Training in progress, step 1500, 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 +711 -3
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 791869518
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f95abb4f35b22a7e758afd27aca2a69f419698ad1276e97da477eb487b8e37f1
|
| 3 |
size 791869518
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2375752250
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:048bbd1cabc0a45ad09ea3651c3b6e3d5d885845f3d358d903b243b17193ba3a
|
| 3 |
size 2375752250
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0c461c9d337dfc684e9352ec72bfa344e2f5d377f7cfc4475de9acae294dca89
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fae392ec6232cbf9da21d6ed12bc8247d0d24e7f3a3606acd23be00f3e8bbfc5
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cbf3e7ca9991a58b0b16574a3c653483c551c270aa05aba06c162ea593f7b0f2
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c695bebf6bcb75cbe26378bfe0ab7e2a33c49f713b9d6e4d10632b24322977e7
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5ebb13c71265c5464c9aa9bb9b66f07764d73befe6cd63a2aaf8e781bf0a374
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12cc6e245e189be568c8dfd43a4dd8f04bb3dbd9f17f41458107935d2c2a6a9d
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:36086646e9a8f76fea69f8a227112e83bb63524964ccdfb82f4cdad88b90e5e4
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b44153bacf860d0ca6ce4c6b9380a199feab8a72ca613e6745bfb671b02c4e4
|
| 3 |
size 15984
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1000
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c784d1e14f175eecd1cb8a33bf7e3edbddb5399a3760000ee27c1d7309b565d
|
| 3 |
size 1000
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1423,6 +1423,714 @@
|
|
| 1423 |
"eval_samples_per_second": 609.845,
|
| 1424 |
"eval_steps_per_second": 38.116,
|
| 1425 |
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1426 |
}
|
| 1427 |
],
|
| 1428 |
"logging_steps": 5,
|
|
@@ -1442,7 +2150,7 @@
|
|
| 1442 |
"attributes": {}
|
| 1443 |
}
|
| 1444 |
},
|
| 1445 |
-
"total_flos":
|
| 1446 |
"train_batch_size": 4,
|
| 1447 |
"trial_name": null,
|
| 1448 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.34676318241910664,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1423 |
"eval_samples_per_second": 609.845,
|
| 1424 |
"eval_steps_per_second": 38.116,
|
| 1425 |
"step": 1000
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 0.23233133222080146,
|
| 1429 |
+
"grad_norm": 166.75,
|
| 1430 |
+
"learning_rate": 8.530318602261049e-06,
|
| 1431 |
+
"loss": 72.1535,
|
| 1432 |
+
"step": 1005
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 0.2334872094955318,
|
| 1436 |
+
"grad_norm": 178.5,
|
| 1437 |
+
"learning_rate": 8.517471736896197e-06,
|
| 1438 |
+
"loss": 74.0213,
|
| 1439 |
+
"step": 1010
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 0.23464308677026216,
|
| 1443 |
+
"grad_norm": 156.25,
|
| 1444 |
+
"learning_rate": 8.504624871531348e-06,
|
| 1445 |
+
"loss": 74.9422,
|
| 1446 |
+
"step": 1015
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 0.2357989640449925,
|
| 1450 |
+
"grad_norm": 174.0,
|
| 1451 |
+
"learning_rate": 8.491778006166495e-06,
|
| 1452 |
+
"loss": 72.5082,
|
| 1453 |
+
"step": 1020
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 0.2369548413197229,
|
| 1457 |
+
"grad_norm": 162.5,
|
| 1458 |
+
"learning_rate": 8.478931140801646e-06,
|
| 1459 |
+
"loss": 73.302,
|
| 1460 |
+
"step": 1025
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 0.23811071859445324,
|
| 1464 |
+
"grad_norm": 182.625,
|
| 1465 |
+
"learning_rate": 8.466084275436795e-06,
|
| 1466 |
+
"loss": 73.0291,
|
| 1467 |
+
"step": 1030
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 0.23926659586918358,
|
| 1471 |
+
"grad_norm": 162.875,
|
| 1472 |
+
"learning_rate": 8.453237410071943e-06,
|
| 1473 |
+
"loss": 73.237,
|
| 1474 |
+
"step": 1035
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 0.24042247314391393,
|
| 1478 |
+
"grad_norm": 181.75,
|
| 1479 |
+
"learning_rate": 8.440390544707092e-06,
|
| 1480 |
+
"loss": 73.0543,
|
| 1481 |
+
"step": 1040
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 0.2415783504186443,
|
| 1485 |
+
"grad_norm": 194.25,
|
| 1486 |
+
"learning_rate": 8.427543679342241e-06,
|
| 1487 |
+
"loss": 72.1511,
|
| 1488 |
+
"step": 1045
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 0.24273422769337466,
|
| 1492 |
+
"grad_norm": 197.625,
|
| 1493 |
+
"learning_rate": 8.41469681397739e-06,
|
| 1494 |
+
"loss": 72.7044,
|
| 1495 |
+
"step": 1050
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 0.243890104968105,
|
| 1499 |
+
"grad_norm": 205.75,
|
| 1500 |
+
"learning_rate": 8.401849948612539e-06,
|
| 1501 |
+
"loss": 73.4772,
|
| 1502 |
+
"step": 1055
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 0.24504598224283536,
|
| 1506 |
+
"grad_norm": 179.375,
|
| 1507 |
+
"learning_rate": 8.38900308324769e-06,
|
| 1508 |
+
"loss": 73.4765,
|
| 1509 |
+
"step": 1060
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 0.24620185951756574,
|
| 1513 |
+
"grad_norm": 163.25,
|
| 1514 |
+
"learning_rate": 8.376156217882836e-06,
|
| 1515 |
+
"loss": 71.4076,
|
| 1516 |
+
"step": 1065
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 0.24735773679229608,
|
| 1520 |
+
"grad_norm": 159.75,
|
| 1521 |
+
"learning_rate": 8.363309352517987e-06,
|
| 1522 |
+
"loss": 74.3268,
|
| 1523 |
+
"step": 1070
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 0.24851361406702643,
|
| 1527 |
+
"grad_norm": 174.625,
|
| 1528 |
+
"learning_rate": 8.350462487153134e-06,
|
| 1529 |
+
"loss": 71.9133,
|
| 1530 |
+
"step": 1075
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 0.24966949134175678,
|
| 1534 |
+
"grad_norm": 224.125,
|
| 1535 |
+
"learning_rate": 8.337615621788285e-06,
|
| 1536 |
+
"loss": 72.5681,
|
| 1537 |
+
"step": 1080
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 0.25082536861648713,
|
| 1541 |
+
"grad_norm": 196.25,
|
| 1542 |
+
"learning_rate": 8.324768756423434e-06,
|
| 1543 |
+
"loss": 72.7169,
|
| 1544 |
+
"step": 1085
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 0.2519812458912175,
|
| 1548 |
+
"grad_norm": 163.5,
|
| 1549 |
+
"learning_rate": 8.311921891058582e-06,
|
| 1550 |
+
"loss": 73.2148,
|
| 1551 |
+
"step": 1090
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 0.25313712316594783,
|
| 1555 |
+
"grad_norm": 181.5,
|
| 1556 |
+
"learning_rate": 8.299075025693731e-06,
|
| 1557 |
+
"loss": 74.4619,
|
| 1558 |
+
"step": 1095
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 0.2542930004406782,
|
| 1562 |
+
"grad_norm": 201.125,
|
| 1563 |
+
"learning_rate": 8.28622816032888e-06,
|
| 1564 |
+
"loss": 72.7498,
|
| 1565 |
+
"step": 1100
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 0.2554488777154086,
|
| 1569 |
+
"grad_norm": 191.75,
|
| 1570 |
+
"learning_rate": 8.273381294964029e-06,
|
| 1571 |
+
"loss": 73.7849,
|
| 1572 |
+
"step": 1105
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 0.2566047549901389,
|
| 1576 |
+
"grad_norm": 187.375,
|
| 1577 |
+
"learning_rate": 8.260534429599178e-06,
|
| 1578 |
+
"loss": 73.536,
|
| 1579 |
+
"step": 1110
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 0.2577606322648693,
|
| 1583 |
+
"grad_norm": 161.25,
|
| 1584 |
+
"learning_rate": 8.247687564234328e-06,
|
| 1585 |
+
"loss": 73.2744,
|
| 1586 |
+
"step": 1115
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.25891650953959966,
|
| 1590 |
+
"grad_norm": 170.875,
|
| 1591 |
+
"learning_rate": 8.234840698869475e-06,
|
| 1592 |
+
"loss": 74.2862,
|
| 1593 |
+
"step": 1120
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 0.26007238681433,
|
| 1597 |
+
"grad_norm": 175.375,
|
| 1598 |
+
"learning_rate": 8.221993833504626e-06,
|
| 1599 |
+
"loss": 73.767,
|
| 1600 |
+
"step": 1125
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 0.26122826408906036,
|
| 1604 |
+
"grad_norm": 175.25,
|
| 1605 |
+
"learning_rate": 8.209146968139775e-06,
|
| 1606 |
+
"loss": 72.3037,
|
| 1607 |
+
"step": 1130
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 0.2623841413637907,
|
| 1611 |
+
"grad_norm": 180.125,
|
| 1612 |
+
"learning_rate": 8.196300102774924e-06,
|
| 1613 |
+
"loss": 73.3881,
|
| 1614 |
+
"step": 1135
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 0.26354001863852106,
|
| 1618 |
+
"grad_norm": 165.25,
|
| 1619 |
+
"learning_rate": 8.183453237410073e-06,
|
| 1620 |
+
"loss": 74.2604,
|
| 1621 |
+
"step": 1140
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 0.26469589591325143,
|
| 1625 |
+
"grad_norm": 208.875,
|
| 1626 |
+
"learning_rate": 8.170606372045221e-06,
|
| 1627 |
+
"loss": 73.6266,
|
| 1628 |
+
"step": 1145
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 0.26585177318798175,
|
| 1632 |
+
"grad_norm": 164.375,
|
| 1633 |
+
"learning_rate": 8.15775950668037e-06,
|
| 1634 |
+
"loss": 71.7306,
|
| 1635 |
+
"step": 1150
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 0.26700765046271213,
|
| 1639 |
+
"grad_norm": 180.5,
|
| 1640 |
+
"learning_rate": 8.14491264131552e-06,
|
| 1641 |
+
"loss": 73.3496,
|
| 1642 |
+
"step": 1155
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 0.26816352773744245,
|
| 1646 |
+
"grad_norm": 161.75,
|
| 1647 |
+
"learning_rate": 8.132065775950668e-06,
|
| 1648 |
+
"loss": 71.9542,
|
| 1649 |
+
"step": 1160
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 0.26931940501217283,
|
| 1653 |
+
"grad_norm": 160.375,
|
| 1654 |
+
"learning_rate": 8.119218910585818e-06,
|
| 1655 |
+
"loss": 72.852,
|
| 1656 |
+
"step": 1165
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 0.2704752822869032,
|
| 1660 |
+
"grad_norm": 166.875,
|
| 1661 |
+
"learning_rate": 8.106372045220967e-06,
|
| 1662 |
+
"loss": 73.498,
|
| 1663 |
+
"step": 1170
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 0.27163115956163353,
|
| 1667 |
+
"grad_norm": 152.5,
|
| 1668 |
+
"learning_rate": 8.093525179856116e-06,
|
| 1669 |
+
"loss": 72.89,
|
| 1670 |
+
"step": 1175
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 0.2727870368363639,
|
| 1674 |
+
"grad_norm": 179.625,
|
| 1675 |
+
"learning_rate": 8.080678314491265e-06,
|
| 1676 |
+
"loss": 73.7572,
|
| 1677 |
+
"step": 1180
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 0.2739429141110943,
|
| 1681 |
+
"grad_norm": 178.625,
|
| 1682 |
+
"learning_rate": 8.067831449126414e-06,
|
| 1683 |
+
"loss": 72.1453,
|
| 1684 |
+
"step": 1185
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 0.2750987913858246,
|
| 1688 |
+
"grad_norm": 165.0,
|
| 1689 |
+
"learning_rate": 8.054984583761563e-06,
|
| 1690 |
+
"loss": 73.9004,
|
| 1691 |
+
"step": 1190
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 0.276254668660555,
|
| 1695 |
+
"grad_norm": 170.875,
|
| 1696 |
+
"learning_rate": 8.042137718396712e-06,
|
| 1697 |
+
"loss": 72.7976,
|
| 1698 |
+
"step": 1195
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 0.2774105459352853,
|
| 1702 |
+
"grad_norm": 163.125,
|
| 1703 |
+
"learning_rate": 8.029290853031862e-06,
|
| 1704 |
+
"loss": 73.1718,
|
| 1705 |
+
"step": 1200
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 0.2785664232100157,
|
| 1709 |
+
"grad_norm": 179.25,
|
| 1710 |
+
"learning_rate": 8.01644398766701e-06,
|
| 1711 |
+
"loss": 72.7056,
|
| 1712 |
+
"step": 1205
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 0.27972230048474606,
|
| 1716 |
+
"grad_norm": 190.625,
|
| 1717 |
+
"learning_rate": 8.00359712230216e-06,
|
| 1718 |
+
"loss": 73.2504,
|
| 1719 |
+
"step": 1210
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 0.2808781777594764,
|
| 1723 |
+
"grad_norm": 171.375,
|
| 1724 |
+
"learning_rate": 7.990750256937307e-06,
|
| 1725 |
+
"loss": 72.886,
|
| 1726 |
+
"step": 1215
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 0.28203405503420675,
|
| 1730 |
+
"grad_norm": 168.25,
|
| 1731 |
+
"learning_rate": 7.977903391572457e-06,
|
| 1732 |
+
"loss": 73.0636,
|
| 1733 |
+
"step": 1220
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.2831899323089371,
|
| 1737 |
+
"grad_norm": 170.875,
|
| 1738 |
+
"learning_rate": 7.965056526207606e-06,
|
| 1739 |
+
"loss": 73.4683,
|
| 1740 |
+
"step": 1225
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 0.28434580958366745,
|
| 1744 |
+
"grad_norm": 170.375,
|
| 1745 |
+
"learning_rate": 7.952209660842755e-06,
|
| 1746 |
+
"loss": 72.7721,
|
| 1747 |
+
"step": 1230
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 0.28550168685839783,
|
| 1751 |
+
"grad_norm": 187.875,
|
| 1752 |
+
"learning_rate": 7.939362795477904e-06,
|
| 1753 |
+
"loss": 74.0967,
|
| 1754 |
+
"step": 1235
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 0.28665756413312815,
|
| 1758 |
+
"grad_norm": 206.0,
|
| 1759 |
+
"learning_rate": 7.926515930113053e-06,
|
| 1760 |
+
"loss": 72.2294,
|
| 1761 |
+
"step": 1240
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 0.28781344140785853,
|
| 1765 |
+
"grad_norm": 174.625,
|
| 1766 |
+
"learning_rate": 7.913669064748202e-06,
|
| 1767 |
+
"loss": 71.7166,
|
| 1768 |
+
"step": 1245
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.28896931868258885,
|
| 1772 |
+
"grad_norm": 195.0,
|
| 1773 |
+
"learning_rate": 7.90082219938335e-06,
|
| 1774 |
+
"loss": 72.4936,
|
| 1775 |
+
"step": 1250
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 0.2901251959573192,
|
| 1779 |
+
"grad_norm": 174.25,
|
| 1780 |
+
"learning_rate": 7.887975334018501e-06,
|
| 1781 |
+
"loss": 71.8098,
|
| 1782 |
+
"step": 1255
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 0.2912810732320496,
|
| 1786 |
+
"grad_norm": 162.25,
|
| 1787 |
+
"learning_rate": 7.875128468653648e-06,
|
| 1788 |
+
"loss": 73.6741,
|
| 1789 |
+
"step": 1260
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 0.2924369505067799,
|
| 1793 |
+
"grad_norm": 163.0,
|
| 1794 |
+
"learning_rate": 7.862281603288799e-06,
|
| 1795 |
+
"loss": 74.2419,
|
| 1796 |
+
"step": 1265
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.2935928277815103,
|
| 1800 |
+
"grad_norm": 156.625,
|
| 1801 |
+
"learning_rate": 7.849434737923948e-06,
|
| 1802 |
+
"loss": 71.8441,
|
| 1803 |
+
"step": 1270
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.2947487050562407,
|
| 1807 |
+
"grad_norm": 159.125,
|
| 1808 |
+
"learning_rate": 7.836587872559096e-06,
|
| 1809 |
+
"loss": 72.4638,
|
| 1810 |
+
"step": 1275
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 0.295904582330971,
|
| 1814 |
+
"grad_norm": 163.75,
|
| 1815 |
+
"learning_rate": 7.823741007194245e-06,
|
| 1816 |
+
"loss": 72.4716,
|
| 1817 |
+
"step": 1280
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 0.2970604596057014,
|
| 1821 |
+
"grad_norm": 149.25,
|
| 1822 |
+
"learning_rate": 7.810894141829394e-06,
|
| 1823 |
+
"loss": 71.0848,
|
| 1824 |
+
"step": 1285
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 0.2982163368804317,
|
| 1828 |
+
"grad_norm": 157.625,
|
| 1829 |
+
"learning_rate": 7.798047276464543e-06,
|
| 1830 |
+
"loss": 71.6886,
|
| 1831 |
+
"step": 1290
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 0.2993722141551621,
|
| 1835 |
+
"grad_norm": 161.125,
|
| 1836 |
+
"learning_rate": 7.785200411099692e-06,
|
| 1837 |
+
"loss": 72.6646,
|
| 1838 |
+
"step": 1295
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 0.30052809142989245,
|
| 1842 |
+
"grad_norm": 189.25,
|
| 1843 |
+
"learning_rate": 7.77235354573484e-06,
|
| 1844 |
+
"loss": 72.4807,
|
| 1845 |
+
"step": 1300
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 0.3016839687046228,
|
| 1849 |
+
"grad_norm": 207.0,
|
| 1850 |
+
"learning_rate": 7.75950668036999e-06,
|
| 1851 |
+
"loss": 72.7943,
|
| 1852 |
+
"step": 1305
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 0.30283984597935315,
|
| 1856 |
+
"grad_norm": 160.625,
|
| 1857 |
+
"learning_rate": 7.74665981500514e-06,
|
| 1858 |
+
"loss": 72.9901,
|
| 1859 |
+
"step": 1310
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.30399572325408347,
|
| 1863 |
+
"grad_norm": 218.375,
|
| 1864 |
+
"learning_rate": 7.733812949640287e-06,
|
| 1865 |
+
"loss": 73.4908,
|
| 1866 |
+
"step": 1315
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 0.30515160052881385,
|
| 1870 |
+
"grad_norm": 166.125,
|
| 1871 |
+
"learning_rate": 7.720966084275438e-06,
|
| 1872 |
+
"loss": 72.9456,
|
| 1873 |
+
"step": 1320
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 0.3063074778035442,
|
| 1877 |
+
"grad_norm": 171.0,
|
| 1878 |
+
"learning_rate": 7.708119218910587e-06,
|
| 1879 |
+
"loss": 72.6809,
|
| 1880 |
+
"step": 1325
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 0.30746335507827455,
|
| 1884 |
+
"grad_norm": 162.75,
|
| 1885 |
+
"learning_rate": 7.695272353545736e-06,
|
| 1886 |
+
"loss": 71.7408,
|
| 1887 |
+
"step": 1330
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 0.3086192323530049,
|
| 1891 |
+
"grad_norm": 155.875,
|
| 1892 |
+
"learning_rate": 7.682425488180884e-06,
|
| 1893 |
+
"loss": 73.5375,
|
| 1894 |
+
"step": 1335
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 0.3097751096277353,
|
| 1898 |
+
"grad_norm": 155.625,
|
| 1899 |
+
"learning_rate": 7.669578622816033e-06,
|
| 1900 |
+
"loss": 72.8354,
|
| 1901 |
+
"step": 1340
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 0.3109309869024656,
|
| 1905 |
+
"grad_norm": 176.75,
|
| 1906 |
+
"learning_rate": 7.656731757451182e-06,
|
| 1907 |
+
"loss": 72.5769,
|
| 1908 |
+
"step": 1345
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 0.312086864177196,
|
| 1912 |
+
"grad_norm": 187.75,
|
| 1913 |
+
"learning_rate": 7.643884892086333e-06,
|
| 1914 |
+
"loss": 72.8171,
|
| 1915 |
+
"step": 1350
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 0.3132427414519263,
|
| 1919 |
+
"grad_norm": 159.25,
|
| 1920 |
+
"learning_rate": 7.63103802672148e-06,
|
| 1921 |
+
"loss": 72.2052,
|
| 1922 |
+
"step": 1355
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 0.3143986187266567,
|
| 1926 |
+
"grad_norm": 163.125,
|
| 1927 |
+
"learning_rate": 7.6181911613566294e-06,
|
| 1928 |
+
"loss": 71.9153,
|
| 1929 |
+
"step": 1360
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 0.3155544960013871,
|
| 1933 |
+
"grad_norm": 171.375,
|
| 1934 |
+
"learning_rate": 7.605344295991779e-06,
|
| 1935 |
+
"loss": 70.9257,
|
| 1936 |
+
"step": 1365
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 0.3167103732761174,
|
| 1940 |
+
"grad_norm": 180.0,
|
| 1941 |
+
"learning_rate": 7.592497430626927e-06,
|
| 1942 |
+
"loss": 72.8817,
|
| 1943 |
+
"step": 1370
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 0.3178662505508478,
|
| 1947 |
+
"grad_norm": 173.625,
|
| 1948 |
+
"learning_rate": 7.579650565262077e-06,
|
| 1949 |
+
"loss": 72.5966,
|
| 1950 |
+
"step": 1375
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 0.3190221278255781,
|
| 1954 |
+
"grad_norm": 160.125,
|
| 1955 |
+
"learning_rate": 7.566803699897226e-06,
|
| 1956 |
+
"loss": 72.0935,
|
| 1957 |
+
"step": 1380
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 0.32017800510030847,
|
| 1961 |
+
"grad_norm": 158.875,
|
| 1962 |
+
"learning_rate": 7.5539568345323745e-06,
|
| 1963 |
+
"loss": 72.1024,
|
| 1964 |
+
"step": 1385
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 0.32133388237503885,
|
| 1968 |
+
"grad_norm": 155.375,
|
| 1969 |
+
"learning_rate": 7.541109969167524e-06,
|
| 1970 |
+
"loss": 71.8397,
|
| 1971 |
+
"step": 1390
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 0.32248975964976917,
|
| 1975 |
+
"grad_norm": 165.5,
|
| 1976 |
+
"learning_rate": 7.528263103802673e-06,
|
| 1977 |
+
"loss": 71.5119,
|
| 1978 |
+
"step": 1395
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 0.32364563692449955,
|
| 1982 |
+
"grad_norm": 170.875,
|
| 1983 |
+
"learning_rate": 7.515416238437822e-06,
|
| 1984 |
+
"loss": 73.3594,
|
| 1985 |
+
"step": 1400
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 0.3248015141992299,
|
| 1989 |
+
"grad_norm": 184.125,
|
| 1990 |
+
"learning_rate": 7.502569373072971e-06,
|
| 1991 |
+
"loss": 73.504,
|
| 1992 |
+
"step": 1405
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 0.32595739147396025,
|
| 1996 |
+
"grad_norm": 165.25,
|
| 1997 |
+
"learning_rate": 7.4897225077081205e-06,
|
| 1998 |
+
"loss": 71.1564,
|
| 1999 |
+
"step": 1410
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 0.3271132687486906,
|
| 2003 |
+
"grad_norm": 175.5,
|
| 2004 |
+
"learning_rate": 7.4768756423432685e-06,
|
| 2005 |
+
"loss": 72.8789,
|
| 2006 |
+
"step": 1415
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 0.32826914602342094,
|
| 2010 |
+
"grad_norm": 168.375,
|
| 2011 |
+
"learning_rate": 7.464028776978418e-06,
|
| 2012 |
+
"loss": 73.2979,
|
| 2013 |
+
"step": 1420
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 0.3294250232981513,
|
| 2017 |
+
"grad_norm": 177.5,
|
| 2018 |
+
"learning_rate": 7.451181911613566e-06,
|
| 2019 |
+
"loss": 71.187,
|
| 2020 |
+
"step": 1425
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 0.3305809005728817,
|
| 2024 |
+
"grad_norm": 160.375,
|
| 2025 |
+
"learning_rate": 7.438335046248716e-06,
|
| 2026 |
+
"loss": 73.9314,
|
| 2027 |
+
"step": 1430
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 0.331736777847612,
|
| 2031 |
+
"grad_norm": 167.625,
|
| 2032 |
+
"learning_rate": 7.4254881808838655e-06,
|
| 2033 |
+
"loss": 72.4073,
|
| 2034 |
+
"step": 1435
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 0.3328926551223424,
|
| 2038 |
+
"grad_norm": 167.0,
|
| 2039 |
+
"learning_rate": 7.4126413155190135e-06,
|
| 2040 |
+
"loss": 72.5102,
|
| 2041 |
+
"step": 1440
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 0.3340485323970727,
|
| 2045 |
+
"grad_norm": 173.5,
|
| 2046 |
+
"learning_rate": 7.399794450154163e-06,
|
| 2047 |
+
"loss": 72.4247,
|
| 2048 |
+
"step": 1445
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 0.3352044096718031,
|
| 2052 |
+
"grad_norm": 169.125,
|
| 2053 |
+
"learning_rate": 7.386947584789312e-06,
|
| 2054 |
+
"loss": 71.9371,
|
| 2055 |
+
"step": 1450
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 0.33636028694653347,
|
| 2059 |
+
"grad_norm": 157.375,
|
| 2060 |
+
"learning_rate": 7.374100719424461e-06,
|
| 2061 |
+
"loss": 71.7102,
|
| 2062 |
+
"step": 1455
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 0.3375161642212638,
|
| 2066 |
+
"grad_norm": 173.625,
|
| 2067 |
+
"learning_rate": 7.36125385405961e-06,
|
| 2068 |
+
"loss": 72.5739,
|
| 2069 |
+
"step": 1460
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 0.33867204149599417,
|
| 2073 |
+
"grad_norm": 196.125,
|
| 2074 |
+
"learning_rate": 7.3484069886947595e-06,
|
| 2075 |
+
"loss": 72.1263,
|
| 2076 |
+
"step": 1465
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 0.3398279187707245,
|
| 2080 |
+
"grad_norm": 154.125,
|
| 2081 |
+
"learning_rate": 7.3355601233299075e-06,
|
| 2082 |
+
"loss": 72.4053,
|
| 2083 |
+
"step": 1470
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 0.34098379604545487,
|
| 2087 |
+
"grad_norm": 182.125,
|
| 2088 |
+
"learning_rate": 7.322713257965057e-06,
|
| 2089 |
+
"loss": 72.6574,
|
| 2090 |
+
"step": 1475
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 0.34213967332018524,
|
| 2094 |
+
"grad_norm": 182.375,
|
| 2095 |
+
"learning_rate": 7.309866392600207e-06,
|
| 2096 |
+
"loss": 72.0569,
|
| 2097 |
+
"step": 1480
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 0.34329555059491557,
|
| 2101 |
+
"grad_norm": 161.375,
|
| 2102 |
+
"learning_rate": 7.297019527235355e-06,
|
| 2103 |
+
"loss": 72.4482,
|
| 2104 |
+
"step": 1485
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 0.34445142786964594,
|
| 2108 |
+
"grad_norm": 165.25,
|
| 2109 |
+
"learning_rate": 7.2841726618705045e-06,
|
| 2110 |
+
"loss": 73.3294,
|
| 2111 |
+
"step": 1490
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 0.3456073051443763,
|
| 2115 |
+
"grad_norm": 162.625,
|
| 2116 |
+
"learning_rate": 7.2713257965056525e-06,
|
| 2117 |
+
"loss": 72.414,
|
| 2118 |
+
"step": 1495
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 0.34676318241910664,
|
| 2122 |
+
"grad_norm": 165.875,
|
| 2123 |
+
"learning_rate": 7.258478931140802e-06,
|
| 2124 |
+
"loss": 70.1396,
|
| 2125 |
+
"step": 1500
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 0.34676318241910664,
|
| 2129 |
+
"eval_loss": NaN,
|
| 2130 |
+
"eval_runtime": 382.2578,
|
| 2131 |
+
"eval_samples_per_second": 609.887,
|
| 2132 |
+
"eval_steps_per_second": 38.118,
|
| 2133 |
+
"step": 1500
|
| 2134 |
}
|
| 2135 |
],
|
| 2136 |
"logging_steps": 5,
|
|
|
|
| 2150 |
"attributes": {}
|
| 2151 |
}
|
| 2152 |
},
|
| 2153 |
+
"total_flos": 6.498536989183181e+18,
|
| 2154 |
"train_batch_size": 4,
|
| 2155 |
"trial_name": null,
|
| 2156 |
"trial_params": null
|