Commit
•
1553183
1
Parent(s):
1098d3f
Training in progress, step 102000, 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 +1411 -3
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2997015224
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7821e784728fa25fcfdc9d782f516d1e1f9b338613feccdf922f1d690b910a85
|
| 3 |
size 2997015224
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5994123294
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9f7fc10eaccc24db892ba90a3c868629f6f74d56d4af8eec371960b9460a18fd
|
| 3 |
size 5994123294
|
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 15920
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2cf990d87a7056e9e4b0a252156f6b30795db2a686df2fb1c99f6e3a11d9838f
|
| 3 |
size 15920
|
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 15920
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:76db20779ac73e63d7d686b30d532d98ecbc31c9d6b323e5308a241a5f958a43
|
| 3 |
size 15920
|
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 15920
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ce61d981d04468713b611e91a2f0d4a82eab3612f8080a73371b4286cb63626d
|
| 3 |
size 15920
|
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 15920
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f071c6bab394ba16aa2e81723b39d293147108a946add4b7cb9e8995f7a2eda6
|
| 3 |
size 15920
|
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 15920
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7365d7bb904d56209b9f093837e31cc34422778fa119f818fd18ab35415e1bdc
|
| 3 |
size 15920
|
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 15920
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9adc39521416cf27c3bd1677c4843d13bc25278518dfb7775d844fc53cf3a78f
|
| 3 |
size 15920
|
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 15920
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4647715470bd8450f77c0b9ab529748ba4914076bb6ec38c8c5c5b31cf8666dd
|
| 3 |
size 15920
|
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 15920
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e7904bd2bef78f87b42aca6356cfdbb66809fe1ebc3b4a680d13a9faa13505e0
|
| 3 |
size 15920
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a01d2518bf51c6d331e2a2fd21d75544661ef5a07c1d1563fa31a405709da4a5
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 2000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -70407,6 +70407,1414 @@
|
|
| 70407 |
"eval_samples_per_second": 27.748,
|
| 70408 |
"eval_steps_per_second": 0.444,
|
| 70409 |
"step": 100000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 70410 |
}
|
| 70411 |
],
|
| 70412 |
"logging_steps": 10,
|
|
@@ -70426,7 +71834,7 @@
|
|
| 70426 |
"attributes": {}
|
| 70427 |
}
|
| 70428 |
},
|
| 70429 |
-
"total_flos": 2.
|
| 70430 |
"train_batch_size": 8,
|
| 70431 |
"trial_name": null,
|
| 70432 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.51,
|
| 5 |
"eval_steps": 2000,
|
| 6 |
+
"global_step": 102000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 70407 |
"eval_samples_per_second": 27.748,
|
| 70408 |
"eval_steps_per_second": 0.444,
|
| 70409 |
"step": 100000
|
| 70410 |
+
},
|
| 70411 |
+
{
|
| 70412 |
+
"epoch": 0.50005,
|
| 70413 |
+
"grad_norm": 2.046875,
|
| 70414 |
+
"learning_rate": 0.00015073869346733667,
|
| 70415 |
+
"loss": 2.0577,
|
| 70416 |
+
"step": 100010
|
| 70417 |
+
},
|
| 70418 |
+
{
|
| 70419 |
+
"epoch": 0.5001,
|
| 70420 |
+
"grad_norm": 1.96875,
|
| 70421 |
+
"learning_rate": 0.00015072361809045226,
|
| 70422 |
+
"loss": 2.04,
|
| 70423 |
+
"step": 100020
|
| 70424 |
+
},
|
| 70425 |
+
{
|
| 70426 |
+
"epoch": 0.50015,
|
| 70427 |
+
"grad_norm": 2.046875,
|
| 70428 |
+
"learning_rate": 0.00015070854271356782,
|
| 70429 |
+
"loss": 2.0467,
|
| 70430 |
+
"step": 100030
|
| 70431 |
+
},
|
| 70432 |
+
{
|
| 70433 |
+
"epoch": 0.5002,
|
| 70434 |
+
"grad_norm": 2.078125,
|
| 70435 |
+
"learning_rate": 0.0001506934673366834,
|
| 70436 |
+
"loss": 2.0113,
|
| 70437 |
+
"step": 100040
|
| 70438 |
+
},
|
| 70439 |
+
{
|
| 70440 |
+
"epoch": 0.50025,
|
| 70441 |
+
"grad_norm": 2.078125,
|
| 70442 |
+
"learning_rate": 0.000150678391959799,
|
| 70443 |
+
"loss": 2.0461,
|
| 70444 |
+
"step": 100050
|
| 70445 |
+
},
|
| 70446 |
+
{
|
| 70447 |
+
"epoch": 0.5003,
|
| 70448 |
+
"grad_norm": 1.859375,
|
| 70449 |
+
"learning_rate": 0.00015066331658291455,
|
| 70450 |
+
"loss": 1.9994,
|
| 70451 |
+
"step": 100060
|
| 70452 |
+
},
|
| 70453 |
+
{
|
| 70454 |
+
"epoch": 0.50035,
|
| 70455 |
+
"grad_norm": 2.0625,
|
| 70456 |
+
"learning_rate": 0.00015064824120603016,
|
| 70457 |
+
"loss": 2.0283,
|
| 70458 |
+
"step": 100070
|
| 70459 |
+
},
|
| 70460 |
+
{
|
| 70461 |
+
"epoch": 0.5004,
|
| 70462 |
+
"grad_norm": 2.0625,
|
| 70463 |
+
"learning_rate": 0.00015063316582914572,
|
| 70464 |
+
"loss": 2.0629,
|
| 70465 |
+
"step": 100080
|
| 70466 |
+
},
|
| 70467 |
+
{
|
| 70468 |
+
"epoch": 0.50045,
|
| 70469 |
+
"grad_norm": 2.09375,
|
| 70470 |
+
"learning_rate": 0.00015061809045226128,
|
| 70471 |
+
"loss": 2.0144,
|
| 70472 |
+
"step": 100090
|
| 70473 |
+
},
|
| 70474 |
+
{
|
| 70475 |
+
"epoch": 0.5005,
|
| 70476 |
+
"grad_norm": 1.828125,
|
| 70477 |
+
"learning_rate": 0.0001506030150753769,
|
| 70478 |
+
"loss": 2.082,
|
| 70479 |
+
"step": 100100
|
| 70480 |
+
},
|
| 70481 |
+
{
|
| 70482 |
+
"epoch": 0.50055,
|
| 70483 |
+
"grad_norm": 2.140625,
|
| 70484 |
+
"learning_rate": 0.00015058793969849245,
|
| 70485 |
+
"loss": 2.0549,
|
| 70486 |
+
"step": 100110
|
| 70487 |
+
},
|
| 70488 |
+
{
|
| 70489 |
+
"epoch": 0.5006,
|
| 70490 |
+
"grad_norm": 1.9296875,
|
| 70491 |
+
"learning_rate": 0.000150572864321608,
|
| 70492 |
+
"loss": 2.0591,
|
| 70493 |
+
"step": 100120
|
| 70494 |
+
},
|
| 70495 |
+
{
|
| 70496 |
+
"epoch": 0.50065,
|
| 70497 |
+
"grad_norm": 1.9296875,
|
| 70498 |
+
"learning_rate": 0.0001505577889447236,
|
| 70499 |
+
"loss": 1.9906,
|
| 70500 |
+
"step": 100130
|
| 70501 |
+
},
|
| 70502 |
+
{
|
| 70503 |
+
"epoch": 0.5007,
|
| 70504 |
+
"grad_norm": 1.9453125,
|
| 70505 |
+
"learning_rate": 0.00015054271356783919,
|
| 70506 |
+
"loss": 2.0959,
|
| 70507 |
+
"step": 100140
|
| 70508 |
+
},
|
| 70509 |
+
{
|
| 70510 |
+
"epoch": 0.50075,
|
| 70511 |
+
"grad_norm": 1.875,
|
| 70512 |
+
"learning_rate": 0.00015052763819095477,
|
| 70513 |
+
"loss": 2.0214,
|
| 70514 |
+
"step": 100150
|
| 70515 |
+
},
|
| 70516 |
+
{
|
| 70517 |
+
"epoch": 0.5008,
|
| 70518 |
+
"grad_norm": 2.0625,
|
| 70519 |
+
"learning_rate": 0.00015051256281407033,
|
| 70520 |
+
"loss": 2.0448,
|
| 70521 |
+
"step": 100160
|
| 70522 |
+
},
|
| 70523 |
+
{
|
| 70524 |
+
"epoch": 0.50085,
|
| 70525 |
+
"grad_norm": 1.96875,
|
| 70526 |
+
"learning_rate": 0.00015049748743718592,
|
| 70527 |
+
"loss": 2.05,
|
| 70528 |
+
"step": 100170
|
| 70529 |
+
},
|
| 70530 |
+
{
|
| 70531 |
+
"epoch": 0.5009,
|
| 70532 |
+
"grad_norm": 2.21875,
|
| 70533 |
+
"learning_rate": 0.0001504824120603015,
|
| 70534 |
+
"loss": 2.014,
|
| 70535 |
+
"step": 100180
|
| 70536 |
+
},
|
| 70537 |
+
{
|
| 70538 |
+
"epoch": 0.50095,
|
| 70539 |
+
"grad_norm": 1.7734375,
|
| 70540 |
+
"learning_rate": 0.00015046733668341706,
|
| 70541 |
+
"loss": 1.9987,
|
| 70542 |
+
"step": 100190
|
| 70543 |
+
},
|
| 70544 |
+
{
|
| 70545 |
+
"epoch": 0.501,
|
| 70546 |
+
"grad_norm": 2.0,
|
| 70547 |
+
"learning_rate": 0.00015045226130653267,
|
| 70548 |
+
"loss": 2.0231,
|
| 70549 |
+
"step": 100200
|
| 70550 |
+
},
|
| 70551 |
+
{
|
| 70552 |
+
"epoch": 0.50105,
|
| 70553 |
+
"grad_norm": 1.859375,
|
| 70554 |
+
"learning_rate": 0.00015043718592964823,
|
| 70555 |
+
"loss": 2.0519,
|
| 70556 |
+
"step": 100210
|
| 70557 |
+
},
|
| 70558 |
+
{
|
| 70559 |
+
"epoch": 0.5011,
|
| 70560 |
+
"grad_norm": 2.015625,
|
| 70561 |
+
"learning_rate": 0.0001504221105527638,
|
| 70562 |
+
"loss": 2.0005,
|
| 70563 |
+
"step": 100220
|
| 70564 |
+
},
|
| 70565 |
+
{
|
| 70566 |
+
"epoch": 0.50115,
|
| 70567 |
+
"grad_norm": 1.9296875,
|
| 70568 |
+
"learning_rate": 0.0001504070351758794,
|
| 70569 |
+
"loss": 2.0114,
|
| 70570 |
+
"step": 100230
|
| 70571 |
+
},
|
| 70572 |
+
{
|
| 70573 |
+
"epoch": 0.5012,
|
| 70574 |
+
"grad_norm": 1.8359375,
|
| 70575 |
+
"learning_rate": 0.00015039195979899496,
|
| 70576 |
+
"loss": 2.0219,
|
| 70577 |
+
"step": 100240
|
| 70578 |
+
},
|
| 70579 |
+
{
|
| 70580 |
+
"epoch": 0.50125,
|
| 70581 |
+
"grad_norm": 2.140625,
|
| 70582 |
+
"learning_rate": 0.00015037688442211052,
|
| 70583 |
+
"loss": 2.0079,
|
| 70584 |
+
"step": 100250
|
| 70585 |
+
},
|
| 70586 |
+
{
|
| 70587 |
+
"epoch": 0.5013,
|
| 70588 |
+
"grad_norm": 1.9453125,
|
| 70589 |
+
"learning_rate": 0.00015036180904522614,
|
| 70590 |
+
"loss": 2.0236,
|
| 70591 |
+
"step": 100260
|
| 70592 |
+
},
|
| 70593 |
+
{
|
| 70594 |
+
"epoch": 0.50135,
|
| 70595 |
+
"grad_norm": 2.09375,
|
| 70596 |
+
"learning_rate": 0.0001503467336683417,
|
| 70597 |
+
"loss": 2.0025,
|
| 70598 |
+
"step": 100270
|
| 70599 |
+
},
|
| 70600 |
+
{
|
| 70601 |
+
"epoch": 0.5014,
|
| 70602 |
+
"grad_norm": 1.90625,
|
| 70603 |
+
"learning_rate": 0.00015033165829145726,
|
| 70604 |
+
"loss": 2.0407,
|
| 70605 |
+
"step": 100280
|
| 70606 |
+
},
|
| 70607 |
+
{
|
| 70608 |
+
"epoch": 0.50145,
|
| 70609 |
+
"grad_norm": 2.296875,
|
| 70610 |
+
"learning_rate": 0.00015031658291457284,
|
| 70611 |
+
"loss": 2.0013,
|
| 70612 |
+
"step": 100290
|
| 70613 |
+
},
|
| 70614 |
+
{
|
| 70615 |
+
"epoch": 0.5015,
|
| 70616 |
+
"grad_norm": 2.046875,
|
| 70617 |
+
"learning_rate": 0.00015030150753768843,
|
| 70618 |
+
"loss": 1.9921,
|
| 70619 |
+
"step": 100300
|
| 70620 |
+
},
|
| 70621 |
+
{
|
| 70622 |
+
"epoch": 0.50155,
|
| 70623 |
+
"grad_norm": 1.9453125,
|
| 70624 |
+
"learning_rate": 0.000150286432160804,
|
| 70625 |
+
"loss": 2.0336,
|
| 70626 |
+
"step": 100310
|
| 70627 |
+
},
|
| 70628 |
+
{
|
| 70629 |
+
"epoch": 0.5016,
|
| 70630 |
+
"grad_norm": 2.171875,
|
| 70631 |
+
"learning_rate": 0.00015027135678391957,
|
| 70632 |
+
"loss": 2.0329,
|
| 70633 |
+
"step": 100320
|
| 70634 |
+
},
|
| 70635 |
+
{
|
| 70636 |
+
"epoch": 0.50165,
|
| 70637 |
+
"grad_norm": 1.9921875,
|
| 70638 |
+
"learning_rate": 0.00015025628140703519,
|
| 70639 |
+
"loss": 2.0327,
|
| 70640 |
+
"step": 100330
|
| 70641 |
+
},
|
| 70642 |
+
{
|
| 70643 |
+
"epoch": 0.5017,
|
| 70644 |
+
"grad_norm": 1.78125,
|
| 70645 |
+
"learning_rate": 0.00015024120603015074,
|
| 70646 |
+
"loss": 2.0726,
|
| 70647 |
+
"step": 100340
|
| 70648 |
+
},
|
| 70649 |
+
{
|
| 70650 |
+
"epoch": 0.50175,
|
| 70651 |
+
"grad_norm": 1.828125,
|
| 70652 |
+
"learning_rate": 0.0001502261306532663,
|
| 70653 |
+
"loss": 2.0025,
|
| 70654 |
+
"step": 100350
|
| 70655 |
+
},
|
| 70656 |
+
{
|
| 70657 |
+
"epoch": 0.5018,
|
| 70658 |
+
"grad_norm": 1.9140625,
|
| 70659 |
+
"learning_rate": 0.00015021105527638192,
|
| 70660 |
+
"loss": 2.0309,
|
| 70661 |
+
"step": 100360
|
| 70662 |
+
},
|
| 70663 |
+
{
|
| 70664 |
+
"epoch": 0.50185,
|
| 70665 |
+
"grad_norm": 1.9609375,
|
| 70666 |
+
"learning_rate": 0.00015019597989949748,
|
| 70667 |
+
"loss": 2.0078,
|
| 70668 |
+
"step": 100370
|
| 70669 |
+
},
|
| 70670 |
+
{
|
| 70671 |
+
"epoch": 0.5019,
|
| 70672 |
+
"grad_norm": 1.921875,
|
| 70673 |
+
"learning_rate": 0.00015018090452261303,
|
| 70674 |
+
"loss": 2.0121,
|
| 70675 |
+
"step": 100380
|
| 70676 |
+
},
|
| 70677 |
+
{
|
| 70678 |
+
"epoch": 0.50195,
|
| 70679 |
+
"grad_norm": 1.8984375,
|
| 70680 |
+
"learning_rate": 0.00015016582914572865,
|
| 70681 |
+
"loss": 2.0177,
|
| 70682 |
+
"step": 100390
|
| 70683 |
+
},
|
| 70684 |
+
{
|
| 70685 |
+
"epoch": 0.502,
|
| 70686 |
+
"grad_norm": 1.8515625,
|
| 70687 |
+
"learning_rate": 0.0001501507537688442,
|
| 70688 |
+
"loss": 2.012,
|
| 70689 |
+
"step": 100400
|
| 70690 |
+
},
|
| 70691 |
+
{
|
| 70692 |
+
"epoch": 0.50205,
|
| 70693 |
+
"grad_norm": 1.9453125,
|
| 70694 |
+
"learning_rate": 0.00015013567839195977,
|
| 70695 |
+
"loss": 2.0336,
|
| 70696 |
+
"step": 100410
|
| 70697 |
+
},
|
| 70698 |
+
{
|
| 70699 |
+
"epoch": 0.5021,
|
| 70700 |
+
"grad_norm": 2.03125,
|
| 70701 |
+
"learning_rate": 0.00015012060301507538,
|
| 70702 |
+
"loss": 2.0294,
|
| 70703 |
+
"step": 100420
|
| 70704 |
+
},
|
| 70705 |
+
{
|
| 70706 |
+
"epoch": 0.50215,
|
| 70707 |
+
"grad_norm": 1.875,
|
| 70708 |
+
"learning_rate": 0.00015010552763819094,
|
| 70709 |
+
"loss": 2.0701,
|
| 70710 |
+
"step": 100430
|
| 70711 |
+
},
|
| 70712 |
+
{
|
| 70713 |
+
"epoch": 0.5022,
|
| 70714 |
+
"grad_norm": 2.140625,
|
| 70715 |
+
"learning_rate": 0.00015009045226130652,
|
| 70716 |
+
"loss": 2.042,
|
| 70717 |
+
"step": 100440
|
| 70718 |
+
},
|
| 70719 |
+
{
|
| 70720 |
+
"epoch": 0.50225,
|
| 70721 |
+
"grad_norm": 2.15625,
|
| 70722 |
+
"learning_rate": 0.00015007537688442208,
|
| 70723 |
+
"loss": 2.0562,
|
| 70724 |
+
"step": 100450
|
| 70725 |
+
},
|
| 70726 |
+
{
|
| 70727 |
+
"epoch": 0.5023,
|
| 70728 |
+
"grad_norm": 1.71875,
|
| 70729 |
+
"learning_rate": 0.0001500603015075377,
|
| 70730 |
+
"loss": 1.9848,
|
| 70731 |
+
"step": 100460
|
| 70732 |
+
},
|
| 70733 |
+
{
|
| 70734 |
+
"epoch": 0.50235,
|
| 70735 |
+
"grad_norm": 2.03125,
|
| 70736 |
+
"learning_rate": 0.00015004522613065326,
|
| 70737 |
+
"loss": 2.0385,
|
| 70738 |
+
"step": 100470
|
| 70739 |
+
},
|
| 70740 |
+
{
|
| 70741 |
+
"epoch": 0.5024,
|
| 70742 |
+
"grad_norm": 1.8515625,
|
| 70743 |
+
"learning_rate": 0.00015003015075376881,
|
| 70744 |
+
"loss": 2.0441,
|
| 70745 |
+
"step": 100480
|
| 70746 |
+
},
|
| 70747 |
+
{
|
| 70748 |
+
"epoch": 0.50245,
|
| 70749 |
+
"grad_norm": 1.8828125,
|
| 70750 |
+
"learning_rate": 0.00015001507537688443,
|
| 70751 |
+
"loss": 1.9888,
|
| 70752 |
+
"step": 100490
|
| 70753 |
+
},
|
| 70754 |
+
{
|
| 70755 |
+
"epoch": 0.5025,
|
| 70756 |
+
"grad_norm": 2.109375,
|
| 70757 |
+
"learning_rate": 0.00015,
|
| 70758 |
+
"loss": 2.0933,
|
| 70759 |
+
"step": 100500
|
| 70760 |
+
},
|
| 70761 |
+
{
|
| 70762 |
+
"epoch": 0.50255,
|
| 70763 |
+
"grad_norm": 1.921875,
|
| 70764 |
+
"learning_rate": 0.00014998492462311557,
|
| 70765 |
+
"loss": 1.9895,
|
| 70766 |
+
"step": 100510
|
| 70767 |
+
},
|
| 70768 |
+
{
|
| 70769 |
+
"epoch": 0.5026,
|
| 70770 |
+
"grad_norm": 2.0625,
|
| 70771 |
+
"learning_rate": 0.00014996984924623113,
|
| 70772 |
+
"loss": 2.0241,
|
| 70773 |
+
"step": 100520
|
| 70774 |
+
},
|
| 70775 |
+
{
|
| 70776 |
+
"epoch": 0.50265,
|
| 70777 |
+
"grad_norm": 1.9296875,
|
| 70778 |
+
"learning_rate": 0.00014995477386934672,
|
| 70779 |
+
"loss": 2.0868,
|
| 70780 |
+
"step": 100530
|
| 70781 |
+
},
|
| 70782 |
+
{
|
| 70783 |
+
"epoch": 0.5027,
|
| 70784 |
+
"grad_norm": 1.90625,
|
| 70785 |
+
"learning_rate": 0.00014993969849246228,
|
| 70786 |
+
"loss": 1.9649,
|
| 70787 |
+
"step": 100540
|
| 70788 |
+
},
|
| 70789 |
+
{
|
| 70790 |
+
"epoch": 0.50275,
|
| 70791 |
+
"grad_norm": 2.09375,
|
| 70792 |
+
"learning_rate": 0.00014992462311557786,
|
| 70793 |
+
"loss": 2.0709,
|
| 70794 |
+
"step": 100550
|
| 70795 |
+
},
|
| 70796 |
+
{
|
| 70797 |
+
"epoch": 0.5028,
|
| 70798 |
+
"grad_norm": 2.046875,
|
| 70799 |
+
"learning_rate": 0.00014990954773869345,
|
| 70800 |
+
"loss": 2.0065,
|
| 70801 |
+
"step": 100560
|
| 70802 |
+
},
|
| 70803 |
+
{
|
| 70804 |
+
"epoch": 0.50285,
|
| 70805 |
+
"grad_norm": 1.828125,
|
| 70806 |
+
"learning_rate": 0.00014989447236180904,
|
| 70807 |
+
"loss": 2.0105,
|
| 70808 |
+
"step": 100570
|
| 70809 |
+
},
|
| 70810 |
+
{
|
| 70811 |
+
"epoch": 0.5029,
|
| 70812 |
+
"grad_norm": 2.09375,
|
| 70813 |
+
"learning_rate": 0.00014987939698492462,
|
| 70814 |
+
"loss": 2.004,
|
| 70815 |
+
"step": 100580
|
| 70816 |
+
},
|
| 70817 |
+
{
|
| 70818 |
+
"epoch": 0.50295,
|
| 70819 |
+
"grad_norm": 1.9921875,
|
| 70820 |
+
"learning_rate": 0.0001498643216080402,
|
| 70821 |
+
"loss": 2.0418,
|
| 70822 |
+
"step": 100590
|
| 70823 |
+
},
|
| 70824 |
+
{
|
| 70825 |
+
"epoch": 0.503,
|
| 70826 |
+
"grad_norm": 1.921875,
|
| 70827 |
+
"learning_rate": 0.00014984924623115577,
|
| 70828 |
+
"loss": 1.9932,
|
| 70829 |
+
"step": 100600
|
| 70830 |
+
},
|
| 70831 |
+
{
|
| 70832 |
+
"epoch": 0.50305,
|
| 70833 |
+
"grad_norm": 2.03125,
|
| 70834 |
+
"learning_rate": 0.00014983417085427135,
|
| 70835 |
+
"loss": 2.0319,
|
| 70836 |
+
"step": 100610
|
| 70837 |
+
},
|
| 70838 |
+
{
|
| 70839 |
+
"epoch": 0.5031,
|
| 70840 |
+
"grad_norm": 2.09375,
|
| 70841 |
+
"learning_rate": 0.0001498190954773869,
|
| 70842 |
+
"loss": 2.0594,
|
| 70843 |
+
"step": 100620
|
| 70844 |
+
},
|
| 70845 |
+
{
|
| 70846 |
+
"epoch": 0.50315,
|
| 70847 |
+
"grad_norm": 1.875,
|
| 70848 |
+
"learning_rate": 0.0001498040201005025,
|
| 70849 |
+
"loss": 2.0186,
|
| 70850 |
+
"step": 100630
|
| 70851 |
+
},
|
| 70852 |
+
{
|
| 70853 |
+
"epoch": 0.5032,
|
| 70854 |
+
"grad_norm": 1.875,
|
| 70855 |
+
"learning_rate": 0.00014978894472361808,
|
| 70856 |
+
"loss": 2.0526,
|
| 70857 |
+
"step": 100640
|
| 70858 |
+
},
|
| 70859 |
+
{
|
| 70860 |
+
"epoch": 0.50325,
|
| 70861 |
+
"grad_norm": 2.265625,
|
| 70862 |
+
"learning_rate": 0.00014977386934673364,
|
| 70863 |
+
"loss": 1.9934,
|
| 70864 |
+
"step": 100650
|
| 70865 |
+
},
|
| 70866 |
+
{
|
| 70867 |
+
"epoch": 0.5033,
|
| 70868 |
+
"grad_norm": 1.9765625,
|
| 70869 |
+
"learning_rate": 0.00014975879396984923,
|
| 70870 |
+
"loss": 1.9944,
|
| 70871 |
+
"step": 100660
|
| 70872 |
+
},
|
| 70873 |
+
{
|
| 70874 |
+
"epoch": 0.50335,
|
| 70875 |
+
"grad_norm": 1.890625,
|
| 70876 |
+
"learning_rate": 0.00014974371859296482,
|
| 70877 |
+
"loss": 2.0696,
|
| 70878 |
+
"step": 100670
|
| 70879 |
+
},
|
| 70880 |
+
{
|
| 70881 |
+
"epoch": 0.5034,
|
| 70882 |
+
"grad_norm": 1.9609375,
|
| 70883 |
+
"learning_rate": 0.00014972864321608037,
|
| 70884 |
+
"loss": 2.0373,
|
| 70885 |
+
"step": 100680
|
| 70886 |
+
},
|
| 70887 |
+
{
|
| 70888 |
+
"epoch": 0.50345,
|
| 70889 |
+
"grad_norm": 2.046875,
|
| 70890 |
+
"learning_rate": 0.00014971356783919596,
|
| 70891 |
+
"loss": 2.0697,
|
| 70892 |
+
"step": 100690
|
| 70893 |
+
},
|
| 70894 |
+
{
|
| 70895 |
+
"epoch": 0.5035,
|
| 70896 |
+
"grad_norm": 2.078125,
|
| 70897 |
+
"learning_rate": 0.00014969849246231155,
|
| 70898 |
+
"loss": 2.015,
|
| 70899 |
+
"step": 100700
|
| 70900 |
+
},
|
| 70901 |
+
{
|
| 70902 |
+
"epoch": 0.50355,
|
| 70903 |
+
"grad_norm": 2.015625,
|
| 70904 |
+
"learning_rate": 0.00014968341708542713,
|
| 70905 |
+
"loss": 2.026,
|
| 70906 |
+
"step": 100710
|
| 70907 |
+
},
|
| 70908 |
+
{
|
| 70909 |
+
"epoch": 0.5036,
|
| 70910 |
+
"grad_norm": 1.859375,
|
| 70911 |
+
"learning_rate": 0.00014966834170854272,
|
| 70912 |
+
"loss": 1.9736,
|
| 70913 |
+
"step": 100720
|
| 70914 |
+
},
|
| 70915 |
+
{
|
| 70916 |
+
"epoch": 0.50365,
|
| 70917 |
+
"grad_norm": 2.109375,
|
| 70918 |
+
"learning_rate": 0.00014965326633165828,
|
| 70919 |
+
"loss": 1.9758,
|
| 70920 |
+
"step": 100730
|
| 70921 |
+
},
|
| 70922 |
+
{
|
| 70923 |
+
"epoch": 0.5037,
|
| 70924 |
+
"grad_norm": 2.03125,
|
| 70925 |
+
"learning_rate": 0.00014963819095477386,
|
| 70926 |
+
"loss": 2.0381,
|
| 70927 |
+
"step": 100740
|
| 70928 |
+
},
|
| 70929 |
+
{
|
| 70930 |
+
"epoch": 0.50375,
|
| 70931 |
+
"grad_norm": 1.984375,
|
| 70932 |
+
"learning_rate": 0.00014962311557788945,
|
| 70933 |
+
"loss": 2.005,
|
| 70934 |
+
"step": 100750
|
| 70935 |
+
},
|
| 70936 |
+
{
|
| 70937 |
+
"epoch": 0.5038,
|
| 70938 |
+
"grad_norm": 2.1875,
|
| 70939 |
+
"learning_rate": 0.000149608040201005,
|
| 70940 |
+
"loss": 2.0392,
|
| 70941 |
+
"step": 100760
|
| 70942 |
+
},
|
| 70943 |
+
{
|
| 70944 |
+
"epoch": 0.50385,
|
| 70945 |
+
"grad_norm": 1.921875,
|
| 70946 |
+
"learning_rate": 0.0001495929648241206,
|
| 70947 |
+
"loss": 1.9888,
|
| 70948 |
+
"step": 100770
|
| 70949 |
+
},
|
| 70950 |
+
{
|
| 70951 |
+
"epoch": 0.5039,
|
| 70952 |
+
"grad_norm": 2.0,
|
| 70953 |
+
"learning_rate": 0.00014957788944723615,
|
| 70954 |
+
"loss": 2.0717,
|
| 70955 |
+
"step": 100780
|
| 70956 |
+
},
|
| 70957 |
+
{
|
| 70958 |
+
"epoch": 0.50395,
|
| 70959 |
+
"grad_norm": 2.09375,
|
| 70960 |
+
"learning_rate": 0.00014956281407035174,
|
| 70961 |
+
"loss": 2.0205,
|
| 70962 |
+
"step": 100790
|
| 70963 |
+
},
|
| 70964 |
+
{
|
| 70965 |
+
"epoch": 0.504,
|
| 70966 |
+
"grad_norm": 2.421875,
|
| 70967 |
+
"learning_rate": 0.00014954773869346733,
|
| 70968 |
+
"loss": 2.0285,
|
| 70969 |
+
"step": 100800
|
| 70970 |
+
},
|
| 70971 |
+
{
|
| 70972 |
+
"epoch": 0.50405,
|
| 70973 |
+
"grad_norm": 1.9453125,
|
| 70974 |
+
"learning_rate": 0.00014953266331658289,
|
| 70975 |
+
"loss": 2.086,
|
| 70976 |
+
"step": 100810
|
| 70977 |
+
},
|
| 70978 |
+
{
|
| 70979 |
+
"epoch": 0.5041,
|
| 70980 |
+
"grad_norm": 2.015625,
|
| 70981 |
+
"learning_rate": 0.00014951758793969847,
|
| 70982 |
+
"loss": 1.9678,
|
| 70983 |
+
"step": 100820
|
| 70984 |
+
},
|
| 70985 |
+
{
|
| 70986 |
+
"epoch": 0.50415,
|
| 70987 |
+
"grad_norm": 1.828125,
|
| 70988 |
+
"learning_rate": 0.00014950251256281406,
|
| 70989 |
+
"loss": 2.1181,
|
| 70990 |
+
"step": 100830
|
| 70991 |
+
},
|
| 70992 |
+
{
|
| 70993 |
+
"epoch": 0.5042,
|
| 70994 |
+
"grad_norm": 1.765625,
|
| 70995 |
+
"learning_rate": 0.00014948743718592964,
|
| 70996 |
+
"loss": 2.0051,
|
| 70997 |
+
"step": 100840
|
| 70998 |
+
},
|
| 70999 |
+
{
|
| 71000 |
+
"epoch": 0.50425,
|
| 71001 |
+
"grad_norm": 1.875,
|
| 71002 |
+
"learning_rate": 0.0001494723618090452,
|
| 71003 |
+
"loss": 2.0193,
|
| 71004 |
+
"step": 100850
|
| 71005 |
+
},
|
| 71006 |
+
{
|
| 71007 |
+
"epoch": 0.5043,
|
| 71008 |
+
"grad_norm": 1.875,
|
| 71009 |
+
"learning_rate": 0.0001494572864321608,
|
| 71010 |
+
"loss": 2.0467,
|
| 71011 |
+
"step": 100860
|
| 71012 |
+
},
|
| 71013 |
+
{
|
| 71014 |
+
"epoch": 0.50435,
|
| 71015 |
+
"grad_norm": 2.09375,
|
| 71016 |
+
"learning_rate": 0.00014944221105527637,
|
| 71017 |
+
"loss": 2.0017,
|
| 71018 |
+
"step": 100870
|
| 71019 |
+
},
|
| 71020 |
+
{
|
| 71021 |
+
"epoch": 0.5044,
|
| 71022 |
+
"grad_norm": 1.90625,
|
| 71023 |
+
"learning_rate": 0.00014942713567839196,
|
| 71024 |
+
"loss": 2.0517,
|
| 71025 |
+
"step": 100880
|
| 71026 |
+
},
|
| 71027 |
+
{
|
| 71028 |
+
"epoch": 0.50445,
|
| 71029 |
+
"grad_norm": 2.078125,
|
| 71030 |
+
"learning_rate": 0.00014941206030150752,
|
| 71031 |
+
"loss": 1.9835,
|
| 71032 |
+
"step": 100890
|
| 71033 |
+
},
|
| 71034 |
+
{
|
| 71035 |
+
"epoch": 0.5045,
|
| 71036 |
+
"grad_norm": 1.8203125,
|
| 71037 |
+
"learning_rate": 0.0001493969849246231,
|
| 71038 |
+
"loss": 2.0614,
|
| 71039 |
+
"step": 100900
|
| 71040 |
+
},
|
| 71041 |
+
{
|
| 71042 |
+
"epoch": 0.50455,
|
| 71043 |
+
"grad_norm": 1.953125,
|
| 71044 |
+
"learning_rate": 0.0001493819095477387,
|
| 71045 |
+
"loss": 2.0655,
|
| 71046 |
+
"step": 100910
|
| 71047 |
+
},
|
| 71048 |
+
{
|
| 71049 |
+
"epoch": 0.5046,
|
| 71050 |
+
"grad_norm": 2.109375,
|
| 71051 |
+
"learning_rate": 0.00014936683417085425,
|
| 71052 |
+
"loss": 1.9778,
|
| 71053 |
+
"step": 100920
|
| 71054 |
+
},
|
| 71055 |
+
{
|
| 71056 |
+
"epoch": 0.50465,
|
| 71057 |
+
"grad_norm": 1.9296875,
|
| 71058 |
+
"learning_rate": 0.00014935175879396984,
|
| 71059 |
+
"loss": 2.0143,
|
| 71060 |
+
"step": 100930
|
| 71061 |
+
},
|
| 71062 |
+
{
|
| 71063 |
+
"epoch": 0.5047,
|
| 71064 |
+
"grad_norm": 2.0625,
|
| 71065 |
+
"learning_rate": 0.0001493366834170854,
|
| 71066 |
+
"loss": 1.9989,
|
| 71067 |
+
"step": 100940
|
| 71068 |
+
},
|
| 71069 |
+
{
|
| 71070 |
+
"epoch": 0.50475,
|
| 71071 |
+
"grad_norm": 2.03125,
|
| 71072 |
+
"learning_rate": 0.00014932160804020098,
|
| 71073 |
+
"loss": 2.0735,
|
| 71074 |
+
"step": 100950
|
| 71075 |
+
},
|
| 71076 |
+
{
|
| 71077 |
+
"epoch": 0.5048,
|
| 71078 |
+
"grad_norm": 1.984375,
|
| 71079 |
+
"learning_rate": 0.00014930653266331657,
|
| 71080 |
+
"loss": 2.0113,
|
| 71081 |
+
"step": 100960
|
| 71082 |
+
},
|
| 71083 |
+
{
|
| 71084 |
+
"epoch": 0.50485,
|
| 71085 |
+
"grad_norm": 1.765625,
|
| 71086 |
+
"learning_rate": 0.00014929145728643215,
|
| 71087 |
+
"loss": 2.046,
|
| 71088 |
+
"step": 100970
|
| 71089 |
+
},
|
| 71090 |
+
{
|
| 71091 |
+
"epoch": 0.5049,
|
| 71092 |
+
"grad_norm": 2.015625,
|
| 71093 |
+
"learning_rate": 0.0001492763819095477,
|
| 71094 |
+
"loss": 1.9569,
|
| 71095 |
+
"step": 100980
|
| 71096 |
+
},
|
| 71097 |
+
{
|
| 71098 |
+
"epoch": 0.50495,
|
| 71099 |
+
"grad_norm": 2.0625,
|
| 71100 |
+
"learning_rate": 0.0001492613065326633,
|
| 71101 |
+
"loss": 2.0329,
|
| 71102 |
+
"step": 100990
|
| 71103 |
+
},
|
| 71104 |
+
{
|
| 71105 |
+
"epoch": 0.505,
|
| 71106 |
+
"grad_norm": 1.96875,
|
| 71107 |
+
"learning_rate": 0.00014924623115577889,
|
| 71108 |
+
"loss": 1.9891,
|
| 71109 |
+
"step": 101000
|
| 71110 |
+
},
|
| 71111 |
+
{
|
| 71112 |
+
"epoch": 0.50505,
|
| 71113 |
+
"grad_norm": 2.015625,
|
| 71114 |
+
"learning_rate": 0.00014923115577889447,
|
| 71115 |
+
"loss": 1.9908,
|
| 71116 |
+
"step": 101010
|
| 71117 |
+
},
|
| 71118 |
+
{
|
| 71119 |
+
"epoch": 0.5051,
|
| 71120 |
+
"grad_norm": 1.7421875,
|
| 71121 |
+
"learning_rate": 0.00014921608040201003,
|
| 71122 |
+
"loss": 2.0593,
|
| 71123 |
+
"step": 101020
|
| 71124 |
+
},
|
| 71125 |
+
{
|
| 71126 |
+
"epoch": 0.50515,
|
| 71127 |
+
"grad_norm": 1.9609375,
|
| 71128 |
+
"learning_rate": 0.00014920100502512562,
|
| 71129 |
+
"loss": 1.9822,
|
| 71130 |
+
"step": 101030
|
| 71131 |
+
},
|
| 71132 |
+
{
|
| 71133 |
+
"epoch": 0.5052,
|
| 71134 |
+
"grad_norm": 1.7734375,
|
| 71135 |
+
"learning_rate": 0.0001491859296482412,
|
| 71136 |
+
"loss": 2.0216,
|
| 71137 |
+
"step": 101040
|
| 71138 |
+
},
|
| 71139 |
+
{
|
| 71140 |
+
"epoch": 0.50525,
|
| 71141 |
+
"grad_norm": 1.8359375,
|
| 71142 |
+
"learning_rate": 0.00014917085427135676,
|
| 71143 |
+
"loss": 1.9841,
|
| 71144 |
+
"step": 101050
|
| 71145 |
+
},
|
| 71146 |
+
{
|
| 71147 |
+
"epoch": 0.5053,
|
| 71148 |
+
"grad_norm": 1.9296875,
|
| 71149 |
+
"learning_rate": 0.00014915577889447235,
|
| 71150 |
+
"loss": 1.9997,
|
| 71151 |
+
"step": 101060
|
| 71152 |
+
},
|
| 71153 |
+
{
|
| 71154 |
+
"epoch": 0.50535,
|
| 71155 |
+
"grad_norm": 1.984375,
|
| 71156 |
+
"learning_rate": 0.00014914070351758793,
|
| 71157 |
+
"loss": 2.044,
|
| 71158 |
+
"step": 101070
|
| 71159 |
+
},
|
| 71160 |
+
{
|
| 71161 |
+
"epoch": 0.5054,
|
| 71162 |
+
"grad_norm": 2.03125,
|
| 71163 |
+
"learning_rate": 0.0001491256281407035,
|
| 71164 |
+
"loss": 2.0443,
|
| 71165 |
+
"step": 101080
|
| 71166 |
+
},
|
| 71167 |
+
{
|
| 71168 |
+
"epoch": 0.50545,
|
| 71169 |
+
"grad_norm": 1.9453125,
|
| 71170 |
+
"learning_rate": 0.00014911055276381908,
|
| 71171 |
+
"loss": 2.0243,
|
| 71172 |
+
"step": 101090
|
| 71173 |
+
},
|
| 71174 |
+
{
|
| 71175 |
+
"epoch": 0.5055,
|
| 71176 |
+
"grad_norm": 1.8359375,
|
| 71177 |
+
"learning_rate": 0.00014909547738693467,
|
| 71178 |
+
"loss": 2.0287,
|
| 71179 |
+
"step": 101100
|
| 71180 |
+
},
|
| 71181 |
+
{
|
| 71182 |
+
"epoch": 0.50555,
|
| 71183 |
+
"grad_norm": 1.8046875,
|
| 71184 |
+
"learning_rate": 0.00014908040201005022,
|
| 71185 |
+
"loss": 2.036,
|
| 71186 |
+
"step": 101110
|
| 71187 |
+
},
|
| 71188 |
+
{
|
| 71189 |
+
"epoch": 0.5056,
|
| 71190 |
+
"grad_norm": 2.046875,
|
| 71191 |
+
"learning_rate": 0.0001490653266331658,
|
| 71192 |
+
"loss": 2.0173,
|
| 71193 |
+
"step": 101120
|
| 71194 |
+
},
|
| 71195 |
+
{
|
| 71196 |
+
"epoch": 0.50565,
|
| 71197 |
+
"grad_norm": 1.921875,
|
| 71198 |
+
"learning_rate": 0.0001490502512562814,
|
| 71199 |
+
"loss": 1.9824,
|
| 71200 |
+
"step": 101130
|
| 71201 |
+
},
|
| 71202 |
+
{
|
| 71203 |
+
"epoch": 0.5057,
|
| 71204 |
+
"grad_norm": 1.765625,
|
| 71205 |
+
"learning_rate": 0.00014903517587939698,
|
| 71206 |
+
"loss": 2.0331,
|
| 71207 |
+
"step": 101140
|
| 71208 |
+
},
|
| 71209 |
+
{
|
| 71210 |
+
"epoch": 0.50575,
|
| 71211 |
+
"grad_norm": 2.21875,
|
| 71212 |
+
"learning_rate": 0.00014902010050251257,
|
| 71213 |
+
"loss": 2.0081,
|
| 71214 |
+
"step": 101150
|
| 71215 |
+
},
|
| 71216 |
+
{
|
| 71217 |
+
"epoch": 0.5058,
|
| 71218 |
+
"grad_norm": 1.9921875,
|
| 71219 |
+
"learning_rate": 0.00014900502512562813,
|
| 71220 |
+
"loss": 2.0302,
|
| 71221 |
+
"step": 101160
|
| 71222 |
+
},
|
| 71223 |
+
{
|
| 71224 |
+
"epoch": 0.50585,
|
| 71225 |
+
"grad_norm": 1.8203125,
|
| 71226 |
+
"learning_rate": 0.00014898994974874371,
|
| 71227 |
+
"loss": 2.0202,
|
| 71228 |
+
"step": 101170
|
| 71229 |
+
},
|
| 71230 |
+
{
|
| 71231 |
+
"epoch": 0.5059,
|
| 71232 |
+
"grad_norm": 2.125,
|
| 71233 |
+
"learning_rate": 0.00014897487437185927,
|
| 71234 |
+
"loss": 2.0086,
|
| 71235 |
+
"step": 101180
|
| 71236 |
+
},
|
| 71237 |
+
{
|
| 71238 |
+
"epoch": 0.50595,
|
| 71239 |
+
"grad_norm": 1.96875,
|
| 71240 |
+
"learning_rate": 0.00014895979899497486,
|
| 71241 |
+
"loss": 2.0516,
|
| 71242 |
+
"step": 101190
|
| 71243 |
+
},
|
| 71244 |
+
{
|
| 71245 |
+
"epoch": 0.506,
|
| 71246 |
+
"grad_norm": 1.9453125,
|
| 71247 |
+
"learning_rate": 0.00014894472361809044,
|
| 71248 |
+
"loss": 2.0198,
|
| 71249 |
+
"step": 101200
|
| 71250 |
+
},
|
| 71251 |
+
{
|
| 71252 |
+
"epoch": 0.50605,
|
| 71253 |
+
"grad_norm": 1.8984375,
|
| 71254 |
+
"learning_rate": 0.000148929648241206,
|
| 71255 |
+
"loss": 2.068,
|
| 71256 |
+
"step": 101210
|
| 71257 |
+
},
|
| 71258 |
+
{
|
| 71259 |
+
"epoch": 0.5061,
|
| 71260 |
+
"grad_norm": 1.90625,
|
| 71261 |
+
"learning_rate": 0.0001489145728643216,
|
| 71262 |
+
"loss": 1.9807,
|
| 71263 |
+
"step": 101220
|
| 71264 |
+
},
|
| 71265 |
+
{
|
| 71266 |
+
"epoch": 0.50615,
|
| 71267 |
+
"grad_norm": 1.8828125,
|
| 71268 |
+
"learning_rate": 0.00014889949748743718,
|
| 71269 |
+
"loss": 2.0408,
|
| 71270 |
+
"step": 101230
|
| 71271 |
+
},
|
| 71272 |
+
{
|
| 71273 |
+
"epoch": 0.5062,
|
| 71274 |
+
"grad_norm": 1.9609375,
|
| 71275 |
+
"learning_rate": 0.00014888442211055274,
|
| 71276 |
+
"loss": 2.0004,
|
| 71277 |
+
"step": 101240
|
| 71278 |
+
},
|
| 71279 |
+
{
|
| 71280 |
+
"epoch": 0.50625,
|
| 71281 |
+
"grad_norm": 1.8515625,
|
| 71282 |
+
"learning_rate": 0.00014886934673366832,
|
| 71283 |
+
"loss": 2.0351,
|
| 71284 |
+
"step": 101250
|
| 71285 |
+
},
|
| 71286 |
+
{
|
| 71287 |
+
"epoch": 0.5063,
|
| 71288 |
+
"grad_norm": 2.203125,
|
| 71289 |
+
"learning_rate": 0.0001488542713567839,
|
| 71290 |
+
"loss": 2.0074,
|
| 71291 |
+
"step": 101260
|
| 71292 |
+
},
|
| 71293 |
+
{
|
| 71294 |
+
"epoch": 0.50635,
|
| 71295 |
+
"grad_norm": 1.9296875,
|
| 71296 |
+
"learning_rate": 0.0001488391959798995,
|
| 71297 |
+
"loss": 2.0361,
|
| 71298 |
+
"step": 101270
|
| 71299 |
+
},
|
| 71300 |
+
{
|
| 71301 |
+
"epoch": 0.5064,
|
| 71302 |
+
"grad_norm": 1.8984375,
|
| 71303 |
+
"learning_rate": 0.00014882412060301508,
|
| 71304 |
+
"loss": 2.0219,
|
| 71305 |
+
"step": 101280
|
| 71306 |
+
},
|
| 71307 |
+
{
|
| 71308 |
+
"epoch": 0.50645,
|
| 71309 |
+
"grad_norm": 2.0,
|
| 71310 |
+
"learning_rate": 0.00014880904522613064,
|
| 71311 |
+
"loss": 2.0133,
|
| 71312 |
+
"step": 101290
|
| 71313 |
+
},
|
| 71314 |
+
{
|
| 71315 |
+
"epoch": 0.5065,
|
| 71316 |
+
"grad_norm": 1.796875,
|
| 71317 |
+
"learning_rate": 0.00014879396984924622,
|
| 71318 |
+
"loss": 1.9861,
|
| 71319 |
+
"step": 101300
|
| 71320 |
+
},
|
| 71321 |
+
{
|
| 71322 |
+
"epoch": 0.50655,
|
| 71323 |
+
"grad_norm": 1.9375,
|
| 71324 |
+
"learning_rate": 0.0001487788944723618,
|
| 71325 |
+
"loss": 2.0125,
|
| 71326 |
+
"step": 101310
|
| 71327 |
+
},
|
| 71328 |
+
{
|
| 71329 |
+
"epoch": 0.5066,
|
| 71330 |
+
"grad_norm": 1.7265625,
|
| 71331 |
+
"learning_rate": 0.00014876381909547737,
|
| 71332 |
+
"loss": 2.0377,
|
| 71333 |
+
"step": 101320
|
| 71334 |
+
},
|
| 71335 |
+
{
|
| 71336 |
+
"epoch": 0.50665,
|
| 71337 |
+
"grad_norm": 2.0,
|
| 71338 |
+
"learning_rate": 0.00014874874371859296,
|
| 71339 |
+
"loss": 2.0377,
|
| 71340 |
+
"step": 101330
|
| 71341 |
+
},
|
| 71342 |
+
{
|
| 71343 |
+
"epoch": 0.5067,
|
| 71344 |
+
"grad_norm": 1.921875,
|
| 71345 |
+
"learning_rate": 0.00014873366834170851,
|
| 71346 |
+
"loss": 2.0038,
|
| 71347 |
+
"step": 101340
|
| 71348 |
+
},
|
| 71349 |
+
{
|
| 71350 |
+
"epoch": 0.50675,
|
| 71351 |
+
"grad_norm": 1.828125,
|
| 71352 |
+
"learning_rate": 0.0001487185929648241,
|
| 71353 |
+
"loss": 2.0242,
|
| 71354 |
+
"step": 101350
|
| 71355 |
+
},
|
| 71356 |
+
{
|
| 71357 |
+
"epoch": 0.5068,
|
| 71358 |
+
"grad_norm": 1.8984375,
|
| 71359 |
+
"learning_rate": 0.0001487035175879397,
|
| 71360 |
+
"loss": 2.0106,
|
| 71361 |
+
"step": 101360
|
| 71362 |
+
},
|
| 71363 |
+
{
|
| 71364 |
+
"epoch": 0.50685,
|
| 71365 |
+
"grad_norm": 1.8125,
|
| 71366 |
+
"learning_rate": 0.00014868844221105525,
|
| 71367 |
+
"loss": 2.0362,
|
| 71368 |
+
"step": 101370
|
| 71369 |
+
},
|
| 71370 |
+
{
|
| 71371 |
+
"epoch": 0.5069,
|
| 71372 |
+
"grad_norm": 2.078125,
|
| 71373 |
+
"learning_rate": 0.00014867336683417083,
|
| 71374 |
+
"loss": 2.0324,
|
| 71375 |
+
"step": 101380
|
| 71376 |
+
},
|
| 71377 |
+
{
|
| 71378 |
+
"epoch": 0.50695,
|
| 71379 |
+
"grad_norm": 1.921875,
|
| 71380 |
+
"learning_rate": 0.00014865829145728642,
|
| 71381 |
+
"loss": 2.0529,
|
| 71382 |
+
"step": 101390
|
| 71383 |
+
},
|
| 71384 |
+
{
|
| 71385 |
+
"epoch": 0.507,
|
| 71386 |
+
"grad_norm": 1.9765625,
|
| 71387 |
+
"learning_rate": 0.000148643216080402,
|
| 71388 |
+
"loss": 2.0232,
|
| 71389 |
+
"step": 101400
|
| 71390 |
+
},
|
| 71391 |
+
{
|
| 71392 |
+
"epoch": 0.50705,
|
| 71393 |
+
"grad_norm": 1.9296875,
|
| 71394 |
+
"learning_rate": 0.0001486281407035176,
|
| 71395 |
+
"loss": 2.0135,
|
| 71396 |
+
"step": 101410
|
| 71397 |
+
},
|
| 71398 |
+
{
|
| 71399 |
+
"epoch": 0.5071,
|
| 71400 |
+
"grad_norm": 1.875,
|
| 71401 |
+
"learning_rate": 0.00014861306532663315,
|
| 71402 |
+
"loss": 1.9847,
|
| 71403 |
+
"step": 101420
|
| 71404 |
+
},
|
| 71405 |
+
{
|
| 71406 |
+
"epoch": 0.50715,
|
| 71407 |
+
"grad_norm": 2.046875,
|
| 71408 |
+
"learning_rate": 0.00014859798994974874,
|
| 71409 |
+
"loss": 2.0022,
|
| 71410 |
+
"step": 101430
|
| 71411 |
+
},
|
| 71412 |
+
{
|
| 71413 |
+
"epoch": 0.5072,
|
| 71414 |
+
"grad_norm": 2.09375,
|
| 71415 |
+
"learning_rate": 0.00014858291457286432,
|
| 71416 |
+
"loss": 2.0359,
|
| 71417 |
+
"step": 101440
|
| 71418 |
+
},
|
| 71419 |
+
{
|
| 71420 |
+
"epoch": 0.50725,
|
| 71421 |
+
"grad_norm": 2.0,
|
| 71422 |
+
"learning_rate": 0.00014856783919597988,
|
| 71423 |
+
"loss": 1.9862,
|
| 71424 |
+
"step": 101450
|
| 71425 |
+
},
|
| 71426 |
+
{
|
| 71427 |
+
"epoch": 0.5073,
|
| 71428 |
+
"grad_norm": 2.15625,
|
| 71429 |
+
"learning_rate": 0.00014855276381909547,
|
| 71430 |
+
"loss": 2.0993,
|
| 71431 |
+
"step": 101460
|
| 71432 |
+
},
|
| 71433 |
+
{
|
| 71434 |
+
"epoch": 0.50735,
|
| 71435 |
+
"grad_norm": 1.84375,
|
| 71436 |
+
"learning_rate": 0.00014853768844221105,
|
| 71437 |
+
"loss": 1.9999,
|
| 71438 |
+
"step": 101470
|
| 71439 |
+
},
|
| 71440 |
+
{
|
| 71441 |
+
"epoch": 0.5074,
|
| 71442 |
+
"grad_norm": 2.21875,
|
| 71443 |
+
"learning_rate": 0.0001485226130653266,
|
| 71444 |
+
"loss": 2.0576,
|
| 71445 |
+
"step": 101480
|
| 71446 |
+
},
|
| 71447 |
+
{
|
| 71448 |
+
"epoch": 0.50745,
|
| 71449 |
+
"grad_norm": 1.96875,
|
| 71450 |
+
"learning_rate": 0.0001485075376884422,
|
| 71451 |
+
"loss": 2.0463,
|
| 71452 |
+
"step": 101490
|
| 71453 |
+
},
|
| 71454 |
+
{
|
| 71455 |
+
"epoch": 0.5075,
|
| 71456 |
+
"grad_norm": 2.046875,
|
| 71457 |
+
"learning_rate": 0.00014849246231155776,
|
| 71458 |
+
"loss": 1.9991,
|
| 71459 |
+
"step": 101500
|
| 71460 |
+
},
|
| 71461 |
+
{
|
| 71462 |
+
"epoch": 0.50755,
|
| 71463 |
+
"grad_norm": 2.015625,
|
| 71464 |
+
"learning_rate": 0.00014847738693467334,
|
| 71465 |
+
"loss": 2.0225,
|
| 71466 |
+
"step": 101510
|
| 71467 |
+
},
|
| 71468 |
+
{
|
| 71469 |
+
"epoch": 0.5076,
|
| 71470 |
+
"grad_norm": 2.0,
|
| 71471 |
+
"learning_rate": 0.00014846231155778893,
|
| 71472 |
+
"loss": 1.9933,
|
| 71473 |
+
"step": 101520
|
| 71474 |
+
},
|
| 71475 |
+
{
|
| 71476 |
+
"epoch": 0.50765,
|
| 71477 |
+
"grad_norm": 2.03125,
|
| 71478 |
+
"learning_rate": 0.00014844723618090452,
|
| 71479 |
+
"loss": 2.068,
|
| 71480 |
+
"step": 101530
|
| 71481 |
+
},
|
| 71482 |
+
{
|
| 71483 |
+
"epoch": 0.5077,
|
| 71484 |
+
"grad_norm": 2.140625,
|
| 71485 |
+
"learning_rate": 0.0001484321608040201,
|
| 71486 |
+
"loss": 2.0123,
|
| 71487 |
+
"step": 101540
|
| 71488 |
+
},
|
| 71489 |
+
{
|
| 71490 |
+
"epoch": 0.50775,
|
| 71491 |
+
"grad_norm": 2.046875,
|
| 71492 |
+
"learning_rate": 0.00014841708542713566,
|
| 71493 |
+
"loss": 2.0615,
|
| 71494 |
+
"step": 101550
|
| 71495 |
+
},
|
| 71496 |
+
{
|
| 71497 |
+
"epoch": 0.5078,
|
| 71498 |
+
"grad_norm": 1.78125,
|
| 71499 |
+
"learning_rate": 0.00014840201005025125,
|
| 71500 |
+
"loss": 2.0239,
|
| 71501 |
+
"step": 101560
|
| 71502 |
+
},
|
| 71503 |
+
{
|
| 71504 |
+
"epoch": 0.50785,
|
| 71505 |
+
"grad_norm": 1.9140625,
|
| 71506 |
+
"learning_rate": 0.00014838693467336683,
|
| 71507 |
+
"loss": 2.0751,
|
| 71508 |
+
"step": 101570
|
| 71509 |
+
},
|
| 71510 |
+
{
|
| 71511 |
+
"epoch": 0.5079,
|
| 71512 |
+
"grad_norm": 2.0625,
|
| 71513 |
+
"learning_rate": 0.0001483718592964824,
|
| 71514 |
+
"loss": 2.0723,
|
| 71515 |
+
"step": 101580
|
| 71516 |
+
},
|
| 71517 |
+
{
|
| 71518 |
+
"epoch": 0.50795,
|
| 71519 |
+
"grad_norm": 2.265625,
|
| 71520 |
+
"learning_rate": 0.00014835678391959798,
|
| 71521 |
+
"loss": 1.9995,
|
| 71522 |
+
"step": 101590
|
| 71523 |
+
},
|
| 71524 |
+
{
|
| 71525 |
+
"epoch": 0.508,
|
| 71526 |
+
"grad_norm": 1.9765625,
|
| 71527 |
+
"learning_rate": 0.00014834170854271356,
|
| 71528 |
+
"loss": 1.9753,
|
| 71529 |
+
"step": 101600
|
| 71530 |
+
},
|
| 71531 |
+
{
|
| 71532 |
+
"epoch": 0.50805,
|
| 71533 |
+
"grad_norm": 2.046875,
|
| 71534 |
+
"learning_rate": 0.00014832663316582912,
|
| 71535 |
+
"loss": 1.977,
|
| 71536 |
+
"step": 101610
|
| 71537 |
+
},
|
| 71538 |
+
{
|
| 71539 |
+
"epoch": 0.5081,
|
| 71540 |
+
"grad_norm": 2.09375,
|
| 71541 |
+
"learning_rate": 0.0001483115577889447,
|
| 71542 |
+
"loss": 2.0894,
|
| 71543 |
+
"step": 101620
|
| 71544 |
+
},
|
| 71545 |
+
{
|
| 71546 |
+
"epoch": 0.50815,
|
| 71547 |
+
"grad_norm": 1.9140625,
|
| 71548 |
+
"learning_rate": 0.0001482964824120603,
|
| 71549 |
+
"loss": 1.9872,
|
| 71550 |
+
"step": 101630
|
| 71551 |
+
},
|
| 71552 |
+
{
|
| 71553 |
+
"epoch": 0.5082,
|
| 71554 |
+
"grad_norm": 2.03125,
|
| 71555 |
+
"learning_rate": 0.00014828140703517585,
|
| 71556 |
+
"loss": 2.0307,
|
| 71557 |
+
"step": 101640
|
| 71558 |
+
},
|
| 71559 |
+
{
|
| 71560 |
+
"epoch": 0.50825,
|
| 71561 |
+
"grad_norm": 1.7109375,
|
| 71562 |
+
"learning_rate": 0.00014826633165829144,
|
| 71563 |
+
"loss": 2.0058,
|
| 71564 |
+
"step": 101650
|
| 71565 |
+
},
|
| 71566 |
+
{
|
| 71567 |
+
"epoch": 0.5083,
|
| 71568 |
+
"grad_norm": 1.8828125,
|
| 71569 |
+
"learning_rate": 0.00014825125628140703,
|
| 71570 |
+
"loss": 2.0163,
|
| 71571 |
+
"step": 101660
|
| 71572 |
+
},
|
| 71573 |
+
{
|
| 71574 |
+
"epoch": 0.50835,
|
| 71575 |
+
"grad_norm": 1.7734375,
|
| 71576 |
+
"learning_rate": 0.0001482361809045226,
|
| 71577 |
+
"loss": 2.0325,
|
| 71578 |
+
"step": 101670
|
| 71579 |
+
},
|
| 71580 |
+
{
|
| 71581 |
+
"epoch": 0.5084,
|
| 71582 |
+
"grad_norm": 1.9296875,
|
| 71583 |
+
"learning_rate": 0.00014822110552763817,
|
| 71584 |
+
"loss": 2.0308,
|
| 71585 |
+
"step": 101680
|
| 71586 |
+
},
|
| 71587 |
+
{
|
| 71588 |
+
"epoch": 0.50845,
|
| 71589 |
+
"grad_norm": 1.9921875,
|
| 71590 |
+
"learning_rate": 0.00014820603015075376,
|
| 71591 |
+
"loss": 2.0286,
|
| 71592 |
+
"step": 101690
|
| 71593 |
+
},
|
| 71594 |
+
{
|
| 71595 |
+
"epoch": 0.5085,
|
| 71596 |
+
"grad_norm": 2.015625,
|
| 71597 |
+
"learning_rate": 0.00014819095477386934,
|
| 71598 |
+
"loss": 2.0159,
|
| 71599 |
+
"step": 101700
|
| 71600 |
+
},
|
| 71601 |
+
{
|
| 71602 |
+
"epoch": 0.50855,
|
| 71603 |
+
"grad_norm": 1.890625,
|
| 71604 |
+
"learning_rate": 0.0001481758793969849,
|
| 71605 |
+
"loss": 2.04,
|
| 71606 |
+
"step": 101710
|
| 71607 |
+
},
|
| 71608 |
+
{
|
| 71609 |
+
"epoch": 0.5086,
|
| 71610 |
+
"grad_norm": 1.9296875,
|
| 71611 |
+
"learning_rate": 0.0001481608040201005,
|
| 71612 |
+
"loss": 2.0119,
|
| 71613 |
+
"step": 101720
|
| 71614 |
+
},
|
| 71615 |
+
{
|
| 71616 |
+
"epoch": 0.50865,
|
| 71617 |
+
"grad_norm": 2.0,
|
| 71618 |
+
"learning_rate": 0.00014814572864321607,
|
| 71619 |
+
"loss": 1.9724,
|
| 71620 |
+
"step": 101730
|
| 71621 |
+
},
|
| 71622 |
+
{
|
| 71623 |
+
"epoch": 0.5087,
|
| 71624 |
+
"grad_norm": 2.078125,
|
| 71625 |
+
"learning_rate": 0.00014813065326633163,
|
| 71626 |
+
"loss": 2.0856,
|
| 71627 |
+
"step": 101740
|
| 71628 |
+
},
|
| 71629 |
+
{
|
| 71630 |
+
"epoch": 0.50875,
|
| 71631 |
+
"grad_norm": 1.9375,
|
| 71632 |
+
"learning_rate": 0.00014811557788944722,
|
| 71633 |
+
"loss": 2.027,
|
| 71634 |
+
"step": 101750
|
| 71635 |
+
},
|
| 71636 |
+
{
|
| 71637 |
+
"epoch": 0.5088,
|
| 71638 |
+
"grad_norm": 2.015625,
|
| 71639 |
+
"learning_rate": 0.0001481005025125628,
|
| 71640 |
+
"loss": 2.0297,
|
| 71641 |
+
"step": 101760
|
| 71642 |
+
},
|
| 71643 |
+
{
|
| 71644 |
+
"epoch": 0.50885,
|
| 71645 |
+
"grad_norm": 1.9375,
|
| 71646 |
+
"learning_rate": 0.00014808542713567836,
|
| 71647 |
+
"loss": 2.0425,
|
| 71648 |
+
"step": 101770
|
| 71649 |
+
},
|
| 71650 |
+
{
|
| 71651 |
+
"epoch": 0.5089,
|
| 71652 |
+
"grad_norm": 2.40625,
|
| 71653 |
+
"learning_rate": 0.00014807035175879395,
|
| 71654 |
+
"loss": 2.0579,
|
| 71655 |
+
"step": 101780
|
| 71656 |
+
},
|
| 71657 |
+
{
|
| 71658 |
+
"epoch": 0.50895,
|
| 71659 |
+
"grad_norm": 1.9140625,
|
| 71660 |
+
"learning_rate": 0.00014805527638190954,
|
| 71661 |
+
"loss": 1.9863,
|
| 71662 |
+
"step": 101790
|
| 71663 |
+
},
|
| 71664 |
+
{
|
| 71665 |
+
"epoch": 0.509,
|
| 71666 |
+
"grad_norm": 2.0625,
|
| 71667 |
+
"learning_rate": 0.00014804020100502512,
|
| 71668 |
+
"loss": 1.9923,
|
| 71669 |
+
"step": 101800
|
| 71670 |
+
},
|
| 71671 |
+
{
|
| 71672 |
+
"epoch": 0.50905,
|
| 71673 |
+
"grad_norm": 2.109375,
|
| 71674 |
+
"learning_rate": 0.00014802512562814068,
|
| 71675 |
+
"loss": 2.0216,
|
| 71676 |
+
"step": 101810
|
| 71677 |
+
},
|
| 71678 |
+
{
|
| 71679 |
+
"epoch": 0.5091,
|
| 71680 |
+
"grad_norm": 2.140625,
|
| 71681 |
+
"learning_rate": 0.00014801005025125627,
|
| 71682 |
+
"loss": 2.0225,
|
| 71683 |
+
"step": 101820
|
| 71684 |
+
},
|
| 71685 |
+
{
|
| 71686 |
+
"epoch": 0.50915,
|
| 71687 |
+
"grad_norm": 1.921875,
|
| 71688 |
+
"learning_rate": 0.00014799497487437185,
|
| 71689 |
+
"loss": 1.9862,
|
| 71690 |
+
"step": 101830
|
| 71691 |
+
},
|
| 71692 |
+
{
|
| 71693 |
+
"epoch": 0.5092,
|
| 71694 |
+
"grad_norm": 1.984375,
|
| 71695 |
+
"learning_rate": 0.00014797989949748744,
|
| 71696 |
+
"loss": 1.9726,
|
| 71697 |
+
"step": 101840
|
| 71698 |
+
},
|
| 71699 |
+
{
|
| 71700 |
+
"epoch": 0.50925,
|
| 71701 |
+
"grad_norm": 1.8359375,
|
| 71702 |
+
"learning_rate": 0.000147964824120603,
|
| 71703 |
+
"loss": 2.0245,
|
| 71704 |
+
"step": 101850
|
| 71705 |
+
},
|
| 71706 |
+
{
|
| 71707 |
+
"epoch": 0.5093,
|
| 71708 |
+
"grad_norm": 1.890625,
|
| 71709 |
+
"learning_rate": 0.00014794974874371859,
|
| 71710 |
+
"loss": 2.0225,
|
| 71711 |
+
"step": 101860
|
| 71712 |
+
},
|
| 71713 |
+
{
|
| 71714 |
+
"epoch": 0.50935,
|
| 71715 |
+
"grad_norm": 1.9609375,
|
| 71716 |
+
"learning_rate": 0.00014793467336683414,
|
| 71717 |
+
"loss": 2.0797,
|
| 71718 |
+
"step": 101870
|
| 71719 |
+
},
|
| 71720 |
+
{
|
| 71721 |
+
"epoch": 0.5094,
|
| 71722 |
+
"grad_norm": 1.9765625,
|
| 71723 |
+
"learning_rate": 0.00014791959798994973,
|
| 71724 |
+
"loss": 2.0775,
|
| 71725 |
+
"step": 101880
|
| 71726 |
+
},
|
| 71727 |
+
{
|
| 71728 |
+
"epoch": 0.50945,
|
| 71729 |
+
"grad_norm": 2.03125,
|
| 71730 |
+
"learning_rate": 0.00014790452261306532,
|
| 71731 |
+
"loss": 2.0884,
|
| 71732 |
+
"step": 101890
|
| 71733 |
+
},
|
| 71734 |
+
{
|
| 71735 |
+
"epoch": 0.5095,
|
| 71736 |
+
"grad_norm": 1.71875,
|
| 71737 |
+
"learning_rate": 0.00014788944723618088,
|
| 71738 |
+
"loss": 2.0254,
|
| 71739 |
+
"step": 101900
|
| 71740 |
+
},
|
| 71741 |
+
{
|
| 71742 |
+
"epoch": 0.50955,
|
| 71743 |
+
"grad_norm": 1.90625,
|
| 71744 |
+
"learning_rate": 0.00014787437185929646,
|
| 71745 |
+
"loss": 1.9893,
|
| 71746 |
+
"step": 101910
|
| 71747 |
+
},
|
| 71748 |
+
{
|
| 71749 |
+
"epoch": 0.5096,
|
| 71750 |
+
"grad_norm": 1.9296875,
|
| 71751 |
+
"learning_rate": 0.00014785929648241205,
|
| 71752 |
+
"loss": 2.0477,
|
| 71753 |
+
"step": 101920
|
| 71754 |
+
},
|
| 71755 |
+
{
|
| 71756 |
+
"epoch": 0.50965,
|
| 71757 |
+
"grad_norm": 2.125,
|
| 71758 |
+
"learning_rate": 0.00014784422110552763,
|
| 71759 |
+
"loss": 2.0333,
|
| 71760 |
+
"step": 101930
|
| 71761 |
+
},
|
| 71762 |
+
{
|
| 71763 |
+
"epoch": 0.5097,
|
| 71764 |
+
"grad_norm": 1.828125,
|
| 71765 |
+
"learning_rate": 0.0001478291457286432,
|
| 71766 |
+
"loss": 2.009,
|
| 71767 |
+
"step": 101940
|
| 71768 |
+
},
|
| 71769 |
+
{
|
| 71770 |
+
"epoch": 0.50975,
|
| 71771 |
+
"grad_norm": 2.046875,
|
| 71772 |
+
"learning_rate": 0.00014781407035175878,
|
| 71773 |
+
"loss": 2.0322,
|
| 71774 |
+
"step": 101950
|
| 71775 |
+
},
|
| 71776 |
+
{
|
| 71777 |
+
"epoch": 0.5098,
|
| 71778 |
+
"grad_norm": 2.03125,
|
| 71779 |
+
"learning_rate": 0.00014779899497487437,
|
| 71780 |
+
"loss": 1.9683,
|
| 71781 |
+
"step": 101960
|
| 71782 |
+
},
|
| 71783 |
+
{
|
| 71784 |
+
"epoch": 0.50985,
|
| 71785 |
+
"grad_norm": 1.828125,
|
| 71786 |
+
"learning_rate": 0.00014778391959798995,
|
| 71787 |
+
"loss": 2.0255,
|
| 71788 |
+
"step": 101970
|
| 71789 |
+
},
|
| 71790 |
+
{
|
| 71791 |
+
"epoch": 0.5099,
|
| 71792 |
+
"grad_norm": 2.09375,
|
| 71793 |
+
"learning_rate": 0.0001477688442211055,
|
| 71794 |
+
"loss": 2.016,
|
| 71795 |
+
"step": 101980
|
| 71796 |
+
},
|
| 71797 |
+
{
|
| 71798 |
+
"epoch": 0.50995,
|
| 71799 |
+
"grad_norm": 1.8515625,
|
| 71800 |
+
"learning_rate": 0.0001477537688442211,
|
| 71801 |
+
"loss": 2.0768,
|
| 71802 |
+
"step": 101990
|
| 71803 |
+
},
|
| 71804 |
+
{
|
| 71805 |
+
"epoch": 0.51,
|
| 71806 |
+
"grad_norm": 2.109375,
|
| 71807 |
+
"learning_rate": 0.00014773869346733668,
|
| 71808 |
+
"loss": 1.9938,
|
| 71809 |
+
"step": 102000
|
| 71810 |
+
},
|
| 71811 |
+
{
|
| 71812 |
+
"epoch": 0.51,
|
| 71813 |
+
"eval_loss": 2.0225799083709717,
|
| 71814 |
+
"eval_runtime": 90.2278,
|
| 71815 |
+
"eval_samples_per_second": 27.708,
|
| 71816 |
+
"eval_steps_per_second": 0.443,
|
| 71817 |
+
"step": 102000
|
| 71818 |
}
|
| 71819 |
],
|
| 71820 |
"logging_steps": 10,
|
|
|
|
| 71834 |
"attributes": {}
|
| 71835 |
}
|
| 71836 |
},
|
| 71837 |
+
"total_flos": 2.471196266234852e+19,
|
| 71838 |
"train_batch_size": 8,
|
| 71839 |
"trial_name": null,
|
| 71840 |
"trial_params": null
|