Training in progress, step 10000, 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 sha256:2e37ecaf23c566abac1c5ee98c4146e26d8c4ba5c6a88161df70c0d2b752dc7c
|
| 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 sha256:2d46c8aaa115fcd4327ec8b1b175907bb3814e7aa02b98c3227dfcc9b5bf5f4b
|
| 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 sha256:213d68056a85c0df08e945b26d69348b0399610f7d9d5c7d627e4f131e544345
|
| 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:dd7a47e842972f28f9f116aeec69116f078e9141f9b533871ca95d65f1a4b341
|
| 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:d5fb0549dec6485ba72461a3921f0d8d9f46d9566606316de09ad419d70b65e3
|
| 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 sha256:082abf7017f5f44577de9434942671a6c463a6ee15d5e82ccef96f18aa764420
|
| 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:b26a3cf6ee652fa5262f0bc3dc8594e75f5cf498d42d683da2a1c0f55da32f4c
|
| 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 sha256:1174c09b73b86c6cef6b37654360d23f03f1591b209dc8550fe82e6616fdb483
|
| 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 sha256:4f3beb1efbd57c929e4bc9609eea220ab72e734b7467fca176c3d751540705e8
|
| 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 sha256:757b50a256539469371ecdb8e378d979cfd235031b7a187d1341d66808f7ab34
|
| 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:f407668770ce0d5844aba1969955a95b4faf45e4ab15e542f507c7d79348f384
|
| 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,
|
|
@@ -7028,6 +7028,786 @@
|
|
| 7028 |
"eval_samples_per_second": 16.401,
|
| 7029 |
"eval_steps_per_second": 0.911,
|
| 7030 |
"step": 9000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7031 |
}
|
| 7032 |
],
|
| 7033 |
"logging_steps": 10,
|
|
@@ -7047,7 +7827,7 @@
|
|
| 7047 |
"attributes": {}
|
| 7048 |
}
|
| 7049 |
},
|
| 7050 |
-
"total_flos":
|
| 7051 |
"train_batch_size": 4,
|
| 7052 |
"trial_name": null,
|
| 7053 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.9090909090909091,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 10000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": false,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 7028 |
"eval_samples_per_second": 16.401,
|
| 7029 |
"eval_steps_per_second": 0.911,
|
| 7030 |
"step": 9000
|
| 7031 |
+
},
|
| 7032 |
+
{
|
| 7033 |
+
"epoch": 0.8190909090909091,
|
| 7034 |
+
"grad_norm": 216634107101184.0,
|
| 7035 |
+
"learning_rate": 7.867979391978396e-08,
|
| 7036 |
+
"loss": 26.5111083984375,
|
| 7037 |
+
"step": 9010
|
| 7038 |
+
},
|
| 7039 |
+
{
|
| 7040 |
+
"epoch": 0.82,
|
| 7041 |
+
"grad_norm": 460943641280512.0,
|
| 7042 |
+
"learning_rate": 7.791257035266719e-08,
|
| 7043 |
+
"loss": 26.373800659179686,
|
| 7044 |
+
"step": 9020
|
| 7045 |
+
},
|
| 7046 |
+
{
|
| 7047 |
+
"epoch": 0.8209090909090909,
|
| 7048 |
+
"grad_norm": 207550905581568.0,
|
| 7049 |
+
"learning_rate": 7.71487896228411e-08,
|
| 7050 |
+
"loss": 26.439486694335937,
|
| 7051 |
+
"step": 9030
|
| 7052 |
+
},
|
| 7053 |
+
{
|
| 7054 |
+
"epoch": 0.8218181818181818,
|
| 7055 |
+
"grad_norm": 164937984901120.0,
|
| 7056 |
+
"learning_rate": 7.638845796023002e-08,
|
| 7057 |
+
"loss": 26.551513671875,
|
| 7058 |
+
"step": 9040
|
| 7059 |
+
},
|
| 7060 |
+
{
|
| 7061 |
+
"epoch": 0.8227272727272728,
|
| 7062 |
+
"grad_norm": 467092792934400.0,
|
| 7063 |
+
"learning_rate": 7.563158156662552e-08,
|
| 7064 |
+
"loss": 26.31507568359375,
|
| 7065 |
+
"step": 9050
|
| 7066 |
+
},
|
| 7067 |
+
{
|
| 7068 |
+
"epoch": 0.8236363636363636,
|
| 7069 |
+
"grad_norm": 270167032463360.0,
|
| 7070 |
+
"learning_rate": 7.487816661563523e-08,
|
| 7071 |
+
"loss": 26.83199462890625,
|
| 7072 |
+
"step": 9060
|
| 7073 |
+
},
|
| 7074 |
+
{
|
| 7075 |
+
"epoch": 0.8245454545454546,
|
| 7076 |
+
"grad_norm": 329480295940096.0,
|
| 7077 |
+
"learning_rate": 7.412821925263324e-08,
|
| 7078 |
+
"loss": 26.249203491210938,
|
| 7079 |
+
"step": 9070
|
| 7080 |
+
},
|
| 7081 |
+
{
|
| 7082 |
+
"epoch": 0.8254545454545454,
|
| 7083 |
+
"grad_norm": 176404423507968.0,
|
| 7084 |
+
"learning_rate": 7.338174559470933e-08,
|
| 7085 |
+
"loss": 26.442239379882814,
|
| 7086 |
+
"step": 9080
|
| 7087 |
+
},
|
| 7088 |
+
{
|
| 7089 |
+
"epoch": 0.8263636363636364,
|
| 7090 |
+
"grad_norm": 232708089315328.0,
|
| 7091 |
+
"learning_rate": 7.263875173061968e-08,
|
| 7092 |
+
"loss": 26.62685546875,
|
| 7093 |
+
"step": 9090
|
| 7094 |
+
},
|
| 7095 |
+
{
|
| 7096 |
+
"epoch": 0.8272727272727273,
|
| 7097 |
+
"grad_norm": 264557184417792.0,
|
| 7098 |
+
"learning_rate": 7.189924372073631e-08,
|
| 7099 |
+
"loss": 26.332183837890625,
|
| 7100 |
+
"step": 9100
|
| 7101 |
+
},
|
| 7102 |
+
{
|
| 7103 |
+
"epoch": 0.8272727272727273,
|
| 7104 |
+
"eval_loss": 26.257953643798828,
|
| 7105 |
+
"eval_runtime": 2.2005,
|
| 7106 |
+
"eval_samples_per_second": 16.36,
|
| 7107 |
+
"eval_steps_per_second": 0.909,
|
| 7108 |
+
"step": 9100
|
| 7109 |
+
},
|
| 7110 |
+
{
|
| 7111 |
+
"epoch": 0.8281818181818181,
|
| 7112 |
+
"grad_norm": 170234988199936.0,
|
| 7113 |
+
"learning_rate": 7.116322759699867e-08,
|
| 7114 |
+
"loss": 26.33367919921875,
|
| 7115 |
+
"step": 9110
|
| 7116 |
+
},
|
| 7117 |
+
{
|
| 7118 |
+
"epoch": 0.8290909090909091,
|
| 7119 |
+
"grad_norm": 260837423972352.0,
|
| 7120 |
+
"learning_rate": 7.043070936286393e-08,
|
| 7121 |
+
"loss": 26.51810302734375,
|
| 7122 |
+
"step": 9120
|
| 7123 |
+
},
|
| 7124 |
+
{
|
| 7125 |
+
"epoch": 0.83,
|
| 7126 |
+
"grad_norm": 208293901369344.0,
|
| 7127 |
+
"learning_rate": 6.970169499325757e-08,
|
| 7128 |
+
"loss": 26.384710693359374,
|
| 7129 |
+
"step": 9130
|
| 7130 |
+
},
|
| 7131 |
+
{
|
| 7132 |
+
"epoch": 0.8309090909090909,
|
| 7133 |
+
"grad_norm": 140174285602816.0,
|
| 7134 |
+
"learning_rate": 6.89761904345259e-08,
|
| 7135 |
+
"loss": 26.527005004882813,
|
| 7136 |
+
"step": 9140
|
| 7137 |
+
},
|
| 7138 |
+
{
|
| 7139 |
+
"epoch": 0.8318181818181818,
|
| 7140 |
+
"grad_norm": 260585463742464.0,
|
| 7141 |
+
"learning_rate": 6.825420160438617e-08,
|
| 7142 |
+
"loss": 26.76141357421875,
|
| 7143 |
+
"step": 9150
|
| 7144 |
+
},
|
| 7145 |
+
{
|
| 7146 |
+
"epoch": 0.8327272727272728,
|
| 7147 |
+
"grad_norm": 127458984591360.0,
|
| 7148 |
+
"learning_rate": 6.753573439187915e-08,
|
| 7149 |
+
"loss": 26.833413696289064,
|
| 7150 |
+
"step": 9160
|
| 7151 |
+
},
|
| 7152 |
+
{
|
| 7153 |
+
"epoch": 0.8336363636363636,
|
| 7154 |
+
"grad_norm": 94211844079616.0,
|
| 7155 |
+
"learning_rate": 6.682079465732076e-08,
|
| 7156 |
+
"loss": 26.543435668945314,
|
| 7157 |
+
"step": 9170
|
| 7158 |
+
},
|
| 7159 |
+
{
|
| 7160 |
+
"epoch": 0.8345454545454546,
|
| 7161 |
+
"grad_norm": 84640207470592.0,
|
| 7162 |
+
"learning_rate": 6.61093882322547e-08,
|
| 7163 |
+
"loss": 25.99791564941406,
|
| 7164 |
+
"step": 9180
|
| 7165 |
+
},
|
| 7166 |
+
{
|
| 7167 |
+
"epoch": 0.8354545454545454,
|
| 7168 |
+
"grad_norm": 375480100323328.0,
|
| 7169 |
+
"learning_rate": 6.540152091940393e-08,
|
| 7170 |
+
"loss": 26.548089599609376,
|
| 7171 |
+
"step": 9190
|
| 7172 |
+
},
|
| 7173 |
+
{
|
| 7174 |
+
"epoch": 0.8363636363636363,
|
| 7175 |
+
"grad_norm": 322556707995648.0,
|
| 7176 |
+
"learning_rate": 6.469719849262445e-08,
|
| 7177 |
+
"loss": 26.38921203613281,
|
| 7178 |
+
"step": 9200
|
| 7179 |
+
},
|
| 7180 |
+
{
|
| 7181 |
+
"epoch": 0.8363636363636363,
|
| 7182 |
+
"eval_loss": 26.251873016357422,
|
| 7183 |
+
"eval_runtime": 2.1916,
|
| 7184 |
+
"eval_samples_per_second": 16.426,
|
| 7185 |
+
"eval_steps_per_second": 0.913,
|
| 7186 |
+
"step": 9200
|
| 7187 |
+
},
|
| 7188 |
+
{
|
| 7189 |
+
"epoch": 0.8372727272727273,
|
| 7190 |
+
"grad_norm": 52027614822400.0,
|
| 7191 |
+
"learning_rate": 6.399642669685773e-08,
|
| 7192 |
+
"loss": 27.601669311523438,
|
| 7193 |
+
"step": 9210
|
| 7194 |
+
},
|
| 7195 |
+
{
|
| 7196 |
+
"epoch": 0.8381818181818181,
|
| 7197 |
+
"grad_norm": 116149614280704.0,
|
| 7198 |
+
"learning_rate": 6.329921124808335e-08,
|
| 7199 |
+
"loss": 26.614300537109376,
|
| 7200 |
+
"step": 9220
|
| 7201 |
+
},
|
| 7202 |
+
{
|
| 7203 |
+
"epoch": 0.8390909090909091,
|
| 7204 |
+
"grad_norm": 124626520768512.0,
|
| 7205 |
+
"learning_rate": 6.260555783327364e-08,
|
| 7206 |
+
"loss": 26.363275146484376,
|
| 7207 |
+
"step": 9230
|
| 7208 |
+
},
|
| 7209 |
+
{
|
| 7210 |
+
"epoch": 0.84,
|
| 7211 |
+
"grad_norm": 194277174935552.0,
|
| 7212 |
+
"learning_rate": 6.19154721103457e-08,
|
| 7213 |
+
"loss": 26.312432861328126,
|
| 7214 |
+
"step": 9240
|
| 7215 |
+
},
|
| 7216 |
+
{
|
| 7217 |
+
"epoch": 0.8409090909090909,
|
| 7218 |
+
"grad_norm": 203244177457152.0,
|
| 7219 |
+
"learning_rate": 6.122895970811659e-08,
|
| 7220 |
+
"loss": 26.483282470703124,
|
| 7221 |
+
"step": 9250
|
| 7222 |
+
},
|
| 7223 |
+
{
|
| 7224 |
+
"epoch": 0.8418181818181818,
|
| 7225 |
+
"grad_norm": 96889949126656.0,
|
| 7226 |
+
"learning_rate": 6.054602622625676e-08,
|
| 7227 |
+
"loss": 26.408782958984375,
|
| 7228 |
+
"step": 9260
|
| 7229 |
+
},
|
| 7230 |
+
{
|
| 7231 |
+
"epoch": 0.8427272727272728,
|
| 7232 |
+
"grad_norm": 228295429849088.0,
|
| 7233 |
+
"learning_rate": 5.986667723524452e-08,
|
| 7234 |
+
"loss": 26.440960693359376,
|
| 7235 |
+
"step": 9270
|
| 7236 |
+
},
|
| 7237 |
+
{
|
| 7238 |
+
"epoch": 0.8436363636363636,
|
| 7239 |
+
"grad_norm": 259944691531776.0,
|
| 7240 |
+
"learning_rate": 5.919091827632039e-08,
|
| 7241 |
+
"loss": 26.412326049804687,
|
| 7242 |
+
"step": 9280
|
| 7243 |
+
},
|
| 7244 |
+
{
|
| 7245 |
+
"epoch": 0.8445454545454546,
|
| 7246 |
+
"grad_norm": 123276919570432.0,
|
| 7247 |
+
"learning_rate": 5.8518754861442424e-08,
|
| 7248 |
+
"loss": 26.436785888671874,
|
| 7249 |
+
"step": 9290
|
| 7250 |
+
},
|
| 7251 |
+
{
|
| 7252 |
+
"epoch": 0.8454545454545455,
|
| 7253 |
+
"grad_norm": 175674060963840.0,
|
| 7254 |
+
"learning_rate": 5.78501924732408e-08,
|
| 7255 |
+
"loss": 26.9662353515625,
|
| 7256 |
+
"step": 9300
|
| 7257 |
+
},
|
| 7258 |
+
{
|
| 7259 |
+
"epoch": 0.8454545454545455,
|
| 7260 |
+
"eval_loss": 26.25527000427246,
|
| 7261 |
+
"eval_runtime": 2.1682,
|
| 7262 |
+
"eval_samples_per_second": 16.604,
|
| 7263 |
+
"eval_steps_per_second": 0.922,
|
| 7264 |
+
"step": 9300
|
| 7265 |
+
},
|
| 7266 |
+
{
|
| 7267 |
+
"epoch": 0.8463636363636363,
|
| 7268 |
+
"grad_norm": 107211535679488.0,
|
| 7269 |
+
"learning_rate": 5.7185236564973035e-08,
|
| 7270 |
+
"loss": 26.846856689453126,
|
| 7271 |
+
"step": 9310
|
| 7272 |
+
},
|
| 7273 |
+
{
|
| 7274 |
+
"epoch": 0.8472727272727273,
|
| 7275 |
+
"grad_norm": 466499886120960.0,
|
| 7276 |
+
"learning_rate": 5.6523892560479957e-08,
|
| 7277 |
+
"loss": 26.620623779296874,
|
| 7278 |
+
"step": 9320
|
| 7279 |
+
},
|
| 7280 |
+
{
|
| 7281 |
+
"epoch": 0.8481818181818181,
|
| 7282 |
+
"grad_norm": 94389649014784.0,
|
| 7283 |
+
"learning_rate": 5.586616585414117e-08,
|
| 7284 |
+
"loss": 26.862652587890626,
|
| 7285 |
+
"step": 9330
|
| 7286 |
+
},
|
| 7287 |
+
{
|
| 7288 |
+
"epoch": 0.8490909090909091,
|
| 7289 |
+
"grad_norm": 437626666483712.0,
|
| 7290 |
+
"learning_rate": 5.52120618108311e-08,
|
| 7291 |
+
"loss": 26.546505737304688,
|
| 7292 |
+
"step": 9340
|
| 7293 |
+
},
|
| 7294 |
+
{
|
| 7295 |
+
"epoch": 0.85,
|
| 7296 |
+
"grad_norm": 238070909632512.0,
|
| 7297 |
+
"learning_rate": 5.456158576587488e-08,
|
| 7298 |
+
"loss": 26.345050048828124,
|
| 7299 |
+
"step": 9350
|
| 7300 |
+
},
|
| 7301 |
+
{
|
| 7302 |
+
"epoch": 0.850909090909091,
|
| 7303 |
+
"grad_norm": 214509339803648.0,
|
| 7304 |
+
"learning_rate": 5.391474302500581e-08,
|
| 7305 |
+
"loss": 26.43468322753906,
|
| 7306 |
+
"step": 9360
|
| 7307 |
+
},
|
| 7308 |
+
{
|
| 7309 |
+
"epoch": 0.8518181818181818,
|
| 7310 |
+
"grad_norm": 208559971237888.0,
|
| 7311 |
+
"learning_rate": 5.32715388643209e-08,
|
| 7312 |
+
"loss": 26.35069885253906,
|
| 7313 |
+
"step": 9370
|
| 7314 |
+
},
|
| 7315 |
+
{
|
| 7316 |
+
"epoch": 0.8527272727272728,
|
| 7317 |
+
"grad_norm": 621268160217088.0,
|
| 7318 |
+
"learning_rate": 5.26319785302386e-08,
|
| 7319 |
+
"loss": 26.375735473632812,
|
| 7320 |
+
"step": 9380
|
| 7321 |
+
},
|
| 7322 |
+
{
|
| 7323 |
+
"epoch": 0.8536363636363636,
|
| 7324 |
+
"grad_norm": 419072072220672.0,
|
| 7325 |
+
"learning_rate": 5.199606723945593e-08,
|
| 7326 |
+
"loss": 26.826263427734375,
|
| 7327 |
+
"step": 9390
|
| 7328 |
+
},
|
| 7329 |
+
{
|
| 7330 |
+
"epoch": 0.8545454545454545,
|
| 7331 |
+
"grad_norm": 682285284196352.0,
|
| 7332 |
+
"learning_rate": 5.1363810178905344e-08,
|
| 7333 |
+
"loss": 26.34708251953125,
|
| 7334 |
+
"step": 9400
|
| 7335 |
+
},
|
| 7336 |
+
{
|
| 7337 |
+
"epoch": 0.8545454545454545,
|
| 7338 |
+
"eval_loss": 26.255863189697266,
|
| 7339 |
+
"eval_runtime": 2.1871,
|
| 7340 |
+
"eval_samples_per_second": 16.46,
|
| 7341 |
+
"eval_steps_per_second": 0.914,
|
| 7342 |
+
"step": 9400
|
| 7343 |
+
},
|
| 7344 |
+
{
|
| 7345 |
+
"epoch": 0.8554545454545455,
|
| 7346 |
+
"grad_norm": 89044629323776.0,
|
| 7347 |
+
"learning_rate": 5.073521250571328e-08,
|
| 7348 |
+
"loss": 26.50580749511719,
|
| 7349 |
+
"step": 9410
|
| 7350 |
+
},
|
| 7351 |
+
{
|
| 7352 |
+
"epoch": 0.8563636363636363,
|
| 7353 |
+
"grad_norm": 216007075430400.0,
|
| 7354 |
+
"learning_rate": 5.0110279347157555e-08,
|
| 7355 |
+
"loss": 26.41748046875,
|
| 7356 |
+
"step": 9420
|
| 7357 |
+
},
|
| 7358 |
+
{
|
| 7359 |
+
"epoch": 0.8572727272727273,
|
| 7360 |
+
"grad_norm": 474563418783744.0,
|
| 7361 |
+
"learning_rate": 4.9489015800625755e-08,
|
| 7362 |
+
"loss": 26.493084716796876,
|
| 7363 |
+
"step": 9430
|
| 7364 |
+
},
|
| 7365 |
+
{
|
| 7366 |
+
"epoch": 0.8581818181818182,
|
| 7367 |
+
"grad_norm": 379049620799488.0,
|
| 7368 |
+
"learning_rate": 4.887142693357332e-08,
|
| 7369 |
+
"loss": 26.464553833007812,
|
| 7370 |
+
"step": 9440
|
| 7371 |
+
},
|
| 7372 |
+
{
|
| 7373 |
+
"epoch": 0.8590909090909091,
|
| 7374 |
+
"grad_norm": 80948934737920.0,
|
| 7375 |
+
"learning_rate": 4.825751778348258e-08,
|
| 7376 |
+
"loss": 26.62518310546875,
|
| 7377 |
+
"step": 9450
|
| 7378 |
+
},
|
| 7379 |
+
{
|
| 7380 |
+
"epoch": 0.86,
|
| 7381 |
+
"grad_norm": 145902236860416.0,
|
| 7382 |
+
"learning_rate": 4.76472933578217e-08,
|
| 7383 |
+
"loss": 26.272061157226563,
|
| 7384 |
+
"step": 9460
|
| 7385 |
+
},
|
| 7386 |
+
{
|
| 7387 |
+
"epoch": 0.860909090909091,
|
| 7388 |
+
"grad_norm": 443239853195264.0,
|
| 7389 |
+
"learning_rate": 4.7040758634003585e-08,
|
| 7390 |
+
"loss": 26.57243957519531,
|
| 7391 |
+
"step": 9470
|
| 7392 |
+
},
|
| 7393 |
+
{
|
| 7394 |
+
"epoch": 0.8618181818181818,
|
| 7395 |
+
"grad_norm": 133928698511360.0,
|
| 7396 |
+
"learning_rate": 4.6437918559345257e-08,
|
| 7397 |
+
"loss": 26.390667724609376,
|
| 7398 |
+
"step": 9480
|
| 7399 |
+
},
|
| 7400 |
+
{
|
| 7401 |
+
"epoch": 0.8627272727272727,
|
| 7402 |
+
"grad_norm": 188859761557504.0,
|
| 7403 |
+
"learning_rate": 4.583877805102765e-08,
|
| 7404 |
+
"loss": 26.467236328125,
|
| 7405 |
+
"step": 9490
|
| 7406 |
+
},
|
| 7407 |
+
{
|
| 7408 |
+
"epoch": 0.8636363636363636,
|
| 7409 |
+
"grad_norm": 305761909473280.0,
|
| 7410 |
+
"learning_rate": 4.5243341996055785e-08,
|
| 7411 |
+
"loss": 26.252215576171874,
|
| 7412 |
+
"step": 9500
|
| 7413 |
+
},
|
| 7414 |
+
{
|
| 7415 |
+
"epoch": 0.8636363636363636,
|
| 7416 |
+
"eval_loss": 26.259260177612305,
|
| 7417 |
+
"eval_runtime": 2.1624,
|
| 7418 |
+
"eval_samples_per_second": 16.648,
|
| 7419 |
+
"eval_steps_per_second": 0.925,
|
| 7420 |
+
"step": 9500
|
| 7421 |
+
},
|
| 7422 |
+
{
|
| 7423 |
+
"epoch": 0.8645454545454545,
|
| 7424 |
+
"grad_norm": 203417435766784.0,
|
| 7425 |
+
"learning_rate": 4.465161525121797e-08,
|
| 7426 |
+
"loss": 26.584759521484376,
|
| 7427 |
+
"step": 9510
|
| 7428 |
+
},
|
| 7429 |
+
{
|
| 7430 |
+
"epoch": 0.8654545454545455,
|
| 7431 |
+
"grad_norm": 80916454047744.0,
|
| 7432 |
+
"learning_rate": 4.406360264304759e-08,
|
| 7433 |
+
"loss": 26.7392333984375,
|
| 7434 |
+
"step": 9520
|
| 7435 |
+
},
|
| 7436 |
+
{
|
| 7437 |
+
"epoch": 0.8663636363636363,
|
| 7438 |
+
"grad_norm": 93687598022656.0,
|
| 7439 |
+
"learning_rate": 4.3479308967782345e-08,
|
| 7440 |
+
"loss": 26.5811767578125,
|
| 7441 |
+
"step": 9530
|
| 7442 |
+
},
|
| 7443 |
+
{
|
| 7444 |
+
"epoch": 0.8672727272727273,
|
| 7445 |
+
"grad_norm": 391817216393216.0,
|
| 7446 |
+
"learning_rate": 4.289873899132601e-08,
|
| 7447 |
+
"loss": 26.5900146484375,
|
| 7448 |
+
"step": 9540
|
| 7449 |
+
},
|
| 7450 |
+
{
|
| 7451 |
+
"epoch": 0.8681818181818182,
|
| 7452 |
+
"grad_norm": 220526203109376.0,
|
| 7453 |
+
"learning_rate": 4.232189744920922e-08,
|
| 7454 |
+
"loss": 26.462973022460936,
|
| 7455 |
+
"step": 9550
|
| 7456 |
+
},
|
| 7457 |
+
{
|
| 7458 |
+
"epoch": 0.8690909090909091,
|
| 7459 |
+
"grad_norm": 294352429514752.0,
|
| 7460 |
+
"learning_rate": 4.1748789046551046e-08,
|
| 7461 |
+
"loss": 26.616650390625,
|
| 7462 |
+
"step": 9560
|
| 7463 |
+
},
|
| 7464 |
+
{
|
| 7465 |
+
"epoch": 0.87,
|
| 7466 |
+
"grad_norm": 173306795786240.0,
|
| 7467 |
+
"learning_rate": 4.117941845802026e-08,
|
| 7468 |
+
"loss": 26.646002197265624,
|
| 7469 |
+
"step": 9570
|
| 7470 |
+
},
|
| 7471 |
+
{
|
| 7472 |
+
"epoch": 0.8709090909090909,
|
| 7473 |
+
"grad_norm": 351010866331648.0,
|
| 7474 |
+
"learning_rate": 4.06137903277976e-08,
|
| 7475 |
+
"loss": 26.28143310546875,
|
| 7476 |
+
"step": 9580
|
| 7477 |
+
},
|
| 7478 |
+
{
|
| 7479 |
+
"epoch": 0.8718181818181818,
|
| 7480 |
+
"grad_norm": 102902408413184.0,
|
| 7481 |
+
"learning_rate": 4.0051909269537776e-08,
|
| 7482 |
+
"loss": 26.641571044921875,
|
| 7483 |
+
"step": 9590
|
| 7484 |
+
},
|
| 7485 |
+
{
|
| 7486 |
+
"epoch": 0.8727272727272727,
|
| 7487 |
+
"grad_norm": 173646198865920.0,
|
| 7488 |
+
"learning_rate": 3.9493779866331457e-08,
|
| 7489 |
+
"loss": 26.6406982421875,
|
| 7490 |
+
"step": 9600
|
| 7491 |
+
},
|
| 7492 |
+
{
|
| 7493 |
+
"epoch": 0.8727272727272727,
|
| 7494 |
+
"eval_loss": 26.25297737121582,
|
| 7495 |
+
"eval_runtime": 2.2018,
|
| 7496 |
+
"eval_samples_per_second": 16.35,
|
| 7497 |
+
"eval_steps_per_second": 0.908,
|
| 7498 |
+
"step": 9600
|
| 7499 |
+
},
|
| 7500 |
+
{
|
| 7501 |
+
"epoch": 0.8736363636363637,
|
| 7502 |
+
"grad_norm": 186953400057856.0,
|
| 7503 |
+
"learning_rate": 3.893940667066875e-08,
|
| 7504 |
+
"loss": 26.639181518554686,
|
| 7505 |
+
"step": 9610
|
| 7506 |
+
},
|
| 7507 |
+
{
|
| 7508 |
+
"epoch": 0.8745454545454545,
|
| 7509 |
+
"grad_norm": 119041461059584.0,
|
| 7510 |
+
"learning_rate": 3.8388794204401055e-08,
|
| 7511 |
+
"loss": 26.684365844726564,
|
| 7512 |
+
"step": 9620
|
| 7513 |
+
},
|
| 7514 |
+
{
|
| 7515 |
+
"epoch": 0.8754545454545455,
|
| 7516 |
+
"grad_norm": 393604358995968.0,
|
| 7517 |
+
"learning_rate": 3.784194695870485e-08,
|
| 7518 |
+
"loss": 26.53135986328125,
|
| 7519 |
+
"step": 9630
|
| 7520 |
+
},
|
| 7521 |
+
{
|
| 7522 |
+
"epoch": 0.8763636363636363,
|
| 7523 |
+
"grad_norm": 265332392460288.0,
|
| 7524 |
+
"learning_rate": 3.7298869394044865e-08,
|
| 7525 |
+
"loss": 26.410488891601563,
|
| 7526 |
+
"step": 9640
|
| 7527 |
+
},
|
| 7528 |
+
{
|
| 7529 |
+
"epoch": 0.8772727272727273,
|
| 7530 |
+
"grad_norm": 469139042861056.0,
|
| 7531 |
+
"learning_rate": 3.6759565940137846e-08,
|
| 7532 |
+
"loss": 26.624411010742186,
|
| 7533 |
+
"step": 9650
|
| 7534 |
+
},
|
| 7535 |
+
{
|
| 7536 |
+
"epoch": 0.8781818181818182,
|
| 7537 |
+
"grad_norm": 233339919269888.0,
|
| 7538 |
+
"learning_rate": 3.6224040995915915e-08,
|
| 7539 |
+
"loss": 26.61990966796875,
|
| 7540 |
+
"step": 9660
|
| 7541 |
+
},
|
| 7542 |
+
{
|
| 7543 |
+
"epoch": 0.8790909090909091,
|
| 7544 |
+
"grad_norm": 160247746396160.0,
|
| 7545 |
+
"learning_rate": 3.569229892949133e-08,
|
| 7546 |
+
"loss": 26.281167602539064,
|
| 7547 |
+
"step": 9670
|
| 7548 |
+
},
|
| 7549 |
+
{
|
| 7550 |
+
"epoch": 0.88,
|
| 7551 |
+
"grad_norm": 231272144175104.0,
|
| 7552 |
+
"learning_rate": 3.516434407812063e-08,
|
| 7553 |
+
"loss": 26.265048217773437,
|
| 7554 |
+
"step": 9680
|
| 7555 |
+
},
|
| 7556 |
+
{
|
| 7557 |
+
"epoch": 0.8809090909090909,
|
| 7558 |
+
"grad_norm": 162941127098368.0,
|
| 7559 |
+
"learning_rate": 3.464018074816893e-08,
|
| 7560 |
+
"loss": 26.53309326171875,
|
| 7561 |
+
"step": 9690
|
| 7562 |
+
},
|
| 7563 |
+
{
|
| 7564 |
+
"epoch": 0.8818181818181818,
|
| 7565 |
+
"grad_norm": 235812629250048.0,
|
| 7566 |
+
"learning_rate": 3.411981321507523e-08,
|
| 7567 |
+
"loss": 26.598037719726562,
|
| 7568 |
+
"step": 9700
|
| 7569 |
+
},
|
| 7570 |
+
{
|
| 7571 |
+
"epoch": 0.8818181818181818,
|
| 7572 |
+
"eval_loss": 26.255329132080078,
|
| 7573 |
+
"eval_runtime": 2.1559,
|
| 7574 |
+
"eval_samples_per_second": 16.698,
|
| 7575 |
+
"eval_steps_per_second": 0.928,
|
| 7576 |
+
"step": 9700
|
| 7577 |
+
},
|
| 7578 |
+
{
|
| 7579 |
+
"epoch": 0.8827272727272727,
|
| 7580 |
+
"grad_norm": 529596579577856.0,
|
| 7581 |
+
"learning_rate": 3.360324572331752e-08,
|
| 7582 |
+
"loss": 26.505670166015626,
|
| 7583 |
+
"step": 9710
|
| 7584 |
+
},
|
| 7585 |
+
{
|
| 7586 |
+
"epoch": 0.8836363636363637,
|
| 7587 |
+
"grad_norm": 209934092337152.0,
|
| 7588 |
+
"learning_rate": 3.3090482486377854e-08,
|
| 7589 |
+
"loss": 26.292486572265624,
|
| 7590 |
+
"step": 9720
|
| 7591 |
+
},
|
| 7592 |
+
{
|
| 7593 |
+
"epoch": 0.8845454545454545,
|
| 7594 |
+
"grad_norm": 68720860856320.0,
|
| 7595 |
+
"learning_rate": 3.258152768670802e-08,
|
| 7596 |
+
"loss": 26.464907836914062,
|
| 7597 |
+
"step": 9730
|
| 7598 |
+
},
|
| 7599 |
+
{
|
| 7600 |
+
"epoch": 0.8854545454545455,
|
| 7601 |
+
"grad_norm": 120226192556032.0,
|
| 7602 |
+
"learning_rate": 3.207638547569597e-08,
|
| 7603 |
+
"loss": 26.43253173828125,
|
| 7604 |
+
"step": 9740
|
| 7605 |
+
},
|
| 7606 |
+
{
|
| 7607 |
+
"epoch": 0.8863636363636364,
|
| 7608 |
+
"grad_norm": 203593093218304.0,
|
| 7609 |
+
"learning_rate": 3.157505997363108e-08,
|
| 7610 |
+
"loss": 26.225360107421874,
|
| 7611 |
+
"step": 9750
|
| 7612 |
+
},
|
| 7613 |
+
{
|
| 7614 |
+
"epoch": 0.8872727272727273,
|
| 7615 |
+
"grad_norm": 110963525156864.0,
|
| 7616 |
+
"learning_rate": 3.107755526967132e-08,
|
| 7617 |
+
"loss": 26.272589111328124,
|
| 7618 |
+
"step": 9760
|
| 7619 |
+
},
|
| 7620 |
+
{
|
| 7621 |
+
"epoch": 0.8881818181818182,
|
| 7622 |
+
"grad_norm": 228354837970944.0,
|
| 7623 |
+
"learning_rate": 3.058387542180951e-08,
|
| 7624 |
+
"loss": 26.39100341796875,
|
| 7625 |
+
"step": 9770
|
| 7626 |
+
},
|
| 7627 |
+
{
|
| 7628 |
+
"epoch": 0.889090909090909,
|
| 7629 |
+
"grad_norm": 211785055469568.0,
|
| 7630 |
+
"learning_rate": 3.009402445684017e-08,
|
| 7631 |
+
"loss": 26.599038696289064,
|
| 7632 |
+
"step": 9780
|
| 7633 |
+
},
|
| 7634 |
+
{
|
| 7635 |
+
"epoch": 0.89,
|
| 7636 |
+
"grad_norm": 439669493858304.0,
|
| 7637 |
+
"learning_rate": 2.9608006370326955e-08,
|
| 7638 |
+
"loss": 26.343658447265625,
|
| 7639 |
+
"step": 9790
|
| 7640 |
+
},
|
| 7641 |
+
{
|
| 7642 |
+
"epoch": 0.8909090909090909,
|
| 7643 |
+
"grad_norm": 124145425711104.0,
|
| 7644 |
+
"learning_rate": 2.912582512656986e-08,
|
| 7645 |
+
"loss": 26.353958129882812,
|
| 7646 |
+
"step": 9800
|
| 7647 |
+
},
|
| 7648 |
+
{
|
| 7649 |
+
"epoch": 0.8909090909090909,
|
| 7650 |
+
"eval_loss": 26.251920700073242,
|
| 7651 |
+
"eval_runtime": 2.1837,
|
| 7652 |
+
"eval_samples_per_second": 16.485,
|
| 7653 |
+
"eval_steps_per_second": 0.916,
|
| 7654 |
+
"step": 9800
|
| 7655 |
+
},
|
| 7656 |
+
{
|
| 7657 |
+
"epoch": 0.8918181818181818,
|
| 7658 |
+
"grad_norm": 218182426034176.0,
|
| 7659 |
+
"learning_rate": 2.8647484658572918e-08,
|
| 7660 |
+
"loss": 26.384439086914064,
|
| 7661 |
+
"step": 9810
|
| 7662 |
+
},
|
| 7663 |
+
{
|
| 7664 |
+
"epoch": 0.8927272727272727,
|
| 7665 |
+
"grad_norm": 458707439714304.0,
|
| 7666 |
+
"learning_rate": 2.8172988868012117e-08,
|
| 7667 |
+
"loss": 26.731353759765625,
|
| 7668 |
+
"step": 9820
|
| 7669 |
+
},
|
| 7670 |
+
{
|
| 7671 |
+
"epoch": 0.8936363636363637,
|
| 7672 |
+
"grad_norm": 363114386161664.0,
|
| 7673 |
+
"learning_rate": 2.770234162520363e-08,
|
| 7674 |
+
"loss": 26.34778747558594,
|
| 7675 |
+
"step": 9830
|
| 7676 |
+
},
|
| 7677 |
+
{
|
| 7678 |
+
"epoch": 0.8945454545454545,
|
| 7679 |
+
"grad_norm": 473513165062144.0,
|
| 7680 |
+
"learning_rate": 2.7235546769072247e-08,
|
| 7681 |
+
"loss": 25.989871215820312,
|
| 7682 |
+
"step": 9840
|
| 7683 |
+
},
|
| 7684 |
+
{
|
| 7685 |
+
"epoch": 0.8954545454545455,
|
| 7686 |
+
"grad_norm": 453134988083200.0,
|
| 7687 |
+
"learning_rate": 2.6772608107119954e-08,
|
| 7688 |
+
"loss": 26.36603698730469,
|
| 7689 |
+
"step": 9850
|
| 7690 |
+
},
|
| 7691 |
+
{
|
| 7692 |
+
"epoch": 0.8963636363636364,
|
| 7693 |
+
"grad_norm": 246036849229824.0,
|
| 7694 |
+
"learning_rate": 2.6313529415394942e-08,
|
| 7695 |
+
"loss": 26.3845703125,
|
| 7696 |
+
"step": 9860
|
| 7697 |
+
},
|
| 7698 |
+
{
|
| 7699 |
+
"epoch": 0.8972727272727272,
|
| 7700 |
+
"grad_norm": 72789880995840.0,
|
| 7701 |
+
"learning_rate": 2.585831443846087e-08,
|
| 7702 |
+
"loss": 26.44983825683594,
|
| 7703 |
+
"step": 9870
|
| 7704 |
+
},
|
| 7705 |
+
{
|
| 7706 |
+
"epoch": 0.8981818181818182,
|
| 7707 |
+
"grad_norm": 548986779860992.0,
|
| 7708 |
+
"learning_rate": 2.540696688936622e-08,
|
| 7709 |
+
"loss": 26.52098388671875,
|
| 7710 |
+
"step": 9880
|
| 7711 |
+
},
|
| 7712 |
+
{
|
| 7713 |
+
"epoch": 0.899090909090909,
|
| 7714 |
+
"grad_norm": 388335675637760.0,
|
| 7715 |
+
"learning_rate": 2.4959490449614094e-08,
|
| 7716 |
+
"loss": 26.495932006835936,
|
| 7717 |
+
"step": 9890
|
| 7718 |
+
},
|
| 7719 |
+
{
|
| 7720 |
+
"epoch": 0.9,
|
| 7721 |
+
"grad_norm": 53976842108928.0,
|
| 7722 |
+
"learning_rate": 2.4515888769132244e-08,
|
| 7723 |
+
"loss": 26.7628173828125,
|
| 7724 |
+
"step": 9900
|
| 7725 |
+
},
|
| 7726 |
+
{
|
| 7727 |
+
"epoch": 0.9,
|
| 7728 |
+
"eval_loss": 26.25650978088379,
|
| 7729 |
+
"eval_runtime": 2.1722,
|
| 7730 |
+
"eval_samples_per_second": 16.573,
|
| 7731 |
+
"eval_steps_per_second": 0.921,
|
| 7732 |
+
"step": 9900
|
| 7733 |
+
},
|
| 7734 |
+
{
|
| 7735 |
+
"epoch": 0.9009090909090909,
|
| 7736 |
+
"grad_norm": 222135859544064.0,
|
| 7737 |
+
"learning_rate": 2.407616546624286e-08,
|
| 7738 |
+
"loss": 26.65426025390625,
|
| 7739 |
+
"step": 9910
|
| 7740 |
+
},
|
| 7741 |
+
{
|
| 7742 |
+
"epoch": 0.9018181818181819,
|
| 7743 |
+
"grad_norm": 156349862248448.0,
|
| 7744 |
+
"learning_rate": 2.3640324127633726e-08,
|
| 7745 |
+
"loss": 26.53638610839844,
|
| 7746 |
+
"step": 9920
|
| 7747 |
+
},
|
| 7748 |
+
{
|
| 7749 |
+
"epoch": 0.9027272727272727,
|
| 7750 |
+
"grad_norm": 406163145359360.0,
|
| 7751 |
+
"learning_rate": 2.3208368308328562e-08,
|
| 7752 |
+
"loss": 26.672116088867188,
|
| 7753 |
+
"step": 9930
|
| 7754 |
+
},
|
| 7755 |
+
{
|
| 7756 |
+
"epoch": 0.9036363636363637,
|
| 7757 |
+
"grad_norm": 290215168049152.0,
|
| 7758 |
+
"learning_rate": 2.278030153165811e-08,
|
| 7759 |
+
"loss": 26.395840454101563,
|
| 7760 |
+
"step": 9940
|
| 7761 |
+
},
|
| 7762 |
+
{
|
| 7763 |
+
"epoch": 0.9045454545454545,
|
| 7764 |
+
"grad_norm": 323777854439424.0,
|
| 7765 |
+
"learning_rate": 2.2356127289231198e-08,
|
| 7766 |
+
"loss": 26.753033447265626,
|
| 7767 |
+
"step": 9950
|
| 7768 |
+
},
|
| 7769 |
+
{
|
| 7770 |
+
"epoch": 0.9054545454545454,
|
| 7771 |
+
"grad_norm": 353928558411776.0,
|
| 7772 |
+
"learning_rate": 2.1935849040906785e-08,
|
| 7773 |
+
"loss": 26.23077392578125,
|
| 7774 |
+
"step": 9960
|
| 7775 |
+
},
|
| 7776 |
+
{
|
| 7777 |
+
"epoch": 0.9063636363636364,
|
| 7778 |
+
"grad_norm": 323219877789696.0,
|
| 7779 |
+
"learning_rate": 2.1519470214765134e-08,
|
| 7780 |
+
"loss": 26.30245361328125,
|
| 7781 |
+
"step": 9970
|
| 7782 |
+
},
|
| 7783 |
+
{
|
| 7784 |
+
"epoch": 0.9072727272727272,
|
| 7785 |
+
"grad_norm": 102580596244480.0,
|
| 7786 |
+
"learning_rate": 2.1106994207080176e-08,
|
| 7787 |
+
"loss": 26.48880615234375,
|
| 7788 |
+
"step": 9980
|
| 7789 |
+
},
|
| 7790 |
+
{
|
| 7791 |
+
"epoch": 0.9081818181818182,
|
| 7792 |
+
"grad_norm": 560100477501440.0,
|
| 7793 |
+
"learning_rate": 2.0698424382291913e-08,
|
| 7794 |
+
"loss": 26.237164306640626,
|
| 7795 |
+
"step": 9990
|
| 7796 |
+
},
|
| 7797 |
+
{
|
| 7798 |
+
"epoch": 0.9090909090909091,
|
| 7799 |
+
"grad_norm": 500201991176192.0,
|
| 7800 |
+
"learning_rate": 2.0293764072978616e-08,
|
| 7801 |
+
"loss": 26.076568603515625,
|
| 7802 |
+
"step": 10000
|
| 7803 |
+
},
|
| 7804 |
+
{
|
| 7805 |
+
"epoch": 0.9090909090909091,
|
| 7806 |
+
"eval_loss": 26.25713348388672,
|
| 7807 |
+
"eval_runtime": 2.2039,
|
| 7808 |
+
"eval_samples_per_second": 16.335,
|
| 7809 |
+
"eval_steps_per_second": 0.907,
|
| 7810 |
+
"step": 10000
|
| 7811 |
}
|
| 7812 |
],
|
| 7813 |
"logging_steps": 10,
|
|
|
|
| 7827 |
"attributes": {}
|
| 7828 |
}
|
| 7829 |
},
|
| 7830 |
+
"total_flos": 9.059847634944e+16,
|
| 7831 |
"train_batch_size": 4,
|
| 7832 |
"trial_name": null,
|
| 7833 |
"trial_params": null
|