Training in progress, step 4000, 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 sha256:3fb7bbba242fd1734bca682ecab7164010d011d4bae150f9c01b35c4e5ae214b
|
| 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 sha256:07069eb476da26ea80f4f21b5c52a128f07b59f88d65cea432ed4249827df3da
|
| 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 sha256:1103853b7e1b9edd1b48dc1d37f3d3e771a7bed8926820b7285cfab008291eee
|
| 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:f6e4157ae1d527e6faa3a1b54958e4ae79ca685866a1b9d3709ada96cdc5544b
|
| 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 sha256:3b3bc5cd9e8e23fa067544f348f6a8002cb45b6bc46d49ad8514270519590978
|
| 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:e0ed2314742cc9d6eae821f0a37ec34c1acb9b5f1a15a96a7124b4b01078fc4d
|
| 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 sha256:010b49e7e7f05f696c7422ea191c7f01febd0be1e95a28d27a82ba29233a5c3c
|
| 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:c344937f1eb80674ed90208df930eefdfa96d1305c5ecdd6f71c88884d9618c4
|
| 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 sha256:47845359879e4287eb6b33abc23646b7b3c47e35ecfc2924ee9249c6266386da
|
| 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 sha256:9b9614a31010cd351d1cda78d31fd350abfc95e0db8e88f38df316e5bf0e3457
|
| 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:ccc95881d63aeb232a0d233120305836db2087b7d53cc84df6ba9d3559169ce0
|
| 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,
|
|
@@ -2348,6 +2348,786 @@
|
|
| 2348 |
"eval_samples_per_second": 1.784,
|
| 2349 |
"eval_steps_per_second": 0.446,
|
| 2350 |
"step": 3000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2351 |
}
|
| 2352 |
],
|
| 2353 |
"logging_steps": 10,
|
|
@@ -2367,7 +3147,7 @@
|
|
| 2367 |
"attributes": {}
|
| 2368 |
}
|
| 2369 |
},
|
| 2370 |
-
"total_flos": 1.
|
| 2371 |
"train_batch_size": 1,
|
| 2372 |
"trial_name": null,
|
| 2373 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.36363636363636365,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 4000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": false,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 2348 |
"eval_samples_per_second": 1.784,
|
| 2349 |
"eval_steps_per_second": 0.446,
|
| 2350 |
"step": 3000
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 0.2736363636363636,
|
| 2354 |
+
"grad_norm": 44478215749632.0,
|
| 2355 |
+
"learning_rate": 8.26457999307125e-07,
|
| 2356 |
+
"loss": 30.9502685546875,
|
| 2357 |
+
"step": 3010
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 0.27454545454545454,
|
| 2361 |
+
"grad_norm": 75005027155968.0,
|
| 2362 |
+
"learning_rate": 8.253750613925234e-07,
|
| 2363 |
+
"loss": 31.542059326171874,
|
| 2364 |
+
"step": 3020
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 0.27545454545454545,
|
| 2368 |
+
"grad_norm": 47266672934912.0,
|
| 2369 |
+
"learning_rate": 8.242894694936614e-07,
|
| 2370 |
+
"loss": 30.897921752929687,
|
| 2371 |
+
"step": 3030
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.27636363636363637,
|
| 2375 |
+
"grad_norm": 45886398791680.0,
|
| 2376 |
+
"learning_rate": 8.23201232465378e-07,
|
| 2377 |
+
"loss": 31.505227661132814,
|
| 2378 |
+
"step": 3040
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 0.2772727272727273,
|
| 2382 |
+
"grad_norm": 127583672860672.0,
|
| 2383 |
+
"learning_rate": 8.221103591840881e-07,
|
| 2384 |
+
"loss": 31.139892578125,
|
| 2385 |
+
"step": 3050
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 0.2781818181818182,
|
| 2389 |
+
"grad_norm": 174062391263232.0,
|
| 2390 |
+
"learning_rate": 8.210168585477091e-07,
|
| 2391 |
+
"loss": 31.437637329101562,
|
| 2392 |
+
"step": 3060
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 0.2790909090909091,
|
| 2396 |
+
"grad_norm": 210010848100352.0,
|
| 2397 |
+
"learning_rate": 8.199207394755891e-07,
|
| 2398 |
+
"loss": 30.945309448242188,
|
| 2399 |
+
"step": 3070
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 0.28,
|
| 2403 |
+
"grad_norm": 141527191912448.0,
|
| 2404 |
+
"learning_rate": 8.188220109084346e-07,
|
| 2405 |
+
"loss": 31.963729858398438,
|
| 2406 |
+
"step": 3080
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 0.2809090909090909,
|
| 2410 |
+
"grad_norm": 269273947701248.0,
|
| 2411 |
+
"learning_rate": 8.17720681808236e-07,
|
| 2412 |
+
"loss": 31.625100708007814,
|
| 2413 |
+
"step": 3090
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 0.2818181818181818,
|
| 2417 |
+
"grad_norm": 90777153699840.0,
|
| 2418 |
+
"learning_rate": 8.16616761158196e-07,
|
| 2419 |
+
"loss": 31.37896728515625,
|
| 2420 |
+
"step": 3100
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 0.2818181818181818,
|
| 2424 |
+
"eval_loss": 31.08748435974121,
|
| 2425 |
+
"eval_runtime": 2.2472,
|
| 2426 |
+
"eval_samples_per_second": 1.78,
|
| 2427 |
+
"eval_steps_per_second": 0.445,
|
| 2428 |
+
"step": 3100
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 0.2827272727272727,
|
| 2432 |
+
"grad_norm": 85872594649088.0,
|
| 2433 |
+
"learning_rate": 8.155102579626558e-07,
|
| 2434 |
+
"loss": 31.415878295898438,
|
| 2435 |
+
"step": 3110
|
| 2436 |
+
},
|
| 2437 |
+
{
|
| 2438 |
+
"epoch": 0.28363636363636363,
|
| 2439 |
+
"grad_norm": 128033721679872.0,
|
| 2440 |
+
"learning_rate": 8.144011812470215e-07,
|
| 2441 |
+
"loss": 31.359136962890624,
|
| 2442 |
+
"step": 3120
|
| 2443 |
+
},
|
| 2444 |
+
{
|
| 2445 |
+
"epoch": 0.28454545454545455,
|
| 2446 |
+
"grad_norm": 58961826938880.0,
|
| 2447 |
+
"learning_rate": 8.132895400576904e-07,
|
| 2448 |
+
"loss": 31.53883056640625,
|
| 2449 |
+
"step": 3130
|
| 2450 |
+
},
|
| 2451 |
+
{
|
| 2452 |
+
"epoch": 0.28545454545454546,
|
| 2453 |
+
"grad_norm": 130452350304256.0,
|
| 2454 |
+
"learning_rate": 8.121753434619778e-07,
|
| 2455 |
+
"loss": 31.31527099609375,
|
| 2456 |
+
"step": 3140
|
| 2457 |
+
},
|
| 2458 |
+
{
|
| 2459 |
+
"epoch": 0.2863636363636364,
|
| 2460 |
+
"grad_norm": 39132558524416.0,
|
| 2461 |
+
"learning_rate": 8.110586005480426e-07,
|
| 2462 |
+
"loss": 30.843875122070312,
|
| 2463 |
+
"step": 3150
|
| 2464 |
+
},
|
| 2465 |
+
{
|
| 2466 |
+
"epoch": 0.2872727272727273,
|
| 2467 |
+
"grad_norm": 162498661580800.0,
|
| 2468 |
+
"learning_rate": 8.09939320424813e-07,
|
| 2469 |
+
"loss": 30.928125,
|
| 2470 |
+
"step": 3160
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 0.2881818181818182,
|
| 2474 |
+
"grad_norm": 30503648362496.0,
|
| 2475 |
+
"learning_rate": 8.088175122219127e-07,
|
| 2476 |
+
"loss": 31.2546142578125,
|
| 2477 |
+
"step": 3170
|
| 2478 |
+
},
|
| 2479 |
+
{
|
| 2480 |
+
"epoch": 0.28909090909090907,
|
| 2481 |
+
"grad_norm": 264652227346432.0,
|
| 2482 |
+
"learning_rate": 8.076931850895858e-07,
|
| 2483 |
+
"loss": 31.682107543945314,
|
| 2484 |
+
"step": 3180
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 0.29,
|
| 2488 |
+
"grad_norm": 43181974487040.0,
|
| 2489 |
+
"learning_rate": 8.065663481986229e-07,
|
| 2490 |
+
"loss": 31.056671142578125,
|
| 2491 |
+
"step": 3190
|
| 2492 |
+
},
|
| 2493 |
+
{
|
| 2494 |
+
"epoch": 0.2909090909090909,
|
| 2495 |
+
"grad_norm": 81322110353408.0,
|
| 2496 |
+
"learning_rate": 8.054370107402858e-07,
|
| 2497 |
+
"loss": 31.157467651367188,
|
| 2498 |
+
"step": 3200
|
| 2499 |
+
},
|
| 2500 |
+
{
|
| 2501 |
+
"epoch": 0.2909090909090909,
|
| 2502 |
+
"eval_loss": 31.095945358276367,
|
| 2503 |
+
"eval_runtime": 2.2422,
|
| 2504 |
+
"eval_samples_per_second": 1.784,
|
| 2505 |
+
"eval_steps_per_second": 0.446,
|
| 2506 |
+
"step": 3200
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 0.2918181818181818,
|
| 2510 |
+
"grad_norm": 82796911525888.0,
|
| 2511 |
+
"learning_rate": 8.043051819262327e-07,
|
| 2512 |
+
"loss": 31.241659545898436,
|
| 2513 |
+
"step": 3210
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 0.2927272727272727,
|
| 2517 |
+
"grad_norm": 66819532521472.0,
|
| 2518 |
+
"learning_rate": 8.031708709884429e-07,
|
| 2519 |
+
"loss": 31.24657287597656,
|
| 2520 |
+
"step": 3220
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 0.29363636363636364,
|
| 2524 |
+
"grad_norm": 151652912857088.0,
|
| 2525 |
+
"learning_rate": 8.020340871791417e-07,
|
| 2526 |
+
"loss": 31.857916259765624,
|
| 2527 |
+
"step": 3230
|
| 2528 |
+
},
|
| 2529 |
+
{
|
| 2530 |
+
"epoch": 0.29454545454545455,
|
| 2531 |
+
"grad_norm": 33395514015744.0,
|
| 2532 |
+
"learning_rate": 8.008948397707252e-07,
|
| 2533 |
+
"loss": 31.493441772460937,
|
| 2534 |
+
"step": 3240
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 0.29545454545454547,
|
| 2538 |
+
"grad_norm": 186848542457856.0,
|
| 2539 |
+
"learning_rate": 7.997531380556833e-07,
|
| 2540 |
+
"loss": 31.500466918945314,
|
| 2541 |
+
"step": 3250
|
| 2542 |
+
},
|
| 2543 |
+
{
|
| 2544 |
+
"epoch": 0.2963636363636364,
|
| 2545 |
+
"grad_norm": 87677269442560.0,
|
| 2546 |
+
"learning_rate": 7.986089913465261e-07,
|
| 2547 |
+
"loss": 30.997119140625,
|
| 2548 |
+
"step": 3260
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 0.2972727272727273,
|
| 2552 |
+
"grad_norm": 73329570480128.0,
|
| 2553 |
+
"learning_rate": 7.974624089757064e-07,
|
| 2554 |
+
"loss": 31.911077880859374,
|
| 2555 |
+
"step": 3270
|
| 2556 |
+
},
|
| 2557 |
+
{
|
| 2558 |
+
"epoch": 0.29818181818181816,
|
| 2559 |
+
"grad_norm": 134268890120192.0,
|
| 2560 |
+
"learning_rate": 7.963134002955431e-07,
|
| 2561 |
+
"loss": 31.09755859375,
|
| 2562 |
+
"step": 3280
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 0.2990909090909091,
|
| 2566 |
+
"grad_norm": 57569158627328.0,
|
| 2567 |
+
"learning_rate": 7.951619746781472e-07,
|
| 2568 |
+
"loss": 31.29901123046875,
|
| 2569 |
+
"step": 3290
|
| 2570 |
+
},
|
| 2571 |
+
{
|
| 2572 |
+
"epoch": 0.3,
|
| 2573 |
+
"grad_norm": 58610235211776.0,
|
| 2574 |
+
"learning_rate": 7.940081415153428e-07,
|
| 2575 |
+
"loss": 31.30848388671875,
|
| 2576 |
+
"step": 3300
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"epoch": 0.3,
|
| 2580 |
+
"eval_loss": 31.100173950195312,
|
| 2581 |
+
"eval_runtime": 2.2403,
|
| 2582 |
+
"eval_samples_per_second": 1.785,
|
| 2583 |
+
"eval_steps_per_second": 0.446,
|
| 2584 |
+
"step": 3300
|
| 2585 |
+
},
|
| 2586 |
+
{
|
| 2587 |
+
"epoch": 0.3009090909090909,
|
| 2588 |
+
"grad_norm": 386970849116160.0,
|
| 2589 |
+
"learning_rate": 7.928519102185922e-07,
|
| 2590 |
+
"loss": 31.135516357421874,
|
| 2591 |
+
"step": 3310
|
| 2592 |
+
},
|
| 2593 |
+
{
|
| 2594 |
+
"epoch": 0.3018181818181818,
|
| 2595 |
+
"grad_norm": 288033458880512.0,
|
| 2596 |
+
"learning_rate": 7.91693290218918e-07,
|
| 2597 |
+
"loss": 31.321795654296874,
|
| 2598 |
+
"step": 3320
|
| 2599 |
+
},
|
| 2600 |
+
{
|
| 2601 |
+
"epoch": 0.30272727272727273,
|
| 2602 |
+
"grad_norm": 83541618589696.0,
|
| 2603 |
+
"learning_rate": 7.905322909668272e-07,
|
| 2604 |
+
"loss": 30.953836059570314,
|
| 2605 |
+
"step": 3330
|
| 2606 |
+
},
|
| 2607 |
+
{
|
| 2608 |
+
"epoch": 0.30363636363636365,
|
| 2609 |
+
"grad_norm": 77461295464448.0,
|
| 2610 |
+
"learning_rate": 7.893689219322336e-07,
|
| 2611 |
+
"loss": 31.53773193359375,
|
| 2612 |
+
"step": 3340
|
| 2613 |
+
},
|
| 2614 |
+
{
|
| 2615 |
+
"epoch": 0.30454545454545456,
|
| 2616 |
+
"grad_norm": 75457584168960.0,
|
| 2617 |
+
"learning_rate": 7.882031926043803e-07,
|
| 2618 |
+
"loss": 31.441778564453124,
|
| 2619 |
+
"step": 3350
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 0.3054545454545455,
|
| 2623 |
+
"grad_norm": 52945127211008.0,
|
| 2624 |
+
"learning_rate": 7.870351124917629e-07,
|
| 2625 |
+
"loss": 31.60285339355469,
|
| 2626 |
+
"step": 3360
|
| 2627 |
+
},
|
| 2628 |
+
{
|
| 2629 |
+
"epoch": 0.30636363636363634,
|
| 2630 |
+
"grad_norm": 81810528665600.0,
|
| 2631 |
+
"learning_rate": 7.858646911220512e-07,
|
| 2632 |
+
"loss": 31.419757080078124,
|
| 2633 |
+
"step": 3370
|
| 2634 |
+
},
|
| 2635 |
+
{
|
| 2636 |
+
"epoch": 0.30727272727272725,
|
| 2637 |
+
"grad_norm": 103102694817792.0,
|
| 2638 |
+
"learning_rate": 7.846919380420125e-07,
|
| 2639 |
+
"loss": 31.411013793945312,
|
| 2640 |
+
"step": 3380
|
| 2641 |
+
},
|
| 2642 |
+
{
|
| 2643 |
+
"epoch": 0.30818181818181817,
|
| 2644 |
+
"grad_norm": 181314292547584.0,
|
| 2645 |
+
"learning_rate": 7.835168628174325e-07,
|
| 2646 |
+
"loss": 30.9938232421875,
|
| 2647 |
+
"step": 3390
|
| 2648 |
+
},
|
| 2649 |
+
{
|
| 2650 |
+
"epoch": 0.3090909090909091,
|
| 2651 |
+
"grad_norm": 64268049317888.0,
|
| 2652 |
+
"learning_rate": 7.823394750330386e-07,
|
| 2653 |
+
"loss": 31.72479248046875,
|
| 2654 |
+
"step": 3400
|
| 2655 |
+
},
|
| 2656 |
+
{
|
| 2657 |
+
"epoch": 0.3090909090909091,
|
| 2658 |
+
"eval_loss": 31.11138153076172,
|
| 2659 |
+
"eval_runtime": 2.2524,
|
| 2660 |
+
"eval_samples_per_second": 1.776,
|
| 2661 |
+
"eval_steps_per_second": 0.444,
|
| 2662 |
+
"step": 3400
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.31,
|
| 2666 |
+
"grad_norm": 223006160846848.0,
|
| 2667 |
+
"learning_rate": 7.811597842924203e-07,
|
| 2668 |
+
"loss": 31.49818115234375,
|
| 2669 |
+
"step": 3410
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.3109090909090909,
|
| 2673 |
+
"grad_norm": 152988916121600.0,
|
| 2674 |
+
"learning_rate": 7.799778002179523e-07,
|
| 2675 |
+
"loss": 31.330535888671875,
|
| 2676 |
+
"step": 3420
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.3118181818181818,
|
| 2680 |
+
"grad_norm": 115639947624448.0,
|
| 2681 |
+
"learning_rate": 7.787935324507149e-07,
|
| 2682 |
+
"loss": 31.248580932617188,
|
| 2683 |
+
"step": 3430
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.31272727272727274,
|
| 2687 |
+
"grad_norm": 52082308546560.0,
|
| 2688 |
+
"learning_rate": 7.776069906504159e-07,
|
| 2689 |
+
"loss": 30.949420166015624,
|
| 2690 |
+
"step": 3440
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.31363636363636366,
|
| 2694 |
+
"grad_norm": 70850158002176.0,
|
| 2695 |
+
"learning_rate": 7.764181844953116e-07,
|
| 2696 |
+
"loss": 31.362625122070312,
|
| 2697 |
+
"step": 3450
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.3145454545454546,
|
| 2701 |
+
"grad_norm": 78591157075968.0,
|
| 2702 |
+
"learning_rate": 7.752271236821282e-07,
|
| 2703 |
+
"loss": 31.381210327148438,
|
| 2704 |
+
"step": 3460
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.31545454545454543,
|
| 2708 |
+
"grad_norm": 153667252518912.0,
|
| 2709 |
+
"learning_rate": 7.74033817925982e-07,
|
| 2710 |
+
"loss": 31.4740478515625,
|
| 2711 |
+
"step": 3470
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.31636363636363635,
|
| 2715 |
+
"grad_norm": 52006664273920.0,
|
| 2716 |
+
"learning_rate": 7.728382769603011e-07,
|
| 2717 |
+
"loss": 31.210354614257813,
|
| 2718 |
+
"step": 3480
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.31727272727272726,
|
| 2722 |
+
"grad_norm": 186123246632960.0,
|
| 2723 |
+
"learning_rate": 7.716405105367452e-07,
|
| 2724 |
+
"loss": 30.999267578125,
|
| 2725 |
+
"step": 3490
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.3181818181818182,
|
| 2729 |
+
"grad_norm": 232983285989376.0,
|
| 2730 |
+
"learning_rate": 7.704405284251267e-07,
|
| 2731 |
+
"loss": 30.719741821289062,
|
| 2732 |
+
"step": 3500
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.3181818181818182,
|
| 2736 |
+
"eval_loss": 31.12489891052246,
|
| 2737 |
+
"eval_runtime": 2.2583,
|
| 2738 |
+
"eval_samples_per_second": 1.771,
|
| 2739 |
+
"eval_steps_per_second": 0.443,
|
| 2740 |
+
"step": 3500
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 0.3190909090909091,
|
| 2744 |
+
"grad_norm": 142378786619392.0,
|
| 2745 |
+
"learning_rate": 7.6923834041333e-07,
|
| 2746 |
+
"loss": 31.163070678710938,
|
| 2747 |
+
"step": 3510
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 0.32,
|
| 2751 |
+
"grad_norm": 373457472716800.0,
|
| 2752 |
+
"learning_rate": 7.680339563072333e-07,
|
| 2753 |
+
"loss": 30.965121459960937,
|
| 2754 |
+
"step": 3520
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.3209090909090909,
|
| 2758 |
+
"grad_norm": 59119444688896.0,
|
| 2759 |
+
"learning_rate": 7.668273859306269e-07,
|
| 2760 |
+
"loss": 31.44709167480469,
|
| 2761 |
+
"step": 3530
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.32181818181818184,
|
| 2765 |
+
"grad_norm": 96141030981632.0,
|
| 2766 |
+
"learning_rate": 7.656186391251341e-07,
|
| 2767 |
+
"loss": 31.224566650390624,
|
| 2768 |
+
"step": 3540
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 0.32272727272727275,
|
| 2772 |
+
"grad_norm": 277230374617088.0,
|
| 2773 |
+
"learning_rate": 7.644077257501308e-07,
|
| 2774 |
+
"loss": 31.22548828125,
|
| 2775 |
+
"step": 3550
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 0.3236363636363636,
|
| 2779 |
+
"grad_norm": 238855160594432.0,
|
| 2780 |
+
"learning_rate": 7.631946556826647e-07,
|
| 2781 |
+
"loss": 31.30469665527344,
|
| 2782 |
+
"step": 3560
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 0.3245454545454545,
|
| 2786 |
+
"grad_norm": 65425522032640.0,
|
| 2787 |
+
"learning_rate": 7.619794388173751e-07,
|
| 2788 |
+
"loss": 31.599795532226562,
|
| 2789 |
+
"step": 3570
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 0.32545454545454544,
|
| 2793 |
+
"grad_norm": 51770529153024.0,
|
| 2794 |
+
"learning_rate": 7.60762085066412e-07,
|
| 2795 |
+
"loss": 30.914794921875,
|
| 2796 |
+
"step": 3580
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.32636363636363636,
|
| 2800 |
+
"grad_norm": 100869840306176.0,
|
| 2801 |
+
"learning_rate": 7.595426043593556e-07,
|
| 2802 |
+
"loss": 31.482501220703124,
|
| 2803 |
+
"step": 3590
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 0.32727272727272727,
|
| 2807 |
+
"grad_norm": 71489554481152.0,
|
| 2808 |
+
"learning_rate": 7.583210066431346e-07,
|
| 2809 |
+
"loss": 31.595437622070314,
|
| 2810 |
+
"step": 3600
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 0.32727272727272727,
|
| 2814 |
+
"eval_loss": 31.137432098388672,
|
| 2815 |
+
"eval_runtime": 2.2584,
|
| 2816 |
+
"eval_samples_per_second": 1.771,
|
| 2817 |
+
"eval_steps_per_second": 0.443,
|
| 2818 |
+
"step": 3600
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 0.3281818181818182,
|
| 2822 |
+
"grad_norm": 128388249419776.0,
|
| 2823 |
+
"learning_rate": 7.570973018819458e-07,
|
| 2824 |
+
"loss": 31.67999267578125,
|
| 2825 |
+
"step": 3610
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 0.3290909090909091,
|
| 2829 |
+
"grad_norm": 48917999779840.0,
|
| 2830 |
+
"learning_rate": 7.558715000571725e-07,
|
| 2831 |
+
"loss": 31.7310791015625,
|
| 2832 |
+
"step": 3620
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 0.33,
|
| 2836 |
+
"grad_norm": 81801485746176.0,
|
| 2837 |
+
"learning_rate": 7.546436111673027e-07,
|
| 2838 |
+
"loss": 31.565594482421876,
|
| 2839 |
+
"step": 3630
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 0.33090909090909093,
|
| 2843 |
+
"grad_norm": 70281771089920.0,
|
| 2844 |
+
"learning_rate": 7.534136452278486e-07,
|
| 2845 |
+
"loss": 31.4003173828125,
|
| 2846 |
+
"step": 3640
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 0.33181818181818185,
|
| 2850 |
+
"grad_norm": 268885236383744.0,
|
| 2851 |
+
"learning_rate": 7.521816122712637e-07,
|
| 2852 |
+
"loss": 30.847274780273438,
|
| 2853 |
+
"step": 3650
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 0.3327272727272727,
|
| 2857 |
+
"grad_norm": 188701753737216.0,
|
| 2858 |
+
"learning_rate": 7.509475223468618e-07,
|
| 2859 |
+
"loss": 31.30968017578125,
|
| 2860 |
+
"step": 3660
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 0.3336363636363636,
|
| 2864 |
+
"grad_norm": 203268269539328.0,
|
| 2865 |
+
"learning_rate": 7.497113855207347e-07,
|
| 2866 |
+
"loss": 31.730929565429687,
|
| 2867 |
+
"step": 3670
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 0.33454545454545453,
|
| 2871 |
+
"grad_norm": 40551332184064.0,
|
| 2872 |
+
"learning_rate": 7.4847321187567e-07,
|
| 2873 |
+
"loss": 31.19182434082031,
|
| 2874 |
+
"step": 3680
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 0.33545454545454545,
|
| 2878 |
+
"grad_norm": 120100707368960.0,
|
| 2879 |
+
"learning_rate": 7.472330115110696e-07,
|
| 2880 |
+
"loss": 31.216766357421875,
|
| 2881 |
+
"step": 3690
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 0.33636363636363636,
|
| 2885 |
+
"grad_norm": 123142987055104.0,
|
| 2886 |
+
"learning_rate": 7.459907945428657e-07,
|
| 2887 |
+
"loss": 31.759628295898438,
|
| 2888 |
+
"step": 3700
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 0.33636363636363636,
|
| 2892 |
+
"eval_loss": 31.1546630859375,
|
| 2893 |
+
"eval_runtime": 2.2729,
|
| 2894 |
+
"eval_samples_per_second": 1.76,
|
| 2895 |
+
"eval_steps_per_second": 0.44,
|
| 2896 |
+
"step": 3700
|
| 2897 |
+
},
|
| 2898 |
+
{
|
| 2899 |
+
"epoch": 0.3372727272727273,
|
| 2900 |
+
"grad_norm": 65411789881344.0,
|
| 2901 |
+
"learning_rate": 7.447465711034404e-07,
|
| 2902 |
+
"loss": 31.438787841796874,
|
| 2903 |
+
"step": 3710
|
| 2904 |
+
},
|
| 2905 |
+
{
|
| 2906 |
+
"epoch": 0.3381818181818182,
|
| 2907 |
+
"grad_norm": 179227156545536.0,
|
| 2908 |
+
"learning_rate": 7.43500351341541e-07,
|
| 2909 |
+
"loss": 31.2360595703125,
|
| 2910 |
+
"step": 3720
|
| 2911 |
+
},
|
| 2912 |
+
{
|
| 2913 |
+
"epoch": 0.3390909090909091,
|
| 2914 |
+
"grad_norm": 259052831506432.0,
|
| 2915 |
+
"learning_rate": 7.422521454221989e-07,
|
| 2916 |
+
"loss": 31.520187377929688,
|
| 2917 |
+
"step": 3730
|
| 2918 |
+
},
|
| 2919 |
+
{
|
| 2920 |
+
"epoch": 0.34,
|
| 2921 |
+
"grad_norm": 102227595231232.0,
|
| 2922 |
+
"learning_rate": 7.410019635266459e-07,
|
| 2923 |
+
"loss": 31.39177551269531,
|
| 2924 |
+
"step": 3740
|
| 2925 |
+
},
|
| 2926 |
+
{
|
| 2927 |
+
"epoch": 0.3409090909090909,
|
| 2928 |
+
"grad_norm": 68988482617344.0,
|
| 2929 |
+
"learning_rate": 7.397498158522306e-07,
|
| 2930 |
+
"loss": 31.17947998046875,
|
| 2931 |
+
"step": 3750
|
| 2932 |
+
},
|
| 2933 |
+
{
|
| 2934 |
+
"epoch": 0.3418181818181818,
|
| 2935 |
+
"grad_norm": 46946144223232.0,
|
| 2936 |
+
"learning_rate": 7.384957126123365e-07,
|
| 2937 |
+
"loss": 30.96776428222656,
|
| 2938 |
+
"step": 3760
|
| 2939 |
+
},
|
| 2940 |
+
{
|
| 2941 |
+
"epoch": 0.3427272727272727,
|
| 2942 |
+
"grad_norm": 178196901265408.0,
|
| 2943 |
+
"learning_rate": 7.37239664036298e-07,
|
| 2944 |
+
"loss": 31.246176147460936,
|
| 2945 |
+
"step": 3770
|
| 2946 |
+
},
|
| 2947 |
+
{
|
| 2948 |
+
"epoch": 0.34363636363636363,
|
| 2949 |
+
"grad_norm": 77915882520576.0,
|
| 2950 |
+
"learning_rate": 7.359816803693166e-07,
|
| 2951 |
+
"loss": 30.79324951171875,
|
| 2952 |
+
"step": 3780
|
| 2953 |
+
},
|
| 2954 |
+
{
|
| 2955 |
+
"epoch": 0.34454545454545454,
|
| 2956 |
+
"grad_norm": 78789707038720.0,
|
| 2957 |
+
"learning_rate": 7.347217718723783e-07,
|
| 2958 |
+
"loss": 31.38133850097656,
|
| 2959 |
+
"step": 3790
|
| 2960 |
+
},
|
| 2961 |
+
{
|
| 2962 |
+
"epoch": 0.34545454545454546,
|
| 2963 |
+
"grad_norm": 26140632678400.0,
|
| 2964 |
+
"learning_rate": 7.334599488221689e-07,
|
| 2965 |
+
"loss": 31.2718505859375,
|
| 2966 |
+
"step": 3800
|
| 2967 |
+
},
|
| 2968 |
+
{
|
| 2969 |
+
"epoch": 0.34545454545454546,
|
| 2970 |
+
"eval_loss": 31.167020797729492,
|
| 2971 |
+
"eval_runtime": 2.2566,
|
| 2972 |
+
"eval_samples_per_second": 1.773,
|
| 2973 |
+
"eval_steps_per_second": 0.443,
|
| 2974 |
+
"step": 3800
|
| 2975 |
+
},
|
| 2976 |
+
{
|
| 2977 |
+
"epoch": 0.3463636363636364,
|
| 2978 |
+
"grad_norm": 154810032586752.0,
|
| 2979 |
+
"learning_rate": 7.321962215109906e-07,
|
| 2980 |
+
"loss": 30.819577026367188,
|
| 2981 |
+
"step": 3810
|
| 2982 |
+
},
|
| 2983 |
+
{
|
| 2984 |
+
"epoch": 0.3472727272727273,
|
| 2985 |
+
"grad_norm": 118487854874624.0,
|
| 2986 |
+
"learning_rate": 7.309306002466787e-07,
|
| 2987 |
+
"loss": 31.2808837890625,
|
| 2988 |
+
"step": 3820
|
| 2989 |
+
},
|
| 2990 |
+
{
|
| 2991 |
+
"epoch": 0.3481818181818182,
|
| 2992 |
+
"grad_norm": 108455390085120.0,
|
| 2993 |
+
"learning_rate": 7.296630953525158e-07,
|
| 2994 |
+
"loss": 31.507012939453126,
|
| 2995 |
+
"step": 3830
|
| 2996 |
+
},
|
| 2997 |
+
{
|
| 2998 |
+
"epoch": 0.3490909090909091,
|
| 2999 |
+
"grad_norm": 174323612516352.0,
|
| 3000 |
+
"learning_rate": 7.283937171671497e-07,
|
| 3001 |
+
"loss": 31.171493530273438,
|
| 3002 |
+
"step": 3840
|
| 3003 |
+
},
|
| 3004 |
+
{
|
| 3005 |
+
"epoch": 0.35,
|
| 3006 |
+
"grad_norm": 52621549240320.0,
|
| 3007 |
+
"learning_rate": 7.271224760445079e-07,
|
| 3008 |
+
"loss": 31.58785400390625,
|
| 3009 |
+
"step": 3850
|
| 3010 |
+
},
|
| 3011 |
+
{
|
| 3012 |
+
"epoch": 0.3509090909090909,
|
| 3013 |
+
"grad_norm": 245216510476288.0,
|
| 3014 |
+
"learning_rate": 7.258493823537124e-07,
|
| 3015 |
+
"loss": 31.41130676269531,
|
| 3016 |
+
"step": 3860
|
| 3017 |
+
},
|
| 3018 |
+
{
|
| 3019 |
+
"epoch": 0.3518181818181818,
|
| 3020 |
+
"grad_norm": 40472269553664.0,
|
| 3021 |
+
"learning_rate": 7.245744464789973e-07,
|
| 3022 |
+
"loss": 31.25943603515625,
|
| 3023 |
+
"step": 3870
|
| 3024 |
+
},
|
| 3025 |
+
{
|
| 3026 |
+
"epoch": 0.3527272727272727,
|
| 3027 |
+
"grad_norm": 152244192280576.0,
|
| 3028 |
+
"learning_rate": 7.232976788196221e-07,
|
| 3029 |
+
"loss": 31.55677490234375,
|
| 3030 |
+
"step": 3880
|
| 3031 |
+
},
|
| 3032 |
+
{
|
| 3033 |
+
"epoch": 0.35363636363636364,
|
| 3034 |
+
"grad_norm": 163018369400832.0,
|
| 3035 |
+
"learning_rate": 7.220190897897877e-07,
|
| 3036 |
+
"loss": 31.163226318359374,
|
| 3037 |
+
"step": 3890
|
| 3038 |
+
},
|
| 3039 |
+
{
|
| 3040 |
+
"epoch": 0.35454545454545455,
|
| 3041 |
+
"grad_norm": 31649326694400.0,
|
| 3042 |
+
"learning_rate": 7.207386898185515e-07,
|
| 3043 |
+
"loss": 31.222012329101563,
|
| 3044 |
+
"step": 3900
|
| 3045 |
+
},
|
| 3046 |
+
{
|
| 3047 |
+
"epoch": 0.35454545454545455,
|
| 3048 |
+
"eval_loss": 31.183151245117188,
|
| 3049 |
+
"eval_runtime": 2.2888,
|
| 3050 |
+
"eval_samples_per_second": 1.748,
|
| 3051 |
+
"eval_steps_per_second": 0.437,
|
| 3052 |
+
"step": 3900
|
| 3053 |
+
},
|
| 3054 |
+
{
|
| 3055 |
+
"epoch": 0.35545454545454547,
|
| 3056 |
+
"grad_norm": 120699922415616.0,
|
| 3057 |
+
"learning_rate": 7.194564893497419e-07,
|
| 3058 |
+
"loss": 31.556930541992188,
|
| 3059 |
+
"step": 3910
|
| 3060 |
+
},
|
| 3061 |
+
{
|
| 3062 |
+
"epoch": 0.3563636363636364,
|
| 3063 |
+
"grad_norm": 143552671645696.0,
|
| 3064 |
+
"learning_rate": 7.181724988418737e-07,
|
| 3065 |
+
"loss": 31.260910034179688,
|
| 3066 |
+
"step": 3920
|
| 3067 |
+
},
|
| 3068 |
+
{
|
| 3069 |
+
"epoch": 0.3572727272727273,
|
| 3070 |
+
"grad_norm": 66354916884480.0,
|
| 3071 |
+
"learning_rate": 7.168867287680627e-07,
|
| 3072 |
+
"loss": 31.252059936523438,
|
| 3073 |
+
"step": 3930
|
| 3074 |
+
},
|
| 3075 |
+
{
|
| 3076 |
+
"epoch": 0.35818181818181816,
|
| 3077 |
+
"grad_norm": 142235978956800.0,
|
| 3078 |
+
"learning_rate": 7.155991896159392e-07,
|
| 3079 |
+
"loss": 31.511947631835938,
|
| 3080 |
+
"step": 3940
|
| 3081 |
+
},
|
| 3082 |
+
{
|
| 3083 |
+
"epoch": 0.35909090909090907,
|
| 3084 |
+
"grad_norm": 331125603958784.0,
|
| 3085 |
+
"learning_rate": 7.143098918875642e-07,
|
| 3086 |
+
"loss": 31.4046630859375,
|
| 3087 |
+
"step": 3950
|
| 3088 |
+
},
|
| 3089 |
+
{
|
| 3090 |
+
"epoch": 0.36,
|
| 3091 |
+
"grad_norm": 163280043638784.0,
|
| 3092 |
+
"learning_rate": 7.130188460993426e-07,
|
| 3093 |
+
"loss": 31.372503662109374,
|
| 3094 |
+
"step": 3960
|
| 3095 |
+
},
|
| 3096 |
+
{
|
| 3097 |
+
"epoch": 0.3609090909090909,
|
| 3098 |
+
"grad_norm": 78036493926400.0,
|
| 3099 |
+
"learning_rate": 7.117260627819376e-07,
|
| 3100 |
+
"loss": 31.743682861328125,
|
| 3101 |
+
"step": 3970
|
| 3102 |
+
},
|
| 3103 |
+
{
|
| 3104 |
+
"epoch": 0.3618181818181818,
|
| 3105 |
+
"grad_norm": 141398275784704.0,
|
| 3106 |
+
"learning_rate": 7.104315524801848e-07,
|
| 3107 |
+
"loss": 31.348150634765624,
|
| 3108 |
+
"step": 3980
|
| 3109 |
+
},
|
| 3110 |
+
{
|
| 3111 |
+
"epoch": 0.36272727272727273,
|
| 3112 |
+
"grad_norm": 34922429415424.0,
|
| 3113 |
+
"learning_rate": 7.091353257530068e-07,
|
| 3114 |
+
"loss": 31.071649169921876,
|
| 3115 |
+
"step": 3990
|
| 3116 |
+
},
|
| 3117 |
+
{
|
| 3118 |
+
"epoch": 0.36363636363636365,
|
| 3119 |
+
"grad_norm": 114211183132672.0,
|
| 3120 |
+
"learning_rate": 7.078373931733258e-07,
|
| 3121 |
+
"loss": 31.188916015625,
|
| 3122 |
+
"step": 4000
|
| 3123 |
+
},
|
| 3124 |
+
{
|
| 3125 |
+
"epoch": 0.36363636363636365,
|
| 3126 |
+
"eval_loss": 31.198307037353516,
|
| 3127 |
+
"eval_runtime": 2.261,
|
| 3128 |
+
"eval_samples_per_second": 1.769,
|
| 3129 |
+
"eval_steps_per_second": 0.442,
|
| 3130 |
+
"step": 4000
|
| 3131 |
}
|
| 3132 |
],
|
| 3133 |
"logging_steps": 10,
|
|
|
|
| 3147 |
"attributes": {}
|
| 3148 |
}
|
| 3149 |
},
|
| 3150 |
+
"total_flos": 1.8119695269888e+16,
|
| 3151 |
"train_batch_size": 1,
|
| 3152 |
"trial_name": null,
|
| 3153 |
"trial_params": null
|