Training in progress, step 8000, 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:b70a13b36c334ea1f54ffb5f3083227095b25b5a7b40c9345fd167df916b4f37
|
| 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:cfba6df2db077359f1b6a3b4c158123cf247fbe3722ab6673662579a6d3ece88
|
| 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 sha256:007c0d0a2eb71534a8a1baf1f475fbb7a66ae8beddb78e51908b49e21c5a509a
|
| 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 sha256:58c96086aeb31fae909f44e88cd0e5bbb7f83885f15203654fb87d48097549d3
|
| 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:b5c22022412951b29e6b4330f406b1870d817188c07bdd9fa6a45c3da0d1f54d
|
| 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 sha256:1202abbc6d8d2b5239bcc7f56278a3fd03bbbbb920178d0e0d6537473daa4dd2
|
| 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 sha256:4fc776887b6da6bd34428b5a85b7023e965964cac8547ee61d3ff0c605d1b3db
|
| 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 sha256:4067b82a397b1b9cfb160a353dd480602084dc0ff6c8083d08871025370291dd
|
| 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:bc6596052b9d37c1297136d9615160f12ea20996b7d0fc9eea96275cff5bc705
|
| 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 sha256:69301e8b8bfa56b7f3e0c8ea3e281698b644a7523f1065bd5823da4af4a55fd9
|
| 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 sha256:64bf1268db145ba00529ee988b56673a5a34ca5dc516afb868dd3555f40e2ba2
|
| 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,
|
|
@@ -5468,6 +5468,786 @@
|
|
| 5468 |
"eval_samples_per_second": 16.559,
|
| 5469 |
"eval_steps_per_second": 0.92,
|
| 5470 |
"step": 7000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5471 |
}
|
| 5472 |
],
|
| 5473 |
"logging_steps": 10,
|
|
@@ -5487,7 +6267,7 @@
|
|
| 5487 |
"attributes": {}
|
| 5488 |
}
|
| 5489 |
},
|
| 5490 |
-
"total_flos":
|
| 5491 |
"train_batch_size": 4,
|
| 5492 |
"trial_name": null,
|
| 5493 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.7272727272727273,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 8000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": false,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 5468 |
"eval_samples_per_second": 16.559,
|
| 5469 |
"eval_steps_per_second": 0.92,
|
| 5470 |
"step": 7000
|
| 5471 |
+
},
|
| 5472 |
+
{
|
| 5473 |
+
"epoch": 0.6372727272727273,
|
| 5474 |
+
"grad_norm": 208354400010240.0,
|
| 5475 |
+
"learning_rate": 2.9112412463330807e-07,
|
| 5476 |
+
"loss": 26.38590087890625,
|
| 5477 |
+
"step": 7010
|
| 5478 |
+
},
|
| 5479 |
+
{
|
| 5480 |
+
"epoch": 0.6381818181818182,
|
| 5481 |
+
"grad_norm": 615660476432384.0,
|
| 5482 |
+
"learning_rate": 2.8982755588934925e-07,
|
| 5483 |
+
"loss": 26.374246215820314,
|
| 5484 |
+
"step": 7020
|
| 5485 |
+
},
|
| 5486 |
+
{
|
| 5487 |
+
"epoch": 0.639090909090909,
|
| 5488 |
+
"grad_norm": 227794512510976.0,
|
| 5489 |
+
"learning_rate": 2.885327014573484e-07,
|
| 5490 |
+
"loss": 26.50068359375,
|
| 5491 |
+
"step": 7030
|
| 5492 |
+
},
|
| 5493 |
+
{
|
| 5494 |
+
"epoch": 0.64,
|
| 5495 |
+
"grad_norm": 649978070433792.0,
|
| 5496 |
+
"learning_rate": 2.8723957189903495e-07,
|
| 5497 |
+
"loss": 26.3570556640625,
|
| 5498 |
+
"step": 7040
|
| 5499 |
+
},
|
| 5500 |
+
{
|
| 5501 |
+
"epoch": 0.6409090909090909,
|
| 5502 |
+
"grad_norm": 92747755159552.0,
|
| 5503 |
+
"learning_rate": 2.859481777620688e-07,
|
| 5504 |
+
"loss": 26.351593017578125,
|
| 5505 |
+
"step": 7050
|
| 5506 |
+
},
|
| 5507 |
+
{
|
| 5508 |
+
"epoch": 0.6418181818181818,
|
| 5509 |
+
"grad_norm": 547461831589888.0,
|
| 5510 |
+
"learning_rate": 2.8465852957995485e-07,
|
| 5511 |
+
"loss": 26.504672241210937,
|
| 5512 |
+
"step": 7060
|
| 5513 |
+
},
|
| 5514 |
+
{
|
| 5515 |
+
"epoch": 0.6427272727272727,
|
| 5516 |
+
"grad_norm": 256846124285952.0,
|
| 5517 |
+
"learning_rate": 2.833706378719559e-07,
|
| 5518 |
+
"loss": 27.03226318359375,
|
| 5519 |
+
"step": 7070
|
| 5520 |
+
},
|
| 5521 |
+
{
|
| 5522 |
+
"epoch": 0.6436363636363637,
|
| 5523 |
+
"grad_norm": 146097339105280.0,
|
| 5524 |
+
"learning_rate": 2.8208451314300876e-07,
|
| 5525 |
+
"loss": 26.550949096679688,
|
| 5526 |
+
"step": 7080
|
| 5527 |
+
},
|
| 5528 |
+
{
|
| 5529 |
+
"epoch": 0.6445454545454545,
|
| 5530 |
+
"grad_norm": 463059382435840.0,
|
| 5531 |
+
"learning_rate": 2.80800165883637e-07,
|
| 5532 |
+
"loss": 26.586752319335936,
|
| 5533 |
+
"step": 7090
|
| 5534 |
+
},
|
| 5535 |
+
{
|
| 5536 |
+
"epoch": 0.6454545454545455,
|
| 5537 |
+
"grad_norm": 184904549662720.0,
|
| 5538 |
+
"learning_rate": 2.7951760656986625e-07,
|
| 5539 |
+
"loss": 26.651223754882814,
|
| 5540 |
+
"step": 7100
|
| 5541 |
+
},
|
| 5542 |
+
{
|
| 5543 |
+
"epoch": 0.6454545454545455,
|
| 5544 |
+
"eval_loss": 26.276838302612305,
|
| 5545 |
+
"eval_runtime": 2.1813,
|
| 5546 |
+
"eval_samples_per_second": 16.504,
|
| 5547 |
+
"eval_steps_per_second": 0.917,
|
| 5548 |
+
"step": 7100
|
| 5549 |
+
},
|
| 5550 |
+
{
|
| 5551 |
+
"epoch": 0.6463636363636364,
|
| 5552 |
+
"grad_norm": 497082469187584.0,
|
| 5553 |
+
"learning_rate": 2.782368456631378e-07,
|
| 5554 |
+
"loss": 26.286236572265626,
|
| 5555 |
+
"step": 7110
|
| 5556 |
+
},
|
| 5557 |
+
{
|
| 5558 |
+
"epoch": 0.6472727272727272,
|
| 5559 |
+
"grad_norm": 164335598960640.0,
|
| 5560 |
+
"learning_rate": 2.769578936102245e-07,
|
| 5561 |
+
"loss": 26.465399169921874,
|
| 5562 |
+
"step": 7120
|
| 5563 |
+
},
|
| 5564 |
+
{
|
| 5565 |
+
"epoch": 0.6481818181818182,
|
| 5566 |
+
"grad_norm": 223688372781056.0,
|
| 5567 |
+
"learning_rate": 2.756807608431447e-07,
|
| 5568 |
+
"loss": 26.540838623046874,
|
| 5569 |
+
"step": 7130
|
| 5570 |
+
},
|
| 5571 |
+
{
|
| 5572 |
+
"epoch": 0.649090909090909,
|
| 5573 |
+
"grad_norm": 362925608927232.0,
|
| 5574 |
+
"learning_rate": 2.744054577790774e-07,
|
| 5575 |
+
"loss": 26.485153198242188,
|
| 5576 |
+
"step": 7140
|
| 5577 |
+
},
|
| 5578 |
+
{
|
| 5579 |
+
"epoch": 0.65,
|
| 5580 |
+
"grad_norm": 118669761839104.0,
|
| 5581 |
+
"learning_rate": 2.731319948202776e-07,
|
| 5582 |
+
"loss": 26.7018798828125,
|
| 5583 |
+
"step": 7150
|
| 5584 |
+
},
|
| 5585 |
+
{
|
| 5586 |
+
"epoch": 0.6509090909090909,
|
| 5587 |
+
"grad_norm": 440564222787584.0,
|
| 5588 |
+
"learning_rate": 2.7186038235399027e-07,
|
| 5589 |
+
"loss": 26.294155883789063,
|
| 5590 |
+
"step": 7160
|
| 5591 |
+
},
|
| 5592 |
+
{
|
| 5593 |
+
"epoch": 0.6518181818181819,
|
| 5594 |
+
"grad_norm": 391212498419712.0,
|
| 5595 |
+
"learning_rate": 2.705906307523673e-07,
|
| 5596 |
+
"loss": 26.423309326171875,
|
| 5597 |
+
"step": 7170
|
| 5598 |
+
},
|
| 5599 |
+
{
|
| 5600 |
+
"epoch": 0.6527272727272727,
|
| 5601 |
+
"grad_norm": 291606871670784.0,
|
| 5602 |
+
"learning_rate": 2.693227503723819e-07,
|
| 5603 |
+
"loss": 26.983016967773438,
|
| 5604 |
+
"step": 7180
|
| 5605 |
+
},
|
| 5606 |
+
{
|
| 5607 |
+
"epoch": 0.6536363636363637,
|
| 5608 |
+
"grad_norm": 239193305382912.0,
|
| 5609 |
+
"learning_rate": 2.6805675155574425e-07,
|
| 5610 |
+
"loss": 26.53931884765625,
|
| 5611 |
+
"step": 7190
|
| 5612 |
+
},
|
| 5613 |
+
{
|
| 5614 |
+
"epoch": 0.6545454545454545,
|
| 5615 |
+
"grad_norm": 95832472813568.0,
|
| 5616 |
+
"learning_rate": 2.667926446288169e-07,
|
| 5617 |
+
"loss": 26.326837158203126,
|
| 5618 |
+
"step": 7200
|
| 5619 |
+
},
|
| 5620 |
+
{
|
| 5621 |
+
"epoch": 0.6545454545454545,
|
| 5622 |
+
"eval_loss": 26.273149490356445,
|
| 5623 |
+
"eval_runtime": 2.1822,
|
| 5624 |
+
"eval_samples_per_second": 16.497,
|
| 5625 |
+
"eval_steps_per_second": 0.916,
|
| 5626 |
+
"step": 7200
|
| 5627 |
+
},
|
| 5628 |
+
{
|
| 5629 |
+
"epoch": 0.6554545454545454,
|
| 5630 |
+
"grad_norm": 246679919919104.0,
|
| 5631 |
+
"learning_rate": 2.655304399025312e-07,
|
| 5632 |
+
"loss": 26.533935546875,
|
| 5633 |
+
"step": 7210
|
| 5634 |
+
},
|
| 5635 |
+
{
|
| 5636 |
+
"epoch": 0.6563636363636364,
|
| 5637 |
+
"grad_norm": 521899461312512.0,
|
| 5638 |
+
"learning_rate": 2.6427014767230294e-07,
|
| 5639 |
+
"loss": 26.514251708984375,
|
| 5640 |
+
"step": 7220
|
| 5641 |
+
},
|
| 5642 |
+
{
|
| 5643 |
+
"epoch": 0.6572727272727272,
|
| 5644 |
+
"grad_norm": 228885098659840.0,
|
| 5645 |
+
"learning_rate": 2.630117782179475e-07,
|
| 5646 |
+
"loss": 26.712460327148438,
|
| 5647 |
+
"step": 7230
|
| 5648 |
+
},
|
| 5649 |
+
{
|
| 5650 |
+
"epoch": 0.6581818181818182,
|
| 5651 |
+
"grad_norm": 271884381847552.0,
|
| 5652 |
+
"learning_rate": 2.6175534180359813e-07,
|
| 5653 |
+
"loss": 26.149929809570313,
|
| 5654 |
+
"step": 7240
|
| 5655 |
+
},
|
| 5656 |
+
{
|
| 5657 |
+
"epoch": 0.6590909090909091,
|
| 5658 |
+
"grad_norm": 474217908797440.0,
|
| 5659 |
+
"learning_rate": 2.605008486776195e-07,
|
| 5660 |
+
"loss": 26.453692626953124,
|
| 5661 |
+
"step": 7250
|
| 5662 |
+
},
|
| 5663 |
+
{
|
| 5664 |
+
"epoch": 0.66,
|
| 5665 |
+
"grad_norm": 357043785433088.0,
|
| 5666 |
+
"learning_rate": 2.5924830907252614e-07,
|
| 5667 |
+
"loss": 26.58966064453125,
|
| 5668 |
+
"step": 7260
|
| 5669 |
+
},
|
| 5670 |
+
{
|
| 5671 |
+
"epoch": 0.6609090909090909,
|
| 5672 |
+
"grad_norm": 641387800297472.0,
|
| 5673 |
+
"learning_rate": 2.5799773320489827e-07,
|
| 5674 |
+
"loss": 26.886062622070312,
|
| 5675 |
+
"step": 7270
|
| 5676 |
+
},
|
| 5677 |
+
{
|
| 5678 |
+
"epoch": 0.6618181818181819,
|
| 5679 |
+
"grad_norm": 117462062333952.0,
|
| 5680 |
+
"learning_rate": 2.5674913127529867e-07,
|
| 5681 |
+
"loss": 26.308395385742188,
|
| 5682 |
+
"step": 7280
|
| 5683 |
+
},
|
| 5684 |
+
{
|
| 5685 |
+
"epoch": 0.6627272727272727,
|
| 5686 |
+
"grad_norm": 351057137893376.0,
|
| 5687 |
+
"learning_rate": 2.555025134681886e-07,
|
| 5688 |
+
"loss": 26.34556884765625,
|
| 5689 |
+
"step": 7290
|
| 5690 |
+
},
|
| 5691 |
+
{
|
| 5692 |
+
"epoch": 0.6636363636363637,
|
| 5693 |
+
"grad_norm": 148489820438528.0,
|
| 5694 |
+
"learning_rate": 2.542578899518463e-07,
|
| 5695 |
+
"loss": 26.660372924804687,
|
| 5696 |
+
"step": 7300
|
| 5697 |
+
},
|
| 5698 |
+
{
|
| 5699 |
+
"epoch": 0.6636363636363637,
|
| 5700 |
+
"eval_loss": 26.275470733642578,
|
| 5701 |
+
"eval_runtime": 2.1726,
|
| 5702 |
+
"eval_samples_per_second": 16.57,
|
| 5703 |
+
"eval_steps_per_second": 0.921,
|
| 5704 |
+
"step": 7300
|
| 5705 |
+
},
|
| 5706 |
+
{
|
| 5707 |
+
"epoch": 0.6645454545454546,
|
| 5708 |
+
"grad_norm": 291709548232704.0,
|
| 5709 |
+
"learning_rate": 2.5301527087828287e-07,
|
| 5710 |
+
"loss": 26.54521789550781,
|
| 5711 |
+
"step": 7310
|
| 5712 |
+
},
|
| 5713 |
+
{
|
| 5714 |
+
"epoch": 0.6654545454545454,
|
| 5715 |
+
"grad_norm": 360566128377856.0,
|
| 5716 |
+
"learning_rate": 2.5177466638315945e-07,
|
| 5717 |
+
"loss": 26.633468627929688,
|
| 5718 |
+
"step": 7320
|
| 5719 |
+
},
|
| 5720 |
+
{
|
| 5721 |
+
"epoch": 0.6663636363636364,
|
| 5722 |
+
"grad_norm": 262026844700672.0,
|
| 5723 |
+
"learning_rate": 2.505360865857054e-07,
|
| 5724 |
+
"loss": 26.330783081054687,
|
| 5725 |
+
"step": 7330
|
| 5726 |
+
},
|
| 5727 |
+
{
|
| 5728 |
+
"epoch": 0.6672727272727272,
|
| 5729 |
+
"grad_norm": 365513158950912.0,
|
| 5730 |
+
"learning_rate": 2.492995415886349e-07,
|
| 5731 |
+
"loss": 26.3236572265625,
|
| 5732 |
+
"step": 7340
|
| 5733 |
+
},
|
| 5734 |
+
{
|
| 5735 |
+
"epoch": 0.6681818181818182,
|
| 5736 |
+
"grad_norm": 80057695469568.0,
|
| 5737 |
+
"learning_rate": 2.480650414780653e-07,
|
| 5738 |
+
"loss": 26.489703369140624,
|
| 5739 |
+
"step": 7350
|
| 5740 |
+
},
|
| 5741 |
+
{
|
| 5742 |
+
"epoch": 0.6690909090909091,
|
| 5743 |
+
"grad_norm": 269574847070208.0,
|
| 5744 |
+
"learning_rate": 2.4683259632343364e-07,
|
| 5745 |
+
"loss": 26.589083862304687,
|
| 5746 |
+
"step": 7360
|
| 5747 |
+
},
|
| 5748 |
+
{
|
| 5749 |
+
"epoch": 0.67,
|
| 5750 |
+
"grad_norm": 106464077152256.0,
|
| 5751 |
+
"learning_rate": 2.456022161774164e-07,
|
| 5752 |
+
"loss": 26.390362548828126,
|
| 5753 |
+
"step": 7370
|
| 5754 |
+
},
|
| 5755 |
+
{
|
| 5756 |
+
"epoch": 0.6709090909090909,
|
| 5757 |
+
"grad_norm": 200526067138560.0,
|
| 5758 |
+
"learning_rate": 2.4437391107584556e-07,
|
| 5759 |
+
"loss": 26.3364990234375,
|
| 5760 |
+
"step": 7380
|
| 5761 |
+
},
|
| 5762 |
+
{
|
| 5763 |
+
"epoch": 0.6718181818181819,
|
| 5764 |
+
"grad_norm": 174280209858560.0,
|
| 5765 |
+
"learning_rate": 2.4314769103762784e-07,
|
| 5766 |
+
"loss": 26.455654907226563,
|
| 5767 |
+
"step": 7390
|
| 5768 |
+
},
|
| 5769 |
+
{
|
| 5770 |
+
"epoch": 0.6727272727272727,
|
| 5771 |
+
"grad_norm": 114393173983232.0,
|
| 5772 |
+
"learning_rate": 2.4192356606466303e-07,
|
| 5773 |
+
"loss": 26.521585083007814,
|
| 5774 |
+
"step": 7400
|
| 5775 |
+
},
|
| 5776 |
+
{
|
| 5777 |
+
"epoch": 0.6727272727272727,
|
| 5778 |
+
"eval_loss": 26.27696418762207,
|
| 5779 |
+
"eval_runtime": 2.1578,
|
| 5780 |
+
"eval_samples_per_second": 16.684,
|
| 5781 |
+
"eval_steps_per_second": 0.927,
|
| 5782 |
+
"step": 7400
|
| 5783 |
+
},
|
| 5784 |
+
{
|
| 5785 |
+
"epoch": 0.6736363636363636,
|
| 5786 |
+
"grad_norm": 203887600467968.0,
|
| 5787 |
+
"learning_rate": 2.407015461417616e-07,
|
| 5788 |
+
"loss": 26.397149658203126,
|
| 5789 |
+
"step": 7410
|
| 5790 |
+
},
|
| 5791 |
+
{
|
| 5792 |
+
"epoch": 0.6745454545454546,
|
| 5793 |
+
"grad_norm": 152426476732416.0,
|
| 5794 |
+
"learning_rate": 2.394816412365642e-07,
|
| 5795 |
+
"loss": 26.610260009765625,
|
| 5796 |
+
"step": 7420
|
| 5797 |
+
},
|
| 5798 |
+
{
|
| 5799 |
+
"epoch": 0.6754545454545454,
|
| 5800 |
+
"grad_norm": 281785892077568.0,
|
| 5801 |
+
"learning_rate": 2.3826386129945976e-07,
|
| 5802 |
+
"loss": 26.886895751953126,
|
| 5803 |
+
"step": 7430
|
| 5804 |
+
},
|
| 5805 |
+
{
|
| 5806 |
+
"epoch": 0.6763636363636364,
|
| 5807 |
+
"grad_norm": 123647452774400.0,
|
| 5808 |
+
"learning_rate": 2.3704821626350468e-07,
|
| 5809 |
+
"loss": 26.532232666015624,
|
| 5810 |
+
"step": 7440
|
| 5811 |
+
},
|
| 5812 |
+
{
|
| 5813 |
+
"epoch": 0.6772727272727272,
|
| 5814 |
+
"grad_norm": 195468189499392.0,
|
| 5815 |
+
"learning_rate": 2.358347160443412e-07,
|
| 5816 |
+
"loss": 26.519247436523436,
|
| 5817 |
+
"step": 7450
|
| 5818 |
+
},
|
| 5819 |
+
{
|
| 5820 |
+
"epoch": 0.6781818181818182,
|
| 5821 |
+
"grad_norm": 55197703340032.0,
|
| 5822 |
+
"learning_rate": 2.3462337054011732e-07,
|
| 5823 |
+
"loss": 26.758151245117187,
|
| 5824 |
+
"step": 7460
|
| 5825 |
+
},
|
| 5826 |
+
{
|
| 5827 |
+
"epoch": 0.6790909090909091,
|
| 5828 |
+
"grad_norm": 67209904783360.0,
|
| 5829 |
+
"learning_rate": 2.334141896314057e-07,
|
| 5830 |
+
"loss": 26.347998046875,
|
| 5831 |
+
"step": 7470
|
| 5832 |
+
},
|
| 5833 |
+
{
|
| 5834 |
+
"epoch": 0.68,
|
| 5835 |
+
"grad_norm": 134699796135936.0,
|
| 5836 |
+
"learning_rate": 2.3220718318112281e-07,
|
| 5837 |
+
"loss": 26.302090454101563,
|
| 5838 |
+
"step": 7480
|
| 5839 |
+
},
|
| 5840 |
+
{
|
| 5841 |
+
"epoch": 0.6809090909090909,
|
| 5842 |
+
"grad_norm": 224456467283968.0,
|
| 5843 |
+
"learning_rate": 2.310023610344492e-07,
|
| 5844 |
+
"loss": 26.276641845703125,
|
| 5845 |
+
"step": 7490
|
| 5846 |
+
},
|
| 5847 |
+
{
|
| 5848 |
+
"epoch": 0.6818181818181818,
|
| 5849 |
+
"grad_norm": 669328441606144.0,
|
| 5850 |
+
"learning_rate": 2.2979973301874795e-07,
|
| 5851 |
+
"loss": 26.991085815429688,
|
| 5852 |
+
"step": 7500
|
| 5853 |
+
},
|
| 5854 |
+
{
|
| 5855 |
+
"epoch": 0.6818181818181818,
|
| 5856 |
+
"eval_loss": 26.273746490478516,
|
| 5857 |
+
"eval_runtime": 2.155,
|
| 5858 |
+
"eval_samples_per_second": 16.705,
|
| 5859 |
+
"eval_steps_per_second": 0.928,
|
| 5860 |
+
"step": 7500
|
| 5861 |
+
},
|
| 5862 |
+
{
|
| 5863 |
+
"epoch": 0.6827272727272727,
|
| 5864 |
+
"grad_norm": 246346388865024.0,
|
| 5865 |
+
"learning_rate": 2.2859930894348594e-07,
|
| 5866 |
+
"loss": 26.169650268554687,
|
| 5867 |
+
"step": 7510
|
| 5868 |
+
},
|
| 5869 |
+
{
|
| 5870 |
+
"epoch": 0.6836363636363636,
|
| 5871 |
+
"grad_norm": 117372857876480.0,
|
| 5872 |
+
"learning_rate": 2.2740109860015304e-07,
|
| 5873 |
+
"loss": 26.467178344726562,
|
| 5874 |
+
"step": 7520
|
| 5875 |
+
},
|
| 5876 |
+
{
|
| 5877 |
+
"epoch": 0.6845454545454546,
|
| 5878 |
+
"grad_norm": 186215957528576.0,
|
| 5879 |
+
"learning_rate": 2.2620511176218255e-07,
|
| 5880 |
+
"loss": 26.1698486328125,
|
| 5881 |
+
"step": 7530
|
| 5882 |
+
},
|
| 5883 |
+
{
|
| 5884 |
+
"epoch": 0.6854545454545454,
|
| 5885 |
+
"grad_norm": 167226934034432.0,
|
| 5886 |
+
"learning_rate": 2.250113581848706e-07,
|
| 5887 |
+
"loss": 26.436212158203126,
|
| 5888 |
+
"step": 7540
|
| 5889 |
+
},
|
| 5890 |
+
{
|
| 5891 |
+
"epoch": 0.6863636363636364,
|
| 5892 |
+
"grad_norm": 54176792969216.0,
|
| 5893 |
+
"learning_rate": 2.2381984760529816e-07,
|
| 5894 |
+
"loss": 26.699081420898438,
|
| 5895 |
+
"step": 7550
|
| 5896 |
+
},
|
| 5897 |
+
{
|
| 5898 |
+
"epoch": 0.6872727272727273,
|
| 5899 |
+
"grad_norm": 131126349791232.0,
|
| 5900 |
+
"learning_rate": 2.2263058974225023e-07,
|
| 5901 |
+
"loss": 26.650469970703124,
|
| 5902 |
+
"step": 7560
|
| 5903 |
+
},
|
| 5904 |
+
{
|
| 5905 |
+
"epoch": 0.6881818181818182,
|
| 5906 |
+
"grad_norm": 190212038721536.0,
|
| 5907 |
+
"learning_rate": 2.214435942961377e-07,
|
| 5908 |
+
"loss": 26.493695068359376,
|
| 5909 |
+
"step": 7570
|
| 5910 |
+
},
|
| 5911 |
+
{
|
| 5912 |
+
"epoch": 0.6890909090909091,
|
| 5913 |
+
"grad_norm": 263704901844992.0,
|
| 5914 |
+
"learning_rate": 2.2025887094891655e-07,
|
| 5915 |
+
"loss": 26.46141662597656,
|
| 5916 |
+
"step": 7580
|
| 5917 |
+
},
|
| 5918 |
+
{
|
| 5919 |
+
"epoch": 0.69,
|
| 5920 |
+
"grad_norm": 391329167179776.0,
|
| 5921 |
+
"learning_rate": 2.19076429364011e-07,
|
| 5922 |
+
"loss": 26.3736083984375,
|
| 5923 |
+
"step": 7590
|
| 5924 |
+
},
|
| 5925 |
+
{
|
| 5926 |
+
"epoch": 0.6909090909090909,
|
| 5927 |
+
"grad_norm": 119184780427264.0,
|
| 5928 |
+
"learning_rate": 2.178962791862333e-07,
|
| 5929 |
+
"loss": 26.546969604492187,
|
| 5930 |
+
"step": 7600
|
| 5931 |
+
},
|
| 5932 |
+
{
|
| 5933 |
+
"epoch": 0.6909090909090909,
|
| 5934 |
+
"eval_loss": 26.27129364013672,
|
| 5935 |
+
"eval_runtime": 2.1527,
|
| 5936 |
+
"eval_samples_per_second": 16.723,
|
| 5937 |
+
"eval_steps_per_second": 0.929,
|
| 5938 |
+
"step": 7600
|
| 5939 |
+
},
|
| 5940 |
+
{
|
| 5941 |
+
"epoch": 0.6918181818181818,
|
| 5942 |
+
"grad_norm": 177234845368320.0,
|
| 5943 |
+
"learning_rate": 2.1671843004170483e-07,
|
| 5944 |
+
"loss": 26.3050048828125,
|
| 5945 |
+
"step": 7610
|
| 5946 |
+
},
|
| 5947 |
+
{
|
| 5948 |
+
"epoch": 0.6927272727272727,
|
| 5949 |
+
"grad_norm": 323850130685952.0,
|
| 5950 |
+
"learning_rate": 2.1554289153777933e-07,
|
| 5951 |
+
"loss": 26.890570068359374,
|
| 5952 |
+
"step": 7620
|
| 5953 |
+
},
|
| 5954 |
+
{
|
| 5955 |
+
"epoch": 0.6936363636363636,
|
| 5956 |
+
"grad_norm": 153877135491072.0,
|
| 5957 |
+
"learning_rate": 2.1436967326296213e-07,
|
| 5958 |
+
"loss": 26.495651245117188,
|
| 5959 |
+
"step": 7630
|
| 5960 |
+
},
|
| 5961 |
+
{
|
| 5962 |
+
"epoch": 0.6945454545454546,
|
| 5963 |
+
"grad_norm": 196137852076032.0,
|
| 5964 |
+
"learning_rate": 2.1319878478683396e-07,
|
| 5965 |
+
"loss": 26.359036254882813,
|
| 5966 |
+
"step": 7640
|
| 5967 |
+
},
|
| 5968 |
+
{
|
| 5969 |
+
"epoch": 0.6954545454545454,
|
| 5970 |
+
"grad_norm": 99828335902720.0,
|
| 5971 |
+
"learning_rate": 2.1203023565997173e-07,
|
| 5972 |
+
"loss": 26.25531005859375,
|
| 5973 |
+
"step": 7650
|
| 5974 |
+
},
|
| 5975 |
+
{
|
| 5976 |
+
"epoch": 0.6963636363636364,
|
| 5977 |
+
"grad_norm": 566579267895296.0,
|
| 5978 |
+
"learning_rate": 2.1086403541387144e-07,
|
| 5979 |
+
"loss": 26.096768188476563,
|
| 5980 |
+
"step": 7660
|
| 5981 |
+
},
|
| 5982 |
+
{
|
| 5983 |
+
"epoch": 0.6972727272727273,
|
| 5984 |
+
"grad_norm": 206705417781248.0,
|
| 5985 |
+
"learning_rate": 2.097001935608692e-07,
|
| 5986 |
+
"loss": 26.354949951171875,
|
| 5987 |
+
"step": 7670
|
| 5988 |
+
},
|
| 5989 |
+
{
|
| 5990 |
+
"epoch": 0.6981818181818182,
|
| 5991 |
+
"grad_norm": 88147518029824.0,
|
| 5992 |
+
"learning_rate": 2.085387195940651e-07,
|
| 5993 |
+
"loss": 26.522039794921874,
|
| 5994 |
+
"step": 7680
|
| 5995 |
+
},
|
| 5996 |
+
{
|
| 5997 |
+
"epoch": 0.6990909090909091,
|
| 5998 |
+
"grad_norm": 223890320130048.0,
|
| 5999 |
+
"learning_rate": 2.073796229872451e-07,
|
| 6000 |
+
"loss": 26.2290283203125,
|
| 6001 |
+
"step": 7690
|
| 6002 |
+
},
|
| 6003 |
+
{
|
| 6004 |
+
"epoch": 0.7,
|
| 6005 |
+
"grad_norm": 142860158500864.0,
|
| 6006 |
+
"learning_rate": 2.0622291319480316e-07,
|
| 6007 |
+
"loss": 26.3935791015625,
|
| 6008 |
+
"step": 7700
|
| 6009 |
+
},
|
| 6010 |
+
{
|
| 6011 |
+
"epoch": 0.7,
|
| 6012 |
+
"eval_loss": 26.27055549621582,
|
| 6013 |
+
"eval_runtime": 2.1622,
|
| 6014 |
+
"eval_samples_per_second": 16.649,
|
| 6015 |
+
"eval_steps_per_second": 0.925,
|
| 6016 |
+
"step": 7700
|
| 6017 |
+
},
|
| 6018 |
+
{
|
| 6019 |
+
"epoch": 0.7009090909090909,
|
| 6020 |
+
"grad_norm": 76660552499200.0,
|
| 6021 |
+
"learning_rate": 2.0506859965166546e-07,
|
| 6022 |
+
"loss": 26.559854125976564,
|
| 6023 |
+
"step": 7710
|
| 6024 |
+
},
|
| 6025 |
+
{
|
| 6026 |
+
"epoch": 0.7018181818181818,
|
| 6027 |
+
"grad_norm": 165078661857280.0,
|
| 6028 |
+
"learning_rate": 2.039166917732123e-07,
|
| 6029 |
+
"loss": 26.336550903320312,
|
| 6030 |
+
"step": 7720
|
| 6031 |
+
},
|
| 6032 |
+
{
|
| 6033 |
+
"epoch": 0.7027272727272728,
|
| 6034 |
+
"grad_norm": 131228850192384.0,
|
| 6035 |
+
"learning_rate": 2.02767198955202e-07,
|
| 6036 |
+
"loss": 26.307232666015626,
|
| 6037 |
+
"step": 7730
|
| 6038 |
+
},
|
| 6039 |
+
{
|
| 6040 |
+
"epoch": 0.7036363636363636,
|
| 6041 |
+
"grad_norm": 158765814906880.0,
|
| 6042 |
+
"learning_rate": 2.0162013057369327e-07,
|
| 6043 |
+
"loss": 26.472689819335937,
|
| 6044 |
+
"step": 7740
|
| 6045 |
+
},
|
| 6046 |
+
{
|
| 6047 |
+
"epoch": 0.7045454545454546,
|
| 6048 |
+
"grad_norm": 415826083577856.0,
|
| 6049 |
+
"learning_rate": 2.0047549598497064e-07,
|
| 6050 |
+
"loss": 26.57115478515625,
|
| 6051 |
+
"step": 7750
|
| 6052 |
+
},
|
| 6053 |
+
{
|
| 6054 |
+
"epoch": 0.7054545454545454,
|
| 6055 |
+
"grad_norm": 340710863667200.0,
|
| 6056 |
+
"learning_rate": 1.9933330452546575e-07,
|
| 6057 |
+
"loss": 26.334005737304686,
|
| 6058 |
+
"step": 7760
|
| 6059 |
+
},
|
| 6060 |
+
{
|
| 6061 |
+
"epoch": 0.7063636363636364,
|
| 6062 |
+
"grad_norm": 200871073808384.0,
|
| 6063 |
+
"learning_rate": 1.9819356551168304e-07,
|
| 6064 |
+
"loss": 26.425982666015624,
|
| 6065 |
+
"step": 7770
|
| 6066 |
+
},
|
| 6067 |
+
{
|
| 6068 |
+
"epoch": 0.7072727272727273,
|
| 6069 |
+
"grad_norm": 326121094643712.0,
|
| 6070 |
+
"learning_rate": 1.9705628824012315e-07,
|
| 6071 |
+
"loss": 26.586672973632812,
|
| 6072 |
+
"step": 7780
|
| 6073 |
+
},
|
| 6074 |
+
{
|
| 6075 |
+
"epoch": 0.7081818181818181,
|
| 6076 |
+
"grad_norm": 311439185149952.0,
|
| 6077 |
+
"learning_rate": 1.9592148198720655e-07,
|
| 6078 |
+
"loss": 26.54600830078125,
|
| 6079 |
+
"step": 7790
|
| 6080 |
+
},
|
| 6081 |
+
{
|
| 6082 |
+
"epoch": 0.7090909090909091,
|
| 6083 |
+
"grad_norm": 154627949461504.0,
|
| 6084 |
+
"learning_rate": 1.9478915600919877e-07,
|
| 6085 |
+
"loss": 26.687026977539062,
|
| 6086 |
+
"step": 7800
|
| 6087 |
+
},
|
| 6088 |
+
{
|
| 6089 |
+
"epoch": 0.7090909090909091,
|
| 6090 |
+
"eval_loss": 26.266042709350586,
|
| 6091 |
+
"eval_runtime": 2.1833,
|
| 6092 |
+
"eval_samples_per_second": 16.489,
|
| 6093 |
+
"eval_steps_per_second": 0.916,
|
| 6094 |
+
"step": 7800
|
| 6095 |
+
},
|
| 6096 |
+
{
|
| 6097 |
+
"epoch": 0.71,
|
| 6098 |
+
"grad_norm": 325412290822144.0,
|
| 6099 |
+
"learning_rate": 1.9365931954213443e-07,
|
| 6100 |
+
"loss": 26.345108032226562,
|
| 6101 |
+
"step": 7810
|
| 6102 |
+
},
|
| 6103 |
+
{
|
| 6104 |
+
"epoch": 0.7109090909090909,
|
| 6105 |
+
"grad_norm": 103427434610688.0,
|
| 6106 |
+
"learning_rate": 1.9253198180174213e-07,
|
| 6107 |
+
"loss": 26.439935302734376,
|
| 6108 |
+
"step": 7820
|
| 6109 |
+
},
|
| 6110 |
+
{
|
| 6111 |
+
"epoch": 0.7118181818181818,
|
| 6112 |
+
"grad_norm": 491774460035072.0,
|
| 6113 |
+
"learning_rate": 1.9140715198336844e-07,
|
| 6114 |
+
"loss": 26.82066650390625,
|
| 6115 |
+
"step": 7830
|
| 6116 |
+
},
|
| 6117 |
+
{
|
| 6118 |
+
"epoch": 0.7127272727272728,
|
| 6119 |
+
"grad_norm": 210541125566464.0,
|
| 6120 |
+
"learning_rate": 1.9028483926190474e-07,
|
| 6121 |
+
"loss": 26.39219970703125,
|
| 6122 |
+
"step": 7840
|
| 6123 |
+
},
|
| 6124 |
+
{
|
| 6125 |
+
"epoch": 0.7136363636363636,
|
| 6126 |
+
"grad_norm": 856512243171328.0,
|
| 6127 |
+
"learning_rate": 1.8916505279171007e-07,
|
| 6128 |
+
"loss": 26.36131591796875,
|
| 6129 |
+
"step": 7850
|
| 6130 |
+
},
|
| 6131 |
+
{
|
| 6132 |
+
"epoch": 0.7145454545454546,
|
| 6133 |
+
"grad_norm": 417027902668800.0,
|
| 6134 |
+
"learning_rate": 1.8804780170653833e-07,
|
| 6135 |
+
"loss": 26.504092407226562,
|
| 6136 |
+
"step": 7860
|
| 6137 |
+
},
|
| 6138 |
+
{
|
| 6139 |
+
"epoch": 0.7154545454545455,
|
| 6140 |
+
"grad_norm": 292983341580288.0,
|
| 6141 |
+
"learning_rate": 1.8693309511946304e-07,
|
| 6142 |
+
"loss": 26.41829833984375,
|
| 6143 |
+
"step": 7870
|
| 6144 |
+
},
|
| 6145 |
+
{
|
| 6146 |
+
"epoch": 0.7163636363636363,
|
| 6147 |
+
"grad_norm": 327994371473408.0,
|
| 6148 |
+
"learning_rate": 1.8582094212280237e-07,
|
| 6149 |
+
"loss": 26.64545593261719,
|
| 6150 |
+
"step": 7880
|
| 6151 |
+
},
|
| 6152 |
+
{
|
| 6153 |
+
"epoch": 0.7172727272727273,
|
| 6154 |
+
"grad_norm": 218186469343232.0,
|
| 6155 |
+
"learning_rate": 1.8471135178804641e-07,
|
| 6156 |
+
"loss": 26.180670166015624,
|
| 6157 |
+
"step": 7890
|
| 6158 |
+
},
|
| 6159 |
+
{
|
| 6160 |
+
"epoch": 0.7181818181818181,
|
| 6161 |
+
"grad_norm": 82476139544576.0,
|
| 6162 |
+
"learning_rate": 1.8360433316578205e-07,
|
| 6163 |
+
"loss": 26.653558349609376,
|
| 6164 |
+
"step": 7900
|
| 6165 |
+
},
|
| 6166 |
+
{
|
| 6167 |
+
"epoch": 0.7181818181818181,
|
| 6168 |
+
"eval_loss": 26.26190948486328,
|
| 6169 |
+
"eval_runtime": 2.1559,
|
| 6170 |
+
"eval_samples_per_second": 16.698,
|
| 6171 |
+
"eval_steps_per_second": 0.928,
|
| 6172 |
+
"step": 7900
|
| 6173 |
+
},
|
| 6174 |
+
{
|
| 6175 |
+
"epoch": 0.7190909090909091,
|
| 6176 |
+
"grad_norm": 118652909125632.0,
|
| 6177 |
+
"learning_rate": 1.8249989528561976e-07,
|
| 6178 |
+
"loss": 26.234396362304686,
|
| 6179 |
+
"step": 7910
|
| 6180 |
+
},
|
| 6181 |
+
{
|
| 6182 |
+
"epoch": 0.72,
|
| 6183 |
+
"grad_norm": 125824606928896.0,
|
| 6184 |
+
"learning_rate": 1.8139804715611916e-07,
|
| 6185 |
+
"loss": 26.377532958984375,
|
| 6186 |
+
"step": 7920
|
| 6187 |
+
},
|
| 6188 |
+
{
|
| 6189 |
+
"epoch": 0.7209090909090909,
|
| 6190 |
+
"grad_norm": 130430497980416.0,
|
| 6191 |
+
"learning_rate": 1.8029879776471661e-07,
|
| 6192 |
+
"loss": 26.323135375976562,
|
| 6193 |
+
"step": 7930
|
| 6194 |
+
},
|
| 6195 |
+
{
|
| 6196 |
+
"epoch": 0.7218181818181818,
|
| 6197 |
+
"grad_norm": 189826951282688.0,
|
| 6198 |
+
"learning_rate": 1.792021560776511e-07,
|
| 6199 |
+
"loss": 26.728903198242186,
|
| 6200 |
+
"step": 7940
|
| 6201 |
+
},
|
| 6202 |
+
{
|
| 6203 |
+
"epoch": 0.7227272727272728,
|
| 6204 |
+
"grad_norm": 148400968302592.0,
|
| 6205 |
+
"learning_rate": 1.781081310398917e-07,
|
| 6206 |
+
"loss": 26.242092895507813,
|
| 6207 |
+
"step": 7950
|
| 6208 |
+
},
|
| 6209 |
+
{
|
| 6210 |
+
"epoch": 0.7236363636363636,
|
| 6211 |
+
"grad_norm": 134777038438400.0,
|
| 6212 |
+
"learning_rate": 1.770167315750636e-07,
|
| 6213 |
+
"loss": 26.503350830078126,
|
| 6214 |
+
"step": 7960
|
| 6215 |
+
},
|
| 6216 |
+
{
|
| 6217 |
+
"epoch": 0.7245454545454545,
|
| 6218 |
+
"grad_norm": 511105268973568.0,
|
| 6219 |
+
"learning_rate": 1.7592796658537658e-07,
|
| 6220 |
+
"loss": 26.567254638671876,
|
| 6221 |
+
"step": 7970
|
| 6222 |
+
},
|
| 6223 |
+
{
|
| 6224 |
+
"epoch": 0.7254545454545455,
|
| 6225 |
+
"grad_norm": 110472221163520.0,
|
| 6226 |
+
"learning_rate": 1.748418449515519e-07,
|
| 6227 |
+
"loss": 26.81078186035156,
|
| 6228 |
+
"step": 7980
|
| 6229 |
+
},
|
| 6230 |
+
{
|
| 6231 |
+
"epoch": 0.7263636363636363,
|
| 6232 |
+
"grad_norm": 329754871857152.0,
|
| 6233 |
+
"learning_rate": 1.7375837553274897e-07,
|
| 6234 |
+
"loss": 26.368780517578124,
|
| 6235 |
+
"step": 7990
|
| 6236 |
+
},
|
| 6237 |
+
{
|
| 6238 |
+
"epoch": 0.7272727272727273,
|
| 6239 |
+
"grad_norm": 74490444775424.0,
|
| 6240 |
+
"learning_rate": 1.7267756716649523e-07,
|
| 6241 |
+
"loss": 26.744158935546874,
|
| 6242 |
+
"step": 8000
|
| 6243 |
+
},
|
| 6244 |
+
{
|
| 6245 |
+
"epoch": 0.7272727272727273,
|
| 6246 |
+
"eval_loss": 26.262548446655273,
|
| 6247 |
+
"eval_runtime": 2.1821,
|
| 6248 |
+
"eval_samples_per_second": 16.498,
|
| 6249 |
+
"eval_steps_per_second": 0.917,
|
| 6250 |
+
"step": 8000
|
| 6251 |
}
|
| 6252 |
],
|
| 6253 |
"logging_steps": 10,
|
|
|
|
| 6267 |
"attributes": {}
|
| 6268 |
}
|
| 6269 |
},
|
| 6270 |
+
"total_flos": 7.2478781079552e+16,
|
| 6271 |
"train_batch_size": 4,
|
| 6272 |
"trial_name": null,
|
| 6273 |
"trial_params": null
|