5000 step
Browse files- model-00001-of-00002.safetensors +1 -1
- model-00002-of-00002.safetensors +1 -1
- optimizer.pt +3 -0
- rng_state.pth +1 -1
- scheduler.pt +3 -0
- trainer_state.json +2103 -3
model-00001-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4999367032
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f87528be2ea496f5f8aac75903db600fc23c39cddccca19012c4b0394cde70d2
|
| 3 |
size 4999367032
|
model-00002-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2586705312
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d1dc1629c5a55632cfcda67f0f2fd9087c659ab137d1d788822a10c260c74d1
|
| 3 |
size 2586705312
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61850a34b537cd1759afafae41a04802926682e757ec560c8699796b8a612445
|
| 3 |
+
size 8550720062
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:500fba54bae9e962bc1226f244528dc6f5c3463aab3afacd9f43c2c82d41aac3
|
| 3 |
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50875594b892047c8e46a61b82a18becc1a9e9c9af6ce75ec5a292af0d5a8cd2
|
| 3 |
+
size 1064
|
trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1408,6 +1408,2106 @@
|
|
| 1408 |
"learning_rate": 7.011505090866913e-05,
|
| 1409 |
"loss": 0.0308,
|
| 1410 |
"step": 2000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1411 |
}
|
| 1412 |
],
|
| 1413 |
"logging_steps": 10,
|
|
@@ -1422,7 +3522,7 @@
|
|
| 1422 |
"should_evaluate": false,
|
| 1423 |
"should_log": false,
|
| 1424 |
"should_save": true,
|
| 1425 |
-
"should_training_stop":
|
| 1426 |
},
|
| 1427 |
"attributes": {}
|
| 1428 |
}
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 7.194244604316546,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 5000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1408 |
"learning_rate": 7.011505090866913e-05,
|
| 1409 |
"loss": 0.0308,
|
| 1410 |
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 2.8920863309352516,
|
| 1414 |
+
"grad_norm": 0.3075130581855774,
|
| 1415 |
+
"learning_rate": 6.981186037991271e-05,
|
| 1416 |
+
"loss": 0.0294,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 2.906474820143885,
|
| 1421 |
+
"grad_norm": 0.3712781071662903,
|
| 1422 |
+
"learning_rate": 6.950780321620174e-05,
|
| 1423 |
+
"loss": 0.0333,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 2.920863309352518,
|
| 1428 |
+
"grad_norm": 0.664781391620636,
|
| 1429 |
+
"learning_rate": 6.920289271798157e-05,
|
| 1430 |
+
"loss": 0.0468,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 2.935251798561151,
|
| 1435 |
+
"grad_norm": 0.48870983719825745,
|
| 1436 |
+
"learning_rate": 6.889714222302517e-05,
|
| 1437 |
+
"loss": 0.0353,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 2.949640287769784,
|
| 1442 |
+
"grad_norm": 0.5128960013389587,
|
| 1443 |
+
"learning_rate": 6.85905651058497e-05,
|
| 1444 |
+
"loss": 0.0348,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 2.9640287769784175,
|
| 1449 |
+
"grad_norm": 0.3929479122161865,
|
| 1450 |
+
"learning_rate": 6.82831747771314e-05,
|
| 1451 |
+
"loss": 0.0255,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 2.9784172661870505,
|
| 1456 |
+
"grad_norm": 0.47159093618392944,
|
| 1457 |
+
"learning_rate": 6.797498468311907e-05,
|
| 1458 |
+
"loss": 0.031,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 2.9928057553956835,
|
| 1463 |
+
"grad_norm": 0.34051942825317383,
|
| 1464 |
+
"learning_rate": 6.766600830504585e-05,
|
| 1465 |
+
"loss": 0.03,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 3.0071942446043165,
|
| 1470 |
+
"grad_norm": 0.440397709608078,
|
| 1471 |
+
"learning_rate": 6.735625915853942e-05,
|
| 1472 |
+
"loss": 0.0264,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 3.0215827338129495,
|
| 1477 |
+
"grad_norm": 0.43398475646972656,
|
| 1478 |
+
"learning_rate": 6.70457507930309e-05,
|
| 1479 |
+
"loss": 0.0298,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 3.0359712230215825,
|
| 1484 |
+
"grad_norm": 0.47332024574279785,
|
| 1485 |
+
"learning_rate": 6.673449679116215e-05,
|
| 1486 |
+
"loss": 0.0273,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 3.050359712230216,
|
| 1491 |
+
"grad_norm": 0.47281894087791443,
|
| 1492 |
+
"learning_rate": 6.642251076819148e-05,
|
| 1493 |
+
"loss": 0.0291,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 3.064748201438849,
|
| 1498 |
+
"grad_norm": 0.4297357499599457,
|
| 1499 |
+
"learning_rate": 6.610980637139827e-05,
|
| 1500 |
+
"loss": 0.0356,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 3.079136690647482,
|
| 1505 |
+
"grad_norm": 0.3897243142127991,
|
| 1506 |
+
"learning_rate": 6.579639727948583e-05,
|
| 1507 |
+
"loss": 0.0299,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 3.093525179856115,
|
| 1512 |
+
"grad_norm": 0.4756819009780884,
|
| 1513 |
+
"learning_rate": 6.548229720198315e-05,
|
| 1514 |
+
"loss": 0.0372,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 3.1079136690647484,
|
| 1519 |
+
"grad_norm": 0.2841879427433014,
|
| 1520 |
+
"learning_rate": 6.516751987864517e-05,
|
| 1521 |
+
"loss": 0.0264,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 3.1223021582733814,
|
| 1526 |
+
"grad_norm": 0.3094269037246704,
|
| 1527 |
+
"learning_rate": 6.485207907885175e-05,
|
| 1528 |
+
"loss": 0.0284,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 3.1366906474820144,
|
| 1533 |
+
"grad_norm": 0.30119588971138,
|
| 1534 |
+
"learning_rate": 6.453598860100536e-05,
|
| 1535 |
+
"loss": 0.0312,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 3.1510791366906474,
|
| 1540 |
+
"grad_norm": 0.30222269892692566,
|
| 1541 |
+
"learning_rate": 6.421926227192749e-05,
|
| 1542 |
+
"loss": 0.0277,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 3.1654676258992804,
|
| 1547 |
+
"grad_norm": 0.26771458983421326,
|
| 1548 |
+
"learning_rate": 6.390191394625381e-05,
|
| 1549 |
+
"loss": 0.0301,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 3.1798561151079134,
|
| 1554 |
+
"grad_norm": 0.36821985244750977,
|
| 1555 |
+
"learning_rate": 6.358395750582817e-05,
|
| 1556 |
+
"loss": 0.0231,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 3.194244604316547,
|
| 1561 |
+
"grad_norm": 0.3953895568847656,
|
| 1562 |
+
"learning_rate": 6.326540685909532e-05,
|
| 1563 |
+
"loss": 0.0263,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 3.20863309352518,
|
| 1568 |
+
"grad_norm": 0.41217997670173645,
|
| 1569 |
+
"learning_rate": 6.294627594049249e-05,
|
| 1570 |
+
"loss": 0.0273,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 3.223021582733813,
|
| 1575 |
+
"grad_norm": 0.37473732233047485,
|
| 1576 |
+
"learning_rate": 6.262657870983989e-05,
|
| 1577 |
+
"loss": 0.0306,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 3.237410071942446,
|
| 1582 |
+
"grad_norm": 0.3851197361946106,
|
| 1583 |
+
"learning_rate": 6.230632915173009e-05,
|
| 1584 |
+
"loss": 0.0283,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 3.2517985611510793,
|
| 1589 |
+
"grad_norm": 0.33268433809280396,
|
| 1590 |
+
"learning_rate": 6.198554127491622e-05,
|
| 1591 |
+
"loss": 0.0338,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 3.2661870503597124,
|
| 1596 |
+
"grad_norm": 0.37359151244163513,
|
| 1597 |
+
"learning_rate": 6.166422911169923e-05,
|
| 1598 |
+
"loss": 0.0276,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 3.2805755395683454,
|
| 1603 |
+
"grad_norm": 0.4748283624649048,
|
| 1604 |
+
"learning_rate": 6.1342406717314e-05,
|
| 1605 |
+
"loss": 0.0333,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 3.2949640287769784,
|
| 1610 |
+
"grad_norm": 0.25305020809173584,
|
| 1611 |
+
"learning_rate": 6.102008816931466e-05,
|
| 1612 |
+
"loss": 0.0315,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 3.3093525179856114,
|
| 1617 |
+
"grad_norm": 0.3776066303253174,
|
| 1618 |
+
"learning_rate": 6.069728756695866e-05,
|
| 1619 |
+
"loss": 0.0243,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 3.3237410071942444,
|
| 1624 |
+
"grad_norm": 0.33795878291130066,
|
| 1625 |
+
"learning_rate": 6.037401903059008e-05,
|
| 1626 |
+
"loss": 0.0338,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 3.338129496402878,
|
| 1631 |
+
"grad_norm": 0.5109716057777405,
|
| 1632 |
+
"learning_rate": 6.005029670102195e-05,
|
| 1633 |
+
"loss": 0.0283,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 3.352517985611511,
|
| 1638 |
+
"grad_norm": 0.3659425675868988,
|
| 1639 |
+
"learning_rate": 5.972613473891766e-05,
|
| 1640 |
+
"loss": 0.0318,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 3.366906474820144,
|
| 1645 |
+
"grad_norm": 0.3820067048072815,
|
| 1646 |
+
"learning_rate": 5.940154732417158e-05,
|
| 1647 |
+
"loss": 0.0345,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 3.381294964028777,
|
| 1652 |
+
"grad_norm": 0.46796146035194397,
|
| 1653 |
+
"learning_rate": 5.907654865528876e-05,
|
| 1654 |
+
"loss": 0.0297,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 3.3956834532374103,
|
| 1659 |
+
"grad_norm": 0.28651493787765503,
|
| 1660 |
+
"learning_rate": 5.875115294876381e-05,
|
| 1661 |
+
"loss": 0.0338,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 3.4100719424460433,
|
| 1666 |
+
"grad_norm": 0.3362789452075958,
|
| 1667 |
+
"learning_rate": 5.842537443845908e-05,
|
| 1668 |
+
"loss": 0.0289,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 3.4244604316546763,
|
| 1673 |
+
"grad_norm": 0.2535592317581177,
|
| 1674 |
+
"learning_rate": 5.809922737498198e-05,
|
| 1675 |
+
"loss": 0.0264,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 3.4388489208633093,
|
| 1680 |
+
"grad_norm": 0.2152898758649826,
|
| 1681 |
+
"learning_rate": 5.777272602506165e-05,
|
| 1682 |
+
"loss": 0.0312,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 3.4532374100719423,
|
| 1687 |
+
"grad_norm": 0.31355392932891846,
|
| 1688 |
+
"learning_rate": 5.744588467092483e-05,
|
| 1689 |
+
"loss": 0.032,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 3.4676258992805753,
|
| 1694 |
+
"grad_norm": 0.18381516635417938,
|
| 1695 |
+
"learning_rate": 5.7118717609671194e-05,
|
| 1696 |
+
"loss": 0.0258,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 3.4820143884892087,
|
| 1701 |
+
"grad_norm": 0.5383989214897156,
|
| 1702 |
+
"learning_rate": 5.679123915264786e-05,
|
| 1703 |
+
"loss": 0.0252,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 3.4964028776978417,
|
| 1708 |
+
"grad_norm": 0.3286333680152893,
|
| 1709 |
+
"learning_rate": 5.646346362482342e-05,
|
| 1710 |
+
"loss": 0.029,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 3.5107913669064748,
|
| 1715 |
+
"grad_norm": 0.39701902866363525,
|
| 1716 |
+
"learning_rate": 5.613540536416132e-05,
|
| 1717 |
+
"loss": 0.0302,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 3.5251798561151078,
|
| 1722 |
+
"grad_norm": 0.2979109585285187,
|
| 1723 |
+
"learning_rate": 5.5807078720992645e-05,
|
| 1724 |
+
"loss": 0.0284,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 3.539568345323741,
|
| 1729 |
+
"grad_norm": 0.37951502203941345,
|
| 1730 |
+
"learning_rate": 5.547849805738836e-05,
|
| 1731 |
+
"loss": 0.0303,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 3.553956834532374,
|
| 1736 |
+
"grad_norm": 0.36232998967170715,
|
| 1737 |
+
"learning_rate": 5.514967774653118e-05,
|
| 1738 |
+
"loss": 0.0295,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 3.568345323741007,
|
| 1743 |
+
"grad_norm": 0.3650036156177521,
|
| 1744 |
+
"learning_rate": 5.482063217208674e-05,
|
| 1745 |
+
"loss": 0.0296,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 3.58273381294964,
|
| 1750 |
+
"grad_norm": 0.3585372269153595,
|
| 1751 |
+
"learning_rate": 5.449137572757439e-05,
|
| 1752 |
+
"loss": 0.0255,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 3.597122302158273,
|
| 1757 |
+
"grad_norm": 0.284637987613678,
|
| 1758 |
+
"learning_rate": 5.4161922815737696e-05,
|
| 1759 |
+
"loss": 0.0283,
|
| 1760 |
+
"step": 2500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 3.6115107913669062,
|
| 1764 |
+
"grad_norm": 0.33122116327285767,
|
| 1765 |
+
"learning_rate": 5.3832287847914276e-05,
|
| 1766 |
+
"loss": 0.0248,
|
| 1767 |
+
"step": 2510
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 3.6258992805755397,
|
| 1771 |
+
"grad_norm": 0.3594101667404175,
|
| 1772 |
+
"learning_rate": 5.35024852434055e-05,
|
| 1773 |
+
"loss": 0.0318,
|
| 1774 |
+
"step": 2520
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 3.6402877697841727,
|
| 1778 |
+
"grad_norm": 0.3645528554916382,
|
| 1779 |
+
"learning_rate": 5.317252942884567e-05,
|
| 1780 |
+
"loss": 0.0251,
|
| 1781 |
+
"step": 2530
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 3.6546762589928057,
|
| 1785 |
+
"grad_norm": 0.3929426074028015,
|
| 1786 |
+
"learning_rate": 5.284243483757109e-05,
|
| 1787 |
+
"loss": 0.0277,
|
| 1788 |
+
"step": 2540
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 3.6690647482014387,
|
| 1792 |
+
"grad_norm": 0.31914427876472473,
|
| 1793 |
+
"learning_rate": 5.2512215908988484e-05,
|
| 1794 |
+
"loss": 0.0268,
|
| 1795 |
+
"step": 2550
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 3.683453237410072,
|
| 1799 |
+
"grad_norm": 0.27050238847732544,
|
| 1800 |
+
"learning_rate": 5.218188708794357e-05,
|
| 1801 |
+
"loss": 0.0302,
|
| 1802 |
+
"step": 2560
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 3.697841726618705,
|
| 1806 |
+
"grad_norm": 0.3745627701282501,
|
| 1807 |
+
"learning_rate": 5.18514628240891e-05,
|
| 1808 |
+
"loss": 0.0292,
|
| 1809 |
+
"step": 2570
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 3.712230215827338,
|
| 1813 |
+
"grad_norm": 0.4727473855018616,
|
| 1814 |
+
"learning_rate": 5.1520957571252795e-05,
|
| 1815 |
+
"loss": 0.0303,
|
| 1816 |
+
"step": 2580
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 3.726618705035971,
|
| 1820 |
+
"grad_norm": 0.42518627643585205,
|
| 1821 |
+
"learning_rate": 5.1190385786805106e-05,
|
| 1822 |
+
"loss": 0.0341,
|
| 1823 |
+
"step": 2590
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 3.741007194244604,
|
| 1827 |
+
"grad_norm": 0.37858137488365173,
|
| 1828 |
+
"learning_rate": 5.085976193102677e-05,
|
| 1829 |
+
"loss": 0.0239,
|
| 1830 |
+
"step": 2600
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 3.755395683453237,
|
| 1834 |
+
"grad_norm": 0.43805086612701416,
|
| 1835 |
+
"learning_rate": 5.052910046647634e-05,
|
| 1836 |
+
"loss": 0.0274,
|
| 1837 |
+
"step": 2610
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 3.7697841726618706,
|
| 1841 |
+
"grad_norm": 0.3793890178203583,
|
| 1842 |
+
"learning_rate": 5.0198415857357464e-05,
|
| 1843 |
+
"loss": 0.03,
|
| 1844 |
+
"step": 2620
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 3.7841726618705036,
|
| 1848 |
+
"grad_norm": 0.25346389412879944,
|
| 1849 |
+
"learning_rate": 4.9867722568886223e-05,
|
| 1850 |
+
"loss": 0.0217,
|
| 1851 |
+
"step": 2630
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 3.7985611510791366,
|
| 1855 |
+
"grad_norm": 0.3088151812553406,
|
| 1856 |
+
"learning_rate": 4.9537035066658314e-05,
|
| 1857 |
+
"loss": 0.0229,
|
| 1858 |
+
"step": 2640
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 3.81294964028777,
|
| 1862 |
+
"grad_norm": 0.4379916191101074,
|
| 1863 |
+
"learning_rate": 4.920636781601638e-05,
|
| 1864 |
+
"loss": 0.025,
|
| 1865 |
+
"step": 2650
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 3.827338129496403,
|
| 1869 |
+
"grad_norm": 0.26715871691703796,
|
| 1870 |
+
"learning_rate": 4.88757352814172e-05,
|
| 1871 |
+
"loss": 0.0363,
|
| 1872 |
+
"step": 2660
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 3.841726618705036,
|
| 1876 |
+
"grad_norm": 0.3415953516960144,
|
| 1877 |
+
"learning_rate": 4.8545151925798924e-05,
|
| 1878 |
+
"loss": 0.0297,
|
| 1879 |
+
"step": 2670
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 3.856115107913669,
|
| 1883 |
+
"grad_norm": 0.40429043769836426,
|
| 1884 |
+
"learning_rate": 4.821463220994848e-05,
|
| 1885 |
+
"loss": 0.0273,
|
| 1886 |
+
"step": 2680
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 3.870503597122302,
|
| 1890 |
+
"grad_norm": 0.3280932903289795,
|
| 1891 |
+
"learning_rate": 4.788419059186895e-05,
|
| 1892 |
+
"loss": 0.0247,
|
| 1893 |
+
"step": 2690
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 3.884892086330935,
|
| 1897 |
+
"grad_norm": 0.3966505229473114,
|
| 1898 |
+
"learning_rate": 4.7553841526147205e-05,
|
| 1899 |
+
"loss": 0.0278,
|
| 1900 |
+
"step": 2700
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 3.899280575539568,
|
| 1904 |
+
"grad_norm": 0.32661014795303345,
|
| 1905 |
+
"learning_rate": 4.722359946332156e-05,
|
| 1906 |
+
"loss": 0.027,
|
| 1907 |
+
"step": 2710
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 3.9136690647482015,
|
| 1911 |
+
"grad_norm": 0.3605490028858185,
|
| 1912 |
+
"learning_rate": 4.6893478849249654e-05,
|
| 1913 |
+
"loss": 0.0269,
|
| 1914 |
+
"step": 2720
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 3.9280575539568345,
|
| 1918 |
+
"grad_norm": 0.3466671109199524,
|
| 1919 |
+
"learning_rate": 4.656349412447664e-05,
|
| 1920 |
+
"loss": 0.027,
|
| 1921 |
+
"step": 2730
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 3.9424460431654675,
|
| 1925 |
+
"grad_norm": 0.3359682857990265,
|
| 1926 |
+
"learning_rate": 4.623365972360337e-05,
|
| 1927 |
+
"loss": 0.0333,
|
| 1928 |
+
"step": 2740
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 3.956834532374101,
|
| 1932 |
+
"grad_norm": 0.32730910181999207,
|
| 1933 |
+
"learning_rate": 4.590399007465503e-05,
|
| 1934 |
+
"loss": 0.0233,
|
| 1935 |
+
"step": 2750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 3.971223021582734,
|
| 1939 |
+
"grad_norm": 0.3271855115890503,
|
| 1940 |
+
"learning_rate": 4.557449959845005e-05,
|
| 1941 |
+
"loss": 0.0248,
|
| 1942 |
+
"step": 2760
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 3.985611510791367,
|
| 1946 |
+
"grad_norm": 0.3906365633010864,
|
| 1947 |
+
"learning_rate": 4.524520270796927e-05,
|
| 1948 |
+
"loss": 0.0263,
|
| 1949 |
+
"step": 2770
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 4.0,
|
| 1953 |
+
"grad_norm": 0.6004611253738403,
|
| 1954 |
+
"learning_rate": 4.491611380772545e-05,
|
| 1955 |
+
"loss": 0.0253,
|
| 1956 |
+
"step": 2780
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 4.014388489208633,
|
| 1960 |
+
"grad_norm": 0.3637600541114807,
|
| 1961 |
+
"learning_rate": 4.458724729313318e-05,
|
| 1962 |
+
"loss": 0.0247,
|
| 1963 |
+
"step": 2790
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 4.028776978417266,
|
| 1967 |
+
"grad_norm": 0.3008747100830078,
|
| 1968 |
+
"learning_rate": 4.42586175498792e-05,
|
| 1969 |
+
"loss": 0.0237,
|
| 1970 |
+
"step": 2800
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 4.043165467625899,
|
| 1974 |
+
"grad_norm": 0.20412537455558777,
|
| 1975 |
+
"learning_rate": 4.3930238953293094e-05,
|
| 1976 |
+
"loss": 0.0266,
|
| 1977 |
+
"step": 2810
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 4.057553956834532,
|
| 1981 |
+
"grad_norm": 0.20325446128845215,
|
| 1982 |
+
"learning_rate": 4.360212586771847e-05,
|
| 1983 |
+
"loss": 0.0206,
|
| 1984 |
+
"step": 2820
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 4.071942446043165,
|
| 1988 |
+
"grad_norm": 0.3927724063396454,
|
| 1989 |
+
"learning_rate": 4.327429264588463e-05,
|
| 1990 |
+
"loss": 0.0265,
|
| 1991 |
+
"step": 2830
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 4.086330935251799,
|
| 1995 |
+
"grad_norm": 0.30222776532173157,
|
| 1996 |
+
"learning_rate": 4.2946753628278725e-05,
|
| 1997 |
+
"loss": 0.0252,
|
| 1998 |
+
"step": 2840
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 4.100719424460432,
|
| 2002 |
+
"grad_norm": 0.3702770471572876,
|
| 2003 |
+
"learning_rate": 4.2619523142518474e-05,
|
| 2004 |
+
"loss": 0.0214,
|
| 2005 |
+
"step": 2850
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 4.115107913669065,
|
| 2009 |
+
"grad_norm": 0.40920382738113403,
|
| 2010 |
+
"learning_rate": 4.229261550272539e-05,
|
| 2011 |
+
"loss": 0.0286,
|
| 2012 |
+
"step": 2860
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 4.129496402877698,
|
| 2016 |
+
"grad_norm": 0.3616887629032135,
|
| 2017 |
+
"learning_rate": 4.196604500889868e-05,
|
| 2018 |
+
"loss": 0.0246,
|
| 2019 |
+
"step": 2870
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 4.143884892086331,
|
| 2023 |
+
"grad_norm": 0.155805766582489,
|
| 2024 |
+
"learning_rate": 4.163982594628969e-05,
|
| 2025 |
+
"loss": 0.0249,
|
| 2026 |
+
"step": 2880
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 4.158273381294964,
|
| 2030 |
+
"grad_norm": 0.3218766152858734,
|
| 2031 |
+
"learning_rate": 4.131397258477702e-05,
|
| 2032 |
+
"loss": 0.0241,
|
| 2033 |
+
"step": 2890
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 4.172661870503597,
|
| 2037 |
+
"grad_norm": 0.30419379472732544,
|
| 2038 |
+
"learning_rate": 4.0988499178242315e-05,
|
| 2039 |
+
"loss": 0.0279,
|
| 2040 |
+
"step": 2900
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 4.18705035971223,
|
| 2044 |
+
"grad_norm": 0.3262396454811096,
|
| 2045 |
+
"learning_rate": 4.066341996394678e-05,
|
| 2046 |
+
"loss": 0.0313,
|
| 2047 |
+
"step": 2910
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 4.201438848920863,
|
| 2051 |
+
"grad_norm": 0.3692721426486969,
|
| 2052 |
+
"learning_rate": 4.033874916190833e-05,
|
| 2053 |
+
"loss": 0.0224,
|
| 2054 |
+
"step": 2920
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 4.215827338129497,
|
| 2058 |
+
"grad_norm": 0.28412437438964844,
|
| 2059 |
+
"learning_rate": 4.001450097427966e-05,
|
| 2060 |
+
"loss": 0.022,
|
| 2061 |
+
"step": 2930
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 4.23021582733813,
|
| 2065 |
+
"grad_norm": 0.3362281620502472,
|
| 2066 |
+
"learning_rate": 3.9690689584726894e-05,
|
| 2067 |
+
"loss": 0.0267,
|
| 2068 |
+
"step": 2940
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 4.244604316546763,
|
| 2072 |
+
"grad_norm": 0.2920989394187927,
|
| 2073 |
+
"learning_rate": 3.936732915780923e-05,
|
| 2074 |
+
"loss": 0.0215,
|
| 2075 |
+
"step": 2950
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 4.258992805755396,
|
| 2079 |
+
"grad_norm": 0.27789244055747986,
|
| 2080 |
+
"learning_rate": 3.904443383835929e-05,
|
| 2081 |
+
"loss": 0.0278,
|
| 2082 |
+
"step": 2960
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 4.273381294964029,
|
| 2086 |
+
"grad_norm": 0.32670503854751587,
|
| 2087 |
+
"learning_rate": 3.872201775086437e-05,
|
| 2088 |
+
"loss": 0.0232,
|
| 2089 |
+
"step": 2970
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 4.287769784172662,
|
| 2093 |
+
"grad_norm": 0.25809797644615173,
|
| 2094 |
+
"learning_rate": 3.8400094998848616e-05,
|
| 2095 |
+
"loss": 0.0238,
|
| 2096 |
+
"step": 2980
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 4.302158273381295,
|
| 2100 |
+
"grad_norm": 0.21341978013515472,
|
| 2101 |
+
"learning_rate": 3.807867966425611e-05,
|
| 2102 |
+
"loss": 0.0238,
|
| 2103 |
+
"step": 2990
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 4.316546762589928,
|
| 2107 |
+
"grad_norm": 0.2576282024383545,
|
| 2108 |
+
"learning_rate": 3.775778580683481e-05,
|
| 2109 |
+
"loss": 0.0256,
|
| 2110 |
+
"step": 3000
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 4.330935251798561,
|
| 2114 |
+
"grad_norm": 0.2862774729728699,
|
| 2115 |
+
"learning_rate": 3.743742746352156e-05,
|
| 2116 |
+
"loss": 0.0243,
|
| 2117 |
+
"step": 3010
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 4.345323741007194,
|
| 2121 |
+
"grad_norm": 0.38920220732688904,
|
| 2122 |
+
"learning_rate": 3.711761864782817e-05,
|
| 2123 |
+
"loss": 0.0345,
|
| 2124 |
+
"step": 3020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 4.359712230215827,
|
| 2128 |
+
"grad_norm": 0.3424301743507385,
|
| 2129 |
+
"learning_rate": 3.679837334922825e-05,
|
| 2130 |
+
"loss": 0.0222,
|
| 2131 |
+
"step": 3030
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 4.374100719424461,
|
| 2135 |
+
"grad_norm": 0.21217834949493408,
|
| 2136 |
+
"learning_rate": 3.647970553254538e-05,
|
| 2137 |
+
"loss": 0.0216,
|
| 2138 |
+
"step": 3040
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 4.388489208633094,
|
| 2142 |
+
"grad_norm": 0.2746034562587738,
|
| 2143 |
+
"learning_rate": 3.61616291373422e-05,
|
| 2144 |
+
"loss": 0.0226,
|
| 2145 |
+
"step": 3050
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 4.402877697841727,
|
| 2149 |
+
"grad_norm": 0.2410103976726532,
|
| 2150 |
+
"learning_rate": 3.584415807731065e-05,
|
| 2151 |
+
"loss": 0.0268,
|
| 2152 |
+
"step": 3060
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 4.41726618705036,
|
| 2156 |
+
"grad_norm": 0.26465702056884766,
|
| 2157 |
+
"learning_rate": 3.552730623966337e-05,
|
| 2158 |
+
"loss": 0.0245,
|
| 2159 |
+
"step": 3070
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 4.431654676258993,
|
| 2163 |
+
"grad_norm": 0.37287843227386475,
|
| 2164 |
+
"learning_rate": 3.521108748452617e-05,
|
| 2165 |
+
"loss": 0.0222,
|
| 2166 |
+
"step": 3080
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 4.446043165467626,
|
| 2170 |
+
"grad_norm": 0.3482387661933899,
|
| 2171 |
+
"learning_rate": 3.489551564433186e-05,
|
| 2172 |
+
"loss": 0.024,
|
| 2173 |
+
"step": 3090
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 4.460431654676259,
|
| 2177 |
+
"grad_norm": 0.386384516954422,
|
| 2178 |
+
"learning_rate": 3.4580604523215006e-05,
|
| 2179 |
+
"loss": 0.0253,
|
| 2180 |
+
"step": 3100
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 4.474820143884892,
|
| 2184 |
+
"grad_norm": 0.4487895667552948,
|
| 2185 |
+
"learning_rate": 3.4266367896408216e-05,
|
| 2186 |
+
"loss": 0.0259,
|
| 2187 |
+
"step": 3110
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 4.489208633093525,
|
| 2191 |
+
"grad_norm": 0.3981205224990845,
|
| 2192 |
+
"learning_rate": 3.3952819509639534e-05,
|
| 2193 |
+
"loss": 0.0289,
|
| 2194 |
+
"step": 3120
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 4.503597122302159,
|
| 2198 |
+
"grad_norm": 0.3784792721271515,
|
| 2199 |
+
"learning_rate": 3.3639973078531165e-05,
|
| 2200 |
+
"loss": 0.0205,
|
| 2201 |
+
"step": 3130
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 4.517985611510792,
|
| 2205 |
+
"grad_norm": 0.2165733426809311,
|
| 2206 |
+
"learning_rate": 3.332784228799947e-05,
|
| 2207 |
+
"loss": 0.0246,
|
| 2208 |
+
"step": 3140
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 4.532374100719425,
|
| 2212 |
+
"grad_norm": 0.20944079756736755,
|
| 2213 |
+
"learning_rate": 3.301644079165638e-05,
|
| 2214 |
+
"loss": 0.0229,
|
| 2215 |
+
"step": 3150
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 4.546762589928058,
|
| 2219 |
+
"grad_norm": 0.3632758557796478,
|
| 2220 |
+
"learning_rate": 3.27057822112122e-05,
|
| 2221 |
+
"loss": 0.0199,
|
| 2222 |
+
"step": 3160
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 4.561151079136691,
|
| 2226 |
+
"grad_norm": 0.2845008671283722,
|
| 2227 |
+
"learning_rate": 3.239588013587958e-05,
|
| 2228 |
+
"loss": 0.0247,
|
| 2229 |
+
"step": 3170
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 4.575539568345324,
|
| 2233 |
+
"grad_norm": 0.19694705307483673,
|
| 2234 |
+
"learning_rate": 3.208674812177926e-05,
|
| 2235 |
+
"loss": 0.0239,
|
| 2236 |
+
"step": 3180
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 4.589928057553957,
|
| 2240 |
+
"grad_norm": 0.2823171019554138,
|
| 2241 |
+
"learning_rate": 3.177839969134698e-05,
|
| 2242 |
+
"loss": 0.0228,
|
| 2243 |
+
"step": 3190
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 4.60431654676259,
|
| 2247 |
+
"grad_norm": 0.3103242516517639,
|
| 2248 |
+
"learning_rate": 3.1470848332742e-05,
|
| 2249 |
+
"loss": 0.0245,
|
| 2250 |
+
"step": 3200
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 4.618705035971223,
|
| 2254 |
+
"grad_norm": 0.25528964400291443,
|
| 2255 |
+
"learning_rate": 3.116410749925708e-05,
|
| 2256 |
+
"loss": 0.0258,
|
| 2257 |
+
"step": 3210
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 4.633093525179856,
|
| 2261 |
+
"grad_norm": 0.22488592565059662,
|
| 2262 |
+
"learning_rate": 3.085819060872995e-05,
|
| 2263 |
+
"loss": 0.0227,
|
| 2264 |
+
"step": 3220
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 4.647482014388489,
|
| 2268 |
+
"grad_norm": 0.19623073935508728,
|
| 2269 |
+
"learning_rate": 3.055311104295648e-05,
|
| 2270 |
+
"loss": 0.0228,
|
| 2271 |
+
"step": 3230
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 4.661870503597123,
|
| 2275 |
+
"grad_norm": 0.261444091796875,
|
| 2276 |
+
"learning_rate": 3.024888214710517e-05,
|
| 2277 |
+
"loss": 0.018,
|
| 2278 |
+
"step": 3240
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 4.676258992805756,
|
| 2282 |
+
"grad_norm": 0.2911510169506073,
|
| 2283 |
+
"learning_rate": 2.994551722913349e-05,
|
| 2284 |
+
"loss": 0.0194,
|
| 2285 |
+
"step": 3250
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 4.690647482014389,
|
| 2289 |
+
"grad_norm": 0.24137957394123077,
|
| 2290 |
+
"learning_rate": 2.9643029559205727e-05,
|
| 2291 |
+
"loss": 0.0273,
|
| 2292 |
+
"step": 3260
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 4.705035971223022,
|
| 2296 |
+
"grad_norm": 0.3094216287136078,
|
| 2297 |
+
"learning_rate": 2.934143236911248e-05,
|
| 2298 |
+
"loss": 0.0224,
|
| 2299 |
+
"step": 3270
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 4.719424460431655,
|
| 2303 |
+
"grad_norm": 0.2885383665561676,
|
| 2304 |
+
"learning_rate": 2.90407388516919e-05,
|
| 2305 |
+
"loss": 0.021,
|
| 2306 |
+
"step": 3280
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 4.733812949640288,
|
| 2310 |
+
"grad_norm": 0.19373250007629395,
|
| 2311 |
+
"learning_rate": 2.8740962160252495e-05,
|
| 2312 |
+
"loss": 0.0218,
|
| 2313 |
+
"step": 3290
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 4.748201438848921,
|
| 2317 |
+
"grad_norm": 0.21961501240730286,
|
| 2318 |
+
"learning_rate": 2.844211540799797e-05,
|
| 2319 |
+
"loss": 0.0207,
|
| 2320 |
+
"step": 3300
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 4.762589928057554,
|
| 2324 |
+
"grad_norm": 0.2696077227592468,
|
| 2325 |
+
"learning_rate": 2.8144211667453368e-05,
|
| 2326 |
+
"loss": 0.0194,
|
| 2327 |
+
"step": 3310
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 4.7769784172661875,
|
| 2331 |
+
"grad_norm": 0.15248015522956848,
|
| 2332 |
+
"learning_rate": 2.7847263969893344e-05,
|
| 2333 |
+
"loss": 0.0217,
|
| 2334 |
+
"step": 3320
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 4.7913669064748206,
|
| 2338 |
+
"grad_norm": 0.3766297399997711,
|
| 2339 |
+
"learning_rate": 2.7551285304772206e-05,
|
| 2340 |
+
"loss": 0.0196,
|
| 2341 |
+
"step": 3330
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 4.805755395683454,
|
| 2345 |
+
"grad_norm": 0.26399049162864685,
|
| 2346 |
+
"learning_rate": 2.7256288619155567e-05,
|
| 2347 |
+
"loss": 0.022,
|
| 2348 |
+
"step": 3340
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 4.820143884892087,
|
| 2352 |
+
"grad_norm": 0.2961912453174591,
|
| 2353 |
+
"learning_rate": 2.6962286817154158e-05,
|
| 2354 |
+
"loss": 0.0218,
|
| 2355 |
+
"step": 3350
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 4.83453237410072,
|
| 2359 |
+
"grad_norm": 0.2859618663787842,
|
| 2360 |
+
"learning_rate": 2.6669292759359166e-05,
|
| 2361 |
+
"loss": 0.022,
|
| 2362 |
+
"step": 3360
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 4.848920863309353,
|
| 2366 |
+
"grad_norm": 0.27274298667907715,
|
| 2367 |
+
"learning_rate": 2.637731926227993e-05,
|
| 2368 |
+
"loss": 0.0264,
|
| 2369 |
+
"step": 3370
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 4.863309352517986,
|
| 2373 |
+
"grad_norm": 0.2099636048078537,
|
| 2374 |
+
"learning_rate": 2.6086379097783033e-05,
|
| 2375 |
+
"loss": 0.0212,
|
| 2376 |
+
"step": 3380
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 4.877697841726619,
|
| 2380 |
+
"grad_norm": 0.29768136143684387,
|
| 2381 |
+
"learning_rate": 2.579648499253377e-05,
|
| 2382 |
+
"loss": 0.0278,
|
| 2383 |
+
"step": 3390
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 4.892086330935252,
|
| 2387 |
+
"grad_norm": 0.21766944229602814,
|
| 2388 |
+
"learning_rate": 2.5507649627439466e-05,
|
| 2389 |
+
"loss": 0.0221,
|
| 2390 |
+
"step": 3400
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 4.906474820143885,
|
| 2394 |
+
"grad_norm": 0.259082168340683,
|
| 2395 |
+
"learning_rate": 2.5219885637094653e-05,
|
| 2396 |
+
"loss": 0.0263,
|
| 2397 |
+
"step": 3410
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 4.920863309352518,
|
| 2401 |
+
"grad_norm": 0.3283446729183197,
|
| 2402 |
+
"learning_rate": 2.4933205609228533e-05,
|
| 2403 |
+
"loss": 0.0233,
|
| 2404 |
+
"step": 3420
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 4.935251798561151,
|
| 2408 |
+
"grad_norm": 0.2639141082763672,
|
| 2409 |
+
"learning_rate": 2.464762208415419e-05,
|
| 2410 |
+
"loss": 0.0166,
|
| 2411 |
+
"step": 3430
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 4.9496402877697845,
|
| 2415 |
+
"grad_norm": 0.24496804177761078,
|
| 2416 |
+
"learning_rate": 2.4363147554220213e-05,
|
| 2417 |
+
"loss": 0.0199,
|
| 2418 |
+
"step": 3440
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 4.9640287769784175,
|
| 2422 |
+
"grad_norm": 0.22914527356624603,
|
| 2423 |
+
"learning_rate": 2.407979446326411e-05,
|
| 2424 |
+
"loss": 0.0233,
|
| 2425 |
+
"step": 3450
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 4.9784172661870505,
|
| 2429 |
+
"grad_norm": 0.2796647250652313,
|
| 2430 |
+
"learning_rate": 2.379757520606799e-05,
|
| 2431 |
+
"loss": 0.0251,
|
| 2432 |
+
"step": 3460
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 4.9928057553956835,
|
| 2436 |
+
"grad_norm": 0.3473125398159027,
|
| 2437 |
+
"learning_rate": 2.3516502127816455e-05,
|
| 2438 |
+
"loss": 0.0189,
|
| 2439 |
+
"step": 3470
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 5.0071942446043165,
|
| 2443 |
+
"grad_norm": 0.2519746720790863,
|
| 2444 |
+
"learning_rate": 2.323658752355647e-05,
|
| 2445 |
+
"loss": 0.0194,
|
| 2446 |
+
"step": 3480
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 5.0215827338129495,
|
| 2450 |
+
"grad_norm": 0.3052903115749359,
|
| 2451 |
+
"learning_rate": 2.2957843637659654e-05,
|
| 2452 |
+
"loss": 0.0218,
|
| 2453 |
+
"step": 3490
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 5.0359712230215825,
|
| 2457 |
+
"grad_norm": 0.31738007068634033,
|
| 2458 |
+
"learning_rate": 2.2680282663286552e-05,
|
| 2459 |
+
"loss": 0.0277,
|
| 2460 |
+
"step": 3500
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 5.0503597122302155,
|
| 2464 |
+
"grad_norm": 0.27347511053085327,
|
| 2465 |
+
"learning_rate": 2.2403916741853364e-05,
|
| 2466 |
+
"loss": 0.0216,
|
| 2467 |
+
"step": 3510
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 5.0647482014388485,
|
| 2471 |
+
"grad_norm": 0.1651657223701477,
|
| 2472 |
+
"learning_rate": 2.2128757962500817e-05,
|
| 2473 |
+
"loss": 0.0193,
|
| 2474 |
+
"step": 3520
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 5.079136690647482,
|
| 2478 |
+
"grad_norm": 0.2815980911254883,
|
| 2479 |
+
"learning_rate": 2.1854818361565275e-05,
|
| 2480 |
+
"loss": 0.0214,
|
| 2481 |
+
"step": 3530
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 5.093525179856115,
|
| 2485 |
+
"grad_norm": 0.21221508085727692,
|
| 2486 |
+
"learning_rate": 2.1582109922052364e-05,
|
| 2487 |
+
"loss": 0.0185,
|
| 2488 |
+
"step": 3540
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 5.107913669064748,
|
| 2492 |
+
"grad_norm": 0.2501748204231262,
|
| 2493 |
+
"learning_rate": 2.1310644573112635e-05,
|
| 2494 |
+
"loss": 0.022,
|
| 2495 |
+
"step": 3550
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 5.122302158273381,
|
| 2499 |
+
"grad_norm": 0.3282177150249481,
|
| 2500 |
+
"learning_rate": 2.1040434189519924e-05,
|
| 2501 |
+
"loss": 0.0209,
|
| 2502 |
+
"step": 3560
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 5.136690647482014,
|
| 2506 |
+
"grad_norm": 0.3199025094509125,
|
| 2507 |
+
"learning_rate": 2.0771490591151733e-05,
|
| 2508 |
+
"loss": 0.0246,
|
| 2509 |
+
"step": 3570
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 5.151079136690647,
|
| 2513 |
+
"grad_norm": 0.2805894613265991,
|
| 2514 |
+
"learning_rate": 2.0503825542472317e-05,
|
| 2515 |
+
"loss": 0.0215,
|
| 2516 |
+
"step": 3580
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 5.16546762589928,
|
| 2520 |
+
"grad_norm": 0.31447431445121765,
|
| 2521 |
+
"learning_rate": 2.023745075201805e-05,
|
| 2522 |
+
"loss": 0.0227,
|
| 2523 |
+
"step": 3590
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 5.179856115107913,
|
| 2527 |
+
"grad_norm": 0.2621252238750458,
|
| 2528 |
+
"learning_rate": 1.9972377871885157e-05,
|
| 2529 |
+
"loss": 0.0182,
|
| 2530 |
+
"step": 3600
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 5.194244604316546,
|
| 2534 |
+
"grad_norm": 0.3012264370918274,
|
| 2535 |
+
"learning_rate": 1.970861849722017e-05,
|
| 2536 |
+
"loss": 0.0209,
|
| 2537 |
+
"step": 3610
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 5.2086330935251794,
|
| 2541 |
+
"grad_norm": 0.2724980115890503,
|
| 2542 |
+
"learning_rate": 1.9446184165712587e-05,
|
| 2543 |
+
"loss": 0.0194,
|
| 2544 |
+
"step": 3620
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 5.223021582733813,
|
| 2548 |
+
"grad_norm": 0.37012600898742676,
|
| 2549 |
+
"learning_rate": 1.9185086357090214e-05,
|
| 2550 |
+
"loss": 0.0204,
|
| 2551 |
+
"step": 3630
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 5.237410071942446,
|
| 2555 |
+
"grad_norm": 0.25987115502357483,
|
| 2556 |
+
"learning_rate": 1.8925336492617057e-05,
|
| 2557 |
+
"loss": 0.0216,
|
| 2558 |
+
"step": 3640
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 5.251798561151079,
|
| 2562 |
+
"grad_norm": 0.16672657430171967,
|
| 2563 |
+
"learning_rate": 1.8666945934593666e-05,
|
| 2564 |
+
"loss": 0.0194,
|
| 2565 |
+
"step": 3650
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 5.266187050359712,
|
| 2569 |
+
"grad_norm": 0.2172255963087082,
|
| 2570 |
+
"learning_rate": 1.8409925985860126e-05,
|
| 2571 |
+
"loss": 0.0187,
|
| 2572 |
+
"step": 3660
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 5.280575539568345,
|
| 2576 |
+
"grad_norm": 0.2498919516801834,
|
| 2577 |
+
"learning_rate": 1.8154287889301603e-05,
|
| 2578 |
+
"loss": 0.019,
|
| 2579 |
+
"step": 3670
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 5.294964028776978,
|
| 2583 |
+
"grad_norm": 0.2137516736984253,
|
| 2584 |
+
"learning_rate": 1.7900042827356612e-05,
|
| 2585 |
+
"loss": 0.0197,
|
| 2586 |
+
"step": 3680
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 5.309352517985611,
|
| 2590 |
+
"grad_norm": 0.2527858018875122,
|
| 2591 |
+
"learning_rate": 1.76472019215278e-05,
|
| 2592 |
+
"loss": 0.0234,
|
| 2593 |
+
"step": 3690
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 5.323741007194244,
|
| 2597 |
+
"grad_norm": 0.24032394587993622,
|
| 2598 |
+
"learning_rate": 1.739577623189545e-05,
|
| 2599 |
+
"loss": 0.0228,
|
| 2600 |
+
"step": 3700
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 5.338129496402877,
|
| 2604 |
+
"grad_norm": 0.2509600818157196,
|
| 2605 |
+
"learning_rate": 1.7145776756633768e-05,
|
| 2606 |
+
"loss": 0.019,
|
| 2607 |
+
"step": 3710
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 5.35251798561151,
|
| 2611 |
+
"grad_norm": 0.17037628591060638,
|
| 2612 |
+
"learning_rate": 1.6897214431529646e-05,
|
| 2613 |
+
"loss": 0.0227,
|
| 2614 |
+
"step": 3720
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 5.366906474820144,
|
| 2618 |
+
"grad_norm": 0.24370284378528595,
|
| 2619 |
+
"learning_rate": 1.6650100129504475e-05,
|
| 2620 |
+
"loss": 0.0192,
|
| 2621 |
+
"step": 3730
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 5.381294964028777,
|
| 2625 |
+
"grad_norm": 0.15974493324756622,
|
| 2626 |
+
"learning_rate": 1.6404444660138335e-05,
|
| 2627 |
+
"loss": 0.0172,
|
| 2628 |
+
"step": 3740
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 5.39568345323741,
|
| 2632 |
+
"grad_norm": 0.21873117983341217,
|
| 2633 |
+
"learning_rate": 1.616025876919725e-05,
|
| 2634 |
+
"loss": 0.0235,
|
| 2635 |
+
"step": 3750
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 5.410071942446043,
|
| 2639 |
+
"grad_norm": 0.31240856647491455,
|
| 2640 |
+
"learning_rate": 1.5917553138163172e-05,
|
| 2641 |
+
"loss": 0.0178,
|
| 2642 |
+
"step": 3760
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 5.424460431654676,
|
| 2646 |
+
"grad_norm": 0.2524794042110443,
|
| 2647 |
+
"learning_rate": 1.5676338383766632e-05,
|
| 2648 |
+
"loss": 0.0276,
|
| 2649 |
+
"step": 3770
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 5.438848920863309,
|
| 2653 |
+
"grad_norm": 0.38405412435531616,
|
| 2654 |
+
"learning_rate": 1.5436625057522447e-05,
|
| 2655 |
+
"loss": 0.0198,
|
| 2656 |
+
"step": 3780
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 5.453237410071942,
|
| 2660 |
+
"grad_norm": 0.2832338511943817,
|
| 2661 |
+
"learning_rate": 1.519842364526804e-05,
|
| 2662 |
+
"loss": 0.0182,
|
| 2663 |
+
"step": 3790
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 5.467625899280575,
|
| 2667 |
+
"grad_norm": 0.18746420741081238,
|
| 2668 |
+
"learning_rate": 1.4961744566704855e-05,
|
| 2669 |
+
"loss": 0.0212,
|
| 2670 |
+
"step": 3800
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 5.482014388489208,
|
| 2674 |
+
"grad_norm": 0.1788494884967804,
|
| 2675 |
+
"learning_rate": 1.4726598174942551e-05,
|
| 2676 |
+
"loss": 0.0194,
|
| 2677 |
+
"step": 3810
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 5.496402877697841,
|
| 2681 |
+
"grad_norm": 0.28818562626838684,
|
| 2682 |
+
"learning_rate": 1.4492994756046035e-05,
|
| 2683 |
+
"loss": 0.0201,
|
| 2684 |
+
"step": 3820
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 5.510791366906475,
|
| 2688 |
+
"grad_norm": 0.2759857773780823,
|
| 2689 |
+
"learning_rate": 1.4260944528585645e-05,
|
| 2690 |
+
"loss": 0.0191,
|
| 2691 |
+
"step": 3830
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 5.525179856115108,
|
| 2695 |
+
"grad_norm": 0.23128828406333923,
|
| 2696 |
+
"learning_rate": 1.4030457643190048e-05,
|
| 2697 |
+
"loss": 0.0176,
|
| 2698 |
+
"step": 3840
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 5.539568345323741,
|
| 2702 |
+
"grad_norm": 0.20284543931484222,
|
| 2703 |
+
"learning_rate": 1.3801544182102311e-05,
|
| 2704 |
+
"loss": 0.0179,
|
| 2705 |
+
"step": 3850
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 5.553956834532374,
|
| 2709 |
+
"grad_norm": 0.320852130651474,
|
| 2710 |
+
"learning_rate": 1.3574214158738763e-05,
|
| 2711 |
+
"loss": 0.0172,
|
| 2712 |
+
"step": 3860
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 5.568345323741007,
|
| 2716 |
+
"grad_norm": 0.1807006448507309,
|
| 2717 |
+
"learning_rate": 1.3348477517251101e-05,
|
| 2718 |
+
"loss": 0.0216,
|
| 2719 |
+
"step": 3870
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 5.58273381294964,
|
| 2723 |
+
"grad_norm": 0.2539335787296295,
|
| 2724 |
+
"learning_rate": 1.312434413209131e-05,
|
| 2725 |
+
"loss": 0.0203,
|
| 2726 |
+
"step": 3880
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 5.597122302158273,
|
| 2730 |
+
"grad_norm": 0.2468968778848648,
|
| 2731 |
+
"learning_rate": 1.2901823807579727e-05,
|
| 2732 |
+
"loss": 0.0174,
|
| 2733 |
+
"step": 3890
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 5.611510791366906,
|
| 2737 |
+
"grad_norm": 0.3068462908267975,
|
| 2738 |
+
"learning_rate": 1.2680926277476245e-05,
|
| 2739 |
+
"loss": 0.0226,
|
| 2740 |
+
"step": 3900
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 5.625899280575539,
|
| 2744 |
+
"grad_norm": 0.2476012408733368,
|
| 2745 |
+
"learning_rate": 1.2461661204554397e-05,
|
| 2746 |
+
"loss": 0.0205,
|
| 2747 |
+
"step": 3910
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 5.640287769784173,
|
| 2751 |
+
"grad_norm": 0.22999592125415802,
|
| 2752 |
+
"learning_rate": 1.2244038180178835e-05,
|
| 2753 |
+
"loss": 0.0215,
|
| 2754 |
+
"step": 3920
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 5.654676258992806,
|
| 2758 |
+
"grad_norm": 0.30130302906036377,
|
| 2759 |
+
"learning_rate": 1.2028066723885612e-05,
|
| 2760 |
+
"loss": 0.0192,
|
| 2761 |
+
"step": 3930
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 5.669064748201439,
|
| 2765 |
+
"grad_norm": 0.17271891236305237,
|
| 2766 |
+
"learning_rate": 1.1813756282965888e-05,
|
| 2767 |
+
"loss": 0.0199,
|
| 2768 |
+
"step": 3940
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 5.683453237410072,
|
| 2772 |
+
"grad_norm": 0.26495277881622314,
|
| 2773 |
+
"learning_rate": 1.1601116232052638e-05,
|
| 2774 |
+
"loss": 0.0229,
|
| 2775 |
+
"step": 3950
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 5.697841726618705,
|
| 2779 |
+
"grad_norm": 0.3486734926700592,
|
| 2780 |
+
"learning_rate": 1.1390155872710517e-05,
|
| 2781 |
+
"loss": 0.0162,
|
| 2782 |
+
"step": 3960
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 5.712230215827338,
|
| 2786 |
+
"grad_norm": 0.2978843152523041,
|
| 2787 |
+
"learning_rate": 1.1180884433029087e-05,
|
| 2788 |
+
"loss": 0.0177,
|
| 2789 |
+
"step": 3970
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 5.726618705035971,
|
| 2793 |
+
"grad_norm": 0.24050843715667725,
|
| 2794 |
+
"learning_rate": 1.097331106721904e-05,
|
| 2795 |
+
"loss": 0.0168,
|
| 2796 |
+
"step": 3980
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 5.741007194244604,
|
| 2800 |
+
"grad_norm": 0.3195205330848694,
|
| 2801 |
+
"learning_rate": 1.0767444855211862e-05,
|
| 2802 |
+
"loss": 0.0168,
|
| 2803 |
+
"step": 3990
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 5.755395683453237,
|
| 2807 |
+
"grad_norm": 0.3221238851547241,
|
| 2808 |
+
"learning_rate": 1.0563294802262558e-05,
|
| 2809 |
+
"loss": 0.0193,
|
| 2810 |
+
"step": 4000
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 5.76978417266187,
|
| 2814 |
+
"grad_norm": 0.16008298099040985,
|
| 2815 |
+
"learning_rate": 1.0360869838555809e-05,
|
| 2816 |
+
"loss": 0.0191,
|
| 2817 |
+
"step": 4010
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 5.784172661870503,
|
| 2821 |
+
"grad_norm": 0.17987611889839172,
|
| 2822 |
+
"learning_rate": 1.0160178818815313e-05,
|
| 2823 |
+
"loss": 0.0178,
|
| 2824 |
+
"step": 4020
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 5.798561151079137,
|
| 2828 |
+
"grad_norm": 0.2427978813648224,
|
| 2829 |
+
"learning_rate": 9.961230521916387e-06,
|
| 2830 |
+
"loss": 0.0199,
|
| 2831 |
+
"step": 4030
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 5.81294964028777,
|
| 2835 |
+
"grad_norm": 0.21939091384410858,
|
| 2836 |
+
"learning_rate": 9.764033650502074e-06,
|
| 2837 |
+
"loss": 0.0212,
|
| 2838 |
+
"step": 4040
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 5.827338129496403,
|
| 2842 |
+
"grad_norm": 0.15761235356330872,
|
| 2843 |
+
"learning_rate": 9.568596830602344e-06,
|
| 2844 |
+
"loss": 0.0148,
|
| 2845 |
+
"step": 4050
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 5.841726618705036,
|
| 2849 |
+
"grad_norm": 0.3346870541572571,
|
| 2850 |
+
"learning_rate": 9.37492861125681e-06,
|
| 2851 |
+
"loss": 0.0211,
|
| 2852 |
+
"step": 4060
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 5.856115107913669,
|
| 2856 |
+
"grad_norm": 0.18777605891227722,
|
| 2857 |
+
"learning_rate": 9.183037464140804e-06,
|
| 2858 |
+
"loss": 0.0209,
|
| 2859 |
+
"step": 4070
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 5.870503597122302,
|
| 2863 |
+
"grad_norm": 0.2534259557723999,
|
| 2864 |
+
"learning_rate": 8.992931783194735e-06,
|
| 2865 |
+
"loss": 0.0176,
|
| 2866 |
+
"step": 4080
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 5.884892086330935,
|
| 2870 |
+
"grad_norm": 0.24573186039924622,
|
| 2871 |
+
"learning_rate": 8.80461988425696e-06,
|
| 2872 |
+
"loss": 0.0187,
|
| 2873 |
+
"step": 4090
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 5.899280575539568,
|
| 2877 |
+
"grad_norm": 0.18632175028324127,
|
| 2878 |
+
"learning_rate": 8.618110004699974e-06,
|
| 2879 |
+
"loss": 0.0179,
|
| 2880 |
+
"step": 4100
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 5.913669064748201,
|
| 2884 |
+
"grad_norm": 0.36280566453933716,
|
| 2885 |
+
"learning_rate": 8.4334103030701e-06,
|
| 2886 |
+
"loss": 0.0208,
|
| 2887 |
+
"step": 4110
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 5.928057553956835,
|
| 2891 |
+
"grad_norm": 0.3409193158149719,
|
| 2892 |
+
"learning_rate": 8.25052885873066e-06,
|
| 2893 |
+
"loss": 0.0178,
|
| 2894 |
+
"step": 4120
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 5.942446043165468,
|
| 2898 |
+
"grad_norm": 0.18552973866462708,
|
| 2899 |
+
"learning_rate": 8.06947367150846e-06,
|
| 2900 |
+
"loss": 0.0184,
|
| 2901 |
+
"step": 4130
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 5.956834532374101,
|
| 2905 |
+
"grad_norm": 0.27096259593963623,
|
| 2906 |
+
"learning_rate": 7.890252661343938e-06,
|
| 2907 |
+
"loss": 0.0185,
|
| 2908 |
+
"step": 4140
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 5.971223021582734,
|
| 2912 |
+
"grad_norm": 0.22975075244903564,
|
| 2913 |
+
"learning_rate": 7.712873667944681e-06,
|
| 2914 |
+
"loss": 0.0196,
|
| 2915 |
+
"step": 4150
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 5.985611510791367,
|
| 2919 |
+
"grad_norm": 0.17989930510520935,
|
| 2920 |
+
"learning_rate": 7.537344450442469e-06,
|
| 2921 |
+
"loss": 0.0134,
|
| 2922 |
+
"step": 4160
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 6.0,
|
| 2926 |
+
"grad_norm": 0.3224349021911621,
|
| 2927 |
+
"learning_rate": 7.36367268705393e-06,
|
| 2928 |
+
"loss": 0.02,
|
| 2929 |
+
"step": 4170
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 6.014388489208633,
|
| 2933 |
+
"grad_norm": 0.3323955237865448,
|
| 2934 |
+
"learning_rate": 7.1918659747446e-06,
|
| 2935 |
+
"loss": 0.0221,
|
| 2936 |
+
"step": 4180
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 6.028776978417266,
|
| 2940 |
+
"grad_norm": 0.1492568701505661,
|
| 2941 |
+
"learning_rate": 7.021931828896666e-06,
|
| 2942 |
+
"loss": 0.0226,
|
| 2943 |
+
"step": 4190
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 6.043165467625899,
|
| 2947 |
+
"grad_norm": 0.18576060235500336,
|
| 2948 |
+
"learning_rate": 6.8538776829801584e-06,
|
| 2949 |
+
"loss": 0.0186,
|
| 2950 |
+
"step": 4200
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 6.057553956834532,
|
| 2954 |
+
"grad_norm": 0.1899133026599884,
|
| 2955 |
+
"learning_rate": 6.687710888227849e-06,
|
| 2956 |
+
"loss": 0.0171,
|
| 2957 |
+
"step": 4210
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 6.071942446043165,
|
| 2961 |
+
"grad_norm": 0.22415462136268616,
|
| 2962 |
+
"learning_rate": 6.5234387133136565e-06,
|
| 2963 |
+
"loss": 0.0183,
|
| 2964 |
+
"step": 4220
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 6.086330935251799,
|
| 2968 |
+
"grad_norm": 0.22749356925487518,
|
| 2969 |
+
"learning_rate": 6.361068344034665e-06,
|
| 2970 |
+
"loss": 0.0173,
|
| 2971 |
+
"step": 4230
|
| 2972 |
+
},
|
| 2973 |
+
{
|
| 2974 |
+
"epoch": 6.100719424460432,
|
| 2975 |
+
"grad_norm": 0.22299259901046753,
|
| 2976 |
+
"learning_rate": 6.200606882996846e-06,
|
| 2977 |
+
"loss": 0.0179,
|
| 2978 |
+
"step": 4240
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"epoch": 6.115107913669065,
|
| 2982 |
+
"grad_norm": 0.28247034549713135,
|
| 2983 |
+
"learning_rate": 6.042061349304312e-06,
|
| 2984 |
+
"loss": 0.0182,
|
| 2985 |
+
"step": 4250
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 6.129496402877698,
|
| 2989 |
+
"grad_norm": 0.1768667995929718,
|
| 2990 |
+
"learning_rate": 5.885438678252342e-06,
|
| 2991 |
+
"loss": 0.0226,
|
| 2992 |
+
"step": 4260
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 6.143884892086331,
|
| 2996 |
+
"grad_norm": 0.1978299468755722,
|
| 2997 |
+
"learning_rate": 5.730745721023939e-06,
|
| 2998 |
+
"loss": 0.0145,
|
| 2999 |
+
"step": 4270
|
| 3000 |
+
},
|
| 3001 |
+
{
|
| 3002 |
+
"epoch": 6.158273381294964,
|
| 3003 |
+
"grad_norm": 0.1870502531528473,
|
| 3004 |
+
"learning_rate": 5.577989244390192e-06,
|
| 3005 |
+
"loss": 0.0183,
|
| 3006 |
+
"step": 4280
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 6.172661870503597,
|
| 3010 |
+
"grad_norm": 0.24501974880695343,
|
| 3011 |
+
"learning_rate": 5.4271759304142635e-06,
|
| 3012 |
+
"loss": 0.0166,
|
| 3013 |
+
"step": 4290
|
| 3014 |
+
},
|
| 3015 |
+
{
|
| 3016 |
+
"epoch": 6.18705035971223,
|
| 3017 |
+
"grad_norm": 0.15945523977279663,
|
| 3018 |
+
"learning_rate": 5.278312376159051e-06,
|
| 3019 |
+
"loss": 0.0158,
|
| 3020 |
+
"step": 4300
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 6.201438848920863,
|
| 3024 |
+
"grad_norm": 0.2202870100736618,
|
| 3025 |
+
"learning_rate": 5.1314050933986944e-06,
|
| 3026 |
+
"loss": 0.0167,
|
| 3027 |
+
"step": 4310
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 6.215827338129497,
|
| 3031 |
+
"grad_norm": 0.2053917497396469,
|
| 3032 |
+
"learning_rate": 4.986460508333634e-06,
|
| 3033 |
+
"loss": 0.0181,
|
| 3034 |
+
"step": 4320
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 6.23021582733813,
|
| 3038 |
+
"grad_norm": 0.27106422185897827,
|
| 3039 |
+
"learning_rate": 4.843484961309597e-06,
|
| 3040 |
+
"loss": 0.0168,
|
| 3041 |
+
"step": 4330
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 6.244604316546763,
|
| 3045 |
+
"grad_norm": 0.2498120814561844,
|
| 3046 |
+
"learning_rate": 4.702484706540161e-06,
|
| 3047 |
+
"loss": 0.0138,
|
| 3048 |
+
"step": 4340
|
| 3049 |
+
},
|
| 3050 |
+
{
|
| 3051 |
+
"epoch": 6.258992805755396,
|
| 3052 |
+
"grad_norm": 0.21487227082252502,
|
| 3053 |
+
"learning_rate": 4.563465911833259e-06,
|
| 3054 |
+
"loss": 0.0155,
|
| 3055 |
+
"step": 4350
|
| 3056 |
+
},
|
| 3057 |
+
{
|
| 3058 |
+
"epoch": 6.273381294964029,
|
| 3059 |
+
"grad_norm": 0.17530597746372223,
|
| 3060 |
+
"learning_rate": 4.426434658321344e-06,
|
| 3061 |
+
"loss": 0.0159,
|
| 3062 |
+
"step": 4360
|
| 3063 |
+
},
|
| 3064 |
+
{
|
| 3065 |
+
"epoch": 6.287769784172662,
|
| 3066 |
+
"grad_norm": 0.14909881353378296,
|
| 3067 |
+
"learning_rate": 4.2913969401953466e-06,
|
| 3068 |
+
"loss": 0.018,
|
| 3069 |
+
"step": 4370
|
| 3070 |
+
},
|
| 3071 |
+
{
|
| 3072 |
+
"epoch": 6.302158273381295,
|
| 3073 |
+
"grad_norm": 0.14552098512649536,
|
| 3074 |
+
"learning_rate": 4.15835866444253e-06,
|
| 3075 |
+
"loss": 0.0135,
|
| 3076 |
+
"step": 4380
|
| 3077 |
+
},
|
| 3078 |
+
{
|
| 3079 |
+
"epoch": 6.316546762589928,
|
| 3080 |
+
"grad_norm": 0.13056610524654388,
|
| 3081 |
+
"learning_rate": 4.027325650588043e-06,
|
| 3082 |
+
"loss": 0.0188,
|
| 3083 |
+
"step": 4390
|
| 3084 |
+
},
|
| 3085 |
+
{
|
| 3086 |
+
"epoch": 6.330935251798561,
|
| 3087 |
+
"grad_norm": 0.22001564502716064,
|
| 3088 |
+
"learning_rate": 3.898303630440419e-06,
|
| 3089 |
+
"loss": 0.0229,
|
| 3090 |
+
"step": 4400
|
| 3091 |
+
},
|
| 3092 |
+
{
|
| 3093 |
+
"epoch": 6.345323741007194,
|
| 3094 |
+
"grad_norm": 0.21187953650951385,
|
| 3095 |
+
"learning_rate": 3.7712982478407877e-06,
|
| 3096 |
+
"loss": 0.0221,
|
| 3097 |
+
"step": 4410
|
| 3098 |
+
},
|
| 3099 |
+
{
|
| 3100 |
+
"epoch": 6.359712230215827,
|
| 3101 |
+
"grad_norm": 0.2211296558380127,
|
| 3102 |
+
"learning_rate": 3.6463150584160053e-06,
|
| 3103 |
+
"loss": 0.0192,
|
| 3104 |
+
"step": 4420
|
| 3105 |
+
},
|
| 3106 |
+
{
|
| 3107 |
+
"epoch": 6.374100719424461,
|
| 3108 |
+
"grad_norm": 0.3535507917404175,
|
| 3109 |
+
"learning_rate": 3.5233595293356957e-06,
|
| 3110 |
+
"loss": 0.0185,
|
| 3111 |
+
"step": 4430
|
| 3112 |
+
},
|
| 3113 |
+
{
|
| 3114 |
+
"epoch": 6.388489208633094,
|
| 3115 |
+
"grad_norm": 0.1929875761270523,
|
| 3116 |
+
"learning_rate": 3.4024370390730033e-06,
|
| 3117 |
+
"loss": 0.0195,
|
| 3118 |
+
"step": 4440
|
| 3119 |
+
},
|
| 3120 |
+
{
|
| 3121 |
+
"epoch": 6.402877697841727,
|
| 3122 |
+
"grad_norm": 0.13576443493366241,
|
| 3123 |
+
"learning_rate": 3.2835528771693992e-06,
|
| 3124 |
+
"loss": 0.0192,
|
| 3125 |
+
"step": 4450
|
| 3126 |
+
},
|
| 3127 |
+
{
|
| 3128 |
+
"epoch": 6.41726618705036,
|
| 3129 |
+
"grad_norm": 0.18413235247135162,
|
| 3130 |
+
"learning_rate": 3.1667122440032505e-06,
|
| 3131 |
+
"loss": 0.0143,
|
| 3132 |
+
"step": 4460
|
| 3133 |
+
},
|
| 3134 |
+
{
|
| 3135 |
+
"epoch": 6.431654676258993,
|
| 3136 |
+
"grad_norm": 0.2805648744106293,
|
| 3137 |
+
"learning_rate": 3.051920250562351e-06,
|
| 3138 |
+
"loss": 0.0173,
|
| 3139 |
+
"step": 4470
|
| 3140 |
+
},
|
| 3141 |
+
{
|
| 3142 |
+
"epoch": 6.446043165467626,
|
| 3143 |
+
"grad_norm": 0.2387954592704773,
|
| 3144 |
+
"learning_rate": 2.939181918220385e-06,
|
| 3145 |
+
"loss": 0.0173,
|
| 3146 |
+
"step": 4480
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 6.460431654676259,
|
| 3150 |
+
"grad_norm": 0.4205700159072876,
|
| 3151 |
+
"learning_rate": 2.8285021785172226e-06,
|
| 3152 |
+
"loss": 0.0193,
|
| 3153 |
+
"step": 4490
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 6.474820143884892,
|
| 3157 |
+
"grad_norm": 0.2456294447183609,
|
| 3158 |
+
"learning_rate": 2.7198858729432288e-06,
|
| 3159 |
+
"loss": 0.0171,
|
| 3160 |
+
"step": 4500
|
| 3161 |
+
},
|
| 3162 |
+
{
|
| 3163 |
+
"epoch": 6.489208633093525,
|
| 3164 |
+
"grad_norm": 0.25572633743286133,
|
| 3165 |
+
"learning_rate": 2.6133377527274905e-06,
|
| 3166 |
+
"loss": 0.0247,
|
| 3167 |
+
"step": 4510
|
| 3168 |
+
},
|
| 3169 |
+
{
|
| 3170 |
+
"epoch": 6.503597122302159,
|
| 3171 |
+
"grad_norm": 0.22960637509822845,
|
| 3172 |
+
"learning_rate": 2.5088624786299366e-06,
|
| 3173 |
+
"loss": 0.0189,
|
| 3174 |
+
"step": 4520
|
| 3175 |
+
},
|
| 3176 |
+
{
|
| 3177 |
+
"epoch": 6.517985611510792,
|
| 3178 |
+
"grad_norm": 0.2636794447898865,
|
| 3179 |
+
"learning_rate": 2.406464620737531e-06,
|
| 3180 |
+
"loss": 0.0184,
|
| 3181 |
+
"step": 4530
|
| 3182 |
+
},
|
| 3183 |
+
{
|
| 3184 |
+
"epoch": 6.532374100719425,
|
| 3185 |
+
"grad_norm": 0.18809296190738678,
|
| 3186 |
+
"learning_rate": 2.3061486582642734e-06,
|
| 3187 |
+
"loss": 0.0151,
|
| 3188 |
+
"step": 4540
|
| 3189 |
+
},
|
| 3190 |
+
{
|
| 3191 |
+
"epoch": 6.546762589928058,
|
| 3192 |
+
"grad_norm": 0.21880139410495758,
|
| 3193 |
+
"learning_rate": 2.2079189793553667e-06,
|
| 3194 |
+
"loss": 0.0147,
|
| 3195 |
+
"step": 4550
|
| 3196 |
+
},
|
| 3197 |
+
{
|
| 3198 |
+
"epoch": 6.561151079136691,
|
| 3199 |
+
"grad_norm": 0.24475565552711487,
|
| 3200 |
+
"learning_rate": 2.111779880895165e-06,
|
| 3201 |
+
"loss": 0.0142,
|
| 3202 |
+
"step": 4560
|
| 3203 |
+
},
|
| 3204 |
+
{
|
| 3205 |
+
"epoch": 6.575539568345324,
|
| 3206 |
+
"grad_norm": 0.19628864526748657,
|
| 3207 |
+
"learning_rate": 2.01773556831929e-06,
|
| 3208 |
+
"loss": 0.0163,
|
| 3209 |
+
"step": 4570
|
| 3210 |
+
},
|
| 3211 |
+
{
|
| 3212 |
+
"epoch": 6.589928057553957,
|
| 3213 |
+
"grad_norm": 0.2752235531806946,
|
| 3214 |
+
"learning_rate": 1.9257901554306513e-06,
|
| 3215 |
+
"loss": 0.0171,
|
| 3216 |
+
"step": 4580
|
| 3217 |
+
},
|
| 3218 |
+
{
|
| 3219 |
+
"epoch": 6.60431654676259,
|
| 3220 |
+
"grad_norm": 0.19138553738594055,
|
| 3221 |
+
"learning_rate": 1.835947664219445e-06,
|
| 3222 |
+
"loss": 0.0171,
|
| 3223 |
+
"step": 4590
|
| 3224 |
+
},
|
| 3225 |
+
{
|
| 3226 |
+
"epoch": 6.618705035971223,
|
| 3227 |
+
"grad_norm": 0.16835470497608185,
|
| 3228 |
+
"learning_rate": 1.748212024687307e-06,
|
| 3229 |
+
"loss": 0.0144,
|
| 3230 |
+
"step": 4600
|
| 3231 |
+
},
|
| 3232 |
+
{
|
| 3233 |
+
"epoch": 6.633093525179856,
|
| 3234 |
+
"grad_norm": 0.286734938621521,
|
| 3235 |
+
"learning_rate": 1.6625870746753147e-06,
|
| 3236 |
+
"loss": 0.0177,
|
| 3237 |
+
"step": 4610
|
| 3238 |
+
},
|
| 3239 |
+
{
|
| 3240 |
+
"epoch": 6.647482014388489,
|
| 3241 |
+
"grad_norm": 0.21962714195251465,
|
| 3242 |
+
"learning_rate": 1.5790765596961853e-06,
|
| 3243 |
+
"loss": 0.0154,
|
| 3244 |
+
"step": 4620
|
| 3245 |
+
},
|
| 3246 |
+
{
|
| 3247 |
+
"epoch": 6.661870503597123,
|
| 3248 |
+
"grad_norm": 0.19300445914268494,
|
| 3249 |
+
"learning_rate": 1.4976841327703717e-06,
|
| 3250 |
+
"loss": 0.0144,
|
| 3251 |
+
"step": 4630
|
| 3252 |
+
},
|
| 3253 |
+
{
|
| 3254 |
+
"epoch": 6.676258992805756,
|
| 3255 |
+
"grad_norm": 0.327727347612381,
|
| 3256 |
+
"learning_rate": 1.4184133542663014e-06,
|
| 3257 |
+
"loss": 0.0138,
|
| 3258 |
+
"step": 4640
|
| 3259 |
+
},
|
| 3260 |
+
{
|
| 3261 |
+
"epoch": 6.690647482014389,
|
| 3262 |
+
"grad_norm": 0.21031884849071503,
|
| 3263 |
+
"learning_rate": 1.341267691744641e-06,
|
| 3264 |
+
"loss": 0.0177,
|
| 3265 |
+
"step": 4650
|
| 3266 |
+
},
|
| 3267 |
+
{
|
| 3268 |
+
"epoch": 6.705035971223022,
|
| 3269 |
+
"grad_norm": 0.19146843254566193,
|
| 3270 |
+
"learning_rate": 1.2662505198065666e-06,
|
| 3271 |
+
"loss": 0.0172,
|
| 3272 |
+
"step": 4660
|
| 3273 |
+
},
|
| 3274 |
+
{
|
| 3275 |
+
"epoch": 6.719424460431655,
|
| 3276 |
+
"grad_norm": 0.20976972579956055,
|
| 3277 |
+
"learning_rate": 1.193365119946216e-06,
|
| 3278 |
+
"loss": 0.0189,
|
| 3279 |
+
"step": 4670
|
| 3280 |
+
},
|
| 3281 |
+
{
|
| 3282 |
+
"epoch": 6.733812949640288,
|
| 3283 |
+
"grad_norm": 0.15400229394435883,
|
| 3284 |
+
"learning_rate": 1.1226146804070859e-06,
|
| 3285 |
+
"loss": 0.0155,
|
| 3286 |
+
"step": 4680
|
| 3287 |
+
},
|
| 3288 |
+
{
|
| 3289 |
+
"epoch": 6.748201438848921,
|
| 3290 |
+
"grad_norm": 0.27561676502227783,
|
| 3291 |
+
"learning_rate": 1.0540022960426111e-06,
|
| 3292 |
+
"loss": 0.0181,
|
| 3293 |
+
"step": 4690
|
| 3294 |
+
},
|
| 3295 |
+
{
|
| 3296 |
+
"epoch": 6.762589928057554,
|
| 3297 |
+
"grad_norm": 0.152504563331604,
|
| 3298 |
+
"learning_rate": 9.875309681807443e-07,
|
| 3299 |
+
"loss": 0.0215,
|
| 3300 |
+
"step": 4700
|
| 3301 |
+
},
|
| 3302 |
+
{
|
| 3303 |
+
"epoch": 6.7769784172661875,
|
| 3304 |
+
"grad_norm": 0.2141709178686142,
|
| 3305 |
+
"learning_rate": 9.232036044927061e-07,
|
| 3306 |
+
"loss": 0.0142,
|
| 3307 |
+
"step": 4710
|
| 3308 |
+
},
|
| 3309 |
+
{
|
| 3310 |
+
"epoch": 6.7913669064748206,
|
| 3311 |
+
"grad_norm": 0.2436617612838745,
|
| 3312 |
+
"learning_rate": 8.610230188657919e-07,
|
| 3313 |
+
"loss": 0.016,
|
| 3314 |
+
"step": 4720
|
| 3315 |
+
},
|
| 3316 |
+
{
|
| 3317 |
+
"epoch": 6.805755395683454,
|
| 3318 |
+
"grad_norm": 0.20569947361946106,
|
| 3319 |
+
"learning_rate": 8.009919312802372e-07,
|
| 3320 |
+
"loss": 0.0175,
|
| 3321 |
+
"step": 4730
|
| 3322 |
+
},
|
| 3323 |
+
{
|
| 3324 |
+
"epoch": 6.820143884892087,
|
| 3325 |
+
"grad_norm": 0.19237467646598816,
|
| 3326 |
+
"learning_rate": 7.431129676902904e-07,
|
| 3327 |
+
"loss": 0.0161,
|
| 3328 |
+
"step": 4740
|
| 3329 |
+
},
|
| 3330 |
+
{
|
| 3331 |
+
"epoch": 6.83453237410072,
|
| 3332 |
+
"grad_norm": 0.2221413552761078,
|
| 3333 |
+
"learning_rate": 6.873886599093215e-07,
|
| 3334 |
+
"loss": 0.0169,
|
| 3335 |
+
"step": 4750
|
| 3336 |
+
},
|
| 3337 |
+
{
|
| 3338 |
+
"epoch": 6.848920863309353,
|
| 3339 |
+
"grad_norm": 0.1079205796122551,
|
| 3340 |
+
"learning_rate": 6.338214454990776e-07,
|
| 3341 |
+
"loss": 0.017,
|
| 3342 |
+
"step": 4760
|
| 3343 |
+
},
|
| 3344 |
+
{
|
| 3345 |
+
"epoch": 6.863309352517986,
|
| 3346 |
+
"grad_norm": 0.15608228743076324,
|
| 3347 |
+
"learning_rate": 5.82413667663051e-07,
|
| 3348 |
+
"loss": 0.0166,
|
| 3349 |
+
"step": 4770
|
| 3350 |
+
},
|
| 3351 |
+
{
|
| 3352 |
+
"epoch": 6.877697841726619,
|
| 3353 |
+
"grad_norm": 0.2794274389743805,
|
| 3354 |
+
"learning_rate": 5.331675751439725e-07,
|
| 3355 |
+
"loss": 0.0171,
|
| 3356 |
+
"step": 4780
|
| 3357 |
+
},
|
| 3358 |
+
{
|
| 3359 |
+
"epoch": 6.892086330935252,
|
| 3360 |
+
"grad_norm": 0.15306247770786285,
|
| 3361 |
+
"learning_rate": 4.86085322125479e-07,
|
| 3362 |
+
"loss": 0.0132,
|
| 3363 |
+
"step": 4790
|
| 3364 |
+
},
|
| 3365 |
+
{
|
| 3366 |
+
"epoch": 6.906474820143885,
|
| 3367 |
+
"grad_norm": 0.2512068450450897,
|
| 3368 |
+
"learning_rate": 4.411689681378284e-07,
|
| 3369 |
+
"loss": 0.0145,
|
| 3370 |
+
"step": 4800
|
| 3371 |
+
},
|
| 3372 |
+
{
|
| 3373 |
+
"epoch": 6.920863309352518,
|
| 3374 |
+
"grad_norm": 0.2012811005115509,
|
| 3375 |
+
"learning_rate": 3.9842047796786466e-07,
|
| 3376 |
+
"loss": 0.0163,
|
| 3377 |
+
"step": 4810
|
| 3378 |
+
},
|
| 3379 |
+
{
|
| 3380 |
+
"epoch": 6.935251798561151,
|
| 3381 |
+
"grad_norm": 0.1467149555683136,
|
| 3382 |
+
"learning_rate": 3.578417215730323e-07,
|
| 3383 |
+
"loss": 0.0153,
|
| 3384 |
+
"step": 4820
|
| 3385 |
+
},
|
| 3386 |
+
{
|
| 3387 |
+
"epoch": 6.9496402877697845,
|
| 3388 |
+
"grad_norm": 0.1585516780614853,
|
| 3389 |
+
"learning_rate": 3.1943447399958027e-07,
|
| 3390 |
+
"loss": 0.0178,
|
| 3391 |
+
"step": 4830
|
| 3392 |
+
},
|
| 3393 |
+
{
|
| 3394 |
+
"epoch": 6.9640287769784175,
|
| 3395 |
+
"grad_norm": 0.21342824399471283,
|
| 3396 |
+
"learning_rate": 2.8320041530495724e-07,
|
| 3397 |
+
"loss": 0.0245,
|
| 3398 |
+
"step": 4840
|
| 3399 |
+
},
|
| 3400 |
+
{
|
| 3401 |
+
"epoch": 6.9784172661870505,
|
| 3402 |
+
"grad_norm": 0.24616077542304993,
|
| 3403 |
+
"learning_rate": 2.491411304842539e-07,
|
| 3404 |
+
"loss": 0.0189,
|
| 3405 |
+
"step": 4850
|
| 3406 |
+
},
|
| 3407 |
+
{
|
| 3408 |
+
"epoch": 6.9928057553956835,
|
| 3409 |
+
"grad_norm": 0.24037420749664307,
|
| 3410 |
+
"learning_rate": 2.1725810940094183e-07,
|
| 3411 |
+
"loss": 0.0156,
|
| 3412 |
+
"step": 4860
|
| 3413 |
+
},
|
| 3414 |
+
{
|
| 3415 |
+
"epoch": 7.0071942446043165,
|
| 3416 |
+
"grad_norm": 0.1024838462471962,
|
| 3417 |
+
"learning_rate": 1.8755274672164202e-07,
|
| 3418 |
+
"loss": 0.0158,
|
| 3419 |
+
"step": 4870
|
| 3420 |
+
},
|
| 3421 |
+
{
|
| 3422 |
+
"epoch": 7.0215827338129495,
|
| 3423 |
+
"grad_norm": 0.09105559438467026,
|
| 3424 |
+
"learning_rate": 1.600263418551573e-07,
|
| 3425 |
+
"loss": 0.0198,
|
| 3426 |
+
"step": 4880
|
| 3427 |
+
},
|
| 3428 |
+
{
|
| 3429 |
+
"epoch": 7.0359712230215825,
|
| 3430 |
+
"grad_norm": 0.1912888139486313,
|
| 3431 |
+
"learning_rate": 1.346800988955954e-07,
|
| 3432 |
+
"loss": 0.0166,
|
| 3433 |
+
"step": 4890
|
| 3434 |
+
},
|
| 3435 |
+
{
|
| 3436 |
+
"epoch": 7.0503597122302155,
|
| 3437 |
+
"grad_norm": 0.1934279054403305,
|
| 3438 |
+
"learning_rate": 1.1151512656975005e-07,
|
| 3439 |
+
"loss": 0.0199,
|
| 3440 |
+
"step": 4900
|
| 3441 |
+
},
|
| 3442 |
+
{
|
| 3443 |
+
"epoch": 7.0647482014388485,
|
| 3444 |
+
"grad_norm": 0.13366839289665222,
|
| 3445 |
+
"learning_rate": 9.053243818853973e-08,
|
| 3446 |
+
"loss": 0.0186,
|
| 3447 |
+
"step": 4910
|
| 3448 |
+
},
|
| 3449 |
+
{
|
| 3450 |
+
"epoch": 7.079136690647482,
|
| 3451 |
+
"grad_norm": 0.15921102464199066,
|
| 3452 |
+
"learning_rate": 7.173295160273763e-08,
|
| 3453 |
+
"loss": 0.0145,
|
| 3454 |
+
"step": 4920
|
| 3455 |
+
},
|
| 3456 |
+
{
|
| 3457 |
+
"epoch": 7.093525179856115,
|
| 3458 |
+
"grad_norm": 0.21049778163433075,
|
| 3459 |
+
"learning_rate": 5.511748916279258e-08,
|
| 3460 |
+
"loss": 0.0158,
|
| 3461 |
+
"step": 4930
|
| 3462 |
+
},
|
| 3463 |
+
{
|
| 3464 |
+
"epoch": 7.107913669064748,
|
| 3465 |
+
"grad_norm": 0.15058083832263947,
|
| 3466 |
+
"learning_rate": 4.068677768285234e-08,
|
| 3467 |
+
"loss": 0.0177,
|
| 3468 |
+
"step": 4940
|
| 3469 |
+
},
|
| 3470 |
+
{
|
| 3471 |
+
"epoch": 7.122302158273381,
|
| 3472 |
+
"grad_norm": 0.183713898062706,
|
| 3473 |
+
"learning_rate": 2.844144840898344e-08,
|
| 3474 |
+
"loss": 0.0152,
|
| 3475 |
+
"step": 4950
|
| 3476 |
+
},
|
| 3477 |
+
{
|
| 3478 |
+
"epoch": 7.136690647482014,
|
| 3479 |
+
"grad_norm": 0.17629458010196686,
|
| 3480 |
+
"learning_rate": 1.8382036991559936e-08,
|
| 3481 |
+
"loss": 0.0153,
|
| 3482 |
+
"step": 4960
|
| 3483 |
+
},
|
| 3484 |
+
{
|
| 3485 |
+
"epoch": 7.151079136690647,
|
| 3486 |
+
"grad_norm": 0.20629747211933136,
|
| 3487 |
+
"learning_rate": 1.0508983461832156e-08,
|
| 3488 |
+
"loss": 0.0167,
|
| 3489 |
+
"step": 4970
|
| 3490 |
+
},
|
| 3491 |
+
{
|
| 3492 |
+
"epoch": 7.16546762589928,
|
| 3493 |
+
"grad_norm": 0.27550452947616577,
|
| 3494 |
+
"learning_rate": 4.822632212653222e-09,
|
| 3495 |
+
"loss": 0.0231,
|
| 3496 |
+
"step": 4980
|
| 3497 |
+
},
|
| 3498 |
+
{
|
| 3499 |
+
"epoch": 7.179856115107913,
|
| 3500 |
+
"grad_norm": 0.3242543041706085,
|
| 3501 |
+
"learning_rate": 1.3232319834632912e-09,
|
| 3502 |
+
"loss": 0.0152,
|
| 3503 |
+
"step": 4990
|
| 3504 |
+
},
|
| 3505 |
+
{
|
| 3506 |
+
"epoch": 7.194244604316546,
|
| 3507 |
+
"grad_norm": 0.23857244849205017,
|
| 3508 |
+
"learning_rate": 1.0935849353854721e-11,
|
| 3509 |
+
"loss": 0.0177,
|
| 3510 |
+
"step": 5000
|
| 3511 |
}
|
| 3512 |
],
|
| 3513 |
"logging_steps": 10,
|
|
|
|
| 3522 |
"should_evaluate": false,
|
| 3523 |
"should_log": false,
|
| 3524 |
"should_save": true,
|
| 3525 |
+
"should_training_stop": true
|
| 3526 |
},
|
| 3527 |
"attributes": {}
|
| 3528 |
}
|