Training in progress, step 3000, 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:b9d9ea30bd9c708493572bd44e3256c83dc3c2f0da6636321410d4a296da4fce
|
| 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 sha256:8c473a88f8a30921081abbf5456efaacd00f88eda77799e15230de46b74af750
|
| 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 sha256:5b3ee6aa25e2d997147f110f0106bfa09c947f3a17a42ac631fe9fd10d359c64
|
| 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 sha256:312862b75420285bfb1144ee02e5d98151fa644e025e2b8bdf98a1cfb6bf1784
|
| 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 sha256:08c57e79b84a1cd74caf1436d7ba51e307367bec5d0e696893db9df56dd4534a
|
| 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:f85ac9ae48e62b6c6e8aa06bdecb611b6e6ee537c6d82b6b083d470d2f737981
|
| 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:aff24c8c76bc594a4ec846173c30674664beb048676e34f910d356116d9444f2
|
| 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 sha256:9bf16c3d4ba315c00cb40b017cfa96465f22cde05a0b5413db32688107d9373b
|
| 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 sha256:9c75e30f4407f49e03a0397235b46bd9e402554a12147d65fe85d21a0fbfedb6
|
| 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:d9413bda3b0c7ad2808ddc0f627a8614796347dd04936bc178f8cb262d5460fd
|
| 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:aa66f2dbce1093003d024195070a3c39d20fa11c87dce6abb88ed9710e3bf6f3
|
| 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,
|
|
@@ -1568,6 +1568,786 @@
|
|
| 1568 |
"eval_samples_per_second": 1.775,
|
| 1569 |
"eval_steps_per_second": 0.444,
|
| 1570 |
"step": 2000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1571 |
}
|
| 1572 |
],
|
| 1573 |
"logging_steps": 10,
|
|
@@ -1587,7 +2367,7 @@
|
|
| 1587 |
"attributes": {}
|
| 1588 |
}
|
| 1589 |
},
|
| 1590 |
-
"total_flos":
|
| 1591 |
"train_batch_size": 1,
|
| 1592 |
"trial_name": null,
|
| 1593 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.2727272727272727,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 3000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": false,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1568 |
"eval_samples_per_second": 1.775,
|
| 1569 |
"eval_steps_per_second": 0.444,
|
| 1570 |
"step": 2000
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.18272727272727274,
|
| 1574 |
+
"grad_norm": 71538795610112.0,
|
| 1575 |
+
"learning_rate": 9.199305477028869e-07,
|
| 1576 |
+
"loss": 31.124655151367186,
|
| 1577 |
+
"step": 2010
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.18363636363636363,
|
| 1581 |
+
"grad_norm": 117176044355584.0,
|
| 1582 |
+
"learning_rate": 9.19153717573394e-07,
|
| 1583 |
+
"loss": 31.152999877929688,
|
| 1584 |
+
"step": 2020
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.18454545454545454,
|
| 1588 |
+
"grad_norm": 88894632624128.0,
|
| 1589 |
+
"learning_rate": 9.183734685359972e-07,
|
| 1590 |
+
"loss": 31.35181884765625,
|
| 1591 |
+
"step": 2030
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.18545454545454546,
|
| 1595 |
+
"grad_norm": 56644360732672.0,
|
| 1596 |
+
"learning_rate": 9.175898069549476e-07,
|
| 1597 |
+
"loss": 31.333255004882812,
|
| 1598 |
+
"step": 2040
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.18636363636363637,
|
| 1602 |
+
"grad_norm": 104232355102720.0,
|
| 1603 |
+
"learning_rate": 9.168027392223319e-07,
|
| 1604 |
+
"loss": 31.229754638671874,
|
| 1605 |
+
"step": 2050
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.18727272727272729,
|
| 1609 |
+
"grad_norm": 93866015326208.0,
|
| 1610 |
+
"learning_rate": 9.160122717580194e-07,
|
| 1611 |
+
"loss": 31.03287353515625,
|
| 1612 |
+
"step": 2060
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.18818181818181817,
|
| 1616 |
+
"grad_norm": 45236478803968.0,
|
| 1617 |
+
"learning_rate": 9.152184110096097e-07,
|
| 1618 |
+
"loss": 31.709677124023436,
|
| 1619 |
+
"step": 2070
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.1890909090909091,
|
| 1623 |
+
"grad_norm": 46076950216704.0,
|
| 1624 |
+
"learning_rate": 9.144211634523808e-07,
|
| 1625 |
+
"loss": 31.2308837890625,
|
| 1626 |
+
"step": 2080
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.19,
|
| 1630 |
+
"grad_norm": 72614668140544.0,
|
| 1631 |
+
"learning_rate": 9.13620535589236e-07,
|
| 1632 |
+
"loss": 31.108331298828126,
|
| 1633 |
+
"step": 2090
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.19090909090909092,
|
| 1637 |
+
"grad_norm": 42438555074560.0,
|
| 1638 |
+
"learning_rate": 9.128165339506502e-07,
|
| 1639 |
+
"loss": 31.92352294921875,
|
| 1640 |
+
"step": 2100
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.19090909090909092,
|
| 1644 |
+
"eval_loss": 31.03235626220703,
|
| 1645 |
+
"eval_runtime": 2.2586,
|
| 1646 |
+
"eval_samples_per_second": 1.771,
|
| 1647 |
+
"eval_steps_per_second": 0.443,
|
| 1648 |
+
"step": 2100
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.1918181818181818,
|
| 1652 |
+
"grad_norm": 63270400557056.0,
|
| 1653 |
+
"learning_rate": 9.120091650946171e-07,
|
| 1654 |
+
"loss": 31.78459777832031,
|
| 1655 |
+
"step": 2110
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.19272727272727272,
|
| 1659 |
+
"grad_norm": 79119404498944.0,
|
| 1660 |
+
"learning_rate": 9.111984356065966e-07,
|
| 1661 |
+
"loss": 31.268167114257814,
|
| 1662 |
+
"step": 2120
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.19363636363636363,
|
| 1666 |
+
"grad_norm": 29490430345216.0,
|
| 1667 |
+
"learning_rate": 9.103843520994592e-07,
|
| 1668 |
+
"loss": 31.48218994140625,
|
| 1669 |
+
"step": 2130
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.19454545454545455,
|
| 1673 |
+
"grad_norm": 34231048732672.0,
|
| 1674 |
+
"learning_rate": 9.095669212134339e-07,
|
| 1675 |
+
"loss": 30.843756103515624,
|
| 1676 |
+
"step": 2140
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.19545454545454546,
|
| 1680 |
+
"grad_norm": 78581879275520.0,
|
| 1681 |
+
"learning_rate": 9.087461496160528e-07,
|
| 1682 |
+
"loss": 30.995782470703126,
|
| 1683 |
+
"step": 2150
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.19636363636363635,
|
| 1687 |
+
"grad_norm": 44611347152896.0,
|
| 1688 |
+
"learning_rate": 9.079220440020972e-07,
|
| 1689 |
+
"loss": 31.502972412109376,
|
| 1690 |
+
"step": 2160
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.19727272727272727,
|
| 1694 |
+
"grad_norm": 37552992026624.0,
|
| 1695 |
+
"learning_rate": 9.070946110935431e-07,
|
| 1696 |
+
"loss": 31.0704833984375,
|
| 1697 |
+
"step": 2170
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.19818181818181818,
|
| 1701 |
+
"grad_norm": 47458872721408.0,
|
| 1702 |
+
"learning_rate": 9.062638576395061e-07,
|
| 1703 |
+
"loss": 31.406292724609376,
|
| 1704 |
+
"step": 2180
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.1990909090909091,
|
| 1708 |
+
"grad_norm": 92816332029952.0,
|
| 1709 |
+
"learning_rate": 9.054297904161867e-07,
|
| 1710 |
+
"loss": 31.4571044921875,
|
| 1711 |
+
"step": 2190
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.2,
|
| 1715 |
+
"grad_norm": 29963092754432.0,
|
| 1716 |
+
"learning_rate": 9.045924162268144e-07,
|
| 1717 |
+
"loss": 31.258853149414062,
|
| 1718 |
+
"step": 2200
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.2,
|
| 1722 |
+
"eval_loss": 31.03536605834961,
|
| 1723 |
+
"eval_runtime": 2.2498,
|
| 1724 |
+
"eval_samples_per_second": 1.778,
|
| 1725 |
+
"eval_steps_per_second": 0.444,
|
| 1726 |
+
"step": 2200
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 0.2009090909090909,
|
| 1730 |
+
"grad_norm": 259804182347776.0,
|
| 1731 |
+
"learning_rate": 9.037517419015928e-07,
|
| 1732 |
+
"loss": 31.453973388671876,
|
| 1733 |
+
"step": 2210
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.2018181818181818,
|
| 1737 |
+
"grad_norm": 98955845173248.0,
|
| 1738 |
+
"learning_rate": 9.029077742976437e-07,
|
| 1739 |
+
"loss": 30.824130249023437,
|
| 1740 |
+
"step": 2220
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 0.20272727272727273,
|
| 1744 |
+
"grad_norm": 35170289713152.0,
|
| 1745 |
+
"learning_rate": 9.020605202989513e-07,
|
| 1746 |
+
"loss": 31.691207885742188,
|
| 1747 |
+
"step": 2230
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 0.20363636363636364,
|
| 1751 |
+
"grad_norm": 25527377199104.0,
|
| 1752 |
+
"learning_rate": 9.012099868163056e-07,
|
| 1753 |
+
"loss": 31.00306091308594,
|
| 1754 |
+
"step": 2240
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 0.20454545454545456,
|
| 1758 |
+
"grad_norm": 136539543699456.0,
|
| 1759 |
+
"learning_rate": 9.003561807872469e-07,
|
| 1760 |
+
"loss": 31.796575927734374,
|
| 1761 |
+
"step": 2250
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 0.20545454545454545,
|
| 1765 |
+
"grad_norm": 46742410100736.0,
|
| 1766 |
+
"learning_rate": 8.994991091760076e-07,
|
| 1767 |
+
"loss": 31.3921142578125,
|
| 1768 |
+
"step": 2260
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.20636363636363636,
|
| 1772 |
+
"grad_norm": 30860340363264.0,
|
| 1773 |
+
"learning_rate": 8.986387789734576e-07,
|
| 1774 |
+
"loss": 31.278704833984374,
|
| 1775 |
+
"step": 2270
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 0.20727272727272728,
|
| 1779 |
+
"grad_norm": 31539043762176.0,
|
| 1780 |
+
"learning_rate": 8.977751971970456e-07,
|
| 1781 |
+
"loss": 30.982208251953125,
|
| 1782 |
+
"step": 2280
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 0.2081818181818182,
|
| 1786 |
+
"grad_norm": 134957301235712.0,
|
| 1787 |
+
"learning_rate": 8.969083708907424e-07,
|
| 1788 |
+
"loss": 31.08028564453125,
|
| 1789 |
+
"step": 2290
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 0.20909090909090908,
|
| 1793 |
+
"grad_norm": 86353354162176.0,
|
| 1794 |
+
"learning_rate": 8.960383071249835e-07,
|
| 1795 |
+
"loss": 31.018426513671876,
|
| 1796 |
+
"step": 2300
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.20909090909090908,
|
| 1800 |
+
"eval_loss": 31.035140991210938,
|
| 1801 |
+
"eval_runtime": 2.2639,
|
| 1802 |
+
"eval_samples_per_second": 1.767,
|
| 1803 |
+
"eval_steps_per_second": 0.442,
|
| 1804 |
+
"step": 2300
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 0.21,
|
| 1808 |
+
"grad_norm": 155203122757632.0,
|
| 1809 |
+
"learning_rate": 8.951650129966112e-07,
|
| 1810 |
+
"loss": 31.086285400390626,
|
| 1811 |
+
"step": 2310
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 0.2109090909090909,
|
| 1815 |
+
"grad_norm": 109969449943040.0,
|
| 1816 |
+
"learning_rate": 8.942884956288171e-07,
|
| 1817 |
+
"loss": 30.969015502929686,
|
| 1818 |
+
"step": 2320
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 0.21181818181818182,
|
| 1822 |
+
"grad_norm": 101191014940672.0,
|
| 1823 |
+
"learning_rate": 8.934087621710835e-07,
|
| 1824 |
+
"loss": 30.67431640625,
|
| 1825 |
+
"step": 2330
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 0.21272727272727274,
|
| 1829 |
+
"grad_norm": 146619093745664.0,
|
| 1830 |
+
"learning_rate": 8.925258197991258e-07,
|
| 1831 |
+
"loss": 31.378811645507813,
|
| 1832 |
+
"step": 2340
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 0.21363636363636362,
|
| 1836 |
+
"grad_norm": 91310996324352.0,
|
| 1837 |
+
"learning_rate": 8.91639675714833e-07,
|
| 1838 |
+
"loss": 31.404779052734376,
|
| 1839 |
+
"step": 2350
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 0.21454545454545454,
|
| 1843 |
+
"grad_norm": 62529711636480.0,
|
| 1844 |
+
"learning_rate": 8.907503371462099e-07,
|
| 1845 |
+
"loss": 30.654598999023438,
|
| 1846 |
+
"step": 2360
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 0.21545454545454545,
|
| 1850 |
+
"grad_norm": 70472276377600.0,
|
| 1851 |
+
"learning_rate": 8.898578113473176e-07,
|
| 1852 |
+
"loss": 31.602188110351562,
|
| 1853 |
+
"step": 2370
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 0.21636363636363637,
|
| 1857 |
+
"grad_norm": 158925886324736.0,
|
| 1858 |
+
"learning_rate": 8.889621055982144e-07,
|
| 1859 |
+
"loss": 31.739437866210938,
|
| 1860 |
+
"step": 2380
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 0.21727272727272728,
|
| 1864 |
+
"grad_norm": 70523740487680.0,
|
| 1865 |
+
"learning_rate": 8.880632272048962e-07,
|
| 1866 |
+
"loss": 31.94945373535156,
|
| 1867 |
+
"step": 2390
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 0.21818181818181817,
|
| 1871 |
+
"grad_norm": 60489816080384.0,
|
| 1872 |
+
"learning_rate": 8.871611834992379e-07,
|
| 1873 |
+
"loss": 31.090286254882812,
|
| 1874 |
+
"step": 2400
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 0.21818181818181817,
|
| 1878 |
+
"eval_loss": 31.047500610351562,
|
| 1879 |
+
"eval_runtime": 2.2365,
|
| 1880 |
+
"eval_samples_per_second": 1.788,
|
| 1881 |
+
"eval_steps_per_second": 0.447,
|
| 1882 |
+
"step": 2400
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 0.2190909090909091,
|
| 1886 |
+
"grad_norm": 146665382084608.0,
|
| 1887 |
+
"learning_rate": 8.862559818389321e-07,
|
| 1888 |
+
"loss": 31.17398681640625,
|
| 1889 |
+
"step": 2410
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 0.22,
|
| 1893 |
+
"grad_norm": 56466086035456.0,
|
| 1894 |
+
"learning_rate": 8.853476296074305e-07,
|
| 1895 |
+
"loss": 31.538543701171875,
|
| 1896 |
+
"step": 2420
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 0.22090909090909092,
|
| 1900 |
+
"grad_norm": 205152065683456.0,
|
| 1901 |
+
"learning_rate": 8.844361342138827e-07,
|
| 1902 |
+
"loss": 31.731301879882814,
|
| 1903 |
+
"step": 2430
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 0.22181818181818183,
|
| 1907 |
+
"grad_norm": 65588269416448.0,
|
| 1908 |
+
"learning_rate": 8.835215030930761e-07,
|
| 1909 |
+
"loss": 30.90111083984375,
|
| 1910 |
+
"step": 2440
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 0.22272727272727272,
|
| 1914 |
+
"grad_norm": 34737408180224.0,
|
| 1915 |
+
"learning_rate": 8.82603743705375e-07,
|
| 1916 |
+
"loss": 31.98955078125,
|
| 1917 |
+
"step": 2450
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 0.22363636363636363,
|
| 1921 |
+
"grad_norm": 154878517182464.0,
|
| 1922 |
+
"learning_rate": 8.81682863536661e-07,
|
| 1923 |
+
"loss": 31.334310913085936,
|
| 1924 |
+
"step": 2460
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 0.22454545454545455,
|
| 1928 |
+
"grad_norm": 166664142323712.0,
|
| 1929 |
+
"learning_rate": 8.807588700982699e-07,
|
| 1930 |
+
"loss": 31.236883544921874,
|
| 1931 |
+
"step": 2470
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 0.22545454545454546,
|
| 1935 |
+
"grad_norm": 145991323877376.0,
|
| 1936 |
+
"learning_rate": 8.798317709269318e-07,
|
| 1937 |
+
"loss": 31.298379516601564,
|
| 1938 |
+
"step": 2480
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 0.22636363636363635,
|
| 1942 |
+
"grad_norm": 70725335515136.0,
|
| 1943 |
+
"learning_rate": 8.789015735847098e-07,
|
| 1944 |
+
"loss": 31.637060546875,
|
| 1945 |
+
"step": 2490
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 0.22727272727272727,
|
| 1949 |
+
"grad_norm": 164278623535104.0,
|
| 1950 |
+
"learning_rate": 8.779682856589367e-07,
|
| 1951 |
+
"loss": 31.241668701171875,
|
| 1952 |
+
"step": 2500
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 0.22727272727272727,
|
| 1956 |
+
"eval_loss": 31.05630111694336,
|
| 1957 |
+
"eval_runtime": 2.28,
|
| 1958 |
+
"eval_samples_per_second": 1.754,
|
| 1959 |
+
"eval_steps_per_second": 0.439,
|
| 1960 |
+
"step": 2500
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 0.22818181818181818,
|
| 1964 |
+
"grad_norm": 65195573510144.0,
|
| 1965 |
+
"learning_rate": 8.770319147621557e-07,
|
| 1966 |
+
"loss": 31.208535766601564,
|
| 1967 |
+
"step": 2510
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 0.2290909090909091,
|
| 1971 |
+
"grad_norm": 38720564625408.0,
|
| 1972 |
+
"learning_rate": 8.760924685320557e-07,
|
| 1973 |
+
"loss": 31.412356567382812,
|
| 1974 |
+
"step": 2520
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 0.23,
|
| 1978 |
+
"grad_norm": 20436991606784.0,
|
| 1979 |
+
"learning_rate": 8.751499546314105e-07,
|
| 1980 |
+
"loss": 31.454071044921875,
|
| 1981 |
+
"step": 2530
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 0.2309090909090909,
|
| 1985 |
+
"grad_norm": 135642474348544.0,
|
| 1986 |
+
"learning_rate": 8.742043807480162e-07,
|
| 1987 |
+
"loss": 31.459384155273437,
|
| 1988 |
+
"step": 2540
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 0.2318181818181818,
|
| 1992 |
+
"grad_norm": 44671283757056.0,
|
| 1993 |
+
"learning_rate": 8.732557545946278e-07,
|
| 1994 |
+
"loss": 31.1458251953125,
|
| 1995 |
+
"step": 2550
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.23272727272727273,
|
| 1999 |
+
"grad_norm": 86082586673152.0,
|
| 2000 |
+
"learning_rate": 8.723040839088969e-07,
|
| 2001 |
+
"loss": 30.845843505859374,
|
| 2002 |
+
"step": 2560
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 0.23363636363636364,
|
| 2006 |
+
"grad_norm": 158711691608064.0,
|
| 2007 |
+
"learning_rate": 8.713493764533088e-07,
|
| 2008 |
+
"loss": 31.084823608398438,
|
| 2009 |
+
"step": 2570
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 0.23454545454545456,
|
| 2013 |
+
"grad_norm": 57405924704256.0,
|
| 2014 |
+
"learning_rate": 8.703916400151183e-07,
|
| 2015 |
+
"loss": 30.91130676269531,
|
| 2016 |
+
"step": 2580
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 0.23545454545454544,
|
| 2020 |
+
"grad_norm": 82000975233024.0,
|
| 2021 |
+
"learning_rate": 8.694308824062867e-07,
|
| 2022 |
+
"loss": 31.061813354492188,
|
| 2023 |
+
"step": 2590
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 0.23636363636363636,
|
| 2027 |
+
"grad_norm": 79047270858752.0,
|
| 2028 |
+
"learning_rate": 8.684671114634183e-07,
|
| 2029 |
+
"loss": 31.08419189453125,
|
| 2030 |
+
"step": 2600
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 0.23636363636363636,
|
| 2034 |
+
"eval_loss": 31.05272102355957,
|
| 2035 |
+
"eval_runtime": 2.2261,
|
| 2036 |
+
"eval_samples_per_second": 1.797,
|
| 2037 |
+
"eval_steps_per_second": 0.449,
|
| 2038 |
+
"step": 2600
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 0.23727272727272727,
|
| 2042 |
+
"grad_norm": 49400906776576.0,
|
| 2043 |
+
"learning_rate": 8.675003350476961e-07,
|
| 2044 |
+
"loss": 30.686038208007812,
|
| 2045 |
+
"step": 2610
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 0.2381818181818182,
|
| 2049 |
+
"grad_norm": 94167459954688.0,
|
| 2050 |
+
"learning_rate": 8.66530561044818e-07,
|
| 2051 |
+
"loss": 31.08385314941406,
|
| 2052 |
+
"step": 2620
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 0.2390909090909091,
|
| 2056 |
+
"grad_norm": 40438371188736.0,
|
| 2057 |
+
"learning_rate": 8.655577973649321e-07,
|
| 2058 |
+
"loss": 31.391049194335938,
|
| 2059 |
+
"step": 2630
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 0.24,
|
| 2063 |
+
"grad_norm": 67886441824256.0,
|
| 2064 |
+
"learning_rate": 8.645820519425722e-07,
|
| 2065 |
+
"loss": 30.860980224609374,
|
| 2066 |
+
"step": 2640
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 0.2409090909090909,
|
| 2070 |
+
"grad_norm": 156507970732032.0,
|
| 2071 |
+
"learning_rate": 8.636033327365937e-07,
|
| 2072 |
+
"loss": 31.027520751953126,
|
| 2073 |
+
"step": 2650
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 0.24181818181818182,
|
| 2077 |
+
"grad_norm": 33284687921152.0,
|
| 2078 |
+
"learning_rate": 8.626216477301081e-07,
|
| 2079 |
+
"loss": 31.069308471679687,
|
| 2080 |
+
"step": 2660
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 0.24272727272727274,
|
| 2084 |
+
"grad_norm": 202165503131648.0,
|
| 2085 |
+
"learning_rate": 8.616370049304175e-07,
|
| 2086 |
+
"loss": 31.114059448242188,
|
| 2087 |
+
"step": 2670
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 0.24363636363636362,
|
| 2091 |
+
"grad_norm": 32541889265664.0,
|
| 2092 |
+
"learning_rate": 8.606494123689508e-07,
|
| 2093 |
+
"loss": 30.974282836914064,
|
| 2094 |
+
"step": 2680
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 0.24454545454545454,
|
| 2098 |
+
"grad_norm": 133315969417216.0,
|
| 2099 |
+
"learning_rate": 8.596588781011963e-07,
|
| 2100 |
+
"loss": 31.376223754882812,
|
| 2101 |
+
"step": 2690
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"epoch": 0.24545454545454545,
|
| 2105 |
+
"grad_norm": 120080675373056.0,
|
| 2106 |
+
"learning_rate": 8.586654102066374e-07,
|
| 2107 |
+
"loss": 31.123995971679687,
|
| 2108 |
+
"step": 2700
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 0.24545454545454545,
|
| 2112 |
+
"eval_loss": 31.06121826171875,
|
| 2113 |
+
"eval_runtime": 2.263,
|
| 2114 |
+
"eval_samples_per_second": 1.768,
|
| 2115 |
+
"eval_steps_per_second": 0.442,
|
| 2116 |
+
"step": 2700
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.24636363636363637,
|
| 2120 |
+
"grad_norm": 203449278922752.0,
|
| 2121 |
+
"learning_rate": 8.576690167886859e-07,
|
| 2122 |
+
"loss": 31.33082275390625,
|
| 2123 |
+
"step": 2710
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.24727272727272728,
|
| 2127 |
+
"grad_norm": 63677252239360.0,
|
| 2128 |
+
"learning_rate": 8.566697059746166e-07,
|
| 2129 |
+
"loss": 31.389141845703126,
|
| 2130 |
+
"step": 2720
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.24818181818181817,
|
| 2134 |
+
"grad_norm": 112063229722624.0,
|
| 2135 |
+
"learning_rate": 8.556674859155e-07,
|
| 2136 |
+
"loss": 31.25843200683594,
|
| 2137 |
+
"step": 2730
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.24909090909090909,
|
| 2141 |
+
"grad_norm": 153532078489600.0,
|
| 2142 |
+
"learning_rate": 8.546623647861369e-07,
|
| 2143 |
+
"loss": 31.412567138671875,
|
| 2144 |
+
"step": 2740
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.25,
|
| 2148 |
+
"grad_norm": 196838988709888.0,
|
| 2149 |
+
"learning_rate": 8.536543507849911e-07,
|
| 2150 |
+
"loss": 31.105191040039063,
|
| 2151 |
+
"step": 2750
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.2509090909090909,
|
| 2155 |
+
"grad_norm": 92189401022464.0,
|
| 2156 |
+
"learning_rate": 8.526434521341226e-07,
|
| 2157 |
+
"loss": 30.956512451171875,
|
| 2158 |
+
"step": 2760
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.25181818181818183,
|
| 2162 |
+
"grad_norm": 127660478955520.0,
|
| 2163 |
+
"learning_rate": 8.516296770791207e-07,
|
| 2164 |
+
"loss": 31.13167724609375,
|
| 2165 |
+
"step": 2770
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.25272727272727274,
|
| 2169 |
+
"grad_norm": 103122206720000.0,
|
| 2170 |
+
"learning_rate": 8.506130338890365e-07,
|
| 2171 |
+
"loss": 31.019964599609374,
|
| 2172 |
+
"step": 2780
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.25363636363636366,
|
| 2176 |
+
"grad_norm": 238406386843648.0,
|
| 2177 |
+
"learning_rate": 8.495935308563158e-07,
|
| 2178 |
+
"loss": 31.423956298828124,
|
| 2179 |
+
"step": 2790
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.2545454545454545,
|
| 2183 |
+
"grad_norm": 134973164093440.0,
|
| 2184 |
+
"learning_rate": 8.485711762967312e-07,
|
| 2185 |
+
"loss": 30.945632934570312,
|
| 2186 |
+
"step": 2800
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.2545454545454545,
|
| 2190 |
+
"eval_loss": 31.074186325073242,
|
| 2191 |
+
"eval_runtime": 2.2625,
|
| 2192 |
+
"eval_samples_per_second": 1.768,
|
| 2193 |
+
"eval_steps_per_second": 0.442,
|
| 2194 |
+
"step": 2800
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 0.25545454545454543,
|
| 2198 |
+
"grad_norm": 50747676819456.0,
|
| 2199 |
+
"learning_rate": 8.47545978549314e-07,
|
| 2200 |
+
"loss": 31.677044677734376,
|
| 2201 |
+
"step": 2810
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.25636363636363635,
|
| 2205 |
+
"grad_norm": 138435553656832.0,
|
| 2206 |
+
"learning_rate": 8.46517945976287e-07,
|
| 2207 |
+
"loss": 31.338131713867188,
|
| 2208 |
+
"step": 2820
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.25727272727272726,
|
| 2212 |
+
"grad_norm": 152003741220864.0,
|
| 2213 |
+
"learning_rate": 8.454870869629955e-07,
|
| 2214 |
+
"loss": 30.898031616210936,
|
| 2215 |
+
"step": 2830
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.2581818181818182,
|
| 2219 |
+
"grad_norm": 159471129067520.0,
|
| 2220 |
+
"learning_rate": 8.444534099178393e-07,
|
| 2221 |
+
"loss": 30.950164794921875,
|
| 2222 |
+
"step": 2840
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.2590909090909091,
|
| 2226 |
+
"grad_norm": 141189164564480.0,
|
| 2227 |
+
"learning_rate": 8.434169232722041e-07,
|
| 2228 |
+
"loss": 31.282101440429688,
|
| 2229 |
+
"step": 2850
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.26,
|
| 2233 |
+
"grad_norm": 130600384069632.0,
|
| 2234 |
+
"learning_rate": 8.423776354803925e-07,
|
| 2235 |
+
"loss": 31.540997314453126,
|
| 2236 |
+
"step": 2860
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 0.2609090909090909,
|
| 2240 |
+
"grad_norm": 199170954625024.0,
|
| 2241 |
+
"learning_rate": 8.41335555019555e-07,
|
| 2242 |
+
"loss": 31.349322509765624,
|
| 2243 |
+
"step": 2870
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.26181818181818184,
|
| 2247 |
+
"grad_norm": 208188993437696.0,
|
| 2248 |
+
"learning_rate": 8.402906903896216e-07,
|
| 2249 |
+
"loss": 30.875155639648437,
|
| 2250 |
+
"step": 2880
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 0.26272727272727275,
|
| 2254 |
+
"grad_norm": 125149995073536.0,
|
| 2255 |
+
"learning_rate": 8.392430501132316e-07,
|
| 2256 |
+
"loss": 31.10132751464844,
|
| 2257 |
+
"step": 2890
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.2636363636363636,
|
| 2261 |
+
"grad_norm": 47845277171712.0,
|
| 2262 |
+
"learning_rate": 8.381926427356642e-07,
|
| 2263 |
+
"loss": 31.629965209960936,
|
| 2264 |
+
"step": 2900
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 0.2636363636363636,
|
| 2268 |
+
"eval_loss": 31.083770751953125,
|
| 2269 |
+
"eval_runtime": 2.258,
|
| 2270 |
+
"eval_samples_per_second": 1.771,
|
| 2271 |
+
"eval_steps_per_second": 0.443,
|
| 2272 |
+
"step": 2900
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 0.26454545454545453,
|
| 2276 |
+
"grad_norm": 114591514230784.0,
|
| 2277 |
+
"learning_rate": 8.37139476824769e-07,
|
| 2278 |
+
"loss": 30.9484130859375,
|
| 2279 |
+
"step": 2910
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 0.26545454545454544,
|
| 2283 |
+
"grad_norm": 76647743094784.0,
|
| 2284 |
+
"learning_rate": 8.360835609708967e-07,
|
| 2285 |
+
"loss": 31.01370849609375,
|
| 2286 |
+
"step": 2920
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 0.26636363636363636,
|
| 2290 |
+
"grad_norm": 190012658286592.0,
|
| 2291 |
+
"learning_rate": 8.35024903786828e-07,
|
| 2292 |
+
"loss": 31.296466064453124,
|
| 2293 |
+
"step": 2930
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 0.2672727272727273,
|
| 2297 |
+
"grad_norm": 152463738929152.0,
|
| 2298 |
+
"learning_rate": 8.339635139077035e-07,
|
| 2299 |
+
"loss": 31.273577880859374,
|
| 2300 |
+
"step": 2940
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 0.2681818181818182,
|
| 2304 |
+
"grad_norm": 216430482030592.0,
|
| 2305 |
+
"learning_rate": 8.32899399990954e-07,
|
| 2306 |
+
"loss": 31.327499389648438,
|
| 2307 |
+
"step": 2950
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 0.2690909090909091,
|
| 2311 |
+
"grad_norm": 48228892409856.0,
|
| 2312 |
+
"learning_rate": 8.318325707162293e-07,
|
| 2313 |
+
"loss": 31.284075927734374,
|
| 2314 |
+
"step": 2960
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 0.27,
|
| 2318 |
+
"grad_norm": 109718655729664.0,
|
| 2319 |
+
"learning_rate": 8.307630347853273e-07,
|
| 2320 |
+
"loss": 31.343002319335938,
|
| 2321 |
+
"step": 2970
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 0.27090909090909093,
|
| 2325 |
+
"grad_norm": 102808120459264.0,
|
| 2326 |
+
"learning_rate": 8.296908009221242e-07,
|
| 2327 |
+
"loss": 31.049209594726562,
|
| 2328 |
+
"step": 2980
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 0.2718181818181818,
|
| 2332 |
+
"grad_norm": 69456642441216.0,
|
| 2333 |
+
"learning_rate": 8.286158778725011e-07,
|
| 2334 |
+
"loss": 31.43966064453125,
|
| 2335 |
+
"step": 2990
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 0.2727272727272727,
|
| 2339 |
+
"grad_norm": 92835936206848.0,
|
| 2340 |
+
"learning_rate": 8.275382744042747e-07,
|
| 2341 |
+
"loss": 30.909332275390625,
|
| 2342 |
+
"step": 3000
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 0.2727272727272727,
|
| 2346 |
+
"eval_loss": 31.07646369934082,
|
| 2347 |
+
"eval_runtime": 2.242,
|
| 2348 |
+
"eval_samples_per_second": 1.784,
|
| 2349 |
+
"eval_steps_per_second": 0.446,
|
| 2350 |
+
"step": 3000
|
| 2351 |
}
|
| 2352 |
],
|
| 2353 |
"logging_steps": 10,
|
|
|
|
| 2367 |
"attributes": {}
|
| 2368 |
}
|
| 2369 |
},
|
| 2370 |
+
"total_flos": 1.3589771452416e+16,
|
| 2371 |
"train_batch_size": 1,
|
| 2372 |
"trial_name": null,
|
| 2373 |
"trial_params": null
|