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 +783 -3
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 570000656
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:abf51b0bde4f4516e3a2bf61ff16b0db2f00ab75455543141c7b71cb35a30750
|
| 3 |
size 570000656
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1622091
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ed728decad04a96e1b4e776374d534768a07aea8d060a2fb392cdcc62864c814
|
| 3 |
size 1622091
|
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 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5e5867d58a148bc81e08159e2e3724e0d0a99698f308c9342157abe9ddd58156
|
| 3 |
size 14469
|
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 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4980c8e9b7686eab28fc9789b9110330c13c0f23ed7c460116a946fbe7757faf
|
| 3 |
size 14469
|
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 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9a3a3c027333b6aea30cc3906e13408d60c9550a941487df9779594e03c20cd2
|
| 3 |
size 14469
|
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 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fdbb82c6d6029243dc741ed0885b78c4a844d17fb22e4ca9ed1920563ebf5afb
|
| 3 |
size 14469
|
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 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06c23439f83824a35eb66cda4f7406531987bb08dcc042e35d59248b94616e64
|
| 3 |
size 14469
|
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 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:96cb367a2058a5c675a1660d3af4ce2050cfb404bb097783d807641980f8dd39
|
| 3 |
size 14469
|
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 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e5a04bd93fe87df2320cde20a3158f9a585e5e972b775bc706879bb3a119221c
|
| 3 |
size 14469
|
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 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9af5b9224b7f5300ba0da9631c43a8601e21c96d7096e62f310c75c81f13b0e0
|
| 3 |
size 14469
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:451841ec4cf498c8f72515e07aa50b5a61b703c955791debe1c9eb115681dd44
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 100,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": false,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -3908,6 +3908,786 @@
|
|
| 3908 |
"eval_samples_per_second": 16.528,
|
| 3909 |
"eval_steps_per_second": 0.918,
|
| 3910 |
"step": 5000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3911 |
}
|
| 3912 |
],
|
| 3913 |
"logging_steps": 10,
|
|
@@ -3927,7 +4707,7 @@
|
|
| 3927 |
"attributes": {}
|
| 3928 |
}
|
| 3929 |
},
|
| 3930 |
-
"total_flos":
|
| 3931 |
"train_batch_size": 4,
|
| 3932 |
"trial_name": null,
|
| 3933 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.5454545454545454,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 6000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": false,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 3908 |
"eval_samples_per_second": 16.528,
|
| 3909 |
"eval_steps_per_second": 0.918,
|
| 3910 |
"step": 5000
|
| 3911 |
+
},
|
| 3912 |
+
{
|
| 3913 |
+
"epoch": 0.45545454545454545,
|
| 3914 |
+
"grad_norm": 80346641072128.0,
|
| 3915 |
+
"learning_rate": 5.69885069929346e-07,
|
| 3916 |
+
"loss": 26.32060546875,
|
| 3917 |
+
"step": 5010
|
| 3918 |
+
},
|
| 3919 |
+
{
|
| 3920 |
+
"epoch": 0.45636363636363636,
|
| 3921 |
+
"grad_norm": 422829933723648.0,
|
| 3922 |
+
"learning_rate": 5.684708074236349e-07,
|
| 3923 |
+
"loss": 26.738665771484374,
|
| 3924 |
+
"step": 5020
|
| 3925 |
+
},
|
| 3926 |
+
{
|
| 3927 |
+
"epoch": 0.4572727272727273,
|
| 3928 |
+
"grad_norm": 424351761432576.0,
|
| 3929 |
+
"learning_rate": 5.670559864226153e-07,
|
| 3930 |
+
"loss": 26.33757629394531,
|
| 3931 |
+
"step": 5030
|
| 3932 |
+
},
|
| 3933 |
+
{
|
| 3934 |
+
"epoch": 0.4581818181818182,
|
| 3935 |
+
"grad_norm": 121131155587072.0,
|
| 3936 |
+
"learning_rate": 5.656406184665472e-07,
|
| 3937 |
+
"loss": 26.6727783203125,
|
| 3938 |
+
"step": 5040
|
| 3939 |
+
},
|
| 3940 |
+
{
|
| 3941 |
+
"epoch": 0.4590909090909091,
|
| 3942 |
+
"grad_norm": 118593920434176.0,
|
| 3943 |
+
"learning_rate": 5.642247151001515e-07,
|
| 3944 |
+
"loss": 26.53905029296875,
|
| 3945 |
+
"step": 5050
|
| 3946 |
+
},
|
| 3947 |
+
{
|
| 3948 |
+
"epoch": 0.46,
|
| 3949 |
+
"grad_norm": 74622355636224.0,
|
| 3950 |
+
"learning_rate": 5.628082878725159e-07,
|
| 3951 |
+
"loss": 26.59415283203125,
|
| 3952 |
+
"step": 5060
|
| 3953 |
+
},
|
| 3954 |
+
{
|
| 3955 |
+
"epoch": 0.46090909090909093,
|
| 3956 |
+
"grad_norm": 131786524852224.0,
|
| 3957 |
+
"learning_rate": 5.61391348337002e-07,
|
| 3958 |
+
"loss": 26.788510131835938,
|
| 3959 |
+
"step": 5070
|
| 3960 |
+
},
|
| 3961 |
+
{
|
| 3962 |
+
"epoch": 0.4618181818181818,
|
| 3963 |
+
"grad_norm": 196731346092032.0,
|
| 3964 |
+
"learning_rate": 5.599739080511492e-07,
|
| 3965 |
+
"loss": 26.669586181640625,
|
| 3966 |
+
"step": 5080
|
| 3967 |
+
},
|
| 3968 |
+
{
|
| 3969 |
+
"epoch": 0.4627272727272727,
|
| 3970 |
+
"grad_norm": 182333021880320.0,
|
| 3971 |
+
"learning_rate": 5.585559785765819e-07,
|
| 3972 |
+
"loss": 26.678341674804688,
|
| 3973 |
+
"step": 5090
|
| 3974 |
+
},
|
| 3975 |
+
{
|
| 3976 |
+
"epoch": 0.4636363636363636,
|
| 3977 |
+
"grad_norm": 352578864939008.0,
|
| 3978 |
+
"learning_rate": 5.571375714789147e-07,
|
| 3979 |
+
"loss": 26.814114379882813,
|
| 3980 |
+
"step": 5100
|
| 3981 |
+
},
|
| 3982 |
+
{
|
| 3983 |
+
"epoch": 0.4636363636363636,
|
| 3984 |
+
"eval_loss": 26.3111515045166,
|
| 3985 |
+
"eval_runtime": 2.2022,
|
| 3986 |
+
"eval_samples_per_second": 16.348,
|
| 3987 |
+
"eval_steps_per_second": 0.908,
|
| 3988 |
+
"step": 5100
|
| 3989 |
+
},
|
| 3990 |
+
{
|
| 3991 |
+
"epoch": 0.46454545454545454,
|
| 3992 |
+
"grad_norm": 518492847603712.0,
|
| 3993 |
+
"learning_rate": 5.557186983276576e-07,
|
| 3994 |
+
"loss": 26.460382080078126,
|
| 3995 |
+
"step": 5110
|
| 3996 |
+
},
|
| 3997 |
+
{
|
| 3998 |
+
"epoch": 0.46545454545454545,
|
| 3999 |
+
"grad_norm": 190987498422272.0,
|
| 4000 |
+
"learning_rate": 5.542993706961225e-07,
|
| 4001 |
+
"loss": 26.442779541015625,
|
| 4002 |
+
"step": 5120
|
| 4003 |
+
},
|
| 4004 |
+
{
|
| 4005 |
+
"epoch": 0.46636363636363637,
|
| 4006 |
+
"grad_norm": 172608712605696.0,
|
| 4007 |
+
"learning_rate": 5.528796001613282e-07,
|
| 4008 |
+
"loss": 26.353005981445314,
|
| 4009 |
+
"step": 5130
|
| 4010 |
+
},
|
| 4011 |
+
{
|
| 4012 |
+
"epoch": 0.4672727272727273,
|
| 4013 |
+
"grad_norm": 422501872041984.0,
|
| 4014 |
+
"learning_rate": 5.514593983039058e-07,
|
| 4015 |
+
"loss": 26.381781005859374,
|
| 4016 |
+
"step": 5140
|
| 4017 |
+
},
|
| 4018 |
+
{
|
| 4019 |
+
"epoch": 0.4681818181818182,
|
| 4020 |
+
"grad_norm": 560177182932992.0,
|
| 4021 |
+
"learning_rate": 5.500387767080053e-07,
|
| 4022 |
+
"loss": 26.472189331054686,
|
| 4023 |
+
"step": 5150
|
| 4024 |
+
},
|
| 4025 |
+
{
|
| 4026 |
+
"epoch": 0.4690909090909091,
|
| 4027 |
+
"grad_norm": 265595475984384.0,
|
| 4028 |
+
"learning_rate": 5.486177469611998e-07,
|
| 4029 |
+
"loss": 26.500082397460936,
|
| 4030 |
+
"step": 5160
|
| 4031 |
+
},
|
| 4032 |
+
{
|
| 4033 |
+
"epoch": 0.47,
|
| 4034 |
+
"grad_norm": 230636791005184.0,
|
| 4035 |
+
"learning_rate": 5.471963206543913e-07,
|
| 4036 |
+
"loss": 26.208493041992188,
|
| 4037 |
+
"step": 5170
|
| 4038 |
+
},
|
| 4039 |
+
{
|
| 4040 |
+
"epoch": 0.4709090909090909,
|
| 4041 |
+
"grad_norm": 236991077679104.0,
|
| 4042 |
+
"learning_rate": 5.457745093817174e-07,
|
| 4043 |
+
"loss": 26.544073486328124,
|
| 4044 |
+
"step": 5180
|
| 4045 |
+
},
|
| 4046 |
+
{
|
| 4047 |
+
"epoch": 0.4718181818181818,
|
| 4048 |
+
"grad_norm": 428384232407040.0,
|
| 4049 |
+
"learning_rate": 5.44352324740455e-07,
|
| 4050 |
+
"loss": 26.5481201171875,
|
| 4051 |
+
"step": 5190
|
| 4052 |
+
},
|
| 4053 |
+
{
|
| 4054 |
+
"epoch": 0.4727272727272727,
|
| 4055 |
+
"grad_norm": 211124435812352.0,
|
| 4056 |
+
"learning_rate": 5.429297783309264e-07,
|
| 4057 |
+
"loss": 26.475244140625,
|
| 4058 |
+
"step": 5200
|
| 4059 |
+
},
|
| 4060 |
+
{
|
| 4061 |
+
"epoch": 0.4727272727272727,
|
| 4062 |
+
"eval_loss": 26.310951232910156,
|
| 4063 |
+
"eval_runtime": 2.153,
|
| 4064 |
+
"eval_samples_per_second": 16.721,
|
| 4065 |
+
"eval_steps_per_second": 0.929,
|
| 4066 |
+
"step": 5200
|
| 4067 |
+
},
|
| 4068 |
+
{
|
| 4069 |
+
"epoch": 0.47363636363636363,
|
| 4070 |
+
"grad_norm": 181685639446528.0,
|
| 4071 |
+
"learning_rate": 5.415068817564053e-07,
|
| 4072 |
+
"loss": 26.813641357421876,
|
| 4073 |
+
"step": 5210
|
| 4074 |
+
},
|
| 4075 |
+
{
|
| 4076 |
+
"epoch": 0.47454545454545455,
|
| 4077 |
+
"grad_norm": 324085380808704.0,
|
| 4078 |
+
"learning_rate": 5.400836466230208e-07,
|
| 4079 |
+
"loss": 26.601318359375,
|
| 4080 |
+
"step": 5220
|
| 4081 |
+
},
|
| 4082 |
+
{
|
| 4083 |
+
"epoch": 0.47545454545454546,
|
| 4084 |
+
"grad_norm": 476272278896640.0,
|
| 4085 |
+
"learning_rate": 5.386600845396642e-07,
|
| 4086 |
+
"loss": 26.603497314453126,
|
| 4087 |
+
"step": 5230
|
| 4088 |
+
},
|
| 4089 |
+
{
|
| 4090 |
+
"epoch": 0.4763636363636364,
|
| 4091 |
+
"grad_norm": 261529517686784.0,
|
| 4092 |
+
"learning_rate": 5.372362071178936e-07,
|
| 4093 |
+
"loss": 26.458706665039063,
|
| 4094 |
+
"step": 5240
|
| 4095 |
+
},
|
| 4096 |
+
{
|
| 4097 |
+
"epoch": 0.4772727272727273,
|
| 4098 |
+
"grad_norm": 607072420888576.0,
|
| 4099 |
+
"learning_rate": 5.358120259718388e-07,
|
| 4100 |
+
"loss": 26.715435791015626,
|
| 4101 |
+
"step": 5250
|
| 4102 |
+
},
|
| 4103 |
+
{
|
| 4104 |
+
"epoch": 0.4781818181818182,
|
| 4105 |
+
"grad_norm": 98573517586432.0,
|
| 4106 |
+
"learning_rate": 5.343875527181072e-07,
|
| 4107 |
+
"loss": 26.47364501953125,
|
| 4108 |
+
"step": 5260
|
| 4109 |
+
},
|
| 4110 |
+
{
|
| 4111 |
+
"epoch": 0.47909090909090907,
|
| 4112 |
+
"grad_norm": 135659385782272.0,
|
| 4113 |
+
"learning_rate": 5.32962798975689e-07,
|
| 4114 |
+
"loss": 26.717742919921875,
|
| 4115 |
+
"step": 5270
|
| 4116 |
+
},
|
| 4117 |
+
{
|
| 4118 |
+
"epoch": 0.48,
|
| 4119 |
+
"grad_norm": 387802562822144.0,
|
| 4120 |
+
"learning_rate": 5.315377763658617e-07,
|
| 4121 |
+
"loss": 26.490786743164062,
|
| 4122 |
+
"step": 5280
|
| 4123 |
+
},
|
| 4124 |
+
{
|
| 4125 |
+
"epoch": 0.4809090909090909,
|
| 4126 |
+
"grad_norm": 418252572327936.0,
|
| 4127 |
+
"learning_rate": 5.301124965120968e-07,
|
| 4128 |
+
"loss": 26.737353515625,
|
| 4129 |
+
"step": 5290
|
| 4130 |
+
},
|
| 4131 |
+
{
|
| 4132 |
+
"epoch": 0.4818181818181818,
|
| 4133 |
+
"grad_norm": 214300849340416.0,
|
| 4134 |
+
"learning_rate": 5.28686971039963e-07,
|
| 4135 |
+
"loss": 26.633428955078124,
|
| 4136 |
+
"step": 5300
|
| 4137 |
+
},
|
| 4138 |
+
{
|
| 4139 |
+
"epoch": 0.4818181818181818,
|
| 4140 |
+
"eval_loss": 26.304325103759766,
|
| 4141 |
+
"eval_runtime": 2.1933,
|
| 4142 |
+
"eval_samples_per_second": 16.414,
|
| 4143 |
+
"eval_steps_per_second": 0.912,
|
| 4144 |
+
"step": 5300
|
| 4145 |
+
},
|
| 4146 |
+
{
|
| 4147 |
+
"epoch": 0.4827272727272727,
|
| 4148 |
+
"grad_norm": 477280589578240.0,
|
| 4149 |
+
"learning_rate": 5.272612115770332e-07,
|
| 4150 |
+
"loss": 26.399496459960936,
|
| 4151 |
+
"step": 5310
|
| 4152 |
+
},
|
| 4153 |
+
{
|
| 4154 |
+
"epoch": 0.48363636363636364,
|
| 4155 |
+
"grad_norm": 231165105537024.0,
|
| 4156 |
+
"learning_rate": 5.258352297527887e-07,
|
| 4157 |
+
"loss": 26.705911254882814,
|
| 4158 |
+
"step": 5320
|
| 4159 |
+
},
|
| 4160 |
+
{
|
| 4161 |
+
"epoch": 0.48454545454545456,
|
| 4162 |
+
"grad_norm": 193821488971776.0,
|
| 4163 |
+
"learning_rate": 5.244090371985243e-07,
|
| 4164 |
+
"loss": 27.068051147460938,
|
| 4165 |
+
"step": 5330
|
| 4166 |
+
},
|
| 4167 |
+
{
|
| 4168 |
+
"epoch": 0.48545454545454547,
|
| 4169 |
+
"grad_norm": 370139140718592.0,
|
| 4170 |
+
"learning_rate": 5.229826455472541e-07,
|
| 4171 |
+
"loss": 26.2793212890625,
|
| 4172 |
+
"step": 5340
|
| 4173 |
+
},
|
| 4174 |
+
{
|
| 4175 |
+
"epoch": 0.4863636363636364,
|
| 4176 |
+
"grad_norm": 66449884315648.0,
|
| 4177 |
+
"learning_rate": 5.215560664336157e-07,
|
| 4178 |
+
"loss": 26.93189392089844,
|
| 4179 |
+
"step": 5350
|
| 4180 |
+
},
|
| 4181 |
+
{
|
| 4182 |
+
"epoch": 0.48727272727272725,
|
| 4183 |
+
"grad_norm": 114903520116736.0,
|
| 4184 |
+
"learning_rate": 5.201293114937759e-07,
|
| 4185 |
+
"loss": 26.350509643554688,
|
| 4186 |
+
"step": 5360
|
| 4187 |
+
},
|
| 4188 |
+
{
|
| 4189 |
+
"epoch": 0.48818181818181816,
|
| 4190 |
+
"grad_norm": 163418724106240.0,
|
| 4191 |
+
"learning_rate": 5.18702392365336e-07,
|
| 4192 |
+
"loss": 26.36776123046875,
|
| 4193 |
+
"step": 5370
|
| 4194 |
+
},
|
| 4195 |
+
{
|
| 4196 |
+
"epoch": 0.4890909090909091,
|
| 4197 |
+
"grad_norm": 386242080407552.0,
|
| 4198 |
+
"learning_rate": 5.172753206872362e-07,
|
| 4199 |
+
"loss": 26.549008178710938,
|
| 4200 |
+
"step": 5380
|
| 4201 |
+
},
|
| 4202 |
+
{
|
| 4203 |
+
"epoch": 0.49,
|
| 4204 |
+
"grad_norm": 84084361527296.0,
|
| 4205 |
+
"learning_rate": 5.158481080996609e-07,
|
| 4206 |
+
"loss": 26.612332153320313,
|
| 4207 |
+
"step": 5390
|
| 4208 |
+
},
|
| 4209 |
+
{
|
| 4210 |
+
"epoch": 0.4909090909090909,
|
| 4211 |
+
"grad_norm": 408571950923776.0,
|
| 4212 |
+
"learning_rate": 5.144207662439443e-07,
|
| 4213 |
+
"loss": 26.4984619140625,
|
| 4214 |
+
"step": 5400
|
| 4215 |
+
},
|
| 4216 |
+
{
|
| 4217 |
+
"epoch": 0.4909090909090909,
|
| 4218 |
+
"eval_loss": 26.30891990661621,
|
| 4219 |
+
"eval_runtime": 2.2008,
|
| 4220 |
+
"eval_samples_per_second": 16.358,
|
| 4221 |
+
"eval_steps_per_second": 0.909,
|
| 4222 |
+
"step": 5400
|
| 4223 |
+
},
|
| 4224 |
+
{
|
| 4225 |
+
"epoch": 0.4918181818181818,
|
| 4226 |
+
"grad_norm": 392391265615872.0,
|
| 4227 |
+
"learning_rate": 5.129933067624745e-07,
|
| 4228 |
+
"loss": 26.757540893554687,
|
| 4229 |
+
"step": 5410
|
| 4230 |
+
},
|
| 4231 |
+
{
|
| 4232 |
+
"epoch": 0.49272727272727274,
|
| 4233 |
+
"grad_norm": 343983997845504.0,
|
| 4234 |
+
"learning_rate": 5.115657412985993e-07,
|
| 4235 |
+
"loss": 26.5795654296875,
|
| 4236 |
+
"step": 5420
|
| 4237 |
+
},
|
| 4238 |
+
{
|
| 4239 |
+
"epoch": 0.49363636363636365,
|
| 4240 |
+
"grad_norm": 276295011270656.0,
|
| 4241 |
+
"learning_rate": 5.101380814965313e-07,
|
| 4242 |
+
"loss": 26.277252197265625,
|
| 4243 |
+
"step": 5430
|
| 4244 |
+
},
|
| 4245 |
+
{
|
| 4246 |
+
"epoch": 0.49454545454545457,
|
| 4247 |
+
"grad_norm": 225630402641920.0,
|
| 4248 |
+
"learning_rate": 5.087103390012517e-07,
|
| 4249 |
+
"loss": 26.9617919921875,
|
| 4250 |
+
"step": 5440
|
| 4251 |
+
},
|
| 4252 |
+
{
|
| 4253 |
+
"epoch": 0.4954545454545455,
|
| 4254 |
+
"grad_norm": 110936799051776.0,
|
| 4255 |
+
"learning_rate": 5.072825254584171e-07,
|
| 4256 |
+
"loss": 26.697900390625,
|
| 4257 |
+
"step": 5450
|
| 4258 |
+
},
|
| 4259 |
+
{
|
| 4260 |
+
"epoch": 0.49636363636363634,
|
| 4261 |
+
"grad_norm": 429198766243840.0,
|
| 4262 |
+
"learning_rate": 5.058546525142631e-07,
|
| 4263 |
+
"loss": 26.221038818359375,
|
| 4264 |
+
"step": 5460
|
| 4265 |
+
},
|
| 4266 |
+
{
|
| 4267 |
+
"epoch": 0.49727272727272726,
|
| 4268 |
+
"grad_norm": 307539992379392.0,
|
| 4269 |
+
"learning_rate": 5.044267318155098e-07,
|
| 4270 |
+
"loss": 26.596530151367187,
|
| 4271 |
+
"step": 5470
|
| 4272 |
+
},
|
| 4273 |
+
{
|
| 4274 |
+
"epoch": 0.49818181818181817,
|
| 4275 |
+
"grad_norm": 377537322549248.0,
|
| 4276 |
+
"learning_rate": 5.029987750092675e-07,
|
| 4277 |
+
"loss": 26.367523193359375,
|
| 4278 |
+
"step": 5480
|
| 4279 |
+
},
|
| 4280 |
+
{
|
| 4281 |
+
"epoch": 0.4990909090909091,
|
| 4282 |
+
"grad_norm": 248897851097088.0,
|
| 4283 |
+
"learning_rate": 5.015707937429397e-07,
|
| 4284 |
+
"loss": 26.702847290039063,
|
| 4285 |
+
"step": 5490
|
| 4286 |
+
},
|
| 4287 |
+
{
|
| 4288 |
+
"epoch": 0.5,
|
| 4289 |
+
"grad_norm": 572252919693312.0,
|
| 4290 |
+
"learning_rate": 5.00142799664131e-07,
|
| 4291 |
+
"loss": 26.316940307617188,
|
| 4292 |
+
"step": 5500
|
| 4293 |
+
},
|
| 4294 |
+
{
|
| 4295 |
+
"epoch": 0.5,
|
| 4296 |
+
"eval_loss": 26.30450439453125,
|
| 4297 |
+
"eval_runtime": 2.1679,
|
| 4298 |
+
"eval_samples_per_second": 16.606,
|
| 4299 |
+
"eval_steps_per_second": 0.923,
|
| 4300 |
+
"step": 5500
|
| 4301 |
+
},
|
| 4302 |
+
{
|
| 4303 |
+
"epoch": 0.5009090909090909,
|
| 4304 |
+
"grad_norm": 194077274406912.0,
|
| 4305 |
+
"learning_rate": 4.987148044205492e-07,
|
| 4306 |
+
"loss": 26.6015625,
|
| 4307 |
+
"step": 5510
|
| 4308 |
+
},
|
| 4309 |
+
{
|
| 4310 |
+
"epoch": 0.5018181818181818,
|
| 4311 |
+
"grad_norm": 262825356296192.0,
|
| 4312 |
+
"learning_rate": 4.972868196599125e-07,
|
| 4313 |
+
"loss": 26.622930908203124,
|
| 4314 |
+
"step": 5520
|
| 4315 |
+
},
|
| 4316 |
+
{
|
| 4317 |
+
"epoch": 0.5027272727272727,
|
| 4318 |
+
"grad_norm": 173687571480576.0,
|
| 4319 |
+
"learning_rate": 4.958588570298526e-07,
|
| 4320 |
+
"loss": 26.34784851074219,
|
| 4321 |
+
"step": 5530
|
| 4322 |
+
},
|
| 4323 |
+
{
|
| 4324 |
+
"epoch": 0.5036363636363637,
|
| 4325 |
+
"grad_norm": 280642273148928.0,
|
| 4326 |
+
"learning_rate": 4.944309281778223e-07,
|
| 4327 |
+
"loss": 26.642620849609376,
|
| 4328 |
+
"step": 5540
|
| 4329 |
+
},
|
| 4330 |
+
{
|
| 4331 |
+
"epoch": 0.5045454545454545,
|
| 4332 |
+
"grad_norm": 293827940188160.0,
|
| 4333 |
+
"learning_rate": 4.930030447509968e-07,
|
| 4334 |
+
"loss": 26.564447021484376,
|
| 4335 |
+
"step": 5550
|
| 4336 |
+
},
|
| 4337 |
+
{
|
| 4338 |
+
"epoch": 0.5054545454545455,
|
| 4339 |
+
"grad_norm": 277824522944512.0,
|
| 4340 |
+
"learning_rate": 4.915752183961826e-07,
|
| 4341 |
+
"loss": 26.670132446289063,
|
| 4342 |
+
"step": 5560
|
| 4343 |
+
},
|
| 4344 |
+
{
|
| 4345 |
+
"epoch": 0.5063636363636363,
|
| 4346 |
+
"grad_norm": 107426208546816.0,
|
| 4347 |
+
"learning_rate": 4.901474607597196e-07,
|
| 4348 |
+
"loss": 26.282070922851563,
|
| 4349 |
+
"step": 5570
|
| 4350 |
+
},
|
| 4351 |
+
{
|
| 4352 |
+
"epoch": 0.5072727272727273,
|
| 4353 |
+
"grad_norm": 258411438538752.0,
|
| 4354 |
+
"learning_rate": 4.887197834873877e-07,
|
| 4355 |
+
"loss": 26.355166625976562,
|
| 4356 |
+
"step": 5580
|
| 4357 |
+
},
|
| 4358 |
+
{
|
| 4359 |
+
"epoch": 0.5081818181818182,
|
| 4360 |
+
"grad_norm": 140088537251840.0,
|
| 4361 |
+
"learning_rate": 4.872921982243111e-07,
|
| 4362 |
+
"loss": 26.558563232421875,
|
| 4363 |
+
"step": 5590
|
| 4364 |
+
},
|
| 4365 |
+
{
|
| 4366 |
+
"epoch": 0.509090909090909,
|
| 4367 |
+
"grad_norm": 138965688516608.0,
|
| 4368 |
+
"learning_rate": 4.858647166148634e-07,
|
| 4369 |
+
"loss": 26.477615356445312,
|
| 4370 |
+
"step": 5600
|
| 4371 |
+
},
|
| 4372 |
+
{
|
| 4373 |
+
"epoch": 0.509090909090909,
|
| 4374 |
+
"eval_loss": 26.29946517944336,
|
| 4375 |
+
"eval_runtime": 2.216,
|
| 4376 |
+
"eval_samples_per_second": 16.245,
|
| 4377 |
+
"eval_steps_per_second": 0.903,
|
| 4378 |
+
"step": 5600
|
| 4379 |
+
},
|
| 4380 |
+
{
|
| 4381 |
+
"epoch": 0.51,
|
| 4382 |
+
"grad_norm": 627078646988800.0,
|
| 4383 |
+
"learning_rate": 4.84437350302573e-07,
|
| 4384 |
+
"loss": 26.314041137695312,
|
| 4385 |
+
"step": 5610
|
| 4386 |
+
},
|
| 4387 |
+
{
|
| 4388 |
+
"epoch": 0.5109090909090909,
|
| 4389 |
+
"grad_norm": 376094112874496.0,
|
| 4390 |
+
"learning_rate": 4.830101109300277e-07,
|
| 4391 |
+
"loss": 26.738397216796876,
|
| 4392 |
+
"step": 5620
|
| 4393 |
+
},
|
| 4394 |
+
{
|
| 4395 |
+
"epoch": 0.5118181818181818,
|
| 4396 |
+
"grad_norm": 111444955758592.0,
|
| 4397 |
+
"learning_rate": 4.815830101387799e-07,
|
| 4398 |
+
"loss": 26.68155517578125,
|
| 4399 |
+
"step": 5630
|
| 4400 |
+
},
|
| 4401 |
+
{
|
| 4402 |
+
"epoch": 0.5127272727272727,
|
| 4403 |
+
"grad_norm": 182908782379008.0,
|
| 4404 |
+
"learning_rate": 4.801560595692515e-07,
|
| 4405 |
+
"loss": 26.290713500976562,
|
| 4406 |
+
"step": 5640
|
| 4407 |
+
},
|
| 4408 |
+
{
|
| 4409 |
+
"epoch": 0.5136363636363637,
|
| 4410 |
+
"grad_norm": 321908033716224.0,
|
| 4411 |
+
"learning_rate": 4.787292708606394e-07,
|
| 4412 |
+
"loss": 26.5162353515625,
|
| 4413 |
+
"step": 5650
|
| 4414 |
+
},
|
| 4415 |
+
{
|
| 4416 |
+
"epoch": 0.5145454545454545,
|
| 4417 |
+
"grad_norm": 203153714708480.0,
|
| 4418 |
+
"learning_rate": 4.773026556508201e-07,
|
| 4419 |
+
"loss": 26.120401000976564,
|
| 4420 |
+
"step": 5660
|
| 4421 |
+
},
|
| 4422 |
+
{
|
| 4423 |
+
"epoch": 0.5154545454545455,
|
| 4424 |
+
"grad_norm": 395124106330112.0,
|
| 4425 |
+
"learning_rate": 4.7587622557625485e-07,
|
| 4426 |
+
"loss": 26.68848876953125,
|
| 4427 |
+
"step": 5670
|
| 4428 |
+
},
|
| 4429 |
+
{
|
| 4430 |
+
"epoch": 0.5163636363636364,
|
| 4431 |
+
"grad_norm": 515408322887680.0,
|
| 4432 |
+
"learning_rate": 4.744499922718948e-07,
|
| 4433 |
+
"loss": 26.975100708007812,
|
| 4434 |
+
"step": 5680
|
| 4435 |
+
},
|
| 4436 |
+
{
|
| 4437 |
+
"epoch": 0.5172727272727272,
|
| 4438 |
+
"grad_norm": 415264516603904.0,
|
| 4439 |
+
"learning_rate": 4.7302396737108643e-07,
|
| 4440 |
+
"loss": 26.722061157226562,
|
| 4441 |
+
"step": 5690
|
| 4442 |
+
},
|
| 4443 |
+
{
|
| 4444 |
+
"epoch": 0.5181818181818182,
|
| 4445 |
+
"grad_norm": 198320131670016.0,
|
| 4446 |
+
"learning_rate": 4.7159816250547593e-07,
|
| 4447 |
+
"loss": 27.05069580078125,
|
| 4448 |
+
"step": 5700
|
| 4449 |
+
},
|
| 4450 |
+
{
|
| 4451 |
+
"epoch": 0.5181818181818182,
|
| 4452 |
+
"eval_loss": 26.303539276123047,
|
| 4453 |
+
"eval_runtime": 2.1678,
|
| 4454 |
+
"eval_samples_per_second": 16.607,
|
| 4455 |
+
"eval_steps_per_second": 0.923,
|
| 4456 |
+
"step": 5700
|
| 4457 |
+
},
|
| 4458 |
+
{
|
| 4459 |
+
"epoch": 0.519090909090909,
|
| 4460 |
+
"grad_norm": 350612004798464.0,
|
| 4461 |
+
"learning_rate": 4.7017258930491466e-07,
|
| 4462 |
+
"loss": 26.506210327148438,
|
| 4463 |
+
"step": 5710
|
| 4464 |
+
},
|
| 4465 |
+
{
|
| 4466 |
+
"epoch": 0.52,
|
| 4467 |
+
"grad_norm": 353382929793024.0,
|
| 4468 |
+
"learning_rate": 4.687472593973649e-07,
|
| 4469 |
+
"loss": 26.487466430664064,
|
| 4470 |
+
"step": 5720
|
| 4471 |
+
},
|
| 4472 |
+
{
|
| 4473 |
+
"epoch": 0.5209090909090909,
|
| 4474 |
+
"grad_norm": 128485766987776.0,
|
| 4475 |
+
"learning_rate": 4.673221844088039e-07,
|
| 4476 |
+
"loss": 26.61768798828125,
|
| 4477 |
+
"step": 5730
|
| 4478 |
+
},
|
| 4479 |
+
{
|
| 4480 |
+
"epoch": 0.5218181818181818,
|
| 4481 |
+
"grad_norm": 176263411007488.0,
|
| 4482 |
+
"learning_rate": 4.6589737596313005e-07,
|
| 4483 |
+
"loss": 26.494561767578126,
|
| 4484 |
+
"step": 5740
|
| 4485 |
+
},
|
| 4486 |
+
{
|
| 4487 |
+
"epoch": 0.5227272727272727,
|
| 4488 |
+
"grad_norm": 490709543026688.0,
|
| 4489 |
+
"learning_rate": 4.6447284568206727e-07,
|
| 4490 |
+
"loss": 26.74195556640625,
|
| 4491 |
+
"step": 5750
|
| 4492 |
+
},
|
| 4493 |
+
{
|
| 4494 |
+
"epoch": 0.5236363636363637,
|
| 4495 |
+
"grad_norm": 439582084562944.0,
|
| 4496 |
+
"learning_rate": 4.63048605185071e-07,
|
| 4497 |
+
"loss": 26.69200744628906,
|
| 4498 |
+
"step": 5760
|
| 4499 |
+
},
|
| 4500 |
+
{
|
| 4501 |
+
"epoch": 0.5245454545454545,
|
| 4502 |
+
"grad_norm": 155888790798336.0,
|
| 4503 |
+
"learning_rate": 4.616246660892323e-07,
|
| 4504 |
+
"loss": 26.530010986328126,
|
| 4505 |
+
"step": 5770
|
| 4506 |
+
},
|
| 4507 |
+
{
|
| 4508 |
+
"epoch": 0.5254545454545455,
|
| 4509 |
+
"grad_norm": 313929695756288.0,
|
| 4510 |
+
"learning_rate": 4.6020104000918487e-07,
|
| 4511 |
+
"loss": 26.343841552734375,
|
| 4512 |
+
"step": 5780
|
| 4513 |
+
},
|
| 4514 |
+
{
|
| 4515 |
+
"epoch": 0.5263636363636364,
|
| 4516 |
+
"grad_norm": 59153334665216.0,
|
| 4517 |
+
"learning_rate": 4.587777385570081e-07,
|
| 4518 |
+
"loss": 26.92261962890625,
|
| 4519 |
+
"step": 5790
|
| 4520 |
+
},
|
| 4521 |
+
{
|
| 4522 |
+
"epoch": 0.5272727272727272,
|
| 4523 |
+
"grad_norm": 321196008669184.0,
|
| 4524 |
+
"learning_rate": 4.5735477334213403e-07,
|
| 4525 |
+
"loss": 26.5276123046875,
|
| 4526 |
+
"step": 5800
|
| 4527 |
+
},
|
| 4528 |
+
{
|
| 4529 |
+
"epoch": 0.5272727272727272,
|
| 4530 |
+
"eval_loss": 26.29823112487793,
|
| 4531 |
+
"eval_runtime": 2.1826,
|
| 4532 |
+
"eval_samples_per_second": 16.494,
|
| 4533 |
+
"eval_steps_per_second": 0.916,
|
| 4534 |
+
"step": 5800
|
| 4535 |
+
},
|
| 4536 |
+
{
|
| 4537 |
+
"epoch": 0.5281818181818182,
|
| 4538 |
+
"grad_norm": 115456698482688.0,
|
| 4539 |
+
"learning_rate": 4.5593215597125246e-07,
|
| 4540 |
+
"loss": 26.302325439453124,
|
| 4541 |
+
"step": 5810
|
| 4542 |
+
},
|
| 4543 |
+
{
|
| 4544 |
+
"epoch": 0.5290909090909091,
|
| 4545 |
+
"grad_norm": 54434067382272.0,
|
| 4546 |
+
"learning_rate": 4.54509898048215e-07,
|
| 4547 |
+
"loss": 26.550509643554687,
|
| 4548 |
+
"step": 5820
|
| 4549 |
+
},
|
| 4550 |
+
{
|
| 4551 |
+
"epoch": 0.53,
|
| 4552 |
+
"grad_norm": 117245166485504.0,
|
| 4553 |
+
"learning_rate": 4.530880111739424e-07,
|
| 4554 |
+
"loss": 26.198770141601564,
|
| 4555 |
+
"step": 5830
|
| 4556 |
+
},
|
| 4557 |
+
{
|
| 4558 |
+
"epoch": 0.5309090909090909,
|
| 4559 |
+
"grad_norm": 174198236381184.0,
|
| 4560 |
+
"learning_rate": 4.5166650694632813e-07,
|
| 4561 |
+
"loss": 26.7167236328125,
|
| 4562 |
+
"step": 5840
|
| 4563 |
+
},
|
| 4564 |
+
{
|
| 4565 |
+
"epoch": 0.5318181818181819,
|
| 4566 |
+
"grad_norm": 202826961649664.0,
|
| 4567 |
+
"learning_rate": 4.5024539696014496e-07,
|
| 4568 |
+
"loss": 26.728240966796875,
|
| 4569 |
+
"step": 5850
|
| 4570 |
+
},
|
| 4571 |
+
{
|
| 4572 |
+
"epoch": 0.5327272727272727,
|
| 4573 |
+
"grad_norm": 183662431698944.0,
|
| 4574 |
+
"learning_rate": 4.488246928069496e-07,
|
| 4575 |
+
"loss": 26.686666870117186,
|
| 4576 |
+
"step": 5860
|
| 4577 |
+
},
|
| 4578 |
+
{
|
| 4579 |
+
"epoch": 0.5336363636363637,
|
| 4580 |
+
"grad_norm": 132447521996800.0,
|
| 4581 |
+
"learning_rate": 4.474044060749889e-07,
|
| 4582 |
+
"loss": 26.561785888671874,
|
| 4583 |
+
"step": 5870
|
| 4584 |
+
},
|
| 4585 |
+
{
|
| 4586 |
+
"epoch": 0.5345454545454545,
|
| 4587 |
+
"grad_norm": 294536173584384.0,
|
| 4588 |
+
"learning_rate": 4.459845483491045e-07,
|
| 4589 |
+
"loss": 26.386618041992186,
|
| 4590 |
+
"step": 5880
|
| 4591 |
+
},
|
| 4592 |
+
{
|
| 4593 |
+
"epoch": 0.5354545454545454,
|
| 4594 |
+
"grad_norm": 566112055984128.0,
|
| 4595 |
+
"learning_rate": 4.44565131210639e-07,
|
| 4596 |
+
"loss": 26.43914794921875,
|
| 4597 |
+
"step": 5890
|
| 4598 |
+
},
|
| 4599 |
+
{
|
| 4600 |
+
"epoch": 0.5363636363636364,
|
| 4601 |
+
"grad_norm": 379239069122560.0,
|
| 4602 |
+
"learning_rate": 4.431461662373415e-07,
|
| 4603 |
+
"loss": 26.517977905273437,
|
| 4604 |
+
"step": 5900
|
| 4605 |
+
},
|
| 4606 |
+
{
|
| 4607 |
+
"epoch": 0.5363636363636364,
|
| 4608 |
+
"eval_loss": 26.296443939208984,
|
| 4609 |
+
"eval_runtime": 2.1745,
|
| 4610 |
+
"eval_samples_per_second": 16.555,
|
| 4611 |
+
"eval_steps_per_second": 0.92,
|
| 4612 |
+
"step": 5900
|
| 4613 |
+
},
|
| 4614 |
+
{
|
| 4615 |
+
"epoch": 0.5372727272727272,
|
| 4616 |
+
"grad_norm": 146172048048128.0,
|
| 4617 |
+
"learning_rate": 4.4172766500327227e-07,
|
| 4618 |
+
"loss": 26.98131103515625,
|
| 4619 |
+
"step": 5910
|
| 4620 |
+
},
|
| 4621 |
+
{
|
| 4622 |
+
"epoch": 0.5381818181818182,
|
| 4623 |
+
"grad_norm": 280093792403456.0,
|
| 4624 |
+
"learning_rate": 4.4030963907871e-07,
|
| 4625 |
+
"loss": 26.6470703125,
|
| 4626 |
+
"step": 5920
|
| 4627 |
+
},
|
| 4628 |
+
{
|
| 4629 |
+
"epoch": 0.5390909090909091,
|
| 4630 |
+
"grad_norm": 200643255992320.0,
|
| 4631 |
+
"learning_rate": 4.3889210003005525e-07,
|
| 4632 |
+
"loss": 26.51512451171875,
|
| 4633 |
+
"step": 5930
|
| 4634 |
+
},
|
| 4635 |
+
{
|
| 4636 |
+
"epoch": 0.54,
|
| 4637 |
+
"grad_norm": 298077273456640.0,
|
| 4638 |
+
"learning_rate": 4.374750594197383e-07,
|
| 4639 |
+
"loss": 26.81731262207031,
|
| 4640 |
+
"step": 5940
|
| 4641 |
+
},
|
| 4642 |
+
{
|
| 4643 |
+
"epoch": 0.5409090909090909,
|
| 4644 |
+
"grad_norm": 105089628897280.0,
|
| 4645 |
+
"learning_rate": 4.36058528806123e-07,
|
| 4646 |
+
"loss": 26.3632568359375,
|
| 4647 |
+
"step": 5950
|
| 4648 |
+
},
|
| 4649 |
+
{
|
| 4650 |
+
"epoch": 0.5418181818181819,
|
| 4651 |
+
"grad_norm": 243687518896128.0,
|
| 4652 |
+
"learning_rate": 4.346425197434142e-07,
|
| 4653 |
+
"loss": 26.923687744140626,
|
| 4654 |
+
"step": 5960
|
| 4655 |
+
},
|
| 4656 |
+
{
|
| 4657 |
+
"epoch": 0.5427272727272727,
|
| 4658 |
+
"grad_norm": 667943180435456.0,
|
| 4659 |
+
"learning_rate": 4.3322704378156184e-07,
|
| 4660 |
+
"loss": 26.333908081054688,
|
| 4661 |
+
"step": 5970
|
| 4662 |
+
},
|
| 4663 |
+
{
|
| 4664 |
+
"epoch": 0.5436363636363636,
|
| 4665 |
+
"grad_norm": 232555886411776.0,
|
| 4666 |
+
"learning_rate": 4.3181211246616774e-07,
|
| 4667 |
+
"loss": 26.753402709960938,
|
| 4668 |
+
"step": 5980
|
| 4669 |
+
},
|
| 4670 |
+
{
|
| 4671 |
+
"epoch": 0.5445454545454546,
|
| 4672 |
+
"grad_norm": 504201209708544.0,
|
| 4673 |
+
"learning_rate": 4.3039773733839145e-07,
|
| 4674 |
+
"loss": 26.233575439453126,
|
| 4675 |
+
"step": 5990
|
| 4676 |
+
},
|
| 4677 |
+
{
|
| 4678 |
+
"epoch": 0.5454545454545454,
|
| 4679 |
+
"grad_norm": 184770180939776.0,
|
| 4680 |
+
"learning_rate": 4.2898392993485547e-07,
|
| 4681 |
+
"loss": 26.377456665039062,
|
| 4682 |
+
"step": 6000
|
| 4683 |
+
},
|
| 4684 |
+
{
|
| 4685 |
+
"epoch": 0.5454545454545454,
|
| 4686 |
+
"eval_loss": 26.29542350769043,
|
| 4687 |
+
"eval_runtime": 2.1707,
|
| 4688 |
+
"eval_samples_per_second": 16.585,
|
| 4689 |
+
"eval_steps_per_second": 0.921,
|
| 4690 |
+
"step": 6000
|
| 4691 |
}
|
| 4692 |
],
|
| 4693 |
"logging_steps": 10,
|
|
|
|
| 4707 |
"attributes": {}
|
| 4708 |
}
|
| 4709 |
},
|
| 4710 |
+
"total_flos": 5.4359085809664e+16,
|
| 4711 |
"train_batch_size": 4,
|
| 4712 |
"trial_name": null,
|
| 4713 |
"trial_params": null
|