lowest loss model (#2)
Browse files- Lowest loss (0.081) model - kaggle 2022 data (388f69a32c0e4558a22b0a4ce70b7c021e92bc84)
- model.safetensors +1 -1
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1714 -6
- training_args.bin +2 -2
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1711848436
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5e98686d0d03cbee7a1649e25db03270c82652fb357994baa867fa48d8e78df7
|
| 3 |
size 1711848436
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:120711f8e3a5143c632fd38b9e2d39f4f3f2e29ca1792568088d420c6f196157
|
| 3 |
+
size 3424048316
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3962da8332dca055eb5eac726e4e894ffb5cd43a2a9df05810854ac241ffd9a4
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:02f28bef84c97955eea9d12d8874e6f8cb49f55187bb78549b0ba5bde0b4d2a2
|
| 3 |
+
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 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,
|
|
@@ -1479,12 +1479,1720 @@
|
|
| 1479 |
"train_runtime": 27901.1518,
|
| 1480 |
"train_samples_per_second": 30.098,
|
| 1481 |
"train_steps_per_second": 3.762
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1482 |
}
|
| 1483 |
],
|
| 1484 |
"logging_steps": 500,
|
| 1485 |
-
"max_steps":
|
| 1486 |
"num_input_tokens_seen": 0,
|
| 1487 |
-
"num_train_epochs":
|
| 1488 |
"save_steps": 500,
|
| 1489 |
"stateful_callbacks": {
|
| 1490 |
"TrainerControl": {
|
|
@@ -1493,12 +3201,12 @@
|
|
| 1493 |
"should_evaluate": false,
|
| 1494 |
"should_log": false,
|
| 1495 |
"should_save": true,
|
| 1496 |
-
"should_training_stop":
|
| 1497 |
},
|
| 1498 |
"attributes": {}
|
| 1499 |
}
|
| 1500 |
},
|
| 1501 |
-
"total_flos":
|
| 1502 |
"train_batch_size": 8,
|
| 1503 |
"trial_name": null,
|
| 1504 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.1577182486758373,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 226500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1479 |
"train_runtime": 27901.1518,
|
| 1480 |
"train_samples_per_second": 30.098,
|
| 1481 |
"train_steps_per_second": 3.762
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 1.0002667377967458,
|
| 1485 |
+
"grad_norm": 14.302581787109375,
|
| 1486 |
+
"learning_rate": 2.2480186776155165e-07,
|
| 1487 |
+
"loss": 0.0893,
|
| 1488 |
+
"step": 105000
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 1.005029912738635,
|
| 1492 |
+
"grad_norm": 2.5177512168884277,
|
| 1493 |
+
"learning_rate": 2.2444491248259286e-07,
|
| 1494 |
+
"loss": 0.1055,
|
| 1495 |
+
"step": 105500
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 1.0097930876805243,
|
| 1499 |
+
"grad_norm": 2.1406190395355225,
|
| 1500 |
+
"learning_rate": 2.2408795720363407e-07,
|
| 1501 |
+
"loss": 0.1254,
|
| 1502 |
+
"step": 106000
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 1.0145562626224136,
|
| 1506 |
+
"grad_norm": 1.1113516092300415,
|
| 1507 |
+
"learning_rate": 2.2373100192467528e-07,
|
| 1508 |
+
"loss": 0.0951,
|
| 1509 |
+
"step": 106500
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 1.0193194375643029,
|
| 1513 |
+
"grad_norm": 35.0811653137207,
|
| 1514 |
+
"learning_rate": 2.233740466457165e-07,
|
| 1515 |
+
"loss": 0.11,
|
| 1516 |
+
"step": 107000
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 1.0240826125061921,
|
| 1520 |
+
"grad_norm": 11.133783340454102,
|
| 1521 |
+
"learning_rate": 2.230170913667577e-07,
|
| 1522 |
+
"loss": 0.1132,
|
| 1523 |
+
"step": 107500
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 1.0288457874480814,
|
| 1527 |
+
"grad_norm": 4.060661315917969,
|
| 1528 |
+
"learning_rate": 2.226601360877989e-07,
|
| 1529 |
+
"loss": 0.1202,
|
| 1530 |
+
"step": 108000
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 1.0336089623899707,
|
| 1534 |
+
"grad_norm": 0.13336028158664703,
|
| 1535 |
+
"learning_rate": 2.2230318080884012e-07,
|
| 1536 |
+
"loss": 0.1139,
|
| 1537 |
+
"step": 108500
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 1.03837213733186,
|
| 1541 |
+
"grad_norm": 1.6817339658737183,
|
| 1542 |
+
"learning_rate": 2.2194622552988133e-07,
|
| 1543 |
+
"loss": 0.1093,
|
| 1544 |
+
"step": 109000
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 1.0431353122737492,
|
| 1548 |
+
"grad_norm": 7.226846694946289,
|
| 1549 |
+
"learning_rate": 2.215892702509225e-07,
|
| 1550 |
+
"loss": 0.1019,
|
| 1551 |
+
"step": 109500
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 1.0478984872156385,
|
| 1555 |
+
"grad_norm": 20.337854385375977,
|
| 1556 |
+
"learning_rate": 2.2123231497196372e-07,
|
| 1557 |
+
"loss": 0.1159,
|
| 1558 |
+
"step": 110000
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 1.0526616621575278,
|
| 1562 |
+
"grad_norm": 12.429189682006836,
|
| 1563 |
+
"learning_rate": 2.2087535969300493e-07,
|
| 1564 |
+
"loss": 0.1058,
|
| 1565 |
+
"step": 110500
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 1.057424837099417,
|
| 1569 |
+
"grad_norm": 2.344127655029297,
|
| 1570 |
+
"learning_rate": 2.2051840441404614e-07,
|
| 1571 |
+
"loss": 0.1252,
|
| 1572 |
+
"step": 111000
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 1.0621880120413063,
|
| 1576 |
+
"grad_norm": 3.0213356018066406,
|
| 1577 |
+
"learning_rate": 2.2016144913508735e-07,
|
| 1578 |
+
"loss": 0.1091,
|
| 1579 |
+
"step": 111500
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 1.0669511869831956,
|
| 1583 |
+
"grad_norm": 8.44406509399414,
|
| 1584 |
+
"learning_rate": 2.198044938561286e-07,
|
| 1585 |
+
"loss": 0.1084,
|
| 1586 |
+
"step": 112000
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 1.0717143619250848,
|
| 1590 |
+
"grad_norm": 5.177192211151123,
|
| 1591 |
+
"learning_rate": 2.194475385771698e-07,
|
| 1592 |
+
"loss": 0.111,
|
| 1593 |
+
"step": 112500
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 1.0764775368669741,
|
| 1597 |
+
"grad_norm": 0.7378814816474915,
|
| 1598 |
+
"learning_rate": 2.19090583298211e-07,
|
| 1599 |
+
"loss": 0.1123,
|
| 1600 |
+
"step": 113000
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 1.0812407118088634,
|
| 1604 |
+
"grad_norm": 11.683990478515625,
|
| 1605 |
+
"learning_rate": 2.1873362801925222e-07,
|
| 1606 |
+
"loss": 0.1049,
|
| 1607 |
+
"step": 113500
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 1.0860038867507527,
|
| 1611 |
+
"grad_norm": 1.845595121383667,
|
| 1612 |
+
"learning_rate": 2.1837667274029343e-07,
|
| 1613 |
+
"loss": 0.1359,
|
| 1614 |
+
"step": 114000
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 1.090767061692642,
|
| 1618 |
+
"grad_norm": 18.56471824645996,
|
| 1619 |
+
"learning_rate": 2.1801971746133464e-07,
|
| 1620 |
+
"loss": 0.1101,
|
| 1621 |
+
"step": 114500
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 1.0955302366345312,
|
| 1625 |
+
"grad_norm": 2.1949081420898438,
|
| 1626 |
+
"learning_rate": 2.1766276218237585e-07,
|
| 1627 |
+
"loss": 0.1213,
|
| 1628 |
+
"step": 115000
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 1.1002934115764205,
|
| 1632 |
+
"grad_norm": 8.824433326721191,
|
| 1633 |
+
"learning_rate": 2.1730580690341706e-07,
|
| 1634 |
+
"loss": 0.0972,
|
| 1635 |
+
"step": 115500
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 1.1050565865183097,
|
| 1639 |
+
"grad_norm": 8.95617389678955,
|
| 1640 |
+
"learning_rate": 2.1694885162445827e-07,
|
| 1641 |
+
"loss": 0.1088,
|
| 1642 |
+
"step": 116000
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 1.109819761460199,
|
| 1646 |
+
"grad_norm": 16.236719131469727,
|
| 1647 |
+
"learning_rate": 2.1659189634549948e-07,
|
| 1648 |
+
"loss": 0.126,
|
| 1649 |
+
"step": 116500
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 1.1145829364020883,
|
| 1653 |
+
"grad_norm": 0.8696116805076599,
|
| 1654 |
+
"learning_rate": 2.1623494106654069e-07,
|
| 1655 |
+
"loss": 0.111,
|
| 1656 |
+
"step": 117000
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 1.1193461113439775,
|
| 1660 |
+
"grad_norm": 22.61142349243164,
|
| 1661 |
+
"learning_rate": 2.158779857875819e-07,
|
| 1662 |
+
"loss": 0.1128,
|
| 1663 |
+
"step": 117500
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 1.1241092862858668,
|
| 1667 |
+
"grad_norm": 9.516005516052246,
|
| 1668 |
+
"learning_rate": 2.155210305086231e-07,
|
| 1669 |
+
"loss": 0.1262,
|
| 1670 |
+
"step": 118000
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 1.1288724612277559,
|
| 1674 |
+
"grad_norm": 2.8408925533294678,
|
| 1675 |
+
"learning_rate": 2.1516407522966432e-07,
|
| 1676 |
+
"loss": 0.1254,
|
| 1677 |
+
"step": 118500
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 1.1336356361696454,
|
| 1681 |
+
"grad_norm": 0.08307074010372162,
|
| 1682 |
+
"learning_rate": 2.148071199507055e-07,
|
| 1683 |
+
"loss": 0.1128,
|
| 1684 |
+
"step": 119000
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 1.1383988111115344,
|
| 1688 |
+
"grad_norm": 0.07671812921762466,
|
| 1689 |
+
"learning_rate": 2.144501646717467e-07,
|
| 1690 |
+
"loss": 0.1105,
|
| 1691 |
+
"step": 119500
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 1.1431619860534237,
|
| 1695 |
+
"grad_norm": 23.30893325805664,
|
| 1696 |
+
"learning_rate": 2.1409320939278792e-07,
|
| 1697 |
+
"loss": 0.1016,
|
| 1698 |
+
"step": 120000
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 1.147925160995313,
|
| 1702 |
+
"grad_norm": 2.4784510135650635,
|
| 1703 |
+
"learning_rate": 2.1373625411382913e-07,
|
| 1704 |
+
"loss": 0.121,
|
| 1705 |
+
"step": 120500
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 1.1526883359372022,
|
| 1709 |
+
"grad_norm": 6.876108169555664,
|
| 1710 |
+
"learning_rate": 2.1337929883487034e-07,
|
| 1711 |
+
"loss": 0.1191,
|
| 1712 |
+
"step": 121000
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 1.1574515108790915,
|
| 1716 |
+
"grad_norm": 0.040010467171669006,
|
| 1717 |
+
"learning_rate": 2.1302234355591155e-07,
|
| 1718 |
+
"loss": 0.107,
|
| 1719 |
+
"step": 121500
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 1.1622146858209808,
|
| 1723 |
+
"grad_norm": 0.3573683202266693,
|
| 1724 |
+
"learning_rate": 2.1266538827695276e-07,
|
| 1725 |
+
"loss": 0.1226,
|
| 1726 |
+
"step": 122000
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 1.16697786076287,
|
| 1730 |
+
"grad_norm": 5.653687477111816,
|
| 1731 |
+
"learning_rate": 2.1230843299799397e-07,
|
| 1732 |
+
"loss": 0.1171,
|
| 1733 |
+
"step": 122500
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 1.1717410357047593,
|
| 1737 |
+
"grad_norm": 0.8669099807739258,
|
| 1738 |
+
"learning_rate": 2.1195147771903518e-07,
|
| 1739 |
+
"loss": 0.1194,
|
| 1740 |
+
"step": 123000
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 1.1765042106466486,
|
| 1744 |
+
"grad_norm": 4.056451320648193,
|
| 1745 |
+
"learning_rate": 2.115945224400764e-07,
|
| 1746 |
+
"loss": 0.111,
|
| 1747 |
+
"step": 123500
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 1.1812673855885378,
|
| 1751 |
+
"grad_norm": 20.773231506347656,
|
| 1752 |
+
"learning_rate": 2.112375671611176e-07,
|
| 1753 |
+
"loss": 0.1009,
|
| 1754 |
+
"step": 124000
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 1.186030560530427,
|
| 1758 |
+
"grad_norm": 0.029131252318620682,
|
| 1759 |
+
"learning_rate": 2.108806118821588e-07,
|
| 1760 |
+
"loss": 0.1076,
|
| 1761 |
+
"step": 124500
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 1.1907937354723164,
|
| 1765 |
+
"grad_norm": 0.8366307020187378,
|
| 1766 |
+
"learning_rate": 2.1052365660320002e-07,
|
| 1767 |
+
"loss": 0.1112,
|
| 1768 |
+
"step": 125000
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 1.1955569104142056,
|
| 1772 |
+
"grad_norm": 9.74376106262207,
|
| 1773 |
+
"learning_rate": 2.1016670132424123e-07,
|
| 1774 |
+
"loss": 0.1101,
|
| 1775 |
+
"step": 125500
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 1.200320085356095,
|
| 1779 |
+
"grad_norm": 4.660519123077393,
|
| 1780 |
+
"learning_rate": 2.0980974604528244e-07,
|
| 1781 |
+
"loss": 0.1084,
|
| 1782 |
+
"step": 126000
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 1.2050832602979842,
|
| 1786 |
+
"grad_norm": 14.676050186157227,
|
| 1787 |
+
"learning_rate": 2.0945279076632365e-07,
|
| 1788 |
+
"loss": 0.1186,
|
| 1789 |
+
"step": 126500
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 1.2098464352398735,
|
| 1793 |
+
"grad_norm": 20.924720764160156,
|
| 1794 |
+
"learning_rate": 2.0909583548736486e-07,
|
| 1795 |
+
"loss": 0.1106,
|
| 1796 |
+
"step": 127000
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 1.2146096101817627,
|
| 1800 |
+
"grad_norm": 0.0010393880074843764,
|
| 1801 |
+
"learning_rate": 2.0873888020840607e-07,
|
| 1802 |
+
"loss": 0.1225,
|
| 1803 |
+
"step": 127500
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 1.219372785123652,
|
| 1807 |
+
"grad_norm": 11.950651168823242,
|
| 1808 |
+
"learning_rate": 2.0838192492944728e-07,
|
| 1809 |
+
"loss": 0.1122,
|
| 1810 |
+
"step": 128000
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 1.2241359600655413,
|
| 1814 |
+
"grad_norm": 2.3444035053253174,
|
| 1815 |
+
"learning_rate": 2.0802496965048849e-07,
|
| 1816 |
+
"loss": 0.116,
|
| 1817 |
+
"step": 128500
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 1.2288991350074305,
|
| 1821 |
+
"grad_norm": 18.723051071166992,
|
| 1822 |
+
"learning_rate": 2.0766801437152967e-07,
|
| 1823 |
+
"loss": 0.1091,
|
| 1824 |
+
"step": 129000
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 1.2336623099493198,
|
| 1828 |
+
"grad_norm": 13.435932159423828,
|
| 1829 |
+
"learning_rate": 2.0731105909257088e-07,
|
| 1830 |
+
"loss": 0.1068,
|
| 1831 |
+
"step": 129500
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 1.238425484891209,
|
| 1835 |
+
"grad_norm": 1.0864711999893188,
|
| 1836 |
+
"learning_rate": 2.069541038136121e-07,
|
| 1837 |
+
"loss": 0.1266,
|
| 1838 |
+
"step": 130000
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 1.2431886598330983,
|
| 1842 |
+
"grad_norm": 8.090726852416992,
|
| 1843 |
+
"learning_rate": 2.065971485346533e-07,
|
| 1844 |
+
"loss": 0.1256,
|
| 1845 |
+
"step": 130500
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 1.2479518347749876,
|
| 1849 |
+
"grad_norm": 3.7775135040283203,
|
| 1850 |
+
"learning_rate": 2.062401932556945e-07,
|
| 1851 |
+
"loss": 0.1112,
|
| 1852 |
+
"step": 131000
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 1.2527150097168769,
|
| 1856 |
+
"grad_norm": 0.0022473677527159452,
|
| 1857 |
+
"learning_rate": 2.0588323797673572e-07,
|
| 1858 |
+
"loss": 0.12,
|
| 1859 |
+
"step": 131500
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 1.2574781846587662,
|
| 1863 |
+
"grad_norm": 1.232242465019226,
|
| 1864 |
+
"learning_rate": 2.0552628269777693e-07,
|
| 1865 |
+
"loss": 0.1023,
|
| 1866 |
+
"step": 132000
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 1.2622413596006554,
|
| 1870 |
+
"grad_norm": 0.16098114848136902,
|
| 1871 |
+
"learning_rate": 2.0516932741881814e-07,
|
| 1872 |
+
"loss": 0.1346,
|
| 1873 |
+
"step": 132500
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 1.2670045345425447,
|
| 1877 |
+
"grad_norm": 12.452221870422363,
|
| 1878 |
+
"learning_rate": 2.0481237213985935e-07,
|
| 1879 |
+
"loss": 0.1215,
|
| 1880 |
+
"step": 133000
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 1.271767709484434,
|
| 1884 |
+
"grad_norm": 10.727972984313965,
|
| 1885 |
+
"learning_rate": 2.0445541686090056e-07,
|
| 1886 |
+
"loss": 0.1062,
|
| 1887 |
+
"step": 133500
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 1.2765308844263232,
|
| 1891 |
+
"grad_norm": 0.5605247020721436,
|
| 1892 |
+
"learning_rate": 2.0409846158194177e-07,
|
| 1893 |
+
"loss": 0.1082,
|
| 1894 |
+
"step": 134000
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 1.2812940593682125,
|
| 1898 |
+
"grad_norm": 0.9301961660385132,
|
| 1899 |
+
"learning_rate": 2.0374150630298298e-07,
|
| 1900 |
+
"loss": 0.1273,
|
| 1901 |
+
"step": 134500
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 1.2860572343101018,
|
| 1905 |
+
"grad_norm": 0.5364490747451782,
|
| 1906 |
+
"learning_rate": 2.033845510240242e-07,
|
| 1907 |
+
"loss": 0.117,
|
| 1908 |
+
"step": 135000
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 1.290820409251991,
|
| 1912 |
+
"grad_norm": 11.421608924865723,
|
| 1913 |
+
"learning_rate": 2.030275957450654e-07,
|
| 1914 |
+
"loss": 0.119,
|
| 1915 |
+
"step": 135500
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 1.2955835841938803,
|
| 1919 |
+
"grad_norm": 29.850894927978516,
|
| 1920 |
+
"learning_rate": 2.026706404661066e-07,
|
| 1921 |
+
"loss": 0.0931,
|
| 1922 |
+
"step": 136000
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 1.3003467591357696,
|
| 1926 |
+
"grad_norm": 1.1624155044555664,
|
| 1927 |
+
"learning_rate": 2.0231368518714782e-07,
|
| 1928 |
+
"loss": 0.1055,
|
| 1929 |
+
"step": 136500
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 1.3051099340776589,
|
| 1933 |
+
"grad_norm": 8.889257431030273,
|
| 1934 |
+
"learning_rate": 2.0195672990818903e-07,
|
| 1935 |
+
"loss": 0.1174,
|
| 1936 |
+
"step": 137000
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 1.3098731090195481,
|
| 1940 |
+
"grad_norm": 0.10517348349094391,
|
| 1941 |
+
"learning_rate": 2.0159977462923024e-07,
|
| 1942 |
+
"loss": 0.0913,
|
| 1943 |
+
"step": 137500
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 1.3146362839614374,
|
| 1947 |
+
"grad_norm": 3.1619289075024426e-05,
|
| 1948 |
+
"learning_rate": 2.0124281935027145e-07,
|
| 1949 |
+
"loss": 0.1053,
|
| 1950 |
+
"step": 138000
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 1.3193994589033267,
|
| 1954 |
+
"grad_norm": 8.190518379211426,
|
| 1955 |
+
"learning_rate": 2.0088586407131266e-07,
|
| 1956 |
+
"loss": 0.1017,
|
| 1957 |
+
"step": 138500
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 1.324162633845216,
|
| 1961 |
+
"grad_norm": 8.892984390258789,
|
| 1962 |
+
"learning_rate": 2.0052890879235387e-07,
|
| 1963 |
+
"loss": 0.1225,
|
| 1964 |
+
"step": 139000
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 1.3289258087871052,
|
| 1968 |
+
"grad_norm": 0.019881388172507286,
|
| 1969 |
+
"learning_rate": 2.0017195351339505e-07,
|
| 1970 |
+
"loss": 0.1109,
|
| 1971 |
+
"step": 139500
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 1.3336889837289945,
|
| 1975 |
+
"grad_norm": 0.39558929204940796,
|
| 1976 |
+
"learning_rate": 1.9981499823443626e-07,
|
| 1977 |
+
"loss": 0.1113,
|
| 1978 |
+
"step": 140000
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 1.3384521586708837,
|
| 1982 |
+
"grad_norm": 0.0027440183330327272,
|
| 1983 |
+
"learning_rate": 1.9945804295547747e-07,
|
| 1984 |
+
"loss": 0.1111,
|
| 1985 |
+
"step": 140500
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 1.343215333612773,
|
| 1989 |
+
"grad_norm": 0.051865462213754654,
|
| 1990 |
+
"learning_rate": 1.9910108767651868e-07,
|
| 1991 |
+
"loss": 0.1001,
|
| 1992 |
+
"step": 141000
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 1.3479785085546623,
|
| 1996 |
+
"grad_norm": 17.542285919189453,
|
| 1997 |
+
"learning_rate": 1.987441323975599e-07,
|
| 1998 |
+
"loss": 0.1093,
|
| 1999 |
+
"step": 141500
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 1.3527416834965513,
|
| 2003 |
+
"grad_norm": 1.5649853944778442,
|
| 2004 |
+
"learning_rate": 1.983871771186011e-07,
|
| 2005 |
+
"loss": 0.1108,
|
| 2006 |
+
"step": 142000
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 1.3575048584384408,
|
| 2010 |
+
"grad_norm": 3.2729179859161377,
|
| 2011 |
+
"learning_rate": 1.980302218396423e-07,
|
| 2012 |
+
"loss": 0.1094,
|
| 2013 |
+
"step": 142500
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 1.3622680333803299,
|
| 2017 |
+
"grad_norm": 19.61039161682129,
|
| 2018 |
+
"learning_rate": 1.9767326656068352e-07,
|
| 2019 |
+
"loss": 0.1232,
|
| 2020 |
+
"step": 143000
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 1.3670312083222194,
|
| 2024 |
+
"grad_norm": 0.48026251792907715,
|
| 2025 |
+
"learning_rate": 1.9731631128172473e-07,
|
| 2026 |
+
"loss": 0.1179,
|
| 2027 |
+
"step": 143500
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 1.3717943832641084,
|
| 2031 |
+
"grad_norm": 0.011149807833135128,
|
| 2032 |
+
"learning_rate": 1.9695935600276594e-07,
|
| 2033 |
+
"loss": 0.1205,
|
| 2034 |
+
"step": 144000
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 1.376557558205998,
|
| 2038 |
+
"grad_norm": 0.02112853154540062,
|
| 2039 |
+
"learning_rate": 1.9660240072380715e-07,
|
| 2040 |
+
"loss": 0.1124,
|
| 2041 |
+
"step": 144500
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 1.381320733147887,
|
| 2045 |
+
"grad_norm": 34.17774200439453,
|
| 2046 |
+
"learning_rate": 1.9624544544484836e-07,
|
| 2047 |
+
"loss": 0.096,
|
| 2048 |
+
"step": 145000
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 1.3860839080897764,
|
| 2052 |
+
"grad_norm": 0.002524700714275241,
|
| 2053 |
+
"learning_rate": 1.9588849016588957e-07,
|
| 2054 |
+
"loss": 0.1279,
|
| 2055 |
+
"step": 145500
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 1.3908470830316655,
|
| 2059 |
+
"grad_norm": 0.4319082498550415,
|
| 2060 |
+
"learning_rate": 1.9553153488693078e-07,
|
| 2061 |
+
"loss": 0.1096,
|
| 2062 |
+
"step": 146000
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 1.395610257973555,
|
| 2066 |
+
"grad_norm": 17.499792098999023,
|
| 2067 |
+
"learning_rate": 1.95174579607972e-07,
|
| 2068 |
+
"loss": 0.1137,
|
| 2069 |
+
"step": 146500
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 1.400373432915444,
|
| 2073 |
+
"grad_norm": 16.145008087158203,
|
| 2074 |
+
"learning_rate": 1.948176243290132e-07,
|
| 2075 |
+
"loss": 0.1051,
|
| 2076 |
+
"step": 147000
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 1.4051366078573335,
|
| 2080 |
+
"grad_norm": 0.8132957816123962,
|
| 2081 |
+
"learning_rate": 1.944606690500544e-07,
|
| 2082 |
+
"loss": 0.1063,
|
| 2083 |
+
"step": 147500
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 1.4098997827992226,
|
| 2087 |
+
"grad_norm": 0.05169007182121277,
|
| 2088 |
+
"learning_rate": 1.9410371377109562e-07,
|
| 2089 |
+
"loss": 0.1047,
|
| 2090 |
+
"step": 148000
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 1.4146629577411118,
|
| 2094 |
+
"grad_norm": 3.716925859451294,
|
| 2095 |
+
"learning_rate": 1.9374675849213683e-07,
|
| 2096 |
+
"loss": 0.1081,
|
| 2097 |
+
"step": 148500
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 1.419426132683001,
|
| 2101 |
+
"grad_norm": 16.134973526000977,
|
| 2102 |
+
"learning_rate": 1.9338980321317804e-07,
|
| 2103 |
+
"loss": 0.127,
|
| 2104 |
+
"step": 149000
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 1.4241893076248904,
|
| 2108 |
+
"grad_norm": 0.8836007118225098,
|
| 2109 |
+
"learning_rate": 1.9303284793421922e-07,
|
| 2110 |
+
"loss": 0.1116,
|
| 2111 |
+
"step": 149500
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 1.4289524825667796,
|
| 2115 |
+
"grad_norm": 4.612446308135986,
|
| 2116 |
+
"learning_rate": 1.9267589265526043e-07,
|
| 2117 |
+
"loss": 0.1263,
|
| 2118 |
+
"step": 150000
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 1.433715657508669,
|
| 2122 |
+
"grad_norm": 0.24257691204547882,
|
| 2123 |
+
"learning_rate": 1.9231893737630164e-07,
|
| 2124 |
+
"loss": 0.1067,
|
| 2125 |
+
"step": 150500
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 1.4384788324505582,
|
| 2129 |
+
"grad_norm": 18.72231101989746,
|
| 2130 |
+
"learning_rate": 1.9196198209734285e-07,
|
| 2131 |
+
"loss": 0.1174,
|
| 2132 |
+
"step": 151000
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 1.4432420073924475,
|
| 2136 |
+
"grad_norm": 0.09871924668550491,
|
| 2137 |
+
"learning_rate": 1.9160502681838409e-07,
|
| 2138 |
+
"loss": 0.1096,
|
| 2139 |
+
"step": 151500
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 1.4480051823343367,
|
| 2143 |
+
"grad_norm": 1.2534486055374146,
|
| 2144 |
+
"learning_rate": 1.912480715394253e-07,
|
| 2145 |
+
"loss": 0.1155,
|
| 2146 |
+
"step": 152000
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 1.452768357276226,
|
| 2150 |
+
"grad_norm": 14.395082473754883,
|
| 2151 |
+
"learning_rate": 1.908911162604665e-07,
|
| 2152 |
+
"loss": 0.1047,
|
| 2153 |
+
"step": 152500
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 1.4575315322181153,
|
| 2157 |
+
"grad_norm": 2.9728424549102783,
|
| 2158 |
+
"learning_rate": 1.9053416098150771e-07,
|
| 2159 |
+
"loss": 0.0942,
|
| 2160 |
+
"step": 153000
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 1.4622947071600045,
|
| 2164 |
+
"grad_norm": 10.138772964477539,
|
| 2165 |
+
"learning_rate": 1.9017720570254892e-07,
|
| 2166 |
+
"loss": 0.1219,
|
| 2167 |
+
"step": 153500
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"epoch": 1.4670578821018938,
|
| 2171 |
+
"grad_norm": 16.113460540771484,
|
| 2172 |
+
"learning_rate": 1.8982025042359013e-07,
|
| 2173 |
+
"loss": 0.1094,
|
| 2174 |
+
"step": 154000
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 1.471821057043783,
|
| 2178 |
+
"grad_norm": 1.4605754613876343,
|
| 2179 |
+
"learning_rate": 1.8946329514463134e-07,
|
| 2180 |
+
"loss": 0.1227,
|
| 2181 |
+
"step": 154500
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 1.4765842319856723,
|
| 2185 |
+
"grad_norm": 19.34916877746582,
|
| 2186 |
+
"learning_rate": 1.8910633986567255e-07,
|
| 2187 |
+
"loss": 0.1032,
|
| 2188 |
+
"step": 155000
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"epoch": 1.4813474069275616,
|
| 2192 |
+
"grad_norm": 5.430410861968994,
|
| 2193 |
+
"learning_rate": 1.8874938458671376e-07,
|
| 2194 |
+
"loss": 0.1107,
|
| 2195 |
+
"step": 155500
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 1.4861105818694509,
|
| 2199 |
+
"grad_norm": 0.01834196224808693,
|
| 2200 |
+
"learning_rate": 1.8839242930775497e-07,
|
| 2201 |
+
"loss": 0.1243,
|
| 2202 |
+
"step": 156000
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 1.4908737568113402,
|
| 2206 |
+
"grad_norm": 22.170907974243164,
|
| 2207 |
+
"learning_rate": 1.8803547402879618e-07,
|
| 2208 |
+
"loss": 0.1019,
|
| 2209 |
+
"step": 156500
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 1.4956369317532294,
|
| 2213 |
+
"grad_norm": 0.28510582447052,
|
| 2214 |
+
"learning_rate": 1.876785187498374e-07,
|
| 2215 |
+
"loss": 0.1354,
|
| 2216 |
+
"step": 157000
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 1.5004001066951187,
|
| 2220 |
+
"grad_norm": 8.662985801696777,
|
| 2221 |
+
"learning_rate": 1.873215634708786e-07,
|
| 2222 |
+
"loss": 0.1043,
|
| 2223 |
+
"step": 157500
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 1.505163281637008,
|
| 2227 |
+
"grad_norm": 5.042104721069336,
|
| 2228 |
+
"learning_rate": 1.869646081919198e-07,
|
| 2229 |
+
"loss": 0.1102,
|
| 2230 |
+
"step": 158000
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 1.5099264565788972,
|
| 2234 |
+
"grad_norm": 0.07364173978567123,
|
| 2235 |
+
"learning_rate": 1.8660765291296102e-07,
|
| 2236 |
+
"loss": 0.129,
|
| 2237 |
+
"step": 158500
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 1.5146896315207865,
|
| 2241 |
+
"grad_norm": 3.405261278152466,
|
| 2242 |
+
"learning_rate": 1.862506976340022e-07,
|
| 2243 |
+
"loss": 0.111,
|
| 2244 |
+
"step": 159000
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 1.5194528064626758,
|
| 2248 |
+
"grad_norm": 3.3657171726226807,
|
| 2249 |
+
"learning_rate": 1.8589374235504342e-07,
|
| 2250 |
+
"loss": 0.1089,
|
| 2251 |
+
"step": 159500
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 1.524215981404565,
|
| 2255 |
+
"grad_norm": 0.028935715556144714,
|
| 2256 |
+
"learning_rate": 1.8553678707608463e-07,
|
| 2257 |
+
"loss": 0.1038,
|
| 2258 |
+
"step": 160000
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 1.5289791563464543,
|
| 2262 |
+
"grad_norm": 1.048443078994751,
|
| 2263 |
+
"learning_rate": 1.8517983179712584e-07,
|
| 2264 |
+
"loss": 0.0938,
|
| 2265 |
+
"step": 160500
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 1.5337423312883436,
|
| 2269 |
+
"grad_norm": 15.945822715759277,
|
| 2270 |
+
"learning_rate": 1.8482287651816705e-07,
|
| 2271 |
+
"loss": 0.1138,
|
| 2272 |
+
"step": 161000
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 1.5385055062302329,
|
| 2276 |
+
"grad_norm": 0.36497658491134644,
|
| 2277 |
+
"learning_rate": 1.8446592123920826e-07,
|
| 2278 |
+
"loss": 0.107,
|
| 2279 |
+
"step": 161500
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 1.5432686811721221,
|
| 2283 |
+
"grad_norm": 0.7298380732536316,
|
| 2284 |
+
"learning_rate": 1.8410896596024947e-07,
|
| 2285 |
+
"loss": 0.1012,
|
| 2286 |
+
"step": 162000
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 1.5480318561140114,
|
| 2290 |
+
"grad_norm": 0.041379645466804504,
|
| 2291 |
+
"learning_rate": 1.8375201068129068e-07,
|
| 2292 |
+
"loss": 0.1171,
|
| 2293 |
+
"step": 162500
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 1.5527950310559007,
|
| 2297 |
+
"grad_norm": 1.361636996269226,
|
| 2298 |
+
"learning_rate": 1.8339505540233188e-07,
|
| 2299 |
+
"loss": 0.0978,
|
| 2300 |
+
"step": 163000
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 1.55755820599779,
|
| 2304 |
+
"grad_norm": 0.0006419995916076005,
|
| 2305 |
+
"learning_rate": 1.830381001233731e-07,
|
| 2306 |
+
"loss": 0.1152,
|
| 2307 |
+
"step": 163500
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 1.5623213809396792,
|
| 2311 |
+
"grad_norm": 29.9415340423584,
|
| 2312 |
+
"learning_rate": 1.826811448444143e-07,
|
| 2313 |
+
"loss": 0.1124,
|
| 2314 |
+
"step": 164000
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 1.5670845558815683,
|
| 2318 |
+
"grad_norm": 14.25601577758789,
|
| 2319 |
+
"learning_rate": 1.8232418956545551e-07,
|
| 2320 |
+
"loss": 0.1096,
|
| 2321 |
+
"step": 164500
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 1.5718477308234577,
|
| 2325 |
+
"grad_norm": 10.357718467712402,
|
| 2326 |
+
"learning_rate": 1.8196723428649672e-07,
|
| 2327 |
+
"loss": 0.1068,
|
| 2328 |
+
"step": 165000
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 1.5766109057653468,
|
| 2332 |
+
"grad_norm": 2.6006975173950195,
|
| 2333 |
+
"learning_rate": 1.8161027900753793e-07,
|
| 2334 |
+
"loss": 0.112,
|
| 2335 |
+
"step": 165500
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 1.5813740807072363,
|
| 2339 |
+
"grad_norm": 0.013517569750547409,
|
| 2340 |
+
"learning_rate": 1.8125332372857914e-07,
|
| 2341 |
+
"loss": 0.1046,
|
| 2342 |
+
"step": 166000
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 1.5861372556491253,
|
| 2346 |
+
"grad_norm": 8.826921463012695,
|
| 2347 |
+
"learning_rate": 1.8089636844962035e-07,
|
| 2348 |
+
"loss": 0.1148,
|
| 2349 |
+
"step": 166500
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 1.5909004305910148,
|
| 2353 |
+
"grad_norm": 30.500505447387695,
|
| 2354 |
+
"learning_rate": 1.8053941317066156e-07,
|
| 2355 |
+
"loss": 0.1022,
|
| 2356 |
+
"step": 167000
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 1.5956636055329039,
|
| 2360 |
+
"grad_norm": 0.0015779563691467047,
|
| 2361 |
+
"learning_rate": 1.8018245789170277e-07,
|
| 2362 |
+
"loss": 0.0967,
|
| 2363 |
+
"step": 167500
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 1.6004267804747934,
|
| 2367 |
+
"grad_norm": 5.478053092956543,
|
| 2368 |
+
"learning_rate": 1.7982550261274398e-07,
|
| 2369 |
+
"loss": 0.1134,
|
| 2370 |
+
"step": 168000
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 1.6051899554166824,
|
| 2374 |
+
"grad_norm": 0.11672214418649673,
|
| 2375 |
+
"learning_rate": 1.794685473337852e-07,
|
| 2376 |
+
"loss": 0.1,
|
| 2377 |
+
"step": 168500
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 1.609953130358572,
|
| 2381 |
+
"grad_norm": 1.3032927513122559,
|
| 2382 |
+
"learning_rate": 1.791115920548264e-07,
|
| 2383 |
+
"loss": 0.1165,
|
| 2384 |
+
"step": 169000
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 1.614716305300461,
|
| 2388 |
+
"grad_norm": 0.20952224731445312,
|
| 2389 |
+
"learning_rate": 1.7875463677586759e-07,
|
| 2390 |
+
"loss": 0.112,
|
| 2391 |
+
"step": 169500
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 1.6194794802423504,
|
| 2395 |
+
"grad_norm": 5.338772296905518,
|
| 2396 |
+
"learning_rate": 1.783976814969088e-07,
|
| 2397 |
+
"loss": 0.1151,
|
| 2398 |
+
"step": 170000
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 1.6242426551842395,
|
| 2402 |
+
"grad_norm": 14.308218955993652,
|
| 2403 |
+
"learning_rate": 1.7804072621795e-07,
|
| 2404 |
+
"loss": 0.1188,
|
| 2405 |
+
"step": 170500
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 1.629005830126129,
|
| 2409 |
+
"grad_norm": 2.5902771949768066,
|
| 2410 |
+
"learning_rate": 1.7768377093899122e-07,
|
| 2411 |
+
"loss": 0.1095,
|
| 2412 |
+
"step": 171000
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 1.633769005068018,
|
| 2416 |
+
"grad_norm": 1.145963191986084,
|
| 2417 |
+
"learning_rate": 1.7732681566003243e-07,
|
| 2418 |
+
"loss": 0.1081,
|
| 2419 |
+
"step": 171500
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 1.6385321800099075,
|
| 2423 |
+
"grad_norm": 0.019109368324279785,
|
| 2424 |
+
"learning_rate": 1.7696986038107364e-07,
|
| 2425 |
+
"loss": 0.1037,
|
| 2426 |
+
"step": 172000
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 1.6432953549517966,
|
| 2430 |
+
"grad_norm": 0.381287544965744,
|
| 2431 |
+
"learning_rate": 1.7661290510211485e-07,
|
| 2432 |
+
"loss": 0.1024,
|
| 2433 |
+
"step": 172500
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 1.648058529893686,
|
| 2437 |
+
"grad_norm": 0.28823208808898926,
|
| 2438 |
+
"learning_rate": 1.7625594982315606e-07,
|
| 2439 |
+
"loss": 0.1112,
|
| 2440 |
+
"step": 173000
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 1.6528217048355751,
|
| 2444 |
+
"grad_norm": 7.33282995223999,
|
| 2445 |
+
"learning_rate": 1.7589899454419726e-07,
|
| 2446 |
+
"loss": 0.1037,
|
| 2447 |
+
"step": 173500
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 1.6575848797774646,
|
| 2451 |
+
"grad_norm": 18.685562133789062,
|
| 2452 |
+
"learning_rate": 1.7554203926523847e-07,
|
| 2453 |
+
"loss": 0.1096,
|
| 2454 |
+
"step": 174000
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 1.6623480547193537,
|
| 2458 |
+
"grad_norm": 5.129574298858643,
|
| 2459 |
+
"learning_rate": 1.7518508398627968e-07,
|
| 2460 |
+
"loss": 0.1028,
|
| 2461 |
+
"step": 174500
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 1.6671112296612431,
|
| 2465 |
+
"grad_norm": 1.4392452239990234,
|
| 2466 |
+
"learning_rate": 1.748281287073209e-07,
|
| 2467 |
+
"loss": 0.1106,
|
| 2468 |
+
"step": 175000
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 1.6718744046031322,
|
| 2472 |
+
"grad_norm": 3.51692795753479,
|
| 2473 |
+
"learning_rate": 1.744711734283621e-07,
|
| 2474 |
+
"loss": 0.1142,
|
| 2475 |
+
"step": 175500
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 1.6766375795450217,
|
| 2479 |
+
"grad_norm": 0.026211030781269073,
|
| 2480 |
+
"learning_rate": 1.7411421814940331e-07,
|
| 2481 |
+
"loss": 0.1083,
|
| 2482 |
+
"step": 176000
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 1.6814007544869107,
|
| 2486 |
+
"grad_norm": 4.911788463592529,
|
| 2487 |
+
"learning_rate": 1.7375726287044452e-07,
|
| 2488 |
+
"loss": 0.1053,
|
| 2489 |
+
"step": 176500
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 1.6861639294288002,
|
| 2493 |
+
"grad_norm": 5.630278587341309,
|
| 2494 |
+
"learning_rate": 1.7340030759148573e-07,
|
| 2495 |
+
"loss": 0.1087,
|
| 2496 |
+
"step": 177000
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 1.6909271043706893,
|
| 2500 |
+
"grad_norm": 1.8357000350952148,
|
| 2501 |
+
"learning_rate": 1.7304335231252694e-07,
|
| 2502 |
+
"loss": 0.104,
|
| 2503 |
+
"step": 177500
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 1.6956902793125785,
|
| 2507 |
+
"grad_norm": 2.734142780303955,
|
| 2508 |
+
"learning_rate": 1.7268639703356815e-07,
|
| 2509 |
+
"loss": 0.1339,
|
| 2510 |
+
"step": 178000
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 1.7004534542544678,
|
| 2514 |
+
"grad_norm": 27.017423629760742,
|
| 2515 |
+
"learning_rate": 1.7232944175460936e-07,
|
| 2516 |
+
"loss": 0.1085,
|
| 2517 |
+
"step": 178500
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 1.705216629196357,
|
| 2521 |
+
"grad_norm": 0.3949691653251648,
|
| 2522 |
+
"learning_rate": 1.7197248647565057e-07,
|
| 2523 |
+
"loss": 0.1074,
|
| 2524 |
+
"step": 179000
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 1.7099798041382464,
|
| 2528 |
+
"grad_norm": 14.20468807220459,
|
| 2529 |
+
"learning_rate": 1.7161553119669176e-07,
|
| 2530 |
+
"loss": 0.099,
|
| 2531 |
+
"step": 179500
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 1.7147429790801356,
|
| 2535 |
+
"grad_norm": 0.0762014091014862,
|
| 2536 |
+
"learning_rate": 1.7125857591773297e-07,
|
| 2537 |
+
"loss": 0.1108,
|
| 2538 |
+
"step": 180000
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 1.719506154022025,
|
| 2542 |
+
"grad_norm": 19.510847091674805,
|
| 2543 |
+
"learning_rate": 1.7090162063877418e-07,
|
| 2544 |
+
"loss": 0.1225,
|
| 2545 |
+
"step": 180500
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 1.7242693289639142,
|
| 2549 |
+
"grad_norm": 0.08469968289136887,
|
| 2550 |
+
"learning_rate": 1.7054466535981539e-07,
|
| 2551 |
+
"loss": 0.1015,
|
| 2552 |
+
"step": 181000
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 1.7290325039058034,
|
| 2556 |
+
"grad_norm": 30.91697120666504,
|
| 2557 |
+
"learning_rate": 1.701877100808566e-07,
|
| 2558 |
+
"loss": 0.108,
|
| 2559 |
+
"step": 181500
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 1.7337956788476927,
|
| 2563 |
+
"grad_norm": 5.243447780609131,
|
| 2564 |
+
"learning_rate": 1.698307548018978e-07,
|
| 2565 |
+
"loss": 0.1279,
|
| 2566 |
+
"step": 182000
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 1.738558853789582,
|
| 2570 |
+
"grad_norm": 19.558759689331055,
|
| 2571 |
+
"learning_rate": 1.6947379952293902e-07,
|
| 2572 |
+
"loss": 0.1043,
|
| 2573 |
+
"step": 182500
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 1.7433220287314712,
|
| 2577 |
+
"grad_norm": 18.18353271484375,
|
| 2578 |
+
"learning_rate": 1.6911684424398023e-07,
|
| 2579 |
+
"loss": 0.1005,
|
| 2580 |
+
"step": 183000
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 1.7480852036733605,
|
| 2584 |
+
"grad_norm": 0.06688349694013596,
|
| 2585 |
+
"learning_rate": 1.6875988896502144e-07,
|
| 2586 |
+
"loss": 0.1021,
|
| 2587 |
+
"step": 183500
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 1.7528483786152498,
|
| 2591 |
+
"grad_norm": 3.162032127380371,
|
| 2592 |
+
"learning_rate": 1.6840293368606264e-07,
|
| 2593 |
+
"loss": 0.1244,
|
| 2594 |
+
"step": 184000
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 1.757611553557139,
|
| 2598 |
+
"grad_norm": 0.6052823662757874,
|
| 2599 |
+
"learning_rate": 1.6804597840710385e-07,
|
| 2600 |
+
"loss": 0.1145,
|
| 2601 |
+
"step": 184500
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 1.7623747284990283,
|
| 2605 |
+
"grad_norm": 12.853279113769531,
|
| 2606 |
+
"learning_rate": 1.6768902312814506e-07,
|
| 2607 |
+
"loss": 0.1004,
|
| 2608 |
+
"step": 185000
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 1.7671379034409176,
|
| 2612 |
+
"grad_norm": 0.010185870341956615,
|
| 2613 |
+
"learning_rate": 1.6733206784918627e-07,
|
| 2614 |
+
"loss": 0.1048,
|
| 2615 |
+
"step": 185500
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 1.7719010783828069,
|
| 2619 |
+
"grad_norm": 7.452764839399606e-05,
|
| 2620 |
+
"learning_rate": 1.6697511257022748e-07,
|
| 2621 |
+
"loss": 0.1,
|
| 2622 |
+
"step": 186000
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 1.7766642533246961,
|
| 2626 |
+
"grad_norm": 17.169069290161133,
|
| 2627 |
+
"learning_rate": 1.666181572912687e-07,
|
| 2628 |
+
"loss": 0.1176,
|
| 2629 |
+
"step": 186500
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 1.7814274282665854,
|
| 2633 |
+
"grad_norm": 12.43346118927002,
|
| 2634 |
+
"learning_rate": 1.662612020123099e-07,
|
| 2635 |
+
"loss": 0.111,
|
| 2636 |
+
"step": 187000
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 1.7861906032084747,
|
| 2640 |
+
"grad_norm": 13.447138786315918,
|
| 2641 |
+
"learning_rate": 1.6590424673335111e-07,
|
| 2642 |
+
"loss": 0.1198,
|
| 2643 |
+
"step": 187500
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 1.790953778150364,
|
| 2647 |
+
"grad_norm": 2.679323673248291,
|
| 2648 |
+
"learning_rate": 1.6554729145439232e-07,
|
| 2649 |
+
"loss": 0.1083,
|
| 2650 |
+
"step": 188000
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 1.7957169530922532,
|
| 2654 |
+
"grad_norm": 12.469381332397461,
|
| 2655 |
+
"learning_rate": 1.6519033617543353e-07,
|
| 2656 |
+
"loss": 0.0948,
|
| 2657 |
+
"step": 188500
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 1.8004801280341425,
|
| 2661 |
+
"grad_norm": 5.650728225708008,
|
| 2662 |
+
"learning_rate": 1.6483338089647474e-07,
|
| 2663 |
+
"loss": 0.1329,
|
| 2664 |
+
"step": 189000
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 1.8052433029760317,
|
| 2668 |
+
"grad_norm": 0.8955023884773254,
|
| 2669 |
+
"learning_rate": 1.6447642561751593e-07,
|
| 2670 |
+
"loss": 0.0933,
|
| 2671 |
+
"step": 189500
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 1.810006477917921,
|
| 2675 |
+
"grad_norm": 26.87681770324707,
|
| 2676 |
+
"learning_rate": 1.6411947033855714e-07,
|
| 2677 |
+
"loss": 0.1015,
|
| 2678 |
+
"step": 190000
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 1.8147696528598103,
|
| 2682 |
+
"grad_norm": 4.95468282699585,
|
| 2683 |
+
"learning_rate": 1.6376251505959835e-07,
|
| 2684 |
+
"loss": 0.1067,
|
| 2685 |
+
"step": 190500
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 1.8195328278016993,
|
| 2689 |
+
"grad_norm": 0.931701123714447,
|
| 2690 |
+
"learning_rate": 1.6340555978063958e-07,
|
| 2691 |
+
"loss": 0.1282,
|
| 2692 |
+
"step": 191000
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 1.8242960027435888,
|
| 2696 |
+
"grad_norm": 7.3879780769348145,
|
| 2697 |
+
"learning_rate": 1.630486045016808e-07,
|
| 2698 |
+
"loss": 0.0935,
|
| 2699 |
+
"step": 191500
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 1.8290591776854779,
|
| 2703 |
+
"grad_norm": 0.022659489884972572,
|
| 2704 |
+
"learning_rate": 1.62691649222722e-07,
|
| 2705 |
+
"loss": 0.1086,
|
| 2706 |
+
"step": 192000
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 1.8338223526273674,
|
| 2710 |
+
"grad_norm": 0.1140231192111969,
|
| 2711 |
+
"learning_rate": 1.623346939437632e-07,
|
| 2712 |
+
"loss": 0.1005,
|
| 2713 |
+
"step": 192500
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 1.8385855275692564,
|
| 2717 |
+
"grad_norm": 0.6795700192451477,
|
| 2718 |
+
"learning_rate": 1.6197773866480442e-07,
|
| 2719 |
+
"loss": 0.11,
|
| 2720 |
+
"step": 193000
|
| 2721 |
+
},
|
| 2722 |
+
{
|
| 2723 |
+
"epoch": 1.843348702511146,
|
| 2724 |
+
"grad_norm": 9.37308406829834,
|
| 2725 |
+
"learning_rate": 1.6162078338584563e-07,
|
| 2726 |
+
"loss": 0.1224,
|
| 2727 |
+
"step": 193500
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 1.848111877453035,
|
| 2731 |
+
"grad_norm": 33.167701721191406,
|
| 2732 |
+
"learning_rate": 1.6126382810688684e-07,
|
| 2733 |
+
"loss": 0.1122,
|
| 2734 |
+
"step": 194000
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 1.8528750523949244,
|
| 2738 |
+
"grad_norm": 0.8446443676948547,
|
| 2739 |
+
"learning_rate": 1.6090687282792805e-07,
|
| 2740 |
+
"loss": 0.1098,
|
| 2741 |
+
"step": 194500
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 1.8576382273368135,
|
| 2745 |
+
"grad_norm": 30.448610305786133,
|
| 2746 |
+
"learning_rate": 1.6054991754896926e-07,
|
| 2747 |
+
"loss": 0.1041,
|
| 2748 |
+
"step": 195000
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 1.862401402278703,
|
| 2752 |
+
"grad_norm": 21.807527542114258,
|
| 2753 |
+
"learning_rate": 1.6019296227001047e-07,
|
| 2754 |
+
"loss": 0.097,
|
| 2755 |
+
"step": 195500
|
| 2756 |
+
},
|
| 2757 |
+
{
|
| 2758 |
+
"epoch": 1.867164577220592,
|
| 2759 |
+
"grad_norm": 0.030451184138655663,
|
| 2760 |
+
"learning_rate": 1.5983600699105168e-07,
|
| 2761 |
+
"loss": 0.1073,
|
| 2762 |
+
"step": 196000
|
| 2763 |
+
},
|
| 2764 |
+
{
|
| 2765 |
+
"epoch": 1.8719277521624815,
|
| 2766 |
+
"grad_norm": 0.03133384510874748,
|
| 2767 |
+
"learning_rate": 1.594790517120929e-07,
|
| 2768 |
+
"loss": 0.1123,
|
| 2769 |
+
"step": 196500
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 1.8766909271043706,
|
| 2773 |
+
"grad_norm": 2.491389513015747,
|
| 2774 |
+
"learning_rate": 1.591220964331341e-07,
|
| 2775 |
+
"loss": 0.0996,
|
| 2776 |
+
"step": 197000
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 1.88145410204626,
|
| 2780 |
+
"grad_norm": 29.67350959777832,
|
| 2781 |
+
"learning_rate": 1.587651411541753e-07,
|
| 2782 |
+
"loss": 0.0885,
|
| 2783 |
+
"step": 197500
|
| 2784 |
+
},
|
| 2785 |
+
{
|
| 2786 |
+
"epoch": 1.8862172769881491,
|
| 2787 |
+
"grad_norm": 0.28274720907211304,
|
| 2788 |
+
"learning_rate": 1.5840818587521652e-07,
|
| 2789 |
+
"loss": 0.1085,
|
| 2790 |
+
"step": 198000
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 1.8909804519300386,
|
| 2794 |
+
"grad_norm": 0.04003256559371948,
|
| 2795 |
+
"learning_rate": 1.5805123059625773e-07,
|
| 2796 |
+
"loss": 0.1044,
|
| 2797 |
+
"step": 198500
|
| 2798 |
+
},
|
| 2799 |
+
{
|
| 2800 |
+
"epoch": 1.8957436268719277,
|
| 2801 |
+
"grad_norm": 44.06182861328125,
|
| 2802 |
+
"learning_rate": 1.5769427531729894e-07,
|
| 2803 |
+
"loss": 0.1161,
|
| 2804 |
+
"step": 199000
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 1.9005068018138171,
|
| 2808 |
+
"grad_norm": 3.635102621046826e-05,
|
| 2809 |
+
"learning_rate": 1.5733732003834012e-07,
|
| 2810 |
+
"loss": 0.112,
|
| 2811 |
+
"step": 199500
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 1.9052699767557062,
|
| 2815 |
+
"grad_norm": 0.03733281418681145,
|
| 2816 |
+
"learning_rate": 1.5698036475938133e-07,
|
| 2817 |
+
"loss": 0.1241,
|
| 2818 |
+
"step": 200000
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 1.9100331516975957,
|
| 2822 |
+
"grad_norm": 0.009642825461924076,
|
| 2823 |
+
"learning_rate": 1.5662340948042254e-07,
|
| 2824 |
+
"loss": 0.1116,
|
| 2825 |
+
"step": 200500
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 1.9147963266394847,
|
| 2829 |
+
"grad_norm": 0.04384785518050194,
|
| 2830 |
+
"learning_rate": 1.5626645420146375e-07,
|
| 2831 |
+
"loss": 0.0975,
|
| 2832 |
+
"step": 201000
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 1.9195595015813742,
|
| 2836 |
+
"grad_norm": 28.704349517822266,
|
| 2837 |
+
"learning_rate": 1.5590949892250496e-07,
|
| 2838 |
+
"loss": 0.1376,
|
| 2839 |
+
"step": 201500
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 1.9243226765232633,
|
| 2843 |
+
"grad_norm": 0.0023307004012167454,
|
| 2844 |
+
"learning_rate": 1.5555254364354617e-07,
|
| 2845 |
+
"loss": 0.1216,
|
| 2846 |
+
"step": 202000
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 1.9290858514651528,
|
| 2850 |
+
"grad_norm": 3.073564291000366,
|
| 2851 |
+
"learning_rate": 1.5519558836458738e-07,
|
| 2852 |
+
"loss": 0.1139,
|
| 2853 |
+
"step": 202500
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 1.9338490264070418,
|
| 2857 |
+
"grad_norm": 0.6808690428733826,
|
| 2858 |
+
"learning_rate": 1.548386330856286e-07,
|
| 2859 |
+
"loss": 0.0976,
|
| 2860 |
+
"step": 203000
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 1.9386122013489313,
|
| 2864 |
+
"grad_norm": 5.560125350952148,
|
| 2865 |
+
"learning_rate": 1.544816778066698e-07,
|
| 2866 |
+
"loss": 0.1022,
|
| 2867 |
+
"step": 203500
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 1.9433753762908204,
|
| 2871 |
+
"grad_norm": 0.019307024776935577,
|
| 2872 |
+
"learning_rate": 1.54124722527711e-07,
|
| 2873 |
+
"loss": 0.1042,
|
| 2874 |
+
"step": 204000
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 1.9481385512327096,
|
| 2878 |
+
"grad_norm": 0.23428542912006378,
|
| 2879 |
+
"learning_rate": 1.5376776724875222e-07,
|
| 2880 |
+
"loss": 0.1173,
|
| 2881 |
+
"step": 204500
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 1.952901726174599,
|
| 2885 |
+
"grad_norm": 5.58413553237915,
|
| 2886 |
+
"learning_rate": 1.5341081196979343e-07,
|
| 2887 |
+
"loss": 0.1216,
|
| 2888 |
+
"step": 205000
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 1.9576649011164882,
|
| 2892 |
+
"grad_norm": 2.8649184703826904,
|
| 2893 |
+
"learning_rate": 1.5305385669083464e-07,
|
| 2894 |
+
"loss": 0.1034,
|
| 2895 |
+
"step": 205500
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"epoch": 1.9624280760583774,
|
| 2899 |
+
"grad_norm": 0.037286777049303055,
|
| 2900 |
+
"learning_rate": 1.5269690141187585e-07,
|
| 2901 |
+
"loss": 0.1011,
|
| 2902 |
+
"step": 206000
|
| 2903 |
+
},
|
| 2904 |
+
{
|
| 2905 |
+
"epoch": 1.9671912510002667,
|
| 2906 |
+
"grad_norm": 0.007163808681070805,
|
| 2907 |
+
"learning_rate": 1.5233994613291706e-07,
|
| 2908 |
+
"loss": 0.1087,
|
| 2909 |
+
"step": 206500
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 1.971954425942156,
|
| 2913 |
+
"grad_norm": 24.465059280395508,
|
| 2914 |
+
"learning_rate": 1.5198299085395827e-07,
|
| 2915 |
+
"loss": 0.1122,
|
| 2916 |
+
"step": 207000
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 1.9767176008840452,
|
| 2920 |
+
"grad_norm": 6.201001167297363,
|
| 2921 |
+
"learning_rate": 1.5162603557499948e-07,
|
| 2922 |
+
"loss": 0.0882,
|
| 2923 |
+
"step": 207500
|
| 2924 |
+
},
|
| 2925 |
+
{
|
| 2926 |
+
"epoch": 1.9814807758259345,
|
| 2927 |
+
"grad_norm": 0.19351300597190857,
|
| 2928 |
+
"learning_rate": 1.512690802960407e-07,
|
| 2929 |
+
"loss": 0.1109,
|
| 2930 |
+
"step": 208000
|
| 2931 |
+
},
|
| 2932 |
+
{
|
| 2933 |
+
"epoch": 1.9862439507678238,
|
| 2934 |
+
"grad_norm": 0.0022748825140297413,
|
| 2935 |
+
"learning_rate": 1.509121250170819e-07,
|
| 2936 |
+
"loss": 0.1045,
|
| 2937 |
+
"step": 208500
|
| 2938 |
+
},
|
| 2939 |
+
{
|
| 2940 |
+
"epoch": 1.991007125709713,
|
| 2941 |
+
"grad_norm": 0.00041236350079998374,
|
| 2942 |
+
"learning_rate": 1.505551697381231e-07,
|
| 2943 |
+
"loss": 0.1095,
|
| 2944 |
+
"step": 209000
|
| 2945 |
+
},
|
| 2946 |
+
{
|
| 2947 |
+
"epoch": 1.9957703006516023,
|
| 2948 |
+
"grad_norm": 11.131386756896973,
|
| 2949 |
+
"learning_rate": 1.501982144591643e-07,
|
| 2950 |
+
"loss": 0.1293,
|
| 2951 |
+
"step": 209500
|
| 2952 |
+
},
|
| 2953 |
+
{
|
| 2954 |
+
"epoch": 2.0005334755934916,
|
| 2955 |
+
"grad_norm": 0.002583843655884266,
|
| 2956 |
+
"learning_rate": 1.498412591802055e-07,
|
| 2957 |
+
"loss": 0.1142,
|
| 2958 |
+
"step": 210000
|
| 2959 |
+
},
|
| 2960 |
+
{
|
| 2961 |
+
"epoch": 2.0052966505353806,
|
| 2962 |
+
"grad_norm": 14.562299728393555,
|
| 2963 |
+
"learning_rate": 1.494843039012467e-07,
|
| 2964 |
+
"loss": 0.1048,
|
| 2965 |
+
"step": 210500
|
| 2966 |
+
},
|
| 2967 |
+
{
|
| 2968 |
+
"epoch": 2.01005982547727,
|
| 2969 |
+
"grad_norm": 4.4846510887146,
|
| 2970 |
+
"learning_rate": 1.4912734862228792e-07,
|
| 2971 |
+
"loss": 0.1019,
|
| 2972 |
+
"step": 211000
|
| 2973 |
+
},
|
| 2974 |
+
{
|
| 2975 |
+
"epoch": 2.014823000419159,
|
| 2976 |
+
"grad_norm": 0.0017532992642372847,
|
| 2977 |
+
"learning_rate": 1.4877039334332913e-07,
|
| 2978 |
+
"loss": 0.1092,
|
| 2979 |
+
"step": 211500
|
| 2980 |
+
},
|
| 2981 |
+
{
|
| 2982 |
+
"epoch": 2.0195861753610487,
|
| 2983 |
+
"grad_norm": 0.005973345600068569,
|
| 2984 |
+
"learning_rate": 1.4841343806437034e-07,
|
| 2985 |
+
"loss": 0.099,
|
| 2986 |
+
"step": 212000
|
| 2987 |
+
},
|
| 2988 |
+
{
|
| 2989 |
+
"epoch": 2.0243493503029377,
|
| 2990 |
+
"grad_norm": 25.91645622253418,
|
| 2991 |
+
"learning_rate": 1.4805648278541155e-07,
|
| 2992 |
+
"loss": 0.1197,
|
| 2993 |
+
"step": 212500
|
| 2994 |
+
},
|
| 2995 |
+
{
|
| 2996 |
+
"epoch": 2.029112525244827,
|
| 2997 |
+
"grad_norm": 0.009553452022373676,
|
| 2998 |
+
"learning_rate": 1.4769952750645276e-07,
|
| 2999 |
+
"loss": 0.1058,
|
| 3000 |
+
"step": 213000
|
| 3001 |
+
},
|
| 3002 |
+
{
|
| 3003 |
+
"epoch": 2.0338757001867163,
|
| 3004 |
+
"grad_norm": 2.3231234550476074,
|
| 3005 |
+
"learning_rate": 1.4734257222749397e-07,
|
| 3006 |
+
"loss": 0.1098,
|
| 3007 |
+
"step": 213500
|
| 3008 |
+
},
|
| 3009 |
+
{
|
| 3010 |
+
"epoch": 2.0386388751286058,
|
| 3011 |
+
"grad_norm": 7.085731506347656,
|
| 3012 |
+
"learning_rate": 1.4698561694853518e-07,
|
| 3013 |
+
"loss": 0.0992,
|
| 3014 |
+
"step": 214000
|
| 3015 |
+
},
|
| 3016 |
+
{
|
| 3017 |
+
"epoch": 2.043402050070495,
|
| 3018 |
+
"grad_norm": 0.2953038215637207,
|
| 3019 |
+
"learning_rate": 1.466286616695764e-07,
|
| 3020 |
+
"loss": 0.1126,
|
| 3021 |
+
"step": 214500
|
| 3022 |
+
},
|
| 3023 |
+
{
|
| 3024 |
+
"epoch": 2.0481652250123843,
|
| 3025 |
+
"grad_norm": 0.0010392339900135994,
|
| 3026 |
+
"learning_rate": 1.462717063906176e-07,
|
| 3027 |
+
"loss": 0.1034,
|
| 3028 |
+
"step": 215000
|
| 3029 |
+
},
|
| 3030 |
+
{
|
| 3031 |
+
"epoch": 2.0529283999542733,
|
| 3032 |
+
"grad_norm": 0.07149825990200043,
|
| 3033 |
+
"learning_rate": 1.459147511116588e-07,
|
| 3034 |
+
"loss": 0.1008,
|
| 3035 |
+
"step": 215500
|
| 3036 |
+
},
|
| 3037 |
+
{
|
| 3038 |
+
"epoch": 2.057691574896163,
|
| 3039 |
+
"grad_norm": 0.594185471534729,
|
| 3040 |
+
"learning_rate": 1.4555779583270002e-07,
|
| 3041 |
+
"loss": 0.1004,
|
| 3042 |
+
"step": 216000
|
| 3043 |
+
},
|
| 3044 |
+
{
|
| 3045 |
+
"epoch": 2.062454749838052,
|
| 3046 |
+
"grad_norm": 2.347266912460327,
|
| 3047 |
+
"learning_rate": 1.4520084055374123e-07,
|
| 3048 |
+
"loss": 0.1087,
|
| 3049 |
+
"step": 216500
|
| 3050 |
+
},
|
| 3051 |
+
{
|
| 3052 |
+
"epoch": 2.0672179247799414,
|
| 3053 |
+
"grad_norm": 0.16203026473522186,
|
| 3054 |
+
"learning_rate": 1.4484388527478244e-07,
|
| 3055 |
+
"loss": 0.1065,
|
| 3056 |
+
"step": 217000
|
| 3057 |
+
},
|
| 3058 |
+
{
|
| 3059 |
+
"epoch": 2.0719810997218304,
|
| 3060 |
+
"grad_norm": 0.03992275521159172,
|
| 3061 |
+
"learning_rate": 1.4448692999582365e-07,
|
| 3062 |
+
"loss": 0.1247,
|
| 3063 |
+
"step": 217500
|
| 3064 |
+
},
|
| 3065 |
+
{
|
| 3066 |
+
"epoch": 2.07674427466372,
|
| 3067 |
+
"grad_norm": 0.5468718409538269,
|
| 3068 |
+
"learning_rate": 1.4412997471686486e-07,
|
| 3069 |
+
"loss": 0.1078,
|
| 3070 |
+
"step": 218000
|
| 3071 |
+
},
|
| 3072 |
+
{
|
| 3073 |
+
"epoch": 2.081507449605609,
|
| 3074 |
+
"grad_norm": 7.58856201171875,
|
| 3075 |
+
"learning_rate": 1.4377301943790607e-07,
|
| 3076 |
+
"loss": 0.0921,
|
| 3077 |
+
"step": 218500
|
| 3078 |
+
},
|
| 3079 |
+
{
|
| 3080 |
+
"epoch": 2.0862706245474985,
|
| 3081 |
+
"grad_norm": 17.591888427734375,
|
| 3082 |
+
"learning_rate": 1.4341606415894728e-07,
|
| 3083 |
+
"loss": 0.1105,
|
| 3084 |
+
"step": 219000
|
| 3085 |
+
},
|
| 3086 |
+
{
|
| 3087 |
+
"epoch": 2.0910337994893875,
|
| 3088 |
+
"grad_norm": 9.395910263061523,
|
| 3089 |
+
"learning_rate": 1.4305910887998846e-07,
|
| 3090 |
+
"loss": 0.1021,
|
| 3091 |
+
"step": 219500
|
| 3092 |
+
},
|
| 3093 |
+
{
|
| 3094 |
+
"epoch": 2.095796974431277,
|
| 3095 |
+
"grad_norm": 0.0014383163070306182,
|
| 3096 |
+
"learning_rate": 1.4270215360102967e-07,
|
| 3097 |
+
"loss": 0.1089,
|
| 3098 |
+
"step": 220000
|
| 3099 |
+
},
|
| 3100 |
+
{
|
| 3101 |
+
"epoch": 2.100560149373166,
|
| 3102 |
+
"grad_norm": 0.11439737677574158,
|
| 3103 |
+
"learning_rate": 1.4234519832207088e-07,
|
| 3104 |
+
"loss": 0.1108,
|
| 3105 |
+
"step": 220500
|
| 3106 |
+
},
|
| 3107 |
+
{
|
| 3108 |
+
"epoch": 2.1053233243150555,
|
| 3109 |
+
"grad_norm": 1.5022146701812744,
|
| 3110 |
+
"learning_rate": 1.419882430431121e-07,
|
| 3111 |
+
"loss": 0.1098,
|
| 3112 |
+
"step": 221000
|
| 3113 |
+
},
|
| 3114 |
+
{
|
| 3115 |
+
"epoch": 2.1100864992569446,
|
| 3116 |
+
"grad_norm": 0.004490518942475319,
|
| 3117 |
+
"learning_rate": 1.416312877641533e-07,
|
| 3118 |
+
"loss": 0.1063,
|
| 3119 |
+
"step": 221500
|
| 3120 |
+
},
|
| 3121 |
+
{
|
| 3122 |
+
"epoch": 2.114849674198834,
|
| 3123 |
+
"grad_norm": 5.6877360343933105,
|
| 3124 |
+
"learning_rate": 1.412743324851945e-07,
|
| 3125 |
+
"loss": 0.1107,
|
| 3126 |
+
"step": 222000
|
| 3127 |
+
},
|
| 3128 |
+
{
|
| 3129 |
+
"epoch": 2.119612849140723,
|
| 3130 |
+
"grad_norm": 1.135617733001709,
|
| 3131 |
+
"learning_rate": 1.4091737720623572e-07,
|
| 3132 |
+
"loss": 0.115,
|
| 3133 |
+
"step": 222500
|
| 3134 |
+
},
|
| 3135 |
+
{
|
| 3136 |
+
"epoch": 2.1243760240826126,
|
| 3137 |
+
"grad_norm": 2.0897531509399414,
|
| 3138 |
+
"learning_rate": 1.4056042192727693e-07,
|
| 3139 |
+
"loss": 0.1242,
|
| 3140 |
+
"step": 223000
|
| 3141 |
+
},
|
| 3142 |
+
{
|
| 3143 |
+
"epoch": 2.1291391990245017,
|
| 3144 |
+
"grad_norm": 12.82266902923584,
|
| 3145 |
+
"learning_rate": 1.4020346664831814e-07,
|
| 3146 |
+
"loss": 0.1066,
|
| 3147 |
+
"step": 223500
|
| 3148 |
+
},
|
| 3149 |
+
{
|
| 3150 |
+
"epoch": 2.133902373966391,
|
| 3151 |
+
"grad_norm": 0.15470212697982788,
|
| 3152 |
+
"learning_rate": 1.3984651136935935e-07,
|
| 3153 |
+
"loss": 0.0952,
|
| 3154 |
+
"step": 224000
|
| 3155 |
+
},
|
| 3156 |
+
{
|
| 3157 |
+
"epoch": 2.13866554890828,
|
| 3158 |
+
"grad_norm": 7.721962174400687e-05,
|
| 3159 |
+
"learning_rate": 1.3948955609040056e-07,
|
| 3160 |
+
"loss": 0.1237,
|
| 3161 |
+
"step": 224500
|
| 3162 |
+
},
|
| 3163 |
+
{
|
| 3164 |
+
"epoch": 2.1434287238501697,
|
| 3165 |
+
"grad_norm": 0.023135853931307793,
|
| 3166 |
+
"learning_rate": 1.3913260081144177e-07,
|
| 3167 |
+
"loss": 0.1118,
|
| 3168 |
+
"step": 225000
|
| 3169 |
+
},
|
| 3170 |
+
{
|
| 3171 |
+
"epoch": 2.1481918987920587,
|
| 3172 |
+
"grad_norm": 1.2089338302612305,
|
| 3173 |
+
"learning_rate": 1.3877564553248298e-07,
|
| 3174 |
+
"loss": 0.0944,
|
| 3175 |
+
"step": 225500
|
| 3176 |
+
},
|
| 3177 |
+
{
|
| 3178 |
+
"epoch": 2.1529550737339482,
|
| 3179 |
+
"grad_norm": 7.6081109046936035,
|
| 3180 |
+
"learning_rate": 1.384186902535242e-07,
|
| 3181 |
+
"loss": 0.1025,
|
| 3182 |
+
"step": 226000
|
| 3183 |
+
},
|
| 3184 |
+
{
|
| 3185 |
+
"epoch": 2.1577182486758373,
|
| 3186 |
+
"grad_norm": 11.101881980895996,
|
| 3187 |
+
"learning_rate": 1.380617349745654e-07,
|
| 3188 |
+
"loss": 0.081,
|
| 3189 |
+
"step": 226500
|
| 3190 |
}
|
| 3191 |
],
|
| 3192 |
"logging_steps": 500,
|
| 3193 |
+
"max_steps": 314916,
|
| 3194 |
"num_input_tokens_seen": 0,
|
| 3195 |
+
"num_train_epochs": 3,
|
| 3196 |
"save_steps": 500,
|
| 3197 |
"stateful_callbacks": {
|
| 3198 |
"TrainerControl": {
|
|
|
|
| 3201 |
"should_evaluate": false,
|
| 3202 |
"should_log": false,
|
| 3203 |
"should_save": true,
|
| 3204 |
+
"should_training_stop": false
|
| 3205 |
},
|
| 3206 |
"attributes": {}
|
| 3207 |
}
|
| 3208 |
},
|
| 3209 |
+
"total_flos": 3.2593962720815494e+17,
|
| 3210 |
"train_batch_size": 8,
|
| 3211 |
"trial_name": null,
|
| 3212 |
"trial_params": null
|
training_args.bin
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 sha256:3cbfdc468a2a0424d8e075f221e8a6abca4bc2ebfd47002af3dd1af0fd638d1d
|
| 3 |
+
size 5240
|