Training in progress, step 2000, 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 791781368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eccb0e2bd66b47391dd89be4927a9aed1333b9e1e0b7d9f1951a4aaeb35401e8
|
| 3 |
size 791781368
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2375487866
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12fb28e52c48bea00a85883fac40f131eedf9cbb7d2564050dd0b64aff066681
|
| 3 |
size 2375487866
|
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:c22704cb68ce2d335ed3aa00237997dfcf49180d12a11fa5fb434d67c714d2da
|
| 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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:41c88f9de084200454883a13c3717941ea3fd433e2f8735507fc30611f9c5501
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:965b00d4cb4710ebab57c8787b9925bb3f77b8eeba94a186ec4bc1c2f326ef3f
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5dc374b8b9a4c45c950f9d136feab85a767081fa59f0c7d68ed3a62060c4949
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c7c212fb779217f1edac0baf44f67b608eefc1e0e4e3f5a9dd7eb557032c1bc
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86e1effd626ce1e95dd68a0c8089fe19218f2b24dfe9e45ed2cab1c0ebc10ba1
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:799cc83f60dfc1c4243cfd6403592112414a2eec494e6832f10221c96ff62c20
|
| 3 |
size 15984
|
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 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:586777c398770c3255d3a1f48c7fef44ea9d89117c627c9ea490e16bfd9a49ba
|
| 3 |
size 15984
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1000
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fd44b9ad3ef5591f2a0671f1ec04c21ad479cbd1d478859e3ba017f1c74bf027
|
| 3 |
size 1000
|
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": 1000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1415,6 +1415,1414 @@
|
|
| 1415 |
"eval_samples_per_second": 998.04,
|
| 1416 |
"eval_steps_per_second": 31.191,
|
| 1417 |
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1418 |
}
|
| 1419 |
],
|
| 1420 |
"logging_steps": 5,
|
|
@@ -1434,7 +2842,7 @@
|
|
| 1434 |
"attributes": {}
|
| 1435 |
}
|
| 1436 |
},
|
| 1437 |
-
"total_flos":
|
| 1438 |
"train_batch_size": 8,
|
| 1439 |
"trial_name": null,
|
| 1440 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.8762202050902917,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 2000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1415 |
"eval_samples_per_second": 998.04,
|
| 1416 |
"eval_steps_per_second": 31.191,
|
| 1417 |
"step": 1000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.4403006530578716,
|
| 1421 |
+
"grad_norm": 158.5,
|
| 1422 |
+
"learning_rate": 1.2440331222601072e-05,
|
| 1423 |
+
"loss": 25.968,
|
| 1424 |
+
"step": 1005
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.44249120357059735,
|
| 1428 |
+
"grad_norm": 66.1875,
|
| 1429 |
+
"learning_rate": 1.2391622016561131e-05,
|
| 1430 |
+
"loss": 25.9174,
|
| 1431 |
+
"step": 1010
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.44468175408332306,
|
| 1435 |
+
"grad_norm": 68.75,
|
| 1436 |
+
"learning_rate": 1.234291281052119e-05,
|
| 1437 |
+
"loss": 25.4954,
|
| 1438 |
+
"step": 1015
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 0.4468723045960488,
|
| 1442 |
+
"grad_norm": 130.75,
|
| 1443 |
+
"learning_rate": 1.2294203604481248e-05,
|
| 1444 |
+
"loss": 25.5188,
|
| 1445 |
+
"step": 1020
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.44906285510877453,
|
| 1449 |
+
"grad_norm": 112.3125,
|
| 1450 |
+
"learning_rate": 1.2245494398441307e-05,
|
| 1451 |
+
"loss": 25.8739,
|
| 1452 |
+
"step": 1025
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.45125340562150024,
|
| 1456 |
+
"grad_norm": 117.0,
|
| 1457 |
+
"learning_rate": 1.2196785192401366e-05,
|
| 1458 |
+
"loss": 25.0384,
|
| 1459 |
+
"step": 1030
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.453443956134226,
|
| 1463 |
+
"grad_norm": 77.375,
|
| 1464 |
+
"learning_rate": 1.2148075986361424e-05,
|
| 1465 |
+
"loss": 25.6608,
|
| 1466 |
+
"step": 1035
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.4556345066469517,
|
| 1470 |
+
"grad_norm": 60.34375,
|
| 1471 |
+
"learning_rate": 1.2099366780321483e-05,
|
| 1472 |
+
"loss": 25.7742,
|
| 1473 |
+
"step": 1040
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.4578250571596774,
|
| 1477 |
+
"grad_norm": 78.5625,
|
| 1478 |
+
"learning_rate": 1.2050657574281542e-05,
|
| 1479 |
+
"loss": 25.5131,
|
| 1480 |
+
"step": 1045
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.4600156076724032,
|
| 1484 |
+
"grad_norm": 115.1875,
|
| 1485 |
+
"learning_rate": 1.2001948368241598e-05,
|
| 1486 |
+
"loss": 25.5009,
|
| 1487 |
+
"step": 1050
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.4622061581851289,
|
| 1491 |
+
"grad_norm": 118.4375,
|
| 1492 |
+
"learning_rate": 1.1953239162201657e-05,
|
| 1493 |
+
"loss": 25.7924,
|
| 1494 |
+
"step": 1055
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.46439670869785465,
|
| 1498 |
+
"grad_norm": 68.375,
|
| 1499 |
+
"learning_rate": 1.1904529956161715e-05,
|
| 1500 |
+
"loss": 25.9603,
|
| 1501 |
+
"step": 1060
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 0.46658725921058036,
|
| 1505 |
+
"grad_norm": 121.375,
|
| 1506 |
+
"learning_rate": 1.1855820750121774e-05,
|
| 1507 |
+
"loss": 25.6717,
|
| 1508 |
+
"step": 1065
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.46877780972330607,
|
| 1512 |
+
"grad_norm": 83.6875,
|
| 1513 |
+
"learning_rate": 1.1807111544081833e-05,
|
| 1514 |
+
"loss": 25.6563,
|
| 1515 |
+
"step": 1070
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.47096836023603184,
|
| 1519 |
+
"grad_norm": 64.5625,
|
| 1520 |
+
"learning_rate": 1.175840233804189e-05,
|
| 1521 |
+
"loss": 26.0566,
|
| 1522 |
+
"step": 1075
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.47315891074875754,
|
| 1526 |
+
"grad_norm": 81.6875,
|
| 1527 |
+
"learning_rate": 1.170969313200195e-05,
|
| 1528 |
+
"loss": 25.7885,
|
| 1529 |
+
"step": 1080
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.47534946126148325,
|
| 1533 |
+
"grad_norm": 100.1875,
|
| 1534 |
+
"learning_rate": 1.1660983925962009e-05,
|
| 1535 |
+
"loss": 25.5735,
|
| 1536 |
+
"step": 1085
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.477540011774209,
|
| 1540 |
+
"grad_norm": 116.0,
|
| 1541 |
+
"learning_rate": 1.1612274719922065e-05,
|
| 1542 |
+
"loss": 25.7891,
|
| 1543 |
+
"step": 1090
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 0.4797305622869347,
|
| 1547 |
+
"grad_norm": 70.125,
|
| 1548 |
+
"learning_rate": 1.1563565513882124e-05,
|
| 1549 |
+
"loss": 25.6365,
|
| 1550 |
+
"step": 1095
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.4819211127996605,
|
| 1554 |
+
"grad_norm": 52.96875,
|
| 1555 |
+
"learning_rate": 1.1514856307842183e-05,
|
| 1556 |
+
"loss": 25.5813,
|
| 1557 |
+
"step": 1100
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.4841116633123862,
|
| 1561 |
+
"grad_norm": 150.125,
|
| 1562 |
+
"learning_rate": 1.1466147101802241e-05,
|
| 1563 |
+
"loss": 26.0147,
|
| 1564 |
+
"step": 1105
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 0.4863022138251119,
|
| 1568 |
+
"grad_norm": 125.0625,
|
| 1569 |
+
"learning_rate": 1.14174378957623e-05,
|
| 1570 |
+
"loss": 26.1209,
|
| 1571 |
+
"step": 1110
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.48849276433783767,
|
| 1575 |
+
"grad_norm": 89.875,
|
| 1576 |
+
"learning_rate": 1.136872868972236e-05,
|
| 1577 |
+
"loss": 25.7375,
|
| 1578 |
+
"step": 1115
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.4906833148505634,
|
| 1582 |
+
"grad_norm": 125.5625,
|
| 1583 |
+
"learning_rate": 1.1320019483682417e-05,
|
| 1584 |
+
"loss": 25.246,
|
| 1585 |
+
"step": 1120
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.4928738653632891,
|
| 1589 |
+
"grad_norm": 96.625,
|
| 1590 |
+
"learning_rate": 1.1271310277642476e-05,
|
| 1591 |
+
"loss": 25.7476,
|
| 1592 |
+
"step": 1125
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.49506441587601485,
|
| 1596 |
+
"grad_norm": 148.375,
|
| 1597 |
+
"learning_rate": 1.1222601071602535e-05,
|
| 1598 |
+
"loss": 25.0299,
|
| 1599 |
+
"step": 1130
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.49725496638874056,
|
| 1603 |
+
"grad_norm": 131.25,
|
| 1604 |
+
"learning_rate": 1.1173891865562591e-05,
|
| 1605 |
+
"loss": 25.7002,
|
| 1606 |
+
"step": 1135
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 0.4994455169014663,
|
| 1610 |
+
"grad_norm": 99.0,
|
| 1611 |
+
"learning_rate": 1.112518265952265e-05,
|
| 1612 |
+
"loss": 25.3308,
|
| 1613 |
+
"step": 1140
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 0.501636067414192,
|
| 1617 |
+
"grad_norm": 52.90625,
|
| 1618 |
+
"learning_rate": 1.107647345348271e-05,
|
| 1619 |
+
"loss": 25.7107,
|
| 1620 |
+
"step": 1145
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 0.5038266179269177,
|
| 1624 |
+
"grad_norm": 92.375,
|
| 1625 |
+
"learning_rate": 1.1027764247442767e-05,
|
| 1626 |
+
"loss": 25.6609,
|
| 1627 |
+
"step": 1150
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 0.5060171684396435,
|
| 1631 |
+
"grad_norm": 59.46875,
|
| 1632 |
+
"learning_rate": 1.0979055041402826e-05,
|
| 1633 |
+
"loss": 25.5677,
|
| 1634 |
+
"step": 1155
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.5082077189523693,
|
| 1638 |
+
"grad_norm": 96.375,
|
| 1639 |
+
"learning_rate": 1.0930345835362886e-05,
|
| 1640 |
+
"loss": 25.3615,
|
| 1641 |
+
"step": 1160
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.5103982694650949,
|
| 1645 |
+
"grad_norm": 75.0,
|
| 1646 |
+
"learning_rate": 1.0881636629322943e-05,
|
| 1647 |
+
"loss": 25.4644,
|
| 1648 |
+
"step": 1165
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.5125888199778207,
|
| 1652 |
+
"grad_norm": 98.5,
|
| 1653 |
+
"learning_rate": 1.0832927423283002e-05,
|
| 1654 |
+
"loss": 25.6878,
|
| 1655 |
+
"step": 1170
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.5147793704905465,
|
| 1659 |
+
"grad_norm": 76.1875,
|
| 1660 |
+
"learning_rate": 1.0784218217243058e-05,
|
| 1661 |
+
"loss": 25.8905,
|
| 1662 |
+
"step": 1175
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.5169699210032721,
|
| 1666 |
+
"grad_norm": 52.4375,
|
| 1667 |
+
"learning_rate": 1.0735509011203117e-05,
|
| 1668 |
+
"loss": 25.7829,
|
| 1669 |
+
"step": 1180
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.5191604715159979,
|
| 1673 |
+
"grad_norm": 105.625,
|
| 1674 |
+
"learning_rate": 1.0686799805163177e-05,
|
| 1675 |
+
"loss": 25.244,
|
| 1676 |
+
"step": 1185
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.5213510220287236,
|
| 1680 |
+
"grad_norm": 59.84375,
|
| 1681 |
+
"learning_rate": 1.0638090599123234e-05,
|
| 1682 |
+
"loss": 25.1215,
|
| 1683 |
+
"step": 1190
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.5235415725414493,
|
| 1687 |
+
"grad_norm": 54.03125,
|
| 1688 |
+
"learning_rate": 1.0589381393083293e-05,
|
| 1689 |
+
"loss": 25.3532,
|
| 1690 |
+
"step": 1195
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.525732123054175,
|
| 1694 |
+
"grad_norm": 134.0,
|
| 1695 |
+
"learning_rate": 1.0540672187043353e-05,
|
| 1696 |
+
"loss": 24.9972,
|
| 1697 |
+
"step": 1200
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.5279226735669008,
|
| 1701 |
+
"grad_norm": 77.5625,
|
| 1702 |
+
"learning_rate": 1.049196298100341e-05,
|
| 1703 |
+
"loss": 24.5776,
|
| 1704 |
+
"step": 1205
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.5301132240796265,
|
| 1708 |
+
"grad_norm": 163.25,
|
| 1709 |
+
"learning_rate": 1.044325377496347e-05,
|
| 1710 |
+
"loss": 25.7603,
|
| 1711 |
+
"step": 1210
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.5323037745923522,
|
| 1715 |
+
"grad_norm": 77.4375,
|
| 1716 |
+
"learning_rate": 1.0394544568923529e-05,
|
| 1717 |
+
"loss": 24.9012,
|
| 1718 |
+
"step": 1215
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.534494325105078,
|
| 1722 |
+
"grad_norm": 101.25,
|
| 1723 |
+
"learning_rate": 1.0345835362883584e-05,
|
| 1724 |
+
"loss": 25.5309,
|
| 1725 |
+
"step": 1220
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.5366848756178036,
|
| 1729 |
+
"grad_norm": 127.625,
|
| 1730 |
+
"learning_rate": 1.0297126156843644e-05,
|
| 1731 |
+
"loss": 25.7696,
|
| 1732 |
+
"step": 1225
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.5388754261305294,
|
| 1736 |
+
"grad_norm": 79.75,
|
| 1737 |
+
"learning_rate": 1.0248416950803703e-05,
|
| 1738 |
+
"loss": 25.5109,
|
| 1739 |
+
"step": 1230
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.5410659766432552,
|
| 1743 |
+
"grad_norm": 151.0,
|
| 1744 |
+
"learning_rate": 1.019970774476376e-05,
|
| 1745 |
+
"loss": 25.7191,
|
| 1746 |
+
"step": 1235
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.5432565271559809,
|
| 1750 |
+
"grad_norm": 54.0,
|
| 1751 |
+
"learning_rate": 1.015099853872382e-05,
|
| 1752 |
+
"loss": 25.1186,
|
| 1753 |
+
"step": 1240
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.5454470776687066,
|
| 1757 |
+
"grad_norm": 54.3125,
|
| 1758 |
+
"learning_rate": 1.0102289332683879e-05,
|
| 1759 |
+
"loss": 24.919,
|
| 1760 |
+
"step": 1245
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.5476376281814324,
|
| 1764 |
+
"grad_norm": 106.8125,
|
| 1765 |
+
"learning_rate": 1.0053580126643936e-05,
|
| 1766 |
+
"loss": 25.1853,
|
| 1767 |
+
"step": 1250
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.5498281786941581,
|
| 1771 |
+
"grad_norm": 97.75,
|
| 1772 |
+
"learning_rate": 1.0004870920603996e-05,
|
| 1773 |
+
"loss": 24.8172,
|
| 1774 |
+
"step": 1255
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.5520187292068838,
|
| 1778 |
+
"grad_norm": 67.8125,
|
| 1779 |
+
"learning_rate": 9.956161714564053e-06,
|
| 1780 |
+
"loss": 24.8634,
|
| 1781 |
+
"step": 1260
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.5542092797196095,
|
| 1785 |
+
"grad_norm": 169.125,
|
| 1786 |
+
"learning_rate": 9.90745250852411e-06,
|
| 1787 |
+
"loss": 25.2829,
|
| 1788 |
+
"step": 1265
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.5563998302323353,
|
| 1792 |
+
"grad_norm": 116.375,
|
| 1793 |
+
"learning_rate": 9.85874330248417e-06,
|
| 1794 |
+
"loss": 25.702,
|
| 1795 |
+
"step": 1270
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.558590380745061,
|
| 1799 |
+
"grad_norm": 229.125,
|
| 1800 |
+
"learning_rate": 9.810034096444229e-06,
|
| 1801 |
+
"loss": 24.9604,
|
| 1802 |
+
"step": 1275
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 0.5607809312577867,
|
| 1806 |
+
"grad_norm": 51.5625,
|
| 1807 |
+
"learning_rate": 9.761324890404287e-06,
|
| 1808 |
+
"loss": 24.766,
|
| 1809 |
+
"step": 1280
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 0.5629714817705125,
|
| 1813 |
+
"grad_norm": 88.5625,
|
| 1814 |
+
"learning_rate": 9.712615684364346e-06,
|
| 1815 |
+
"loss": 24.9662,
|
| 1816 |
+
"step": 1285
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 0.5651620322832381,
|
| 1820 |
+
"grad_norm": 92.375,
|
| 1821 |
+
"learning_rate": 9.663906478324403e-06,
|
| 1822 |
+
"loss": 25.0194,
|
| 1823 |
+
"step": 1290
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 0.5673525827959639,
|
| 1827 |
+
"grad_norm": 83.8125,
|
| 1828 |
+
"learning_rate": 9.615197272284463e-06,
|
| 1829 |
+
"loss": 25.4854,
|
| 1830 |
+
"step": 1295
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 0.5695431333086897,
|
| 1834 |
+
"grad_norm": 76.75,
|
| 1835 |
+
"learning_rate": 9.566488066244522e-06,
|
| 1836 |
+
"loss": 24.8307,
|
| 1837 |
+
"step": 1300
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 0.5717336838214153,
|
| 1841 |
+
"grad_norm": 92.75,
|
| 1842 |
+
"learning_rate": 9.51777886020458e-06,
|
| 1843 |
+
"loss": 24.5626,
|
| 1844 |
+
"step": 1305
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 0.5739242343341411,
|
| 1848 |
+
"grad_norm": 98.6875,
|
| 1849 |
+
"learning_rate": 9.469069654164637e-06,
|
| 1850 |
+
"loss": 24.9844,
|
| 1851 |
+
"step": 1310
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 0.5761147848468668,
|
| 1855 |
+
"grad_norm": 63.28125,
|
| 1856 |
+
"learning_rate": 9.420360448124696e-06,
|
| 1857 |
+
"loss": 25.0084,
|
| 1858 |
+
"step": 1315
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 0.5783053353595926,
|
| 1862 |
+
"grad_norm": 272.75,
|
| 1863 |
+
"learning_rate": 9.371651242084755e-06,
|
| 1864 |
+
"loss": 25.2756,
|
| 1865 |
+
"step": 1320
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 0.5804958858723183,
|
| 1869 |
+
"grad_norm": 214.125,
|
| 1870 |
+
"learning_rate": 9.322942036044813e-06,
|
| 1871 |
+
"loss": 25.1908,
|
| 1872 |
+
"step": 1325
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 0.582686436385044,
|
| 1876 |
+
"grad_norm": 93.875,
|
| 1877 |
+
"learning_rate": 9.27423283000487e-06,
|
| 1878 |
+
"loss": 24.5392,
|
| 1879 |
+
"step": 1330
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 0.5848769868977698,
|
| 1883 |
+
"grad_norm": 119.5625,
|
| 1884 |
+
"learning_rate": 9.22552362396493e-06,
|
| 1885 |
+
"loss": 25.1336,
|
| 1886 |
+
"step": 1335
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 0.5870675374104954,
|
| 1890 |
+
"grad_norm": 104.4375,
|
| 1891 |
+
"learning_rate": 9.176814417924989e-06,
|
| 1892 |
+
"loss": 24.3678,
|
| 1893 |
+
"step": 1340
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 0.5892580879232212,
|
| 1897 |
+
"grad_norm": 77.0625,
|
| 1898 |
+
"learning_rate": 9.128105211885046e-06,
|
| 1899 |
+
"loss": 25.1012,
|
| 1900 |
+
"step": 1345
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 0.591448638435947,
|
| 1904 |
+
"grad_norm": 196.375,
|
| 1905 |
+
"learning_rate": 9.079396005845106e-06,
|
| 1906 |
+
"loss": 24.9593,
|
| 1907 |
+
"step": 1350
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 0.5936391889486726,
|
| 1911 |
+
"grad_norm": 71.4375,
|
| 1912 |
+
"learning_rate": 9.030686799805163e-06,
|
| 1913 |
+
"loss": 24.8451,
|
| 1914 |
+
"step": 1355
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 0.5958297394613984,
|
| 1918 |
+
"grad_norm": 389.5,
|
| 1919 |
+
"learning_rate": 8.981977593765222e-06,
|
| 1920 |
+
"loss": 25.182,
|
| 1921 |
+
"step": 1360
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 0.5980202899741242,
|
| 1925 |
+
"grad_norm": 56.03125,
|
| 1926 |
+
"learning_rate": 8.933268387725282e-06,
|
| 1927 |
+
"loss": 24.3735,
|
| 1928 |
+
"step": 1365
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 0.6002108404868498,
|
| 1932 |
+
"grad_norm": 55.53125,
|
| 1933 |
+
"learning_rate": 8.884559181685339e-06,
|
| 1934 |
+
"loss": 25.3762,
|
| 1935 |
+
"step": 1370
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 0.6024013909995756,
|
| 1939 |
+
"grad_norm": 115.5,
|
| 1940 |
+
"learning_rate": 8.835849975645398e-06,
|
| 1941 |
+
"loss": 23.874,
|
| 1942 |
+
"step": 1375
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 0.6045919415123013,
|
| 1946 |
+
"grad_norm": 76.0,
|
| 1947 |
+
"learning_rate": 8.787140769605456e-06,
|
| 1948 |
+
"loss": 25.0251,
|
| 1949 |
+
"step": 1380
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 0.606782492025027,
|
| 1953 |
+
"grad_norm": 152.625,
|
| 1954 |
+
"learning_rate": 8.738431563565515e-06,
|
| 1955 |
+
"loss": 25.099,
|
| 1956 |
+
"step": 1385
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 0.6089730425377528,
|
| 1960 |
+
"grad_norm": 81.625,
|
| 1961 |
+
"learning_rate": 8.689722357525573e-06,
|
| 1962 |
+
"loss": 24.6503,
|
| 1963 |
+
"step": 1390
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 0.6111635930504785,
|
| 1967 |
+
"grad_norm": 99.9375,
|
| 1968 |
+
"learning_rate": 8.641013151485632e-06,
|
| 1969 |
+
"loss": 25.0892,
|
| 1970 |
+
"step": 1395
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 0.6133541435632043,
|
| 1974 |
+
"grad_norm": 67.5625,
|
| 1975 |
+
"learning_rate": 8.592303945445691e-06,
|
| 1976 |
+
"loss": 25.1785,
|
| 1977 |
+
"step": 1400
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 0.6155446940759299,
|
| 1981 |
+
"grad_norm": 133.125,
|
| 1982 |
+
"learning_rate": 8.543594739405749e-06,
|
| 1983 |
+
"loss": 25.2308,
|
| 1984 |
+
"step": 1405
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 0.6177352445886557,
|
| 1988 |
+
"grad_norm": 127.1875,
|
| 1989 |
+
"learning_rate": 8.494885533365806e-06,
|
| 1990 |
+
"loss": 24.7525,
|
| 1991 |
+
"step": 1410
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 0.6199257951013815,
|
| 1995 |
+
"grad_norm": 88.4375,
|
| 1996 |
+
"learning_rate": 8.446176327325865e-06,
|
| 1997 |
+
"loss": 25.3764,
|
| 1998 |
+
"step": 1415
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 0.6221163456141071,
|
| 2002 |
+
"grad_norm": 68.1875,
|
| 2003 |
+
"learning_rate": 8.397467121285925e-06,
|
| 2004 |
+
"loss": 25.6512,
|
| 2005 |
+
"step": 1420
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 0.6243068961268329,
|
| 2009 |
+
"grad_norm": 105.25,
|
| 2010 |
+
"learning_rate": 8.348757915245982e-06,
|
| 2011 |
+
"loss": 24.5618,
|
| 2012 |
+
"step": 1425
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 0.6264974466395586,
|
| 2016 |
+
"grad_norm": 62.96875,
|
| 2017 |
+
"learning_rate": 8.300048709206041e-06,
|
| 2018 |
+
"loss": 25.5549,
|
| 2019 |
+
"step": 1430
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.6286879971522843,
|
| 2023 |
+
"grad_norm": 112.75,
|
| 2024 |
+
"learning_rate": 8.251339503166099e-06,
|
| 2025 |
+
"loss": 25.7603,
|
| 2026 |
+
"step": 1435
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 0.6308785476650101,
|
| 2030 |
+
"grad_norm": 146.75,
|
| 2031 |
+
"learning_rate": 8.202630297126158e-06,
|
| 2032 |
+
"loss": 24.9323,
|
| 2033 |
+
"step": 1440
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 0.6330690981777358,
|
| 2037 |
+
"grad_norm": 72.8125,
|
| 2038 |
+
"learning_rate": 8.153921091086216e-06,
|
| 2039 |
+
"loss": 25.1751,
|
| 2040 |
+
"step": 1445
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 0.6352596486904615,
|
| 2044 |
+
"grad_norm": 131.75,
|
| 2045 |
+
"learning_rate": 8.105211885046275e-06,
|
| 2046 |
+
"loss": 24.7063,
|
| 2047 |
+
"step": 1450
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 0.6374501992031872,
|
| 2051 |
+
"grad_norm": 81.5,
|
| 2052 |
+
"learning_rate": 8.056502679006332e-06,
|
| 2053 |
+
"loss": 25.0378,
|
| 2054 |
+
"step": 1455
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 0.639640749715913,
|
| 2058 |
+
"grad_norm": 74.0625,
|
| 2059 |
+
"learning_rate": 8.007793472966392e-06,
|
| 2060 |
+
"loss": 24.7719,
|
| 2061 |
+
"step": 1460
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 0.6418313002286387,
|
| 2065 |
+
"grad_norm": 65.0,
|
| 2066 |
+
"learning_rate": 7.95908426692645e-06,
|
| 2067 |
+
"loss": 24.439,
|
| 2068 |
+
"step": 1465
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 0.6440218507413644,
|
| 2072 |
+
"grad_norm": 66.8125,
|
| 2073 |
+
"learning_rate": 7.910375060886508e-06,
|
| 2074 |
+
"loss": 24.7476,
|
| 2075 |
+
"step": 1470
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 0.6462124012540902,
|
| 2079 |
+
"grad_norm": 38.3125,
|
| 2080 |
+
"learning_rate": 7.861665854846566e-06,
|
| 2081 |
+
"loss": 25.3374,
|
| 2082 |
+
"step": 1475
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 0.648402951766816,
|
| 2086 |
+
"grad_norm": 55.5625,
|
| 2087 |
+
"learning_rate": 7.812956648806625e-06,
|
| 2088 |
+
"loss": 24.3505,
|
| 2089 |
+
"step": 1480
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 0.6505935022795416,
|
| 2093 |
+
"grad_norm": 86.25,
|
| 2094 |
+
"learning_rate": 7.764247442766684e-06,
|
| 2095 |
+
"loss": 24.8686,
|
| 2096 |
+
"step": 1485
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.6527840527922674,
|
| 2100 |
+
"grad_norm": 78.4375,
|
| 2101 |
+
"learning_rate": 7.715538236726742e-06,
|
| 2102 |
+
"loss": 24.9231,
|
| 2103 |
+
"step": 1490
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 0.6549746033049931,
|
| 2107 |
+
"grad_norm": 110.9375,
|
| 2108 |
+
"learning_rate": 7.666829030686801e-06,
|
| 2109 |
+
"loss": 25.0029,
|
| 2110 |
+
"step": 1495
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 0.6571651538177188,
|
| 2114 |
+
"grad_norm": 43.84375,
|
| 2115 |
+
"learning_rate": 7.6181198246468595e-06,
|
| 2116 |
+
"loss": 24.4183,
|
| 2117 |
+
"step": 1500
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 0.6593557043304445,
|
| 2121 |
+
"grad_norm": 50.46875,
|
| 2122 |
+
"learning_rate": 7.569410618606917e-06,
|
| 2123 |
+
"loss": 24.3607,
|
| 2124 |
+
"step": 1505
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 0.6615462548431703,
|
| 2128 |
+
"grad_norm": 220.125,
|
| 2129 |
+
"learning_rate": 7.520701412566975e-06,
|
| 2130 |
+
"loss": 24.7162,
|
| 2131 |
+
"step": 1510
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 0.663736805355896,
|
| 2135 |
+
"grad_norm": 89.625,
|
| 2136 |
+
"learning_rate": 7.471992206527035e-06,
|
| 2137 |
+
"loss": 24.8443,
|
| 2138 |
+
"step": 1515
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 0.6659273558686217,
|
| 2142 |
+
"grad_norm": 45.96875,
|
| 2143 |
+
"learning_rate": 7.423283000487093e-06,
|
| 2144 |
+
"loss": 24.0884,
|
| 2145 |
+
"step": 1520
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 0.6681179063813475,
|
| 2149 |
+
"grad_norm": 41.125,
|
| 2150 |
+
"learning_rate": 7.3745737944471505e-06,
|
| 2151 |
+
"loss": 24.1827,
|
| 2152 |
+
"step": 1525
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 0.6703084568940731,
|
| 2156 |
+
"grad_norm": 46.0625,
|
| 2157 |
+
"learning_rate": 7.32586458840721e-06,
|
| 2158 |
+
"loss": 24.6457,
|
| 2159 |
+
"step": 1530
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 0.6724990074067989,
|
| 2163 |
+
"grad_norm": 49.96875,
|
| 2164 |
+
"learning_rate": 7.277155382367268e-06,
|
| 2165 |
+
"loss": 24.2197,
|
| 2166 |
+
"step": 1535
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 0.6746895579195247,
|
| 2170 |
+
"grad_norm": 55.4375,
|
| 2171 |
+
"learning_rate": 7.2284461763273265e-06,
|
| 2172 |
+
"loss": 24.5083,
|
| 2173 |
+
"step": 1540
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 0.6768801084322503,
|
| 2177 |
+
"grad_norm": 147.375,
|
| 2178 |
+
"learning_rate": 7.179736970287386e-06,
|
| 2179 |
+
"loss": 24.5295,
|
| 2180 |
+
"step": 1545
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 0.6790706589449761,
|
| 2184 |
+
"grad_norm": 66.8125,
|
| 2185 |
+
"learning_rate": 7.131027764247443e-06,
|
| 2186 |
+
"loss": 24.3484,
|
| 2187 |
+
"step": 1550
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 0.6812612094577019,
|
| 2191 |
+
"grad_norm": 99.125,
|
| 2192 |
+
"learning_rate": 7.082318558207502e-06,
|
| 2193 |
+
"loss": 24.5569,
|
| 2194 |
+
"step": 1555
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 0.6834517599704276,
|
| 2198 |
+
"grad_norm": 55.34375,
|
| 2199 |
+
"learning_rate": 7.03360935216756e-06,
|
| 2200 |
+
"loss": 24.0178,
|
| 2201 |
+
"step": 1560
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.6856423104831533,
|
| 2205 |
+
"grad_norm": 89.8125,
|
| 2206 |
+
"learning_rate": 6.984900146127619e-06,
|
| 2207 |
+
"loss": 23.5264,
|
| 2208 |
+
"step": 1565
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.687832860995879,
|
| 2212 |
+
"grad_norm": 248.5,
|
| 2213 |
+
"learning_rate": 6.936190940087677e-06,
|
| 2214 |
+
"loss": 25.177,
|
| 2215 |
+
"step": 1570
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.6900234115086048,
|
| 2219 |
+
"grad_norm": 97.75,
|
| 2220 |
+
"learning_rate": 6.887481734047735e-06,
|
| 2221 |
+
"loss": 24.3799,
|
| 2222 |
+
"step": 1575
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.6922139620213305,
|
| 2226 |
+
"grad_norm": 101.0,
|
| 2227 |
+
"learning_rate": 6.838772528007794e-06,
|
| 2228 |
+
"loss": 24.7024,
|
| 2229 |
+
"step": 1580
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.6944045125340562,
|
| 2233 |
+
"grad_norm": 94.1875,
|
| 2234 |
+
"learning_rate": 6.790063321967853e-06,
|
| 2235 |
+
"loss": 24.3573,
|
| 2236 |
+
"step": 1585
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 0.696595063046782,
|
| 2240 |
+
"grad_norm": 71.0625,
|
| 2241 |
+
"learning_rate": 6.74135411592791e-06,
|
| 2242 |
+
"loss": 24.3899,
|
| 2243 |
+
"step": 1590
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.6987856135595076,
|
| 2247 |
+
"grad_norm": 62.90625,
|
| 2248 |
+
"learning_rate": 6.6926449098879695e-06,
|
| 2249 |
+
"loss": 24.4977,
|
| 2250 |
+
"step": 1595
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 0.7009761640722334,
|
| 2254 |
+
"grad_norm": 42.25,
|
| 2255 |
+
"learning_rate": 6.643935703848028e-06,
|
| 2256 |
+
"loss": 24.0119,
|
| 2257 |
+
"step": 1600
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.7031667145849592,
|
| 2261 |
+
"grad_norm": 72.75,
|
| 2262 |
+
"learning_rate": 6.595226497808086e-06,
|
| 2263 |
+
"loss": 24.1055,
|
| 2264 |
+
"step": 1605
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 0.7053572650976848,
|
| 2268 |
+
"grad_norm": 50.1875,
|
| 2269 |
+
"learning_rate": 6.5465172917681454e-06,
|
| 2270 |
+
"loss": 24.6823,
|
| 2271 |
+
"step": 1610
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.7075478156104106,
|
| 2275 |
+
"grad_norm": 76.75,
|
| 2276 |
+
"learning_rate": 6.497808085728203e-06,
|
| 2277 |
+
"loss": 24.3863,
|
| 2278 |
+
"step": 1615
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 0.7097383661231363,
|
| 2282 |
+
"grad_norm": 41.8125,
|
| 2283 |
+
"learning_rate": 6.449098879688261e-06,
|
| 2284 |
+
"loss": 24.5434,
|
| 2285 |
+
"step": 1620
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 0.711928916635862,
|
| 2289 |
+
"grad_norm": 49.71875,
|
| 2290 |
+
"learning_rate": 6.40038967364832e-06,
|
| 2291 |
+
"loss": 23.9173,
|
| 2292 |
+
"step": 1625
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 0.7141194671485878,
|
| 2296 |
+
"grad_norm": 303.75,
|
| 2297 |
+
"learning_rate": 6.351680467608379e-06,
|
| 2298 |
+
"loss": 23.8009,
|
| 2299 |
+
"step": 1630
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.7163100176613135,
|
| 2303 |
+
"grad_norm": 107.0,
|
| 2304 |
+
"learning_rate": 6.3029712615684365e-06,
|
| 2305 |
+
"loss": 24.348,
|
| 2306 |
+
"step": 1635
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.7185005681740393,
|
| 2310 |
+
"grad_norm": 57.5625,
|
| 2311 |
+
"learning_rate": 6.254262055528495e-06,
|
| 2312 |
+
"loss": 24.7125,
|
| 2313 |
+
"step": 1640
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.7206911186867649,
|
| 2317 |
+
"grad_norm": 83.125,
|
| 2318 |
+
"learning_rate": 6.205552849488554e-06,
|
| 2319 |
+
"loss": 24.2265,
|
| 2320 |
+
"step": 1645
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 0.7228816691994907,
|
| 2324 |
+
"grad_norm": 42.28125,
|
| 2325 |
+
"learning_rate": 6.1568436434486125e-06,
|
| 2326 |
+
"loss": 24.2224,
|
| 2327 |
+
"step": 1650
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 0.7250722197122165,
|
| 2331 |
+
"grad_norm": 64.125,
|
| 2332 |
+
"learning_rate": 6.10813443740867e-06,
|
| 2333 |
+
"loss": 24.8299,
|
| 2334 |
+
"step": 1655
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 0.7272627702249421,
|
| 2338 |
+
"grad_norm": 105.125,
|
| 2339 |
+
"learning_rate": 6.059425231368729e-06,
|
| 2340 |
+
"loss": 23.6778,
|
| 2341 |
+
"step": 1660
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.7294533207376679,
|
| 2345 |
+
"grad_norm": 113.6875,
|
| 2346 |
+
"learning_rate": 6.010716025328788e-06,
|
| 2347 |
+
"loss": 24.2767,
|
| 2348 |
+
"step": 1665
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 0.7316438712503937,
|
| 2352 |
+
"grad_norm": 67.9375,
|
| 2353 |
+
"learning_rate": 5.962006819288846e-06,
|
| 2354 |
+
"loss": 23.9942,
|
| 2355 |
+
"step": 1670
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 0.7338344217631193,
|
| 2359 |
+
"grad_norm": 54.03125,
|
| 2360 |
+
"learning_rate": 5.913297613248905e-06,
|
| 2361 |
+
"loss": 23.7793,
|
| 2362 |
+
"step": 1675
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 0.7360249722758451,
|
| 2366 |
+
"grad_norm": 63.625,
|
| 2367 |
+
"learning_rate": 5.8645884072089636e-06,
|
| 2368 |
+
"loss": 23.582,
|
| 2369 |
+
"step": 1680
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 0.7382155227885708,
|
| 2373 |
+
"grad_norm": 58.375,
|
| 2374 |
+
"learning_rate": 5.815879201169021e-06,
|
| 2375 |
+
"loss": 24.222,
|
| 2376 |
+
"step": 1685
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 0.7404060733012965,
|
| 2380 |
+
"grad_norm": 68.8125,
|
| 2381 |
+
"learning_rate": 5.7671699951290795e-06,
|
| 2382 |
+
"loss": 23.639,
|
| 2383 |
+
"step": 1690
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 0.7425966238140222,
|
| 2387 |
+
"grad_norm": 32.53125,
|
| 2388 |
+
"learning_rate": 5.718460789089139e-06,
|
| 2389 |
+
"loss": 23.6367,
|
| 2390 |
+
"step": 1695
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 0.744787174326748,
|
| 2394 |
+
"grad_norm": 62.53125,
|
| 2395 |
+
"learning_rate": 5.669751583049197e-06,
|
| 2396 |
+
"loss": 24.6188,
|
| 2397 |
+
"step": 1700
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 0.7469777248394737,
|
| 2401 |
+
"grad_norm": 108.75,
|
| 2402 |
+
"learning_rate": 5.621042377009255e-06,
|
| 2403 |
+
"loss": 24.1849,
|
| 2404 |
+
"step": 1705
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 0.7491682753521994,
|
| 2408 |
+
"grad_norm": 126.0,
|
| 2409 |
+
"learning_rate": 5.572333170969314e-06,
|
| 2410 |
+
"loss": 24.3651,
|
| 2411 |
+
"step": 1710
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 0.7513588258649252,
|
| 2415 |
+
"grad_norm": 50.84375,
|
| 2416 |
+
"learning_rate": 5.523623964929372e-06,
|
| 2417 |
+
"loss": 24.2613,
|
| 2418 |
+
"step": 1715
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 0.753549376377651,
|
| 2422 |
+
"grad_norm": 74.5625,
|
| 2423 |
+
"learning_rate": 5.474914758889431e-06,
|
| 2424 |
+
"loss": 23.998,
|
| 2425 |
+
"step": 1720
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 0.7557399268903766,
|
| 2429 |
+
"grad_norm": 86.125,
|
| 2430 |
+
"learning_rate": 5.42620555284949e-06,
|
| 2431 |
+
"loss": 23.9456,
|
| 2432 |
+
"step": 1725
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 0.7579304774031024,
|
| 2436 |
+
"grad_norm": 61.09375,
|
| 2437 |
+
"learning_rate": 5.377496346809547e-06,
|
| 2438 |
+
"loss": 23.3448,
|
| 2439 |
+
"step": 1730
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 0.7601210279158281,
|
| 2443 |
+
"grad_norm": 65.125,
|
| 2444 |
+
"learning_rate": 5.328787140769606e-06,
|
| 2445 |
+
"loss": 24.1856,
|
| 2446 |
+
"step": 1735
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 0.7623115784285538,
|
| 2450 |
+
"grad_norm": 54.375,
|
| 2451 |
+
"learning_rate": 5.280077934729664e-06,
|
| 2452 |
+
"loss": 23.7022,
|
| 2453 |
+
"step": 1740
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 0.7645021289412796,
|
| 2457 |
+
"grad_norm": 73.8125,
|
| 2458 |
+
"learning_rate": 5.231368728689723e-06,
|
| 2459 |
+
"loss": 24.0991,
|
| 2460 |
+
"step": 1745
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 0.7666926794540053,
|
| 2464 |
+
"grad_norm": 41.6875,
|
| 2465 |
+
"learning_rate": 5.182659522649781e-06,
|
| 2466 |
+
"loss": 23.627,
|
| 2467 |
+
"step": 1750
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 0.768883229966731,
|
| 2471 |
+
"grad_norm": 95.5625,
|
| 2472 |
+
"learning_rate": 5.133950316609839e-06,
|
| 2473 |
+
"loss": 24.0103,
|
| 2474 |
+
"step": 1755
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 0.7710737804794567,
|
| 2478 |
+
"grad_norm": 58.09375,
|
| 2479 |
+
"learning_rate": 5.0852411105698985e-06,
|
| 2480 |
+
"loss": 24.4259,
|
| 2481 |
+
"step": 1760
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 0.7732643309921825,
|
| 2485 |
+
"grad_norm": 51.1875,
|
| 2486 |
+
"learning_rate": 5.036531904529957e-06,
|
| 2487 |
+
"loss": 24.3509,
|
| 2488 |
+
"step": 1765
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 0.7754548815049082,
|
| 2492 |
+
"grad_norm": 26.265625,
|
| 2493 |
+
"learning_rate": 4.987822698490015e-06,
|
| 2494 |
+
"loss": 24.2639,
|
| 2495 |
+
"step": 1770
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 0.7776454320176339,
|
| 2499 |
+
"grad_norm": 37.03125,
|
| 2500 |
+
"learning_rate": 4.939113492450074e-06,
|
| 2501 |
+
"loss": 23.9353,
|
| 2502 |
+
"step": 1775
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 0.7798359825303597,
|
| 2506 |
+
"grad_norm": 71.6875,
|
| 2507 |
+
"learning_rate": 4.890404286410132e-06,
|
| 2508 |
+
"loss": 23.7745,
|
| 2509 |
+
"step": 1780
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 0.7820265330430853,
|
| 2513 |
+
"grad_norm": 53.875,
|
| 2514 |
+
"learning_rate": 4.84169508037019e-06,
|
| 2515 |
+
"loss": 24.0737,
|
| 2516 |
+
"step": 1785
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 0.7842170835558111,
|
| 2520 |
+
"grad_norm": 62.15625,
|
| 2521 |
+
"learning_rate": 4.792985874330249e-06,
|
| 2522 |
+
"loss": 24.3623,
|
| 2523 |
+
"step": 1790
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 0.7864076340685369,
|
| 2527 |
+
"grad_norm": 46.3125,
|
| 2528 |
+
"learning_rate": 4.744276668290307e-06,
|
| 2529 |
+
"loss": 24.1168,
|
| 2530 |
+
"step": 1795
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 0.7885981845812626,
|
| 2534 |
+
"grad_norm": 49.90625,
|
| 2535 |
+
"learning_rate": 4.6955674622503655e-06,
|
| 2536 |
+
"loss": 23.9208,
|
| 2537 |
+
"step": 1800
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 0.7907887350939883,
|
| 2541 |
+
"grad_norm": 70.0625,
|
| 2542 |
+
"learning_rate": 4.646858256210424e-06,
|
| 2543 |
+
"loss": 23.9657,
|
| 2544 |
+
"step": 1805
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 0.792979285606714,
|
| 2548 |
+
"grad_norm": 273.5,
|
| 2549 |
+
"learning_rate": 4.598149050170483e-06,
|
| 2550 |
+
"loss": 23.4883,
|
| 2551 |
+
"step": 1810
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 0.7951698361194398,
|
| 2555 |
+
"grad_norm": 36.46875,
|
| 2556 |
+
"learning_rate": 4.549439844130541e-06,
|
| 2557 |
+
"loss": 24.1732,
|
| 2558 |
+
"step": 1815
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 0.7973603866321655,
|
| 2562 |
+
"grad_norm": 98.1875,
|
| 2563 |
+
"learning_rate": 4.5007306380906e-06,
|
| 2564 |
+
"loss": 24.5668,
|
| 2565 |
+
"step": 1820
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 0.7995509371448912,
|
| 2569 |
+
"grad_norm": 79.9375,
|
| 2570 |
+
"learning_rate": 4.452021432050657e-06,
|
| 2571 |
+
"loss": 23.3709,
|
| 2572 |
+
"step": 1825
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 0.801741487657617,
|
| 2576 |
+
"grad_norm": 51.5,
|
| 2577 |
+
"learning_rate": 4.403312226010717e-06,
|
| 2578 |
+
"loss": 23.4818,
|
| 2579 |
+
"step": 1830
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 0.8039320381703426,
|
| 2583 |
+
"grad_norm": 72.25,
|
| 2584 |
+
"learning_rate": 4.354603019970775e-06,
|
| 2585 |
+
"loss": 23.678,
|
| 2586 |
+
"step": 1835
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 0.8061225886830684,
|
| 2590 |
+
"grad_norm": 56.21875,
|
| 2591 |
+
"learning_rate": 4.305893813930833e-06,
|
| 2592 |
+
"loss": 23.3332,
|
| 2593 |
+
"step": 1840
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 0.8083131391957942,
|
| 2597 |
+
"grad_norm": 179.25,
|
| 2598 |
+
"learning_rate": 4.257184607890892e-06,
|
| 2599 |
+
"loss": 23.4972,
|
| 2600 |
+
"step": 1845
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 0.8105036897085198,
|
| 2604 |
+
"grad_norm": 105.875,
|
| 2605 |
+
"learning_rate": 4.20847540185095e-06,
|
| 2606 |
+
"loss": 23.0243,
|
| 2607 |
+
"step": 1850
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 0.8126942402212456,
|
| 2611 |
+
"grad_norm": 49.53125,
|
| 2612 |
+
"learning_rate": 4.1597661958110085e-06,
|
| 2613 |
+
"loss": 23.926,
|
| 2614 |
+
"step": 1855
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 0.8148847907339714,
|
| 2618 |
+
"grad_norm": 42.1875,
|
| 2619 |
+
"learning_rate": 4.111056989771067e-06,
|
| 2620 |
+
"loss": 24.0595,
|
| 2621 |
+
"step": 1860
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 0.817075341246697,
|
| 2625 |
+
"grad_norm": 38.0625,
|
| 2626 |
+
"learning_rate": 4.062347783731125e-06,
|
| 2627 |
+
"loss": 23.3749,
|
| 2628 |
+
"step": 1865
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 0.8192658917594228,
|
| 2632 |
+
"grad_norm": 46.21875,
|
| 2633 |
+
"learning_rate": 4.0136385776911845e-06,
|
| 2634 |
+
"loss": 23.5141,
|
| 2635 |
+
"step": 1870
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 0.8214564422721485,
|
| 2639 |
+
"grad_norm": 64.6875,
|
| 2640 |
+
"learning_rate": 3.964929371651242e-06,
|
| 2641 |
+
"loss": 22.9993,
|
| 2642 |
+
"step": 1875
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 0.8236469927848743,
|
| 2646 |
+
"grad_norm": 104.1875,
|
| 2647 |
+
"learning_rate": 3.916220165611301e-06,
|
| 2648 |
+
"loss": 23.1683,
|
| 2649 |
+
"step": 1880
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 0.8258375432976,
|
| 2653 |
+
"grad_norm": 58.84375,
|
| 2654 |
+
"learning_rate": 3.86751095957136e-06,
|
| 2655 |
+
"loss": 23.006,
|
| 2656 |
+
"step": 1885
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 0.8280280938103257,
|
| 2660 |
+
"grad_norm": 47.625,
|
| 2661 |
+
"learning_rate": 3.818801753531418e-06,
|
| 2662 |
+
"loss": 23.3635,
|
| 2663 |
+
"step": 1890
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 0.8302186443230515,
|
| 2667 |
+
"grad_norm": 37.03125,
|
| 2668 |
+
"learning_rate": 3.7700925474914763e-06,
|
| 2669 |
+
"loss": 23.9395,
|
| 2670 |
+
"step": 1895
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 0.8324091948357771,
|
| 2674 |
+
"grad_norm": 40.0625,
|
| 2675 |
+
"learning_rate": 3.7213833414515347e-06,
|
| 2676 |
+
"loss": 23.6212,
|
| 2677 |
+
"step": 1900
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 0.8345997453485029,
|
| 2681 |
+
"grad_norm": 144.875,
|
| 2682 |
+
"learning_rate": 3.672674135411593e-06,
|
| 2683 |
+
"loss": 23.753,
|
| 2684 |
+
"step": 1905
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 0.8367902958612287,
|
| 2688 |
+
"grad_norm": 34.53125,
|
| 2689 |
+
"learning_rate": 3.6239649293716515e-06,
|
| 2690 |
+
"loss": 22.3497,
|
| 2691 |
+
"step": 1910
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 0.8389808463739543,
|
| 2695 |
+
"grad_norm": 44.71875,
|
| 2696 |
+
"learning_rate": 3.57525572333171e-06,
|
| 2697 |
+
"loss": 22.9936,
|
| 2698 |
+
"step": 1915
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 0.8411713968866801,
|
| 2702 |
+
"grad_norm": 42.5625,
|
| 2703 |
+
"learning_rate": 3.5265465172917687e-06,
|
| 2704 |
+
"loss": 23.6379,
|
| 2705 |
+
"step": 1920
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 0.8433619473994058,
|
| 2709 |
+
"grad_norm": 77.8125,
|
| 2710 |
+
"learning_rate": 3.477837311251827e-06,
|
| 2711 |
+
"loss": 23.9546,
|
| 2712 |
+
"step": 1925
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 0.8455524979121315,
|
| 2716 |
+
"grad_norm": 149.375,
|
| 2717 |
+
"learning_rate": 3.4291281052118854e-06,
|
| 2718 |
+
"loss": 23.1372,
|
| 2719 |
+
"step": 1930
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 0.8477430484248573,
|
| 2723 |
+
"grad_norm": 105.0625,
|
| 2724 |
+
"learning_rate": 3.3804188991719438e-06,
|
| 2725 |
+
"loss": 23.1796,
|
| 2726 |
+
"step": 1935
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 0.849933598937583,
|
| 2730 |
+
"grad_norm": 27.484375,
|
| 2731 |
+
"learning_rate": 3.331709693132002e-06,
|
| 2732 |
+
"loss": 23.3445,
|
| 2733 |
+
"step": 1940
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 0.8521241494503087,
|
| 2737 |
+
"grad_norm": 50.78125,
|
| 2738 |
+
"learning_rate": 3.2830004870920605e-06,
|
| 2739 |
+
"loss": 23.4366,
|
| 2740 |
+
"step": 1945
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 0.8543146999630344,
|
| 2744 |
+
"grad_norm": 33.125,
|
| 2745 |
+
"learning_rate": 3.2342912810521193e-06,
|
| 2746 |
+
"loss": 23.522,
|
| 2747 |
+
"step": 1950
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 0.8565052504757602,
|
| 2751 |
+
"grad_norm": 36.71875,
|
| 2752 |
+
"learning_rate": 3.1855820750121773e-06,
|
| 2753 |
+
"loss": 23.6913,
|
| 2754 |
+
"step": 1955
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.858695800988486,
|
| 2758 |
+
"grad_norm": 48.59375,
|
| 2759 |
+
"learning_rate": 3.136872868972236e-06,
|
| 2760 |
+
"loss": 23.2535,
|
| 2761 |
+
"step": 1960
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.8608863515012116,
|
| 2765 |
+
"grad_norm": 90.1875,
|
| 2766 |
+
"learning_rate": 3.088163662932294e-06,
|
| 2767 |
+
"loss": 23.2659,
|
| 2768 |
+
"step": 1965
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 0.8630769020139374,
|
| 2772 |
+
"grad_norm": 49.5625,
|
| 2773 |
+
"learning_rate": 3.039454456892353e-06,
|
| 2774 |
+
"loss": 22.8549,
|
| 2775 |
+
"step": 1970
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 0.8652674525266632,
|
| 2779 |
+
"grad_norm": 134.625,
|
| 2780 |
+
"learning_rate": 2.9907452508524117e-06,
|
| 2781 |
+
"loss": 23.9329,
|
| 2782 |
+
"step": 1975
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 0.8674580030393888,
|
| 2786 |
+
"grad_norm": 43.875,
|
| 2787 |
+
"learning_rate": 2.9420360448124696e-06,
|
| 2788 |
+
"loss": 22.7826,
|
| 2789 |
+
"step": 1980
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 0.8696485535521146,
|
| 2793 |
+
"grad_norm": 31.65625,
|
| 2794 |
+
"learning_rate": 2.8933268387725284e-06,
|
| 2795 |
+
"loss": 23.0787,
|
| 2796 |
+
"step": 1985
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.8718391040648403,
|
| 2800 |
+
"grad_norm": 56.9375,
|
| 2801 |
+
"learning_rate": 2.8446176327325868e-06,
|
| 2802 |
+
"loss": 23.0323,
|
| 2803 |
+
"step": 1990
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 0.874029654577566,
|
| 2807 |
+
"grad_norm": 39.78125,
|
| 2808 |
+
"learning_rate": 2.795908426692645e-06,
|
| 2809 |
+
"loss": 23.3068,
|
| 2810 |
+
"step": 1995
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 0.8762202050902917,
|
| 2814 |
+
"grad_norm": 48.09375,
|
| 2815 |
+
"learning_rate": 2.7471992206527035e-06,
|
| 2816 |
+
"loss": 23.4063,
|
| 2817 |
+
"step": 2000
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 0.8762202050902917,
|
| 2821 |
+
"eval_loss": NaN,
|
| 2822 |
+
"eval_runtime": 243.6077,
|
| 2823 |
+
"eval_samples_per_second": 1009.956,
|
| 2824 |
+
"eval_steps_per_second": 31.563,
|
| 2825 |
+
"step": 2000
|
| 2826 |
}
|
| 2827 |
],
|
| 2828 |
"logging_steps": 5,
|
|
|
|
| 2842 |
"attributes": {}
|
| 2843 |
}
|
| 2844 |
},
|
| 2845 |
+
"total_flos": 1.732943025316823e+19,
|
| 2846 |
"train_batch_size": 8,
|
| 2847 |
"trial_name": null,
|
| 2848 |
"trial_params": null
|