Training in progress, epoch 1, checkpoint
Browse files- last-checkpoint/adapter_model.safetensors +1 -1
- last-checkpoint/optimizer.pt +2 -2
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +1299 -4
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 97307544
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:567aaba69f3cb77d994ba0e0738cbad8878c63da061f21a21d0878bac0f67c8e
|
| 3 |
size 97307544
|
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 sha256:275085713fd0875dd909176096878c72d99476c333a1fe69547ff72b57fd8221
|
| 3 |
+
size 49846644
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eaee7c880dd5922ff2e7624ad963a49aa5a83b227e32cf8dfc2d6dff3f328fca
|
| 3 |
size 14960
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15024
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bf237a9746230efe192a37050e4413a37de7159a210eac68e5d8323d309c6dc5
|
| 3 |
size 15024
|
last-checkpoint/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15024
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:08dcbbe0d4a951d71d2b7bf4badaa3af2f1c5b8375d963fa60da7c9bd52deb90
|
| 3 |
size 15024
|
last-checkpoint/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15024
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f7024f178aa49a15b1aab78b10dd28c7859fa9128b3a6891ca2125db85464c91
|
| 3 |
size 15024
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bc7d6d18ada3aab0aa3019451acdf41b141b4e20aa0b1e0d83d96f7806f2a87d
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 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,
|
|
@@ -1512,6 +1512,1301 @@
|
|
| 1512 |
"learning_rate": 0.00010344732356263808,
|
| 1513 |
"loss": 0.8395,
|
| 1514 |
"step": 215
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1515 |
}
|
| 1516 |
],
|
| 1517 |
"logging_steps": 1,
|
|
@@ -1526,12 +2821,12 @@
|
|
| 1526 |
"should_evaluate": false,
|
| 1527 |
"should_log": false,
|
| 1528 |
"should_save": true,
|
| 1529 |
-
"should_training_stop":
|
| 1530 |
},
|
| 1531 |
"attributes": {}
|
| 1532 |
}
|
| 1533 |
},
|
| 1534 |
-
"total_flos":
|
| 1535 |
"train_batch_size": 6,
|
| 1536 |
"trial_name": null,
|
| 1537 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.8596001859600186,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 400,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1512 |
"learning_rate": 0.00010344732356263808,
|
| 1513 |
"loss": 0.8395,
|
| 1514 |
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.00418410041841,
|
| 1518 |
+
"grad_norm": 0.0985850915312767,
|
| 1519 |
+
"learning_rate": 0.00010272076036747365,
|
| 1520 |
+
"loss": 0.9939,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.00883310088331,
|
| 1525 |
+
"grad_norm": 0.06556614488363266,
|
| 1526 |
+
"learning_rate": 0.00010199467439373956,
|
| 1527 |
+
"loss": 1.0809,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.01348210134821,
|
| 1532 |
+
"grad_norm": 0.07044567912817001,
|
| 1533 |
+
"learning_rate": 0.00010126911324308168,
|
| 1534 |
+
"loss": 1.0733,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.0181311018131103,
|
| 1539 |
+
"grad_norm": 0.08298351615667343,
|
| 1540 |
+
"learning_rate": 0.00010054412448273886,
|
| 1541 |
+
"loss": 1.0562,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.0227801022780103,
|
| 1546 |
+
"grad_norm": 0.09130789339542389,
|
| 1547 |
+
"learning_rate": 9.981975564242443e-05,
|
| 1548 |
+
"loss": 1.0767,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.0274291027429103,
|
| 1553 |
+
"grad_norm": 0.10151571035385132,
|
| 1554 |
+
"learning_rate": 9.909605421121028e-05,
|
| 1555 |
+
"loss": 1.1066,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.0320781032078103,
|
| 1560 |
+
"grad_norm": 0.11376982927322388,
|
| 1561 |
+
"learning_rate": 9.837306763441345e-05,
|
| 1562 |
+
"loss": 1.0909,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.0367271036727104,
|
| 1567 |
+
"grad_norm": 0.14317406713962555,
|
| 1568 |
+
"learning_rate": 9.765084331048567e-05,
|
| 1569 |
+
"loss": 1.0539,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.0413761041376104,
|
| 1574 |
+
"grad_norm": 0.183615580201149,
|
| 1575 |
+
"learning_rate": 9.692942858790591e-05,
|
| 1576 |
+
"loss": 0.9804,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.0460251046025104,
|
| 1581 |
+
"grad_norm": 0.2267308384180069,
|
| 1582 |
+
"learning_rate": 9.620887076207632e-05,
|
| 1583 |
+
"loss": 0.9,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.0506741050674104,
|
| 1588 |
+
"grad_norm": 0.25844672322273254,
|
| 1589 |
+
"learning_rate": 9.548921707222163e-05,
|
| 1590 |
+
"loss": 0.7342,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.0553231055323105,
|
| 1595 |
+
"grad_norm": 0.35192665457725525,
|
| 1596 |
+
"learning_rate": 9.477051469829196e-05,
|
| 1597 |
+
"loss": 0.7048,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.0599721059972107,
|
| 1602 |
+
"grad_norm": 0.28442806005477905,
|
| 1603 |
+
"learning_rate": 9.405281075786995e-05,
|
| 1604 |
+
"loss": 0.7852,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.0646211064621107,
|
| 1609 |
+
"grad_norm": 0.06501670181751251,
|
| 1610 |
+
"learning_rate": 9.333615230308173e-05,
|
| 1611 |
+
"loss": 1.0592,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.0692701069270107,
|
| 1616 |
+
"grad_norm": 0.06888816505670547,
|
| 1617 |
+
"learning_rate": 9.26205863175121e-05,
|
| 1618 |
+
"loss": 1.0582,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.0739191073919108,
|
| 1623 |
+
"grad_norm": 0.07344524562358856,
|
| 1624 |
+
"learning_rate": 9.190615971312446e-05,
|
| 1625 |
+
"loss": 1.0434,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.0785681078568108,
|
| 1630 |
+
"grad_norm": 0.08665701746940613,
|
| 1631 |
+
"learning_rate": 9.119291932718525e-05,
|
| 1632 |
+
"loss": 1.0843,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.0832171083217108,
|
| 1637 |
+
"grad_norm": 0.09859387576580048,
|
| 1638 |
+
"learning_rate": 9.048091191919332e-05,
|
| 1639 |
+
"loss": 1.0871,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.0878661087866108,
|
| 1644 |
+
"grad_norm": 0.11464710533618927,
|
| 1645 |
+
"learning_rate": 8.97701841678144e-05,
|
| 1646 |
+
"loss": 1.0692,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.0925151092515109,
|
| 1651 |
+
"grad_norm": 0.14108416438102722,
|
| 1652 |
+
"learning_rate": 8.906078266782087e-05,
|
| 1653 |
+
"loss": 1.0655,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.0971641097164109,
|
| 1658 |
+
"grad_norm": 0.1800997108221054,
|
| 1659 |
+
"learning_rate": 8.835275392703721e-05,
|
| 1660 |
+
"loss": 1.0205,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.1018131101813111,
|
| 1665 |
+
"grad_norm": 0.2220492660999298,
|
| 1666 |
+
"learning_rate": 8.764614436329066e-05,
|
| 1667 |
+
"loss": 0.9219,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.1064621106462111,
|
| 1672 |
+
"grad_norm": 0.2564367353916168,
|
| 1673 |
+
"learning_rate": 8.694100030136849e-05,
|
| 1674 |
+
"loss": 0.7692,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.1111111111111112,
|
| 1679 |
+
"grad_norm": 0.2938039004802704,
|
| 1680 |
+
"learning_rate": 8.623736796998063e-05,
|
| 1681 |
+
"loss": 0.7331,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.1157601115760112,
|
| 1686 |
+
"grad_norm": 0.37189003825187683,
|
| 1687 |
+
"learning_rate": 8.553529349872916e-05,
|
| 1688 |
+
"loss": 0.5737,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.1204091120409112,
|
| 1693 |
+
"grad_norm": 0.21121616661548615,
|
| 1694 |
+
"learning_rate": 8.4834822915084e-05,
|
| 1695 |
+
"loss": 1.0279,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.1250581125058112,
|
| 1700 |
+
"grad_norm": 0.08440238237380981,
|
| 1701 |
+
"learning_rate": 8.413600214136548e-05,
|
| 1702 |
+
"loss": 1.039,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.1297071129707112,
|
| 1707 |
+
"grad_norm": 0.08833085745573044,
|
| 1708 |
+
"learning_rate": 8.343887699173356e-05,
|
| 1709 |
+
"loss": 1.065,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.1343561134356113,
|
| 1714 |
+
"grad_norm": 0.09478267282247543,
|
| 1715 |
+
"learning_rate": 8.274349316918446e-05,
|
| 1716 |
+
"loss": 1.0549,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 1.1390051139005113,
|
| 1721 |
+
"grad_norm": 0.1044965535402298,
|
| 1722 |
+
"learning_rate": 8.204989626255422e-05,
|
| 1723 |
+
"loss": 1.1372,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.1436541143654115,
|
| 1728 |
+
"grad_norm": 0.13615119457244873,
|
| 1729 |
+
"learning_rate": 8.135813174353008e-05,
|
| 1730 |
+
"loss": 1.1241,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.1483031148303116,
|
| 1735 |
+
"grad_norm": 0.1386057436466217,
|
| 1736 |
+
"learning_rate": 8.066824496366937e-05,
|
| 1737 |
+
"loss": 1.1088,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 1.1529521152952116,
|
| 1742 |
+
"grad_norm": 0.16772480309009552,
|
| 1743 |
+
"learning_rate": 7.998028115142617e-05,
|
| 1744 |
+
"loss": 1.0241,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.1576011157601116,
|
| 1749 |
+
"grad_norm": 0.2133951038122177,
|
| 1750 |
+
"learning_rate": 7.929428540918635e-05,
|
| 1751 |
+
"loss": 0.9239,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.1622501162250116,
|
| 1756 |
+
"grad_norm": 0.2830871343612671,
|
| 1757 |
+
"learning_rate": 7.86103027103106e-05,
|
| 1758 |
+
"loss": 0.8394,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 1.1668991166899116,
|
| 1763 |
+
"grad_norm": 0.29441362619400024,
|
| 1764 |
+
"learning_rate": 7.792837789618595e-05,
|
| 1765 |
+
"loss": 0.7265,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 1.1715481171548117,
|
| 1770 |
+
"grad_norm": 0.33453473448753357,
|
| 1771 |
+
"learning_rate": 7.724855567328613e-05,
|
| 1772 |
+
"loss": 0.6728,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 1.1761971176197117,
|
| 1777 |
+
"grad_norm": 0.3466179370880127,
|
| 1778 |
+
"learning_rate": 7.657088061024054e-05,
|
| 1779 |
+
"loss": 0.7098,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 1.1808461180846117,
|
| 1784 |
+
"grad_norm": 0.0637335404753685,
|
| 1785 |
+
"learning_rate": 7.58953971349124e-05,
|
| 1786 |
+
"loss": 1.0493,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.185495118549512,
|
| 1791 |
+
"grad_norm": 0.07033415138721466,
|
| 1792 |
+
"learning_rate": 7.522214953148618e-05,
|
| 1793 |
+
"loss": 1.04,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 1.190144119014412,
|
| 1798 |
+
"grad_norm": 0.07706312835216522,
|
| 1799 |
+
"learning_rate": 7.455118193756419e-05,
|
| 1800 |
+
"loss": 1.0632,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 1.194793119479312,
|
| 1805 |
+
"grad_norm": 0.09520713984966278,
|
| 1806 |
+
"learning_rate": 7.388253834127306e-05,
|
| 1807 |
+
"loss": 1.1191,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 1.199442119944212,
|
| 1812 |
+
"grad_norm": 0.10701565444469452,
|
| 1813 |
+
"learning_rate": 7.321626257837996e-05,
|
| 1814 |
+
"loss": 1.0896,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.204091120409112,
|
| 1819 |
+
"grad_norm": 0.11862656474113464,
|
| 1820 |
+
"learning_rate": 7.255239832941866e-05,
|
| 1821 |
+
"loss": 1.0893,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 1.208740120874012,
|
| 1826 |
+
"grad_norm": 0.1468852460384369,
|
| 1827 |
+
"learning_rate": 7.189098911682592e-05,
|
| 1828 |
+
"loss": 1.0755,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 1.213389121338912,
|
| 1833 |
+
"grad_norm": 0.18458393216133118,
|
| 1834 |
+
"learning_rate": 7.123207830208806e-05,
|
| 1835 |
+
"loss": 0.951,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 1.218038121803812,
|
| 1840 |
+
"grad_norm": 0.22184807062149048,
|
| 1841 |
+
"learning_rate": 7.05757090828986e-05,
|
| 1842 |
+
"loss": 0.8779,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 1.2226871222687121,
|
| 1847 |
+
"grad_norm": 0.3002167344093323,
|
| 1848 |
+
"learning_rate": 6.992192449032571e-05,
|
| 1849 |
+
"loss": 0.8529,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 1.2273361227336124,
|
| 1854 |
+
"grad_norm": 0.3212537169456482,
|
| 1855 |
+
"learning_rate": 6.927076738599152e-05,
|
| 1856 |
+
"loss": 0.6355,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.2319851231985124,
|
| 1861 |
+
"grad_norm": 0.3631160259246826,
|
| 1862 |
+
"learning_rate": 6.862228045926202e-05,
|
| 1863 |
+
"loss": 0.4947,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 1.2366341236634124,
|
| 1868 |
+
"grad_norm": 0.19289681315422058,
|
| 1869 |
+
"learning_rate": 6.797650622444836e-05,
|
| 1870 |
+
"loss": 0.9966,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 1.2412831241283124,
|
| 1875 |
+
"grad_norm": 0.07654353976249695,
|
| 1876 |
+
"learning_rate": 6.733348701801969e-05,
|
| 1877 |
+
"loss": 1.0453,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 1.2459321245932125,
|
| 1882 |
+
"grad_norm": 0.08304441720247269,
|
| 1883 |
+
"learning_rate": 6.669326499582755e-05,
|
| 1884 |
+
"loss": 1.0742,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 1.2505811250581125,
|
| 1889 |
+
"grad_norm": 0.0931682139635086,
|
| 1890 |
+
"learning_rate": 6.605588213034227e-05,
|
| 1891 |
+
"loss": 1.0973,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 1.2552301255230125,
|
| 1896 |
+
"grad_norm": 0.10273009538650513,
|
| 1897 |
+
"learning_rate": 6.542138020790116e-05,
|
| 1898 |
+
"loss": 1.0871,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.2598791259879125,
|
| 1903 |
+
"grad_norm": 0.1231117695569992,
|
| 1904 |
+
"learning_rate": 6.478980082596905e-05,
|
| 1905 |
+
"loss": 1.1266,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 1.2645281264528125,
|
| 1910 |
+
"grad_norm": 0.13809573650360107,
|
| 1911 |
+
"learning_rate": 6.416118539041135e-05,
|
| 1912 |
+
"loss": 1.0663,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 1.2691771269177128,
|
| 1917 |
+
"grad_norm": 0.17248062789440155,
|
| 1918 |
+
"learning_rate": 6.353557511277928e-05,
|
| 1919 |
+
"loss": 1.0427,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 1.2738261273826128,
|
| 1924 |
+
"grad_norm": 0.22152015566825867,
|
| 1925 |
+
"learning_rate": 6.291301100760829e-05,
|
| 1926 |
+
"loss": 0.9226,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 1.2784751278475128,
|
| 1931 |
+
"grad_norm": 0.2788388729095459,
|
| 1932 |
+
"learning_rate": 6.229353388972901e-05,
|
| 1933 |
+
"loss": 0.8773,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 1.2831241283124128,
|
| 1938 |
+
"grad_norm": 0.30847859382629395,
|
| 1939 |
+
"learning_rate": 6.167718437159147e-05,
|
| 1940 |
+
"loss": 0.7012,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 1.2877731287773129,
|
| 1945 |
+
"grad_norm": 0.3570963442325592,
|
| 1946 |
+
"learning_rate": 6.106400286060274e-05,
|
| 1947 |
+
"loss": 0.6076,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 1.292422129242213,
|
| 1952 |
+
"grad_norm": 0.3086980879306793,
|
| 1953 |
+
"learning_rate": 6.045402955647769e-05,
|
| 1954 |
+
"loss": 0.7205,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 1.297071129707113,
|
| 1959 |
+
"grad_norm": 0.0752180814743042,
|
| 1960 |
+
"learning_rate": 5.98473044486036e-05,
|
| 1961 |
+
"loss": 1.0282,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 1.301720130172013,
|
| 1966 |
+
"grad_norm": 0.07571510970592499,
|
| 1967 |
+
"learning_rate": 5.924386731341842e-05,
|
| 1968 |
+
"loss": 1.0712,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 1.306369130636913,
|
| 1973 |
+
"grad_norm": 0.08388427644968033,
|
| 1974 |
+
"learning_rate": 5.864375771180317e-05,
|
| 1975 |
+
"loss": 1.0425,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 1.3110181311018132,
|
| 1980 |
+
"grad_norm": 0.0906265452504158,
|
| 1981 |
+
"learning_rate": 5.804701498648828e-05,
|
| 1982 |
+
"loss": 1.0463,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 1.3156671315667132,
|
| 1987 |
+
"grad_norm": 0.10397284477949142,
|
| 1988 |
+
"learning_rate": 5.7453678259474234e-05,
|
| 1989 |
+
"loss": 1.0913,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 1.3203161320316132,
|
| 1994 |
+
"grad_norm": 0.12175474315881729,
|
| 1995 |
+
"learning_rate": 5.686378642946699e-05,
|
| 1996 |
+
"loss": 1.0896,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 1.3249651324965133,
|
| 2001 |
+
"grad_norm": 0.14317680895328522,
|
| 2002 |
+
"learning_rate": 5.627737816932754e-05,
|
| 2003 |
+
"loss": 1.0503,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 1.3296141329614133,
|
| 2008 |
+
"grad_norm": 0.19962027668952942,
|
| 2009 |
+
"learning_rate": 5.569449192353678e-05,
|
| 2010 |
+
"loss": 1.0168,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 1.3342631334263133,
|
| 2015 |
+
"grad_norm": 0.253571093082428,
|
| 2016 |
+
"learning_rate": 5.511516590567499e-05,
|
| 2017 |
+
"loss": 0.8714,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 1.3389121338912133,
|
| 2022 |
+
"grad_norm": 0.2986457049846649,
|
| 2023 |
+
"learning_rate": 5.453943809591654e-05,
|
| 2024 |
+
"loss": 0.7573,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 1.3435611343561136,
|
| 2029 |
+
"grad_norm": 0.37284067273139954,
|
| 2030 |
+
"learning_rate": 5.396734623854012e-05,
|
| 2031 |
+
"loss": 0.6708,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 1.3482101348210134,
|
| 2036 |
+
"grad_norm": 0.4108807444572449,
|
| 2037 |
+
"learning_rate": 5.3398927839453996e-05,
|
| 2038 |
+
"loss": 0.549,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 1.3528591352859136,
|
| 2043 |
+
"grad_norm": 0.21761813759803772,
|
| 2044 |
+
"learning_rate": 5.283422016373745e-05,
|
| 2045 |
+
"loss": 0.9454,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 1.3575081357508136,
|
| 2050 |
+
"grad_norm": 0.07017389684915543,
|
| 2051 |
+
"learning_rate": 5.227326023319743e-05,
|
| 2052 |
+
"loss": 1.06,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 1.3621571362157137,
|
| 2057 |
+
"grad_norm": 0.07673922181129456,
|
| 2058 |
+
"learning_rate": 5.17160848239416e-05,
|
| 2059 |
+
"loss": 1.061,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 1.3668061366806137,
|
| 2064 |
+
"grad_norm": 0.08843539655208588,
|
| 2065 |
+
"learning_rate": 5.1162730463967304e-05,
|
| 2066 |
+
"loss": 1.0668,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 1.3714551371455137,
|
| 2071 |
+
"grad_norm": 0.10018595308065414,
|
| 2072 |
+
"learning_rate": 5.061323343076672e-05,
|
| 2073 |
+
"loss": 1.0926,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 1.3761041376104137,
|
| 2078 |
+
"grad_norm": 0.11475757509469986,
|
| 2079 |
+
"learning_rate": 5.006762974894872e-05,
|
| 2080 |
+
"loss": 1.0878,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 1.3807531380753137,
|
| 2085 |
+
"grad_norm": 0.14695028960704803,
|
| 2086 |
+
"learning_rate": 4.9525955187876885e-05,
|
| 2087 |
+
"loss": 1.088,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 1.385402138540214,
|
| 2092 |
+
"grad_norm": 0.18600180745124817,
|
| 2093 |
+
"learning_rate": 4.898824525932471e-05,
|
| 2094 |
+
"loss": 1.0019,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 1.3900511390051138,
|
| 2099 |
+
"grad_norm": 0.21412943303585052,
|
| 2100 |
+
"learning_rate": 4.845453521514738e-05,
|
| 2101 |
+
"loss": 0.9161,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 1.394700139470014,
|
| 2106 |
+
"grad_norm": 0.2769903838634491,
|
| 2107 |
+
"learning_rate": 4.7924860044970615e-05,
|
| 2108 |
+
"loss": 0.8767,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 1.399349139934914,
|
| 2113 |
+
"grad_norm": 0.37584012746810913,
|
| 2114 |
+
"learning_rate": 4.739925447389698e-05,
|
| 2115 |
+
"loss": 0.8159,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 1.403998140399814,
|
| 2120 |
+
"grad_norm": 0.35033103823661804,
|
| 2121 |
+
"learning_rate": 4.687775296022908e-05,
|
| 2122 |
+
"loss": 0.5912,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 1.408647140864714,
|
| 2127 |
+
"grad_norm": 0.3620702624320984,
|
| 2128 |
+
"learning_rate": 4.6360389693210735e-05,
|
| 2129 |
+
"loss": 0.7819,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 1.4132961413296141,
|
| 2134 |
+
"grad_norm": 0.06753652542829514,
|
| 2135 |
+
"learning_rate": 4.5847198590785394e-05,
|
| 2136 |
+
"loss": 1.023,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 1.4179451417945141,
|
| 2141 |
+
"grad_norm": 0.07456682622432709,
|
| 2142 |
+
"learning_rate": 4.5338213297372534e-05,
|
| 2143 |
+
"loss": 1.0321,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 1.4225941422594142,
|
| 2148 |
+
"grad_norm": 0.08220840245485306,
|
| 2149 |
+
"learning_rate": 4.4833467181662086e-05,
|
| 2150 |
+
"loss": 1.0518,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 1.4272431427243144,
|
| 2155 |
+
"grad_norm": 0.09785965085029602,
|
| 2156 |
+
"learning_rate": 4.4332993334426576e-05,
|
| 2157 |
+
"loss": 1.0736,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 1.4318921431892142,
|
| 2162 |
+
"grad_norm": 0.10714032500982285,
|
| 2163 |
+
"learning_rate": 4.383682456635199e-05,
|
| 2164 |
+
"loss": 1.0766,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 1.4365411436541144,
|
| 2169 |
+
"grad_norm": 0.12374605983495712,
|
| 2170 |
+
"learning_rate": 4.3344993405886425e-05,
|
| 2171 |
+
"loss": 1.1013,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 1.4411901441190145,
|
| 2176 |
+
"grad_norm": 0.15557971596717834,
|
| 2177 |
+
"learning_rate": 4.285753209710786e-05,
|
| 2178 |
+
"loss": 1.0673,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 1.4458391445839145,
|
| 2183 |
+
"grad_norm": 0.1834760308265686,
|
| 2184 |
+
"learning_rate": 4.2374472597610044e-05,
|
| 2185 |
+
"loss": 0.9611,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 1.4504881450488145,
|
| 2190 |
+
"grad_norm": 0.24258282780647278,
|
| 2191 |
+
"learning_rate": 4.1895846576407424e-05,
|
| 2192 |
+
"loss": 0.9162,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.4551371455137145,
|
| 2197 |
+
"grad_norm": 0.31023555994033813,
|
| 2198 |
+
"learning_rate": 4.1421685411859046e-05,
|
| 2199 |
+
"loss": 0.8321,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 1.4597861459786146,
|
| 2204 |
+
"grad_norm": 0.3393552601337433,
|
| 2205 |
+
"learning_rate": 4.095202018961125e-05,
|
| 2206 |
+
"loss": 0.6979,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 1.4644351464435146,
|
| 2211 |
+
"grad_norm": 0.3873143792152405,
|
| 2212 |
+
"learning_rate": 4.048688170055989e-05,
|
| 2213 |
+
"loss": 0.5101,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 1.4690841469084148,
|
| 2218 |
+
"grad_norm": 0.20240376889705658,
|
| 2219 |
+
"learning_rate": 4.002630043883159e-05,
|
| 2220 |
+
"loss": 1.0254,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 1.4737331473733146,
|
| 2225 |
+
"grad_norm": 0.08021257072687149,
|
| 2226 |
+
"learning_rate": 3.9570306599784544e-05,
|
| 2227 |
+
"loss": 1.033,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 1.4783821478382149,
|
| 2232 |
+
"grad_norm": 0.079825259745121,
|
| 2233 |
+
"learning_rate": 3.911893007802913e-05,
|
| 2234 |
+
"loss": 1.0105,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 1.4830311483031149,
|
| 2239 |
+
"grad_norm": 0.09591138362884521,
|
| 2240 |
+
"learning_rate": 3.8672200465467765e-05,
|
| 2241 |
+
"loss": 1.0867,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 1.487680148768015,
|
| 2246 |
+
"grad_norm": 0.10192592442035675,
|
| 2247 |
+
"learning_rate": 3.8230147049355147e-05,
|
| 2248 |
+
"loss": 1.1238,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 1.492329149232915,
|
| 2253 |
+
"grad_norm": 0.11998689919710159,
|
| 2254 |
+
"learning_rate": 3.779279881037797e-05,
|
| 2255 |
+
"loss": 1.099,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 1.496978149697815,
|
| 2260 |
+
"grad_norm": 0.14268672466278076,
|
| 2261 |
+
"learning_rate": 3.7360184420755165e-05,
|
| 2262 |
+
"loss": 1.0832,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 1.501627150162715,
|
| 2267 |
+
"grad_norm": 0.1779240220785141,
|
| 2268 |
+
"learning_rate": 3.693233224235806e-05,
|
| 2269 |
+
"loss": 1.0322,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 1.506276150627615,
|
| 2274 |
+
"grad_norm": 0.23235994577407837,
|
| 2275 |
+
"learning_rate": 3.650927032485101e-05,
|
| 2276 |
+
"loss": 0.987,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 1.5109251510925152,
|
| 2281 |
+
"grad_norm": 0.2786919176578522,
|
| 2282 |
+
"learning_rate": 3.609102640385254e-05,
|
| 2283 |
+
"loss": 0.7974,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 1.515574151557415,
|
| 2288 |
+
"grad_norm": 0.3645856976509094,
|
| 2289 |
+
"learning_rate": 3.567762789911693e-05,
|
| 2290 |
+
"loss": 0.7208,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 1.5202231520223153,
|
| 2295 |
+
"grad_norm": 0.4197288751602173,
|
| 2296 |
+
"learning_rate": 3.526910191273665e-05,
|
| 2297 |
+
"loss": 0.5941,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 1.524872152487215,
|
| 2302 |
+
"grad_norm": 0.43531715869903564,
|
| 2303 |
+
"learning_rate": 3.486547522736562e-05,
|
| 2304 |
+
"loss": 0.8169,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 1.5295211529521153,
|
| 2309 |
+
"grad_norm": 0.07292664051055908,
|
| 2310 |
+
"learning_rate": 3.44667743044632e-05,
|
| 2311 |
+
"loss": 1.026,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 1.5341701534170153,
|
| 2316 |
+
"grad_norm": 0.07665824145078659,
|
| 2317 |
+
"learning_rate": 3.407302528255961e-05,
|
| 2318 |
+
"loss": 1.0711,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 1.5388191538819154,
|
| 2323 |
+
"grad_norm": 0.09035459905862808,
|
| 2324 |
+
"learning_rate": 3.36842539755421e-05,
|
| 2325 |
+
"loss": 1.058,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 1.5434681543468154,
|
| 2330 |
+
"grad_norm": 0.10302021354436874,
|
| 2331 |
+
"learning_rate": 3.3300485870962776e-05,
|
| 2332 |
+
"loss": 1.0717,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 1.5481171548117154,
|
| 2337 |
+
"grad_norm": 0.11944916844367981,
|
| 2338 |
+
"learning_rate": 3.292174612836757e-05,
|
| 2339 |
+
"loss": 1.1738,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 1.5527661552766157,
|
| 2344 |
+
"grad_norm": 0.1299082338809967,
|
| 2345 |
+
"learning_rate": 3.254805957764673e-05,
|
| 2346 |
+
"loss": 1.1078,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 1.5574151557415155,
|
| 2351 |
+
"grad_norm": 0.15316098928451538,
|
| 2352 |
+
"learning_rate": 3.217945071740724e-05,
|
| 2353 |
+
"loss": 1.0111,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 1.5620641562064157,
|
| 2358 |
+
"grad_norm": 0.19828684628009796,
|
| 2359 |
+
"learning_rate": 3.1815943713366404e-05,
|
| 2360 |
+
"loss": 1.0292,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 1.5667131566713157,
|
| 2365 |
+
"grad_norm": 0.2431751936674118,
|
| 2366 |
+
"learning_rate": 3.145756239676779e-05,
|
| 2367 |
+
"loss": 0.867,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 1.5713621571362157,
|
| 2372 |
+
"grad_norm": 0.3115411698818207,
|
| 2373 |
+
"learning_rate": 3.110433026281872e-05,
|
| 2374 |
+
"loss": 0.7307,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 1.5760111576011158,
|
| 2379 |
+
"grad_norm": 0.37397709488868713,
|
| 2380 |
+
"learning_rate": 3.075627046915003e-05,
|
| 2381 |
+
"loss": 0.6894,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 1.5806601580660158,
|
| 2386 |
+
"grad_norm": 0.4691689610481262,
|
| 2387 |
+
"learning_rate": 3.041340583429789e-05,
|
| 2388 |
+
"loss": 0.6568,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 1.5853091585309158,
|
| 2393 |
+
"grad_norm": 0.27498671412467957,
|
| 2394 |
+
"learning_rate": 3.0075758836207716e-05,
|
| 2395 |
+
"loss": 1.0342,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 1.5899581589958158,
|
| 2400 |
+
"grad_norm": 0.07807187736034393,
|
| 2401 |
+
"learning_rate": 2.9743351610760716e-05,
|
| 2402 |
+
"loss": 1.0501,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 1.594607159460716,
|
| 2407 |
+
"grad_norm": 0.08518065512180328,
|
| 2408 |
+
"learning_rate": 2.941620595032246e-05,
|
| 2409 |
+
"loss": 1.0657,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 1.5992561599256159,
|
| 2414 |
+
"grad_norm": 0.09096917510032654,
|
| 2415 |
+
"learning_rate": 2.9094343302314432e-05,
|
| 2416 |
+
"loss": 1.0534,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 1.6039051603905161,
|
| 2421 |
+
"grad_norm": 0.10106653720140457,
|
| 2422 |
+
"learning_rate": 2.8777784767807727e-05,
|
| 2423 |
+
"loss": 1.0949,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 1.6085541608554161,
|
| 2428 |
+
"grad_norm": 0.11770515143871307,
|
| 2429 |
+
"learning_rate": 2.846655110013978e-05,
|
| 2430 |
+
"loss": 1.1013,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 1.6132031613203162,
|
| 2435 |
+
"grad_norm": 0.14044031500816345,
|
| 2436 |
+
"learning_rate": 2.816066270355391e-05,
|
| 2437 |
+
"loss": 1.0657,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 1.6178521617852162,
|
| 2442 |
+
"grad_norm": 0.1705469787120819,
|
| 2443 |
+
"learning_rate": 2.78601396318614e-05,
|
| 2444 |
+
"loss": 1.0108,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 1.6225011622501162,
|
| 2449 |
+
"grad_norm": 0.23262286186218262,
|
| 2450 |
+
"learning_rate": 2.7565001587126922e-05,
|
| 2451 |
+
"loss": 0.9104,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 1.6271501627150162,
|
| 2456 |
+
"grad_norm": 0.3084715008735657,
|
| 2457 |
+
"learning_rate": 2.7275267918376912e-05,
|
| 2458 |
+
"loss": 0.8493,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 1.6317991631799162,
|
| 2463 |
+
"grad_norm": 0.34869176149368286,
|
| 2464 |
+
"learning_rate": 2.6990957620330954e-05,
|
| 2465 |
+
"loss": 0.7103,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 1.6364481636448165,
|
| 2470 |
+
"grad_norm": 0.40754231810569763,
|
| 2471 |
+
"learning_rate": 2.6712089332156633e-05,
|
| 2472 |
+
"loss": 0.606,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 1.6410971641097163,
|
| 2477 |
+
"grad_norm": 0.38754644989967346,
|
| 2478 |
+
"learning_rate": 2.6438681336247417e-05,
|
| 2479 |
+
"loss": 0.7465,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 1.6457461645746165,
|
| 2484 |
+
"grad_norm": 0.07341364026069641,
|
| 2485 |
+
"learning_rate": 2.6170751557024197e-05,
|
| 2486 |
+
"loss": 1.0565,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 1.6503951650395166,
|
| 2491 |
+
"grad_norm": 0.08013252913951874,
|
| 2492 |
+
"learning_rate": 2.5908317559760138e-05,
|
| 2493 |
+
"loss": 1.0146,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 1.6550441655044166,
|
| 2498 |
+
"grad_norm": 0.09281094372272491,
|
| 2499 |
+
"learning_rate": 2.5651396549429086e-05,
|
| 2500 |
+
"loss": 1.0903,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 1.6596931659693166,
|
| 2505 |
+
"grad_norm": 0.09770162403583527,
|
| 2506 |
+
"learning_rate": 2.540000536957765e-05,
|
| 2507 |
+
"loss": 1.1007,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 1.6643421664342166,
|
| 2512 |
+
"grad_norm": 0.11755497008562088,
|
| 2513 |
+
"learning_rate": 2.515416050122092e-05,
|
| 2514 |
+
"loss": 1.1027,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 1.6689911668991166,
|
| 2519 |
+
"grad_norm": 0.1314418464899063,
|
| 2520 |
+
"learning_rate": 2.4913878061762094e-05,
|
| 2521 |
+
"loss": 1.0532,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 1.6736401673640167,
|
| 2526 |
+
"grad_norm": 0.1723410040140152,
|
| 2527 |
+
"learning_rate": 2.4679173803935662e-05,
|
| 2528 |
+
"loss": 1.0447,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 1.678289167828917,
|
| 2533 |
+
"grad_norm": 0.22374863922595978,
|
| 2534 |
+
"learning_rate": 2.4450063114774784e-05,
|
| 2535 |
+
"loss": 1.0271,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 1.6829381682938167,
|
| 2540 |
+
"grad_norm": 0.2649782598018646,
|
| 2541 |
+
"learning_rate": 2.4226561014602522e-05,
|
| 2542 |
+
"loss": 0.8288,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 1.687587168758717,
|
| 2547 |
+
"grad_norm": 0.3294927477836609,
|
| 2548 |
+
"learning_rate": 2.400868215604706e-05,
|
| 2549 |
+
"loss": 0.7618,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 1.692236169223617,
|
| 2554 |
+
"grad_norm": 0.390989750623703,
|
| 2555 |
+
"learning_rate": 2.3796440823081167e-05,
|
| 2556 |
+
"loss": 0.6498,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 1.696885169688517,
|
| 2561 |
+
"grad_norm": 0.47515323758125305,
|
| 2562 |
+
"learning_rate": 2.358985093008566e-05,
|
| 2563 |
+
"loss": 0.6206,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 1.701534170153417,
|
| 2568 |
+
"grad_norm": 0.18342038989067078,
|
| 2569 |
+
"learning_rate": 2.3388926020937286e-05,
|
| 2570 |
+
"loss": 0.9768,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 1.706183170618317,
|
| 2575 |
+
"grad_norm": 0.08224090933799744,
|
| 2576 |
+
"learning_rate": 2.3193679268120718e-05,
|
| 2577 |
+
"loss": 1.0393,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 1.710832171083217,
|
| 2582 |
+
"grad_norm": 0.08616173267364502,
|
| 2583 |
+
"learning_rate": 2.3004123471865e-05,
|
| 2584 |
+
"loss": 1.0021,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 1.715481171548117,
|
| 2589 |
+
"grad_norm": 0.08873171359300613,
|
| 2590 |
+
"learning_rate": 2.2820271059304412e-05,
|
| 2591 |
+
"loss": 1.066,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 1.7201301720130173,
|
| 2596 |
+
"grad_norm": 0.10707499831914902,
|
| 2597 |
+
"learning_rate": 2.2642134083663678e-05,
|
| 2598 |
+
"loss": 1.1172,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 1.7247791724779171,
|
| 2603 |
+
"grad_norm": 0.1225440576672554,
|
| 2604 |
+
"learning_rate": 2.2469724223467866e-05,
|
| 2605 |
+
"loss": 1.0865,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 1.7294281729428174,
|
| 2610 |
+
"grad_norm": 0.1396765112876892,
|
| 2611 |
+
"learning_rate": 2.2303052781776664e-05,
|
| 2612 |
+
"loss": 1.0273,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 1.7340771734077174,
|
| 2617 |
+
"grad_norm": 0.17938072979450226,
|
| 2618 |
+
"learning_rate": 2.2142130685443382e-05,
|
| 2619 |
+
"loss": 1.1046,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 1.7387261738726174,
|
| 2624 |
+
"grad_norm": 0.2324485331773758,
|
| 2625 |
+
"learning_rate": 2.198696848439865e-05,
|
| 2626 |
+
"loss": 0.9628,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 1.7433751743375174,
|
| 2631 |
+
"grad_norm": 0.2933482527732849,
|
| 2632 |
+
"learning_rate": 2.1837576350958686e-05,
|
| 2633 |
+
"loss": 0.8228,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 1.7480241748024175,
|
| 2638 |
+
"grad_norm": 0.3423949182033539,
|
| 2639 |
+
"learning_rate": 2.169396407915849e-05,
|
| 2640 |
+
"loss": 0.7043,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 1.7526731752673175,
|
| 2645 |
+
"grad_norm": 0.38756224513053894,
|
| 2646 |
+
"learning_rate": 2.155614108410968e-05,
|
| 2647 |
+
"loss": 0.6579,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 1.7573221757322175,
|
| 2652 |
+
"grad_norm": 0.4097736179828644,
|
| 2653 |
+
"learning_rate": 2.142411640138332e-05,
|
| 2654 |
+
"loss": 0.7569,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 1.7619711761971177,
|
| 2659 |
+
"grad_norm": 0.07299873232841492,
|
| 2660 |
+
"learning_rate": 2.129789868641749e-05,
|
| 2661 |
+
"loss": 1.0571,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 1.7666201766620175,
|
| 2666 |
+
"grad_norm": 0.07895659655332565,
|
| 2667 |
+
"learning_rate": 2.1177496213949837e-05,
|
| 2668 |
+
"loss": 1.0487,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 1.7712691771269178,
|
| 2673 |
+
"grad_norm": 0.09031179547309875,
|
| 2674 |
+
"learning_rate": 2.1062916877475198e-05,
|
| 2675 |
+
"loss": 1.0911,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 1.7759181775918178,
|
| 2680 |
+
"grad_norm": 0.09883815795183182,
|
| 2681 |
+
"learning_rate": 2.0954168188727962e-05,
|
| 2682 |
+
"loss": 1.0619,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 1.7805671780567178,
|
| 2687 |
+
"grad_norm": 0.11488756537437439,
|
| 2688 |
+
"learning_rate": 2.0851257277189703e-05,
|
| 2689 |
+
"loss": 1.0788,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 1.7852161785216178,
|
| 2694 |
+
"grad_norm": 0.14409998059272766,
|
| 2695 |
+
"learning_rate": 2.0754190889621745e-05,
|
| 2696 |
+
"loss": 1.0902,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 1.7898651789865179,
|
| 2701 |
+
"grad_norm": 0.1650954782962799,
|
| 2702 |
+
"learning_rate": 2.0662975389622843e-05,
|
| 2703 |
+
"loss": 0.999,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 1.794514179451418,
|
| 2708 |
+
"grad_norm": 0.22679570317268372,
|
| 2709 |
+
"learning_rate": 2.0577616757212016e-05,
|
| 2710 |
+
"loss": 0.9647,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 1.799163179916318,
|
| 2715 |
+
"grad_norm": 0.284229040145874,
|
| 2716 |
+
"learning_rate": 2.0498120588436466e-05,
|
| 2717 |
+
"loss": 0.85,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 1.8038121803812182,
|
| 2722 |
+
"grad_norm": 0.32825493812561035,
|
| 2723 |
+
"learning_rate": 2.0424492095004746e-05,
|
| 2724 |
+
"loss": 0.7025,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 1.808461180846118,
|
| 2729 |
+
"grad_norm": 0.37925466895103455,
|
| 2730 |
+
"learning_rate": 2.0356736103945047e-05,
|
| 2731 |
+
"loss": 0.6595,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 1.8131101813110182,
|
| 2736 |
+
"grad_norm": 0.45006221532821655,
|
| 2737 |
+
"learning_rate": 2.029485705728876e-05,
|
| 2738 |
+
"loss": 0.4914,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 1.8177591817759182,
|
| 2743 |
+
"grad_norm": 0.2308957725763321,
|
| 2744 |
+
"learning_rate": 2.023885901177926e-05,
|
| 2745 |
+
"loss": 1.0114,
|
| 2746 |
+
"step": 391
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 1.8224081822408182,
|
| 2750 |
+
"grad_norm": 0.07487577944993973,
|
| 2751 |
+
"learning_rate": 2.0188745638605954e-05,
|
| 2752 |
+
"loss": 1.0175,
|
| 2753 |
+
"step": 392
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 1.8270571827057183,
|
| 2757 |
+
"grad_norm": 0.07909037917852402,
|
| 2758 |
+
"learning_rate": 2.014452022316358e-05,
|
| 2759 |
+
"loss": 1.0365,
|
| 2760 |
+
"step": 393
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 1.8317061831706183,
|
| 2764 |
+
"grad_norm": 0.09483642131090164,
|
| 2765 |
+
"learning_rate": 2.010618566483684e-05,
|
| 2766 |
+
"loss": 1.0723,
|
| 2767 |
+
"step": 394
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 1.8363551836355185,
|
| 2771 |
+
"grad_norm": 0.1127537190914154,
|
| 2772 |
+
"learning_rate": 2.00737444768103e-05,
|
| 2773 |
+
"loss": 1.0905,
|
| 2774 |
+
"step": 395
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 1.8410041841004183,
|
| 2778 |
+
"grad_norm": 0.12364522367715836,
|
| 2779 |
+
"learning_rate": 2.0047198785903658e-05,
|
| 2780 |
+
"loss": 1.1045,
|
| 2781 |
+
"step": 396
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 1.8456531845653186,
|
| 2785 |
+
"grad_norm": 0.14588101208209991,
|
| 2786 |
+
"learning_rate": 2.002655033243228e-05,
|
| 2787 |
+
"loss": 1.0753,
|
| 2788 |
+
"step": 397
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 1.8503021850302184,
|
| 2792 |
+
"grad_norm": 0.196581169962883,
|
| 2793 |
+
"learning_rate": 2.0011800470093105e-05,
|
| 2794 |
+
"loss": 1.0165,
|
| 2795 |
+
"step": 398
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 1.8549511854951186,
|
| 2799 |
+
"grad_norm": 0.24468840658664703,
|
| 2800 |
+
"learning_rate": 2.0002950165875934e-05,
|
| 2801 |
+
"loss": 0.9067,
|
| 2802 |
+
"step": 399
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 1.8596001859600186,
|
| 2806 |
+
"grad_norm": 0.32478219270706177,
|
| 2807 |
+
"learning_rate": 2e-05,
|
| 2808 |
+
"loss": 0.8752,
|
| 2809 |
+
"step": 400
|
| 2810 |
}
|
| 2811 |
],
|
| 2812 |
"logging_steps": 1,
|
|
|
|
| 2821 |
"should_evaluate": false,
|
| 2822 |
"should_log": false,
|
| 2823 |
"should_save": true,
|
| 2824 |
+
"should_training_stop": true
|
| 2825 |
},
|
| 2826 |
"attributes": {}
|
| 2827 |
}
|
| 2828 |
},
|
| 2829 |
+
"total_flos": 3.3538115420584673e+18,
|
| 2830 |
"train_batch_size": 6,
|
| 2831 |
"trial_name": null,
|
| 2832 |
"trial_params": null
|