Training in progress, step 5000, 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:c101ac625142f29a6eee5fbfbbd1737d7d0d9ab6c1245846ed65aa0ca3dbcd3e
|
| 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 sha256:0b8ef22d00bb6ace6519a3506e3fd3a360d90e18350c43243854929dc893f639
|
| 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 sha256:4df7d6154cda7a00505b6eddfedfa554866c0af33a7d042dd091c8259eaf0054
|
| 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 sha256:6a62b5c164f986f9d7e084ab3224a2e2fc4938c3db509bf911091becb8eddb7a
|
| 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:d6882a39be52f25d8bb03ff8895311f5da4548bf7a9152c5ef498d3c8a8ff73d
|
| 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 sha256:034c77ddc842f2386858f8350e7548f3f30fbadab51a1bb76f79568b3c5209e2
|
| 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 sha256:08233672e53f8e7c033ec9e5b166eed91d03a7224d96fdf66fafd9dfd76c9dd0
|
| 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 sha256:66187275f250447642ddb60038bb8e413bcb975c1ac8e272d3742b89ed50b9a1
|
| 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 sha256:898f17a0826b05f629ae14d2d48ae45508c41fcab9c163a67b7f4f53dec1d9c4
|
| 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:f7bd0fbe69ea56c936b7fe43ba220db185e5fd561f6d859a9ce0ea7bf56998a1
|
| 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 sha256:40853ecc212e01e867ec0a5405df4309683b8c34e06986a13d1a856a47ba9534
|
| 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,
|
|
@@ -3128,6 +3128,786 @@
|
|
| 3128 |
"eval_samples_per_second": 1.769,
|
| 3129 |
"eval_steps_per_second": 0.442,
|
| 3130 |
"step": 4000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3131 |
}
|
| 3132 |
],
|
| 3133 |
"logging_steps": 10,
|
|
@@ -3147,7 +3927,7 @@
|
|
| 3147 |
"attributes": {}
|
| 3148 |
}
|
| 3149 |
},
|
| 3150 |
-
"total_flos":
|
| 3151 |
"train_batch_size": 1,
|
| 3152 |
"trial_name": null,
|
| 3153 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.45454545454545453,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 5000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": false,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 3128 |
"eval_samples_per_second": 1.769,
|
| 3129 |
"eval_steps_per_second": 0.442,
|
| 3130 |
"step": 4000
|
| 3131 |
+
},
|
| 3132 |
+
{
|
| 3133 |
+
"epoch": 0.36454545454545456,
|
| 3134 |
+
"grad_norm": 331842360180736.0,
|
| 3135 |
+
"learning_rate": 7.065377653279787e-07,
|
| 3136 |
+
"loss": 31.492498779296874,
|
| 3137 |
+
"step": 4010
|
| 3138 |
+
},
|
| 3139 |
+
{
|
| 3140 |
+
"epoch": 0.3654545454545455,
|
| 3141 |
+
"grad_norm": 349987120611328.0,
|
| 3142 |
+
"learning_rate": 7.052364528176298e-07,
|
| 3143 |
+
"loss": 31.934201049804688,
|
| 3144 |
+
"step": 4020
|
| 3145 |
+
},
|
| 3146 |
+
{
|
| 3147 |
+
"epoch": 0.3663636363636364,
|
| 3148 |
+
"grad_norm": 60965357879296.0,
|
| 3149 |
+
"learning_rate": 7.03933466256685e-07,
|
| 3150 |
+
"loss": 31.326864624023436,
|
| 3151 |
+
"step": 4030
|
| 3152 |
+
},
|
| 3153 |
+
{
|
| 3154 |
+
"epoch": 0.36727272727272725,
|
| 3155 |
+
"grad_norm": 48224626802688.0,
|
| 3156 |
+
"learning_rate": 7.026288162732046e-07,
|
| 3157 |
+
"loss": 31.994424438476564,
|
| 3158 |
+
"step": 4040
|
| 3159 |
+
},
|
| 3160 |
+
{
|
| 3161 |
+
"epoch": 0.36818181818181817,
|
| 3162 |
+
"grad_norm": 141879077240832.0,
|
| 3163 |
+
"learning_rate": 7.013225135088171e-07,
|
| 3164 |
+
"loss": 31.8456298828125,
|
| 3165 |
+
"step": 4050
|
| 3166 |
+
},
|
| 3167 |
+
{
|
| 3168 |
+
"epoch": 0.3690909090909091,
|
| 3169 |
+
"grad_norm": 160698214645760.0,
|
| 3170 |
+
"learning_rate": 7.000145686186323e-07,
|
| 3171 |
+
"loss": 31.545791625976562,
|
| 3172 |
+
"step": 4060
|
| 3173 |
+
},
|
| 3174 |
+
{
|
| 3175 |
+
"epoch": 0.37,
|
| 3176 |
+
"grad_norm": 148393821208576.0,
|
| 3177 |
+
"learning_rate": 6.987049922711543e-07,
|
| 3178 |
+
"loss": 31.230282592773438,
|
| 3179 |
+
"step": 4070
|
| 3180 |
+
},
|
| 3181 |
+
{
|
| 3182 |
+
"epoch": 0.3709090909090909,
|
| 3183 |
+
"grad_norm": 45071789457408.0,
|
| 3184 |
+
"learning_rate": 6.973937951481943e-07,
|
| 3185 |
+
"loss": 31.038577270507812,
|
| 3186 |
+
"step": 4080
|
| 3187 |
+
},
|
| 3188 |
+
{
|
| 3189 |
+
"epoch": 0.3718181818181818,
|
| 3190 |
+
"grad_norm": 251187320324096.0,
|
| 3191 |
+
"learning_rate": 6.960809879447838e-07,
|
| 3192 |
+
"loss": 31.384237670898436,
|
| 3193 |
+
"step": 4090
|
| 3194 |
+
},
|
| 3195 |
+
{
|
| 3196 |
+
"epoch": 0.37272727272727274,
|
| 3197 |
+
"grad_norm": 80505772965888.0,
|
| 3198 |
+
"learning_rate": 6.947665813690871e-07,
|
| 3199 |
+
"loss": 31.199725341796874,
|
| 3200 |
+
"step": 4100
|
| 3201 |
+
},
|
| 3202 |
+
{
|
| 3203 |
+
"epoch": 0.37272727272727274,
|
| 3204 |
+
"eval_loss": 31.208812713623047,
|
| 3205 |
+
"eval_runtime": 2.2469,
|
| 3206 |
+
"eval_samples_per_second": 1.78,
|
| 3207 |
+
"eval_steps_per_second": 0.445,
|
| 3208 |
+
"step": 4100
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 0.37363636363636366,
|
| 3212 |
+
"grad_norm": 292126529159168.0,
|
| 3213 |
+
"learning_rate": 6.934505861423144e-07,
|
| 3214 |
+
"loss": 31.544012451171874,
|
| 3215 |
+
"step": 4110
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 0.37454545454545457,
|
| 3219 |
+
"grad_norm": 129792653393920.0,
|
| 3220 |
+
"learning_rate": 6.921330129986338e-07,
|
| 3221 |
+
"loss": 31.441571044921876,
|
| 3222 |
+
"step": 4120
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 0.37545454545454543,
|
| 3226 |
+
"grad_norm": 225189916835840.0,
|
| 3227 |
+
"learning_rate": 6.90813872685084e-07,
|
| 3228 |
+
"loss": 31.419064331054688,
|
| 3229 |
+
"step": 4130
|
| 3230 |
+
},
|
| 3231 |
+
{
|
| 3232 |
+
"epoch": 0.37636363636363634,
|
| 3233 |
+
"grad_norm": 210190464974848.0,
|
| 3234 |
+
"learning_rate": 6.894931759614865e-07,
|
| 3235 |
+
"loss": 31.52581787109375,
|
| 3236 |
+
"step": 4140
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 0.37727272727272726,
|
| 3240 |
+
"grad_norm": 70698005430272.0,
|
| 3241 |
+
"learning_rate": 6.881709336003584e-07,
|
| 3242 |
+
"loss": 31.436795043945313,
|
| 3243 |
+
"step": 4150
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 0.3781818181818182,
|
| 3247 |
+
"grad_norm": 212850693898240.0,
|
| 3248 |
+
"learning_rate": 6.868471563868235e-07,
|
| 3249 |
+
"loss": 31.254794311523437,
|
| 3250 |
+
"step": 4160
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 0.3790909090909091,
|
| 3254 |
+
"grad_norm": 62373582864384.0,
|
| 3255 |
+
"learning_rate": 6.855218551185254e-07,
|
| 3256 |
+
"loss": 30.771942138671875,
|
| 3257 |
+
"step": 4170
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 0.38,
|
| 3261 |
+
"grad_norm": 208657379753984.0,
|
| 3262 |
+
"learning_rate": 6.841950406055389e-07,
|
| 3263 |
+
"loss": 31.365264892578125,
|
| 3264 |
+
"step": 4180
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 0.3809090909090909,
|
| 3268 |
+
"grad_norm": 309057994883072.0,
|
| 3269 |
+
"learning_rate": 6.828667236702815e-07,
|
| 3270 |
+
"loss": 31.405780029296874,
|
| 3271 |
+
"step": 4190
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 0.38181818181818183,
|
| 3275 |
+
"grad_norm": 67258516766720.0,
|
| 3276 |
+
"learning_rate": 6.815369151474256e-07,
|
| 3277 |
+
"loss": 31.415036010742188,
|
| 3278 |
+
"step": 4200
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 0.38181818181818183,
|
| 3282 |
+
"eval_loss": 31.2274169921875,
|
| 3283 |
+
"eval_runtime": 2.2563,
|
| 3284 |
+
"eval_samples_per_second": 1.773,
|
| 3285 |
+
"eval_steps_per_second": 0.443,
|
| 3286 |
+
"step": 4200
|
| 3287 |
+
},
|
| 3288 |
+
{
|
| 3289 |
+
"epoch": 0.38272727272727275,
|
| 3290 |
+
"grad_norm": 254599168524288.0,
|
| 3291 |
+
"learning_rate": 6.802056258838104e-07,
|
| 3292 |
+
"loss": 31.626663208007812,
|
| 3293 |
+
"step": 4210
|
| 3294 |
+
},
|
| 3295 |
+
{
|
| 3296 |
+
"epoch": 0.3836363636363636,
|
| 3297 |
+
"grad_norm": 246138183614464.0,
|
| 3298 |
+
"learning_rate": 6.788728667383528e-07,
|
| 3299 |
+
"loss": 31.129254150390626,
|
| 3300 |
+
"step": 4220
|
| 3301 |
+
},
|
| 3302 |
+
{
|
| 3303 |
+
"epoch": 0.3845454545454545,
|
| 3304 |
+
"grad_norm": 116115715915776.0,
|
| 3305 |
+
"learning_rate": 6.775386485819589e-07,
|
| 3306 |
+
"loss": 31.354595947265626,
|
| 3307 |
+
"step": 4230
|
| 3308 |
+
},
|
| 3309 |
+
{
|
| 3310 |
+
"epoch": 0.38545454545454544,
|
| 3311 |
+
"grad_norm": 37121943404544.0,
|
| 3312 |
+
"learning_rate": 6.762029822974359e-07,
|
| 3313 |
+
"loss": 31.336276245117187,
|
| 3314 |
+
"step": 4240
|
| 3315 |
+
},
|
| 3316 |
+
{
|
| 3317 |
+
"epoch": 0.38636363636363635,
|
| 3318 |
+
"grad_norm": 53457247535104.0,
|
| 3319 |
+
"learning_rate": 6.748658787794025e-07,
|
| 3320 |
+
"loss": 31.495358276367188,
|
| 3321 |
+
"step": 4250
|
| 3322 |
+
},
|
| 3323 |
+
{
|
| 3324 |
+
"epoch": 0.38727272727272727,
|
| 3325 |
+
"grad_norm": 218983068336128.0,
|
| 3326 |
+
"learning_rate": 6.735273489342008e-07,
|
| 3327 |
+
"loss": 31.227752685546875,
|
| 3328 |
+
"step": 4260
|
| 3329 |
+
},
|
| 3330 |
+
{
|
| 3331 |
+
"epoch": 0.3881818181818182,
|
| 3332 |
+
"grad_norm": 164807726596096.0,
|
| 3333 |
+
"learning_rate": 6.721874036798068e-07,
|
| 3334 |
+
"loss": 31.28994140625,
|
| 3335 |
+
"step": 4270
|
| 3336 |
+
},
|
| 3337 |
+
{
|
| 3338 |
+
"epoch": 0.3890909090909091,
|
| 3339 |
+
"grad_norm": 65006854995968.0,
|
| 3340 |
+
"learning_rate": 6.708460539457417e-07,
|
| 3341 |
+
"loss": 31.27008056640625,
|
| 3342 |
+
"step": 4280
|
| 3343 |
+
},
|
| 3344 |
+
{
|
| 3345 |
+
"epoch": 0.39,
|
| 3346 |
+
"grad_norm": 257791369412608.0,
|
| 3347 |
+
"learning_rate": 6.695033106729824e-07,
|
| 3348 |
+
"loss": 31.422894287109376,
|
| 3349 |
+
"step": 4290
|
| 3350 |
+
},
|
| 3351 |
+
{
|
| 3352 |
+
"epoch": 0.39090909090909093,
|
| 3353 |
+
"grad_norm": 200739271999488.0,
|
| 3354 |
+
"learning_rate": 6.681591848138731e-07,
|
| 3355 |
+
"loss": 31.794927978515624,
|
| 3356 |
+
"step": 4300
|
| 3357 |
+
},
|
| 3358 |
+
{
|
| 3359 |
+
"epoch": 0.39090909090909093,
|
| 3360 |
+
"eval_loss": 31.233016967773438,
|
| 3361 |
+
"eval_runtime": 2.2393,
|
| 3362 |
+
"eval_samples_per_second": 1.786,
|
| 3363 |
+
"eval_steps_per_second": 0.447,
|
| 3364 |
+
"step": 4300
|
| 3365 |
+
},
|
| 3366 |
+
{
|
| 3367 |
+
"epoch": 0.39181818181818184,
|
| 3368 |
+
"grad_norm": 236140539936768.0,
|
| 3369 |
+
"learning_rate": 6.668136873320344e-07,
|
| 3370 |
+
"loss": 31.521975708007812,
|
| 3371 |
+
"step": 4310
|
| 3372 |
+
},
|
| 3373 |
+
{
|
| 3374 |
+
"epoch": 0.3927272727272727,
|
| 3375 |
+
"grad_norm": 261026301870080.0,
|
| 3376 |
+
"learning_rate": 6.654668292022754e-07,
|
| 3377 |
+
"loss": 31.655804443359376,
|
| 3378 |
+
"step": 4320
|
| 3379 |
+
},
|
| 3380 |
+
{
|
| 3381 |
+
"epoch": 0.3936363636363636,
|
| 3382 |
+
"grad_norm": 148370215665664.0,
|
| 3383 |
+
"learning_rate": 6.641186214105033e-07,
|
| 3384 |
+
"loss": 31.52177734375,
|
| 3385 |
+
"step": 4330
|
| 3386 |
+
},
|
| 3387 |
+
{
|
| 3388 |
+
"epoch": 0.39454545454545453,
|
| 3389 |
+
"grad_norm": 356924298100736.0,
|
| 3390 |
+
"learning_rate": 6.627690749536346e-07,
|
| 3391 |
+
"loss": 31.177362060546876,
|
| 3392 |
+
"step": 4340
|
| 3393 |
+
},
|
| 3394 |
+
{
|
| 3395 |
+
"epoch": 0.39545454545454545,
|
| 3396 |
+
"grad_norm": 349417802563584.0,
|
| 3397 |
+
"learning_rate": 6.614182008395042e-07,
|
| 3398 |
+
"loss": 31.20628662109375,
|
| 3399 |
+
"step": 4350
|
| 3400 |
+
},
|
| 3401 |
+
{
|
| 3402 |
+
"epoch": 0.39636363636363636,
|
| 3403 |
+
"grad_norm": 114095118352384.0,
|
| 3404 |
+
"learning_rate": 6.600660100867765e-07,
|
| 3405 |
+
"loss": 31.61373291015625,
|
| 3406 |
+
"step": 4360
|
| 3407 |
+
},
|
| 3408 |
+
{
|
| 3409 |
+
"epoch": 0.3972727272727273,
|
| 3410 |
+
"grad_norm": 156622810775552.0,
|
| 3411 |
+
"learning_rate": 6.587125137248558e-07,
|
| 3412 |
+
"loss": 31.190673828125,
|
| 3413 |
+
"step": 4370
|
| 3414 |
+
},
|
| 3415 |
+
{
|
| 3416 |
+
"epoch": 0.3981818181818182,
|
| 3417 |
+
"grad_norm": 238024789065728.0,
|
| 3418 |
+
"learning_rate": 6.573577227937951e-07,
|
| 3419 |
+
"loss": 31.272283935546874,
|
| 3420 |
+
"step": 4380
|
| 3421 |
+
},
|
| 3422 |
+
{
|
| 3423 |
+
"epoch": 0.3990909090909091,
|
| 3424 |
+
"grad_norm": 113410725380096.0,
|
| 3425 |
+
"learning_rate": 6.560016483442075e-07,
|
| 3426 |
+
"loss": 31.559075927734376,
|
| 3427 |
+
"step": 4390
|
| 3428 |
+
},
|
| 3429 |
+
{
|
| 3430 |
+
"epoch": 0.4,
|
| 3431 |
+
"grad_norm": 58008813961216.0,
|
| 3432 |
+
"learning_rate": 6.546443014371745e-07,
|
| 3433 |
+
"loss": 31.474777221679688,
|
| 3434 |
+
"step": 4400
|
| 3435 |
+
},
|
| 3436 |
+
{
|
| 3437 |
+
"epoch": 0.4,
|
| 3438 |
+
"eval_loss": 31.23973846435547,
|
| 3439 |
+
"eval_runtime": 2.2483,
|
| 3440 |
+
"eval_samples_per_second": 1.779,
|
| 3441 |
+
"eval_steps_per_second": 0.445,
|
| 3442 |
+
"step": 4400
|
| 3443 |
+
},
|
| 3444 |
+
{
|
| 3445 |
+
"epoch": 0.4009090909090909,
|
| 3446 |
+
"grad_norm": 143956264353792.0,
|
| 3447 |
+
"learning_rate": 6.53285693144158e-07,
|
| 3448 |
+
"loss": 31.266354370117188,
|
| 3449 |
+
"step": 4410
|
| 3450 |
+
},
|
| 3451 |
+
{
|
| 3452 |
+
"epoch": 0.4018181818181818,
|
| 3453 |
+
"grad_norm": 72287755698176.0,
|
| 3454 |
+
"learning_rate": 6.51925834546907e-07,
|
| 3455 |
+
"loss": 31.40149841308594,
|
| 3456 |
+
"step": 4420
|
| 3457 |
+
},
|
| 3458 |
+
{
|
| 3459 |
+
"epoch": 0.4027272727272727,
|
| 3460 |
+
"grad_norm": 191065126600704.0,
|
| 3461 |
+
"learning_rate": 6.5056473673737e-07,
|
| 3462 |
+
"loss": 30.717300415039062,
|
| 3463 |
+
"step": 4430
|
| 3464 |
+
},
|
| 3465 |
+
{
|
| 3466 |
+
"epoch": 0.4036363636363636,
|
| 3467 |
+
"grad_norm": 268216597217280.0,
|
| 3468 |
+
"learning_rate": 6.49202410817603e-07,
|
| 3469 |
+
"loss": 31.23114929199219,
|
| 3470 |
+
"step": 4440
|
| 3471 |
+
},
|
| 3472 |
+
{
|
| 3473 |
+
"epoch": 0.40454545454545454,
|
| 3474 |
+
"grad_norm": 178938940751872.0,
|
| 3475 |
+
"learning_rate": 6.478388678996796e-07,
|
| 3476 |
+
"loss": 31.212506103515626,
|
| 3477 |
+
"step": 4450
|
| 3478 |
+
},
|
| 3479 |
+
{
|
| 3480 |
+
"epoch": 0.40545454545454546,
|
| 3481 |
+
"grad_norm": 342899149504512.0,
|
| 3482 |
+
"learning_rate": 6.464741191055994e-07,
|
| 3483 |
+
"loss": 31.681393432617188,
|
| 3484 |
+
"step": 4460
|
| 3485 |
+
},
|
| 3486 |
+
{
|
| 3487 |
+
"epoch": 0.40636363636363637,
|
| 3488 |
+
"grad_norm": 184348150071296.0,
|
| 3489 |
+
"learning_rate": 6.451081755671985e-07,
|
| 3490 |
+
"loss": 31.174737548828126,
|
| 3491 |
+
"step": 4470
|
| 3492 |
+
},
|
| 3493 |
+
{
|
| 3494 |
+
"epoch": 0.4072727272727273,
|
| 3495 |
+
"grad_norm": 103112006172672.0,
|
| 3496 |
+
"learning_rate": 6.437410484260583e-07,
|
| 3497 |
+
"loss": 31.140850830078126,
|
| 3498 |
+
"step": 4480
|
| 3499 |
+
},
|
| 3500 |
+
{
|
| 3501 |
+
"epoch": 0.4081818181818182,
|
| 3502 |
+
"grad_norm": 327214902018048.0,
|
| 3503 |
+
"learning_rate": 6.42372748833414e-07,
|
| 3504 |
+
"loss": 31.637176513671875,
|
| 3505 |
+
"step": 4490
|
| 3506 |
+
},
|
| 3507 |
+
{
|
| 3508 |
+
"epoch": 0.4090909090909091,
|
| 3509 |
+
"grad_norm": 162547080626176.0,
|
| 3510 |
+
"learning_rate": 6.410032879500646e-07,
|
| 3511 |
+
"loss": 31.360162353515626,
|
| 3512 |
+
"step": 4500
|
| 3513 |
+
},
|
| 3514 |
+
{
|
| 3515 |
+
"epoch": 0.4090909090909091,
|
| 3516 |
+
"eval_loss": 31.26785659790039,
|
| 3517 |
+
"eval_runtime": 2.2743,
|
| 3518 |
+
"eval_samples_per_second": 1.759,
|
| 3519 |
+
"eval_steps_per_second": 0.44,
|
| 3520 |
+
"step": 4500
|
| 3521 |
+
},
|
| 3522 |
+
{
|
| 3523 |
+
"epoch": 0.41,
|
| 3524 |
+
"grad_norm": 101430937518080.0,
|
| 3525 |
+
"learning_rate": 6.396326769462811e-07,
|
| 3526 |
+
"loss": 30.955780029296875,
|
| 3527 |
+
"step": 4510
|
| 3528 |
+
},
|
| 3529 |
+
{
|
| 3530 |
+
"epoch": 0.4109090909090909,
|
| 3531 |
+
"grad_norm": 411952325066752.0,
|
| 3532 |
+
"learning_rate": 6.382609270017154e-07,
|
| 3533 |
+
"loss": 31.721075439453124,
|
| 3534 |
+
"step": 4520
|
| 3535 |
+
},
|
| 3536 |
+
{
|
| 3537 |
+
"epoch": 0.4118181818181818,
|
| 3538 |
+
"grad_norm": 289259034509312.0,
|
| 3539 |
+
"learning_rate": 6.3688804930531e-07,
|
| 3540 |
+
"loss": 31.32822265625,
|
| 3541 |
+
"step": 4530
|
| 3542 |
+
},
|
| 3543 |
+
{
|
| 3544 |
+
"epoch": 0.4127272727272727,
|
| 3545 |
+
"grad_norm": 55133039755264.0,
|
| 3546 |
+
"learning_rate": 6.355140550552058e-07,
|
| 3547 |
+
"loss": 31.106060791015626,
|
| 3548 |
+
"step": 4540
|
| 3549 |
+
},
|
| 3550 |
+
{
|
| 3551 |
+
"epoch": 0.41363636363636364,
|
| 3552 |
+
"grad_norm": 85135311503360.0,
|
| 3553 |
+
"learning_rate": 6.341389554586511e-07,
|
| 3554 |
+
"loss": 31.692236328125,
|
| 3555 |
+
"step": 4550
|
| 3556 |
+
},
|
| 3557 |
+
{
|
| 3558 |
+
"epoch": 0.41454545454545455,
|
| 3559 |
+
"grad_norm": 67352506925056.0,
|
| 3560 |
+
"learning_rate": 6.327627617319103e-07,
|
| 3561 |
+
"loss": 31.49931640625,
|
| 3562 |
+
"step": 4560
|
| 3563 |
+
},
|
| 3564 |
+
{
|
| 3565 |
+
"epoch": 0.41545454545454547,
|
| 3566 |
+
"grad_norm": 230468012212224.0,
|
| 3567 |
+
"learning_rate": 6.313854851001721e-07,
|
| 3568 |
+
"loss": 31.419146728515624,
|
| 3569 |
+
"step": 4570
|
| 3570 |
+
},
|
| 3571 |
+
{
|
| 3572 |
+
"epoch": 0.4163636363636364,
|
| 3573 |
+
"grad_norm": 191184244834304.0,
|
| 3574 |
+
"learning_rate": 6.300071367974582e-07,
|
| 3575 |
+
"loss": 31.2831298828125,
|
| 3576 |
+
"step": 4580
|
| 3577 |
+
},
|
| 3578 |
+
{
|
| 3579 |
+
"epoch": 0.4172727272727273,
|
| 3580 |
+
"grad_norm": 196244437729280.0,
|
| 3581 |
+
"learning_rate": 6.286277280665315e-07,
|
| 3582 |
+
"loss": 31.070465087890625,
|
| 3583 |
+
"step": 4590
|
| 3584 |
+
},
|
| 3585 |
+
{
|
| 3586 |
+
"epoch": 0.41818181818181815,
|
| 3587 |
+
"grad_norm": 432272285106176.0,
|
| 3588 |
+
"learning_rate": 6.27247270158805e-07,
|
| 3589 |
+
"loss": 31.196286010742188,
|
| 3590 |
+
"step": 4600
|
| 3591 |
+
},
|
| 3592 |
+
{
|
| 3593 |
+
"epoch": 0.41818181818181815,
|
| 3594 |
+
"eval_loss": 31.265504837036133,
|
| 3595 |
+
"eval_runtime": 2.2536,
|
| 3596 |
+
"eval_samples_per_second": 1.775,
|
| 3597 |
+
"eval_steps_per_second": 0.444,
|
| 3598 |
+
"step": 4600
|
| 3599 |
+
},
|
| 3600 |
+
{
|
| 3601 |
+
"epoch": 0.41909090909090907,
|
| 3602 |
+
"grad_norm": 132382501896192.0,
|
| 3603 |
+
"learning_rate": 6.258657743342486e-07,
|
| 3604 |
+
"loss": 31.409124755859374,
|
| 3605 |
+
"step": 4610
|
| 3606 |
+
},
|
| 3607 |
+
{
|
| 3608 |
+
"epoch": 0.42,
|
| 3609 |
+
"grad_norm": 123047918960640.0,
|
| 3610 |
+
"learning_rate": 6.244832518612989e-07,
|
| 3611 |
+
"loss": 31.174249267578126,
|
| 3612 |
+
"step": 4620
|
| 3613 |
+
},
|
| 3614 |
+
{
|
| 3615 |
+
"epoch": 0.4209090909090909,
|
| 3616 |
+
"grad_norm": 101801286172672.0,
|
| 3617 |
+
"learning_rate": 6.230997140167662e-07,
|
| 3618 |
+
"loss": 31.151278686523437,
|
| 3619 |
+
"step": 4630
|
| 3620 |
+
},
|
| 3621 |
+
{
|
| 3622 |
+
"epoch": 0.4218181818181818,
|
| 3623 |
+
"grad_norm": 110819392094208.0,
|
| 3624 |
+
"learning_rate": 6.217151720857432e-07,
|
| 3625 |
+
"loss": 31.5635498046875,
|
| 3626 |
+
"step": 4640
|
| 3627 |
+
},
|
| 3628 |
+
{
|
| 3629 |
+
"epoch": 0.42272727272727273,
|
| 3630 |
+
"grad_norm": 162855865286656.0,
|
| 3631 |
+
"learning_rate": 6.203296373615122e-07,
|
| 3632 |
+
"loss": 31.27996826171875,
|
| 3633 |
+
"step": 4650
|
| 3634 |
+
},
|
| 3635 |
+
{
|
| 3636 |
+
"epoch": 0.42363636363636364,
|
| 3637 |
+
"grad_norm": 113649129619456.0,
|
| 3638 |
+
"learning_rate": 6.189431211454538e-07,
|
| 3639 |
+
"loss": 31.386068725585936,
|
| 3640 |
+
"step": 4660
|
| 3641 |
+
},
|
| 3642 |
+
{
|
| 3643 |
+
"epoch": 0.42454545454545456,
|
| 3644 |
+
"grad_norm": 296968836349952.0,
|
| 3645 |
+
"learning_rate": 6.175556347469541e-07,
|
| 3646 |
+
"loss": 31.396221923828126,
|
| 3647 |
+
"step": 4670
|
| 3648 |
+
},
|
| 3649 |
+
{
|
| 3650 |
+
"epoch": 0.4254545454545455,
|
| 3651 |
+
"grad_norm": 382460730802176.0,
|
| 3652 |
+
"learning_rate": 6.161671894833127e-07,
|
| 3653 |
+
"loss": 31.629803466796876,
|
| 3654 |
+
"step": 4680
|
| 3655 |
+
},
|
| 3656 |
+
{
|
| 3657 |
+
"epoch": 0.4263636363636364,
|
| 3658 |
+
"grad_norm": 50031491022848.0,
|
| 3659 |
+
"learning_rate": 6.147777966796505e-07,
|
| 3660 |
+
"loss": 31.4884765625,
|
| 3661 |
+
"step": 4690
|
| 3662 |
+
},
|
| 3663 |
+
{
|
| 3664 |
+
"epoch": 0.42727272727272725,
|
| 3665 |
+
"grad_norm": 160202313695232.0,
|
| 3666 |
+
"learning_rate": 6.13387467668817e-07,
|
| 3667 |
+
"loss": 31.25142822265625,
|
| 3668 |
+
"step": 4700
|
| 3669 |
+
},
|
| 3670 |
+
{
|
| 3671 |
+
"epoch": 0.42727272727272725,
|
| 3672 |
+
"eval_loss": 31.27532196044922,
|
| 3673 |
+
"eval_runtime": 2.2762,
|
| 3674 |
+
"eval_samples_per_second": 1.757,
|
| 3675 |
+
"eval_steps_per_second": 0.439,
|
| 3676 |
+
"step": 4700
|
| 3677 |
+
},
|
| 3678 |
+
{
|
| 3679 |
+
"epoch": 0.42818181818181816,
|
| 3680 |
+
"grad_norm": 130051811049472.0,
|
| 3681 |
+
"learning_rate": 6.119962137912979e-07,
|
| 3682 |
+
"loss": 31.30537109375,
|
| 3683 |
+
"step": 4710
|
| 3684 |
+
},
|
| 3685 |
+
{
|
| 3686 |
+
"epoch": 0.4290909090909091,
|
| 3687 |
+
"grad_norm": 118681648496640.0,
|
| 3688 |
+
"learning_rate": 6.106040463951236e-07,
|
| 3689 |
+
"loss": 31.411505126953124,
|
| 3690 |
+
"step": 4720
|
| 3691 |
+
},
|
| 3692 |
+
{
|
| 3693 |
+
"epoch": 0.43,
|
| 3694 |
+
"grad_norm": 68806344966144.0,
|
| 3695 |
+
"learning_rate": 6.092109768357747e-07,
|
| 3696 |
+
"loss": 30.8322998046875,
|
| 3697 |
+
"step": 4730
|
| 3698 |
+
},
|
| 3699 |
+
{
|
| 3700 |
+
"epoch": 0.4309090909090909,
|
| 3701 |
+
"grad_norm": 80994853978112.0,
|
| 3702 |
+
"learning_rate": 6.078170164760911e-07,
|
| 3703 |
+
"loss": 31.552935791015624,
|
| 3704 |
+
"step": 4740
|
| 3705 |
+
},
|
| 3706 |
+
{
|
| 3707 |
+
"epoch": 0.4318181818181818,
|
| 3708 |
+
"grad_norm": 192088553553920.0,
|
| 3709 |
+
"learning_rate": 6.064221766861785e-07,
|
| 3710 |
+
"loss": 31.02978515625,
|
| 3711 |
+
"step": 4750
|
| 3712 |
+
},
|
| 3713 |
+
{
|
| 3714 |
+
"epoch": 0.43272727272727274,
|
| 3715 |
+
"grad_norm": 156923173273600.0,
|
| 3716 |
+
"learning_rate": 6.050264688433161e-07,
|
| 3717 |
+
"loss": 31.151821899414063,
|
| 3718 |
+
"step": 4760
|
| 3719 |
+
},
|
| 3720 |
+
{
|
| 3721 |
+
"epoch": 0.43363636363636365,
|
| 3722 |
+
"grad_norm": 306010447151104.0,
|
| 3723 |
+
"learning_rate": 6.036299043318631e-07,
|
| 3724 |
+
"loss": 31.844577026367187,
|
| 3725 |
+
"step": 4770
|
| 3726 |
+
},
|
| 3727 |
+
{
|
| 3728 |
+
"epoch": 0.43454545454545457,
|
| 3729 |
+
"grad_norm": 163939203678208.0,
|
| 3730 |
+
"learning_rate": 6.022324945431665e-07,
|
| 3731 |
+
"loss": 31.494683837890626,
|
| 3732 |
+
"step": 4780
|
| 3733 |
+
},
|
| 3734 |
+
{
|
| 3735 |
+
"epoch": 0.4354545454545454,
|
| 3736 |
+
"grad_norm": 307883992416256.0,
|
| 3737 |
+
"learning_rate": 6.008342508754682e-07,
|
| 3738 |
+
"loss": 31.625247192382812,
|
| 3739 |
+
"step": 4790
|
| 3740 |
+
},
|
| 3741 |
+
{
|
| 3742 |
+
"epoch": 0.43636363636363634,
|
| 3743 |
+
"grad_norm": 161946305298432.0,
|
| 3744 |
+
"learning_rate": 5.994351847338113e-07,
|
| 3745 |
+
"loss": 31.29015197753906,
|
| 3746 |
+
"step": 4800
|
| 3747 |
+
},
|
| 3748 |
+
{
|
| 3749 |
+
"epoch": 0.43636363636363634,
|
| 3750 |
+
"eval_loss": 31.277931213378906,
|
| 3751 |
+
"eval_runtime": 2.2639,
|
| 3752 |
+
"eval_samples_per_second": 1.767,
|
| 3753 |
+
"eval_steps_per_second": 0.442,
|
| 3754 |
+
"step": 4800
|
| 3755 |
+
},
|
| 3756 |
+
{
|
| 3757 |
+
"epoch": 0.43727272727272726,
|
| 3758 |
+
"grad_norm": 106276080058368.0,
|
| 3759 |
+
"learning_rate": 5.980353075299481e-07,
|
| 3760 |
+
"loss": 31.702081298828126,
|
| 3761 |
+
"step": 4810
|
| 3762 |
+
},
|
| 3763 |
+
{
|
| 3764 |
+
"epoch": 0.4381818181818182,
|
| 3765 |
+
"grad_norm": 137356761890816.0,
|
| 3766 |
+
"learning_rate": 5.966346306822458e-07,
|
| 3767 |
+
"loss": 31.010269165039062,
|
| 3768 |
+
"step": 4820
|
| 3769 |
+
},
|
| 3770 |
+
{
|
| 3771 |
+
"epoch": 0.4390909090909091,
|
| 3772 |
+
"grad_norm": 303051114020864.0,
|
| 3773 |
+
"learning_rate": 5.95233165615595e-07,
|
| 3774 |
+
"loss": 31.519564819335937,
|
| 3775 |
+
"step": 4830
|
| 3776 |
+
},
|
| 3777 |
+
{
|
| 3778 |
+
"epoch": 0.44,
|
| 3779 |
+
"grad_norm": 43947078451200.0,
|
| 3780 |
+
"learning_rate": 5.938309237613146e-07,
|
| 3781 |
+
"loss": 31.17099609375,
|
| 3782 |
+
"step": 4840
|
| 3783 |
+
},
|
| 3784 |
+
{
|
| 3785 |
+
"epoch": 0.4409090909090909,
|
| 3786 |
+
"grad_norm": 200443087028224.0,
|
| 3787 |
+
"learning_rate": 5.924279165570602e-07,
|
| 3788 |
+
"loss": 31.536282348632813,
|
| 3789 |
+
"step": 4850
|
| 3790 |
+
},
|
| 3791 |
+
{
|
| 3792 |
+
"epoch": 0.44181818181818183,
|
| 3793 |
+
"grad_norm": 255512100732928.0,
|
| 3794 |
+
"learning_rate": 5.910241554467298e-07,
|
| 3795 |
+
"loss": 31.622979736328126,
|
| 3796 |
+
"step": 4860
|
| 3797 |
+
},
|
| 3798 |
+
{
|
| 3799 |
+
"epoch": 0.44272727272727275,
|
| 3800 |
+
"grad_norm": 122165680668672.0,
|
| 3801 |
+
"learning_rate": 5.896196518803707e-07,
|
| 3802 |
+
"loss": 31.318588256835938,
|
| 3803 |
+
"step": 4870
|
| 3804 |
+
},
|
| 3805 |
+
{
|
| 3806 |
+
"epoch": 0.44363636363636366,
|
| 3807 |
+
"grad_norm": 59236595793920.0,
|
| 3808 |
+
"learning_rate": 5.882144173140867e-07,
|
| 3809 |
+
"loss": 31.511083984375,
|
| 3810 |
+
"step": 4880
|
| 3811 |
+
},
|
| 3812 |
+
{
|
| 3813 |
+
"epoch": 0.4445454545454545,
|
| 3814 |
+
"grad_norm": 160883888095232.0,
|
| 3815 |
+
"learning_rate": 5.868084632099435e-07,
|
| 3816 |
+
"loss": 31.54732666015625,
|
| 3817 |
+
"step": 4890
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 0.44545454545454544,
|
| 3821 |
+
"grad_norm": 257228040830976.0,
|
| 3822 |
+
"learning_rate": 5.854018010358761e-07,
|
| 3823 |
+
"loss": 31.00733337402344,
|
| 3824 |
+
"step": 4900
|
| 3825 |
+
},
|
| 3826 |
+
{
|
| 3827 |
+
"epoch": 0.44545454545454544,
|
| 3828 |
+
"eval_loss": 31.296031951904297,
|
| 3829 |
+
"eval_runtime": 2.2374,
|
| 3830 |
+
"eval_samples_per_second": 1.788,
|
| 3831 |
+
"eval_steps_per_second": 0.447,
|
| 3832 |
+
"step": 4900
|
| 3833 |
+
},
|
| 3834 |
+
{
|
| 3835 |
+
"epoch": 0.44636363636363635,
|
| 3836 |
+
"grad_norm": 92250444922880.0,
|
| 3837 |
+
"learning_rate": 5.839944422655952e-07,
|
| 3838 |
+
"loss": 31.35645446777344,
|
| 3839 |
+
"step": 4910
|
| 3840 |
+
},
|
| 3841 |
+
{
|
| 3842 |
+
"epoch": 0.44727272727272727,
|
| 3843 |
+
"grad_norm": 118262083878912.0,
|
| 3844 |
+
"learning_rate": 5.825863983784931e-07,
|
| 3845 |
+
"loss": 31.456143188476563,
|
| 3846 |
+
"step": 4920
|
| 3847 |
+
},
|
| 3848 |
+
{
|
| 3849 |
+
"epoch": 0.4481818181818182,
|
| 3850 |
+
"grad_norm": 55267186180096.0,
|
| 3851 |
+
"learning_rate": 5.811776808595506e-07,
|
| 3852 |
+
"loss": 31.58196105957031,
|
| 3853 |
+
"step": 4930
|
| 3854 |
+
},
|
| 3855 |
+
{
|
| 3856 |
+
"epoch": 0.4490909090909091,
|
| 3857 |
+
"grad_norm": 242786548842496.0,
|
| 3858 |
+
"learning_rate": 5.797683011992432e-07,
|
| 3859 |
+
"loss": 31.34881591796875,
|
| 3860 |
+
"step": 4940
|
| 3861 |
+
},
|
| 3862 |
+
{
|
| 3863 |
+
"epoch": 0.45,
|
| 3864 |
+
"grad_norm": 193380936056832.0,
|
| 3865 |
+
"learning_rate": 5.783582708934468e-07,
|
| 3866 |
+
"loss": 31.296319580078126,
|
| 3867 |
+
"step": 4950
|
| 3868 |
+
},
|
| 3869 |
+
{
|
| 3870 |
+
"epoch": 0.4509090909090909,
|
| 3871 |
+
"grad_norm": 106026896457728.0,
|
| 3872 |
+
"learning_rate": 5.769476014433451e-07,
|
| 3873 |
+
"loss": 30.779248046875,
|
| 3874 |
+
"step": 4960
|
| 3875 |
+
},
|
| 3876 |
+
{
|
| 3877 |
+
"epoch": 0.45181818181818184,
|
| 3878 |
+
"grad_norm": 87636718911488.0,
|
| 3879 |
+
"learning_rate": 5.755363043553346e-07,
|
| 3880 |
+
"loss": 31.5363037109375,
|
| 3881 |
+
"step": 4970
|
| 3882 |
+
},
|
| 3883 |
+
{
|
| 3884 |
+
"epoch": 0.4527272727272727,
|
| 3885 |
+
"grad_norm": 73219755212800.0,
|
| 3886 |
+
"learning_rate": 5.741243911409314e-07,
|
| 3887 |
+
"loss": 31.977978515625,
|
| 3888 |
+
"step": 4980
|
| 3889 |
+
},
|
| 3890 |
+
{
|
| 3891 |
+
"epoch": 0.4536363636363636,
|
| 3892 |
+
"grad_norm": 236564936392704.0,
|
| 3893 |
+
"learning_rate": 5.72711873316677e-07,
|
| 3894 |
+
"loss": 31.291098022460936,
|
| 3895 |
+
"step": 4990
|
| 3896 |
+
},
|
| 3897 |
+
{
|
| 3898 |
+
"epoch": 0.45454545454545453,
|
| 3899 |
+
"grad_norm": 141793949646848.0,
|
| 3900 |
+
"learning_rate": 5.712987624040451e-07,
|
| 3901 |
+
"loss": 30.713864135742188,
|
| 3902 |
+
"step": 5000
|
| 3903 |
+
},
|
| 3904 |
+
{
|
| 3905 |
+
"epoch": 0.45454545454545453,
|
| 3906 |
+
"eval_loss": 31.29385757446289,
|
| 3907 |
+
"eval_runtime": 2.2444,
|
| 3908 |
+
"eval_samples_per_second": 1.782,
|
| 3909 |
+
"eval_steps_per_second": 0.446,
|
| 3910 |
+
"step": 5000
|
| 3911 |
}
|
| 3912 |
],
|
| 3913 |
"logging_steps": 10,
|
|
|
|
| 3927 |
"attributes": {}
|
| 3928 |
}
|
| 3929 |
},
|
| 3930 |
+
"total_flos": 2.264961908736e+16,
|
| 3931 |
"train_batch_size": 1,
|
| 3932 |
"trial_name": null,
|
| 3933 |
"trial_params": null
|