Training in progress, step 7000, 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:a000fbccdb35daf2266aee700b6416df5d706b2e1ab338bdaf2865bd45ae9cba
|
| 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 sha256:8eb6acce7ccf2cb8305165f2fd9c13795130a0c57087b32b61bb62aecab4b395
|
| 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:d5920ac4f0bba99259462ed725e7e621beec906b352c1c25b8030b235ade5e09
|
| 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:b4804038d40e072898353507105f7ec5bb84f06c500c6c77f9b23c1ffba69da2
|
| 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 sha256:13efeffdbbe1ff32e561220c2dda4e0981a0eea5d6d98be9a53ce5e4516700b4
|
| 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 sha256:2378f6335467d6824d5991555117a905cfacd66d8e20475c21b6217632b4ac67
|
| 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 sha256:55025d46de2da78115d789b4623359037ec52573b190c40bc6dd0fc9edee6cb2
|
| 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:b493d4f290fe42ef5f1784d5efbf107e6b51d92117c6d516dbead5fb235bd991
|
| 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 sha256:55a739c31e021e2247343b9ee6a1dd0907ecba3b8b539617e181504bf40d05fd
|
| 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 sha256:2a27be9081fd95a72cd55c9d8dae191778ab5451c500377cd2773cecbb45418f
|
| 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:fb3660dc7c29e236f64dc7b388f5a9bcf2201e9e232665f64d728ff2de7b4a50
|
| 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,
|
|
@@ -4688,6 +4688,786 @@
|
|
| 4688 |
"eval_samples_per_second": 16.585,
|
| 4689 |
"eval_steps_per_second": 0.921,
|
| 4690 |
"step": 6000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4691 |
}
|
| 4692 |
],
|
| 4693 |
"logging_steps": 10,
|
|
@@ -4707,7 +5487,7 @@
|
|
| 4707 |
"attributes": {}
|
| 4708 |
}
|
| 4709 |
},
|
| 4710 |
-
"total_flos":
|
| 4711 |
"train_batch_size": 4,
|
| 4712 |
"trial_name": null,
|
| 4713 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.6363636363636364,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 7000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": false,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 4688 |
"eval_samples_per_second": 16.585,
|
| 4689 |
"eval_steps_per_second": 0.921,
|
| 4690 |
"step": 6000
|
| 4691 |
+
},
|
| 4692 |
+
{
|
| 4693 |
+
"epoch": 0.5463636363636364,
|
| 4694 |
+
"grad_norm": 306690058616832.0,
|
| 4695 |
+
"learning_rate": 4.2757070178755213e-07,
|
| 4696 |
+
"loss": 26.351678466796876,
|
| 4697 |
+
"step": 6010
|
| 4698 |
+
},
|
| 4699 |
+
{
|
| 4700 |
+
"epoch": 0.5472727272727272,
|
| 4701 |
+
"grad_norm": 103466861068288.0,
|
| 4702 |
+
"learning_rate": 4.261580644237481e-07,
|
| 4703 |
+
"loss": 26.860089111328126,
|
| 4704 |
+
"step": 6020
|
| 4705 |
+
},
|
| 4706 |
+
{
|
| 4707 |
+
"epoch": 0.5481818181818182,
|
| 4708 |
+
"grad_norm": 358580880408576.0,
|
| 4709 |
+
"learning_rate": 4.2474602936589233e-07,
|
| 4710 |
+
"loss": 26.571078491210937,
|
| 4711 |
+
"step": 6030
|
| 4712 |
+
},
|
| 4713 |
+
{
|
| 4714 |
+
"epoch": 0.5490909090909091,
|
| 4715 |
+
"grad_norm": 270877916659712.0,
|
| 4716 |
+
"learning_rate": 4.233346081315196e-07,
|
| 4717 |
+
"loss": 26.6639404296875,
|
| 4718 |
+
"step": 6040
|
| 4719 |
+
},
|
| 4720 |
+
{
|
| 4721 |
+
"epoch": 0.55,
|
| 4722 |
+
"grad_norm": 409891646734336.0,
|
| 4723 |
+
"learning_rate": 4.219238122331593e-07,
|
| 4724 |
+
"loss": 26.780926513671876,
|
| 4725 |
+
"step": 6050
|
| 4726 |
+
},
|
| 4727 |
+
{
|
| 4728 |
+
"epoch": 0.5509090909090909,
|
| 4729 |
+
"grad_norm": 182963593543680.0,
|
| 4730 |
+
"learning_rate": 4.2051365317823893e-07,
|
| 4731 |
+
"loss": 26.43211975097656,
|
| 4732 |
+
"step": 6060
|
| 4733 |
+
},
|
| 4734 |
+
{
|
| 4735 |
+
"epoch": 0.5518181818181818,
|
| 4736 |
+
"grad_norm": 94839890771968.0,
|
| 4737 |
+
"learning_rate": 4.191041424689925e-07,
|
| 4738 |
+
"loss": 27.00631103515625,
|
| 4739 |
+
"step": 6070
|
| 4740 |
+
},
|
| 4741 |
+
{
|
| 4742 |
+
"epoch": 0.5527272727272727,
|
| 4743 |
+
"grad_norm": 415413498281984.0,
|
| 4744 |
+
"learning_rate": 4.176952916023649e-07,
|
| 4745 |
+
"loss": 26.388510131835936,
|
| 4746 |
+
"step": 6080
|
| 4747 |
+
},
|
| 4748 |
+
{
|
| 4749 |
+
"epoch": 0.5536363636363636,
|
| 4750 |
+
"grad_norm": 247516348022784.0,
|
| 4751 |
+
"learning_rate": 4.1628711206991906e-07,
|
| 4752 |
+
"loss": 26.635281372070313,
|
| 4753 |
+
"step": 6090
|
| 4754 |
+
},
|
| 4755 |
+
{
|
| 4756 |
+
"epoch": 0.5545454545454546,
|
| 4757 |
+
"grad_norm": 941664063455232.0,
|
| 4758 |
+
"learning_rate": 4.1487961535774235e-07,
|
| 4759 |
+
"loss": 26.43724365234375,
|
| 4760 |
+
"step": 6100
|
| 4761 |
+
},
|
| 4762 |
+
{
|
| 4763 |
+
"epoch": 0.5545454545454546,
|
| 4764 |
+
"eval_loss": 26.28836441040039,
|
| 4765 |
+
"eval_runtime": 2.1519,
|
| 4766 |
+
"eval_samples_per_second": 16.73,
|
| 4767 |
+
"eval_steps_per_second": 0.929,
|
| 4768 |
+
"step": 6100
|
| 4769 |
+
},
|
| 4770 |
+
{
|
| 4771 |
+
"epoch": 0.5554545454545454,
|
| 4772 |
+
"grad_norm": 295934218993664.0,
|
| 4773 |
+
"learning_rate": 4.134728129463522e-07,
|
| 4774 |
+
"loss": 26.27704772949219,
|
| 4775 |
+
"step": 6110
|
| 4776 |
+
},
|
| 4777 |
+
{
|
| 4778 |
+
"epoch": 0.5563636363636364,
|
| 4779 |
+
"grad_norm": 269193249292288.0,
|
| 4780 |
+
"learning_rate": 4.1206671631060323e-07,
|
| 4781 |
+
"loss": 26.7314697265625,
|
| 4782 |
+
"step": 6120
|
| 4783 |
+
},
|
| 4784 |
+
{
|
| 4785 |
+
"epoch": 0.5572727272727273,
|
| 4786 |
+
"grad_norm": 125726451826688.0,
|
| 4787 |
+
"learning_rate": 4.1066133691959284e-07,
|
| 4788 |
+
"loss": 26.241030883789062,
|
| 4789 |
+
"step": 6130
|
| 4790 |
+
},
|
| 4791 |
+
{
|
| 4792 |
+
"epoch": 0.5581818181818182,
|
| 4793 |
+
"grad_norm": 672323979968512.0,
|
| 4794 |
+
"learning_rate": 4.0925668623656846e-07,
|
| 4795 |
+
"loss": 26.6068115234375,
|
| 4796 |
+
"step": 6140
|
| 4797 |
+
},
|
| 4798 |
+
{
|
| 4799 |
+
"epoch": 0.5590909090909091,
|
| 4800 |
+
"grad_norm": 247083109974016.0,
|
| 4801 |
+
"learning_rate": 4.0785277571883324e-07,
|
| 4802 |
+
"loss": 26.240081787109375,
|
| 4803 |
+
"step": 6150
|
| 4804 |
+
},
|
| 4805 |
+
{
|
| 4806 |
+
"epoch": 0.56,
|
| 4807 |
+
"grad_norm": 347755381784576.0,
|
| 4808 |
+
"learning_rate": 4.0644961681765385e-07,
|
| 4809 |
+
"loss": 26.687600708007814,
|
| 4810 |
+
"step": 6160
|
| 4811 |
+
},
|
| 4812 |
+
{
|
| 4813 |
+
"epoch": 0.5609090909090909,
|
| 4814 |
+
"grad_norm": 193356390989824.0,
|
| 4815 |
+
"learning_rate": 4.0504722097816526e-07,
|
| 4816 |
+
"loss": 26.66230163574219,
|
| 4817 |
+
"step": 6170
|
| 4818 |
+
},
|
| 4819 |
+
{
|
| 4820 |
+
"epoch": 0.5618181818181818,
|
| 4821 |
+
"grad_norm": 264473046679552.0,
|
| 4822 |
+
"learning_rate": 4.0364559963927865e-07,
|
| 4823 |
+
"loss": 26.944580078125,
|
| 4824 |
+
"step": 6180
|
| 4825 |
+
},
|
| 4826 |
+
{
|
| 4827 |
+
"epoch": 0.5627272727272727,
|
| 4828 |
+
"grad_norm": 176868397416448.0,
|
| 4829 |
+
"learning_rate": 4.022447642335885e-07,
|
| 4830 |
+
"loss": 26.1925048828125,
|
| 4831 |
+
"step": 6190
|
| 4832 |
+
},
|
| 4833 |
+
{
|
| 4834 |
+
"epoch": 0.5636363636363636,
|
| 4835 |
+
"grad_norm": 357407649693696.0,
|
| 4836 |
+
"learning_rate": 4.0084472618727795e-07,
|
| 4837 |
+
"loss": 26.477133178710936,
|
| 4838 |
+
"step": 6200
|
| 4839 |
+
},
|
| 4840 |
+
{
|
| 4841 |
+
"epoch": 0.5636363636363636,
|
| 4842 |
+
"eval_loss": 26.29114532470703,
|
| 4843 |
+
"eval_runtime": 2.1728,
|
| 4844 |
+
"eval_samples_per_second": 16.569,
|
| 4845 |
+
"eval_steps_per_second": 0.92,
|
| 4846 |
+
"step": 6200
|
| 4847 |
+
},
|
| 4848 |
+
{
|
| 4849 |
+
"epoch": 0.5645454545454546,
|
| 4850 |
+
"grad_norm": 188899942989824.0,
|
| 4851 |
+
"learning_rate": 3.994454969200266e-07,
|
| 4852 |
+
"loss": 26.23280029296875,
|
| 4853 |
+
"step": 6210
|
| 4854 |
+
},
|
| 4855 |
+
{
|
| 4856 |
+
"epoch": 0.5654545454545454,
|
| 4857 |
+
"grad_norm": 174719772917760.0,
|
| 4858 |
+
"learning_rate": 3.9804708784491697e-07,
|
| 4859 |
+
"loss": 26.93548583984375,
|
| 4860 |
+
"step": 6220
|
| 4861 |
+
},
|
| 4862 |
+
{
|
| 4863 |
+
"epoch": 0.5663636363636364,
|
| 4864 |
+
"grad_norm": 168751630647296.0,
|
| 4865 |
+
"learning_rate": 3.966495103683418e-07,
|
| 4866 |
+
"loss": 26.4197509765625,
|
| 4867 |
+
"step": 6230
|
| 4868 |
+
},
|
| 4869 |
+
{
|
| 4870 |
+
"epoch": 0.5672727272727273,
|
| 4871 |
+
"grad_norm": 353826150285312.0,
|
| 4872 |
+
"learning_rate": 3.952527758899103e-07,
|
| 4873 |
+
"loss": 26.3079345703125,
|
| 4874 |
+
"step": 6240
|
| 4875 |
+
},
|
| 4876 |
+
{
|
| 4877 |
+
"epoch": 0.5681818181818182,
|
| 4878 |
+
"grad_norm": 91272224178176.0,
|
| 4879 |
+
"learning_rate": 3.938568958023561e-07,
|
| 4880 |
+
"loss": 26.32183837890625,
|
| 4881 |
+
"step": 6250
|
| 4882 |
+
},
|
| 4883 |
+
{
|
| 4884 |
+
"epoch": 0.5690909090909091,
|
| 4885 |
+
"grad_norm": 188810235215872.0,
|
| 4886 |
+
"learning_rate": 3.9246188149144344e-07,
|
| 4887 |
+
"loss": 26.427536010742188,
|
| 4888 |
+
"step": 6260
|
| 4889 |
+
},
|
| 4890 |
+
{
|
| 4891 |
+
"epoch": 0.57,
|
| 4892 |
+
"grad_norm": 413852076343296.0,
|
| 4893 |
+
"learning_rate": 3.9106774433587465e-07,
|
| 4894 |
+
"loss": 26.245895385742188,
|
| 4895 |
+
"step": 6270
|
| 4896 |
+
},
|
| 4897 |
+
{
|
| 4898 |
+
"epoch": 0.5709090909090909,
|
| 4899 |
+
"grad_norm": 75952646258688.0,
|
| 4900 |
+
"learning_rate": 3.8967449570719783e-07,
|
| 4901 |
+
"loss": 26.672445678710936,
|
| 4902 |
+
"step": 6280
|
| 4903 |
+
},
|
| 4904 |
+
{
|
| 4905 |
+
"epoch": 0.5718181818181818,
|
| 4906 |
+
"grad_norm": 149835806146560.0,
|
| 4907 |
+
"learning_rate": 3.8828214696971305e-07,
|
| 4908 |
+
"loss": 26.373684692382813,
|
| 4909 |
+
"step": 6290
|
| 4910 |
+
},
|
| 4911 |
+
{
|
| 4912 |
+
"epoch": 0.5727272727272728,
|
| 4913 |
+
"grad_norm": 84309335605248.0,
|
| 4914 |
+
"learning_rate": 3.8689070948038105e-07,
|
| 4915 |
+
"loss": 26.33724365234375,
|
| 4916 |
+
"step": 6300
|
| 4917 |
+
},
|
| 4918 |
+
{
|
| 4919 |
+
"epoch": 0.5727272727272728,
|
| 4920 |
+
"eval_loss": 26.28530502319336,
|
| 4921 |
+
"eval_runtime": 2.1736,
|
| 4922 |
+
"eval_samples_per_second": 16.562,
|
| 4923 |
+
"eval_steps_per_second": 0.92,
|
| 4924 |
+
"step": 6300
|
| 4925 |
+
},
|
| 4926 |
+
{
|
| 4927 |
+
"epoch": 0.5736363636363636,
|
| 4928 |
+
"grad_norm": 184767328813056.0,
|
| 4929 |
+
"learning_rate": 3.8550019458872863e-07,
|
| 4930 |
+
"loss": 26.87149658203125,
|
| 4931 |
+
"step": 6310
|
| 4932 |
+
},
|
| 4933 |
+
{
|
| 4934 |
+
"epoch": 0.5745454545454546,
|
| 4935 |
+
"grad_norm": 166437347917824.0,
|
| 4936 |
+
"learning_rate": 3.841106136367585e-07,
|
| 4937 |
+
"loss": 26.697918701171876,
|
| 4938 |
+
"step": 6320
|
| 4939 |
+
},
|
| 4940 |
+
{
|
| 4941 |
+
"epoch": 0.5754545454545454,
|
| 4942 |
+
"grad_norm": 375701593128960.0,
|
| 4943 |
+
"learning_rate": 3.827219779588546e-07,
|
| 4944 |
+
"loss": 26.319839477539062,
|
| 4945 |
+
"step": 6330
|
| 4946 |
+
},
|
| 4947 |
+
{
|
| 4948 |
+
"epoch": 0.5763636363636364,
|
| 4949 |
+
"grad_norm": 123003073462272.0,
|
| 4950 |
+
"learning_rate": 3.81334298881691e-07,
|
| 4951 |
+
"loss": 26.770510864257812,
|
| 4952 |
+
"step": 6340
|
| 4953 |
+
},
|
| 4954 |
+
{
|
| 4955 |
+
"epoch": 0.5772727272727273,
|
| 4956 |
+
"grad_norm": 429775868919808.0,
|
| 4957 |
+
"learning_rate": 3.7994758772413894e-07,
|
| 4958 |
+
"loss": 26.409426879882812,
|
| 4959 |
+
"step": 6350
|
| 4960 |
+
},
|
| 4961 |
+
{
|
| 4962 |
+
"epoch": 0.5781818181818181,
|
| 4963 |
+
"grad_norm": 180347572584448.0,
|
| 4964 |
+
"learning_rate": 3.7856185579717455e-07,
|
| 4965 |
+
"loss": 26.36715087890625,
|
| 4966 |
+
"step": 6360
|
| 4967 |
+
},
|
| 4968 |
+
{
|
| 4969 |
+
"epoch": 0.5790909090909091,
|
| 4970 |
+
"grad_norm": 1096010222796800.0,
|
| 4971 |
+
"learning_rate": 3.771771144037869e-07,
|
| 4972 |
+
"loss": 26.516082763671875,
|
| 4973 |
+
"step": 6370
|
| 4974 |
+
},
|
| 4975 |
+
{
|
| 4976 |
+
"epoch": 0.58,
|
| 4977 |
+
"grad_norm": 108321600176128.0,
|
| 4978 |
+
"learning_rate": 3.7579337483888535e-07,
|
| 4979 |
+
"loss": 26.694012451171876,
|
| 4980 |
+
"step": 6380
|
| 4981 |
+
},
|
| 4982 |
+
{
|
| 4983 |
+
"epoch": 0.5809090909090909,
|
| 4984 |
+
"grad_norm": 196052774813696.0,
|
| 4985 |
+
"learning_rate": 3.7441064838920793e-07,
|
| 4986 |
+
"loss": 26.63226318359375,
|
| 4987 |
+
"step": 6390
|
| 4988 |
+
},
|
| 4989 |
+
{
|
| 4990 |
+
"epoch": 0.5818181818181818,
|
| 4991 |
+
"grad_norm": 136146621300736.0,
|
| 4992 |
+
"learning_rate": 3.7302894633322857e-07,
|
| 4993 |
+
"loss": 26.535342407226562,
|
| 4994 |
+
"step": 6400
|
| 4995 |
+
},
|
| 4996 |
+
{
|
| 4997 |
+
"epoch": 0.5818181818181818,
|
| 4998 |
+
"eval_loss": 26.28375244140625,
|
| 4999 |
+
"eval_runtime": 2.1824,
|
| 5000 |
+
"eval_samples_per_second": 16.495,
|
| 5001 |
+
"eval_steps_per_second": 0.916,
|
| 5002 |
+
"step": 6400
|
| 5003 |
+
},
|
| 5004 |
+
{
|
| 5005 |
+
"epoch": 0.5827272727272728,
|
| 5006 |
+
"grad_norm": 359597948796928.0,
|
| 5007 |
+
"learning_rate": 3.7164827994106623e-07,
|
| 5008 |
+
"loss": 26.610174560546874,
|
| 5009 |
+
"step": 6410
|
| 5010 |
+
},
|
| 5011 |
+
{
|
| 5012 |
+
"epoch": 0.5836363636363636,
|
| 5013 |
+
"grad_norm": 358815996313600.0,
|
| 5014 |
+
"learning_rate": 3.7026866047439134e-07,
|
| 5015 |
+
"loss": 26.601348876953125,
|
| 5016 |
+
"step": 6420
|
| 5017 |
+
},
|
| 5018 |
+
{
|
| 5019 |
+
"epoch": 0.5845454545454546,
|
| 5020 |
+
"grad_norm": 232237672955904.0,
|
| 5021 |
+
"learning_rate": 3.688900991863357e-07,
|
| 5022 |
+
"loss": 26.58094177246094,
|
| 5023 |
+
"step": 6430
|
| 5024 |
+
},
|
| 5025 |
+
{
|
| 5026 |
+
"epoch": 0.5854545454545454,
|
| 5027 |
+
"grad_norm": 224937939828736.0,
|
| 5028 |
+
"learning_rate": 3.675126073213998e-07,
|
| 5029 |
+
"loss": 26.3585693359375,
|
| 5030 |
+
"step": 6440
|
| 5031 |
+
},
|
| 5032 |
+
{
|
| 5033 |
+
"epoch": 0.5863636363636363,
|
| 5034 |
+
"grad_norm": 170370447441920.0,
|
| 5035 |
+
"learning_rate": 3.661361961153602e-07,
|
| 5036 |
+
"loss": 26.818353271484376,
|
| 5037 |
+
"step": 6450
|
| 5038 |
+
},
|
| 5039 |
+
{
|
| 5040 |
+
"epoch": 0.5872727272727273,
|
| 5041 |
+
"grad_norm": 241997600260096.0,
|
| 5042 |
+
"learning_rate": 3.647608767951803e-07,
|
| 5043 |
+
"loss": 26.691299438476562,
|
| 5044 |
+
"step": 6460
|
| 5045 |
+
},
|
| 5046 |
+
{
|
| 5047 |
+
"epoch": 0.5881818181818181,
|
| 5048 |
+
"grad_norm": 359189289369600.0,
|
| 5049 |
+
"learning_rate": 3.633866605789163e-07,
|
| 5050 |
+
"loss": 26.187884521484374,
|
| 5051 |
+
"step": 6470
|
| 5052 |
+
},
|
| 5053 |
+
{
|
| 5054 |
+
"epoch": 0.5890909090909091,
|
| 5055 |
+
"grad_norm": 318397501931520.0,
|
| 5056 |
+
"learning_rate": 3.6201355867562724e-07,
|
| 5057 |
+
"loss": 26.471282958984375,
|
| 5058 |
+
"step": 6480
|
| 5059 |
+
},
|
| 5060 |
+
{
|
| 5061 |
+
"epoch": 0.59,
|
| 5062 |
+
"grad_norm": 151627210162176.0,
|
| 5063 |
+
"learning_rate": 3.606415822852825e-07,
|
| 5064 |
+
"loss": 26.56259765625,
|
| 5065 |
+
"step": 6490
|
| 5066 |
+
},
|
| 5067 |
+
{
|
| 5068 |
+
"epoch": 0.5909090909090909,
|
| 5069 |
+
"grad_norm": 304659646382080.0,
|
| 5070 |
+
"learning_rate": 3.592707425986717e-07,
|
| 5071 |
+
"loss": 26.369317626953126,
|
| 5072 |
+
"step": 6500
|
| 5073 |
+
},
|
| 5074 |
+
{
|
| 5075 |
+
"epoch": 0.5909090909090909,
|
| 5076 |
+
"eval_loss": 26.286823272705078,
|
| 5077 |
+
"eval_runtime": 2.1707,
|
| 5078 |
+
"eval_samples_per_second": 16.584,
|
| 5079 |
+
"eval_steps_per_second": 0.921,
|
| 5080 |
+
"step": 6500
|
| 5081 |
+
},
|
| 5082 |
+
{
|
| 5083 |
+
"epoch": 0.5918181818181818,
|
| 5084 |
+
"grad_norm": 280783252094976.0,
|
| 5085 |
+
"learning_rate": 3.5790105079731216e-07,
|
| 5086 |
+
"loss": 26.206988525390624,
|
| 5087 |
+
"step": 6510
|
| 5088 |
+
},
|
| 5089 |
+
{
|
| 5090 |
+
"epoch": 0.5927272727272728,
|
| 5091 |
+
"grad_norm": 196329196224512.0,
|
| 5092 |
+
"learning_rate": 3.565325180533586e-07,
|
| 5093 |
+
"loss": 26.52789306640625,
|
| 5094 |
+
"step": 6520
|
| 5095 |
+
},
|
| 5096 |
+
{
|
| 5097 |
+
"epoch": 0.5936363636363636,
|
| 5098 |
+
"grad_norm": 136604035317760.0,
|
| 5099 |
+
"learning_rate": 3.5516515552951124e-07,
|
| 5100 |
+
"loss": 26.288201904296876,
|
| 5101 |
+
"step": 6530
|
| 5102 |
+
},
|
| 5103 |
+
{
|
| 5104 |
+
"epoch": 0.5945454545454546,
|
| 5105 |
+
"grad_norm": 92855758487552.0,
|
| 5106 |
+
"learning_rate": 3.537989743789259e-07,
|
| 5107 |
+
"loss": 26.704229736328124,
|
| 5108 |
+
"step": 6540
|
| 5109 |
+
},
|
| 5110 |
+
{
|
| 5111 |
+
"epoch": 0.5954545454545455,
|
| 5112 |
+
"grad_norm": 160173054230528.0,
|
| 5113 |
+
"learning_rate": 3.5243398574512185e-07,
|
| 5114 |
+
"loss": 26.398892211914063,
|
| 5115 |
+
"step": 6550
|
| 5116 |
+
},
|
| 5117 |
+
{
|
| 5118 |
+
"epoch": 0.5963636363636363,
|
| 5119 |
+
"grad_norm": 205201441030144.0,
|
| 5120 |
+
"learning_rate": 3.5107020076189075e-07,
|
| 5121 |
+
"loss": 26.12825927734375,
|
| 5122 |
+
"step": 6560
|
| 5123 |
+
},
|
| 5124 |
+
{
|
| 5125 |
+
"epoch": 0.5972727272727273,
|
| 5126 |
+
"grad_norm": 381060437245952.0,
|
| 5127 |
+
"learning_rate": 3.4970763055320776e-07,
|
| 5128 |
+
"loss": 26.28996887207031,
|
| 5129 |
+
"step": 6570
|
| 5130 |
+
},
|
| 5131 |
+
{
|
| 5132 |
+
"epoch": 0.5981818181818181,
|
| 5133 |
+
"grad_norm": 101203681738752.0,
|
| 5134 |
+
"learning_rate": 3.4834628623313835e-07,
|
| 5135 |
+
"loss": 26.50263671875,
|
| 5136 |
+
"step": 6580
|
| 5137 |
+
},
|
| 5138 |
+
{
|
| 5139 |
+
"epoch": 0.5990909090909091,
|
| 5140 |
+
"grad_norm": 190785299415040.0,
|
| 5141 |
+
"learning_rate": 3.4698617890574966e-07,
|
| 5142 |
+
"loss": 26.34183349609375,
|
| 5143 |
+
"step": 6590
|
| 5144 |
+
},
|
| 5145 |
+
{
|
| 5146 |
+
"epoch": 0.6,
|
| 5147 |
+
"grad_norm": 127154729779200.0,
|
| 5148 |
+
"learning_rate": 3.4562731966501835e-07,
|
| 5149 |
+
"loss": 26.541549682617188,
|
| 5150 |
+
"step": 6600
|
| 5151 |
+
},
|
| 5152 |
+
{
|
| 5153 |
+
"epoch": 0.6,
|
| 5154 |
+
"eval_loss": 26.287540435791016,
|
| 5155 |
+
"eval_runtime": 2.151,
|
| 5156 |
+
"eval_samples_per_second": 16.737,
|
| 5157 |
+
"eval_steps_per_second": 0.93,
|
| 5158 |
+
"step": 6600
|
| 5159 |
+
},
|
| 5160 |
+
{
|
| 5161 |
+
"epoch": 0.600909090909091,
|
| 5162 |
+
"grad_norm": 294955067113472.0,
|
| 5163 |
+
"learning_rate": 3.442697195947414e-07,
|
| 5164 |
+
"loss": 26.36138000488281,
|
| 5165 |
+
"step": 6610
|
| 5166 |
+
},
|
| 5167 |
+
{
|
| 5168 |
+
"epoch": 0.6018181818181818,
|
| 5169 |
+
"grad_norm": 587266313420800.0,
|
| 5170 |
+
"learning_rate": 3.4291338976844465e-07,
|
| 5171 |
+
"loss": 26.64934387207031,
|
| 5172 |
+
"step": 6620
|
| 5173 |
+
},
|
| 5174 |
+
{
|
| 5175 |
+
"epoch": 0.6027272727272728,
|
| 5176 |
+
"grad_norm": 137201656528896.0,
|
| 5177 |
+
"learning_rate": 3.415583412492933e-07,
|
| 5178 |
+
"loss": 26.626654052734374,
|
| 5179 |
+
"step": 6630
|
| 5180 |
+
},
|
| 5181 |
+
{
|
| 5182 |
+
"epoch": 0.6036363636363636,
|
| 5183 |
+
"grad_norm": 370320368205824.0,
|
| 5184 |
+
"learning_rate": 3.4020458509000117e-07,
|
| 5185 |
+
"loss": 26.581622314453124,
|
| 5186 |
+
"step": 6640
|
| 5187 |
+
},
|
| 5188 |
+
{
|
| 5189 |
+
"epoch": 0.6045454545454545,
|
| 5190 |
+
"grad_norm": 129610041786368.0,
|
| 5191 |
+
"learning_rate": 3.3885213233274053e-07,
|
| 5192 |
+
"loss": 26.582095336914062,
|
| 5193 |
+
"step": 6650
|
| 5194 |
+
},
|
| 5195 |
+
{
|
| 5196 |
+
"epoch": 0.6054545454545455,
|
| 5197 |
+
"grad_norm": 590885427347456.0,
|
| 5198 |
+
"learning_rate": 3.3750099400905256e-07,
|
| 5199 |
+
"loss": 26.447711181640624,
|
| 5200 |
+
"step": 6660
|
| 5201 |
+
},
|
| 5202 |
+
{
|
| 5203 |
+
"epoch": 0.6063636363636363,
|
| 5204 |
+
"grad_norm": 323296281231360.0,
|
| 5205 |
+
"learning_rate": 3.361511811397567e-07,
|
| 5206 |
+
"loss": 26.482095336914064,
|
| 5207 |
+
"step": 6670
|
| 5208 |
+
},
|
| 5209 |
+
{
|
| 5210 |
+
"epoch": 0.6072727272727273,
|
| 5211 |
+
"grad_norm": 210331276148736.0,
|
| 5212 |
+
"learning_rate": 3.348027047348615e-07,
|
| 5213 |
+
"loss": 26.46201171875,
|
| 5214 |
+
"step": 6680
|
| 5215 |
+
},
|
| 5216 |
+
{
|
| 5217 |
+
"epoch": 0.6081818181818182,
|
| 5218 |
+
"grad_norm": 769363464421376.0,
|
| 5219 |
+
"learning_rate": 3.3345557579347353e-07,
|
| 5220 |
+
"loss": 26.45103759765625,
|
| 5221 |
+
"step": 6690
|
| 5222 |
+
},
|
| 5223 |
+
{
|
| 5224 |
+
"epoch": 0.6090909090909091,
|
| 5225 |
+
"grad_norm": 661023115706368.0,
|
| 5226 |
+
"learning_rate": 3.321098053037097e-07,
|
| 5227 |
+
"loss": 26.44871826171875,
|
| 5228 |
+
"step": 6700
|
| 5229 |
+
},
|
| 5230 |
+
{
|
| 5231 |
+
"epoch": 0.6090909090909091,
|
| 5232 |
+
"eval_loss": 26.283708572387695,
|
| 5233 |
+
"eval_runtime": 2.1672,
|
| 5234 |
+
"eval_samples_per_second": 16.611,
|
| 5235 |
+
"eval_steps_per_second": 0.923,
|
| 5236 |
+
"step": 6700
|
| 5237 |
+
},
|
| 5238 |
+
{
|
| 5239 |
+
"epoch": 0.61,
|
| 5240 |
+
"grad_norm": 199447761911808.0,
|
| 5241 |
+
"learning_rate": 3.3076540424260537e-07,
|
| 5242 |
+
"loss": 26.60747375488281,
|
| 5243 |
+
"step": 6710
|
| 5244 |
+
},
|
| 5245 |
+
{
|
| 5246 |
+
"epoch": 0.610909090909091,
|
| 5247 |
+
"grad_norm": 391525628379136.0,
|
| 5248 |
+
"learning_rate": 3.2942238357602657e-07,
|
| 5249 |
+
"loss": 26.542547607421874,
|
| 5250 |
+
"step": 6720
|
| 5251 |
+
},
|
| 5252 |
+
{
|
| 5253 |
+
"epoch": 0.6118181818181818,
|
| 5254 |
+
"grad_norm": 64306838241280.0,
|
| 5255 |
+
"learning_rate": 3.280807542585795e-07,
|
| 5256 |
+
"loss": 26.613406372070312,
|
| 5257 |
+
"step": 6730
|
| 5258 |
+
},
|
| 5259 |
+
{
|
| 5260 |
+
"epoch": 0.6127272727272727,
|
| 5261 |
+
"grad_norm": 70327136681984.0,
|
| 5262 |
+
"learning_rate": 3.2674052723352167e-07,
|
| 5263 |
+
"loss": 26.60184326171875,
|
| 5264 |
+
"step": 6740
|
| 5265 |
+
},
|
| 5266 |
+
{
|
| 5267 |
+
"epoch": 0.6136363636363636,
|
| 5268 |
+
"grad_norm": 62553828884480.0,
|
| 5269 |
+
"learning_rate": 3.2540171343267275e-07,
|
| 5270 |
+
"loss": 26.298440551757814,
|
| 5271 |
+
"step": 6750
|
| 5272 |
+
},
|
| 5273 |
+
{
|
| 5274 |
+
"epoch": 0.6145454545454545,
|
| 5275 |
+
"grad_norm": 251477431943168.0,
|
| 5276 |
+
"learning_rate": 3.240643237763247e-07,
|
| 5277 |
+
"loss": 26.87586669921875,
|
| 5278 |
+
"step": 6760
|
| 5279 |
+
},
|
| 5280 |
+
{
|
| 5281 |
+
"epoch": 0.6154545454545455,
|
| 5282 |
+
"grad_norm": 283658397155328.0,
|
| 5283 |
+
"learning_rate": 3.227283691731537e-07,
|
| 5284 |
+
"loss": 26.486978149414064,
|
| 5285 |
+
"step": 6770
|
| 5286 |
+
},
|
| 5287 |
+
{
|
| 5288 |
+
"epoch": 0.6163636363636363,
|
| 5289 |
+
"grad_norm": 64015141175296.0,
|
| 5290 |
+
"learning_rate": 3.2139386052013e-07,
|
| 5291 |
+
"loss": 26.21527099609375,
|
| 5292 |
+
"step": 6780
|
| 5293 |
+
},
|
| 5294 |
+
{
|
| 5295 |
+
"epoch": 0.6172727272727273,
|
| 5296 |
+
"grad_norm": 412106407018496.0,
|
| 5297 |
+
"learning_rate": 3.200608087024307e-07,
|
| 5298 |
+
"loss": 26.61980285644531,
|
| 5299 |
+
"step": 6790
|
| 5300 |
+
},
|
| 5301 |
+
{
|
| 5302 |
+
"epoch": 0.6181818181818182,
|
| 5303 |
+
"grad_norm": 402233720045568.0,
|
| 5304 |
+
"learning_rate": 3.1872922459334863e-07,
|
| 5305 |
+
"loss": 26.505010986328124,
|
| 5306 |
+
"step": 6800
|
| 5307 |
+
},
|
| 5308 |
+
{
|
| 5309 |
+
"epoch": 0.6181818181818182,
|
| 5310 |
+
"eval_loss": 26.28169059753418,
|
| 5311 |
+
"eval_runtime": 2.1742,
|
| 5312 |
+
"eval_samples_per_second": 16.558,
|
| 5313 |
+
"eval_steps_per_second": 0.92,
|
| 5314 |
+
"step": 6800
|
| 5315 |
+
},
|
| 5316 |
+
{
|
| 5317 |
+
"epoch": 0.6190909090909091,
|
| 5318 |
+
"grad_norm": 501143427874816.0,
|
| 5319 |
+
"learning_rate": 3.1739911905420614e-07,
|
| 5320 |
+
"loss": 26.6177734375,
|
| 5321 |
+
"step": 6810
|
| 5322 |
+
},
|
| 5323 |
+
{
|
| 5324 |
+
"epoch": 0.62,
|
| 5325 |
+
"grad_norm": 320495861891072.0,
|
| 5326 |
+
"learning_rate": 3.1607050293426476e-07,
|
| 5327 |
+
"loss": 26.375076293945312,
|
| 5328 |
+
"step": 6820
|
| 5329 |
+
},
|
| 5330 |
+
{
|
| 5331 |
+
"epoch": 0.6209090909090909,
|
| 5332 |
+
"grad_norm": 578750198579200.0,
|
| 5333 |
+
"learning_rate": 3.147433870706369e-07,
|
| 5334 |
+
"loss": 26.377716064453125,
|
| 5335 |
+
"step": 6830
|
| 5336 |
+
},
|
| 5337 |
+
{
|
| 5338 |
+
"epoch": 0.6218181818181818,
|
| 5339 |
+
"grad_norm": 264358508625920.0,
|
| 5340 |
+
"learning_rate": 3.13417782288199e-07,
|
| 5341 |
+
"loss": 26.565155029296875,
|
| 5342 |
+
"step": 6840
|
| 5343 |
+
},
|
| 5344 |
+
{
|
| 5345 |
+
"epoch": 0.6227272727272727,
|
| 5346 |
+
"grad_norm": 258190952366080.0,
|
| 5347 |
+
"learning_rate": 3.1209369939950093e-07,
|
| 5348 |
+
"loss": 26.477603149414062,
|
| 5349 |
+
"step": 6850
|
| 5350 |
+
},
|
| 5351 |
+
{
|
| 5352 |
+
"epoch": 0.6236363636363637,
|
| 5353 |
+
"grad_norm": 281976514805760.0,
|
| 5354 |
+
"learning_rate": 3.1077114920467975e-07,
|
| 5355 |
+
"loss": 26.77740478515625,
|
| 5356 |
+
"step": 6860
|
| 5357 |
+
},
|
| 5358 |
+
{
|
| 5359 |
+
"epoch": 0.6245454545454545,
|
| 5360 |
+
"grad_norm": 114142262329344.0,
|
| 5361 |
+
"learning_rate": 3.0945014249137e-07,
|
| 5362 |
+
"loss": 26.84779052734375,
|
| 5363 |
+
"step": 6870
|
| 5364 |
+
},
|
| 5365 |
+
{
|
| 5366 |
+
"epoch": 0.6254545454545455,
|
| 5367 |
+
"grad_norm": 101205963440128.0,
|
| 5368 |
+
"learning_rate": 3.081306900346175e-07,
|
| 5369 |
+
"loss": 26.630465698242187,
|
| 5370 |
+
"step": 6880
|
| 5371 |
+
},
|
| 5372 |
+
{
|
| 5373 |
+
"epoch": 0.6263636363636363,
|
| 5374 |
+
"grad_norm": 598499464839168.0,
|
| 5375 |
+
"learning_rate": 3.0681280259678943e-07,
|
| 5376 |
+
"loss": 26.645831298828124,
|
| 5377 |
+
"step": 6890
|
| 5378 |
+
},
|
| 5379 |
+
{
|
| 5380 |
+
"epoch": 0.6272727272727273,
|
| 5381 |
+
"grad_norm": 204581657116672.0,
|
| 5382 |
+
"learning_rate": 3.0549649092748874e-07,
|
| 5383 |
+
"loss": 26.660739135742187,
|
| 5384 |
+
"step": 6900
|
| 5385 |
+
},
|
| 5386 |
+
{
|
| 5387 |
+
"epoch": 0.6272727272727273,
|
| 5388 |
+
"eval_loss": 26.27537727355957,
|
| 5389 |
+
"eval_runtime": 2.1767,
|
| 5390 |
+
"eval_samples_per_second": 16.539,
|
| 5391 |
+
"eval_steps_per_second": 0.919,
|
| 5392 |
+
"step": 6900
|
| 5393 |
+
},
|
| 5394 |
+
{
|
| 5395 |
+
"epoch": 0.6281818181818182,
|
| 5396 |
+
"grad_norm": 235361171144704.0,
|
| 5397 |
+
"learning_rate": 3.0418176576346405e-07,
|
| 5398 |
+
"loss": 26.733432006835937,
|
| 5399 |
+
"step": 6910
|
| 5400 |
+
},
|
| 5401 |
+
{
|
| 5402 |
+
"epoch": 0.6290909090909091,
|
| 5403 |
+
"grad_norm": 122989131595776.0,
|
| 5404 |
+
"learning_rate": 3.028686378285245e-07,
|
| 5405 |
+
"loss": 26.12663879394531,
|
| 5406 |
+
"step": 6920
|
| 5407 |
+
},
|
| 5408 |
+
{
|
| 5409 |
+
"epoch": 0.63,
|
| 5410 |
+
"grad_norm": 654737095524352.0,
|
| 5411 |
+
"learning_rate": 3.0155711783345044e-07,
|
| 5412 |
+
"loss": 26.359912109375,
|
| 5413 |
+
"step": 6930
|
| 5414 |
+
},
|
| 5415 |
+
{
|
| 5416 |
+
"epoch": 0.6309090909090909,
|
| 5417 |
+
"grad_norm": 146589414850560.0,
|
| 5418 |
+
"learning_rate": 3.002472164759067e-07,
|
| 5419 |
+
"loss": 26.213088989257812,
|
| 5420 |
+
"step": 6940
|
| 5421 |
+
},
|
| 5422 |
+
{
|
| 5423 |
+
"epoch": 0.6318181818181818,
|
| 5424 |
+
"grad_norm": 253996732252160.0,
|
| 5425 |
+
"learning_rate": 2.9893894444035584e-07,
|
| 5426 |
+
"loss": 26.507452392578124,
|
| 5427 |
+
"step": 6950
|
| 5428 |
+
},
|
| 5429 |
+
{
|
| 5430 |
+
"epoch": 0.6327272727272727,
|
| 5431 |
+
"grad_norm": 260389656854528.0,
|
| 5432 |
+
"learning_rate": 2.9763231239797025e-07,
|
| 5433 |
+
"loss": 26.148370361328126,
|
| 5434 |
+
"step": 6960
|
| 5435 |
+
},
|
| 5436 |
+
{
|
| 5437 |
+
"epoch": 0.6336363636363637,
|
| 5438 |
+
"grad_norm": 540127503843328.0,
|
| 5439 |
+
"learning_rate": 2.963273310065456e-07,
|
| 5440 |
+
"loss": 26.83206787109375,
|
| 5441 |
+
"step": 6970
|
| 5442 |
+
},
|
| 5443 |
+
{
|
| 5444 |
+
"epoch": 0.6345454545454545,
|
| 5445 |
+
"grad_norm": 71601550786560.0,
|
| 5446 |
+
"learning_rate": 2.9502401091041334e-07,
|
| 5447 |
+
"loss": 26.635586547851563,
|
| 5448 |
+
"step": 6980
|
| 5449 |
+
},
|
| 5450 |
+
{
|
| 5451 |
+
"epoch": 0.6354545454545455,
|
| 5452 |
+
"grad_norm": 550072433508352.0,
|
| 5453 |
+
"learning_rate": 2.9372236274035466e-07,
|
| 5454 |
+
"loss": 26.5653564453125,
|
| 5455 |
+
"step": 6990
|
| 5456 |
+
},
|
| 5457 |
+
{
|
| 5458 |
+
"epoch": 0.6363636363636364,
|
| 5459 |
+
"grad_norm": 277867640389632.0,
|
| 5460 |
+
"learning_rate": 2.9242239711351304e-07,
|
| 5461 |
+
"loss": 26.53182373046875,
|
| 5462 |
+
"step": 7000
|
| 5463 |
+
},
|
| 5464 |
+
{
|
| 5465 |
+
"epoch": 0.6363636363636364,
|
| 5466 |
+
"eval_loss": 26.28166961669922,
|
| 5467 |
+
"eval_runtime": 2.1741,
|
| 5468 |
+
"eval_samples_per_second": 16.559,
|
| 5469 |
+
"eval_steps_per_second": 0.92,
|
| 5470 |
+
"step": 7000
|
| 5471 |
}
|
| 5472 |
],
|
| 5473 |
"logging_steps": 10,
|
|
|
|
| 5487 |
"attributes": {}
|
| 5488 |
}
|
| 5489 |
},
|
| 5490 |
+
"total_flos": 6.3418933444608e+16,
|
| 5491 |
"train_batch_size": 4,
|
| 5492 |
"trial_name": null,
|
| 5493 |
"trial_params": null
|