Training in progress, step 488, 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 3582214344
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c9f68b058e6737e15a24e314cf3497ce1edd16467c55b67887c24bd4d1a035ea
|
| 3 |
size 3582214344
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6fee056d1973b6780b8b93d867a2ff656ca2ed3e48a3387c39f28fc6b52a564
|
| 3 |
+
size 5116251049
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec2cf52e4c3f15759e904ba93b5a10165bb850df54a226c7977bcfa5b79f76fb
|
| 3 |
size 14645
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b6a0b7551e7eea8f9be737f8895d7cea6a958fc4c6adc753e794396a9c28203a
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 244,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1724,6 +1724,1722 @@
|
|
| 1724 |
"eval_samples_per_second": 39.172,
|
| 1725 |
"eval_steps_per_second": 9.793,
|
| 1726 |
"step": 244
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1727 |
}
|
| 1728 |
],
|
| 1729 |
"logging_steps": 1,
|
|
@@ -1743,7 +3459,7 @@
|
|
| 1743 |
"attributes": {}
|
| 1744 |
}
|
| 1745 |
},
|
| 1746 |
-
"total_flos":
|
| 1747 |
"train_batch_size": 4,
|
| 1748 |
"trial_name": null,
|
| 1749 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.4004102564102564,
|
| 6 |
"eval_steps": 244,
|
| 7 |
+
"global_step": 488,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1724 |
"eval_samples_per_second": 39.172,
|
| 1725 |
"eval_steps_per_second": 9.793,
|
| 1726 |
"step": 244
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 0.20102564102564102,
|
| 1730 |
+
"grad_norm": 1.6875,
|
| 1731 |
+
"learning_rate": 1.133308129899189e-05,
|
| 1732 |
+
"loss": 1.128773808479309,
|
| 1733 |
+
"step": 245
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.20184615384615384,
|
| 1737 |
+
"grad_norm": 1.7578125,
|
| 1738 |
+
"learning_rate": 1.1323983010326164e-05,
|
| 1739 |
+
"loss": 1.1320472955703735,
|
| 1740 |
+
"step": 246
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 0.20266666666666666,
|
| 1744 |
+
"grad_norm": 1.6875,
|
| 1745 |
+
"learning_rate": 1.1314849777394252e-05,
|
| 1746 |
+
"loss": 1.1400765180587769,
|
| 1747 |
+
"step": 247
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 0.20348717948717948,
|
| 1751 |
+
"grad_norm": 1.6328125,
|
| 1752 |
+
"learning_rate": 1.1305681662482482e-05,
|
| 1753 |
+
"loss": 1.2028186321258545,
|
| 1754 |
+
"step": 248
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 0.2043076923076923,
|
| 1758 |
+
"grad_norm": 1.5703125,
|
| 1759 |
+
"learning_rate": 1.1296478728115077e-05,
|
| 1760 |
+
"loss": 1.1705706119537354,
|
| 1761 |
+
"step": 249
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 0.20512820512820512,
|
| 1765 |
+
"grad_norm": 1.6015625,
|
| 1766 |
+
"learning_rate": 1.128724103705372e-05,
|
| 1767 |
+
"loss": 1.0424087047576904,
|
| 1768 |
+
"step": 250
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.20594871794871794,
|
| 1772 |
+
"grad_norm": 1.5390625,
|
| 1773 |
+
"learning_rate": 1.1277968652297124e-05,
|
| 1774 |
+
"loss": 1.1476494073867798,
|
| 1775 |
+
"step": 251
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 0.20676923076923076,
|
| 1779 |
+
"grad_norm": 1.6875,
|
| 1780 |
+
"learning_rate": 1.1268661637080603e-05,
|
| 1781 |
+
"loss": 1.0242687463760376,
|
| 1782 |
+
"step": 252
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 0.20758974358974358,
|
| 1786 |
+
"grad_norm": 1.5234375,
|
| 1787 |
+
"learning_rate": 1.1259320054875643e-05,
|
| 1788 |
+
"loss": 1.124701976776123,
|
| 1789 |
+
"step": 253
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 0.2084102564102564,
|
| 1793 |
+
"grad_norm": 1.6953125,
|
| 1794 |
+
"learning_rate": 1.1249943969389469e-05,
|
| 1795 |
+
"loss": 1.1467115879058838,
|
| 1796 |
+
"step": 254
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.20923076923076922,
|
| 1800 |
+
"grad_norm": 1.609375,
|
| 1801 |
+
"learning_rate": 1.1240533444564611e-05,
|
| 1802 |
+
"loss": 1.1093086004257202,
|
| 1803 |
+
"step": 255
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.21005128205128204,
|
| 1807 |
+
"grad_norm": 1.625,
|
| 1808 |
+
"learning_rate": 1.1231088544578462e-05,
|
| 1809 |
+
"loss": 1.122301459312439,
|
| 1810 |
+
"step": 256
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 0.21087179487179486,
|
| 1814 |
+
"grad_norm": 1.6484375,
|
| 1815 |
+
"learning_rate": 1.1221609333842848e-05,
|
| 1816 |
+
"loss": 1.0327812433242798,
|
| 1817 |
+
"step": 257
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 0.21169230769230768,
|
| 1821 |
+
"grad_norm": 1.6015625,
|
| 1822 |
+
"learning_rate": 1.1212095877003584e-05,
|
| 1823 |
+
"loss": 1.0453338623046875,
|
| 1824 |
+
"step": 258
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 0.2125128205128205,
|
| 1828 |
+
"grad_norm": 1.6015625,
|
| 1829 |
+
"learning_rate": 1.1202548238940037e-05,
|
| 1830 |
+
"loss": 1.1212118864059448,
|
| 1831 |
+
"step": 259
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 0.21333333333333335,
|
| 1835 |
+
"grad_norm": 1.6484375,
|
| 1836 |
+
"learning_rate": 1.119296648476468e-05,
|
| 1837 |
+
"loss": 1.1293535232543945,
|
| 1838 |
+
"step": 260
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 0.21415384615384617,
|
| 1842 |
+
"grad_norm": 1.7109375,
|
| 1843 |
+
"learning_rate": 1.1183350679822652e-05,
|
| 1844 |
+
"loss": 1.0833102464675903,
|
| 1845 |
+
"step": 261
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 0.214974358974359,
|
| 1849 |
+
"grad_norm": 1.6640625,
|
| 1850 |
+
"learning_rate": 1.1173700889691301e-05,
|
| 1851 |
+
"loss": 1.099400520324707,
|
| 1852 |
+
"step": 262
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 0.2157948717948718,
|
| 1856 |
+
"grad_norm": 1.5078125,
|
| 1857 |
+
"learning_rate": 1.1164017180179758e-05,
|
| 1858 |
+
"loss": 1.1730022430419922,
|
| 1859 |
+
"step": 263
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.21661538461538463,
|
| 1863 |
+
"grad_norm": 1.546875,
|
| 1864 |
+
"learning_rate": 1.1154299617328462e-05,
|
| 1865 |
+
"loss": 1.0063059329986572,
|
| 1866 |
+
"step": 264
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 0.21743589743589745,
|
| 1870 |
+
"grad_norm": 1.625,
|
| 1871 |
+
"learning_rate": 1.1144548267408738e-05,
|
| 1872 |
+
"loss": 1.1822378635406494,
|
| 1873 |
+
"step": 265
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 0.21825641025641027,
|
| 1877 |
+
"grad_norm": 1.8125,
|
| 1878 |
+
"learning_rate": 1.1134763196922315e-05,
|
| 1879 |
+
"loss": 1.1425609588623047,
|
| 1880 |
+
"step": 266
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 0.21907692307692309,
|
| 1884 |
+
"grad_norm": 1.53125,
|
| 1885 |
+
"learning_rate": 1.1124944472600899e-05,
|
| 1886 |
+
"loss": 1.0716297626495361,
|
| 1887 |
+
"step": 267
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 0.2198974358974359,
|
| 1891 |
+
"grad_norm": 1.6171875,
|
| 1892 |
+
"learning_rate": 1.11150921614057e-05,
|
| 1893 |
+
"loss": 1.2099831104278564,
|
| 1894 |
+
"step": 268
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 0.22071794871794873,
|
| 1898 |
+
"grad_norm": 1.765625,
|
| 1899 |
+
"learning_rate": 1.110520633052699e-05,
|
| 1900 |
+
"loss": 1.1891555786132812,
|
| 1901 |
+
"step": 269
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 0.22153846153846155,
|
| 1905 |
+
"grad_norm": 1.6171875,
|
| 1906 |
+
"learning_rate": 1.1095287047383628e-05,
|
| 1907 |
+
"loss": 1.1714974641799927,
|
| 1908 |
+
"step": 270
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 0.22235897435897437,
|
| 1912 |
+
"grad_norm": 1.5703125,
|
| 1913 |
+
"learning_rate": 1.1085334379622611e-05,
|
| 1914 |
+
"loss": 1.0985504388809204,
|
| 1915 |
+
"step": 271
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 0.22317948717948718,
|
| 1919 |
+
"grad_norm": 1.640625,
|
| 1920 |
+
"learning_rate": 1.1075348395118613e-05,
|
| 1921 |
+
"loss": 1.0896762609481812,
|
| 1922 |
+
"step": 272
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 0.224,
|
| 1926 |
+
"grad_norm": 1.640625,
|
| 1927 |
+
"learning_rate": 1.1065329161973521e-05,
|
| 1928 |
+
"loss": 1.105872392654419,
|
| 1929 |
+
"step": 273
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 0.22482051282051282,
|
| 1933 |
+
"grad_norm": 1.765625,
|
| 1934 |
+
"learning_rate": 1.1055276748515965e-05,
|
| 1935 |
+
"loss": 1.1850640773773193,
|
| 1936 |
+
"step": 274
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 0.22564102564102564,
|
| 1940 |
+
"grad_norm": 1.6171875,
|
| 1941 |
+
"learning_rate": 1.1045191223300857e-05,
|
| 1942 |
+
"loss": 1.066878080368042,
|
| 1943 |
+
"step": 275
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 0.22646153846153846,
|
| 1947 |
+
"grad_norm": 1.6796875,
|
| 1948 |
+
"learning_rate": 1.103507265510893e-05,
|
| 1949 |
+
"loss": 1.2129135131835938,
|
| 1950 |
+
"step": 276
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 0.22728205128205128,
|
| 1954 |
+
"grad_norm": 1.7421875,
|
| 1955 |
+
"learning_rate": 1.1024921112946249e-05,
|
| 1956 |
+
"loss": 1.1626890897750854,
|
| 1957 |
+
"step": 277
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 0.2281025641025641,
|
| 1961 |
+
"grad_norm": 1.609375,
|
| 1962 |
+
"learning_rate": 1.1014736666043765e-05,
|
| 1963 |
+
"loss": 1.1131606101989746,
|
| 1964 |
+
"step": 278
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 0.22892307692307692,
|
| 1968 |
+
"grad_norm": 1.734375,
|
| 1969 |
+
"learning_rate": 1.1004519383856822e-05,
|
| 1970 |
+
"loss": 1.1231145858764648,
|
| 1971 |
+
"step": 279
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 0.22974358974358974,
|
| 1975 |
+
"grad_norm": 1.5703125,
|
| 1976 |
+
"learning_rate": 1.09942693360647e-05,
|
| 1977 |
+
"loss": 1.1255182027816772,
|
| 1978 |
+
"step": 280
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 0.23056410256410256,
|
| 1982 |
+
"grad_norm": 1.703125,
|
| 1983 |
+
"learning_rate": 1.0983986592570127e-05,
|
| 1984 |
+
"loss": 1.2375595569610596,
|
| 1985 |
+
"step": 281
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 0.23138461538461538,
|
| 1989 |
+
"grad_norm": 1.7890625,
|
| 1990 |
+
"learning_rate": 1.097367122349881e-05,
|
| 1991 |
+
"loss": 1.1900784969329834,
|
| 1992 |
+
"step": 282
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 0.2322051282051282,
|
| 1996 |
+
"grad_norm": 1.6796875,
|
| 1997 |
+
"learning_rate": 1.0963323299198952e-05,
|
| 1998 |
+
"loss": 1.0929255485534668,
|
| 1999 |
+
"step": 283
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 0.23302564102564102,
|
| 2003 |
+
"grad_norm": 1.78125,
|
| 2004 |
+
"learning_rate": 1.0952942890240775e-05,
|
| 2005 |
+
"loss": 1.1288514137268066,
|
| 2006 |
+
"step": 284
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 0.23384615384615384,
|
| 2010 |
+
"grad_norm": 1.5703125,
|
| 2011 |
+
"learning_rate": 1.094253006741604e-05,
|
| 2012 |
+
"loss": 1.1498799324035645,
|
| 2013 |
+
"step": 285
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 0.23466666666666666,
|
| 2017 |
+
"grad_norm": 1.671875,
|
| 2018 |
+
"learning_rate": 1.0932084901737562e-05,
|
| 2019 |
+
"loss": 1.1852619647979736,
|
| 2020 |
+
"step": 286
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 0.23548717948717948,
|
| 2024 |
+
"grad_norm": 1.765625,
|
| 2025 |
+
"learning_rate": 1.092160746443872e-05,
|
| 2026 |
+
"loss": 1.1017335653305054,
|
| 2027 |
+
"step": 287
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 0.2363076923076923,
|
| 2031 |
+
"grad_norm": 1.484375,
|
| 2032 |
+
"learning_rate": 1.0911097826972988e-05,
|
| 2033 |
+
"loss": 1.133665919303894,
|
| 2034 |
+
"step": 288
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 0.23712820512820512,
|
| 2038 |
+
"grad_norm": 1.6328125,
|
| 2039 |
+
"learning_rate": 1.090055606101343e-05,
|
| 2040 |
+
"loss": 1.042881727218628,
|
| 2041 |
+
"step": 289
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 0.23794871794871794,
|
| 2045 |
+
"grad_norm": 1.8359375,
|
| 2046 |
+
"learning_rate": 1.0889982238452217e-05,
|
| 2047 |
+
"loss": 1.274213194847107,
|
| 2048 |
+
"step": 290
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 0.23876923076923076,
|
| 2052 |
+
"grad_norm": 1.6640625,
|
| 2053 |
+
"learning_rate": 1.0879376431400145e-05,
|
| 2054 |
+
"loss": 1.0823792219161987,
|
| 2055 |
+
"step": 291
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 0.23958974358974358,
|
| 2059 |
+
"grad_norm": 1.578125,
|
| 2060 |
+
"learning_rate": 1.0868738712186126e-05,
|
| 2061 |
+
"loss": 1.1085994243621826,
|
| 2062 |
+
"step": 292
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 0.2404102564102564,
|
| 2066 |
+
"grad_norm": 1.71875,
|
| 2067 |
+
"learning_rate": 1.0858069153356715e-05,
|
| 2068 |
+
"loss": 1.1221598386764526,
|
| 2069 |
+
"step": 293
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 0.24123076923076922,
|
| 2073 |
+
"grad_norm": 1.5859375,
|
| 2074 |
+
"learning_rate": 1.08473678276756e-05,
|
| 2075 |
+
"loss": 1.0073379278182983,
|
| 2076 |
+
"step": 294
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 0.24205128205128204,
|
| 2080 |
+
"grad_norm": 1.7265625,
|
| 2081 |
+
"learning_rate": 1.083663480812311e-05,
|
| 2082 |
+
"loss": 1.0773701667785645,
|
| 2083 |
+
"step": 295
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 0.24287179487179486,
|
| 2087 |
+
"grad_norm": 1.6953125,
|
| 2088 |
+
"learning_rate": 1.0825870167895717e-05,
|
| 2089 |
+
"loss": 1.1612088680267334,
|
| 2090 |
+
"step": 296
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 0.24369230769230768,
|
| 2094 |
+
"grad_norm": 1.5703125,
|
| 2095 |
+
"learning_rate": 1.0815073980405544e-05,
|
| 2096 |
+
"loss": 1.024916172027588,
|
| 2097 |
+
"step": 297
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 0.2445128205128205,
|
| 2101 |
+
"grad_norm": 1.65625,
|
| 2102 |
+
"learning_rate": 1.0804246319279857e-05,
|
| 2103 |
+
"loss": 1.1280243396759033,
|
| 2104 |
+
"step": 298
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 0.24533333333333332,
|
| 2108 |
+
"grad_norm": 1.75,
|
| 2109 |
+
"learning_rate": 1.0793387258360557e-05,
|
| 2110 |
+
"loss": 1.0987873077392578,
|
| 2111 |
+
"step": 299
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 0.24615384615384617,
|
| 2115 |
+
"grad_norm": 1.7265625,
|
| 2116 |
+
"learning_rate": 1.0782496871703695e-05,
|
| 2117 |
+
"loss": 1.1969925165176392,
|
| 2118 |
+
"step": 300
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 0.246974358974359,
|
| 2122 |
+
"grad_norm": 1.6875,
|
| 2123 |
+
"learning_rate": 1.0771575233578943e-05,
|
| 2124 |
+
"loss": 1.0845359563827515,
|
| 2125 |
+
"step": 301
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 0.2477948717948718,
|
| 2129 |
+
"grad_norm": 1.625,
|
| 2130 |
+
"learning_rate": 1.0760622418469108e-05,
|
| 2131 |
+
"loss": 1.0780291557312012,
|
| 2132 |
+
"step": 302
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 0.24861538461538463,
|
| 2136 |
+
"grad_norm": 1.671875,
|
| 2137 |
+
"learning_rate": 1.0749638501069618e-05,
|
| 2138 |
+
"loss": 1.1431422233581543,
|
| 2139 |
+
"step": 303
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 0.24943589743589745,
|
| 2143 |
+
"grad_norm": 1.703125,
|
| 2144 |
+
"learning_rate": 1.0738623556288004e-05,
|
| 2145 |
+
"loss": 1.1646227836608887,
|
| 2146 |
+
"step": 304
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 0.25025641025641027,
|
| 2150 |
+
"grad_norm": 1.4921875,
|
| 2151 |
+
"learning_rate": 1.0727577659243402e-05,
|
| 2152 |
+
"loss": 1.1574097871780396,
|
| 2153 |
+
"step": 305
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 0.2510769230769231,
|
| 2157 |
+
"grad_norm": 1.6953125,
|
| 2158 |
+
"learning_rate": 1.0716500885266029e-05,
|
| 2159 |
+
"loss": 0.9957287311553955,
|
| 2160 |
+
"step": 306
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 0.2518974358974359,
|
| 2164 |
+
"grad_norm": 1.671875,
|
| 2165 |
+
"learning_rate": 1.0705393309896679e-05,
|
| 2166 |
+
"loss": 1.1173666715621948,
|
| 2167 |
+
"step": 307
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"epoch": 0.2527179487179487,
|
| 2171 |
+
"grad_norm": 1.5703125,
|
| 2172 |
+
"learning_rate": 1.0694255008886203e-05,
|
| 2173 |
+
"loss": 1.0518970489501953,
|
| 2174 |
+
"step": 308
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 0.25353846153846155,
|
| 2178 |
+
"grad_norm": 1.6640625,
|
| 2179 |
+
"learning_rate": 1.0683086058194992e-05,
|
| 2180 |
+
"loss": 1.1639254093170166,
|
| 2181 |
+
"step": 309
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 0.25435897435897437,
|
| 2185 |
+
"grad_norm": 1.71875,
|
| 2186 |
+
"learning_rate": 1.0671886533992459e-05,
|
| 2187 |
+
"loss": 1.0966671705245972,
|
| 2188 |
+
"step": 310
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"epoch": 0.2551794871794872,
|
| 2192 |
+
"grad_norm": 1.6875,
|
| 2193 |
+
"learning_rate": 1.0660656512656526e-05,
|
| 2194 |
+
"loss": 1.158909797668457,
|
| 2195 |
+
"step": 311
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 0.256,
|
| 2199 |
+
"grad_norm": 1.71875,
|
| 2200 |
+
"learning_rate": 1.0649396070773088e-05,
|
| 2201 |
+
"loss": 1.162057876586914,
|
| 2202 |
+
"step": 312
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 0.2568205128205128,
|
| 2206 |
+
"grad_norm": 1.6015625,
|
| 2207 |
+
"learning_rate": 1.0638105285135515e-05,
|
| 2208 |
+
"loss": 1.0730208158493042,
|
| 2209 |
+
"step": 313
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 0.25764102564102564,
|
| 2213 |
+
"grad_norm": 1.984375,
|
| 2214 |
+
"learning_rate": 1.0626784232744097e-05,
|
| 2215 |
+
"loss": 1.252415657043457,
|
| 2216 |
+
"step": 314
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 0.25846153846153846,
|
| 2220 |
+
"grad_norm": 1.6796875,
|
| 2221 |
+
"learning_rate": 1.0615432990805544e-05,
|
| 2222 |
+
"loss": 1.2043737173080444,
|
| 2223 |
+
"step": 315
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 0.2592820512820513,
|
| 2227 |
+
"grad_norm": 1.6171875,
|
| 2228 |
+
"learning_rate": 1.0604051636732456e-05,
|
| 2229 |
+
"loss": 1.1311604976654053,
|
| 2230 |
+
"step": 316
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 0.2601025641025641,
|
| 2234 |
+
"grad_norm": 1.71875,
|
| 2235 |
+
"learning_rate": 1.059264024814278e-05,
|
| 2236 |
+
"loss": 1.2474250793457031,
|
| 2237 |
+
"step": 317
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 0.2609230769230769,
|
| 2241 |
+
"grad_norm": 1.640625,
|
| 2242 |
+
"learning_rate": 1.0581198902859299e-05,
|
| 2243 |
+
"loss": 1.082651138305664,
|
| 2244 |
+
"step": 318
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 0.26174358974358974,
|
| 2248 |
+
"grad_norm": 1.6328125,
|
| 2249 |
+
"learning_rate": 1.0569727678909088e-05,
|
| 2250 |
+
"loss": 1.1605055332183838,
|
| 2251 |
+
"step": 319
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 0.26256410256410256,
|
| 2255 |
+
"grad_norm": 1.5859375,
|
| 2256 |
+
"learning_rate": 1.0558226654522988e-05,
|
| 2257 |
+
"loss": 1.1698611974716187,
|
| 2258 |
+
"step": 320
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 0.2633846153846154,
|
| 2262 |
+
"grad_norm": 1.6953125,
|
| 2263 |
+
"learning_rate": 1.0546695908135073e-05,
|
| 2264 |
+
"loss": 1.2056175470352173,
|
| 2265 |
+
"step": 321
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 0.2642051282051282,
|
| 2269 |
+
"grad_norm": 1.8359375,
|
| 2270 |
+
"learning_rate": 1.0535135518382113e-05,
|
| 2271 |
+
"loss": 1.1495481729507446,
|
| 2272 |
+
"step": 322
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 0.265025641025641,
|
| 2276 |
+
"grad_norm": 1.5859375,
|
| 2277 |
+
"learning_rate": 1.0523545564103036e-05,
|
| 2278 |
+
"loss": 1.0383414030075073,
|
| 2279 |
+
"step": 323
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 0.26584615384615384,
|
| 2283 |
+
"grad_norm": 1.703125,
|
| 2284 |
+
"learning_rate": 1.0511926124338395e-05,
|
| 2285 |
+
"loss": 1.2419593334197998,
|
| 2286 |
+
"step": 324
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 0.26666666666666666,
|
| 2290 |
+
"grad_norm": 1.6875,
|
| 2291 |
+
"learning_rate": 1.0500277278329828e-05,
|
| 2292 |
+
"loss": 1.1622841358184814,
|
| 2293 |
+
"step": 325
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 0.2674871794871795,
|
| 2297 |
+
"grad_norm": 1.6796875,
|
| 2298 |
+
"learning_rate": 1.0488599105519513e-05,
|
| 2299 |
+
"loss": 1.1836302280426025,
|
| 2300 |
+
"step": 326
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 0.2683076923076923,
|
| 2304 |
+
"grad_norm": 1.5859375,
|
| 2305 |
+
"learning_rate": 1.0476891685549626e-05,
|
| 2306 |
+
"loss": 1.088476538658142,
|
| 2307 |
+
"step": 327
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 0.2691282051282051,
|
| 2311 |
+
"grad_norm": 1.6015625,
|
| 2312 |
+
"learning_rate": 1.0465155098261813e-05,
|
| 2313 |
+
"loss": 1.2314226627349854,
|
| 2314 |
+
"step": 328
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 0.26994871794871794,
|
| 2318 |
+
"grad_norm": 1.6875,
|
| 2319 |
+
"learning_rate": 1.0453389423696623e-05,
|
| 2320 |
+
"loss": 1.0652053356170654,
|
| 2321 |
+
"step": 329
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 0.27076923076923076,
|
| 2325 |
+
"grad_norm": 1.5234375,
|
| 2326 |
+
"learning_rate": 1.0441594742092974e-05,
|
| 2327 |
+
"loss": 1.059315800666809,
|
| 2328 |
+
"step": 330
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 0.2715897435897436,
|
| 2332 |
+
"grad_norm": 1.6875,
|
| 2333 |
+
"learning_rate": 1.042977113388761e-05,
|
| 2334 |
+
"loss": 1.0442448854446411,
|
| 2335 |
+
"step": 331
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 0.2724102564102564,
|
| 2339 |
+
"grad_norm": 1.65625,
|
| 2340 |
+
"learning_rate": 1.0417918679714543e-05,
|
| 2341 |
+
"loss": 1.0919098854064941,
|
| 2342 |
+
"step": 332
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 0.2732307692307692,
|
| 2346 |
+
"grad_norm": 1.4765625,
|
| 2347 |
+
"learning_rate": 1.0406037460404507e-05,
|
| 2348 |
+
"loss": 1.0103938579559326,
|
| 2349 |
+
"step": 333
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 0.27405128205128204,
|
| 2353 |
+
"grad_norm": 1.75,
|
| 2354 |
+
"learning_rate": 1.039412755698441e-05,
|
| 2355 |
+
"loss": 1.0885441303253174,
|
| 2356 |
+
"step": 334
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 0.27487179487179486,
|
| 2360 |
+
"grad_norm": 1.6875,
|
| 2361 |
+
"learning_rate": 1.0382189050676774e-05,
|
| 2362 |
+
"loss": 1.197743535041809,
|
| 2363 |
+
"step": 335
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 0.2756923076923077,
|
| 2367 |
+
"grad_norm": 1.5390625,
|
| 2368 |
+
"learning_rate": 1.0370222022899191e-05,
|
| 2369 |
+
"loss": 1.161704659461975,
|
| 2370 |
+
"step": 336
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 0.2765128205128205,
|
| 2374 |
+
"grad_norm": 1.59375,
|
| 2375 |
+
"learning_rate": 1.0358226555263758e-05,
|
| 2376 |
+
"loss": 1.1241810321807861,
|
| 2377 |
+
"step": 337
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 0.2773333333333333,
|
| 2381 |
+
"grad_norm": 1.671875,
|
| 2382 |
+
"learning_rate": 1.0346202729576523e-05,
|
| 2383 |
+
"loss": 1.2113137245178223,
|
| 2384 |
+
"step": 338
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 0.27815384615384614,
|
| 2388 |
+
"grad_norm": 1.5625,
|
| 2389 |
+
"learning_rate": 1.0334150627836934e-05,
|
| 2390 |
+
"loss": 1.12404465675354,
|
| 2391 |
+
"step": 339
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 0.27897435897435896,
|
| 2395 |
+
"grad_norm": 1.515625,
|
| 2396 |
+
"learning_rate": 1.0322070332237274e-05,
|
| 2397 |
+
"loss": 1.0594943761825562,
|
| 2398 |
+
"step": 340
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 0.2797948717948718,
|
| 2402 |
+
"grad_norm": 1.796875,
|
| 2403 |
+
"learning_rate": 1.0309961925162095e-05,
|
| 2404 |
+
"loss": 1.2125948667526245,
|
| 2405 |
+
"step": 341
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 0.2806153846153846,
|
| 2409 |
+
"grad_norm": 1.6328125,
|
| 2410 |
+
"learning_rate": 1.029782548918767e-05,
|
| 2411 |
+
"loss": 1.217134714126587,
|
| 2412 |
+
"step": 342
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 0.2814358974358974,
|
| 2416 |
+
"grad_norm": 1.5625,
|
| 2417 |
+
"learning_rate": 1.0285661107081413e-05,
|
| 2418 |
+
"loss": 1.0295854806900024,
|
| 2419 |
+
"step": 343
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 0.28225641025641024,
|
| 2423 |
+
"grad_norm": 1.6015625,
|
| 2424 |
+
"learning_rate": 1.0273468861801333e-05,
|
| 2425 |
+
"loss": 1.1883779764175415,
|
| 2426 |
+
"step": 344
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 0.28307692307692306,
|
| 2430 |
+
"grad_norm": 1.6328125,
|
| 2431 |
+
"learning_rate": 1.0261248836495455e-05,
|
| 2432 |
+
"loss": 1.1626862287521362,
|
| 2433 |
+
"step": 345
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 0.2838974358974359,
|
| 2437 |
+
"grad_norm": 1.7421875,
|
| 2438 |
+
"learning_rate": 1.024900111450125e-05,
|
| 2439 |
+
"loss": 1.1395612955093384,
|
| 2440 |
+
"step": 346
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 0.2847179487179487,
|
| 2444 |
+
"grad_norm": 1.7734375,
|
| 2445 |
+
"learning_rate": 1.0236725779345086e-05,
|
| 2446 |
+
"loss": 1.1365091800689697,
|
| 2447 |
+
"step": 347
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 0.2855384615384615,
|
| 2451 |
+
"grad_norm": 1.6796875,
|
| 2452 |
+
"learning_rate": 1.0224422914741637e-05,
|
| 2453 |
+
"loss": 1.2354658842086792,
|
| 2454 |
+
"step": 348
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 0.28635897435897434,
|
| 2458 |
+
"grad_norm": 1.6484375,
|
| 2459 |
+
"learning_rate": 1.0212092604593325e-05,
|
| 2460 |
+
"loss": 1.1859526634216309,
|
| 2461 |
+
"step": 349
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 0.28717948717948716,
|
| 2465 |
+
"grad_norm": 1.671875,
|
| 2466 |
+
"learning_rate": 1.0199734932989737e-05,
|
| 2467 |
+
"loss": 1.167510986328125,
|
| 2468 |
+
"step": 350
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 0.288,
|
| 2472 |
+
"grad_norm": 1.921875,
|
| 2473 |
+
"learning_rate": 1.0187349984207067e-05,
|
| 2474 |
+
"loss": 1.1553354263305664,
|
| 2475 |
+
"step": 351
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 0.2888205128205128,
|
| 2479 |
+
"grad_norm": 1.6015625,
|
| 2480 |
+
"learning_rate": 1.017493784270753e-05,
|
| 2481 |
+
"loss": 1.0852011442184448,
|
| 2482 |
+
"step": 352
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 0.2896410256410256,
|
| 2486 |
+
"grad_norm": 1.4921875,
|
| 2487 |
+
"learning_rate": 1.0162498593138783e-05,
|
| 2488 |
+
"loss": 1.1114470958709717,
|
| 2489 |
+
"step": 353
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 0.29046153846153844,
|
| 2493 |
+
"grad_norm": 1.7265625,
|
| 2494 |
+
"learning_rate": 1.0150032320333358e-05,
|
| 2495 |
+
"loss": 1.1002390384674072,
|
| 2496 |
+
"step": 354
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 0.29128205128205126,
|
| 2500 |
+
"grad_norm": 1.921875,
|
| 2501 |
+
"learning_rate": 1.0137539109308078e-05,
|
| 2502 |
+
"loss": 1.1473479270935059,
|
| 2503 |
+
"step": 355
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 0.2921025641025641,
|
| 2507 |
+
"grad_norm": 1.5859375,
|
| 2508 |
+
"learning_rate": 1.0125019045263481e-05,
|
| 2509 |
+
"loss": 1.1331301927566528,
|
| 2510 |
+
"step": 356
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 0.2929230769230769,
|
| 2514 |
+
"grad_norm": 1.796875,
|
| 2515 |
+
"learning_rate": 1.0112472213583227e-05,
|
| 2516 |
+
"loss": 1.167328953742981,
|
| 2517 |
+
"step": 357
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 0.2937435897435897,
|
| 2521 |
+
"grad_norm": 1.5703125,
|
| 2522 |
+
"learning_rate": 1.0099898699833535e-05,
|
| 2523 |
+
"loss": 1.0663187503814697,
|
| 2524 |
+
"step": 358
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 0.29456410256410254,
|
| 2528 |
+
"grad_norm": 1.6328125,
|
| 2529 |
+
"learning_rate": 1.0087298589762579e-05,
|
| 2530 |
+
"loss": 1.0864899158477783,
|
| 2531 |
+
"step": 359
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 0.2953846153846154,
|
| 2535 |
+
"grad_norm": 1.65625,
|
| 2536 |
+
"learning_rate": 1.0074671969299918e-05,
|
| 2537 |
+
"loss": 1.1541707515716553,
|
| 2538 |
+
"step": 360
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 0.29620512820512823,
|
| 2542 |
+
"grad_norm": 1.6171875,
|
| 2543 |
+
"learning_rate": 1.006201892455591e-05,
|
| 2544 |
+
"loss": 1.1286821365356445,
|
| 2545 |
+
"step": 361
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 0.29702564102564105,
|
| 2549 |
+
"grad_norm": 1.703125,
|
| 2550 |
+
"learning_rate": 1.0049339541821114e-05,
|
| 2551 |
+
"loss": 1.111492395401001,
|
| 2552 |
+
"step": 362
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 0.29784615384615387,
|
| 2556 |
+
"grad_norm": 1.6875,
|
| 2557 |
+
"learning_rate": 1.0036633907565703e-05,
|
| 2558 |
+
"loss": 1.1827677488327026,
|
| 2559 |
+
"step": 363
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 0.2986666666666667,
|
| 2563 |
+
"grad_norm": 1.75,
|
| 2564 |
+
"learning_rate": 1.002390210843889e-05,
|
| 2565 |
+
"loss": 1.2569875717163086,
|
| 2566 |
+
"step": 364
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 0.2994871794871795,
|
| 2570 |
+
"grad_norm": 1.71875,
|
| 2571 |
+
"learning_rate": 1.0011144231268316e-05,
|
| 2572 |
+
"loss": 1.0503405332565308,
|
| 2573 |
+
"step": 365
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 0.30030769230769233,
|
| 2577 |
+
"grad_norm": 1.5078125,
|
| 2578 |
+
"learning_rate": 9.998360363059478e-06,
|
| 2579 |
+
"loss": 1.0804475545883179,
|
| 2580 |
+
"step": 366
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 0.30112820512820515,
|
| 2584 |
+
"grad_norm": 1.53125,
|
| 2585 |
+
"learning_rate": 9.98555059099511e-06,
|
| 2586 |
+
"loss": 1.031091332435608,
|
| 2587 |
+
"step": 367
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 0.30194871794871797,
|
| 2591 |
+
"grad_norm": 1.625,
|
| 2592 |
+
"learning_rate": 9.972715002434621e-06,
|
| 2593 |
+
"loss": 1.2711988687515259,
|
| 2594 |
+
"step": 368
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 0.3027692307692308,
|
| 2598 |
+
"grad_norm": 1.5234375,
|
| 2599 |
+
"learning_rate": 9.959853684913468e-06,
|
| 2600 |
+
"loss": 1.1220529079437256,
|
| 2601 |
+
"step": 369
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 0.3035897435897436,
|
| 2605 |
+
"grad_norm": 1.7890625,
|
| 2606 |
+
"learning_rate": 9.946966726142579e-06,
|
| 2607 |
+
"loss": 1.1590451002120972,
|
| 2608 |
+
"step": 370
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 0.30441025641025643,
|
| 2612 |
+
"grad_norm": 1.78125,
|
| 2613 |
+
"learning_rate": 9.934054214007753e-06,
|
| 2614 |
+
"loss": 1.1940088272094727,
|
| 2615 |
+
"step": 371
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 0.30523076923076925,
|
| 2619 |
+
"grad_norm": 1.6640625,
|
| 2620 |
+
"learning_rate": 9.921116236569046e-06,
|
| 2621 |
+
"loss": 1.1365522146224976,
|
| 2622 |
+
"step": 372
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 0.30605128205128207,
|
| 2626 |
+
"grad_norm": 1.578125,
|
| 2627 |
+
"learning_rate": 9.908152882060194e-06,
|
| 2628 |
+
"loss": 1.2056771516799927,
|
| 2629 |
+
"step": 373
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 0.3068717948717949,
|
| 2633 |
+
"grad_norm": 1.5390625,
|
| 2634 |
+
"learning_rate": 9.895164238887982e-06,
|
| 2635 |
+
"loss": 1.111759066581726,
|
| 2636 |
+
"step": 374
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 0.3076923076923077,
|
| 2640 |
+
"grad_norm": 1.5078125,
|
| 2641 |
+
"learning_rate": 9.882150395631673e-06,
|
| 2642 |
+
"loss": 1.0607143640518188,
|
| 2643 |
+
"step": 375
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 0.30851282051282053,
|
| 2647 |
+
"grad_norm": 1.6328125,
|
| 2648 |
+
"learning_rate": 9.86911144104238e-06,
|
| 2649 |
+
"loss": 1.0898137092590332,
|
| 2650 |
+
"step": 376
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 0.30933333333333335,
|
| 2654 |
+
"grad_norm": 1.7890625,
|
| 2655 |
+
"learning_rate": 9.85604746404247e-06,
|
| 2656 |
+
"loss": 1.0215013027191162,
|
| 2657 |
+
"step": 377
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 0.31015384615384617,
|
| 2661 |
+
"grad_norm": 1.78125,
|
| 2662 |
+
"learning_rate": 9.842958553724959e-06,
|
| 2663 |
+
"loss": 1.0680979490280151,
|
| 2664 |
+
"step": 378
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 0.310974358974359,
|
| 2668 |
+
"grad_norm": 1.7578125,
|
| 2669 |
+
"learning_rate": 9.829844799352896e-06,
|
| 2670 |
+
"loss": 1.2479051351547241,
|
| 2671 |
+
"step": 379
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 0.3117948717948718,
|
| 2675 |
+
"grad_norm": 1.53125,
|
| 2676 |
+
"learning_rate": 9.816706290358764e-06,
|
| 2677 |
+
"loss": 1.180200457572937,
|
| 2678 |
+
"step": 380
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 0.31261538461538463,
|
| 2682 |
+
"grad_norm": 1.515625,
|
| 2683 |
+
"learning_rate": 9.80354311634387e-06,
|
| 2684 |
+
"loss": 1.0876795053482056,
|
| 2685 |
+
"step": 381
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 0.31343589743589745,
|
| 2689 |
+
"grad_norm": 1.609375,
|
| 2690 |
+
"learning_rate": 9.790355367077724e-06,
|
| 2691 |
+
"loss": 1.1074776649475098,
|
| 2692 |
+
"step": 382
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 0.31425641025641027,
|
| 2696 |
+
"grad_norm": 1.578125,
|
| 2697 |
+
"learning_rate": 9.777143132497436e-06,
|
| 2698 |
+
"loss": 1.143994688987732,
|
| 2699 |
+
"step": 383
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 0.3150769230769231,
|
| 2703 |
+
"grad_norm": 1.6796875,
|
| 2704 |
+
"learning_rate": 9.7639065027071e-06,
|
| 2705 |
+
"loss": 1.1703780889511108,
|
| 2706 |
+
"step": 384
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 0.3158974358974359,
|
| 2710 |
+
"grad_norm": 1.7265625,
|
| 2711 |
+
"learning_rate": 9.750645567977178e-06,
|
| 2712 |
+
"loss": 1.1306742429733276,
|
| 2713 |
+
"step": 385
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 0.3167179487179487,
|
| 2717 |
+
"grad_norm": 1.5625,
|
| 2718 |
+
"learning_rate": 9.737360418743883e-06,
|
| 2719 |
+
"loss": 1.089449167251587,
|
| 2720 |
+
"step": 386
|
| 2721 |
+
},
|
| 2722 |
+
{
|
| 2723 |
+
"epoch": 0.31753846153846155,
|
| 2724 |
+
"grad_norm": 1.609375,
|
| 2725 |
+
"learning_rate": 9.724051145608572e-06,
|
| 2726 |
+
"loss": 1.0401030778884888,
|
| 2727 |
+
"step": 387
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 0.31835897435897437,
|
| 2731 |
+
"grad_norm": 1.7265625,
|
| 2732 |
+
"learning_rate": 9.710717839337114e-06,
|
| 2733 |
+
"loss": 1.036818265914917,
|
| 2734 |
+
"step": 388
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 0.3191794871794872,
|
| 2738 |
+
"grad_norm": 1.578125,
|
| 2739 |
+
"learning_rate": 9.69736059085928e-06,
|
| 2740 |
+
"loss": 1.0188218355178833,
|
| 2741 |
+
"step": 389
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 0.32,
|
| 2745 |
+
"grad_norm": 1.5625,
|
| 2746 |
+
"learning_rate": 9.68397949126812e-06,
|
| 2747 |
+
"loss": 1.0937597751617432,
|
| 2748 |
+
"step": 390
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 0.3208205128205128,
|
| 2752 |
+
"grad_norm": 1.625,
|
| 2753 |
+
"learning_rate": 9.670574631819347e-06,
|
| 2754 |
+
"loss": 1.1591130495071411,
|
| 2755 |
+
"step": 391
|
| 2756 |
+
},
|
| 2757 |
+
{
|
| 2758 |
+
"epoch": 0.32164102564102565,
|
| 2759 |
+
"grad_norm": 1.6328125,
|
| 2760 |
+
"learning_rate": 9.657146103930706e-06,
|
| 2761 |
+
"loss": 1.1604453325271606,
|
| 2762 |
+
"step": 392
|
| 2763 |
+
},
|
| 2764 |
+
{
|
| 2765 |
+
"epoch": 0.32246153846153847,
|
| 2766 |
+
"grad_norm": 1.8203125,
|
| 2767 |
+
"learning_rate": 9.643693999181355e-06,
|
| 2768 |
+
"loss": 1.242227554321289,
|
| 2769 |
+
"step": 393
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 0.3232820512820513,
|
| 2773 |
+
"grad_norm": 1.59375,
|
| 2774 |
+
"learning_rate": 9.630218409311239e-06,
|
| 2775 |
+
"loss": 1.1561675071716309,
|
| 2776 |
+
"step": 394
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 0.3241025641025641,
|
| 2780 |
+
"grad_norm": 1.6875,
|
| 2781 |
+
"learning_rate": 9.61671942622047e-06,
|
| 2782 |
+
"loss": 1.1183706521987915,
|
| 2783 |
+
"step": 395
|
| 2784 |
+
},
|
| 2785 |
+
{
|
| 2786 |
+
"epoch": 0.3249230769230769,
|
| 2787 |
+
"grad_norm": 1.625,
|
| 2788 |
+
"learning_rate": 9.603197141968693e-06,
|
| 2789 |
+
"loss": 1.163654088973999,
|
| 2790 |
+
"step": 396
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 0.32574358974358975,
|
| 2794 |
+
"grad_norm": 1.640625,
|
| 2795 |
+
"learning_rate": 9.589651648774459e-06,
|
| 2796 |
+
"loss": 1.1973570585250854,
|
| 2797 |
+
"step": 397
|
| 2798 |
+
},
|
| 2799 |
+
{
|
| 2800 |
+
"epoch": 0.32656410256410257,
|
| 2801 |
+
"grad_norm": 1.5546875,
|
| 2802 |
+
"learning_rate": 9.576083039014605e-06,
|
| 2803 |
+
"loss": 1.1929141283035278,
|
| 2804 |
+
"step": 398
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 0.3273846153846154,
|
| 2808 |
+
"grad_norm": 1.59375,
|
| 2809 |
+
"learning_rate": 9.562491405223609e-06,
|
| 2810 |
+
"loss": 1.0391355752944946,
|
| 2811 |
+
"step": 399
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 0.3282051282051282,
|
| 2815 |
+
"grad_norm": 1.6328125,
|
| 2816 |
+
"learning_rate": 9.54887684009297e-06,
|
| 2817 |
+
"loss": 1.1033200025558472,
|
| 2818 |
+
"step": 400
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 0.329025641025641,
|
| 2822 |
+
"grad_norm": 1.6953125,
|
| 2823 |
+
"learning_rate": 9.535239436470579e-06,
|
| 2824 |
+
"loss": 1.096189022064209,
|
| 2825 |
+
"step": 401
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 0.32984615384615384,
|
| 2829 |
+
"grad_norm": 1.7734375,
|
| 2830 |
+
"learning_rate": 9.521579287360066e-06,
|
| 2831 |
+
"loss": 1.1212831735610962,
|
| 2832 |
+
"step": 402
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 0.33066666666666666,
|
| 2836 |
+
"grad_norm": 1.78125,
|
| 2837 |
+
"learning_rate": 9.507896485920193e-06,
|
| 2838 |
+
"loss": 1.1692836284637451,
|
| 2839 |
+
"step": 403
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 0.3314871794871795,
|
| 2843 |
+
"grad_norm": 1.5,
|
| 2844 |
+
"learning_rate": 9.494191125464199e-06,
|
| 2845 |
+
"loss": 1.1815142631530762,
|
| 2846 |
+
"step": 404
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 0.3323076923076923,
|
| 2850 |
+
"grad_norm": 1.625,
|
| 2851 |
+
"learning_rate": 9.48046329945917e-06,
|
| 2852 |
+
"loss": 1.073824167251587,
|
| 2853 |
+
"step": 405
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 0.3331282051282051,
|
| 2857 |
+
"grad_norm": 1.5234375,
|
| 2858 |
+
"learning_rate": 9.466713101525404e-06,
|
| 2859 |
+
"loss": 1.1135120391845703,
|
| 2860 |
+
"step": 406
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 0.33394871794871794,
|
| 2864 |
+
"grad_norm": 1.5625,
|
| 2865 |
+
"learning_rate": 9.452940625435764e-06,
|
| 2866 |
+
"loss": 1.0316897630691528,
|
| 2867 |
+
"step": 407
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 0.33476923076923076,
|
| 2871 |
+
"grad_norm": 1.6328125,
|
| 2872 |
+
"learning_rate": 9.439145965115051e-06,
|
| 2873 |
+
"loss": 1.1632769107818604,
|
| 2874 |
+
"step": 408
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 0.3355897435897436,
|
| 2878 |
+
"grad_norm": 1.5859375,
|
| 2879 |
+
"learning_rate": 9.425329214639357e-06,
|
| 2880 |
+
"loss": 1.1414909362792969,
|
| 2881 |
+
"step": 409
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 0.3364102564102564,
|
| 2885 |
+
"grad_norm": 1.75,
|
| 2886 |
+
"learning_rate": 9.411490468235414e-06,
|
| 2887 |
+
"loss": 1.0593255758285522,
|
| 2888 |
+
"step": 410
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 0.3372307692307692,
|
| 2892 |
+
"grad_norm": 1.578125,
|
| 2893 |
+
"learning_rate": 9.39762982027997e-06,
|
| 2894 |
+
"loss": 1.1819946765899658,
|
| 2895 |
+
"step": 411
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"epoch": 0.33805128205128204,
|
| 2899 |
+
"grad_norm": 1.59375,
|
| 2900 |
+
"learning_rate": 9.383747365299133e-06,
|
| 2901 |
+
"loss": 1.0364710092544556,
|
| 2902 |
+
"step": 412
|
| 2903 |
+
},
|
| 2904 |
+
{
|
| 2905 |
+
"epoch": 0.33887179487179486,
|
| 2906 |
+
"grad_norm": 1.671875,
|
| 2907 |
+
"learning_rate": 9.369843197967733e-06,
|
| 2908 |
+
"loss": 1.0772686004638672,
|
| 2909 |
+
"step": 413
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 0.3396923076923077,
|
| 2913 |
+
"grad_norm": 1.71875,
|
| 2914 |
+
"learning_rate": 9.355917413108662e-06,
|
| 2915 |
+
"loss": 1.1532864570617676,
|
| 2916 |
+
"step": 414
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 0.3405128205128205,
|
| 2920 |
+
"grad_norm": 1.7734375,
|
| 2921 |
+
"learning_rate": 9.34197010569225e-06,
|
| 2922 |
+
"loss": 1.1974714994430542,
|
| 2923 |
+
"step": 415
|
| 2924 |
+
},
|
| 2925 |
+
{
|
| 2926 |
+
"epoch": 0.3413333333333333,
|
| 2927 |
+
"grad_norm": 1.71875,
|
| 2928 |
+
"learning_rate": 9.3280013708356e-06,
|
| 2929 |
+
"loss": 1.1706582307815552,
|
| 2930 |
+
"step": 416
|
| 2931 |
+
},
|
| 2932 |
+
{
|
| 2933 |
+
"epoch": 0.34215384615384614,
|
| 2934 |
+
"grad_norm": 1.59375,
|
| 2935 |
+
"learning_rate": 9.314011303801946e-06,
|
| 2936 |
+
"loss": 1.1343363523483276,
|
| 2937 |
+
"step": 417
|
| 2938 |
+
},
|
| 2939 |
+
{
|
| 2940 |
+
"epoch": 0.34297435897435896,
|
| 2941 |
+
"grad_norm": 1.5859375,
|
| 2942 |
+
"learning_rate": 9.299999999999999e-06,
|
| 2943 |
+
"loss": 1.2338825464248657,
|
| 2944 |
+
"step": 418
|
| 2945 |
+
},
|
| 2946 |
+
{
|
| 2947 |
+
"epoch": 0.3437948717948718,
|
| 2948 |
+
"grad_norm": 1.671875,
|
| 2949 |
+
"learning_rate": 9.285967554983304e-06,
|
| 2950 |
+
"loss": 1.112164855003357,
|
| 2951 |
+
"step": 419
|
| 2952 |
+
},
|
| 2953 |
+
{
|
| 2954 |
+
"epoch": 0.3446153846153846,
|
| 2955 |
+
"grad_norm": 1.578125,
|
| 2956 |
+
"learning_rate": 9.271914064449582e-06,
|
| 2957 |
+
"loss": 1.041966438293457,
|
| 2958 |
+
"step": 420
|
| 2959 |
+
},
|
| 2960 |
+
{
|
| 2961 |
+
"epoch": 0.3454358974358974,
|
| 2962 |
+
"grad_norm": 1.6640625,
|
| 2963 |
+
"learning_rate": 9.257839624240077e-06,
|
| 2964 |
+
"loss": 1.0895839929580688,
|
| 2965 |
+
"step": 421
|
| 2966 |
+
},
|
| 2967 |
+
{
|
| 2968 |
+
"epoch": 0.34625641025641024,
|
| 2969 |
+
"grad_norm": 1.8515625,
|
| 2970 |
+
"learning_rate": 9.243744330338904e-06,
|
| 2971 |
+
"loss": 1.0958746671676636,
|
| 2972 |
+
"step": 422
|
| 2973 |
+
},
|
| 2974 |
+
{
|
| 2975 |
+
"epoch": 0.34707692307692306,
|
| 2976 |
+
"grad_norm": 1.640625,
|
| 2977 |
+
"learning_rate": 9.2296282788724e-06,
|
| 2978 |
+
"loss": 1.0720858573913574,
|
| 2979 |
+
"step": 423
|
| 2980 |
+
},
|
| 2981 |
+
{
|
| 2982 |
+
"epoch": 0.3478974358974359,
|
| 2983 |
+
"grad_norm": 1.6328125,
|
| 2984 |
+
"learning_rate": 9.21549156610846e-06,
|
| 2985 |
+
"loss": 1.1186761856079102,
|
| 2986 |
+
"step": 424
|
| 2987 |
+
},
|
| 2988 |
+
{
|
| 2989 |
+
"epoch": 0.3487179487179487,
|
| 2990 |
+
"grad_norm": 1.609375,
|
| 2991 |
+
"learning_rate": 9.201334288455883e-06,
|
| 2992 |
+
"loss": 1.1753547191619873,
|
| 2993 |
+
"step": 425
|
| 2994 |
+
},
|
| 2995 |
+
{
|
| 2996 |
+
"epoch": 0.3495384615384615,
|
| 2997 |
+
"grad_norm": 1.75,
|
| 2998 |
+
"learning_rate": 9.18715654246372e-06,
|
| 2999 |
+
"loss": 1.1445863246917725,
|
| 3000 |
+
"step": 426
|
| 3001 |
+
},
|
| 3002 |
+
{
|
| 3003 |
+
"epoch": 0.35035897435897434,
|
| 3004 |
+
"grad_norm": 1.8046875,
|
| 3005 |
+
"learning_rate": 9.1729584248206e-06,
|
| 3006 |
+
"loss": 1.0882112979888916,
|
| 3007 |
+
"step": 427
|
| 3008 |
+
},
|
| 3009 |
+
{
|
| 3010 |
+
"epoch": 0.35117948717948716,
|
| 3011 |
+
"grad_norm": 1.7109375,
|
| 3012 |
+
"learning_rate": 9.1587400323541e-06,
|
| 3013 |
+
"loss": 1.1449003219604492,
|
| 3014 |
+
"step": 428
|
| 3015 |
+
},
|
| 3016 |
+
{
|
| 3017 |
+
"epoch": 0.352,
|
| 3018 |
+
"grad_norm": 1.625,
|
| 3019 |
+
"learning_rate": 9.144501462030042e-06,
|
| 3020 |
+
"loss": 1.17095947265625,
|
| 3021 |
+
"step": 429
|
| 3022 |
+
},
|
| 3023 |
+
{
|
| 3024 |
+
"epoch": 0.3528205128205128,
|
| 3025 |
+
"grad_norm": 1.84375,
|
| 3026 |
+
"learning_rate": 9.130242810951877e-06,
|
| 3027 |
+
"loss": 1.0751999616622925,
|
| 3028 |
+
"step": 430
|
| 3029 |
+
},
|
| 3030 |
+
{
|
| 3031 |
+
"epoch": 0.3536410256410256,
|
| 3032 |
+
"grad_norm": 1.6484375,
|
| 3033 |
+
"learning_rate": 9.115964176359993e-06,
|
| 3034 |
+
"loss": 1.1300032138824463,
|
| 3035 |
+
"step": 431
|
| 3036 |
+
},
|
| 3037 |
+
{
|
| 3038 |
+
"epoch": 0.35446153846153844,
|
| 3039 |
+
"grad_norm": 1.7421875,
|
| 3040 |
+
"learning_rate": 9.101665655631059e-06,
|
| 3041 |
+
"loss": 1.1875977516174316,
|
| 3042 |
+
"step": 432
|
| 3043 |
+
},
|
| 3044 |
+
{
|
| 3045 |
+
"epoch": 0.35528205128205126,
|
| 3046 |
+
"grad_norm": 1.65625,
|
| 3047 |
+
"learning_rate": 9.087347346277366e-06,
|
| 3048 |
+
"loss": 1.0804345607757568,
|
| 3049 |
+
"step": 433
|
| 3050 |
+
},
|
| 3051 |
+
{
|
| 3052 |
+
"epoch": 0.3561025641025641,
|
| 3053 |
+
"grad_norm": 1.6484375,
|
| 3054 |
+
"learning_rate": 9.073009345946156e-06,
|
| 3055 |
+
"loss": 1.2072385549545288,
|
| 3056 |
+
"step": 434
|
| 3057 |
+
},
|
| 3058 |
+
{
|
| 3059 |
+
"epoch": 0.3569230769230769,
|
| 3060 |
+
"grad_norm": 1.6171875,
|
| 3061 |
+
"learning_rate": 9.058651752418958e-06,
|
| 3062 |
+
"loss": 1.1018540859222412,
|
| 3063 |
+
"step": 435
|
| 3064 |
+
},
|
| 3065 |
+
{
|
| 3066 |
+
"epoch": 0.3577435897435897,
|
| 3067 |
+
"grad_norm": 1.578125,
|
| 3068 |
+
"learning_rate": 9.044274663610928e-06,
|
| 3069 |
+
"loss": 1.068818211555481,
|
| 3070 |
+
"step": 436
|
| 3071 |
+
},
|
| 3072 |
+
{
|
| 3073 |
+
"epoch": 0.35856410256410254,
|
| 3074 |
+
"grad_norm": 1.78125,
|
| 3075 |
+
"learning_rate": 9.029878177570164e-06,
|
| 3076 |
+
"loss": 1.1553325653076172,
|
| 3077 |
+
"step": 437
|
| 3078 |
+
},
|
| 3079 |
+
{
|
| 3080 |
+
"epoch": 0.35938461538461536,
|
| 3081 |
+
"grad_norm": 1.71875,
|
| 3082 |
+
"learning_rate": 9.015462392477058e-06,
|
| 3083 |
+
"loss": 1.1629064083099365,
|
| 3084 |
+
"step": 438
|
| 3085 |
+
},
|
| 3086 |
+
{
|
| 3087 |
+
"epoch": 0.3602051282051282,
|
| 3088 |
+
"grad_norm": 1.6171875,
|
| 3089 |
+
"learning_rate": 9.001027406643609e-06,
|
| 3090 |
+
"loss": 1.1597505807876587,
|
| 3091 |
+
"step": 439
|
| 3092 |
+
},
|
| 3093 |
+
{
|
| 3094 |
+
"epoch": 0.36102564102564105,
|
| 3095 |
+
"grad_norm": 1.5625,
|
| 3096 |
+
"learning_rate": 8.986573318512766e-06,
|
| 3097 |
+
"loss": 1.1568797826766968,
|
| 3098 |
+
"step": 440
|
| 3099 |
+
},
|
| 3100 |
+
{
|
| 3101 |
+
"epoch": 0.3618461538461539,
|
| 3102 |
+
"grad_norm": 1.734375,
|
| 3103 |
+
"learning_rate": 8.972100226657751e-06,
|
| 3104 |
+
"loss": 1.179506540298462,
|
| 3105 |
+
"step": 441
|
| 3106 |
+
},
|
| 3107 |
+
{
|
| 3108 |
+
"epoch": 0.3626666666666667,
|
| 3109 |
+
"grad_norm": 1.7265625,
|
| 3110 |
+
"learning_rate": 8.957608229781379e-06,
|
| 3111 |
+
"loss": 1.068928599357605,
|
| 3112 |
+
"step": 442
|
| 3113 |
+
},
|
| 3114 |
+
{
|
| 3115 |
+
"epoch": 0.3634871794871795,
|
| 3116 |
+
"grad_norm": 1.6484375,
|
| 3117 |
+
"learning_rate": 8.943097426715405e-06,
|
| 3118 |
+
"loss": 1.1226983070373535,
|
| 3119 |
+
"step": 443
|
| 3120 |
+
},
|
| 3121 |
+
{
|
| 3122 |
+
"epoch": 0.36430769230769233,
|
| 3123 |
+
"grad_norm": 1.8203125,
|
| 3124 |
+
"learning_rate": 8.928567916419824e-06,
|
| 3125 |
+
"loss": 1.155146598815918,
|
| 3126 |
+
"step": 444
|
| 3127 |
+
},
|
| 3128 |
+
{
|
| 3129 |
+
"epoch": 0.36512820512820515,
|
| 3130 |
+
"grad_norm": 1.5703125,
|
| 3131 |
+
"learning_rate": 8.914019797982218e-06,
|
| 3132 |
+
"loss": 1.0993762016296387,
|
| 3133 |
+
"step": 445
|
| 3134 |
+
},
|
| 3135 |
+
{
|
| 3136 |
+
"epoch": 0.36594871794871797,
|
| 3137 |
+
"grad_norm": 1.6015625,
|
| 3138 |
+
"learning_rate": 8.89945317061707e-06,
|
| 3139 |
+
"loss": 1.065480351448059,
|
| 3140 |
+
"step": 446
|
| 3141 |
+
},
|
| 3142 |
+
{
|
| 3143 |
+
"epoch": 0.3667692307692308,
|
| 3144 |
+
"grad_norm": 1.53125,
|
| 3145 |
+
"learning_rate": 8.884868133665088e-06,
|
| 3146 |
+
"loss": 1.1303458213806152,
|
| 3147 |
+
"step": 447
|
| 3148 |
+
},
|
| 3149 |
+
{
|
| 3150 |
+
"epoch": 0.3675897435897436,
|
| 3151 |
+
"grad_norm": 1.6953125,
|
| 3152 |
+
"learning_rate": 8.870264786592529e-06,
|
| 3153 |
+
"loss": 1.1668941974639893,
|
| 3154 |
+
"step": 448
|
| 3155 |
+
},
|
| 3156 |
+
{
|
| 3157 |
+
"epoch": 0.36841025641025643,
|
| 3158 |
+
"grad_norm": 1.6953125,
|
| 3159 |
+
"learning_rate": 8.855643228990524e-06,
|
| 3160 |
+
"loss": 1.083670973777771,
|
| 3161 |
+
"step": 449
|
| 3162 |
+
},
|
| 3163 |
+
{
|
| 3164 |
+
"epoch": 0.36923076923076925,
|
| 3165 |
+
"grad_norm": 1.7578125,
|
| 3166 |
+
"learning_rate": 8.841003560574386e-06,
|
| 3167 |
+
"loss": 1.178176999092102,
|
| 3168 |
+
"step": 450
|
| 3169 |
+
},
|
| 3170 |
+
{
|
| 3171 |
+
"epoch": 0.37005128205128207,
|
| 3172 |
+
"grad_norm": 1.703125,
|
| 3173 |
+
"learning_rate": 8.826345881182953e-06,
|
| 3174 |
+
"loss": 1.05520761013031,
|
| 3175 |
+
"step": 451
|
| 3176 |
+
},
|
| 3177 |
+
{
|
| 3178 |
+
"epoch": 0.3708717948717949,
|
| 3179 |
+
"grad_norm": 1.7578125,
|
| 3180 |
+
"learning_rate": 8.81167029077788e-06,
|
| 3181 |
+
"loss": 1.076585054397583,
|
| 3182 |
+
"step": 452
|
| 3183 |
+
},
|
| 3184 |
+
{
|
| 3185 |
+
"epoch": 0.3716923076923077,
|
| 3186 |
+
"grad_norm": 1.65625,
|
| 3187 |
+
"learning_rate": 8.796976889442977e-06,
|
| 3188 |
+
"loss": 1.1256433725357056,
|
| 3189 |
+
"step": 453
|
| 3190 |
+
},
|
| 3191 |
+
{
|
| 3192 |
+
"epoch": 0.37251282051282053,
|
| 3193 |
+
"grad_norm": 1.7109375,
|
| 3194 |
+
"learning_rate": 8.78226577738352e-06,
|
| 3195 |
+
"loss": 1.1826629638671875,
|
| 3196 |
+
"step": 454
|
| 3197 |
+
},
|
| 3198 |
+
{
|
| 3199 |
+
"epoch": 0.37333333333333335,
|
| 3200 |
+
"grad_norm": 1.6640625,
|
| 3201 |
+
"learning_rate": 8.76753705492557e-06,
|
| 3202 |
+
"loss": 1.131862998008728,
|
| 3203 |
+
"step": 455
|
| 3204 |
+
},
|
| 3205 |
+
{
|
| 3206 |
+
"epoch": 0.37415384615384617,
|
| 3207 |
+
"grad_norm": 1.6953125,
|
| 3208 |
+
"learning_rate": 8.752790822515277e-06,
|
| 3209 |
+
"loss": 1.1642948389053345,
|
| 3210 |
+
"step": 456
|
| 3211 |
+
},
|
| 3212 |
+
{
|
| 3213 |
+
"epoch": 0.374974358974359,
|
| 3214 |
+
"grad_norm": 1.7734375,
|
| 3215 |
+
"learning_rate": 8.738027180718214e-06,
|
| 3216 |
+
"loss": 1.2190356254577637,
|
| 3217 |
+
"step": 457
|
| 3218 |
+
},
|
| 3219 |
+
{
|
| 3220 |
+
"epoch": 0.3757948717948718,
|
| 3221 |
+
"grad_norm": 1.7890625,
|
| 3222 |
+
"learning_rate": 8.72324623021868e-06,
|
| 3223 |
+
"loss": 1.0883055925369263,
|
| 3224 |
+
"step": 458
|
| 3225 |
+
},
|
| 3226 |
+
{
|
| 3227 |
+
"epoch": 0.37661538461538463,
|
| 3228 |
+
"grad_norm": 1.6171875,
|
| 3229 |
+
"learning_rate": 8.70844807181901e-06,
|
| 3230 |
+
"loss": 1.0619423389434814,
|
| 3231 |
+
"step": 459
|
| 3232 |
+
},
|
| 3233 |
+
{
|
| 3234 |
+
"epoch": 0.37743589743589745,
|
| 3235 |
+
"grad_norm": 1.6015625,
|
| 3236 |
+
"learning_rate": 8.693632806438897e-06,
|
| 3237 |
+
"loss": 1.0208218097686768,
|
| 3238 |
+
"step": 460
|
| 3239 |
+
},
|
| 3240 |
+
{
|
| 3241 |
+
"epoch": 0.37825641025641027,
|
| 3242 |
+
"grad_norm": 1.59375,
|
| 3243 |
+
"learning_rate": 8.6788005351147e-06,
|
| 3244 |
+
"loss": 1.0997884273529053,
|
| 3245 |
+
"step": 461
|
| 3246 |
+
},
|
| 3247 |
+
{
|
| 3248 |
+
"epoch": 0.3790769230769231,
|
| 3249 |
+
"grad_norm": 1.5,
|
| 3250 |
+
"learning_rate": 8.663951358998748e-06,
|
| 3251 |
+
"loss": 1.1279151439666748,
|
| 3252 |
+
"step": 462
|
| 3253 |
+
},
|
| 3254 |
+
{
|
| 3255 |
+
"epoch": 0.3798974358974359,
|
| 3256 |
+
"grad_norm": 1.6640625,
|
| 3257 |
+
"learning_rate": 8.649085379358666e-06,
|
| 3258 |
+
"loss": 1.0676960945129395,
|
| 3259 |
+
"step": 463
|
| 3260 |
+
},
|
| 3261 |
+
{
|
| 3262 |
+
"epoch": 0.38071794871794873,
|
| 3263 |
+
"grad_norm": 1.6484375,
|
| 3264 |
+
"learning_rate": 8.634202697576664e-06,
|
| 3265 |
+
"loss": 1.0952509641647339,
|
| 3266 |
+
"step": 464
|
| 3267 |
+
},
|
| 3268 |
+
{
|
| 3269 |
+
"epoch": 0.38153846153846155,
|
| 3270 |
+
"grad_norm": 1.71875,
|
| 3271 |
+
"learning_rate": 8.619303415148866e-06,
|
| 3272 |
+
"loss": 1.1405962705612183,
|
| 3273 |
+
"step": 465
|
| 3274 |
+
},
|
| 3275 |
+
{
|
| 3276 |
+
"epoch": 0.38235897435897437,
|
| 3277 |
+
"grad_norm": 1.734375,
|
| 3278 |
+
"learning_rate": 8.604387633684606e-06,
|
| 3279 |
+
"loss": 1.2375311851501465,
|
| 3280 |
+
"step": 466
|
| 3281 |
+
},
|
| 3282 |
+
{
|
| 3283 |
+
"epoch": 0.3831794871794872,
|
| 3284 |
+
"grad_norm": 1.734375,
|
| 3285 |
+
"learning_rate": 8.589455454905727e-06,
|
| 3286 |
+
"loss": 1.0690231323242188,
|
| 3287 |
+
"step": 467
|
| 3288 |
+
},
|
| 3289 |
+
{
|
| 3290 |
+
"epoch": 0.384,
|
| 3291 |
+
"grad_norm": 1.5625,
|
| 3292 |
+
"learning_rate": 8.574506980645914e-06,
|
| 3293 |
+
"loss": 1.092921257019043,
|
| 3294 |
+
"step": 468
|
| 3295 |
+
},
|
| 3296 |
+
{
|
| 3297 |
+
"epoch": 0.38482051282051283,
|
| 3298 |
+
"grad_norm": 1.671875,
|
| 3299 |
+
"learning_rate": 8.55954231284997e-06,
|
| 3300 |
+
"loss": 1.138890266418457,
|
| 3301 |
+
"step": 469
|
| 3302 |
+
},
|
| 3303 |
+
{
|
| 3304 |
+
"epoch": 0.38564102564102565,
|
| 3305 |
+
"grad_norm": 1.734375,
|
| 3306 |
+
"learning_rate": 8.544561553573142e-06,
|
| 3307 |
+
"loss": 1.206894874572754,
|
| 3308 |
+
"step": 470
|
| 3309 |
+
},
|
| 3310 |
+
{
|
| 3311 |
+
"epoch": 0.38646153846153847,
|
| 3312 |
+
"grad_norm": 1.71875,
|
| 3313 |
+
"learning_rate": 8.529564804980411e-06,
|
| 3314 |
+
"loss": 1.104117512702942,
|
| 3315 |
+
"step": 471
|
| 3316 |
+
},
|
| 3317 |
+
{
|
| 3318 |
+
"epoch": 0.3872820512820513,
|
| 3319 |
+
"grad_norm": 1.5859375,
|
| 3320 |
+
"learning_rate": 8.514552169345805e-06,
|
| 3321 |
+
"loss": 1.1774654388427734,
|
| 3322 |
+
"step": 472
|
| 3323 |
+
},
|
| 3324 |
+
{
|
| 3325 |
+
"epoch": 0.3881025641025641,
|
| 3326 |
+
"grad_norm": 1.7734375,
|
| 3327 |
+
"learning_rate": 8.499523749051696e-06,
|
| 3328 |
+
"loss": 1.15071702003479,
|
| 3329 |
+
"step": 473
|
| 3330 |
+
},
|
| 3331 |
+
{
|
| 3332 |
+
"epoch": 0.3889230769230769,
|
| 3333 |
+
"grad_norm": 1.8046875,
|
| 3334 |
+
"learning_rate": 8.484479646588105e-06,
|
| 3335 |
+
"loss": 1.151330828666687,
|
| 3336 |
+
"step": 474
|
| 3337 |
+
},
|
| 3338 |
+
{
|
| 3339 |
+
"epoch": 0.38974358974358975,
|
| 3340 |
+
"grad_norm": 1.65625,
|
| 3341 |
+
"learning_rate": 8.469419964551997e-06,
|
| 3342 |
+
"loss": 1.1878314018249512,
|
| 3343 |
+
"step": 475
|
| 3344 |
+
},
|
| 3345 |
+
{
|
| 3346 |
+
"epoch": 0.39056410256410257,
|
| 3347 |
+
"grad_norm": 1.7265625,
|
| 3348 |
+
"learning_rate": 8.45434480564659e-06,
|
| 3349 |
+
"loss": 1.2046477794647217,
|
| 3350 |
+
"step": 476
|
| 3351 |
+
},
|
| 3352 |
+
{
|
| 3353 |
+
"epoch": 0.3913846153846154,
|
| 3354 |
+
"grad_norm": 1.578125,
|
| 3355 |
+
"learning_rate": 8.439254272680651e-06,
|
| 3356 |
+
"loss": 1.1146059036254883,
|
| 3357 |
+
"step": 477
|
| 3358 |
+
},
|
| 3359 |
+
{
|
| 3360 |
+
"epoch": 0.3922051282051282,
|
| 3361 |
+
"grad_norm": 1.625,
|
| 3362 |
+
"learning_rate": 8.42414846856779e-06,
|
| 3363 |
+
"loss": 1.0920354127883911,
|
| 3364 |
+
"step": 478
|
| 3365 |
+
},
|
| 3366 |
+
{
|
| 3367 |
+
"epoch": 0.393025641025641,
|
| 3368 |
+
"grad_norm": 1.6953125,
|
| 3369 |
+
"learning_rate": 8.409027496325765e-06,
|
| 3370 |
+
"loss": 1.1231528520584106,
|
| 3371 |
+
"step": 479
|
| 3372 |
+
},
|
| 3373 |
+
{
|
| 3374 |
+
"epoch": 0.39384615384615385,
|
| 3375 |
+
"grad_norm": 1.5078125,
|
| 3376 |
+
"learning_rate": 8.393891459075777e-06,
|
| 3377 |
+
"loss": 1.1461691856384277,
|
| 3378 |
+
"step": 480
|
| 3379 |
+
},
|
| 3380 |
+
{
|
| 3381 |
+
"epoch": 0.39466666666666667,
|
| 3382 |
+
"grad_norm": 1.703125,
|
| 3383 |
+
"learning_rate": 8.378740460041767e-06,
|
| 3384 |
+
"loss": 1.0773999691009521,
|
| 3385 |
+
"step": 481
|
| 3386 |
+
},
|
| 3387 |
+
{
|
| 3388 |
+
"epoch": 0.3954871794871795,
|
| 3389 |
+
"grad_norm": 1.78125,
|
| 3390 |
+
"learning_rate": 8.363574602549707e-06,
|
| 3391 |
+
"loss": 1.0802425146102905,
|
| 3392 |
+
"step": 482
|
| 3393 |
+
},
|
| 3394 |
+
{
|
| 3395 |
+
"epoch": 0.3963076923076923,
|
| 3396 |
+
"grad_norm": 1.640625,
|
| 3397 |
+
"learning_rate": 8.348393990026907e-06,
|
| 3398 |
+
"loss": 1.0606069564819336,
|
| 3399 |
+
"step": 483
|
| 3400 |
+
},
|
| 3401 |
+
{
|
| 3402 |
+
"epoch": 0.3971282051282051,
|
| 3403 |
+
"grad_norm": 1.71875,
|
| 3404 |
+
"learning_rate": 8.333198726001298e-06,
|
| 3405 |
+
"loss": 1.080243468284607,
|
| 3406 |
+
"step": 484
|
| 3407 |
+
},
|
| 3408 |
+
{
|
| 3409 |
+
"epoch": 0.39794871794871794,
|
| 3410 |
+
"grad_norm": 1.65625,
|
| 3411 |
+
"learning_rate": 8.317988914100733e-06,
|
| 3412 |
+
"loss": 0.9865357875823975,
|
| 3413 |
+
"step": 485
|
| 3414 |
+
},
|
| 3415 |
+
{
|
| 3416 |
+
"epoch": 0.39876923076923076,
|
| 3417 |
+
"grad_norm": 1.5,
|
| 3418 |
+
"learning_rate": 8.302764658052273e-06,
|
| 3419 |
+
"loss": 1.112703561782837,
|
| 3420 |
+
"step": 486
|
| 3421 |
+
},
|
| 3422 |
+
{
|
| 3423 |
+
"epoch": 0.3995897435897436,
|
| 3424 |
+
"grad_norm": 1.5390625,
|
| 3425 |
+
"learning_rate": 8.287526061681487e-06,
|
| 3426 |
+
"loss": 1.1482199430465698,
|
| 3427 |
+
"step": 487
|
| 3428 |
+
},
|
| 3429 |
+
{
|
| 3430 |
+
"epoch": 0.4004102564102564,
|
| 3431 |
+
"grad_norm": 1.578125,
|
| 3432 |
+
"learning_rate": 8.27227322891175e-06,
|
| 3433 |
+
"loss": 1.1118931770324707,
|
| 3434 |
+
"step": 488
|
| 3435 |
+
},
|
| 3436 |
+
{
|
| 3437 |
+
"epoch": 0.4004102564102564,
|
| 3438 |
+
"eval_loss": 1.1139229536056519,
|
| 3439 |
+
"eval_runtime": 25.6102,
|
| 3440 |
+
"eval_samples_per_second": 39.047,
|
| 3441 |
+
"eval_steps_per_second": 9.762,
|
| 3442 |
+
"step": 488
|
| 3443 |
}
|
| 3444 |
],
|
| 3445 |
"logging_steps": 1,
|
|
|
|
| 3459 |
"attributes": {}
|
| 3460 |
}
|
| 3461 |
},
|
| 3462 |
+
"total_flos": 1.2715603388752282e+17,
|
| 3463 |
"train_batch_size": 4,
|
| 3464 |
"trial_name": null,
|
| 3465 |
"trial_params": null
|