nadahlberg
commited on
Commit
•
e80a0af
1
Parent(s):
0c19e85
Training in progress, step 6000, checkpoint
Browse files- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/optimizer.pt +1 -1
- 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/rng_state_4.pth +1 -1
- last-checkpoint/rng_state_5.pth +1 -1
- last-checkpoint/rng_state_6.pth +1 -1
- last-checkpoint/rng_state_7.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +1411 -3
last-checkpoint/model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 325690872
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1a9324f1d5bbfda5c5ef464f5ae538c67609b9809f4bb9e7b5ff3d7efd2a04d0
|
3 |
size 325690872
|
last-checkpoint/optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 651550778
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:801f28773de7283c3987b1e383caa6f1e78eb2516d24634a1b704f529ee08071
|
3 |
size 651550778
|
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 15920
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5ff22c72cb50d4b6353cf336b950d2c54115c739606544de9a8d3b0fab0ef188
|
3 |
size 15920
|
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 15920
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c253ad05d3ffa98dc1e5291ec640e6158218602254206e6fb97ea82185040775
|
3 |
size 15920
|
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 15920
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:edcdbb53b869133655de3ceb14325bb161e7aa238640244125393ff4afcb7363
|
3 |
size 15920
|
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 15920
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b095e9807a468e05241a1d389fd326265bee89778655d3c00298c2f6abcee791
|
3 |
size 15920
|
last-checkpoint/rng_state_4.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 15920
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a4da3805c54028ef89f9cce32b640836c54b432b251fae2dea3d0182fa96f403
|
3 |
size 15920
|
last-checkpoint/rng_state_5.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 15920
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:08bb3c8d2c1053a6b93a1342acb589ea4695b3fdb4dfeaf3675191c7e3390d3d
|
3 |
size 15920
|
last-checkpoint/rng_state_6.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 15920
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3cfebda0f5a673869a40a8a82d7e308c843f9b55ec287861a6a0670b4c43c9d4
|
3 |
size 15920
|
last-checkpoint/rng_state_7.pth
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 15920
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:34bc6aae481e0bf889f93cc019a7692fbdaba7925f341ec202dd711aa1d2a724
|
3 |
size 15920
|
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 sha256:7aacbaeb2fb166ed7abcf1ad1b0f61705e2164e64d30abdff36a13f020686241
|
3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
@@ -1,9 +1,9 @@
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
-
"epoch": 0.
|
5 |
"eval_steps": 2000,
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
@@ -2823,6 +2823,1414 @@
|
|
2823 |
"eval_samples_per_second": 28.059,
|
2824 |
"eval_steps_per_second": 0.449,
|
2825 |
"step": 4000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2826 |
}
|
2827 |
],
|
2828 |
"logging_steps": 10,
|
@@ -2842,7 +4250,7 @@
|
|
2842 |
"attributes": {}
|
2843 |
}
|
2844 |
},
|
2845 |
-
"total_flos":
|
2846 |
"train_batch_size": 64,
|
2847 |
"trial_name": null,
|
2848 |
"trial_params": null
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.03,
|
5 |
"eval_steps": 2000,
|
6 |
+
"global_step": 6000,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
|
|
2823 |
"eval_samples_per_second": 28.059,
|
2824 |
"eval_steps_per_second": 0.449,
|
2825 |
"step": 4000
|
2826 |
+
},
|
2827 |
+
{
|
2828 |
+
"epoch": 0.02005,
|
2829 |
+
"grad_norm": 31.375,
|
2830 |
+
"learning_rate": 4.9243718592964825e-05,
|
2831 |
+
"loss": 7.56,
|
2832 |
+
"step": 4010
|
2833 |
+
},
|
2834 |
+
{
|
2835 |
+
"epoch": 0.0201,
|
2836 |
+
"grad_norm": 1.28125,
|
2837 |
+
"learning_rate": 4.9241206030150756e-05,
|
2838 |
+
"loss": 7.5483,
|
2839 |
+
"step": 4020
|
2840 |
+
},
|
2841 |
+
{
|
2842 |
+
"epoch": 0.02015,
|
2843 |
+
"grad_norm": 7.59375,
|
2844 |
+
"learning_rate": 4.923869346733669e-05,
|
2845 |
+
"loss": 7.5583,
|
2846 |
+
"step": 4030
|
2847 |
+
},
|
2848 |
+
{
|
2849 |
+
"epoch": 0.0202,
|
2850 |
+
"grad_norm": 3.21875,
|
2851 |
+
"learning_rate": 4.923618090452262e-05,
|
2852 |
+
"loss": 7.5831,
|
2853 |
+
"step": 4040
|
2854 |
+
},
|
2855 |
+
{
|
2856 |
+
"epoch": 0.02025,
|
2857 |
+
"grad_norm": 3.125,
|
2858 |
+
"learning_rate": 4.923366834170854e-05,
|
2859 |
+
"loss": 7.5884,
|
2860 |
+
"step": 4050
|
2861 |
+
},
|
2862 |
+
{
|
2863 |
+
"epoch": 0.0203,
|
2864 |
+
"grad_norm": 1.625,
|
2865 |
+
"learning_rate": 4.9231155778894474e-05,
|
2866 |
+
"loss": 7.5728,
|
2867 |
+
"step": 4060
|
2868 |
+
},
|
2869 |
+
{
|
2870 |
+
"epoch": 0.02035,
|
2871 |
+
"grad_norm": 73.5,
|
2872 |
+
"learning_rate": 4.9228643216080405e-05,
|
2873 |
+
"loss": 7.5865,
|
2874 |
+
"step": 4070
|
2875 |
+
},
|
2876 |
+
{
|
2877 |
+
"epoch": 0.0204,
|
2878 |
+
"grad_norm": 15.0,
|
2879 |
+
"learning_rate": 4.9226130653266336e-05,
|
2880 |
+
"loss": 7.5648,
|
2881 |
+
"step": 4080
|
2882 |
+
},
|
2883 |
+
{
|
2884 |
+
"epoch": 0.02045,
|
2885 |
+
"grad_norm": 1.359375,
|
2886 |
+
"learning_rate": 4.922361809045226e-05,
|
2887 |
+
"loss": 7.5471,
|
2888 |
+
"step": 4090
|
2889 |
+
},
|
2890 |
+
{
|
2891 |
+
"epoch": 0.0205,
|
2892 |
+
"grad_norm": 56.25,
|
2893 |
+
"learning_rate": 4.92211055276382e-05,
|
2894 |
+
"loss": 7.564,
|
2895 |
+
"step": 4100
|
2896 |
+
},
|
2897 |
+
{
|
2898 |
+
"epoch": 0.02055,
|
2899 |
+
"grad_norm": 50.75,
|
2900 |
+
"learning_rate": 4.921859296482412e-05,
|
2901 |
+
"loss": 7.5555,
|
2902 |
+
"step": 4110
|
2903 |
+
},
|
2904 |
+
{
|
2905 |
+
"epoch": 0.0206,
|
2906 |
+
"grad_norm": 1.7890625,
|
2907 |
+
"learning_rate": 4.921608040201005e-05,
|
2908 |
+
"loss": 7.5561,
|
2909 |
+
"step": 4120
|
2910 |
+
},
|
2911 |
+
{
|
2912 |
+
"epoch": 0.02065,
|
2913 |
+
"grad_norm": 2.21875,
|
2914 |
+
"learning_rate": 4.9213567839195984e-05,
|
2915 |
+
"loss": 7.5605,
|
2916 |
+
"step": 4130
|
2917 |
+
},
|
2918 |
+
{
|
2919 |
+
"epoch": 0.0207,
|
2920 |
+
"grad_norm": 2.71875,
|
2921 |
+
"learning_rate": 4.9211055276381915e-05,
|
2922 |
+
"loss": 7.5899,
|
2923 |
+
"step": 4140
|
2924 |
+
},
|
2925 |
+
{
|
2926 |
+
"epoch": 0.02075,
|
2927 |
+
"grad_norm": 2.25,
|
2928 |
+
"learning_rate": 4.920854271356784e-05,
|
2929 |
+
"loss": 7.5548,
|
2930 |
+
"step": 4150
|
2931 |
+
},
|
2932 |
+
{
|
2933 |
+
"epoch": 0.0208,
|
2934 |
+
"grad_norm": 14.5625,
|
2935 |
+
"learning_rate": 4.920603015075377e-05,
|
2936 |
+
"loss": 7.625,
|
2937 |
+
"step": 4160
|
2938 |
+
},
|
2939 |
+
{
|
2940 |
+
"epoch": 0.02085,
|
2941 |
+
"grad_norm": 1.765625,
|
2942 |
+
"learning_rate": 4.92035175879397e-05,
|
2943 |
+
"loss": 7.5754,
|
2944 |
+
"step": 4170
|
2945 |
+
},
|
2946 |
+
{
|
2947 |
+
"epoch": 0.0209,
|
2948 |
+
"grad_norm": 1.390625,
|
2949 |
+
"learning_rate": 4.920100502512563e-05,
|
2950 |
+
"loss": 7.601,
|
2951 |
+
"step": 4180
|
2952 |
+
},
|
2953 |
+
{
|
2954 |
+
"epoch": 0.02095,
|
2955 |
+
"grad_norm": 13.5,
|
2956 |
+
"learning_rate": 4.919849246231156e-05,
|
2957 |
+
"loss": 7.5853,
|
2958 |
+
"step": 4190
|
2959 |
+
},
|
2960 |
+
{
|
2961 |
+
"epoch": 0.021,
|
2962 |
+
"grad_norm": 3.421875,
|
2963 |
+
"learning_rate": 4.919597989949749e-05,
|
2964 |
+
"loss": 7.5461,
|
2965 |
+
"step": 4200
|
2966 |
+
},
|
2967 |
+
{
|
2968 |
+
"epoch": 0.02105,
|
2969 |
+
"grad_norm": 10.6875,
|
2970 |
+
"learning_rate": 4.919346733668342e-05,
|
2971 |
+
"loss": 7.5906,
|
2972 |
+
"step": 4210
|
2973 |
+
},
|
2974 |
+
{
|
2975 |
+
"epoch": 0.0211,
|
2976 |
+
"grad_norm": 2.140625,
|
2977 |
+
"learning_rate": 4.919095477386935e-05,
|
2978 |
+
"loss": 7.5771,
|
2979 |
+
"step": 4220
|
2980 |
+
},
|
2981 |
+
{
|
2982 |
+
"epoch": 0.02115,
|
2983 |
+
"grad_norm": 4.09375,
|
2984 |
+
"learning_rate": 4.918844221105528e-05,
|
2985 |
+
"loss": 7.6011,
|
2986 |
+
"step": 4230
|
2987 |
+
},
|
2988 |
+
{
|
2989 |
+
"epoch": 0.0212,
|
2990 |
+
"grad_norm": 64.0,
|
2991 |
+
"learning_rate": 4.9185929648241205e-05,
|
2992 |
+
"loss": 7.5948,
|
2993 |
+
"step": 4240
|
2994 |
+
},
|
2995 |
+
{
|
2996 |
+
"epoch": 0.02125,
|
2997 |
+
"grad_norm": 2.21875,
|
2998 |
+
"learning_rate": 4.9183417085427136e-05,
|
2999 |
+
"loss": 7.5682,
|
3000 |
+
"step": 4250
|
3001 |
+
},
|
3002 |
+
{
|
3003 |
+
"epoch": 0.0213,
|
3004 |
+
"grad_norm": 28.625,
|
3005 |
+
"learning_rate": 4.918090452261307e-05,
|
3006 |
+
"loss": 7.5859,
|
3007 |
+
"step": 4260
|
3008 |
+
},
|
3009 |
+
{
|
3010 |
+
"epoch": 0.02135,
|
3011 |
+
"grad_norm": 5.6875,
|
3012 |
+
"learning_rate": 4.9178391959799e-05,
|
3013 |
+
"loss": 7.591,
|
3014 |
+
"step": 4270
|
3015 |
+
},
|
3016 |
+
{
|
3017 |
+
"epoch": 0.0214,
|
3018 |
+
"grad_norm": 60.25,
|
3019 |
+
"learning_rate": 4.917587939698492e-05,
|
3020 |
+
"loss": 7.5652,
|
3021 |
+
"step": 4280
|
3022 |
+
},
|
3023 |
+
{
|
3024 |
+
"epoch": 0.02145,
|
3025 |
+
"grad_norm": 1.75,
|
3026 |
+
"learning_rate": 4.917336683417086e-05,
|
3027 |
+
"loss": 7.5578,
|
3028 |
+
"step": 4290
|
3029 |
+
},
|
3030 |
+
{
|
3031 |
+
"epoch": 0.0215,
|
3032 |
+
"grad_norm": 1.734375,
|
3033 |
+
"learning_rate": 4.9170854271356784e-05,
|
3034 |
+
"loss": 7.6033,
|
3035 |
+
"step": 4300
|
3036 |
+
},
|
3037 |
+
{
|
3038 |
+
"epoch": 0.02155,
|
3039 |
+
"grad_norm": 7.75,
|
3040 |
+
"learning_rate": 4.9168341708542715e-05,
|
3041 |
+
"loss": 7.5929,
|
3042 |
+
"step": 4310
|
3043 |
+
},
|
3044 |
+
{
|
3045 |
+
"epoch": 0.0216,
|
3046 |
+
"grad_norm": 29.5,
|
3047 |
+
"learning_rate": 4.9165829145728646e-05,
|
3048 |
+
"loss": 7.5736,
|
3049 |
+
"step": 4320
|
3050 |
+
},
|
3051 |
+
{
|
3052 |
+
"epoch": 0.02165,
|
3053 |
+
"grad_norm": 11.6875,
|
3054 |
+
"learning_rate": 4.916331658291458e-05,
|
3055 |
+
"loss": 7.5885,
|
3056 |
+
"step": 4330
|
3057 |
+
},
|
3058 |
+
{
|
3059 |
+
"epoch": 0.0217,
|
3060 |
+
"grad_norm": 1.1640625,
|
3061 |
+
"learning_rate": 4.91608040201005e-05,
|
3062 |
+
"loss": 7.6,
|
3063 |
+
"step": 4340
|
3064 |
+
},
|
3065 |
+
{
|
3066 |
+
"epoch": 0.02175,
|
3067 |
+
"grad_norm": 80.0,
|
3068 |
+
"learning_rate": 4.915829145728644e-05,
|
3069 |
+
"loss": 7.5922,
|
3070 |
+
"step": 4350
|
3071 |
+
},
|
3072 |
+
{
|
3073 |
+
"epoch": 0.0218,
|
3074 |
+
"grad_norm": 2.0625,
|
3075 |
+
"learning_rate": 4.9155778894472363e-05,
|
3076 |
+
"loss": 7.5828,
|
3077 |
+
"step": 4360
|
3078 |
+
},
|
3079 |
+
{
|
3080 |
+
"epoch": 0.02185,
|
3081 |
+
"grad_norm": 16.5,
|
3082 |
+
"learning_rate": 4.9153266331658294e-05,
|
3083 |
+
"loss": 7.5455,
|
3084 |
+
"step": 4370
|
3085 |
+
},
|
3086 |
+
{
|
3087 |
+
"epoch": 0.0219,
|
3088 |
+
"grad_norm": 1.8984375,
|
3089 |
+
"learning_rate": 4.915075376884422e-05,
|
3090 |
+
"loss": 7.5869,
|
3091 |
+
"step": 4380
|
3092 |
+
},
|
3093 |
+
{
|
3094 |
+
"epoch": 0.02195,
|
3095 |
+
"grad_norm": 9.25,
|
3096 |
+
"learning_rate": 4.9148241206030156e-05,
|
3097 |
+
"loss": 7.57,
|
3098 |
+
"step": 4390
|
3099 |
+
},
|
3100 |
+
{
|
3101 |
+
"epoch": 0.022,
|
3102 |
+
"grad_norm": 1.234375,
|
3103 |
+
"learning_rate": 4.914572864321608e-05,
|
3104 |
+
"loss": 7.5779,
|
3105 |
+
"step": 4400
|
3106 |
+
},
|
3107 |
+
{
|
3108 |
+
"epoch": 0.02205,
|
3109 |
+
"grad_norm": 42.5,
|
3110 |
+
"learning_rate": 4.914321608040201e-05,
|
3111 |
+
"loss": 7.5756,
|
3112 |
+
"step": 4410
|
3113 |
+
},
|
3114 |
+
{
|
3115 |
+
"epoch": 0.0221,
|
3116 |
+
"grad_norm": 2.40625,
|
3117 |
+
"learning_rate": 4.914070351758794e-05,
|
3118 |
+
"loss": 7.5708,
|
3119 |
+
"step": 4420
|
3120 |
+
},
|
3121 |
+
{
|
3122 |
+
"epoch": 0.02215,
|
3123 |
+
"grad_norm": 4.8125,
|
3124 |
+
"learning_rate": 4.9138190954773874e-05,
|
3125 |
+
"loss": 7.5854,
|
3126 |
+
"step": 4430
|
3127 |
+
},
|
3128 |
+
{
|
3129 |
+
"epoch": 0.0222,
|
3130 |
+
"grad_norm": 1.296875,
|
3131 |
+
"learning_rate": 4.91356783919598e-05,
|
3132 |
+
"loss": 7.6228,
|
3133 |
+
"step": 4440
|
3134 |
+
},
|
3135 |
+
{
|
3136 |
+
"epoch": 0.02225,
|
3137 |
+
"grad_norm": 1.6015625,
|
3138 |
+
"learning_rate": 4.9133165829145736e-05,
|
3139 |
+
"loss": 7.56,
|
3140 |
+
"step": 4450
|
3141 |
+
},
|
3142 |
+
{
|
3143 |
+
"epoch": 0.0223,
|
3144 |
+
"grad_norm": 9.375,
|
3145 |
+
"learning_rate": 4.913065326633166e-05,
|
3146 |
+
"loss": 7.5837,
|
3147 |
+
"step": 4460
|
3148 |
+
},
|
3149 |
+
{
|
3150 |
+
"epoch": 0.02235,
|
3151 |
+
"grad_norm": 26.5,
|
3152 |
+
"learning_rate": 4.912814070351759e-05,
|
3153 |
+
"loss": 7.5659,
|
3154 |
+
"step": 4470
|
3155 |
+
},
|
3156 |
+
{
|
3157 |
+
"epoch": 0.0224,
|
3158 |
+
"grad_norm": 1.5859375,
|
3159 |
+
"learning_rate": 4.912562814070352e-05,
|
3160 |
+
"loss": 7.622,
|
3161 |
+
"step": 4480
|
3162 |
+
},
|
3163 |
+
{
|
3164 |
+
"epoch": 0.02245,
|
3165 |
+
"grad_norm": 3.6875,
|
3166 |
+
"learning_rate": 4.912311557788945e-05,
|
3167 |
+
"loss": 7.5823,
|
3168 |
+
"step": 4490
|
3169 |
+
},
|
3170 |
+
{
|
3171 |
+
"epoch": 0.0225,
|
3172 |
+
"grad_norm": 76.5,
|
3173 |
+
"learning_rate": 4.912060301507538e-05,
|
3174 |
+
"loss": 7.5724,
|
3175 |
+
"step": 4500
|
3176 |
+
},
|
3177 |
+
{
|
3178 |
+
"epoch": 0.02255,
|
3179 |
+
"grad_norm": 14.25,
|
3180 |
+
"learning_rate": 4.911809045226131e-05,
|
3181 |
+
"loss": 7.5786,
|
3182 |
+
"step": 4510
|
3183 |
+
},
|
3184 |
+
{
|
3185 |
+
"epoch": 0.0226,
|
3186 |
+
"grad_norm": 1.53125,
|
3187 |
+
"learning_rate": 4.911557788944724e-05,
|
3188 |
+
"loss": 7.5411,
|
3189 |
+
"step": 4520
|
3190 |
+
},
|
3191 |
+
{
|
3192 |
+
"epoch": 0.02265,
|
3193 |
+
"grad_norm": 28.125,
|
3194 |
+
"learning_rate": 4.911306532663317e-05,
|
3195 |
+
"loss": 7.5757,
|
3196 |
+
"step": 4530
|
3197 |
+
},
|
3198 |
+
{
|
3199 |
+
"epoch": 0.0227,
|
3200 |
+
"grad_norm": 1.4609375,
|
3201 |
+
"learning_rate": 4.9110552763819095e-05,
|
3202 |
+
"loss": 7.5391,
|
3203 |
+
"step": 4540
|
3204 |
+
},
|
3205 |
+
{
|
3206 |
+
"epoch": 0.02275,
|
3207 |
+
"grad_norm": 1.546875,
|
3208 |
+
"learning_rate": 4.9108040201005026e-05,
|
3209 |
+
"loss": 7.6018,
|
3210 |
+
"step": 4550
|
3211 |
+
},
|
3212 |
+
{
|
3213 |
+
"epoch": 0.0228,
|
3214 |
+
"grad_norm": 3.890625,
|
3215 |
+
"learning_rate": 4.910552763819096e-05,
|
3216 |
+
"loss": 7.5935,
|
3217 |
+
"step": 4560
|
3218 |
+
},
|
3219 |
+
{
|
3220 |
+
"epoch": 0.02285,
|
3221 |
+
"grad_norm": 2.28125,
|
3222 |
+
"learning_rate": 4.910301507537689e-05,
|
3223 |
+
"loss": 7.5827,
|
3224 |
+
"step": 4570
|
3225 |
+
},
|
3226 |
+
{
|
3227 |
+
"epoch": 0.0229,
|
3228 |
+
"grad_norm": 8.0,
|
3229 |
+
"learning_rate": 4.910050251256282e-05,
|
3230 |
+
"loss": 7.5591,
|
3231 |
+
"step": 4580
|
3232 |
+
},
|
3233 |
+
{
|
3234 |
+
"epoch": 0.02295,
|
3235 |
+
"grad_norm": 3.328125,
|
3236 |
+
"learning_rate": 4.909798994974874e-05,
|
3237 |
+
"loss": 7.563,
|
3238 |
+
"step": 4590
|
3239 |
+
},
|
3240 |
+
{
|
3241 |
+
"epoch": 0.023,
|
3242 |
+
"grad_norm": 10.875,
|
3243 |
+
"learning_rate": 4.9095477386934674e-05,
|
3244 |
+
"loss": 7.542,
|
3245 |
+
"step": 4600
|
3246 |
+
},
|
3247 |
+
{
|
3248 |
+
"epoch": 0.02305,
|
3249 |
+
"grad_norm": 8.25,
|
3250 |
+
"learning_rate": 4.9092964824120605e-05,
|
3251 |
+
"loss": 7.5628,
|
3252 |
+
"step": 4610
|
3253 |
+
},
|
3254 |
+
{
|
3255 |
+
"epoch": 0.0231,
|
3256 |
+
"grad_norm": 29.25,
|
3257 |
+
"learning_rate": 4.9090452261306536e-05,
|
3258 |
+
"loss": 7.5309,
|
3259 |
+
"step": 4620
|
3260 |
+
},
|
3261 |
+
{
|
3262 |
+
"epoch": 0.02315,
|
3263 |
+
"grad_norm": 2.0,
|
3264 |
+
"learning_rate": 4.908793969849246e-05,
|
3265 |
+
"loss": 7.554,
|
3266 |
+
"step": 4630
|
3267 |
+
},
|
3268 |
+
{
|
3269 |
+
"epoch": 0.0232,
|
3270 |
+
"grad_norm": 74.0,
|
3271 |
+
"learning_rate": 4.90854271356784e-05,
|
3272 |
+
"loss": 7.5653,
|
3273 |
+
"step": 4640
|
3274 |
+
},
|
3275 |
+
{
|
3276 |
+
"epoch": 0.02325,
|
3277 |
+
"grad_norm": 27.25,
|
3278 |
+
"learning_rate": 4.908291457286432e-05,
|
3279 |
+
"loss": 7.5813,
|
3280 |
+
"step": 4650
|
3281 |
+
},
|
3282 |
+
{
|
3283 |
+
"epoch": 0.0233,
|
3284 |
+
"grad_norm": 15.6875,
|
3285 |
+
"learning_rate": 4.908040201005025e-05,
|
3286 |
+
"loss": 7.5811,
|
3287 |
+
"step": 4660
|
3288 |
+
},
|
3289 |
+
{
|
3290 |
+
"epoch": 0.02335,
|
3291 |
+
"grad_norm": 4.65625,
|
3292 |
+
"learning_rate": 4.907788944723618e-05,
|
3293 |
+
"loss": 7.5954,
|
3294 |
+
"step": 4670
|
3295 |
+
},
|
3296 |
+
{
|
3297 |
+
"epoch": 0.0234,
|
3298 |
+
"grad_norm": 1.6875,
|
3299 |
+
"learning_rate": 4.9075376884422115e-05,
|
3300 |
+
"loss": 7.5495,
|
3301 |
+
"step": 4680
|
3302 |
+
},
|
3303 |
+
{
|
3304 |
+
"epoch": 0.02345,
|
3305 |
+
"grad_norm": 138.0,
|
3306 |
+
"learning_rate": 4.907286432160804e-05,
|
3307 |
+
"loss": 7.5763,
|
3308 |
+
"step": 4690
|
3309 |
+
},
|
3310 |
+
{
|
3311 |
+
"epoch": 0.0235,
|
3312 |
+
"grad_norm": 3.234375,
|
3313 |
+
"learning_rate": 4.907035175879397e-05,
|
3314 |
+
"loss": 7.5582,
|
3315 |
+
"step": 4700
|
3316 |
+
},
|
3317 |
+
{
|
3318 |
+
"epoch": 0.02355,
|
3319 |
+
"grad_norm": 10.4375,
|
3320 |
+
"learning_rate": 4.90678391959799e-05,
|
3321 |
+
"loss": 7.562,
|
3322 |
+
"step": 4710
|
3323 |
+
},
|
3324 |
+
{
|
3325 |
+
"epoch": 0.0236,
|
3326 |
+
"grad_norm": 46.25,
|
3327 |
+
"learning_rate": 4.906532663316583e-05,
|
3328 |
+
"loss": 7.5865,
|
3329 |
+
"step": 4720
|
3330 |
+
},
|
3331 |
+
{
|
3332 |
+
"epoch": 0.02365,
|
3333 |
+
"grad_norm": 9.5625,
|
3334 |
+
"learning_rate": 4.906281407035176e-05,
|
3335 |
+
"loss": 7.5701,
|
3336 |
+
"step": 4730
|
3337 |
+
},
|
3338 |
+
{
|
3339 |
+
"epoch": 0.0237,
|
3340 |
+
"grad_norm": 5.3125,
|
3341 |
+
"learning_rate": 4.9060301507537695e-05,
|
3342 |
+
"loss": 7.608,
|
3343 |
+
"step": 4740
|
3344 |
+
},
|
3345 |
+
{
|
3346 |
+
"epoch": 0.02375,
|
3347 |
+
"grad_norm": 9.75,
|
3348 |
+
"learning_rate": 4.905778894472362e-05,
|
3349 |
+
"loss": 7.5944,
|
3350 |
+
"step": 4750
|
3351 |
+
},
|
3352 |
+
{
|
3353 |
+
"epoch": 0.0238,
|
3354 |
+
"grad_norm": 8.125,
|
3355 |
+
"learning_rate": 4.905527638190955e-05,
|
3356 |
+
"loss": 7.6044,
|
3357 |
+
"step": 4760
|
3358 |
+
},
|
3359 |
+
{
|
3360 |
+
"epoch": 0.02385,
|
3361 |
+
"grad_norm": 1.5546875,
|
3362 |
+
"learning_rate": 4.905276381909548e-05,
|
3363 |
+
"loss": 7.594,
|
3364 |
+
"step": 4770
|
3365 |
+
},
|
3366 |
+
{
|
3367 |
+
"epoch": 0.0239,
|
3368 |
+
"grad_norm": 5.1875,
|
3369 |
+
"learning_rate": 4.905025125628141e-05,
|
3370 |
+
"loss": 7.5989,
|
3371 |
+
"step": 4780
|
3372 |
+
},
|
3373 |
+
{
|
3374 |
+
"epoch": 0.02395,
|
3375 |
+
"grad_norm": 1.5390625,
|
3376 |
+
"learning_rate": 4.9047738693467336e-05,
|
3377 |
+
"loss": 7.5679,
|
3378 |
+
"step": 4790
|
3379 |
+
},
|
3380 |
+
{
|
3381 |
+
"epoch": 0.024,
|
3382 |
+
"grad_norm": 16.875,
|
3383 |
+
"learning_rate": 4.9045226130653274e-05,
|
3384 |
+
"loss": 7.5629,
|
3385 |
+
"step": 4800
|
3386 |
+
},
|
3387 |
+
{
|
3388 |
+
"epoch": 0.02405,
|
3389 |
+
"grad_norm": 1.671875,
|
3390 |
+
"learning_rate": 4.90427135678392e-05,
|
3391 |
+
"loss": 7.5778,
|
3392 |
+
"step": 4810
|
3393 |
+
},
|
3394 |
+
{
|
3395 |
+
"epoch": 0.0241,
|
3396 |
+
"grad_norm": 1.78125,
|
3397 |
+
"learning_rate": 4.904020100502513e-05,
|
3398 |
+
"loss": 7.5648,
|
3399 |
+
"step": 4820
|
3400 |
+
},
|
3401 |
+
{
|
3402 |
+
"epoch": 0.02415,
|
3403 |
+
"grad_norm": 5.84375,
|
3404 |
+
"learning_rate": 4.9037688442211053e-05,
|
3405 |
+
"loss": 7.5534,
|
3406 |
+
"step": 4830
|
3407 |
+
},
|
3408 |
+
{
|
3409 |
+
"epoch": 0.0242,
|
3410 |
+
"grad_norm": 1.640625,
|
3411 |
+
"learning_rate": 4.903517587939699e-05,
|
3412 |
+
"loss": 7.6181,
|
3413 |
+
"step": 4840
|
3414 |
+
},
|
3415 |
+
{
|
3416 |
+
"epoch": 0.02425,
|
3417 |
+
"grad_norm": 52.25,
|
3418 |
+
"learning_rate": 4.9032663316582916e-05,
|
3419 |
+
"loss": 7.5699,
|
3420 |
+
"step": 4850
|
3421 |
+
},
|
3422 |
+
{
|
3423 |
+
"epoch": 0.0243,
|
3424 |
+
"grad_norm": 1.9453125,
|
3425 |
+
"learning_rate": 4.9030150753768847e-05,
|
3426 |
+
"loss": 7.6262,
|
3427 |
+
"step": 4860
|
3428 |
+
},
|
3429 |
+
{
|
3430 |
+
"epoch": 0.02435,
|
3431 |
+
"grad_norm": 1.625,
|
3432 |
+
"learning_rate": 4.902763819095478e-05,
|
3433 |
+
"loss": 7.5909,
|
3434 |
+
"step": 4870
|
3435 |
+
},
|
3436 |
+
{
|
3437 |
+
"epoch": 0.0244,
|
3438 |
+
"grad_norm": 12.9375,
|
3439 |
+
"learning_rate": 4.902512562814071e-05,
|
3440 |
+
"loss": 7.5804,
|
3441 |
+
"step": 4880
|
3442 |
+
},
|
3443 |
+
{
|
3444 |
+
"epoch": 0.02445,
|
3445 |
+
"grad_norm": 1.84375,
|
3446 |
+
"learning_rate": 4.902261306532663e-05,
|
3447 |
+
"loss": 7.5537,
|
3448 |
+
"step": 4890
|
3449 |
+
},
|
3450 |
+
{
|
3451 |
+
"epoch": 0.0245,
|
3452 |
+
"grad_norm": 9.0,
|
3453 |
+
"learning_rate": 4.9020100502512564e-05,
|
3454 |
+
"loss": 7.5913,
|
3455 |
+
"step": 4900
|
3456 |
+
},
|
3457 |
+
{
|
3458 |
+
"epoch": 0.02455,
|
3459 |
+
"grad_norm": 9.3125,
|
3460 |
+
"learning_rate": 4.9017587939698495e-05,
|
3461 |
+
"loss": 7.5744,
|
3462 |
+
"step": 4910
|
3463 |
+
},
|
3464 |
+
{
|
3465 |
+
"epoch": 0.0246,
|
3466 |
+
"grad_norm": 1.5234375,
|
3467 |
+
"learning_rate": 4.9015075376884426e-05,
|
3468 |
+
"loss": 7.6023,
|
3469 |
+
"step": 4920
|
3470 |
+
},
|
3471 |
+
{
|
3472 |
+
"epoch": 0.02465,
|
3473 |
+
"grad_norm": 1.8046875,
|
3474 |
+
"learning_rate": 4.901256281407036e-05,
|
3475 |
+
"loss": 7.575,
|
3476 |
+
"step": 4930
|
3477 |
+
},
|
3478 |
+
{
|
3479 |
+
"epoch": 0.0247,
|
3480 |
+
"grad_norm": 17.875,
|
3481 |
+
"learning_rate": 4.901005025125628e-05,
|
3482 |
+
"loss": 7.5894,
|
3483 |
+
"step": 4940
|
3484 |
+
},
|
3485 |
+
{
|
3486 |
+
"epoch": 0.02475,
|
3487 |
+
"grad_norm": 33.25,
|
3488 |
+
"learning_rate": 4.900753768844221e-05,
|
3489 |
+
"loss": 7.6012,
|
3490 |
+
"step": 4950
|
3491 |
+
},
|
3492 |
+
{
|
3493 |
+
"epoch": 0.0248,
|
3494 |
+
"grad_norm": 2.53125,
|
3495 |
+
"learning_rate": 4.900502512562814e-05,
|
3496 |
+
"loss": 7.5584,
|
3497 |
+
"step": 4960
|
3498 |
+
},
|
3499 |
+
{
|
3500 |
+
"epoch": 0.02485,
|
3501 |
+
"grad_norm": 2.078125,
|
3502 |
+
"learning_rate": 4.9002512562814074e-05,
|
3503 |
+
"loss": 7.5945,
|
3504 |
+
"step": 4970
|
3505 |
+
},
|
3506 |
+
{
|
3507 |
+
"epoch": 0.0249,
|
3508 |
+
"grad_norm": 3.0625,
|
3509 |
+
"learning_rate": 4.9e-05,
|
3510 |
+
"loss": 7.5507,
|
3511 |
+
"step": 4980
|
3512 |
+
},
|
3513 |
+
{
|
3514 |
+
"epoch": 0.02495,
|
3515 |
+
"grad_norm": 2.859375,
|
3516 |
+
"learning_rate": 4.899748743718593e-05,
|
3517 |
+
"loss": 7.585,
|
3518 |
+
"step": 4990
|
3519 |
+
},
|
3520 |
+
{
|
3521 |
+
"epoch": 0.025,
|
3522 |
+
"grad_norm": 7.1875,
|
3523 |
+
"learning_rate": 4.899497487437186e-05,
|
3524 |
+
"loss": 7.5395,
|
3525 |
+
"step": 5000
|
3526 |
+
},
|
3527 |
+
{
|
3528 |
+
"epoch": 0.02505,
|
3529 |
+
"grad_norm": 1.2109375,
|
3530 |
+
"learning_rate": 4.899246231155779e-05,
|
3531 |
+
"loss": 7.5543,
|
3532 |
+
"step": 5010
|
3533 |
+
},
|
3534 |
+
{
|
3535 |
+
"epoch": 0.0251,
|
3536 |
+
"grad_norm": 104.0,
|
3537 |
+
"learning_rate": 4.8989949748743716e-05,
|
3538 |
+
"loss": 7.5663,
|
3539 |
+
"step": 5020
|
3540 |
+
},
|
3541 |
+
{
|
3542 |
+
"epoch": 0.02515,
|
3543 |
+
"grad_norm": 35.25,
|
3544 |
+
"learning_rate": 4.8987437185929654e-05,
|
3545 |
+
"loss": 7.5931,
|
3546 |
+
"step": 5030
|
3547 |
+
},
|
3548 |
+
{
|
3549 |
+
"epoch": 0.0252,
|
3550 |
+
"grad_norm": 1.8046875,
|
3551 |
+
"learning_rate": 4.898492462311558e-05,
|
3552 |
+
"loss": 7.5548,
|
3553 |
+
"step": 5040
|
3554 |
+
},
|
3555 |
+
{
|
3556 |
+
"epoch": 0.02525,
|
3557 |
+
"grad_norm": 29.5,
|
3558 |
+
"learning_rate": 4.898241206030151e-05,
|
3559 |
+
"loss": 7.5911,
|
3560 |
+
"step": 5050
|
3561 |
+
},
|
3562 |
+
{
|
3563 |
+
"epoch": 0.0253,
|
3564 |
+
"grad_norm": 4.34375,
|
3565 |
+
"learning_rate": 4.897989949748744e-05,
|
3566 |
+
"loss": 7.5819,
|
3567 |
+
"step": 5060
|
3568 |
+
},
|
3569 |
+
{
|
3570 |
+
"epoch": 0.02535,
|
3571 |
+
"grad_norm": 1.5234375,
|
3572 |
+
"learning_rate": 4.897738693467337e-05,
|
3573 |
+
"loss": 7.6094,
|
3574 |
+
"step": 5070
|
3575 |
+
},
|
3576 |
+
{
|
3577 |
+
"epoch": 0.0254,
|
3578 |
+
"grad_norm": 49.25,
|
3579 |
+
"learning_rate": 4.8974874371859295e-05,
|
3580 |
+
"loss": 7.5936,
|
3581 |
+
"step": 5080
|
3582 |
+
},
|
3583 |
+
{
|
3584 |
+
"epoch": 0.02545,
|
3585 |
+
"grad_norm": 2.671875,
|
3586 |
+
"learning_rate": 4.897236180904523e-05,
|
3587 |
+
"loss": 7.5785,
|
3588 |
+
"step": 5090
|
3589 |
+
},
|
3590 |
+
{
|
3591 |
+
"epoch": 0.0255,
|
3592 |
+
"grad_norm": 4.1875,
|
3593 |
+
"learning_rate": 4.896984924623116e-05,
|
3594 |
+
"loss": 7.5704,
|
3595 |
+
"step": 5100
|
3596 |
+
},
|
3597 |
+
{
|
3598 |
+
"epoch": 0.02555,
|
3599 |
+
"grad_norm": 1.4296875,
|
3600 |
+
"learning_rate": 4.896733668341709e-05,
|
3601 |
+
"loss": 7.5756,
|
3602 |
+
"step": 5110
|
3603 |
+
},
|
3604 |
+
{
|
3605 |
+
"epoch": 0.0256,
|
3606 |
+
"grad_norm": 51.75,
|
3607 |
+
"learning_rate": 4.896482412060302e-05,
|
3608 |
+
"loss": 7.594,
|
3609 |
+
"step": 5120
|
3610 |
+
},
|
3611 |
+
{
|
3612 |
+
"epoch": 0.02565,
|
3613 |
+
"grad_norm": 3.484375,
|
3614 |
+
"learning_rate": 4.896231155778895e-05,
|
3615 |
+
"loss": 7.5914,
|
3616 |
+
"step": 5130
|
3617 |
+
},
|
3618 |
+
{
|
3619 |
+
"epoch": 0.0257,
|
3620 |
+
"grad_norm": 6.625,
|
3621 |
+
"learning_rate": 4.8959798994974874e-05,
|
3622 |
+
"loss": 7.5849,
|
3623 |
+
"step": 5140
|
3624 |
+
},
|
3625 |
+
{
|
3626 |
+
"epoch": 0.02575,
|
3627 |
+
"grad_norm": 1.765625,
|
3628 |
+
"learning_rate": 4.8957286432160805e-05,
|
3629 |
+
"loss": 7.592,
|
3630 |
+
"step": 5150
|
3631 |
+
},
|
3632 |
+
{
|
3633 |
+
"epoch": 0.0258,
|
3634 |
+
"grad_norm": 1.765625,
|
3635 |
+
"learning_rate": 4.8954773869346736e-05,
|
3636 |
+
"loss": 7.5583,
|
3637 |
+
"step": 5160
|
3638 |
+
},
|
3639 |
+
{
|
3640 |
+
"epoch": 0.02585,
|
3641 |
+
"grad_norm": 1.5625,
|
3642 |
+
"learning_rate": 4.895226130653267e-05,
|
3643 |
+
"loss": 7.6054,
|
3644 |
+
"step": 5170
|
3645 |
+
},
|
3646 |
+
{
|
3647 |
+
"epoch": 0.0259,
|
3648 |
+
"grad_norm": 1.6796875,
|
3649 |
+
"learning_rate": 4.894974874371859e-05,
|
3650 |
+
"loss": 7.5911,
|
3651 |
+
"step": 5180
|
3652 |
+
},
|
3653 |
+
{
|
3654 |
+
"epoch": 0.02595,
|
3655 |
+
"grad_norm": 2.921875,
|
3656 |
+
"learning_rate": 4.894723618090453e-05,
|
3657 |
+
"loss": 7.5316,
|
3658 |
+
"step": 5190
|
3659 |
+
},
|
3660 |
+
{
|
3661 |
+
"epoch": 0.026,
|
3662 |
+
"grad_norm": 18.125,
|
3663 |
+
"learning_rate": 4.8944723618090454e-05,
|
3664 |
+
"loss": 7.6079,
|
3665 |
+
"step": 5200
|
3666 |
+
},
|
3667 |
+
{
|
3668 |
+
"epoch": 0.02605,
|
3669 |
+
"grad_norm": 2.6875,
|
3670 |
+
"learning_rate": 4.8942211055276385e-05,
|
3671 |
+
"loss": 7.5652,
|
3672 |
+
"step": 5210
|
3673 |
+
},
|
3674 |
+
{
|
3675 |
+
"epoch": 0.0261,
|
3676 |
+
"grad_norm": 13.1875,
|
3677 |
+
"learning_rate": 4.8939698492462316e-05,
|
3678 |
+
"loss": 7.593,
|
3679 |
+
"step": 5220
|
3680 |
+
},
|
3681 |
+
{
|
3682 |
+
"epoch": 0.02615,
|
3683 |
+
"grad_norm": 1.734375,
|
3684 |
+
"learning_rate": 4.893718592964825e-05,
|
3685 |
+
"loss": 7.5381,
|
3686 |
+
"step": 5230
|
3687 |
+
},
|
3688 |
+
{
|
3689 |
+
"epoch": 0.0262,
|
3690 |
+
"grad_norm": 1.390625,
|
3691 |
+
"learning_rate": 4.893467336683417e-05,
|
3692 |
+
"loss": 7.5803,
|
3693 |
+
"step": 5240
|
3694 |
+
},
|
3695 |
+
{
|
3696 |
+
"epoch": 0.02625,
|
3697 |
+
"grad_norm": 32.75,
|
3698 |
+
"learning_rate": 4.89321608040201e-05,
|
3699 |
+
"loss": 7.5725,
|
3700 |
+
"step": 5250
|
3701 |
+
},
|
3702 |
+
{
|
3703 |
+
"epoch": 0.0263,
|
3704 |
+
"grad_norm": 27.375,
|
3705 |
+
"learning_rate": 4.892964824120603e-05,
|
3706 |
+
"loss": 7.5863,
|
3707 |
+
"step": 5260
|
3708 |
+
},
|
3709 |
+
{
|
3710 |
+
"epoch": 0.02635,
|
3711 |
+
"grad_norm": 1.8046875,
|
3712 |
+
"learning_rate": 4.8927135678391964e-05,
|
3713 |
+
"loss": 7.5594,
|
3714 |
+
"step": 5270
|
3715 |
+
},
|
3716 |
+
{
|
3717 |
+
"epoch": 0.0264,
|
3718 |
+
"grad_norm": 84.0,
|
3719 |
+
"learning_rate": 4.8924623115577895e-05,
|
3720 |
+
"loss": 7.5623,
|
3721 |
+
"step": 5280
|
3722 |
+
},
|
3723 |
+
{
|
3724 |
+
"epoch": 0.02645,
|
3725 |
+
"grad_norm": 1.8203125,
|
3726 |
+
"learning_rate": 4.892211055276382e-05,
|
3727 |
+
"loss": 7.5728,
|
3728 |
+
"step": 5290
|
3729 |
+
},
|
3730 |
+
{
|
3731 |
+
"epoch": 0.0265,
|
3732 |
+
"grad_norm": 3.1875,
|
3733 |
+
"learning_rate": 4.891959798994975e-05,
|
3734 |
+
"loss": 7.5856,
|
3735 |
+
"step": 5300
|
3736 |
+
},
|
3737 |
+
{
|
3738 |
+
"epoch": 0.02655,
|
3739 |
+
"grad_norm": 53.5,
|
3740 |
+
"learning_rate": 4.891708542713568e-05,
|
3741 |
+
"loss": 7.577,
|
3742 |
+
"step": 5310
|
3743 |
+
},
|
3744 |
+
{
|
3745 |
+
"epoch": 0.0266,
|
3746 |
+
"grad_norm": 1.7265625,
|
3747 |
+
"learning_rate": 4.891457286432161e-05,
|
3748 |
+
"loss": 7.591,
|
3749 |
+
"step": 5320
|
3750 |
+
},
|
3751 |
+
{
|
3752 |
+
"epoch": 0.02665,
|
3753 |
+
"grad_norm": 19.375,
|
3754 |
+
"learning_rate": 4.8912060301507537e-05,
|
3755 |
+
"loss": 7.5813,
|
3756 |
+
"step": 5330
|
3757 |
+
},
|
3758 |
+
{
|
3759 |
+
"epoch": 0.0267,
|
3760 |
+
"grad_norm": 37.75,
|
3761 |
+
"learning_rate": 4.890954773869347e-05,
|
3762 |
+
"loss": 7.5753,
|
3763 |
+
"step": 5340
|
3764 |
+
},
|
3765 |
+
{
|
3766 |
+
"epoch": 0.02675,
|
3767 |
+
"grad_norm": 4.0,
|
3768 |
+
"learning_rate": 4.89070351758794e-05,
|
3769 |
+
"loss": 7.5711,
|
3770 |
+
"step": 5350
|
3771 |
+
},
|
3772 |
+
{
|
3773 |
+
"epoch": 0.0268,
|
3774 |
+
"grad_norm": 7.84375,
|
3775 |
+
"learning_rate": 4.890452261306533e-05,
|
3776 |
+
"loss": 7.5626,
|
3777 |
+
"step": 5360
|
3778 |
+
},
|
3779 |
+
{
|
3780 |
+
"epoch": 0.02685,
|
3781 |
+
"grad_norm": 13.0625,
|
3782 |
+
"learning_rate": 4.8902010050251254e-05,
|
3783 |
+
"loss": 7.6023,
|
3784 |
+
"step": 5370
|
3785 |
+
},
|
3786 |
+
{
|
3787 |
+
"epoch": 0.0269,
|
3788 |
+
"grad_norm": 33.0,
|
3789 |
+
"learning_rate": 4.889949748743719e-05,
|
3790 |
+
"loss": 7.5754,
|
3791 |
+
"step": 5380
|
3792 |
+
},
|
3793 |
+
{
|
3794 |
+
"epoch": 0.02695,
|
3795 |
+
"grad_norm": 8.3125,
|
3796 |
+
"learning_rate": 4.8896984924623116e-05,
|
3797 |
+
"loss": 7.5874,
|
3798 |
+
"step": 5390
|
3799 |
+
},
|
3800 |
+
{
|
3801 |
+
"epoch": 0.027,
|
3802 |
+
"grad_norm": 2.046875,
|
3803 |
+
"learning_rate": 4.889447236180905e-05,
|
3804 |
+
"loss": 7.5645,
|
3805 |
+
"step": 5400
|
3806 |
+
},
|
3807 |
+
{
|
3808 |
+
"epoch": 0.02705,
|
3809 |
+
"grad_norm": 5.5,
|
3810 |
+
"learning_rate": 4.889195979899498e-05,
|
3811 |
+
"loss": 7.5483,
|
3812 |
+
"step": 5410
|
3813 |
+
},
|
3814 |
+
{
|
3815 |
+
"epoch": 0.0271,
|
3816 |
+
"grad_norm": 20.875,
|
3817 |
+
"learning_rate": 4.888944723618091e-05,
|
3818 |
+
"loss": 7.5784,
|
3819 |
+
"step": 5420
|
3820 |
+
},
|
3821 |
+
{
|
3822 |
+
"epoch": 0.02715,
|
3823 |
+
"grad_norm": 1.6796875,
|
3824 |
+
"learning_rate": 4.888693467336683e-05,
|
3825 |
+
"loss": 7.5691,
|
3826 |
+
"step": 5430
|
3827 |
+
},
|
3828 |
+
{
|
3829 |
+
"epoch": 0.0272,
|
3830 |
+
"grad_norm": 3.828125,
|
3831 |
+
"learning_rate": 4.888442211055277e-05,
|
3832 |
+
"loss": 7.5517,
|
3833 |
+
"step": 5440
|
3834 |
+
},
|
3835 |
+
{
|
3836 |
+
"epoch": 0.02725,
|
3837 |
+
"grad_norm": 2.265625,
|
3838 |
+
"learning_rate": 4.8881909547738695e-05,
|
3839 |
+
"loss": 7.6301,
|
3840 |
+
"step": 5450
|
3841 |
+
},
|
3842 |
+
{
|
3843 |
+
"epoch": 0.0273,
|
3844 |
+
"grad_norm": 2.390625,
|
3845 |
+
"learning_rate": 4.8879396984924626e-05,
|
3846 |
+
"loss": 7.6033,
|
3847 |
+
"step": 5460
|
3848 |
+
},
|
3849 |
+
{
|
3850 |
+
"epoch": 0.02735,
|
3851 |
+
"grad_norm": 2.578125,
|
3852 |
+
"learning_rate": 4.887688442211055e-05,
|
3853 |
+
"loss": 7.5836,
|
3854 |
+
"step": 5470
|
3855 |
+
},
|
3856 |
+
{
|
3857 |
+
"epoch": 0.0274,
|
3858 |
+
"grad_norm": 3.609375,
|
3859 |
+
"learning_rate": 4.887437185929649e-05,
|
3860 |
+
"loss": 7.5596,
|
3861 |
+
"step": 5480
|
3862 |
+
},
|
3863 |
+
{
|
3864 |
+
"epoch": 0.02745,
|
3865 |
+
"grad_norm": 12.25,
|
3866 |
+
"learning_rate": 4.887185929648241e-05,
|
3867 |
+
"loss": 7.5735,
|
3868 |
+
"step": 5490
|
3869 |
+
},
|
3870 |
+
{
|
3871 |
+
"epoch": 0.0275,
|
3872 |
+
"grad_norm": 1.8046875,
|
3873 |
+
"learning_rate": 4.8869346733668344e-05,
|
3874 |
+
"loss": 7.5858,
|
3875 |
+
"step": 5500
|
3876 |
+
},
|
3877 |
+
{
|
3878 |
+
"epoch": 0.02755,
|
3879 |
+
"grad_norm": 12.4375,
|
3880 |
+
"learning_rate": 4.8866834170854275e-05,
|
3881 |
+
"loss": 7.6097,
|
3882 |
+
"step": 5510
|
3883 |
+
},
|
3884 |
+
{
|
3885 |
+
"epoch": 0.0276,
|
3886 |
+
"grad_norm": 1.40625,
|
3887 |
+
"learning_rate": 4.8864321608040206e-05,
|
3888 |
+
"loss": 7.5665,
|
3889 |
+
"step": 5520
|
3890 |
+
},
|
3891 |
+
{
|
3892 |
+
"epoch": 0.02765,
|
3893 |
+
"grad_norm": 2.875,
|
3894 |
+
"learning_rate": 4.886180904522613e-05,
|
3895 |
+
"loss": 7.6013,
|
3896 |
+
"step": 5530
|
3897 |
+
},
|
3898 |
+
{
|
3899 |
+
"epoch": 0.0277,
|
3900 |
+
"grad_norm": 5.40625,
|
3901 |
+
"learning_rate": 4.885929648241207e-05,
|
3902 |
+
"loss": 7.5832,
|
3903 |
+
"step": 5540
|
3904 |
+
},
|
3905 |
+
{
|
3906 |
+
"epoch": 0.02775,
|
3907 |
+
"grad_norm": 1.2578125,
|
3908 |
+
"learning_rate": 4.885678391959799e-05,
|
3909 |
+
"loss": 7.5888,
|
3910 |
+
"step": 5550
|
3911 |
+
},
|
3912 |
+
{
|
3913 |
+
"epoch": 0.0278,
|
3914 |
+
"grad_norm": 4.65625,
|
3915 |
+
"learning_rate": 4.885427135678392e-05,
|
3916 |
+
"loss": 7.5835,
|
3917 |
+
"step": 5560
|
3918 |
+
},
|
3919 |
+
{
|
3920 |
+
"epoch": 0.02785,
|
3921 |
+
"grad_norm": 3.140625,
|
3922 |
+
"learning_rate": 4.8851758793969854e-05,
|
3923 |
+
"loss": 7.5709,
|
3924 |
+
"step": 5570
|
3925 |
+
},
|
3926 |
+
{
|
3927 |
+
"epoch": 0.0279,
|
3928 |
+
"grad_norm": 2.28125,
|
3929 |
+
"learning_rate": 4.8849246231155785e-05,
|
3930 |
+
"loss": 7.5727,
|
3931 |
+
"step": 5580
|
3932 |
+
},
|
3933 |
+
{
|
3934 |
+
"epoch": 0.02795,
|
3935 |
+
"grad_norm": 3.890625,
|
3936 |
+
"learning_rate": 4.884673366834171e-05,
|
3937 |
+
"loss": 7.572,
|
3938 |
+
"step": 5590
|
3939 |
+
},
|
3940 |
+
{
|
3941 |
+
"epoch": 0.028,
|
3942 |
+
"grad_norm": 2.421875,
|
3943 |
+
"learning_rate": 4.884422110552764e-05,
|
3944 |
+
"loss": 7.5692,
|
3945 |
+
"step": 5600
|
3946 |
+
},
|
3947 |
+
{
|
3948 |
+
"epoch": 0.02805,
|
3949 |
+
"grad_norm": 2.109375,
|
3950 |
+
"learning_rate": 4.884170854271357e-05,
|
3951 |
+
"loss": 7.6236,
|
3952 |
+
"step": 5610
|
3953 |
+
},
|
3954 |
+
{
|
3955 |
+
"epoch": 0.0281,
|
3956 |
+
"grad_norm": 1.46875,
|
3957 |
+
"learning_rate": 4.88391959798995e-05,
|
3958 |
+
"loss": 7.5949,
|
3959 |
+
"step": 5620
|
3960 |
+
},
|
3961 |
+
{
|
3962 |
+
"epoch": 0.02815,
|
3963 |
+
"grad_norm": 3.625,
|
3964 |
+
"learning_rate": 4.8836683417085426e-05,
|
3965 |
+
"loss": 7.5883,
|
3966 |
+
"step": 5630
|
3967 |
+
},
|
3968 |
+
{
|
3969 |
+
"epoch": 0.0282,
|
3970 |
+
"grad_norm": 1.484375,
|
3971 |
+
"learning_rate": 4.883417085427136e-05,
|
3972 |
+
"loss": 7.5879,
|
3973 |
+
"step": 5640
|
3974 |
+
},
|
3975 |
+
{
|
3976 |
+
"epoch": 0.02825,
|
3977 |
+
"grad_norm": 1.3359375,
|
3978 |
+
"learning_rate": 4.883165829145729e-05,
|
3979 |
+
"loss": 7.5975,
|
3980 |
+
"step": 5650
|
3981 |
+
},
|
3982 |
+
{
|
3983 |
+
"epoch": 0.0283,
|
3984 |
+
"grad_norm": 3.75,
|
3985 |
+
"learning_rate": 4.882914572864322e-05,
|
3986 |
+
"loss": 7.5806,
|
3987 |
+
"step": 5660
|
3988 |
+
},
|
3989 |
+
{
|
3990 |
+
"epoch": 0.02835,
|
3991 |
+
"grad_norm": 1.421875,
|
3992 |
+
"learning_rate": 4.882663316582915e-05,
|
3993 |
+
"loss": 7.5494,
|
3994 |
+
"step": 5670
|
3995 |
+
},
|
3996 |
+
{
|
3997 |
+
"epoch": 0.0284,
|
3998 |
+
"grad_norm": 1.796875,
|
3999 |
+
"learning_rate": 4.8824120603015075e-05,
|
4000 |
+
"loss": 7.5537,
|
4001 |
+
"step": 5680
|
4002 |
+
},
|
4003 |
+
{
|
4004 |
+
"epoch": 0.02845,
|
4005 |
+
"grad_norm": 5.6875,
|
4006 |
+
"learning_rate": 4.8821608040201006e-05,
|
4007 |
+
"loss": 7.5803,
|
4008 |
+
"step": 5690
|
4009 |
+
},
|
4010 |
+
{
|
4011 |
+
"epoch": 0.0285,
|
4012 |
+
"grad_norm": 1.6171875,
|
4013 |
+
"learning_rate": 4.881909547738694e-05,
|
4014 |
+
"loss": 7.6241,
|
4015 |
+
"step": 5700
|
4016 |
+
},
|
4017 |
+
{
|
4018 |
+
"epoch": 0.02855,
|
4019 |
+
"grad_norm": 25.125,
|
4020 |
+
"learning_rate": 4.881658291457287e-05,
|
4021 |
+
"loss": 7.5613,
|
4022 |
+
"step": 5710
|
4023 |
+
},
|
4024 |
+
{
|
4025 |
+
"epoch": 0.0286,
|
4026 |
+
"grad_norm": 5.65625,
|
4027 |
+
"learning_rate": 4.881407035175879e-05,
|
4028 |
+
"loss": 7.6088,
|
4029 |
+
"step": 5720
|
4030 |
+
},
|
4031 |
+
{
|
4032 |
+
"epoch": 0.02865,
|
4033 |
+
"grad_norm": 60.25,
|
4034 |
+
"learning_rate": 4.881155778894473e-05,
|
4035 |
+
"loss": 7.5541,
|
4036 |
+
"step": 5730
|
4037 |
+
},
|
4038 |
+
{
|
4039 |
+
"epoch": 0.0287,
|
4040 |
+
"grad_norm": 1.2890625,
|
4041 |
+
"learning_rate": 4.8809045226130654e-05,
|
4042 |
+
"loss": 7.5828,
|
4043 |
+
"step": 5740
|
4044 |
+
},
|
4045 |
+
{
|
4046 |
+
"epoch": 0.02875,
|
4047 |
+
"grad_norm": 21.375,
|
4048 |
+
"learning_rate": 4.8806532663316585e-05,
|
4049 |
+
"loss": 7.5942,
|
4050 |
+
"step": 5750
|
4051 |
+
},
|
4052 |
+
{
|
4053 |
+
"epoch": 0.0288,
|
4054 |
+
"grad_norm": 102.0,
|
4055 |
+
"learning_rate": 4.880402010050251e-05,
|
4056 |
+
"loss": 7.5712,
|
4057 |
+
"step": 5760
|
4058 |
+
},
|
4059 |
+
{
|
4060 |
+
"epoch": 0.02885,
|
4061 |
+
"grad_norm": 6.78125,
|
4062 |
+
"learning_rate": 4.880150753768845e-05,
|
4063 |
+
"loss": 7.5548,
|
4064 |
+
"step": 5770
|
4065 |
+
},
|
4066 |
+
{
|
4067 |
+
"epoch": 0.0289,
|
4068 |
+
"grad_norm": 1.6328125,
|
4069 |
+
"learning_rate": 4.879899497487437e-05,
|
4070 |
+
"loss": 7.5928,
|
4071 |
+
"step": 5780
|
4072 |
+
},
|
4073 |
+
{
|
4074 |
+
"epoch": 0.02895,
|
4075 |
+
"grad_norm": 1.71875,
|
4076 |
+
"learning_rate": 4.87964824120603e-05,
|
4077 |
+
"loss": 7.5883,
|
4078 |
+
"step": 5790
|
4079 |
+
},
|
4080 |
+
{
|
4081 |
+
"epoch": 0.029,
|
4082 |
+
"grad_norm": 24.125,
|
4083 |
+
"learning_rate": 4.8793969849246233e-05,
|
4084 |
+
"loss": 7.5893,
|
4085 |
+
"step": 5800
|
4086 |
+
},
|
4087 |
+
{
|
4088 |
+
"epoch": 0.02905,
|
4089 |
+
"grad_norm": 20.875,
|
4090 |
+
"learning_rate": 4.8791457286432164e-05,
|
4091 |
+
"loss": 7.5664,
|
4092 |
+
"step": 5810
|
4093 |
+
},
|
4094 |
+
{
|
4095 |
+
"epoch": 0.0291,
|
4096 |
+
"grad_norm": 51.0,
|
4097 |
+
"learning_rate": 4.878894472361809e-05,
|
4098 |
+
"loss": 7.5849,
|
4099 |
+
"step": 5820
|
4100 |
+
},
|
4101 |
+
{
|
4102 |
+
"epoch": 0.02915,
|
4103 |
+
"grad_norm": 12.25,
|
4104 |
+
"learning_rate": 4.8786432160804026e-05,
|
4105 |
+
"loss": 7.5563,
|
4106 |
+
"step": 5830
|
4107 |
+
},
|
4108 |
+
{
|
4109 |
+
"epoch": 0.0292,
|
4110 |
+
"grad_norm": 1.9921875,
|
4111 |
+
"learning_rate": 4.878391959798995e-05,
|
4112 |
+
"loss": 7.5565,
|
4113 |
+
"step": 5840
|
4114 |
+
},
|
4115 |
+
{
|
4116 |
+
"epoch": 0.02925,
|
4117 |
+
"grad_norm": 11.75,
|
4118 |
+
"learning_rate": 4.878140703517588e-05,
|
4119 |
+
"loss": 7.5521,
|
4120 |
+
"step": 5850
|
4121 |
+
},
|
4122 |
+
{
|
4123 |
+
"epoch": 0.0293,
|
4124 |
+
"grad_norm": 15.3125,
|
4125 |
+
"learning_rate": 4.877889447236181e-05,
|
4126 |
+
"loss": 7.6012,
|
4127 |
+
"step": 5860
|
4128 |
+
},
|
4129 |
+
{
|
4130 |
+
"epoch": 0.02935,
|
4131 |
+
"grad_norm": 45.25,
|
4132 |
+
"learning_rate": 4.8776381909547744e-05,
|
4133 |
+
"loss": 7.557,
|
4134 |
+
"step": 5870
|
4135 |
+
},
|
4136 |
+
{
|
4137 |
+
"epoch": 0.0294,
|
4138 |
+
"grad_norm": 1.8359375,
|
4139 |
+
"learning_rate": 4.877386934673367e-05,
|
4140 |
+
"loss": 7.5793,
|
4141 |
+
"step": 5880
|
4142 |
+
},
|
4143 |
+
{
|
4144 |
+
"epoch": 0.02945,
|
4145 |
+
"grad_norm": 6.96875,
|
4146 |
+
"learning_rate": 4.8771356783919606e-05,
|
4147 |
+
"loss": 7.5896,
|
4148 |
+
"step": 5890
|
4149 |
+
},
|
4150 |
+
{
|
4151 |
+
"epoch": 0.0295,
|
4152 |
+
"grad_norm": 1.6953125,
|
4153 |
+
"learning_rate": 4.876884422110553e-05,
|
4154 |
+
"loss": 7.5965,
|
4155 |
+
"step": 5900
|
4156 |
+
},
|
4157 |
+
{
|
4158 |
+
"epoch": 0.02955,
|
4159 |
+
"grad_norm": 2.046875,
|
4160 |
+
"learning_rate": 4.876633165829146e-05,
|
4161 |
+
"loss": 7.6091,
|
4162 |
+
"step": 5910
|
4163 |
+
},
|
4164 |
+
{
|
4165 |
+
"epoch": 0.0296,
|
4166 |
+
"grad_norm": 1.953125,
|
4167 |
+
"learning_rate": 4.8763819095477385e-05,
|
4168 |
+
"loss": 7.577,
|
4169 |
+
"step": 5920
|
4170 |
+
},
|
4171 |
+
{
|
4172 |
+
"epoch": 0.02965,
|
4173 |
+
"grad_norm": 1.53125,
|
4174 |
+
"learning_rate": 4.876130653266332e-05,
|
4175 |
+
"loss": 7.6014,
|
4176 |
+
"step": 5930
|
4177 |
+
},
|
4178 |
+
{
|
4179 |
+
"epoch": 0.0297,
|
4180 |
+
"grad_norm": 2.546875,
|
4181 |
+
"learning_rate": 4.875879396984925e-05,
|
4182 |
+
"loss": 7.6076,
|
4183 |
+
"step": 5940
|
4184 |
+
},
|
4185 |
+
{
|
4186 |
+
"epoch": 0.02975,
|
4187 |
+
"grad_norm": 48.5,
|
4188 |
+
"learning_rate": 4.875628140703518e-05,
|
4189 |
+
"loss": 7.5915,
|
4190 |
+
"step": 5950
|
4191 |
+
},
|
4192 |
+
{
|
4193 |
+
"epoch": 0.0298,
|
4194 |
+
"grad_norm": 16.375,
|
4195 |
+
"learning_rate": 4.875376884422111e-05,
|
4196 |
+
"loss": 7.5711,
|
4197 |
+
"step": 5960
|
4198 |
+
},
|
4199 |
+
{
|
4200 |
+
"epoch": 0.02985,
|
4201 |
+
"grad_norm": 2.140625,
|
4202 |
+
"learning_rate": 4.875125628140704e-05,
|
4203 |
+
"loss": 7.5665,
|
4204 |
+
"step": 5970
|
4205 |
+
},
|
4206 |
+
{
|
4207 |
+
"epoch": 0.0299,
|
4208 |
+
"grad_norm": 1.8125,
|
4209 |
+
"learning_rate": 4.8748743718592965e-05,
|
4210 |
+
"loss": 7.5443,
|
4211 |
+
"step": 5980
|
4212 |
+
},
|
4213 |
+
{
|
4214 |
+
"epoch": 0.02995,
|
4215 |
+
"grad_norm": 9.125,
|
4216 |
+
"learning_rate": 4.8746231155778896e-05,
|
4217 |
+
"loss": 7.5808,
|
4218 |
+
"step": 5990
|
4219 |
+
},
|
4220 |
+
{
|
4221 |
+
"epoch": 0.03,
|
4222 |
+
"grad_norm": 1.5703125,
|
4223 |
+
"learning_rate": 4.874371859296483e-05,
|
4224 |
+
"loss": 7.5749,
|
4225 |
+
"step": 6000
|
4226 |
+
},
|
4227 |
+
{
|
4228 |
+
"epoch": 0.03,
|
4229 |
+
"eval_loss": 7.596262454986572,
|
4230 |
+
"eval_runtime": 84.3987,
|
4231 |
+
"eval_samples_per_second": 29.621,
|
4232 |
+
"eval_steps_per_second": 0.474,
|
4233 |
+
"step": 6000
|
4234 |
}
|
4235 |
],
|
4236 |
"logging_steps": 10,
|
|
|
4250 |
"attributes": {}
|
4251 |
}
|
4252 |
},
|
4253 |
+
"total_flos": 2.380225711177728e+17,
|
4254 |
"train_batch_size": 64,
|
4255 |
"trial_name": null,
|
4256 |
"trial_params": null
|