Training in progress, step 300, checkpoint
Browse files- last-checkpoint/adapter_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 +713 -5
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 341141032
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e6fd325c9c8f4d7dde06717edc7511070769418008c28d00119c090d6f0b7f68
|
| 3 |
size 341141032
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 682673458
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1348db24f3e478f81922e312fe908cb739c4ee9e150ad2f036ab69be712443a9
|
| 3 |
size 682673458
|
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 16048
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8829ed70b2030302656436ad3ce9c06239dc272781dfeb2182af49dba9382ced
|
| 3 |
size 16048
|
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 16048
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:07ca875b62a1014d462f3dfbedd9672aacac54f278aad3dbfa56502e6d8980ab
|
| 3 |
size 16048
|
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 16048
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0c68c3fb6c675fb82352432cbb00106f3cc3d144b27bc31460fa7671329b310e
|
| 3 |
size 16048
|
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 16048
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:995034fad8cd98fd8cc7b08e9388f5c2c3ffc6f5ffc8b38220e45ff9f33fe319
|
| 3 |
size 16048
|
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 16048
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6ff4478e39984dc2091dfafc378a1bae8f663f5ae6eb146d15efe3a27b8e0984
|
| 3 |
size 16048
|
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 16048
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac6cbcc62b40b33b44f268770d73327c7327607c19cfdd79854e8fb00ff2459d
|
| 3 |
size 16048
|
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 16048
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fd4125a97cf9417ec9195cefebfe2739888b6110a9ec0efda3deecf9d22e99e1
|
| 3 |
size 16048
|
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 16048
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5f72bd0d8d41bcc931626834be67ca7ce72d528b57449f01d7b596f92453186a
|
| 3 |
size 16048
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fa3cb9fdd7c70b3d8d2084f26b1ba813283d06e83c4489733dc372cf0bfcf36f
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": 200,
|
| 3 |
"best_metric": 0.09608737379312515,
|
| 4 |
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 100,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1432,6 +1432,714 @@
|
|
| 1432 |
"eval_samples_per_second": 42.213,
|
| 1433 |
"eval_steps_per_second": 0.331,
|
| 1434 |
"step": 200
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1435 |
}
|
| 1436 |
],
|
| 1437 |
"logging_steps": 1,
|
|
@@ -1446,7 +2154,7 @@
|
|
| 1446 |
"early_stopping_threshold": 0.0
|
| 1447 |
},
|
| 1448 |
"attributes": {
|
| 1449 |
-
"early_stopping_patience_counter":
|
| 1450 |
}
|
| 1451 |
},
|
| 1452 |
"TrainerControl": {
|
|
@@ -1455,12 +2163,12 @@
|
|
| 1455 |
"should_evaluate": false,
|
| 1456 |
"should_log": false,
|
| 1457 |
"should_save": true,
|
| 1458 |
-
"should_training_stop":
|
| 1459 |
},
|
| 1460 |
"attributes": {}
|
| 1461 |
}
|
| 1462 |
},
|
| 1463 |
-
"total_flos":
|
| 1464 |
"train_batch_size": 3,
|
| 1465 |
"trial_name": null,
|
| 1466 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": 200,
|
| 3 |
"best_metric": 0.09608737379312515,
|
| 4 |
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 5 |
+
"epoch": 6.37696335078534,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 300,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1432 |
"eval_samples_per_second": 42.213,
|
| 1433 |
"eval_steps_per_second": 0.331,
|
| 1434 |
"step": 200
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 4.272251308900524,
|
| 1438 |
+
"grad_norm": 0.10457887500524521,
|
| 1439 |
+
"learning_rate": 0.0001654277183310921,
|
| 1440 |
+
"loss": 0.082,
|
| 1441 |
+
"step": 201
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 4.293193717277487,
|
| 1445 |
+
"grad_norm": 0.10059670358896255,
|
| 1446 |
+
"learning_rate": 0.00016478324743627101,
|
| 1447 |
+
"loss": 0.0858,
|
| 1448 |
+
"step": 202
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 4.31413612565445,
|
| 1452 |
+
"grad_norm": 0.11438459903001785,
|
| 1453 |
+
"learning_rate": 0.00016413410611644825,
|
| 1454 |
+
"loss": 0.084,
|
| 1455 |
+
"step": 203
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 4.335078534031414,
|
| 1459 |
+
"grad_norm": 0.06561236828565598,
|
| 1460 |
+
"learning_rate": 0.00016348034117023258,
|
| 1461 |
+
"loss": 0.0822,
|
| 1462 |
+
"step": 204
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 4.356020942408377,
|
| 1466 |
+
"grad_norm": 0.11459755152463913,
|
| 1467 |
+
"learning_rate": 0.00016282199972956425,
|
| 1468 |
+
"loss": 0.0826,
|
| 1469 |
+
"step": 205
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 4.37696335078534,
|
| 1473 |
+
"grad_norm": 0.07275859266519547,
|
| 1474 |
+
"learning_rate": 0.00016215912925631723,
|
| 1475 |
+
"loss": 0.081,
|
| 1476 |
+
"step": 206
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 4.397905759162303,
|
| 1480 |
+
"grad_norm": 0.10688365250825882,
|
| 1481 |
+
"learning_rate": 0.00016149177753887746,
|
| 1482 |
+
"loss": 0.0804,
|
| 1483 |
+
"step": 207
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 4.418848167539267,
|
| 1487 |
+
"grad_norm": 0.07950358837842941,
|
| 1488 |
+
"learning_rate": 0.00016081999268869766,
|
| 1489 |
+
"loss": 0.0817,
|
| 1490 |
+
"step": 208
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 4.439790575916231,
|
| 1494 |
+
"grad_norm": 0.09884137660264969,
|
| 1495 |
+
"learning_rate": 0.00016014382313682881,
|
| 1496 |
+
"loss": 0.0818,
|
| 1497 |
+
"step": 209
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 4.460732984293194,
|
| 1501 |
+
"grad_norm": 0.09773270040750504,
|
| 1502 |
+
"learning_rate": 0.00015946331763042867,
|
| 1503 |
+
"loss": 0.0841,
|
| 1504 |
+
"step": 210
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 4.481675392670157,
|
| 1508 |
+
"grad_norm": 0.07958532869815826,
|
| 1509 |
+
"learning_rate": 0.00015877852522924732,
|
| 1510 |
+
"loss": 0.081,
|
| 1511 |
+
"step": 211
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 4.50261780104712,
|
| 1515 |
+
"grad_norm": 0.10386484861373901,
|
| 1516 |
+
"learning_rate": 0.0001580894953020904,
|
| 1517 |
+
"loss": 0.0821,
|
| 1518 |
+
"step": 212
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 4.523560209424084,
|
| 1522 |
+
"grad_norm": 0.0741114616394043,
|
| 1523 |
+
"learning_rate": 0.00015739627752325996,
|
| 1524 |
+
"loss": 0.081,
|
| 1525 |
+
"step": 213
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 4.544502617801047,
|
| 1529 |
+
"grad_norm": 0.10303428024053574,
|
| 1530 |
+
"learning_rate": 0.00015669892186897318,
|
| 1531 |
+
"loss": 0.0811,
|
| 1532 |
+
"step": 214
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 4.56544502617801,
|
| 1536 |
+
"grad_norm": 0.07302123308181763,
|
| 1537 |
+
"learning_rate": 0.00015599747861375955,
|
| 1538 |
+
"loss": 0.0824,
|
| 1539 |
+
"step": 215
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 4.5863874345549736,
|
| 1543 |
+
"grad_norm": 0.09889702498912811,
|
| 1544 |
+
"learning_rate": 0.00015529199832683635,
|
| 1545 |
+
"loss": 0.0798,
|
| 1546 |
+
"step": 216
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 4.607329842931938,
|
| 1550 |
+
"grad_norm": 0.0839948058128357,
|
| 1551 |
+
"learning_rate": 0.00015458253186846301,
|
| 1552 |
+
"loss": 0.084,
|
| 1553 |
+
"step": 217
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 4.628272251308901,
|
| 1557 |
+
"grad_norm": 0.07695591449737549,
|
| 1558 |
+
"learning_rate": 0.0001538691303862744,
|
| 1559 |
+
"loss": 0.0823,
|
| 1560 |
+
"step": 218
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 4.649214659685864,
|
| 1564 |
+
"grad_norm": 0.09040986001491547,
|
| 1565 |
+
"learning_rate": 0.0001531518453115934,
|
| 1566 |
+
"loss": 0.0783,
|
| 1567 |
+
"step": 219
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 4.670157068062827,
|
| 1571 |
+
"grad_norm": 0.09041474014520645,
|
| 1572 |
+
"learning_rate": 0.00015243072835572318,
|
| 1573 |
+
"loss": 0.0825,
|
| 1574 |
+
"step": 220
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 4.69109947643979,
|
| 1578 |
+
"grad_norm": 0.08503681421279907,
|
| 1579 |
+
"learning_rate": 0.00015170583150621905,
|
| 1580 |
+
"loss": 0.0818,
|
| 1581 |
+
"step": 221
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 4.712041884816754,
|
| 1585 |
+
"grad_norm": 0.08206664770841599,
|
| 1586 |
+
"learning_rate": 0.00015097720702314055,
|
| 1587 |
+
"loss": 0.0799,
|
| 1588 |
+
"step": 222
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 4.732984293193717,
|
| 1592 |
+
"grad_norm": 0.08691777288913727,
|
| 1593 |
+
"learning_rate": 0.00015024490743528393,
|
| 1594 |
+
"loss": 0.0818,
|
| 1595 |
+
"step": 223
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 4.7539267015706805,
|
| 1599 |
+
"grad_norm": 0.07108013331890106,
|
| 1600 |
+
"learning_rate": 0.00014950898553639505,
|
| 1601 |
+
"loss": 0.0796,
|
| 1602 |
+
"step": 224
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 4.774869109947644,
|
| 1606 |
+
"grad_norm": 0.09734012186527252,
|
| 1607 |
+
"learning_rate": 0.00014876949438136347,
|
| 1608 |
+
"loss": 0.0848,
|
| 1609 |
+
"step": 225
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 4.795811518324607,
|
| 1613 |
+
"grad_norm": 0.07660133391618729,
|
| 1614 |
+
"learning_rate": 0.00014802648728239742,
|
| 1615 |
+
"loss": 0.0823,
|
| 1616 |
+
"step": 226
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 4.816753926701571,
|
| 1620 |
+
"grad_norm": 0.0771099254488945,
|
| 1621 |
+
"learning_rate": 0.0001472800178051805,
|
| 1622 |
+
"loss": 0.0816,
|
| 1623 |
+
"step": 227
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 4.837696335078534,
|
| 1627 |
+
"grad_norm": 0.08631302416324615,
|
| 1628 |
+
"learning_rate": 0.00014653013976500975,
|
| 1629 |
+
"loss": 0.0824,
|
| 1630 |
+
"step": 228
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 4.858638743455497,
|
| 1634 |
+
"grad_norm": 0.07344726473093033,
|
| 1635 |
+
"learning_rate": 0.00014577690722291622,
|
| 1636 |
+
"loss": 0.0785,
|
| 1637 |
+
"step": 229
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 4.879581151832461,
|
| 1641 |
+
"grad_norm": 0.08363424241542816,
|
| 1642 |
+
"learning_rate": 0.00014502037448176734,
|
| 1643 |
+
"loss": 0.0796,
|
| 1644 |
+
"step": 230
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 4.900523560209424,
|
| 1648 |
+
"grad_norm": 0.07857396453619003,
|
| 1649 |
+
"learning_rate": 0.00014426059608235208,
|
| 1650 |
+
"loss": 0.0806,
|
| 1651 |
+
"step": 231
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 4.9214659685863875,
|
| 1655 |
+
"grad_norm": 0.08594755083322525,
|
| 1656 |
+
"learning_rate": 0.00014349762679944896,
|
| 1657 |
+
"loss": 0.0812,
|
| 1658 |
+
"step": 232
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 4.942408376963351,
|
| 1662 |
+
"grad_norm": 0.06925872713327408,
|
| 1663 |
+
"learning_rate": 0.00014273152163787726,
|
| 1664 |
+
"loss": 0.0808,
|
| 1665 |
+
"step": 233
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 4.963350785340314,
|
| 1669 |
+
"grad_norm": 0.08095414191484451,
|
| 1670 |
+
"learning_rate": 0.0001419623358285314,
|
| 1671 |
+
"loss": 0.0796,
|
| 1672 |
+
"step": 234
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 4.984293193717278,
|
| 1676 |
+
"grad_norm": 0.07342205196619034,
|
| 1677 |
+
"learning_rate": 0.0001411901248243993,
|
| 1678 |
+
"loss": 0.0806,
|
| 1679 |
+
"step": 235
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 5.020942408376963,
|
| 1683 |
+
"grad_norm": 0.23300093412399292,
|
| 1684 |
+
"learning_rate": 0.00014041494429656442,
|
| 1685 |
+
"loss": 0.1565,
|
| 1686 |
+
"step": 236
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 5.041884816753926,
|
| 1690 |
+
"grad_norm": 0.12772393226623535,
|
| 1691 |
+
"learning_rate": 0.0001396368501301925,
|
| 1692 |
+
"loss": 0.0732,
|
| 1693 |
+
"step": 237
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 5.06282722513089,
|
| 1697 |
+
"grad_norm": 0.10052972286939621,
|
| 1698 |
+
"learning_rate": 0.00013885589842050253,
|
| 1699 |
+
"loss": 0.0738,
|
| 1700 |
+
"step": 238
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 5.0837696335078535,
|
| 1704 |
+
"grad_norm": 0.1047709733247757,
|
| 1705 |
+
"learning_rate": 0.00013807214546872256,
|
| 1706 |
+
"loss": 0.075,
|
| 1707 |
+
"step": 239
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 5.104712041884817,
|
| 1711 |
+
"grad_norm": 0.10601655393838882,
|
| 1712 |
+
"learning_rate": 0.00013728564777803088,
|
| 1713 |
+
"loss": 0.0737,
|
| 1714 |
+
"step": 240
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 5.12565445026178,
|
| 1718 |
+
"grad_norm": 0.08795251697301865,
|
| 1719 |
+
"learning_rate": 0.00013649646204948255,
|
| 1720 |
+
"loss": 0.0717,
|
| 1721 |
+
"step": 241
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 5.146596858638744,
|
| 1725 |
+
"grad_norm": 0.10085717588663101,
|
| 1726 |
+
"learning_rate": 0.00013570464517792153,
|
| 1727 |
+
"loss": 0.0751,
|
| 1728 |
+
"step": 242
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 5.167539267015707,
|
| 1732 |
+
"grad_norm": 0.09512262046337128,
|
| 1733 |
+
"learning_rate": 0.00013491025424787915,
|
| 1734 |
+
"loss": 0.073,
|
| 1735 |
+
"step": 243
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 5.18848167539267,
|
| 1739 |
+
"grad_norm": 0.08350583910942078,
|
| 1740 |
+
"learning_rate": 0.0001341133465294585,
|
| 1741 |
+
"loss": 0.0761,
|
| 1742 |
+
"step": 244
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 5.209424083769633,
|
| 1746 |
+
"grad_norm": 0.09553380310535431,
|
| 1747 |
+
"learning_rate": 0.00013331397947420576,
|
| 1748 |
+
"loss": 0.0747,
|
| 1749 |
+
"step": 245
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 5.230366492146596,
|
| 1753 |
+
"grad_norm": 0.07822317630052567,
|
| 1754 |
+
"learning_rate": 0.00013251221071096836,
|
| 1755 |
+
"loss": 0.0745,
|
| 1756 |
+
"step": 246
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 5.2513089005235605,
|
| 1760 |
+
"grad_norm": 0.10339541733264923,
|
| 1761 |
+
"learning_rate": 0.00013170809804174022,
|
| 1762 |
+
"loss": 0.0762,
|
| 1763 |
+
"step": 247
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 5.272251308900524,
|
| 1767 |
+
"grad_norm": 0.08298144489526749,
|
| 1768 |
+
"learning_rate": 0.00013090169943749476,
|
| 1769 |
+
"loss": 0.0751,
|
| 1770 |
+
"step": 248
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 5.293193717277487,
|
| 1774 |
+
"grad_norm": 0.0966297909617424,
|
| 1775 |
+
"learning_rate": 0.00013009307303400556,
|
| 1776 |
+
"loss": 0.0724,
|
| 1777 |
+
"step": 249
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 5.31413612565445,
|
| 1781 |
+
"grad_norm": 0.09534008800983429,
|
| 1782 |
+
"learning_rate": 0.00012928227712765504,
|
| 1783 |
+
"loss": 0.0731,
|
| 1784 |
+
"step": 250
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 5.335078534031414,
|
| 1788 |
+
"grad_norm": 0.08681947737932205,
|
| 1789 |
+
"learning_rate": 0.00012846937017123197,
|
| 1790 |
+
"loss": 0.075,
|
| 1791 |
+
"step": 251
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 5.356020942408377,
|
| 1795 |
+
"grad_norm": 0.09559471905231476,
|
| 1796 |
+
"learning_rate": 0.00012765441076971712,
|
| 1797 |
+
"loss": 0.0717,
|
| 1798 |
+
"step": 252
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 5.37696335078534,
|
| 1802 |
+
"grad_norm": 0.08022520691156387,
|
| 1803 |
+
"learning_rate": 0.00012683745767605846,
|
| 1804 |
+
"loss": 0.0766,
|
| 1805 |
+
"step": 253
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 5.397905759162303,
|
| 1809 |
+
"grad_norm": 0.10284972935914993,
|
| 1810 |
+
"learning_rate": 0.0001260185697869353,
|
| 1811 |
+
"loss": 0.0704,
|
| 1812 |
+
"step": 254
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 5.418848167539267,
|
| 1816 |
+
"grad_norm": 0.08318132907152176,
|
| 1817 |
+
"learning_rate": 0.00012519780613851254,
|
| 1818 |
+
"loss": 0.0746,
|
| 1819 |
+
"step": 255
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 5.439790575916231,
|
| 1823 |
+
"grad_norm": 0.08917541056871414,
|
| 1824 |
+
"learning_rate": 0.00012437522590218417,
|
| 1825 |
+
"loss": 0.0733,
|
| 1826 |
+
"step": 256
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 5.460732984293194,
|
| 1830 |
+
"grad_norm": 0.08834797143936157,
|
| 1831 |
+
"learning_rate": 0.00012355088838030776,
|
| 1832 |
+
"loss": 0.075,
|
| 1833 |
+
"step": 257
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 5.481675392670157,
|
| 1837 |
+
"grad_norm": 0.08001340925693512,
|
| 1838 |
+
"learning_rate": 0.00012272485300192902,
|
| 1839 |
+
"loss": 0.0731,
|
| 1840 |
+
"step": 258
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 5.50261780104712,
|
| 1844 |
+
"grad_norm": 0.07309938222169876,
|
| 1845 |
+
"learning_rate": 0.00012189717931849731,
|
| 1846 |
+
"loss": 0.0719,
|
| 1847 |
+
"step": 259
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 5.523560209424084,
|
| 1851 |
+
"grad_norm": 0.07951314002275467,
|
| 1852 |
+
"learning_rate": 0.00012106792699957263,
|
| 1853 |
+
"loss": 0.0741,
|
| 1854 |
+
"step": 260
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 5.544502617801047,
|
| 1858 |
+
"grad_norm": 0.07957018166780472,
|
| 1859 |
+
"learning_rate": 0.00012023715582852357,
|
| 1860 |
+
"loss": 0.0738,
|
| 1861 |
+
"step": 261
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 5.56544502617801,
|
| 1865 |
+
"grad_norm": 0.076540008187294,
|
| 1866 |
+
"learning_rate": 0.00011940492569821753,
|
| 1867 |
+
"loss": 0.0714,
|
| 1868 |
+
"step": 262
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 5.5863874345549736,
|
| 1872 |
+
"grad_norm": 0.08407393842935562,
|
| 1873 |
+
"learning_rate": 0.00011857129660670281,
|
| 1874 |
+
"loss": 0.0777,
|
| 1875 |
+
"step": 263
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 5.607329842931938,
|
| 1879 |
+
"grad_norm": 0.0788414478302002,
|
| 1880 |
+
"learning_rate": 0.00011773632865288309,
|
| 1881 |
+
"loss": 0.0732,
|
| 1882 |
+
"step": 264
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 5.628272251308901,
|
| 1886 |
+
"grad_norm": 0.0724525973200798,
|
| 1887 |
+
"learning_rate": 0.00011690008203218493,
|
| 1888 |
+
"loss": 0.0783,
|
| 1889 |
+
"step": 265
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 5.649214659685864,
|
| 1893 |
+
"grad_norm": 0.0882321372628212,
|
| 1894 |
+
"learning_rate": 0.00011606261703221772,
|
| 1895 |
+
"loss": 0.0781,
|
| 1896 |
+
"step": 266
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 5.670157068062827,
|
| 1900 |
+
"grad_norm": 0.07683246582746506,
|
| 1901 |
+
"learning_rate": 0.00011522399402842783,
|
| 1902 |
+
"loss": 0.0706,
|
| 1903 |
+
"step": 267
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 5.69109947643979,
|
| 1907 |
+
"grad_norm": 0.07433947920799255,
|
| 1908 |
+
"learning_rate": 0.00011438427347974554,
|
| 1909 |
+
"loss": 0.074,
|
| 1910 |
+
"step": 268
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 5.712041884816754,
|
| 1914 |
+
"grad_norm": 0.07308503985404968,
|
| 1915 |
+
"learning_rate": 0.00011354351592422665,
|
| 1916 |
+
"loss": 0.0729,
|
| 1917 |
+
"step": 269
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 5.732984293193717,
|
| 1921 |
+
"grad_norm": 0.08603333681821823,
|
| 1922 |
+
"learning_rate": 0.00011270178197468789,
|
| 1923 |
+
"loss": 0.0752,
|
| 1924 |
+
"step": 270
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 5.7539267015706805,
|
| 1928 |
+
"grad_norm": 0.07910820841789246,
|
| 1929 |
+
"learning_rate": 0.00011185913231433733,
|
| 1930 |
+
"loss": 0.0752,
|
| 1931 |
+
"step": 271
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 5.774869109947644,
|
| 1935 |
+
"grad_norm": 0.07771284133195877,
|
| 1936 |
+
"learning_rate": 0.00011101562769239946,
|
| 1937 |
+
"loss": 0.0739,
|
| 1938 |
+
"step": 272
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 5.795811518324607,
|
| 1942 |
+
"grad_norm": 0.08137574046850204,
|
| 1943 |
+
"learning_rate": 0.0001101713289197356,
|
| 1944 |
+
"loss": 0.0704,
|
| 1945 |
+
"step": 273
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 5.816753926701571,
|
| 1949 |
+
"grad_norm": 0.07771284133195877,
|
| 1950 |
+
"learning_rate": 0.00010932629686445986,
|
| 1951 |
+
"loss": 0.0766,
|
| 1952 |
+
"step": 274
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 5.837696335078534,
|
| 1956 |
+
"grad_norm": 0.07269325852394104,
|
| 1957 |
+
"learning_rate": 0.00010848059244755093,
|
| 1958 |
+
"loss": 0.0738,
|
| 1959 |
+
"step": 275
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 5.858638743455497,
|
| 1963 |
+
"grad_norm": 0.09434104710817337,
|
| 1964 |
+
"learning_rate": 0.00010763427663846015,
|
| 1965 |
+
"loss": 0.0754,
|
| 1966 |
+
"step": 276
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 5.879581151832461,
|
| 1970 |
+
"grad_norm": 0.07986113429069519,
|
| 1971 |
+
"learning_rate": 0.00010678741045071609,
|
| 1972 |
+
"loss": 0.0727,
|
| 1973 |
+
"step": 277
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 5.900523560209424,
|
| 1977 |
+
"grad_norm": 0.08152402937412262,
|
| 1978 |
+
"learning_rate": 0.00010594005493752568,
|
| 1979 |
+
"loss": 0.0763,
|
| 1980 |
+
"step": 278
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 5.9214659685863875,
|
| 1984 |
+
"grad_norm": 0.08020524680614471,
|
| 1985 |
+
"learning_rate": 0.00010509227118737298,
|
| 1986 |
+
"loss": 0.0728,
|
| 1987 |
+
"step": 279
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 5.942408376963351,
|
| 1991 |
+
"grad_norm": 0.08128321915864944,
|
| 1992 |
+
"learning_rate": 0.00010424412031961484,
|
| 1993 |
+
"loss": 0.0726,
|
| 1994 |
+
"step": 280
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 5.963350785340314,
|
| 1998 |
+
"grad_norm": 0.09842672944068909,
|
| 1999 |
+
"learning_rate": 0.00010339566348007487,
|
| 2000 |
+
"loss": 0.0738,
|
| 2001 |
+
"step": 281
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 5.984293193717278,
|
| 2005 |
+
"grad_norm": 0.0821060761809349,
|
| 2006 |
+
"learning_rate": 0.00010254696183663511,
|
| 2007 |
+
"loss": 0.0741,
|
| 2008 |
+
"step": 282
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 6.020942408376963,
|
| 2012 |
+
"grad_norm": 0.24885313212871552,
|
| 2013 |
+
"learning_rate": 0.00010169807657482623,
|
| 2014 |
+
"loss": 0.1464,
|
| 2015 |
+
"step": 283
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 6.041884816753926,
|
| 2019 |
+
"grad_norm": 0.10097737610340118,
|
| 2020 |
+
"learning_rate": 0.00010084906889341656,
|
| 2021 |
+
"loss": 0.0664,
|
| 2022 |
+
"step": 284
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 6.06282722513089,
|
| 2026 |
+
"grad_norm": 0.10555354505777359,
|
| 2027 |
+
"learning_rate": 0.0001,
|
| 2028 |
+
"loss": 0.0672,
|
| 2029 |
+
"step": 285
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 6.0837696335078535,
|
| 2033 |
+
"grad_norm": 0.10018379241228104,
|
| 2034 |
+
"learning_rate": 9.915093110658346e-05,
|
| 2035 |
+
"loss": 0.0678,
|
| 2036 |
+
"step": 286
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 6.104712041884817,
|
| 2040 |
+
"grad_norm": 0.09475495666265488,
|
| 2041 |
+
"learning_rate": 9.830192342517379e-05,
|
| 2042 |
+
"loss": 0.0662,
|
| 2043 |
+
"step": 287
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 6.12565445026178,
|
| 2047 |
+
"grad_norm": 0.10079528391361237,
|
| 2048 |
+
"learning_rate": 9.745303816336489e-05,
|
| 2049 |
+
"loss": 0.0645,
|
| 2050 |
+
"step": 288
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 6.146596858638744,
|
| 2054 |
+
"grad_norm": 0.08908044546842575,
|
| 2055 |
+
"learning_rate": 9.660433651992514e-05,
|
| 2056 |
+
"loss": 0.0671,
|
| 2057 |
+
"step": 289
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 6.167539267015707,
|
| 2061 |
+
"grad_norm": 0.10220327973365784,
|
| 2062 |
+
"learning_rate": 9.57558796803852e-05,
|
| 2063 |
+
"loss": 0.0652,
|
| 2064 |
+
"step": 290
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 6.18848167539267,
|
| 2068 |
+
"grad_norm": 0.0913078561425209,
|
| 2069 |
+
"learning_rate": 9.490772881262709e-05,
|
| 2070 |
+
"loss": 0.0654,
|
| 2071 |
+
"step": 291
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 6.209424083769633,
|
| 2075 |
+
"grad_norm": 0.09089622646570206,
|
| 2076 |
+
"learning_rate": 9.405994506247432e-05,
|
| 2077 |
+
"loss": 0.0659,
|
| 2078 |
+
"step": 292
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 6.230366492146596,
|
| 2082 |
+
"grad_norm": 0.10402899235486984,
|
| 2083 |
+
"learning_rate": 9.321258954928393e-05,
|
| 2084 |
+
"loss": 0.0672,
|
| 2085 |
+
"step": 293
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 6.2513089005235605,
|
| 2089 |
+
"grad_norm": 0.09207270294427872,
|
| 2090 |
+
"learning_rate": 9.236572336153986e-05,
|
| 2091 |
+
"loss": 0.0688,
|
| 2092 |
+
"step": 294
|
| 2093 |
+
},
|
| 2094 |
+
{
|
| 2095 |
+
"epoch": 6.272251308900524,
|
| 2096 |
+
"grad_norm": 0.10593326389789581,
|
| 2097 |
+
"learning_rate": 9.151940755244912e-05,
|
| 2098 |
+
"loss": 0.0655,
|
| 2099 |
+
"step": 295
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 6.293193717277487,
|
| 2103 |
+
"grad_norm": 0.09085794538259506,
|
| 2104 |
+
"learning_rate": 9.067370313554015e-05,
|
| 2105 |
+
"loss": 0.0663,
|
| 2106 |
+
"step": 296
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 6.31413612565445,
|
| 2110 |
+
"grad_norm": 0.106470987200737,
|
| 2111 |
+
"learning_rate": 8.982867108026442e-05,
|
| 2112 |
+
"loss": 0.0659,
|
| 2113 |
+
"step": 297
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 6.335078534031414,
|
| 2117 |
+
"grad_norm": 0.08805633336305618,
|
| 2118 |
+
"learning_rate": 8.898437230760058e-05,
|
| 2119 |
+
"loss": 0.0672,
|
| 2120 |
+
"step": 298
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 6.356020942408377,
|
| 2124 |
+
"grad_norm": 0.10489070415496826,
|
| 2125 |
+
"learning_rate": 8.814086768566272e-05,
|
| 2126 |
+
"loss": 0.0665,
|
| 2127 |
+
"step": 299
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 6.37696335078534,
|
| 2131 |
+
"grad_norm": 0.11271199584007263,
|
| 2132 |
+
"learning_rate": 8.729821802531212e-05,
|
| 2133 |
+
"loss": 0.0685,
|
| 2134 |
+
"step": 300
|
| 2135 |
+
},
|
| 2136 |
+
{
|
| 2137 |
+
"epoch": 6.37696335078534,
|
| 2138 |
+
"eval_loss": 0.09648442268371582,
|
| 2139 |
+
"eval_runtime": 36.2137,
|
| 2140 |
+
"eval_samples_per_second": 42.222,
|
| 2141 |
+
"eval_steps_per_second": 0.331,
|
| 2142 |
+
"step": 300
|
| 2143 |
}
|
| 2144 |
],
|
| 2145 |
"logging_steps": 1,
|
|
|
|
| 2154 |
"early_stopping_threshold": 0.0
|
| 2155 |
},
|
| 2156 |
"attributes": {
|
| 2157 |
+
"early_stopping_patience_counter": 1
|
| 2158 |
}
|
| 2159 |
},
|
| 2160 |
"TrainerControl": {
|
|
|
|
| 2163 |
"should_evaluate": false,
|
| 2164 |
"should_log": false,
|
| 2165 |
"should_save": true,
|
| 2166 |
+
"should_training_stop": true
|
| 2167 |
},
|
| 2168 |
"attributes": {}
|
| 2169 |
}
|
| 2170 |
},
|
| 2171 |
+
"total_flos": 8.071804808450802e+18,
|
| 2172 |
"train_batch_size": 3,
|
| 2173 |
"trial_name": null,
|
| 2174 |
"trial_params": null
|