| 11157 |
+
"step": 3178
|
| 11158 |
+
},
|
| 11159 |
+
{
|
| 11160 |
+
"epoch": 0.6635367762128326,
|
| 11161 |
+
"grad_norm": 0.17732010781764984,
|
| 11162 |
+
"learning_rate": 0.0004021177203363438,
|
| 11163 |
+
"loss": 1.7067770957946777,
|
| 11164 |
+
"step": 3180
|
| 11165 |
+
},
|
| 11166 |
+
{
|
| 11167 |
+
"epoch": 0.6639540949400105,
|
| 11168 |
+
"grad_norm": 0.20761427283287048,
|
| 11169 |
+
"learning_rate": 0.0004016194331983806,
|
| 11170 |
+
"loss": 1.4576196670532227,
|
| 11171 |
+
"step": 3182
|
| 11172 |
+
},
|
| 11173 |
+
{
|
| 11174 |
+
"epoch": 0.6643714136671883,
|
| 11175 |
+
"grad_norm": 0.16294176876544952,
|
| 11176 |
+
"learning_rate": 0.0004011211460604173,
|
| 11177 |
+
"loss": 1.7235853672027588,
|
| 11178 |
+
"step": 3184
|
| 11179 |
+
},
|
| 11180 |
+
{
|
| 11181 |
+
"epoch": 0.6647887323943662,
|
| 11182 |
+
"grad_norm": 0.20277062058448792,
|
| 11183 |
+
"learning_rate": 0.00040062285892245405,
|
| 11184 |
+
"loss": 1.5236420631408691,
|
| 11185 |
+
"step": 3186
|
| 11186 |
+
},
|
| 11187 |
+
{
|
| 11188 |
+
"epoch": 0.6652060511215441,
|
| 11189 |
+
"grad_norm": 0.21320201456546783,
|
| 11190 |
+
"learning_rate": 0.0004001245717844908,
|
| 11191 |
+
"loss": 1.6254127025604248,
|
| 11192 |
+
"step": 3188
|
| 11193 |
+
},
|
| 11194 |
+
{
|
| 11195 |
+
"epoch": 0.665623369848722,
|
| 11196 |
+
"grad_norm": 0.1975260078907013,
|
| 11197 |
+
"learning_rate": 0.0003996262846465276,
|
| 11198 |
+
"loss": 1.716719627380371,
|
| 11199 |
+
"step": 3190
|
| 11200 |
+
},
|
| 11201 |
+
{
|
| 11202 |
+
"epoch": 0.6660406885758998,
|
| 11203 |
+
"grad_norm": 0.17935079336166382,
|
| 11204 |
+
"learning_rate": 0.0003991279975085643,
|
| 11205 |
+
"loss": 1.7938005924224854,
|
| 11206 |
+
"step": 3192
|
| 11207 |
+
},
|
| 11208 |
+
{
|
| 11209 |
+
"epoch": 0.6664580073030777,
|
| 11210 |
+
"grad_norm": 0.19730965793132782,
|
| 11211 |
+
"learning_rate": 0.00039862971037060106,
|
| 11212 |
+
"loss": 1.5672612190246582,
|
| 11213 |
+
"step": 3194
|
| 11214 |
+
},
|
| 11215 |
+
{
|
| 11216 |
+
"epoch": 0.6668753260302556,
|
| 11217 |
+
"grad_norm": 0.16387251019477844,
|
| 11218 |
+
"learning_rate": 0.0003981314232326378,
|
| 11219 |
+
"loss": 1.551000952720642,
|
| 11220 |
+
"step": 3196
|
| 11221 |
+
},
|
| 11222 |
+
{
|
| 11223 |
+
"epoch": 0.6672926447574334,
|
| 11224 |
+
"grad_norm": 0.18270333111286163,
|
| 11225 |
+
"learning_rate": 0.0003976331360946746,
|
| 11226 |
+
"loss": 1.6228139400482178,
|
| 11227 |
+
"step": 3198
|
| 11228 |
+
},
|
| 11229 |
+
{
|
| 11230 |
+
"epoch": 0.6677099634846114,
|
| 11231 |
+
"grad_norm": 0.19708923995494843,
|
| 11232 |
+
"learning_rate": 0.0003971348489567113,
|
| 11233 |
+
"loss": 1.6735169887542725,
|
| 11234 |
+
"step": 3200
|
| 11235 |
+
},
|
| 11236 |
+
{
|
| 11237 |
+
"epoch": 0.6681272822117893,
|
| 11238 |
+
"grad_norm": 0.20654243230819702,
|
| 11239 |
+
"learning_rate": 0.00039663656181874806,
|
| 11240 |
+
"loss": 1.555161476135254,
|
| 11241 |
+
"step": 3202
|
| 11242 |
+
},
|
| 11243 |
+
{
|
| 11244 |
+
"epoch": 0.6685446009389672,
|
| 11245 |
+
"grad_norm": 0.17933540046215057,
|
| 11246 |
+
"learning_rate": 0.0003961382746807848,
|
| 11247 |
+
"loss": 1.6258504390716553,
|
| 11248 |
+
"step": 3204
|
| 11249 |
+
},
|
| 11250 |
+
{
|
| 11251 |
+
"epoch": 0.668961919666145,
|
| 11252 |
+
"grad_norm": 0.24534529447555542,
|
| 11253 |
+
"learning_rate": 0.0003956399875428216,
|
| 11254 |
+
"loss": 1.6228227615356445,
|
| 11255 |
+
"step": 3206
|
| 11256 |
+
},
|
| 11257 |
+
{
|
| 11258 |
+
"epoch": 0.6693792383933229,
|
| 11259 |
+
"grad_norm": 0.2009853720664978,
|
| 11260 |
+
"learning_rate": 0.0003951417004048583,
|
| 11261 |
+
"loss": 1.673019289970398,
|
| 11262 |
+
"step": 3208
|
| 11263 |
+
},
|
| 11264 |
+
{
|
| 11265 |
+
"epoch": 0.6697965571205008,
|
| 11266 |
+
"grad_norm": 0.18508794903755188,
|
| 11267 |
+
"learning_rate": 0.00039464341326689507,
|
| 11268 |
+
"loss": 1.5413572788238525,
|
| 11269 |
+
"step": 3210
|
| 11270 |
+
},
|
| 11271 |
+
{
|
| 11272 |
+
"epoch": 0.6702138758476787,
|
| 11273 |
+
"grad_norm": 0.1912604719400406,
|
| 11274 |
+
"learning_rate": 0.00039414512612893183,
|
| 11275 |
+
"loss": 1.593103051185608,
|
| 11276 |
+
"step": 3212
|
| 11277 |
+
},
|
| 11278 |
+
{
|
| 11279 |
+
"epoch": 0.6706311945748565,
|
| 11280 |
+
"grad_norm": 0.1695089191198349,
|
| 11281 |
+
"learning_rate": 0.0003936468389909686,
|
| 11282 |
+
"loss": 1.6970627307891846,
|
| 11283 |
+
"step": 3214
|
| 11284 |
+
},
|
| 11285 |
+
{
|
| 11286 |
+
"epoch": 0.6710485133020344,
|
| 11287 |
+
"grad_norm": 0.19945579767227173,
|
| 11288 |
+
"learning_rate": 0.0003931485518530053,
|
| 11289 |
+
"loss": 1.754443883895874,
|
| 11290 |
+
"step": 3216
|
| 11291 |
+
},
|
| 11292 |
+
{
|
| 11293 |
+
"epoch": 0.6714658320292123,
|
| 11294 |
+
"grad_norm": 0.20188401639461517,
|
| 11295 |
+
"learning_rate": 0.00039265026471504207,
|
| 11296 |
+
"loss": 1.630683422088623,
|
| 11297 |
+
"step": 3218
|
| 11298 |
+
},
|
| 11299 |
+
{
|
| 11300 |
+
"epoch": 0.6718831507563902,
|
| 11301 |
+
"grad_norm": 0.18985597789287567,
|
| 11302 |
+
"learning_rate": 0.00039215197757707884,
|
| 11303 |
+
"loss": 1.5402309894561768,
|
| 11304 |
+
"step": 3220
|
| 11305 |
+
},
|
| 11306 |
+
{
|
| 11307 |
+
"epoch": 0.672300469483568,
|
| 11308 |
+
"grad_norm": 0.17440970242023468,
|
| 11309 |
+
"learning_rate": 0.00039165369043911555,
|
| 11310 |
+
"loss": 1.7405648231506348,
|
| 11311 |
+
"step": 3222
|
| 11312 |
+
},
|
| 11313 |
+
{
|
| 11314 |
+
"epoch": 0.6727177882107459,
|
| 11315 |
+
"grad_norm": 0.1797809600830078,
|
| 11316 |
+
"learning_rate": 0.0003911554033011523,
|
| 11317 |
+
"loss": 1.7436715364456177,
|
| 11318 |
+
"step": 3224
|
| 11319 |
+
},
|
| 11320 |
+
{
|
| 11321 |
+
"epoch": 0.6731351069379239,
|
| 11322 |
+
"grad_norm": 0.18043005466461182,
|
| 11323 |
+
"learning_rate": 0.0003906571161631891,
|
| 11324 |
+
"loss": 1.6429089307785034,
|
| 11325 |
+
"step": 3226
|
| 11326 |
+
},
|
| 11327 |
+
{
|
| 11328 |
+
"epoch": 0.6735524256651018,
|
| 11329 |
+
"grad_norm": 0.19399769604206085,
|
| 11330 |
+
"learning_rate": 0.00039015882902522584,
|
| 11331 |
+
"loss": 1.715533971786499,
|
| 11332 |
+
"step": 3228
|
| 11333 |
+
},
|
| 11334 |
+
{
|
| 11335 |
+
"epoch": 0.6739697443922796,
|
| 11336 |
+
"grad_norm": 0.21315521001815796,
|
| 11337 |
+
"learning_rate": 0.00038966054188726255,
|
| 11338 |
+
"loss": 1.5649092197418213,
|
| 11339 |
+
"step": 3230
|
| 11340 |
+
},
|
| 11341 |
+
{
|
| 11342 |
+
"epoch": 0.6743870631194575,
|
| 11343 |
+
"grad_norm": 0.17036151885986328,
|
| 11344 |
+
"learning_rate": 0.00038916225474929926,
|
| 11345 |
+
"loss": 1.6871109008789062,
|
| 11346 |
+
"step": 3232
|
| 11347 |
+
},
|
| 11348 |
+
{
|
| 11349 |
+
"epoch": 0.6748043818466354,
|
| 11350 |
+
"grad_norm": 0.21142089366912842,
|
| 11351 |
+
"learning_rate": 0.00038866396761133603,
|
| 11352 |
+
"loss": 1.5012352466583252,
|
| 11353 |
+
"step": 3234
|
| 11354 |
+
},
|
| 11355 |
+
{
|
| 11356 |
+
"epoch": 0.6752217005738133,
|
| 11357 |
+
"grad_norm": 0.1776968389749527,
|
| 11358 |
+
"learning_rate": 0.0003881656804733728,
|
| 11359 |
+
"loss": 1.715356707572937,
|
| 11360 |
+
"step": 3236
|
| 11361 |
+
},
|
| 11362 |
+
{
|
| 11363 |
+
"epoch": 0.6756390193009911,
|
| 11364 |
+
"grad_norm": 0.167434424161911,
|
| 11365 |
+
"learning_rate": 0.00038766739333540956,
|
| 11366 |
+
"loss": 1.761622428894043,
|
| 11367 |
+
"step": 3238
|
| 11368 |
+
},
|
| 11369 |
+
{
|
| 11370 |
+
"epoch": 0.676056338028169,
|
| 11371 |
+
"grad_norm": 0.19718925654888153,
|
| 11372 |
+
"learning_rate": 0.00038716910619744627,
|
| 11373 |
+
"loss": 1.716399908065796,
|
| 11374 |
+
"step": 3240
|
| 11375 |
+
},
|
| 11376 |
+
{
|
| 11377 |
+
"epoch": 0.6764736567553469,
|
| 11378 |
+
"grad_norm": 0.18885953724384308,
|
| 11379 |
+
"learning_rate": 0.00038667081905948303,
|
| 11380 |
+
"loss": 1.545933485031128,
|
| 11381 |
+
"step": 3242
|
| 11382 |
+
},
|
| 11383 |
+
{
|
| 11384 |
+
"epoch": 0.6768909754825247,
|
| 11385 |
+
"grad_norm": 0.1755514144897461,
|
| 11386 |
+
"learning_rate": 0.0003861725319215198,
|
| 11387 |
+
"loss": 1.606783151626587,
|
| 11388 |
+
"step": 3244
|
| 11389 |
+
},
|
| 11390 |
+
{
|
| 11391 |
+
"epoch": 0.6773082942097026,
|
| 11392 |
+
"grad_norm": 0.19303053617477417,
|
| 11393 |
+
"learning_rate": 0.00038567424478355656,
|
| 11394 |
+
"loss": 1.5963878631591797,
|
| 11395 |
+
"step": 3246
|
| 11396 |
+
},
|
| 11397 |
+
{
|
| 11398 |
+
"epoch": 0.6777256129368805,
|
| 11399 |
+
"grad_norm": 0.16974513232707977,
|
| 11400 |
+
"learning_rate": 0.0003851759576455933,
|
| 11401 |
+
"loss": 1.658547043800354,
|
| 11402 |
+
"step": 3248
|
| 11403 |
+
},
|
| 11404 |
+
{
|
| 11405 |
+
"epoch": 0.6781429316640585,
|
| 11406 |
+
"grad_norm": 0.21708205342292786,
|
| 11407 |
+
"learning_rate": 0.00038467767050763004,
|
| 11408 |
+
"loss": 1.4607315063476562,
|
| 11409 |
+
"step": 3250
|
| 11410 |
+
},
|
| 11411 |
+
{
|
| 11412 |
+
"epoch": 0.6785602503912364,
|
| 11413 |
+
"grad_norm": 0.21690596640110016,
|
| 11414 |
+
"learning_rate": 0.0003841793833696668,
|
| 11415 |
+
"loss": 1.6367452144622803,
|
| 11416 |
+
"step": 3252
|
| 11417 |
+
},
|
| 11418 |
+
{
|
| 11419 |
+
"epoch": 0.6789775691184142,
|
| 11420 |
+
"grad_norm": 0.21128971874713898,
|
| 11421 |
+
"learning_rate": 0.0003836810962317035,
|
| 11422 |
+
"loss": 1.663558006286621,
|
| 11423 |
+
"step": 3254
|
| 11424 |
+
},
|
| 11425 |
+
{
|
| 11426 |
+
"epoch": 0.6793948878455921,
|
| 11427 |
+
"grad_norm": 0.1995653212070465,
|
| 11428 |
+
"learning_rate": 0.0003831828090937403,
|
| 11429 |
+
"loss": 1.718706488609314,
|
| 11430 |
+
"step": 3256
|
| 11431 |
+
},
|
| 11432 |
+
{
|
| 11433 |
+
"epoch": 0.67981220657277,
|
| 11434 |
+
"grad_norm": 0.17688687145709991,
|
| 11435 |
+
"learning_rate": 0.00038268452195577705,
|
| 11436 |
+
"loss": 1.6101213693618774,
|
| 11437 |
+
"step": 3258
|
| 11438 |
+
},
|
| 11439 |
+
{
|
| 11440 |
+
"epoch": 0.6802295252999478,
|
| 11441 |
+
"grad_norm": 0.17573122680187225,
|
| 11442 |
+
"learning_rate": 0.0003821862348178138,
|
| 11443 |
+
"loss": 1.6461971998214722,
|
| 11444 |
+
"step": 3260
|
| 11445 |
+
},
|
| 11446 |
+
{
|
| 11447 |
+
"epoch": 0.6806468440271257,
|
| 11448 |
+
"grad_norm": 0.16721221804618835,
|
| 11449 |
+
"learning_rate": 0.0003816879476798505,
|
| 11450 |
+
"loss": 1.688936710357666,
|
| 11451 |
+
"step": 3262
|
| 11452 |
+
},
|
| 11453 |
+
{
|
| 11454 |
+
"epoch": 0.6810641627543036,
|
| 11455 |
+
"grad_norm": 0.19553811848163605,
|
| 11456 |
+
"learning_rate": 0.0003811896605418873,
|
| 11457 |
+
"loss": 1.7398422956466675,
|
| 11458 |
+
"step": 3264
|
| 11459 |
+
},
|
| 11460 |
+
{
|
| 11461 |
+
"epoch": 0.6814814814814815,
|
| 11462 |
+
"grad_norm": 0.17369849979877472,
|
| 11463 |
+
"learning_rate": 0.00038069137340392405,
|
| 11464 |
+
"loss": 1.6012886762619019,
|
| 11465 |
+
"step": 3266
|
| 11466 |
+
},
|
| 11467 |
+
{
|
| 11468 |
+
"epoch": 0.6818988002086593,
|
| 11469 |
+
"grad_norm": 0.18755407631397247,
|
| 11470 |
+
"learning_rate": 0.0003801930862659608,
|
| 11471 |
+
"loss": 1.5039241313934326,
|
| 11472 |
+
"step": 3268
|
| 11473 |
+
},
|
| 11474 |
+
{
|
| 11475 |
+
"epoch": 0.6823161189358372,
|
| 11476 |
+
"grad_norm": 0.1913551688194275,
|
| 11477 |
+
"learning_rate": 0.0003796947991279975,
|
| 11478 |
+
"loss": 1.6101913452148438,
|
| 11479 |
+
"step": 3270
|
| 11480 |
+
},
|
| 11481 |
+
{
|
| 11482 |
+
"epoch": 0.6827334376630151,
|
| 11483 |
+
"grad_norm": 0.19303348660469055,
|
| 11484 |
+
"learning_rate": 0.00037919651199003424,
|
| 11485 |
+
"loss": 1.6337401866912842,
|
| 11486 |
+
"step": 3272
|
| 11487 |
+
},
|
| 11488 |
+
{
|
| 11489 |
+
"epoch": 0.683150756390193,
|
| 11490 |
+
"grad_norm": 0.21885934472084045,
|
| 11491 |
+
"learning_rate": 0.000378698224852071,
|
| 11492 |
+
"loss": 1.6370759010314941,
|
| 11493 |
+
"step": 3274
|
| 11494 |
+
},
|
| 11495 |
+
{
|
| 11496 |
+
"epoch": 0.6835680751173709,
|
| 11497 |
+
"grad_norm": 0.16648758947849274,
|
| 11498 |
+
"learning_rate": 0.00037819993771410777,
|
| 11499 |
+
"loss": 1.5797863006591797,
|
| 11500 |
+
"step": 3276
|
| 11501 |
+
},
|
| 11502 |
+
{
|
| 11503 |
+
"epoch": 0.6839853938445488,
|
| 11504 |
+
"grad_norm": 0.1816018521785736,
|
| 11505 |
+
"learning_rate": 0.00037770165057614453,
|
| 11506 |
+
"loss": 1.61228346824646,
|
| 11507 |
+
"step": 3278
|
| 11508 |
+
},
|
| 11509 |
+
{
|
| 11510 |
+
"epoch": 0.6844027125717267,
|
| 11511 |
+
"grad_norm": 0.21701984107494354,
|
| 11512 |
+
"learning_rate": 0.00037720336343818124,
|
| 11513 |
+
"loss": 1.5739727020263672,
|
| 11514 |
+
"step": 3280
|
| 11515 |
+
},
|
| 11516 |
+
{
|
| 11517 |
+
"epoch": 0.6848200312989046,
|
| 11518 |
+
"grad_norm": 0.2241283357143402,
|
| 11519 |
+
"learning_rate": 0.000376705076300218,
|
| 11520 |
+
"loss": 1.674843668937683,
|
| 11521 |
+
"step": 3282
|
| 11522 |
+
},
|
| 11523 |
+
{
|
| 11524 |
+
"epoch": 0.6852373500260824,
|
| 11525 |
+
"grad_norm": 0.17072917520999908,
|
| 11526 |
+
"learning_rate": 0.00037620678916225477,
|
| 11527 |
+
"loss": 1.6810898780822754,
|
| 11528 |
+
"step": 3284
|
| 11529 |
+
},
|
| 11530 |
+
{
|
| 11531 |
+
"epoch": 0.6856546687532603,
|
| 11532 |
+
"grad_norm": 0.1833476424217224,
|
| 11533 |
+
"learning_rate": 0.0003757085020242915,
|
| 11534 |
+
"loss": 1.663867712020874,
|
| 11535 |
+
"step": 3286
|
| 11536 |
+
},
|
| 11537 |
+
{
|
| 11538 |
+
"epoch": 0.6860719874804382,
|
| 11539 |
+
"grad_norm": 0.16291077435016632,
|
| 11540 |
+
"learning_rate": 0.00037521021488632825,
|
| 11541 |
+
"loss": 1.611328125,
|
| 11542 |
+
"step": 3288
|
| 11543 |
+
},
|
| 11544 |
+
{
|
| 11545 |
+
"epoch": 0.6864893062076161,
|
| 11546 |
+
"grad_norm": 0.21334248781204224,
|
| 11547 |
+
"learning_rate": 0.000374711927748365,
|
| 11548 |
+
"loss": 1.6744343042373657,
|
| 11549 |
+
"step": 3290
|
| 11550 |
+
},
|
| 11551 |
+
{
|
| 11552 |
+
"epoch": 0.6869066249347939,
|
| 11553 |
+
"grad_norm": 0.18318229913711548,
|
| 11554 |
+
"learning_rate": 0.0003742136406104018,
|
| 11555 |
+
"loss": 1.687151312828064,
|
| 11556 |
+
"step": 3292
|
| 11557 |
+
},
|
| 11558 |
+
{
|
| 11559 |
+
"epoch": 0.6873239436619718,
|
| 11560 |
+
"grad_norm": 0.2117055207490921,
|
| 11561 |
+
"learning_rate": 0.0003737153534724385,
|
| 11562 |
+
"loss": 1.5922801494598389,
|
| 11563 |
+
"step": 3294
|
| 11564 |
+
},
|
| 11565 |
+
{
|
| 11566 |
+
"epoch": 0.6877412623891497,
|
| 11567 |
+
"grad_norm": 0.17517757415771484,
|
| 11568 |
+
"learning_rate": 0.00037321706633447525,
|
| 11569 |
+
"loss": 1.7506741285324097,
|
| 11570 |
+
"step": 3296
|
| 11571 |
+
},
|
| 11572 |
+
{
|
| 11573 |
+
"epoch": 0.6881585811163276,
|
| 11574 |
+
"grad_norm": 0.18518812954425812,
|
| 11575 |
+
"learning_rate": 0.000372718779196512,
|
| 11576 |
+
"loss": 1.6906851530075073,
|
| 11577 |
+
"step": 3298
|
| 11578 |
+
},
|
| 11579 |
+
{
|
| 11580 |
+
"epoch": 0.6885758998435054,
|
| 11581 |
+
"grad_norm": 0.1808939278125763,
|
| 11582 |
+
"learning_rate": 0.0003722204920585488,
|
| 11583 |
+
"loss": 1.5298922061920166,
|
| 11584 |
+
"step": 3300
|
| 11585 |
+
},
|
| 11586 |
+
{
|
| 11587 |
+
"epoch": 0.6889932185706834,
|
| 11588 |
+
"grad_norm": 0.1775629073381424,
|
| 11589 |
+
"learning_rate": 0.0003717222049205855,
|
| 11590 |
+
"loss": 1.6584842205047607,
|
| 11591 |
+
"step": 3302
|
| 11592 |
+
},
|
| 11593 |
+
{
|
| 11594 |
+
"epoch": 0.6894105372978613,
|
| 11595 |
+
"grad_norm": 0.17619644105434418,
|
| 11596 |
+
"learning_rate": 0.00037122391778262226,
|
| 11597 |
+
"loss": 1.658012866973877,
|
| 11598 |
+
"step": 3304
|
| 11599 |
+
},
|
| 11600 |
+
{
|
| 11601 |
+
"epoch": 0.6898278560250392,
|
| 11602 |
+
"grad_norm": 0.18246251344680786,
|
| 11603 |
+
"learning_rate": 0.000370725630644659,
|
| 11604 |
+
"loss": 1.6148688793182373,
|
| 11605 |
+
"step": 3306
|
| 11606 |
+
},
|
| 11607 |
+
{
|
| 11608 |
+
"epoch": 0.690245174752217,
|
| 11609 |
+
"grad_norm": 0.17772823572158813,
|
| 11610 |
+
"learning_rate": 0.0003702273435066958,
|
| 11611 |
+
"loss": 1.6593983173370361,
|
| 11612 |
+
"step": 3308
|
| 11613 |
+
},
|
| 11614 |
+
{
|
| 11615 |
+
"epoch": 0.6906624934793949,
|
| 11616 |
+
"grad_norm": 0.17234240472316742,
|
| 11617 |
+
"learning_rate": 0.0003697290563687325,
|
| 11618 |
+
"loss": 1.6848957538604736,
|
| 11619 |
+
"step": 3310
|
| 11620 |
+
},
|
| 11621 |
+
{
|
| 11622 |
+
"epoch": 0.6910798122065728,
|
| 11623 |
+
"grad_norm": 0.18917346000671387,
|
| 11624 |
+
"learning_rate": 0.00036923076923076927,
|
| 11625 |
+
"loss": 1.5930238962173462,
|
| 11626 |
+
"step": 3312
|
| 11627 |
+
},
|
| 11628 |
+
{
|
| 11629 |
+
"epoch": 0.6914971309337506,
|
| 11630 |
+
"grad_norm": 0.17682476341724396,
|
| 11631 |
+
"learning_rate": 0.000368732482092806,
|
| 11632 |
+
"loss": 1.7607903480529785,
|
| 11633 |
+
"step": 3314
|
| 11634 |
+
},
|
| 11635 |
+
{
|
| 11636 |
+
"epoch": 0.6919144496609285,
|
| 11637 |
+
"grad_norm": 0.17324304580688477,
|
| 11638 |
+
"learning_rate": 0.00036823419495484274,
|
| 11639 |
+
"loss": 1.6287239789962769,
|
| 11640 |
+
"step": 3316
|
| 11641 |
+
},
|
| 11642 |
+
{
|
| 11643 |
+
"epoch": 0.6923317683881064,
|
| 11644 |
+
"grad_norm": 0.18196740746498108,
|
| 11645 |
+
"learning_rate": 0.00036773590781687945,
|
| 11646 |
+
"loss": 1.6950976848602295,
|
| 11647 |
+
"step": 3318
|
| 11648 |
+
},
|
| 11649 |
+
{
|
| 11650 |
+
"epoch": 0.6927490871152843,
|
| 11651 |
+
"grad_norm": 0.18560899794101715,
|
| 11652 |
+
"learning_rate": 0.0003672376206789162,
|
| 11653 |
+
"loss": 1.6717556715011597,
|
| 11654 |
+
"step": 3320
|
| 11655 |
+
},
|
| 11656 |
+
{
|
| 11657 |
+
"epoch": 0.6931664058424621,
|
| 11658 |
+
"grad_norm": 0.1800072193145752,
|
| 11659 |
+
"learning_rate": 0.000366739333540953,
|
| 11660 |
+
"loss": 1.6935051679611206,
|
| 11661 |
+
"step": 3322
|
| 11662 |
+
},
|
| 11663 |
+
{
|
| 11664 |
+
"epoch": 0.69358372456964,
|
| 11665 |
+
"grad_norm": 0.1855444610118866,
|
| 11666 |
+
"learning_rate": 0.00036624104640298975,
|
| 11667 |
+
"loss": 1.6905603408813477,
|
| 11668 |
+
"step": 3324
|
| 11669 |
+
},
|
| 11670 |
+
{
|
| 11671 |
+
"epoch": 0.694001043296818,
|
| 11672 |
+
"grad_norm": 0.20639169216156006,
|
| 11673 |
+
"learning_rate": 0.00036574275926502646,
|
| 11674 |
+
"loss": 1.5048364400863647,
|
| 11675 |
+
"step": 3326
|
| 11676 |
+
},
|
| 11677 |
+
{
|
| 11678 |
+
"epoch": 0.6944183620239959,
|
| 11679 |
+
"grad_norm": 0.17336873710155487,
|
| 11680 |
+
"learning_rate": 0.0003652444721270632,
|
| 11681 |
+
"loss": 1.5049575567245483,
|
| 11682 |
+
"step": 3328
|
| 11683 |
+
},
|
| 11684 |
+
{
|
| 11685 |
+
"epoch": 0.6948356807511737,
|
| 11686 |
+
"grad_norm": 0.18017591536045074,
|
| 11687 |
+
"learning_rate": 0.0003647461849891,
|
| 11688 |
+
"loss": 1.74604332447052,
|
| 11689 |
+
"step": 3330
|
| 11690 |
+
},
|
| 11691 |
+
{
|
| 11692 |
+
"epoch": 0.6952529994783516,
|
| 11693 |
+
"grad_norm": 0.17093011736869812,
|
| 11694 |
+
"learning_rate": 0.00036424789785113675,
|
| 11695 |
+
"loss": 1.5602046251296997,
|
| 11696 |
+
"step": 3332
|
| 11697 |
+
},
|
| 11698 |
+
{
|
| 11699 |
+
"epoch": 0.6956703182055295,
|
| 11700 |
+
"grad_norm": 0.23787932097911835,
|
| 11701 |
+
"learning_rate": 0.00036374961071317346,
|
| 11702 |
+
"loss": 1.6649483442306519,
|
| 11703 |
+
"step": 3334
|
| 11704 |
+
},
|
| 11705 |
+
{
|
| 11706 |
+
"epoch": 0.6960876369327074,
|
| 11707 |
+
"grad_norm": 0.18329249322414398,
|
| 11708 |
+
"learning_rate": 0.00036325132357521023,
|
| 11709 |
+
"loss": 1.697551965713501,
|
| 11710 |
+
"step": 3336
|
| 11711 |
+
},
|
| 11712 |
+
{
|
| 11713 |
+
"epoch": 0.6965049556598852,
|
| 11714 |
+
"grad_norm": 0.17169825732707977,
|
| 11715 |
+
"learning_rate": 0.000362753036437247,
|
| 11716 |
+
"loss": 1.6272770166397095,
|
| 11717 |
+
"step": 3338
|
| 11718 |
+
},
|
| 11719 |
+
{
|
| 11720 |
+
"epoch": 0.6969222743870631,
|
| 11721 |
+
"grad_norm": 0.18326887488365173,
|
| 11722 |
+
"learning_rate": 0.00036225474929928376,
|
| 11723 |
+
"loss": 1.6268222332000732,
|
| 11724 |
+
"step": 3340
|
| 11725 |
+
},
|
| 11726 |
+
{
|
| 11727 |
+
"epoch": 0.697339593114241,
|
| 11728 |
+
"grad_norm": 0.1736290156841278,
|
| 11729 |
+
"learning_rate": 0.00036175646216132047,
|
| 11730 |
+
"loss": 1.592427372932434,
|
| 11731 |
+
"step": 3342
|
| 11732 |
+
},
|
| 11733 |
+
{
|
| 11734 |
+
"epoch": 0.6977569118414189,
|
| 11735 |
+
"grad_norm": 0.21996745467185974,
|
| 11736 |
+
"learning_rate": 0.00036125817502335723,
|
| 11737 |
+
"loss": 1.6270965337753296,
|
| 11738 |
+
"step": 3344
|
| 11739 |
+
},
|
| 11740 |
+
{
|
| 11741 |
+
"epoch": 0.6981742305685967,
|
| 11742 |
+
"grad_norm": 0.1685713231563568,
|
| 11743 |
+
"learning_rate": 0.000360759887885394,
|
| 11744 |
+
"loss": 1.633894920349121,
|
| 11745 |
+
"step": 3346
|
| 11746 |
+
},
|
| 11747 |
+
{
|
| 11748 |
+
"epoch": 0.6985915492957746,
|
| 11749 |
+
"grad_norm": 0.18318989872932434,
|
| 11750 |
+
"learning_rate": 0.00036026160074743076,
|
| 11751 |
+
"loss": 1.701606035232544,
|
| 11752 |
+
"step": 3348
|
| 11753 |
+
},
|
| 11754 |
+
{
|
| 11755 |
+
"epoch": 0.6990088680229525,
|
| 11756 |
+
"grad_norm": 0.22740933299064636,
|
| 11757 |
+
"learning_rate": 0.0003597633136094675,
|
| 11758 |
+
"loss": 1.625685453414917,
|
| 11759 |
+
"step": 3350
|
| 11760 |
+
},
|
| 11761 |
+
{
|
| 11762 |
+
"epoch": 0.6994261867501305,
|
| 11763 |
+
"grad_norm": 0.15949927270412445,
|
| 11764 |
+
"learning_rate": 0.00035926502647150424,
|
| 11765 |
+
"loss": 1.7595469951629639,
|
| 11766 |
+
"step": 3352
|
| 11767 |
+
},
|
| 11768 |
+
{
|
| 11769 |
+
"epoch": 0.6998435054773083,
|
| 11770 |
+
"grad_norm": 0.1841728687286377,
|
| 11771 |
+
"learning_rate": 0.000358766739333541,
|
| 11772 |
+
"loss": 1.6015747785568237,
|
| 11773 |
+
"step": 3354
|
| 11774 |
+
},
|
| 11775 |
+
{
|
| 11776 |
+
"epoch": 0.7002608242044862,
|
| 11777 |
+
"grad_norm": 0.20003744959831238,
|
| 11778 |
+
"learning_rate": 0.0003582684521955777,
|
| 11779 |
+
"loss": 1.5967538356781006,
|
| 11780 |
+
"step": 3356
|
| 11781 |
+
},
|
| 11782 |
+
{
|
| 11783 |
+
"epoch": 0.7006781429316641,
|
| 11784 |
+
"grad_norm": 0.1791495531797409,
|
| 11785 |
+
"learning_rate": 0.0003577701650576144,
|
| 11786 |
+
"loss": 1.5644580125808716,
|
| 11787 |
+
"step": 3358
|
| 11788 |
+
},
|
| 11789 |
+
{
|
| 11790 |
+
"epoch": 0.701095461658842,
|
| 11791 |
+
"grad_norm": 0.176228329539299,
|
| 11792 |
+
"learning_rate": 0.0003572718779196512,
|
| 11793 |
+
"loss": 1.608605980873108,
|
| 11794 |
+
"step": 3360
|
| 11795 |
}
|
| 11796 |
],
|
| 11797 |
"logging_steps": 2,
|
|
|
|
| 11811 |
"attributes": {}
|
| 11812 |
}
|
| 11813 |
},
|
| 11814 |
+
"total_flos": 7.660138095430464e+17,
|
| 11815 |
"train_batch_size": 8,
|
| 11816 |
"trial_name": null,
|
| 11817 |
"trial_params": null
|