Training in progress, step 5736, checkpoint
Browse files- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +1676 -3
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 516802328
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ebef31135b45565df6ed5f3402d41965c0085e3579054017540bd9502336d5bc
|
| 3 |
size 516802328
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9410d6faf0b0b2a20a9f5d7e528ee5c22798ebdce85f3ee5ebc57c74c84b3d40
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f2d74ffd31206d3e81b71ea7d00d130021e2a8ddf2f551c8cf2a0016186f7977
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -38486,6 +38486,1679 @@
|
|
| 38486 |
"learning_rate": 1.4425382207839954e-06,
|
| 38487 |
"loss": 0.0,
|
| 38488 |
"step": 5497
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38489 |
}
|
| 38490 |
],
|
| 38491 |
"logging_steps": 1,
|
|
@@ -38505,7 +40178,7 @@
|
|
| 38505 |
"attributes": {}
|
| 38506 |
}
|
| 38507 |
},
|
| 38508 |
-
"total_flos": 1.
|
| 38509 |
"train_batch_size": 8,
|
| 38510 |
"trial_name": null,
|
| 38511 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9647632663358843,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 5736,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 38486 |
"learning_rate": 1.4425382207839954e-06,
|
| 38487 |
"loss": 0.0,
|
| 38488 |
"step": 5497
|
| 38489 |
+
},
|
| 38490 |
+
{
|
| 38491 |
+
"epoch": 0.9247329913379867,
|
| 38492 |
+
"grad_norm": NaN,
|
| 38493 |
+
"learning_rate": 1.4361364783600895e-06,
|
| 38494 |
+
"loss": 0.0,
|
| 38495 |
+
"step": 5498
|
| 38496 |
+
},
|
| 38497 |
+
{
|
| 38498 |
+
"epoch": 0.924901185770751,
|
| 38499 |
+
"grad_norm": NaN,
|
| 38500 |
+
"learning_rate": 1.4297487654924002e-06,
|
| 38501 |
+
"loss": 0.0,
|
| 38502 |
+
"step": 5499
|
| 38503 |
+
},
|
| 38504 |
+
{
|
| 38505 |
+
"epoch": 0.9250693802035153,
|
| 38506 |
+
"grad_norm": NaN,
|
| 38507 |
+
"learning_rate": 1.423375084026285e-06,
|
| 38508 |
+
"loss": 0.0,
|
| 38509 |
+
"step": 5500
|
| 38510 |
+
},
|
| 38511 |
+
{
|
| 38512 |
+
"epoch": 0.9252375746362795,
|
| 38513 |
+
"grad_norm": NaN,
|
| 38514 |
+
"learning_rate": 1.4170154358030151e-06,
|
| 38515 |
+
"loss": 0.0,
|
| 38516 |
+
"step": 5501
|
| 38517 |
+
},
|
| 38518 |
+
{
|
| 38519 |
+
"epoch": 0.9254057690690438,
|
| 38520 |
+
"grad_norm": NaN,
|
| 38521 |
+
"learning_rate": 1.410669822659827e-06,
|
| 38522 |
+
"loss": 0.0,
|
| 38523 |
+
"step": 5502
|
| 38524 |
+
},
|
| 38525 |
+
{
|
| 38526 |
+
"epoch": 0.9255739635018081,
|
| 38527 |
+
"grad_norm": NaN,
|
| 38528 |
+
"learning_rate": 1.404338246429887e-06,
|
| 38529 |
+
"loss": 0.0,
|
| 38530 |
+
"step": 5503
|
| 38531 |
+
},
|
| 38532 |
+
{
|
| 38533 |
+
"epoch": 0.9257421579345724,
|
| 38534 |
+
"grad_norm": NaN,
|
| 38535 |
+
"learning_rate": 1.3980207089423326e-06,
|
| 38536 |
+
"loss": 0.0,
|
| 38537 |
+
"step": 5504
|
| 38538 |
+
},
|
| 38539 |
+
{
|
| 38540 |
+
"epoch": 0.9259103523673367,
|
| 38541 |
+
"grad_norm": NaN,
|
| 38542 |
+
"learning_rate": 1.39171721202222e-06,
|
| 38543 |
+
"loss": 0.0,
|
| 38544 |
+
"step": 5505
|
| 38545 |
+
},
|
| 38546 |
+
{
|
| 38547 |
+
"epoch": 0.9260785468001009,
|
| 38548 |
+
"grad_norm": NaN,
|
| 38549 |
+
"learning_rate": 1.385427757490565e-06,
|
| 38550 |
+
"loss": 0.0,
|
| 38551 |
+
"step": 5506
|
| 38552 |
+
},
|
| 38553 |
+
{
|
| 38554 |
+
"epoch": 0.9262467412328652,
|
| 38555 |
+
"grad_norm": NaN,
|
| 38556 |
+
"learning_rate": 1.3791523471643141e-06,
|
| 38557 |
+
"loss": 0.0,
|
| 38558 |
+
"step": 5507
|
| 38559 |
+
},
|
| 38560 |
+
{
|
| 38561 |
+
"epoch": 0.9264149356656295,
|
| 38562 |
+
"grad_norm": NaN,
|
| 38563 |
+
"learning_rate": 1.3728909828563619e-06,
|
| 38564 |
+
"loss": 0.0,
|
| 38565 |
+
"step": 5508
|
| 38566 |
+
},
|
| 38567 |
+
{
|
| 38568 |
+
"epoch": 0.9265831300983938,
|
| 38569 |
+
"grad_norm": NaN,
|
| 38570 |
+
"learning_rate": 1.3666436663755555e-06,
|
| 38571 |
+
"loss": 0.0,
|
| 38572 |
+
"step": 5509
|
| 38573 |
+
},
|
| 38574 |
+
{
|
| 38575 |
+
"epoch": 0.926751324531158,
|
| 38576 |
+
"grad_norm": NaN,
|
| 38577 |
+
"learning_rate": 1.3604103995266682e-06,
|
| 38578 |
+
"loss": 0.0,
|
| 38579 |
+
"step": 5510
|
| 38580 |
+
},
|
| 38581 |
+
{
|
| 38582 |
+
"epoch": 0.9269195189639223,
|
| 38583 |
+
"grad_norm": NaN,
|
| 38584 |
+
"learning_rate": 1.3541911841104149e-06,
|
| 38585 |
+
"loss": 0.0,
|
| 38586 |
+
"step": 5511
|
| 38587 |
+
},
|
| 38588 |
+
{
|
| 38589 |
+
"epoch": 0.9270877133966866,
|
| 38590 |
+
"grad_norm": NaN,
|
| 38591 |
+
"learning_rate": 1.3479860219234697e-06,
|
| 38592 |
+
"loss": 0.0,
|
| 38593 |
+
"step": 5512
|
| 38594 |
+
},
|
| 38595 |
+
{
|
| 38596 |
+
"epoch": 0.9272559078294509,
|
| 38597 |
+
"grad_norm": NaN,
|
| 38598 |
+
"learning_rate": 1.3417949147584318e-06,
|
| 38599 |
+
"loss": 0.0,
|
| 38600 |
+
"step": 5513
|
| 38601 |
+
},
|
| 38602 |
+
{
|
| 38603 |
+
"epoch": 0.9274241022622152,
|
| 38604 |
+
"grad_norm": NaN,
|
| 38605 |
+
"learning_rate": 1.335617864403832e-06,
|
| 38606 |
+
"loss": 0.0,
|
| 38607 |
+
"step": 5514
|
| 38608 |
+
},
|
| 38609 |
+
{
|
| 38610 |
+
"epoch": 0.9275922966949794,
|
| 38611 |
+
"grad_norm": NaN,
|
| 38612 |
+
"learning_rate": 1.3294548726441592e-06,
|
| 38613 |
+
"loss": 0.0,
|
| 38614 |
+
"step": 5515
|
| 38615 |
+
},
|
| 38616 |
+
{
|
| 38617 |
+
"epoch": 0.9277604911277437,
|
| 38618 |
+
"grad_norm": NaN,
|
| 38619 |
+
"learning_rate": 1.3233059412598392e-06,
|
| 38620 |
+
"loss": 0.0,
|
| 38621 |
+
"step": 5516
|
| 38622 |
+
},
|
| 38623 |
+
{
|
| 38624 |
+
"epoch": 0.9279286855605079,
|
| 38625 |
+
"grad_norm": NaN,
|
| 38626 |
+
"learning_rate": 1.3171710720272234e-06,
|
| 38627 |
+
"loss": 0.0,
|
| 38628 |
+
"step": 5517
|
| 38629 |
+
},
|
| 38630 |
+
{
|
| 38631 |
+
"epoch": 0.9280968799932722,
|
| 38632 |
+
"grad_norm": NaN,
|
| 38633 |
+
"learning_rate": 1.3110502667185997e-06,
|
| 38634 |
+
"loss": 0.0,
|
| 38635 |
+
"step": 5518
|
| 38636 |
+
},
|
| 38637 |
+
{
|
| 38638 |
+
"epoch": 0.9282650744260365,
|
| 38639 |
+
"grad_norm": NaN,
|
| 38640 |
+
"learning_rate": 1.3049435271022037e-06,
|
| 38641 |
+
"loss": 0.0,
|
| 38642 |
+
"step": 5519
|
| 38643 |
+
},
|
| 38644 |
+
{
|
| 38645 |
+
"epoch": 0.9284332688588007,
|
| 38646 |
+
"grad_norm": NaN,
|
| 38647 |
+
"learning_rate": 1.298850854942213e-06,
|
| 38648 |
+
"loss": 0.0,
|
| 38649 |
+
"step": 5520
|
| 38650 |
+
},
|
| 38651 |
+
{
|
| 38652 |
+
"epoch": 0.928601463291565,
|
| 38653 |
+
"grad_norm": NaN,
|
| 38654 |
+
"learning_rate": 1.2927722519987306e-06,
|
| 38655 |
+
"loss": 0.0,
|
| 38656 |
+
"step": 5521
|
| 38657 |
+
},
|
| 38658 |
+
{
|
| 38659 |
+
"epoch": 0.9287696577243293,
|
| 38660 |
+
"grad_norm": NaN,
|
| 38661 |
+
"learning_rate": 1.2867077200277856e-06,
|
| 38662 |
+
"loss": 0.0,
|
| 38663 |
+
"step": 5522
|
| 38664 |
+
},
|
| 38665 |
+
{
|
| 38666 |
+
"epoch": 0.9289378521570936,
|
| 38667 |
+
"grad_norm": NaN,
|
| 38668 |
+
"learning_rate": 1.2806572607813649e-06,
|
| 38669 |
+
"loss": 0.0,
|
| 38670 |
+
"step": 5523
|
| 38671 |
+
},
|
| 38672 |
+
{
|
| 38673 |
+
"epoch": 0.9291060465898578,
|
| 38674 |
+
"grad_norm": NaN,
|
| 38675 |
+
"learning_rate": 1.2746208760073708e-06,
|
| 38676 |
+
"loss": 0.0,
|
| 38677 |
+
"step": 5524
|
| 38678 |
+
},
|
| 38679 |
+
{
|
| 38680 |
+
"epoch": 0.9292742410226221,
|
| 38681 |
+
"grad_norm": NaN,
|
| 38682 |
+
"learning_rate": 1.268598567449647e-06,
|
| 38683 |
+
"loss": 0.0,
|
| 38684 |
+
"step": 5525
|
| 38685 |
+
},
|
| 38686 |
+
{
|
| 38687 |
+
"epoch": 0.9294424354553864,
|
| 38688 |
+
"grad_norm": NaN,
|
| 38689 |
+
"learning_rate": 1.2625903368479796e-06,
|
| 38690 |
+
"loss": 0.0,
|
| 38691 |
+
"step": 5526
|
| 38692 |
+
},
|
| 38693 |
+
{
|
| 38694 |
+
"epoch": 0.9296106298881507,
|
| 38695 |
+
"grad_norm": NaN,
|
| 38696 |
+
"learning_rate": 1.2565961859380693e-06,
|
| 38697 |
+
"loss": 0.0,
|
| 38698 |
+
"step": 5527
|
| 38699 |
+
},
|
| 38700 |
+
{
|
| 38701 |
+
"epoch": 0.929778824320915,
|
| 38702 |
+
"grad_norm": NaN,
|
| 38703 |
+
"learning_rate": 1.250616116451564e-06,
|
| 38704 |
+
"loss": 0.0,
|
| 38705 |
+
"step": 5528
|
| 38706 |
+
},
|
| 38707 |
+
{
|
| 38708 |
+
"epoch": 0.9299470187536792,
|
| 38709 |
+
"grad_norm": NaN,
|
| 38710 |
+
"learning_rate": 1.2446501301160374e-06,
|
| 38711 |
+
"loss": 0.0,
|
| 38712 |
+
"step": 5529
|
| 38713 |
+
},
|
| 38714 |
+
{
|
| 38715 |
+
"epoch": 0.9301152131864435,
|
| 38716 |
+
"grad_norm": NaN,
|
| 38717 |
+
"learning_rate": 1.2386982286549998e-06,
|
| 38718 |
+
"loss": 0.0,
|
| 38719 |
+
"step": 5530
|
| 38720 |
+
},
|
| 38721 |
+
{
|
| 38722 |
+
"epoch": 0.9302834076192078,
|
| 38723 |
+
"grad_norm": NaN,
|
| 38724 |
+
"learning_rate": 1.2327604137878812e-06,
|
| 38725 |
+
"loss": 0.0,
|
| 38726 |
+
"step": 5531
|
| 38727 |
+
},
|
| 38728 |
+
{
|
| 38729 |
+
"epoch": 0.9304516020519721,
|
| 38730 |
+
"grad_norm": NaN,
|
| 38731 |
+
"learning_rate": 1.2268366872300597e-06,
|
| 38732 |
+
"loss": 0.0,
|
| 38733 |
+
"step": 5532
|
| 38734 |
+
},
|
| 38735 |
+
{
|
| 38736 |
+
"epoch": 0.9306197964847364,
|
| 38737 |
+
"grad_norm": NaN,
|
| 38738 |
+
"learning_rate": 1.2209270506928271e-06,
|
| 38739 |
+
"loss": 0.0,
|
| 38740 |
+
"step": 5533
|
| 38741 |
+
},
|
| 38742 |
+
{
|
| 38743 |
+
"epoch": 0.9307879909175006,
|
| 38744 |
+
"grad_norm": NaN,
|
| 38745 |
+
"learning_rate": 1.2150315058834184e-06,
|
| 38746 |
+
"loss": 0.0,
|
| 38747 |
+
"step": 5534
|
| 38748 |
+
},
|
| 38749 |
+
{
|
| 38750 |
+
"epoch": 0.9309561853502649,
|
| 38751 |
+
"grad_norm": NaN,
|
| 38752 |
+
"learning_rate": 1.2091500545049706e-06,
|
| 38753 |
+
"loss": 0.0,
|
| 38754 |
+
"step": 5535
|
| 38755 |
+
},
|
| 38756 |
+
{
|
| 38757 |
+
"epoch": 0.9311243797830292,
|
| 38758 |
+
"grad_norm": NaN,
|
| 38759 |
+
"learning_rate": 1.203282698256597e-06,
|
| 38760 |
+
"loss": 0.0,
|
| 38761 |
+
"step": 5536
|
| 38762 |
+
},
|
| 38763 |
+
{
|
| 38764 |
+
"epoch": 0.9312925742157935,
|
| 38765 |
+
"grad_norm": NaN,
|
| 38766 |
+
"learning_rate": 1.1974294388332918e-06,
|
| 38767 |
+
"loss": 0.0,
|
| 38768 |
+
"step": 5537
|
| 38769 |
+
},
|
| 38770 |
+
{
|
| 38771 |
+
"epoch": 0.9314607686485578,
|
| 38772 |
+
"grad_norm": NaN,
|
| 38773 |
+
"learning_rate": 1.1915902779260024e-06,
|
| 38774 |
+
"loss": 0.0,
|
| 38775 |
+
"step": 5538
|
| 38776 |
+
},
|
| 38777 |
+
{
|
| 38778 |
+
"epoch": 0.931628963081322,
|
| 38779 |
+
"grad_norm": NaN,
|
| 38780 |
+
"learning_rate": 1.1857652172215905e-06,
|
| 38781 |
+
"loss": 0.0,
|
| 38782 |
+
"step": 5539
|
| 38783 |
+
},
|
| 38784 |
+
{
|
| 38785 |
+
"epoch": 0.9317971575140863,
|
| 38786 |
+
"grad_norm": NaN,
|
| 38787 |
+
"learning_rate": 1.1799542584028656e-06,
|
| 38788 |
+
"loss": 0.0,
|
| 38789 |
+
"step": 5540
|
| 38790 |
+
},
|
| 38791 |
+
{
|
| 38792 |
+
"epoch": 0.9319653519468506,
|
| 38793 |
+
"grad_norm": NaN,
|
| 38794 |
+
"learning_rate": 1.1741574031485347e-06,
|
| 38795 |
+
"loss": 0.0,
|
| 38796 |
+
"step": 5541
|
| 38797 |
+
},
|
| 38798 |
+
{
|
| 38799 |
+
"epoch": 0.9321335463796149,
|
| 38800 |
+
"grad_norm": NaN,
|
| 38801 |
+
"learning_rate": 1.1683746531332529e-06,
|
| 38802 |
+
"loss": 0.0,
|
| 38803 |
+
"step": 5542
|
| 38804 |
+
},
|
| 38805 |
+
{
|
| 38806 |
+
"epoch": 0.9323017408123792,
|
| 38807 |
+
"grad_norm": NaN,
|
| 38808 |
+
"learning_rate": 1.1626060100275837e-06,
|
| 38809 |
+
"loss": 0.0,
|
| 38810 |
+
"step": 5543
|
| 38811 |
+
},
|
| 38812 |
+
{
|
| 38813 |
+
"epoch": 0.9324699352451434,
|
| 38814 |
+
"grad_norm": NaN,
|
| 38815 |
+
"learning_rate": 1.156851475498033e-06,
|
| 38816 |
+
"loss": 0.0,
|
| 38817 |
+
"step": 5544
|
| 38818 |
+
},
|
| 38819 |
+
{
|
| 38820 |
+
"epoch": 0.9326381296779077,
|
| 38821 |
+
"grad_norm": NaN,
|
| 38822 |
+
"learning_rate": 1.1511110512070155e-06,
|
| 38823 |
+
"loss": 0.0,
|
| 38824 |
+
"step": 5545
|
| 38825 |
+
},
|
| 38826 |
+
{
|
| 38827 |
+
"epoch": 0.932806324110672,
|
| 38828 |
+
"grad_norm": NaN,
|
| 38829 |
+
"learning_rate": 1.1453847388128712e-06,
|
| 38830 |
+
"loss": 0.0,
|
| 38831 |
+
"step": 5546
|
| 38832 |
+
},
|
| 38833 |
+
{
|
| 38834 |
+
"epoch": 0.9329745185434362,
|
| 38835 |
+
"grad_norm": NaN,
|
| 38836 |
+
"learning_rate": 1.1396725399698772e-06,
|
| 38837 |
+
"loss": 0.0,
|
| 38838 |
+
"step": 5547
|
| 38839 |
+
},
|
| 38840 |
+
{
|
| 38841 |
+
"epoch": 0.9331427129762004,
|
| 38842 |
+
"grad_norm": NaN,
|
| 38843 |
+
"learning_rate": 1.133974456328213e-06,
|
| 38844 |
+
"loss": 0.0,
|
| 38845 |
+
"step": 5548
|
| 38846 |
+
},
|
| 38847 |
+
{
|
| 38848 |
+
"epoch": 0.9333109074089647,
|
| 38849 |
+
"grad_norm": NaN,
|
| 38850 |
+
"learning_rate": 1.1282904895340064e-06,
|
| 38851 |
+
"loss": 0.0,
|
| 38852 |
+
"step": 5549
|
| 38853 |
+
},
|
| 38854 |
+
{
|
| 38855 |
+
"epoch": 0.933479101841729,
|
| 38856 |
+
"grad_norm": NaN,
|
| 38857 |
+
"learning_rate": 1.1226206412292773e-06,
|
| 38858 |
+
"loss": 0.0,
|
| 38859 |
+
"step": 5550
|
| 38860 |
+
},
|
| 38861 |
+
{
|
| 38862 |
+
"epoch": 0.9336472962744933,
|
| 38863 |
+
"grad_norm": NaN,
|
| 38864 |
+
"learning_rate": 1.1169649130519932e-06,
|
| 38865 |
+
"loss": 0.0,
|
| 38866 |
+
"step": 5551
|
| 38867 |
+
},
|
| 38868 |
+
{
|
| 38869 |
+
"epoch": 0.9338154907072576,
|
| 38870 |
+
"grad_norm": NaN,
|
| 38871 |
+
"learning_rate": 1.111323306636014e-06,
|
| 38872 |
+
"loss": 0.0,
|
| 38873 |
+
"step": 5552
|
| 38874 |
+
},
|
| 38875 |
+
{
|
| 38876 |
+
"epoch": 0.9339836851400218,
|
| 38877 |
+
"grad_norm": NaN,
|
| 38878 |
+
"learning_rate": 1.1056958236111525e-06,
|
| 38879 |
+
"loss": 0.0,
|
| 38880 |
+
"step": 5553
|
| 38881 |
+
},
|
| 38882 |
+
{
|
| 38883 |
+
"epoch": 0.9341518795727861,
|
| 38884 |
+
"grad_norm": NaN,
|
| 38885 |
+
"learning_rate": 1.1000824656031195e-06,
|
| 38886 |
+
"loss": 0.0,
|
| 38887 |
+
"step": 5554
|
| 38888 |
+
},
|
| 38889 |
+
{
|
| 38890 |
+
"epoch": 0.9343200740055504,
|
| 38891 |
+
"grad_norm": NaN,
|
| 38892 |
+
"learning_rate": 1.0944832342335564e-06,
|
| 38893 |
+
"loss": 0.0,
|
| 38894 |
+
"step": 5555
|
| 38895 |
+
},
|
| 38896 |
+
{
|
| 38897 |
+
"epoch": 0.9344882684383147,
|
| 38898 |
+
"grad_norm": NaN,
|
| 38899 |
+
"learning_rate": 1.0888981311200031e-06,
|
| 38900 |
+
"loss": 0.0,
|
| 38901 |
+
"step": 5556
|
| 38902 |
+
},
|
| 38903 |
+
{
|
| 38904 |
+
"epoch": 0.934656462871079,
|
| 38905 |
+
"grad_norm": NaN,
|
| 38906 |
+
"learning_rate": 1.083327157875952e-06,
|
| 38907 |
+
"loss": 0.0,
|
| 38908 |
+
"step": 5557
|
| 38909 |
+
},
|
| 38910 |
+
{
|
| 38911 |
+
"epoch": 0.9348246573038432,
|
| 38912 |
+
"grad_norm": NaN,
|
| 38913 |
+
"learning_rate": 1.0777703161107877e-06,
|
| 38914 |
+
"loss": 0.0,
|
| 38915 |
+
"step": 5558
|
| 38916 |
+
},
|
| 38917 |
+
{
|
| 38918 |
+
"epoch": 0.9349928517366075,
|
| 38919 |
+
"grad_norm": NaN,
|
| 38920 |
+
"learning_rate": 1.0722276074298154e-06,
|
| 38921 |
+
"loss": 0.0,
|
| 38922 |
+
"step": 5559
|
| 38923 |
+
},
|
| 38924 |
+
{
|
| 38925 |
+
"epoch": 0.9351610461693718,
|
| 38926 |
+
"grad_norm": NaN,
|
| 38927 |
+
"learning_rate": 1.0666990334342707e-06,
|
| 38928 |
+
"loss": 0.0,
|
| 38929 |
+
"step": 5560
|
| 38930 |
+
},
|
| 38931 |
+
{
|
| 38932 |
+
"epoch": 0.9353292406021361,
|
| 38933 |
+
"grad_norm": NaN,
|
| 38934 |
+
"learning_rate": 1.0611845957212873e-06,
|
| 38935 |
+
"loss": 0.0,
|
| 38936 |
+
"step": 5561
|
| 38937 |
+
},
|
| 38938 |
+
{
|
| 38939 |
+
"epoch": 0.9354974350349003,
|
| 38940 |
+
"grad_norm": NaN,
|
| 38941 |
+
"learning_rate": 1.0556842958839242e-06,
|
| 38942 |
+
"loss": 0.0,
|
| 38943 |
+
"step": 5562
|
| 38944 |
+
},
|
| 38945 |
+
{
|
| 38946 |
+
"epoch": 0.9356656294676646,
|
| 38947 |
+
"grad_norm": NaN,
|
| 38948 |
+
"learning_rate": 1.0501981355111656e-06,
|
| 38949 |
+
"loss": 0.0,
|
| 38950 |
+
"step": 5563
|
| 38951 |
+
},
|
| 38952 |
+
{
|
| 38953 |
+
"epoch": 0.9358338239004289,
|
| 38954 |
+
"grad_norm": NaN,
|
| 38955 |
+
"learning_rate": 1.0447261161878884e-06,
|
| 38956 |
+
"loss": 0.0,
|
| 38957 |
+
"step": 5564
|
| 38958 |
+
},
|
| 38959 |
+
{
|
| 38960 |
+
"epoch": 0.9360020183331932,
|
| 38961 |
+
"grad_norm": NaN,
|
| 38962 |
+
"learning_rate": 1.0392682394949116e-06,
|
| 38963 |
+
"loss": 0.0,
|
| 38964 |
+
"step": 5565
|
| 38965 |
+
},
|
| 38966 |
+
{
|
| 38967 |
+
"epoch": 0.9361702127659575,
|
| 38968 |
+
"grad_norm": NaN,
|
| 38969 |
+
"learning_rate": 1.0338245070089515e-06,
|
| 38970 |
+
"loss": 0.0,
|
| 38971 |
+
"step": 5566
|
| 38972 |
+
},
|
| 38973 |
+
{
|
| 38974 |
+
"epoch": 0.9363384071987217,
|
| 38975 |
+
"grad_norm": NaN,
|
| 38976 |
+
"learning_rate": 1.0283949203026332e-06,
|
| 38977 |
+
"loss": 0.0,
|
| 38978 |
+
"step": 5567
|
| 38979 |
+
},
|
| 38980 |
+
{
|
| 38981 |
+
"epoch": 0.936506601631486,
|
| 38982 |
+
"grad_norm": NaN,
|
| 38983 |
+
"learning_rate": 1.0229794809445081e-06,
|
| 38984 |
+
"loss": 0.0,
|
| 38985 |
+
"step": 5568
|
| 38986 |
+
},
|
| 38987 |
+
{
|
| 38988 |
+
"epoch": 0.9366747960642503,
|
| 38989 |
+
"grad_norm": NaN,
|
| 38990 |
+
"learning_rate": 1.0175781904990412e-06,
|
| 38991 |
+
"loss": 0.0,
|
| 38992 |
+
"step": 5569
|
| 38993 |
+
},
|
| 38994 |
+
{
|
| 38995 |
+
"epoch": 0.9368429904970146,
|
| 38996 |
+
"grad_norm": NaN,
|
| 38997 |
+
"learning_rate": 1.0121910505266008e-06,
|
| 38998 |
+
"loss": 0.0,
|
| 38999 |
+
"step": 5570
|
| 39000 |
+
},
|
| 39001 |
+
{
|
| 39002 |
+
"epoch": 0.9370111849297789,
|
| 39003 |
+
"grad_norm": NaN,
|
| 39004 |
+
"learning_rate": 1.00681806258347e-06,
|
| 39005 |
+
"loss": 0.0,
|
| 39006 |
+
"step": 5571
|
| 39007 |
+
},
|
| 39008 |
+
{
|
| 39009 |
+
"epoch": 0.9371793793625431,
|
| 39010 |
+
"grad_norm": NaN,
|
| 39011 |
+
"learning_rate": 1.0014592282218404e-06,
|
| 39012 |
+
"loss": 0.0,
|
| 39013 |
+
"step": 5572
|
| 39014 |
+
},
|
| 39015 |
+
{
|
| 39016 |
+
"epoch": 0.9373475737953074,
|
| 39017 |
+
"grad_norm": NaN,
|
| 39018 |
+
"learning_rate": 9.96114548989835e-07,
|
| 39019 |
+
"loss": 0.0,
|
| 39020 |
+
"step": 5573
|
| 39021 |
+
},
|
| 39022 |
+
{
|
| 39023 |
+
"epoch": 0.9375157682280717,
|
| 39024 |
+
"grad_norm": NaN,
|
| 39025 |
+
"learning_rate": 9.907840264314572e-07,
|
| 39026 |
+
"loss": 0.0,
|
| 39027 |
+
"step": 5574
|
| 39028 |
+
},
|
| 39029 |
+
{
|
| 39030 |
+
"epoch": 0.937683962660836,
|
| 39031 |
+
"grad_norm": NaN,
|
| 39032 |
+
"learning_rate": 9.854676620866366e-07,
|
| 39033 |
+
"loss": 0.0,
|
| 39034 |
+
"step": 5575
|
| 39035 |
+
},
|
| 39036 |
+
{
|
| 39037 |
+
"epoch": 0.9378521570936001,
|
| 39038 |
+
"grad_norm": NaN,
|
| 39039 |
+
"learning_rate": 9.80165457491211e-07,
|
| 39040 |
+
"loss": 0.0,
|
| 39041 |
+
"step": 5576
|
| 39042 |
+
},
|
| 39043 |
+
{
|
| 39044 |
+
"epoch": 0.9380203515263644,
|
| 39045 |
+
"grad_norm": NaN,
|
| 39046 |
+
"learning_rate": 9.74877414176939e-07,
|
| 39047 |
+
"loss": 0.0,
|
| 39048 |
+
"step": 5577
|
| 39049 |
+
},
|
| 39050 |
+
{
|
| 39051 |
+
"epoch": 0.9381885459591287,
|
| 39052 |
+
"grad_norm": NaN,
|
| 39053 |
+
"learning_rate": 9.696035336714648e-07,
|
| 39054 |
+
"loss": 0.0,
|
| 39055 |
+
"step": 5578
|
| 39056 |
+
},
|
| 39057 |
+
{
|
| 39058 |
+
"epoch": 0.938356740391893,
|
| 39059 |
+
"grad_norm": NaN,
|
| 39060 |
+
"learning_rate": 9.643438174983589e-07,
|
| 39061 |
+
"loss": 0.0,
|
| 39062 |
+
"step": 5579
|
| 39063 |
+
},
|
| 39064 |
+
{
|
| 39065 |
+
"epoch": 0.9385249348246573,
|
| 39066 |
+
"grad_norm": NaN,
|
| 39067 |
+
"learning_rate": 9.590982671770943e-07,
|
| 39068 |
+
"loss": 0.0,
|
| 39069 |
+
"step": 5580
|
| 39070 |
+
},
|
| 39071 |
+
{
|
| 39072 |
+
"epoch": 0.9386931292574215,
|
| 39073 |
+
"grad_norm": NaN,
|
| 39074 |
+
"learning_rate": 9.538668842230537e-07,
|
| 39075 |
+
"loss": 0.0,
|
| 39076 |
+
"step": 5581
|
| 39077 |
+
},
|
| 39078 |
+
{
|
| 39079 |
+
"epoch": 0.9388613236901858,
|
| 39080 |
+
"grad_norm": NaN,
|
| 39081 |
+
"learning_rate": 9.486496701475167e-07,
|
| 39082 |
+
"loss": 0.0,
|
| 39083 |
+
"step": 5582
|
| 39084 |
+
},
|
| 39085 |
+
{
|
| 39086 |
+
"epoch": 0.9390295181229501,
|
| 39087 |
+
"grad_norm": NaN,
|
| 39088 |
+
"learning_rate": 9.434466264576892e-07,
|
| 39089 |
+
"loss": 0.0,
|
| 39090 |
+
"step": 5583
|
| 39091 |
+
},
|
| 39092 |
+
{
|
| 39093 |
+
"epoch": 0.9391977125557144,
|
| 39094 |
+
"grad_norm": NaN,
|
| 39095 |
+
"learning_rate": 9.382577546566574e-07,
|
| 39096 |
+
"loss": 0.0,
|
| 39097 |
+
"step": 5584
|
| 39098 |
+
},
|
| 39099 |
+
{
|
| 39100 |
+
"epoch": 0.9393659069884787,
|
| 39101 |
+
"grad_norm": NaN,
|
| 39102 |
+
"learning_rate": 9.330830562434445e-07,
|
| 39103 |
+
"loss": 0.0,
|
| 39104 |
+
"step": 5585
|
| 39105 |
+
},
|
| 39106 |
+
{
|
| 39107 |
+
"epoch": 0.9395341014212429,
|
| 39108 |
+
"grad_norm": NaN,
|
| 39109 |
+
"learning_rate": 9.279225327129548e-07,
|
| 39110 |
+
"loss": 0.0,
|
| 39111 |
+
"step": 5586
|
| 39112 |
+
},
|
| 39113 |
+
{
|
| 39114 |
+
"epoch": 0.9397022958540072,
|
| 39115 |
+
"grad_norm": NaN,
|
| 39116 |
+
"learning_rate": 9.227761855560069e-07,
|
| 39117 |
+
"loss": 0.0,
|
| 39118 |
+
"step": 5587
|
| 39119 |
+
},
|
| 39120 |
+
{
|
| 39121 |
+
"epoch": 0.9398704902867715,
|
| 39122 |
+
"grad_norm": NaN,
|
| 39123 |
+
"learning_rate": 9.176440162593169e-07,
|
| 39124 |
+
"loss": 0.0,
|
| 39125 |
+
"step": 5588
|
| 39126 |
+
},
|
| 39127 |
+
{
|
| 39128 |
+
"epoch": 0.9400386847195358,
|
| 39129 |
+
"grad_norm": NaN,
|
| 39130 |
+
"learning_rate": 9.125260263055213e-07,
|
| 39131 |
+
"loss": 0.0,
|
| 39132 |
+
"step": 5589
|
| 39133 |
+
},
|
| 39134 |
+
{
|
| 39135 |
+
"epoch": 0.9402068791523001,
|
| 39136 |
+
"grad_norm": NaN,
|
| 39137 |
+
"learning_rate": 9.074222171731427e-07,
|
| 39138 |
+
"loss": 0.0,
|
| 39139 |
+
"step": 5590
|
| 39140 |
+
},
|
| 39141 |
+
{
|
| 39142 |
+
"epoch": 0.9403750735850643,
|
| 39143 |
+
"grad_norm": NaN,
|
| 39144 |
+
"learning_rate": 9.023325903366242e-07,
|
| 39145 |
+
"loss": 0.0,
|
| 39146 |
+
"step": 5591
|
| 39147 |
+
},
|
| 39148 |
+
{
|
| 39149 |
+
"epoch": 0.9405432680178286,
|
| 39150 |
+
"grad_norm": NaN,
|
| 39151 |
+
"learning_rate": 8.972571472662838e-07,
|
| 39152 |
+
"loss": 0.0,
|
| 39153 |
+
"step": 5592
|
| 39154 |
+
},
|
| 39155 |
+
{
|
| 39156 |
+
"epoch": 0.9407114624505929,
|
| 39157 |
+
"grad_norm": NaN,
|
| 39158 |
+
"learning_rate": 8.921958894283767e-07,
|
| 39159 |
+
"loss": 0.0,
|
| 39160 |
+
"step": 5593
|
| 39161 |
+
},
|
| 39162 |
+
{
|
| 39163 |
+
"epoch": 0.9408796568833572,
|
| 39164 |
+
"grad_norm": NaN,
|
| 39165 |
+
"learning_rate": 8.871488182850441e-07,
|
| 39166 |
+
"loss": 0.0,
|
| 39167 |
+
"step": 5594
|
| 39168 |
+
},
|
| 39169 |
+
{
|
| 39170 |
+
"epoch": 0.9410478513161215,
|
| 39171 |
+
"grad_norm": NaN,
|
| 39172 |
+
"learning_rate": 8.821159352943143e-07,
|
| 39173 |
+
"loss": 0.0,
|
| 39174 |
+
"step": 5595
|
| 39175 |
+
},
|
| 39176 |
+
{
|
| 39177 |
+
"epoch": 0.9412160457488857,
|
| 39178 |
+
"grad_norm": NaN,
|
| 39179 |
+
"learning_rate": 8.770972419101464e-07,
|
| 39180 |
+
"loss": 0.0,
|
| 39181 |
+
"step": 5596
|
| 39182 |
+
},
|
| 39183 |
+
{
|
| 39184 |
+
"epoch": 0.94138424018165,
|
| 39185 |
+
"grad_norm": NaN,
|
| 39186 |
+
"learning_rate": 8.720927395823697e-07,
|
| 39187 |
+
"loss": 0.0,
|
| 39188 |
+
"step": 5597
|
| 39189 |
+
},
|
| 39190 |
+
{
|
| 39191 |
+
"epoch": 0.9415524346144143,
|
| 39192 |
+
"grad_norm": NaN,
|
| 39193 |
+
"learning_rate": 8.671024297567388e-07,
|
| 39194 |
+
"loss": 0.0,
|
| 39195 |
+
"step": 5598
|
| 39196 |
+
},
|
| 39197 |
+
{
|
| 39198 |
+
"epoch": 0.9417206290471786,
|
| 39199 |
+
"grad_norm": NaN,
|
| 39200 |
+
"learning_rate": 8.621263138749002e-07,
|
| 39201 |
+
"loss": 0.0,
|
| 39202 |
+
"step": 5599
|
| 39203 |
+
},
|
| 39204 |
+
{
|
| 39205 |
+
"epoch": 0.9418888234799428,
|
| 39206 |
+
"grad_norm": NaN,
|
| 39207 |
+
"learning_rate": 8.571643933743879e-07,
|
| 39208 |
+
"loss": 0.0,
|
| 39209 |
+
"step": 5600
|
| 39210 |
+
},
|
| 39211 |
+
{
|
| 39212 |
+
"epoch": 0.9420570179127071,
|
| 39213 |
+
"grad_norm": NaN,
|
| 39214 |
+
"learning_rate": 8.522166696886547e-07,
|
| 39215 |
+
"loss": 0.0,
|
| 39216 |
+
"step": 5601
|
| 39217 |
+
},
|
| 39218 |
+
{
|
| 39219 |
+
"epoch": 0.9422252123454714,
|
| 39220 |
+
"grad_norm": NaN,
|
| 39221 |
+
"learning_rate": 8.472831442470408e-07,
|
| 39222 |
+
"loss": 0.0,
|
| 39223 |
+
"step": 5602
|
| 39224 |
+
},
|
| 39225 |
+
{
|
| 39226 |
+
"epoch": 0.9423934067782357,
|
| 39227 |
+
"grad_norm": NaN,
|
| 39228 |
+
"learning_rate": 8.423638184747784e-07,
|
| 39229 |
+
"loss": 0.0,
|
| 39230 |
+
"step": 5603
|
| 39231 |
+
},
|
| 39232 |
+
{
|
| 39233 |
+
"epoch": 0.942561601211,
|
| 39234 |
+
"grad_norm": NaN,
|
| 39235 |
+
"learning_rate": 8.374586937930196e-07,
|
| 39236 |
+
"loss": 0.0,
|
| 39237 |
+
"step": 5604
|
| 39238 |
+
},
|
| 39239 |
+
{
|
| 39240 |
+
"epoch": 0.9427297956437642,
|
| 39241 |
+
"grad_norm": NaN,
|
| 39242 |
+
"learning_rate": 8.325677716187807e-07,
|
| 39243 |
+
"loss": 0.0,
|
| 39244 |
+
"step": 5605
|
| 39245 |
+
},
|
| 39246 |
+
{
|
| 39247 |
+
"epoch": 0.9428979900765284,
|
| 39248 |
+
"grad_norm": NaN,
|
| 39249 |
+
"learning_rate": 8.276910533650151e-07,
|
| 39250 |
+
"loss": 0.0,
|
| 39251 |
+
"step": 5606
|
| 39252 |
+
},
|
| 39253 |
+
{
|
| 39254 |
+
"epoch": 0.9430661845092927,
|
| 39255 |
+
"grad_norm": NaN,
|
| 39256 |
+
"learning_rate": 8.2282854044054e-07,
|
| 39257 |
+
"loss": 0.0,
|
| 39258 |
+
"step": 5607
|
| 39259 |
+
},
|
| 39260 |
+
{
|
| 39261 |
+
"epoch": 0.943234378942057,
|
| 39262 |
+
"grad_norm": NaN,
|
| 39263 |
+
"learning_rate": 8.179802342500876e-07,
|
| 39264 |
+
"loss": 0.0,
|
| 39265 |
+
"step": 5608
|
| 39266 |
+
},
|
| 39267 |
+
{
|
| 39268 |
+
"epoch": 0.9434025733748213,
|
| 39269 |
+
"grad_norm": NaN,
|
| 39270 |
+
"learning_rate": 8.13146136194265e-07,
|
| 39271 |
+
"loss": 0.0,
|
| 39272 |
+
"step": 5609
|
| 39273 |
+
},
|
| 39274 |
+
{
|
| 39275 |
+
"epoch": 0.9435707678075855,
|
| 39276 |
+
"grad_norm": NaN,
|
| 39277 |
+
"learning_rate": 8.083262476696051e-07,
|
| 39278 |
+
"loss": 0.0,
|
| 39279 |
+
"step": 5610
|
| 39280 |
+
},
|
| 39281 |
+
{
|
| 39282 |
+
"epoch": 0.9437389622403498,
|
| 39283 |
+
"grad_norm": NaN,
|
| 39284 |
+
"learning_rate": 8.035205700685167e-07,
|
| 39285 |
+
"loss": 0.0,
|
| 39286 |
+
"step": 5611
|
| 39287 |
+
},
|
| 39288 |
+
{
|
| 39289 |
+
"epoch": 0.9439071566731141,
|
| 39290 |
+
"grad_norm": NaN,
|
| 39291 |
+
"learning_rate": 7.987291047793056e-07,
|
| 39292 |
+
"loss": 0.0,
|
| 39293 |
+
"step": 5612
|
| 39294 |
+
},
|
| 39295 |
+
{
|
| 39296 |
+
"epoch": 0.9440753511058784,
|
| 39297 |
+
"grad_norm": NaN,
|
| 39298 |
+
"learning_rate": 7.93951853186159e-07,
|
| 39299 |
+
"loss": 0.0,
|
| 39300 |
+
"step": 5613
|
| 39301 |
+
},
|
| 39302 |
+
{
|
| 39303 |
+
"epoch": 0.9442435455386426,
|
| 39304 |
+
"grad_norm": NaN,
|
| 39305 |
+
"learning_rate": 7.891888166691952e-07,
|
| 39306 |
+
"loss": 0.0,
|
| 39307 |
+
"step": 5614
|
| 39308 |
+
},
|
| 39309 |
+
{
|
| 39310 |
+
"epoch": 0.9444117399714069,
|
| 39311 |
+
"grad_norm": NaN,
|
| 39312 |
+
"learning_rate": 7.844399966043802e-07,
|
| 39313 |
+
"loss": 0.0,
|
| 39314 |
+
"step": 5615
|
| 39315 |
+
},
|
| 39316 |
+
{
|
| 39317 |
+
"epoch": 0.9445799344041712,
|
| 39318 |
+
"grad_norm": NaN,
|
| 39319 |
+
"learning_rate": 7.797053943636112e-07,
|
| 39320 |
+
"loss": 0.0,
|
| 39321 |
+
"step": 5616
|
| 39322 |
+
},
|
| 39323 |
+
{
|
| 39324 |
+
"epoch": 0.9447481288369355,
|
| 39325 |
+
"grad_norm": NaN,
|
| 39326 |
+
"learning_rate": 7.749850113146551e-07,
|
| 39327 |
+
"loss": 0.0,
|
| 39328 |
+
"step": 5617
|
| 39329 |
+
},
|
| 39330 |
+
{
|
| 39331 |
+
"epoch": 0.9449163232696998,
|
| 39332 |
+
"grad_norm": NaN,
|
| 39333 |
+
"learning_rate": 7.70278848821171e-07,
|
| 39334 |
+
"loss": 0.0,
|
| 39335 |
+
"step": 5618
|
| 39336 |
+
},
|
| 39337 |
+
{
|
| 39338 |
+
"epoch": 0.945084517702464,
|
| 39339 |
+
"grad_norm": NaN,
|
| 39340 |
+
"learning_rate": 7.655869082427269e-07,
|
| 39341 |
+
"loss": 0.0,
|
| 39342 |
+
"step": 5619
|
| 39343 |
+
},
|
| 39344 |
+
{
|
| 39345 |
+
"epoch": 0.9452527121352283,
|
| 39346 |
+
"grad_norm": NaN,
|
| 39347 |
+
"learning_rate": 7.609091909347721e-07,
|
| 39348 |
+
"loss": 0.0,
|
| 39349 |
+
"step": 5620
|
| 39350 |
+
},
|
| 39351 |
+
{
|
| 39352 |
+
"epoch": 0.9454209065679926,
|
| 39353 |
+
"grad_norm": NaN,
|
| 39354 |
+
"learning_rate": 7.56245698248631e-07,
|
| 39355 |
+
"loss": 0.0,
|
| 39356 |
+
"step": 5621
|
| 39357 |
+
},
|
| 39358 |
+
{
|
| 39359 |
+
"epoch": 0.9455891010007569,
|
| 39360 |
+
"grad_norm": NaN,
|
| 39361 |
+
"learning_rate": 7.515964315315538e-07,
|
| 39362 |
+
"loss": 0.0,
|
| 39363 |
+
"step": 5622
|
| 39364 |
+
},
|
| 39365 |
+
{
|
| 39366 |
+
"epoch": 0.9457572954335212,
|
| 39367 |
+
"grad_norm": NaN,
|
| 39368 |
+
"learning_rate": 7.46961392126655e-07,
|
| 39369 |
+
"loss": 0.0,
|
| 39370 |
+
"step": 5623
|
| 39371 |
+
},
|
| 39372 |
+
{
|
| 39373 |
+
"epoch": 0.9459254898662854,
|
| 39374 |
+
"grad_norm": NaN,
|
| 39375 |
+
"learning_rate": 7.423405813729467e-07,
|
| 39376 |
+
"loss": 0.0,
|
| 39377 |
+
"step": 5624
|
| 39378 |
+
},
|
| 39379 |
+
{
|
| 39380 |
+
"epoch": 0.9460936842990497,
|
| 39381 |
+
"grad_norm": NaN,
|
| 39382 |
+
"learning_rate": 7.377340006053169e-07,
|
| 39383 |
+
"loss": 0.0,
|
| 39384 |
+
"step": 5625
|
| 39385 |
+
},
|
| 39386 |
+
{
|
| 39387 |
+
"epoch": 0.946261878731814,
|
| 39388 |
+
"grad_norm": NaN,
|
| 39389 |
+
"learning_rate": 7.33141651154573e-07,
|
| 39390 |
+
"loss": 0.0,
|
| 39391 |
+
"step": 5626
|
| 39392 |
+
},
|
| 39393 |
+
{
|
| 39394 |
+
"epoch": 0.9464300731645783,
|
| 39395 |
+
"grad_norm": NaN,
|
| 39396 |
+
"learning_rate": 7.285635343473818e-07,
|
| 39397 |
+
"loss": 0.0,
|
| 39398 |
+
"step": 5627
|
| 39399 |
+
},
|
| 39400 |
+
{
|
| 39401 |
+
"epoch": 0.9465982675973426,
|
| 39402 |
+
"grad_norm": NaN,
|
| 39403 |
+
"learning_rate": 7.239996515063186e-07,
|
| 39404 |
+
"loss": 0.0,
|
| 39405 |
+
"step": 5628
|
| 39406 |
+
},
|
| 39407 |
+
{
|
| 39408 |
+
"epoch": 0.9467664620301068,
|
| 39409 |
+
"grad_norm": NaN,
|
| 39410 |
+
"learning_rate": 7.194500039498286e-07,
|
| 39411 |
+
"loss": 0.0,
|
| 39412 |
+
"step": 5629
|
| 39413 |
+
},
|
| 39414 |
+
{
|
| 39415 |
+
"epoch": 0.9469346564628711,
|
| 39416 |
+
"grad_norm": NaN,
|
| 39417 |
+
"learning_rate": 7.149145929922607e-07,
|
| 39418 |
+
"loss": 0.0,
|
| 39419 |
+
"step": 5630
|
| 39420 |
+
},
|
| 39421 |
+
{
|
| 39422 |
+
"epoch": 0.9471028508956354,
|
| 39423 |
+
"grad_norm": NaN,
|
| 39424 |
+
"learning_rate": 7.103934199438444e-07,
|
| 39425 |
+
"loss": 0.0,
|
| 39426 |
+
"step": 5631
|
| 39427 |
+
},
|
| 39428 |
+
{
|
| 39429 |
+
"epoch": 0.9472710453283997,
|
| 39430 |
+
"grad_norm": NaN,
|
| 39431 |
+
"learning_rate": 7.058864861106907e-07,
|
| 39432 |
+
"loss": 0.0,
|
| 39433 |
+
"step": 5632
|
| 39434 |
+
},
|
| 39435 |
+
{
|
| 39436 |
+
"epoch": 0.947439239761164,
|
| 39437 |
+
"grad_norm": NaN,
|
| 39438 |
+
"learning_rate": 7.013937927948022e-07,
|
| 39439 |
+
"loss": 0.0,
|
| 39440 |
+
"step": 5633
|
| 39441 |
+
},
|
| 39442 |
+
{
|
| 39443 |
+
"epoch": 0.9476074341939282,
|
| 39444 |
+
"grad_norm": NaN,
|
| 39445 |
+
"learning_rate": 6.969153412940743e-07,
|
| 39446 |
+
"loss": 0.0,
|
| 39447 |
+
"step": 5634
|
| 39448 |
+
},
|
| 39449 |
+
{
|
| 39450 |
+
"epoch": 0.9477756286266924,
|
| 39451 |
+
"grad_norm": NaN,
|
| 39452 |
+
"learning_rate": 6.924511329022831e-07,
|
| 39453 |
+
"loss": 0.0,
|
| 39454 |
+
"step": 5635
|
| 39455 |
+
},
|
| 39456 |
+
{
|
| 39457 |
+
"epoch": 0.9479438230594567,
|
| 39458 |
+
"grad_norm": NaN,
|
| 39459 |
+
"learning_rate": 6.880011689090804e-07,
|
| 39460 |
+
"loss": 0.0,
|
| 39461 |
+
"step": 5636
|
| 39462 |
+
},
|
| 39463 |
+
{
|
| 39464 |
+
"epoch": 0.948112017492221,
|
| 39465 |
+
"grad_norm": NaN,
|
| 39466 |
+
"learning_rate": 6.8356545060001e-07,
|
| 39467 |
+
"loss": 0.0,
|
| 39468 |
+
"step": 5637
|
| 39469 |
+
},
|
| 39470 |
+
{
|
| 39471 |
+
"epoch": 0.9482802119249852,
|
| 39472 |
+
"grad_norm": NaN,
|
| 39473 |
+
"learning_rate": 6.79143979256508e-07,
|
| 39474 |
+
"loss": 0.0,
|
| 39475 |
+
"step": 5638
|
| 39476 |
+
},
|
| 39477 |
+
{
|
| 39478 |
+
"epoch": 0.9484484063577495,
|
| 39479 |
+
"grad_norm": NaN,
|
| 39480 |
+
"learning_rate": 6.747367561558859e-07,
|
| 39481 |
+
"loss": 0.0,
|
| 39482 |
+
"step": 5639
|
| 39483 |
+
},
|
| 39484 |
+
{
|
| 39485 |
+
"epoch": 0.9486166007905138,
|
| 39486 |
+
"grad_norm": NaN,
|
| 39487 |
+
"learning_rate": 6.703437825713421e-07,
|
| 39488 |
+
"loss": 0.0,
|
| 39489 |
+
"step": 5640
|
| 39490 |
+
},
|
| 39491 |
+
{
|
| 39492 |
+
"epoch": 0.9487847952232781,
|
| 39493 |
+
"grad_norm": NaN,
|
| 39494 |
+
"learning_rate": 6.659650597719502e-07,
|
| 39495 |
+
"loss": 0.0,
|
| 39496 |
+
"step": 5641
|
| 39497 |
+
},
|
| 39498 |
+
{
|
| 39499 |
+
"epoch": 0.9489529896560424,
|
| 39500 |
+
"grad_norm": NaN,
|
| 39501 |
+
"learning_rate": 6.616005890226817e-07,
|
| 39502 |
+
"loss": 0.0,
|
| 39503 |
+
"step": 5642
|
| 39504 |
+
},
|
| 39505 |
+
{
|
| 39506 |
+
"epoch": 0.9491211840888066,
|
| 39507 |
+
"grad_norm": NaN,
|
| 39508 |
+
"learning_rate": 6.572503715843836e-07,
|
| 39509 |
+
"loss": 0.0,
|
| 39510 |
+
"step": 5643
|
| 39511 |
+
},
|
| 39512 |
+
{
|
| 39513 |
+
"epoch": 0.9492893785215709,
|
| 39514 |
+
"grad_norm": NaN,
|
| 39515 |
+
"learning_rate": 6.52914408713784e-07,
|
| 39516 |
+
"loss": 0.0,
|
| 39517 |
+
"step": 5644
|
| 39518 |
+
},
|
| 39519 |
+
{
|
| 39520 |
+
"epoch": 0.9494575729543352,
|
| 39521 |
+
"grad_norm": NaN,
|
| 39522 |
+
"learning_rate": 6.485927016634863e-07,
|
| 39523 |
+
"loss": 0.0,
|
| 39524 |
+
"step": 5645
|
| 39525 |
+
},
|
| 39526 |
+
{
|
| 39527 |
+
"epoch": 0.9496257673870995,
|
| 39528 |
+
"grad_norm": NaN,
|
| 39529 |
+
"learning_rate": 6.44285251681992e-07,
|
| 39530 |
+
"loss": 0.0,
|
| 39531 |
+
"step": 5646
|
| 39532 |
+
},
|
| 39533 |
+
{
|
| 39534 |
+
"epoch": 0.9497939618198638,
|
| 39535 |
+
"grad_norm": NaN,
|
| 39536 |
+
"learning_rate": 6.399920600136722e-07,
|
| 39537 |
+
"loss": 0.0,
|
| 39538 |
+
"step": 5647
|
| 39539 |
+
},
|
| 39540 |
+
{
|
| 39541 |
+
"epoch": 0.949962156252628,
|
| 39542 |
+
"grad_norm": NaN,
|
| 39543 |
+
"learning_rate": 6.357131278987849e-07,
|
| 39544 |
+
"loss": 0.0,
|
| 39545 |
+
"step": 5648
|
| 39546 |
+
},
|
| 39547 |
+
{
|
| 39548 |
+
"epoch": 0.9501303506853923,
|
| 39549 |
+
"grad_norm": NaN,
|
| 39550 |
+
"learning_rate": 6.314484565734636e-07,
|
| 39551 |
+
"loss": 0.0,
|
| 39552 |
+
"step": 5649
|
| 39553 |
+
},
|
| 39554 |
+
{
|
| 39555 |
+
"epoch": 0.9502985451181566,
|
| 39556 |
+
"grad_norm": NaN,
|
| 39557 |
+
"learning_rate": 6.271980472697225e-07,
|
| 39558 |
+
"loss": 0.0,
|
| 39559 |
+
"step": 5650
|
| 39560 |
+
},
|
| 39561 |
+
{
|
| 39562 |
+
"epoch": 0.9504667395509209,
|
| 39563 |
+
"grad_norm": NaN,
|
| 39564 |
+
"learning_rate": 6.229619012154575e-07,
|
| 39565 |
+
"loss": 0.0,
|
| 39566 |
+
"step": 5651
|
| 39567 |
+
},
|
| 39568 |
+
{
|
| 39569 |
+
"epoch": 0.9506349339836851,
|
| 39570 |
+
"grad_norm": NaN,
|
| 39571 |
+
"learning_rate": 6.18740019634445e-07,
|
| 39572 |
+
"loss": 0.0,
|
| 39573 |
+
"step": 5652
|
| 39574 |
+
},
|
| 39575 |
+
{
|
| 39576 |
+
"epoch": 0.9508031284164494,
|
| 39577 |
+
"grad_norm": NaN,
|
| 39578 |
+
"learning_rate": 6.145324037463429e-07,
|
| 39579 |
+
"loss": 0.0,
|
| 39580 |
+
"step": 5653
|
| 39581 |
+
},
|
| 39582 |
+
{
|
| 39583 |
+
"epoch": 0.9509713228492137,
|
| 39584 |
+
"grad_norm": NaN,
|
| 39585 |
+
"learning_rate": 6.103390547666788e-07,
|
| 39586 |
+
"loss": 0.0,
|
| 39587 |
+
"step": 5654
|
| 39588 |
+
},
|
| 39589 |
+
{
|
| 39590 |
+
"epoch": 0.951139517281978,
|
| 39591 |
+
"grad_norm": NaN,
|
| 39592 |
+
"learning_rate": 6.061599739068668e-07,
|
| 39593 |
+
"loss": 0.0,
|
| 39594 |
+
"step": 5655
|
| 39595 |
+
},
|
| 39596 |
+
{
|
| 39597 |
+
"epoch": 0.9513077117147423,
|
| 39598 |
+
"grad_norm": NaN,
|
| 39599 |
+
"learning_rate": 6.019951623741916e-07,
|
| 39600 |
+
"loss": 0.0,
|
| 39601 |
+
"step": 5656
|
| 39602 |
+
},
|
| 39603 |
+
{
|
| 39604 |
+
"epoch": 0.9514759061475065,
|
| 39605 |
+
"grad_norm": NaN,
|
| 39606 |
+
"learning_rate": 5.978446213718291e-07,
|
| 39607 |
+
"loss": 0.0,
|
| 39608 |
+
"step": 5657
|
| 39609 |
+
},
|
| 39610 |
+
{
|
| 39611 |
+
"epoch": 0.9516441005802708,
|
| 39612 |
+
"grad_norm": NaN,
|
| 39613 |
+
"learning_rate": 5.937083520988151e-07,
|
| 39614 |
+
"loss": 0.0,
|
| 39615 |
+
"step": 5658
|
| 39616 |
+
},
|
| 39617 |
+
{
|
| 39618 |
+
"epoch": 0.9518122950130351,
|
| 39619 |
+
"grad_norm": NaN,
|
| 39620 |
+
"learning_rate": 5.895863557500769e-07,
|
| 39621 |
+
"loss": 0.0,
|
| 39622 |
+
"step": 5659
|
| 39623 |
+
},
|
| 39624 |
+
{
|
| 39625 |
+
"epoch": 0.9519804894457994,
|
| 39626 |
+
"grad_norm": NaN,
|
| 39627 |
+
"learning_rate": 5.85478633516412e-07,
|
| 39628 |
+
"loss": 0.0,
|
| 39629 |
+
"step": 5660
|
| 39630 |
+
},
|
| 39631 |
+
{
|
| 39632 |
+
"epoch": 0.9521486838785637,
|
| 39633 |
+
"grad_norm": NaN,
|
| 39634 |
+
"learning_rate": 5.813851865844988e-07,
|
| 39635 |
+
"loss": 0.0,
|
| 39636 |
+
"step": 5661
|
| 39637 |
+
},
|
| 39638 |
+
{
|
| 39639 |
+
"epoch": 0.9523168783113279,
|
| 39640 |
+
"grad_norm": NaN,
|
| 39641 |
+
"learning_rate": 5.773060161368804e-07,
|
| 39642 |
+
"loss": 0.0,
|
| 39643 |
+
"step": 5662
|
| 39644 |
+
},
|
| 39645 |
+
{
|
| 39646 |
+
"epoch": 0.9524850727440922,
|
| 39647 |
+
"grad_norm": NaN,
|
| 39648 |
+
"learning_rate": 5.732411233519919e-07,
|
| 39649 |
+
"loss": 0.0,
|
| 39650 |
+
"step": 5663
|
| 39651 |
+
},
|
| 39652 |
+
{
|
| 39653 |
+
"epoch": 0.9526532671768565,
|
| 39654 |
+
"grad_norm": NaN,
|
| 39655 |
+
"learning_rate": 5.691905094041272e-07,
|
| 39656 |
+
"loss": 0.0,
|
| 39657 |
+
"step": 5664
|
| 39658 |
+
},
|
| 39659 |
+
{
|
| 39660 |
+
"epoch": 0.9528214616096207,
|
| 39661 |
+
"grad_norm": NaN,
|
| 39662 |
+
"learning_rate": 5.651541754634726e-07,
|
| 39663 |
+
"loss": 0.0,
|
| 39664 |
+
"step": 5665
|
| 39665 |
+
},
|
| 39666 |
+
{
|
| 39667 |
+
"epoch": 0.952989656042385,
|
| 39668 |
+
"grad_norm": NaN,
|
| 39669 |
+
"learning_rate": 5.611321226960675e-07,
|
| 39670 |
+
"loss": 0.0,
|
| 39671 |
+
"step": 5666
|
| 39672 |
+
},
|
| 39673 |
+
{
|
| 39674 |
+
"epoch": 0.9531578504751492,
|
| 39675 |
+
"grad_norm": NaN,
|
| 39676 |
+
"learning_rate": 5.571243522638547e-07,
|
| 39677 |
+
"loss": 0.0,
|
| 39678 |
+
"step": 5667
|
| 39679 |
+
},
|
| 39680 |
+
{
|
| 39681 |
+
"epoch": 0.9533260449079135,
|
| 39682 |
+
"grad_norm": NaN,
|
| 39683 |
+
"learning_rate": 5.531308653246192e-07,
|
| 39684 |
+
"loss": 0.0,
|
| 39685 |
+
"step": 5668
|
| 39686 |
+
},
|
| 39687 |
+
{
|
| 39688 |
+
"epoch": 0.9534942393406778,
|
| 39689 |
+
"grad_norm": NaN,
|
| 39690 |
+
"learning_rate": 5.491516630320381e-07,
|
| 39691 |
+
"loss": 0.0,
|
| 39692 |
+
"step": 5669
|
| 39693 |
+
},
|
| 39694 |
+
{
|
| 39695 |
+
"epoch": 0.9536624337734421,
|
| 39696 |
+
"grad_norm": NaN,
|
| 39697 |
+
"learning_rate": 5.451867465356641e-07,
|
| 39698 |
+
"loss": 0.0,
|
| 39699 |
+
"step": 5670
|
| 39700 |
+
},
|
| 39701 |
+
{
|
| 39702 |
+
"epoch": 0.9538306282062063,
|
| 39703 |
+
"grad_norm": NaN,
|
| 39704 |
+
"learning_rate": 5.412361169809088e-07,
|
| 39705 |
+
"loss": 0.0,
|
| 39706 |
+
"step": 5671
|
| 39707 |
+
},
|
| 39708 |
+
{
|
| 39709 |
+
"epoch": 0.9539988226389706,
|
| 39710 |
+
"grad_norm": NaN,
|
| 39711 |
+
"learning_rate": 5.372997755090759e-07,
|
| 39712 |
+
"loss": 0.0,
|
| 39713 |
+
"step": 5672
|
| 39714 |
+
},
|
| 39715 |
+
{
|
| 39716 |
+
"epoch": 0.9541670170717349,
|
| 39717 |
+
"grad_norm": NaN,
|
| 39718 |
+
"learning_rate": 5.333777232573223e-07,
|
| 39719 |
+
"loss": 0.0,
|
| 39720 |
+
"step": 5673
|
| 39721 |
+
},
|
| 39722 |
+
{
|
| 39723 |
+
"epoch": 0.9543352115044992,
|
| 39724 |
+
"grad_norm": NaN,
|
| 39725 |
+
"learning_rate": 5.294699613586862e-07,
|
| 39726 |
+
"loss": 0.0,
|
| 39727 |
+
"step": 5674
|
| 39728 |
+
},
|
| 39729 |
+
{
|
| 39730 |
+
"epoch": 0.9545034059372635,
|
| 39731 |
+
"grad_norm": NaN,
|
| 39732 |
+
"learning_rate": 5.255764909420757e-07,
|
| 39733 |
+
"loss": 0.0,
|
| 39734 |
+
"step": 5675
|
| 39735 |
+
},
|
| 39736 |
+
{
|
| 39737 |
+
"epoch": 0.9546716003700277,
|
| 39738 |
+
"grad_norm": NaN,
|
| 39739 |
+
"learning_rate": 5.216973131322689e-07,
|
| 39740 |
+
"loss": 0.0,
|
| 39741 |
+
"step": 5676
|
| 39742 |
+
},
|
| 39743 |
+
{
|
| 39744 |
+
"epoch": 0.954839794802792,
|
| 39745 |
+
"grad_norm": NaN,
|
| 39746 |
+
"learning_rate": 5.178324290499248e-07,
|
| 39747 |
+
"loss": 0.0,
|
| 39748 |
+
"step": 5677
|
| 39749 |
+
},
|
| 39750 |
+
{
|
| 39751 |
+
"epoch": 0.9550079892355563,
|
| 39752 |
+
"grad_norm": NaN,
|
| 39753 |
+
"learning_rate": 5.139818398115559e-07,
|
| 39754 |
+
"loss": 0.0,
|
| 39755 |
+
"step": 5678
|
| 39756 |
+
},
|
| 39757 |
+
{
|
| 39758 |
+
"epoch": 0.9551761836683206,
|
| 39759 |
+
"grad_norm": NaN,
|
| 39760 |
+
"learning_rate": 5.101455465295557e-07,
|
| 39761 |
+
"loss": 0.0,
|
| 39762 |
+
"step": 5679
|
| 39763 |
+
},
|
| 39764 |
+
{
|
| 39765 |
+
"epoch": 0.9553443781010849,
|
| 39766 |
+
"grad_norm": NaN,
|
| 39767 |
+
"learning_rate": 5.063235503121933e-07,
|
| 39768 |
+
"loss": 0.0,
|
| 39769 |
+
"step": 5680
|
| 39770 |
+
},
|
| 39771 |
+
{
|
| 39772 |
+
"epoch": 0.9555125725338491,
|
| 39773 |
+
"grad_norm": NaN,
|
| 39774 |
+
"learning_rate": 5.025158522635964e-07,
|
| 39775 |
+
"loss": 0.0,
|
| 39776 |
+
"step": 5681
|
| 39777 |
+
},
|
| 39778 |
+
{
|
| 39779 |
+
"epoch": 0.9556807669666134,
|
| 39780 |
+
"grad_norm": NaN,
|
| 39781 |
+
"learning_rate": 4.987224534837631e-07,
|
| 39782 |
+
"loss": 0.0,
|
| 39783 |
+
"step": 5682
|
| 39784 |
+
},
|
| 39785 |
+
{
|
| 39786 |
+
"epoch": 0.9558489613993777,
|
| 39787 |
+
"grad_norm": NaN,
|
| 39788 |
+
"learning_rate": 4.949433550685722e-07,
|
| 39789 |
+
"loss": 0.0,
|
| 39790 |
+
"step": 5683
|
| 39791 |
+
},
|
| 39792 |
+
{
|
| 39793 |
+
"epoch": 0.956017155832142,
|
| 39794 |
+
"grad_norm": NaN,
|
| 39795 |
+
"learning_rate": 4.911785581097561e-07,
|
| 39796 |
+
"loss": 0.0,
|
| 39797 |
+
"step": 5684
|
| 39798 |
+
},
|
| 39799 |
+
{
|
| 39800 |
+
"epoch": 0.9561853502649063,
|
| 39801 |
+
"grad_norm": NaN,
|
| 39802 |
+
"learning_rate": 4.874280636949225e-07,
|
| 39803 |
+
"loss": 0.0,
|
| 39804 |
+
"step": 5685
|
| 39805 |
+
},
|
| 39806 |
+
{
|
| 39807 |
+
"epoch": 0.9563535446976705,
|
| 39808 |
+
"grad_norm": NaN,
|
| 39809 |
+
"learning_rate": 4.836918729075435e-07,
|
| 39810 |
+
"loss": 0.0,
|
| 39811 |
+
"step": 5686
|
| 39812 |
+
},
|
| 39813 |
+
{
|
| 39814 |
+
"epoch": 0.9565217391304348,
|
| 39815 |
+
"grad_norm": NaN,
|
| 39816 |
+
"learning_rate": 4.79969986826978e-07,
|
| 39817 |
+
"loss": 0.0,
|
| 39818 |
+
"step": 5687
|
| 39819 |
+
},
|
| 39820 |
+
{
|
| 39821 |
+
"epoch": 0.9566899335631991,
|
| 39822 |
+
"grad_norm": NaN,
|
| 39823 |
+
"learning_rate": 4.7626240652842155e-07,
|
| 39824 |
+
"loss": 0.0,
|
| 39825 |
+
"step": 5688
|
| 39826 |
+
},
|
| 39827 |
+
{
|
| 39828 |
+
"epoch": 0.9568581279959634,
|
| 39829 |
+
"grad_norm": NaN,
|
| 39830 |
+
"learning_rate": 4.7256913308295627e-07,
|
| 39831 |
+
"loss": 0.0,
|
| 39832 |
+
"step": 5689
|
| 39833 |
+
},
|
| 39834 |
+
{
|
| 39835 |
+
"epoch": 0.9570263224287276,
|
| 39836 |
+
"grad_norm": NaN,
|
| 39837 |
+
"learning_rate": 4.688901675575341e-07,
|
| 39838 |
+
"loss": 0.0,
|
| 39839 |
+
"step": 5690
|
| 39840 |
+
},
|
| 39841 |
+
{
|
| 39842 |
+
"epoch": 0.9571945168614919,
|
| 39843 |
+
"grad_norm": NaN,
|
| 39844 |
+
"learning_rate": 4.6522551101496057e-07,
|
| 39845 |
+
"loss": 0.0,
|
| 39846 |
+
"step": 5691
|
| 39847 |
+
},
|
| 39848 |
+
{
|
| 39849 |
+
"epoch": 0.9573627112942562,
|
| 39850 |
+
"grad_norm": NaN,
|
| 39851 |
+
"learning_rate": 4.6157516451391656e-07,
|
| 39852 |
+
"loss": 0.0,
|
| 39853 |
+
"step": 5692
|
| 39854 |
+
},
|
| 39855 |
+
{
|
| 39856 |
+
"epoch": 0.9575309057270205,
|
| 39857 |
+
"grad_norm": NaN,
|
| 39858 |
+
"learning_rate": 4.579391291089419e-07,
|
| 39859 |
+
"loss": 0.0,
|
| 39860 |
+
"step": 5693
|
| 39861 |
+
},
|
| 39862 |
+
{
|
| 39863 |
+
"epoch": 0.9576991001597848,
|
| 39864 |
+
"grad_norm": NaN,
|
| 39865 |
+
"learning_rate": 4.543174058504518e-07,
|
| 39866 |
+
"loss": 0.0,
|
| 39867 |
+
"step": 5694
|
| 39868 |
+
},
|
| 39869 |
+
{
|
| 39870 |
+
"epoch": 0.9578672945925489,
|
| 39871 |
+
"grad_norm": NaN,
|
| 39872 |
+
"learning_rate": 4.507099957847205e-07,
|
| 39873 |
+
"loss": 0.0,
|
| 39874 |
+
"step": 5695
|
| 39875 |
+
},
|
| 39876 |
+
{
|
| 39877 |
+
"epoch": 0.9580354890253132,
|
| 39878 |
+
"grad_norm": NaN,
|
| 39879 |
+
"learning_rate": 4.4711689995389216e-07,
|
| 39880 |
+
"loss": 0.0,
|
| 39881 |
+
"step": 5696
|
| 39882 |
+
},
|
| 39883 |
+
{
|
| 39884 |
+
"epoch": 0.9582036834580775,
|
| 39885 |
+
"grad_norm": NaN,
|
| 39886 |
+
"learning_rate": 4.435381193959587e-07,
|
| 39887 |
+
"loss": 0.0,
|
| 39888 |
+
"step": 5697
|
| 39889 |
+
},
|
| 39890 |
+
{
|
| 39891 |
+
"epoch": 0.9583718778908418,
|
| 39892 |
+
"grad_norm": NaN,
|
| 39893 |
+
"learning_rate": 4.3997365514480416e-07,
|
| 39894 |
+
"loss": 0.0,
|
| 39895 |
+
"step": 5698
|
| 39896 |
+
},
|
| 39897 |
+
{
|
| 39898 |
+
"epoch": 0.958540072323606,
|
| 39899 |
+
"grad_norm": NaN,
|
| 39900 |
+
"learning_rate": 4.364235082301549e-07,
|
| 39901 |
+
"loss": 0.0,
|
| 39902 |
+
"step": 5699
|
| 39903 |
+
},
|
| 39904 |
+
{
|
| 39905 |
+
"epoch": 0.9587082667563703,
|
| 39906 |
+
"grad_norm": NaN,
|
| 39907 |
+
"learning_rate": 4.3288767967760715e-07,
|
| 39908 |
+
"loss": 0.0,
|
| 39909 |
+
"step": 5700
|
| 39910 |
+
},
|
| 39911 |
+
{
|
| 39912 |
+
"epoch": 0.9588764611891346,
|
| 39913 |
+
"grad_norm": NaN,
|
| 39914 |
+
"learning_rate": 4.293661705086327e-07,
|
| 39915 |
+
"loss": 0.0,
|
| 39916 |
+
"step": 5701
|
| 39917 |
+
},
|
| 39918 |
+
{
|
| 39919 |
+
"epoch": 0.9590446556218989,
|
| 39920 |
+
"grad_norm": NaN,
|
| 39921 |
+
"learning_rate": 4.258589817405401e-07,
|
| 39922 |
+
"loss": 0.0,
|
| 39923 |
+
"step": 5702
|
| 39924 |
+
},
|
| 39925 |
+
{
|
| 39926 |
+
"epoch": 0.9592128500546632,
|
| 39927 |
+
"grad_norm": NaN,
|
| 39928 |
+
"learning_rate": 4.2236611438652986e-07,
|
| 39929 |
+
"loss": 0.0,
|
| 39930 |
+
"step": 5703
|
| 39931 |
+
},
|
| 39932 |
+
{
|
| 39933 |
+
"epoch": 0.9593810444874274,
|
| 39934 |
+
"grad_norm": NaN,
|
| 39935 |
+
"learning_rate": 4.188875694556449e-07,
|
| 39936 |
+
"loss": 0.0,
|
| 39937 |
+
"step": 5704
|
| 39938 |
+
},
|
| 39939 |
+
{
|
| 39940 |
+
"epoch": 0.9595492389201917,
|
| 39941 |
+
"grad_norm": NaN,
|
| 39942 |
+
"learning_rate": 4.1542334795280355e-07,
|
| 39943 |
+
"loss": 0.0,
|
| 39944 |
+
"step": 5705
|
| 39945 |
+
},
|
| 39946 |
+
{
|
| 39947 |
+
"epoch": 0.959717433352956,
|
| 39948 |
+
"grad_norm": NaN,
|
| 39949 |
+
"learning_rate": 4.119734508787776e-07,
|
| 39950 |
+
"loss": 0.0,
|
| 39951 |
+
"step": 5706
|
| 39952 |
+
},
|
| 39953 |
+
{
|
| 39954 |
+
"epoch": 0.9598856277857203,
|
| 39955 |
+
"grad_norm": NaN,
|
| 39956 |
+
"learning_rate": 4.0853787923020303e-07,
|
| 39957 |
+
"loss": 0.0,
|
| 39958 |
+
"step": 5707
|
| 39959 |
+
},
|
| 39960 |
+
{
|
| 39961 |
+
"epoch": 0.9600538222184846,
|
| 39962 |
+
"grad_norm": NaN,
|
| 39963 |
+
"learning_rate": 4.0511663399958044e-07,
|
| 39964 |
+
"loss": 0.0,
|
| 39965 |
+
"step": 5708
|
| 39966 |
+
},
|
| 39967 |
+
{
|
| 39968 |
+
"epoch": 0.9602220166512488,
|
| 39969 |
+
"grad_norm": NaN,
|
| 39970 |
+
"learning_rate": 4.017097161752692e-07,
|
| 39971 |
+
"loss": 0.0,
|
| 39972 |
+
"step": 5709
|
| 39973 |
+
},
|
| 39974 |
+
{
|
| 39975 |
+
"epoch": 0.9603902110840131,
|
| 39976 |
+
"grad_norm": NaN,
|
| 39977 |
+
"learning_rate": 3.983171267414876e-07,
|
| 39978 |
+
"loss": 0.0,
|
| 39979 |
+
"step": 5710
|
| 39980 |
+
},
|
| 39981 |
+
{
|
| 39982 |
+
"epoch": 0.9605584055167774,
|
| 39983 |
+
"grad_norm": NaN,
|
| 39984 |
+
"learning_rate": 3.949388666783127e-07,
|
| 39985 |
+
"loss": 0.0,
|
| 39986 |
+
"step": 5711
|
| 39987 |
+
},
|
| 39988 |
+
{
|
| 39989 |
+
"epoch": 0.9607265999495417,
|
| 39990 |
+
"grad_norm": NaN,
|
| 39991 |
+
"learning_rate": 3.9157493696169724e-07,
|
| 39992 |
+
"loss": 0.0,
|
| 39993 |
+
"step": 5712
|
| 39994 |
+
},
|
| 39995 |
+
{
|
| 39996 |
+
"epoch": 0.960894794382306,
|
| 39997 |
+
"grad_norm": NaN,
|
| 39998 |
+
"learning_rate": 3.8822533856343044e-07,
|
| 39999 |
+
"loss": 0.0,
|
| 40000 |
+
"step": 5713
|
| 40001 |
+
},
|
| 40002 |
+
{
|
| 40003 |
+
"epoch": 0.9610629888150702,
|
| 40004 |
+
"grad_norm": NaN,
|
| 40005 |
+
"learning_rate": 3.848900724511828e-07,
|
| 40006 |
+
"loss": 0.0,
|
| 40007 |
+
"step": 5714
|
| 40008 |
+
},
|
| 40009 |
+
{
|
| 40010 |
+
"epoch": 0.9612311832478345,
|
| 40011 |
+
"grad_norm": NaN,
|
| 40012 |
+
"learning_rate": 3.815691395884724e-07,
|
| 40013 |
+
"loss": 0.0,
|
| 40014 |
+
"step": 5715
|
| 40015 |
+
},
|
| 40016 |
+
{
|
| 40017 |
+
"epoch": 0.9613993776805988,
|
| 40018 |
+
"grad_norm": NaN,
|
| 40019 |
+
"learning_rate": 3.782625409346763e-07,
|
| 40020 |
+
"loss": 0.0,
|
| 40021 |
+
"step": 5716
|
| 40022 |
+
},
|
| 40023 |
+
{
|
| 40024 |
+
"epoch": 0.9615675721133631,
|
| 40025 |
+
"grad_norm": NaN,
|
| 40026 |
+
"learning_rate": 3.749702774450414e-07,
|
| 40027 |
+
"loss": 0.0,
|
| 40028 |
+
"step": 5717
|
| 40029 |
+
},
|
| 40030 |
+
{
|
| 40031 |
+
"epoch": 0.9617357665461274,
|
| 40032 |
+
"grad_norm": NaN,
|
| 40033 |
+
"learning_rate": 3.7169235007065707e-07,
|
| 40034 |
+
"loss": 0.0,
|
| 40035 |
+
"step": 5718
|
| 40036 |
+
},
|
| 40037 |
+
{
|
| 40038 |
+
"epoch": 0.9619039609788916,
|
| 40039 |
+
"grad_norm": NaN,
|
| 40040 |
+
"learning_rate": 3.684287597584879e-07,
|
| 40041 |
+
"loss": 0.0,
|
| 40042 |
+
"step": 5719
|
| 40043 |
+
},
|
| 40044 |
+
{
|
| 40045 |
+
"epoch": 0.9620721554116559,
|
| 40046 |
+
"grad_norm": NaN,
|
| 40047 |
+
"learning_rate": 3.651795074513409e-07,
|
| 40048 |
+
"loss": 0.0,
|
| 40049 |
+
"step": 5720
|
| 40050 |
+
},
|
| 40051 |
+
{
|
| 40052 |
+
"epoch": 0.9622403498444202,
|
| 40053 |
+
"grad_norm": NaN,
|
| 40054 |
+
"learning_rate": 3.619445940878929e-07,
|
| 40055 |
+
"loss": 0.0,
|
| 40056 |
+
"step": 5721
|
| 40057 |
+
},
|
| 40058 |
+
{
|
| 40059 |
+
"epoch": 0.9624085442771845,
|
| 40060 |
+
"grad_norm": NaN,
|
| 40061 |
+
"learning_rate": 3.587240206026743e-07,
|
| 40062 |
+
"loss": 0.0,
|
| 40063 |
+
"step": 5722
|
| 40064 |
+
},
|
| 40065 |
+
{
|
| 40066 |
+
"epoch": 0.9625767387099488,
|
| 40067 |
+
"grad_norm": NaN,
|
| 40068 |
+
"learning_rate": 3.555177879260685e-07,
|
| 40069 |
+
"loss": 0.0,
|
| 40070 |
+
"step": 5723
|
| 40071 |
+
},
|
| 40072 |
+
{
|
| 40073 |
+
"epoch": 0.9627449331427129,
|
| 40074 |
+
"grad_norm": NaN,
|
| 40075 |
+
"learning_rate": 3.5232589698432907e-07,
|
| 40076 |
+
"loss": 0.0,
|
| 40077 |
+
"step": 5724
|
| 40078 |
+
},
|
| 40079 |
+
{
|
| 40080 |
+
"epoch": 0.9629131275754772,
|
| 40081 |
+
"grad_norm": NaN,
|
| 40082 |
+
"learning_rate": 3.491483486995517e-07,
|
| 40083 |
+
"loss": 0.0,
|
| 40084 |
+
"step": 5725
|
| 40085 |
+
},
|
| 40086 |
+
{
|
| 40087 |
+
"epoch": 0.9630813220082415,
|
| 40088 |
+
"grad_norm": NaN,
|
| 40089 |
+
"learning_rate": 3.459851439896966e-07,
|
| 40090 |
+
"loss": 0.0,
|
| 40091 |
+
"step": 5726
|
| 40092 |
+
},
|
| 40093 |
+
{
|
| 40094 |
+
"epoch": 0.9632495164410058,
|
| 40095 |
+
"grad_norm": NaN,
|
| 40096 |
+
"learning_rate": 3.428362837685717e-07,
|
| 40097 |
+
"loss": 0.0,
|
| 40098 |
+
"step": 5727
|
| 40099 |
+
},
|
| 40100 |
+
{
|
| 40101 |
+
"epoch": 0.96341771087377,
|
| 40102 |
+
"grad_norm": NaN,
|
| 40103 |
+
"learning_rate": 3.397017689458548e-07,
|
| 40104 |
+
"loss": 0.0,
|
| 40105 |
+
"step": 5728
|
| 40106 |
+
},
|
| 40107 |
+
{
|
| 40108 |
+
"epoch": 0.9635859053065343,
|
| 40109 |
+
"grad_norm": NaN,
|
| 40110 |
+
"learning_rate": 3.365816004270661e-07,
|
| 40111 |
+
"loss": 0.0,
|
| 40112 |
+
"step": 5729
|
| 40113 |
+
},
|
| 40114 |
+
{
|
| 40115 |
+
"epoch": 0.9637540997392986,
|
| 40116 |
+
"grad_norm": NaN,
|
| 40117 |
+
"learning_rate": 3.334757791135956e-07,
|
| 40118 |
+
"loss": 0.0,
|
| 40119 |
+
"step": 5730
|
| 40120 |
+
},
|
| 40121 |
+
{
|
| 40122 |
+
"epoch": 0.9639222941720629,
|
| 40123 |
+
"grad_norm": NaN,
|
| 40124 |
+
"learning_rate": 3.303843059026757e-07,
|
| 40125 |
+
"loss": 0.0,
|
| 40126 |
+
"step": 5731
|
| 40127 |
+
},
|
| 40128 |
+
{
|
| 40129 |
+
"epoch": 0.9640904886048272,
|
| 40130 |
+
"grad_norm": NaN,
|
| 40131 |
+
"learning_rate": 3.273071816873974e-07,
|
| 40132 |
+
"loss": 0.0,
|
| 40133 |
+
"step": 5732
|
| 40134 |
+
},
|
| 40135 |
+
{
|
| 40136 |
+
"epoch": 0.9642586830375914,
|
| 40137 |
+
"grad_norm": NaN,
|
| 40138 |
+
"learning_rate": 3.2424440735670526e-07,
|
| 40139 |
+
"loss": 0.0,
|
| 40140 |
+
"step": 5733
|
| 40141 |
+
},
|
| 40142 |
+
{
|
| 40143 |
+
"epoch": 0.9644268774703557,
|
| 40144 |
+
"grad_norm": NaN,
|
| 40145 |
+
"learning_rate": 3.211959837954026e-07,
|
| 40146 |
+
"loss": 0.0,
|
| 40147 |
+
"step": 5734
|
| 40148 |
+
},
|
| 40149 |
+
{
|
| 40150 |
+
"epoch": 0.96459507190312,
|
| 40151 |
+
"grad_norm": NaN,
|
| 40152 |
+
"learning_rate": 3.1816191188415166e-07,
|
| 40153 |
+
"loss": 0.0,
|
| 40154 |
+
"step": 5735
|
| 40155 |
+
},
|
| 40156 |
+
{
|
| 40157 |
+
"epoch": 0.9647632663358843,
|
| 40158 |
+
"grad_norm": NaN,
|
| 40159 |
+
"learning_rate": 3.151421924994513e-07,
|
| 40160 |
+
"loss": 0.0,
|
| 40161 |
+
"step": 5736
|
| 40162 |
}
|
| 40163 |
],
|
| 40164 |
"logging_steps": 1,
|
|
|
|
| 40178 |
"attributes": {}
|
| 40179 |
}
|
| 40180 |
},
|
| 40181 |
+
"total_flos": 1.0797842128699392e+17,
|
| 40182 |
"train_batch_size": 8,
|
| 40183 |
"trial_name": null,
|
| 40184 |
"trial_params": null
|