Upload folder using huggingface_hub
Browse files- model-00001-of-00009.safetensors +1 -1
- model-00009-of-00009.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state_0.pth +1 -1
- rng_state_1.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1431 -3
model-00001-of-00009.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4913763168
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4094f490d340236f6ab7658065a8bc228deacc1a4ad7e1f7b1ff29e731661dfe
|
| 3 |
size 4913763168
|
model-00009-of-00009.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3873591000
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a1a1fc2452362572d2435ec51578548c7a4008252bcec5ace7deb9271389ee2
|
| 3 |
size 3873591000
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3145730908
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e0ea31fb5824200b0dac909a6caedbc0f3d726f9e6b0abca674d35b48747bc88
|
| 3 |
size 3145730908
|
rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:34bcae41c589c7e4cab7b2ef263b878c90c2741404a6af11994dc31537b2319b
|
| 3 |
size 14512
|
rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d05dc84075e8f7dd1191c36f3be9dda12073208e12f7d2cef433c38d6336774a
|
| 3 |
size 14512
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:263f9ba4d1ea5aafd961965649939bc24679533d480d13b25adbc1c55d668440
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1195,6 +1195,1434 @@
|
|
| 1195 |
"learning_rate": 0.00019639601779707655,
|
| 1196 |
"loss": 0.6144,
|
| 1197 |
"step": 19800
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1198 |
}
|
| 1199 |
],
|
| 1200 |
"logging_steps": 100,
|
|
@@ -1202,7 +2630,7 @@
|
|
| 1202 |
"num_input_tokens_seen": 0,
|
| 1203 |
"num_train_epochs": 3,
|
| 1204 |
"save_steps": 100,
|
| 1205 |
-
"total_flos":
|
| 1206 |
"train_batch_size": 3,
|
| 1207 |
"trial_name": null,
|
| 1208 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.6425043183126764,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 43600,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1195 |
"learning_rate": 0.00019639601779707655,
|
| 1196 |
"loss": 0.6144,
|
| 1197 |
"step": 19800
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 1.21,
|
| 1201 |
+
"learning_rate": 0.00019548953715356758,
|
| 1202 |
+
"loss": 0.6123,
|
| 1203 |
+
"step": 19900
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 1.21,
|
| 1207 |
+
"learning_rate": 0.0001945812232353326,
|
| 1208 |
+
"loss": 0.6138,
|
| 1209 |
+
"step": 20000
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 1.22,
|
| 1213 |
+
"learning_rate": 0.00019367111264834846,
|
| 1214 |
+
"loss": 0.6173,
|
| 1215 |
+
"step": 20100
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 1.22,
|
| 1219 |
+
"learning_rate": 0.0001927592420709998,
|
| 1220 |
+
"loss": 0.5983,
|
| 1221 |
+
"step": 20200
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.23,
|
| 1225 |
+
"learning_rate": 0.00019184564825260053,
|
| 1226 |
+
"loss": 0.6037,
|
| 1227 |
+
"step": 20300
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.24,
|
| 1231 |
+
"learning_rate": 0.000190930368011913,
|
| 1232 |
+
"loss": 0.602,
|
| 1233 |
+
"step": 20400
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 1.24,
|
| 1237 |
+
"learning_rate": 0.00019001343823566412,
|
| 1238 |
+
"loss": 0.6095,
|
| 1239 |
+
"step": 20500
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 1.25,
|
| 1243 |
+
"learning_rate": 0.0001890948958770587,
|
| 1244 |
+
"loss": 0.6122,
|
| 1245 |
+
"step": 20600
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 1.25,
|
| 1249 |
+
"learning_rate": 0.00018817477795429028,
|
| 1250 |
+
"loss": 0.61,
|
| 1251 |
+
"step": 20700
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 1.26,
|
| 1255 |
+
"learning_rate": 0.00018725312154904925,
|
| 1256 |
+
"loss": 0.6034,
|
| 1257 |
+
"step": 20800
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 1.27,
|
| 1261 |
+
"learning_rate": 0.00018632996380502846,
|
| 1262 |
+
"loss": 0.6085,
|
| 1263 |
+
"step": 20900
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.27,
|
| 1267 |
+
"learning_rate": 0.00018540534192642614,
|
| 1268 |
+
"loss": 0.5977,
|
| 1269 |
+
"step": 21000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.28,
|
| 1273 |
+
"learning_rate": 0.00018447929317644672,
|
| 1274 |
+
"loss": 0.6071,
|
| 1275 |
+
"step": 21100
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 1.28,
|
| 1279 |
+
"learning_rate": 0.00018355185487579898,
|
| 1280 |
+
"loss": 0.6118,
|
| 1281 |
+
"step": 21200
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 1.29,
|
| 1285 |
+
"learning_rate": 0.00018262306440119198,
|
| 1286 |
+
"loss": 0.6078,
|
| 1287 |
+
"step": 21300
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 1.3,
|
| 1291 |
+
"learning_rate": 0.00018169295918382883,
|
| 1292 |
+
"loss": 0.6029,
|
| 1293 |
+
"step": 21400
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 1.3,
|
| 1297 |
+
"learning_rate": 0.00018076157670789803,
|
| 1298 |
+
"loss": 0.5974,
|
| 1299 |
+
"step": 21500
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 1.31,
|
| 1303 |
+
"learning_rate": 0.00017982895450906303,
|
| 1304 |
+
"loss": 0.6077,
|
| 1305 |
+
"step": 21600
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.32,
|
| 1309 |
+
"learning_rate": 0.00017889513017294923,
|
| 1310 |
+
"loss": 0.5953,
|
| 1311 |
+
"step": 21700
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.32,
|
| 1315 |
+
"learning_rate": 0.00017796014133362946,
|
| 1316 |
+
"loss": 0.5985,
|
| 1317 |
+
"step": 21800
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 1.33,
|
| 1321 |
+
"learning_rate": 0.00017702402567210723,
|
| 1322 |
+
"loss": 0.5987,
|
| 1323 |
+
"step": 21900
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 1.33,
|
| 1327 |
+
"learning_rate": 0.00017608682091479813,
|
| 1328 |
+
"loss": 0.6017,
|
| 1329 |
+
"step": 22000
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 1.34,
|
| 1333 |
+
"learning_rate": 0.00017514856483200937,
|
| 1334 |
+
"loss": 0.5985,
|
| 1335 |
+
"step": 22100
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 1.35,
|
| 1339 |
+
"learning_rate": 0.00017420929523641766,
|
| 1340 |
+
"loss": 0.5999,
|
| 1341 |
+
"step": 22200
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 1.35,
|
| 1345 |
+
"learning_rate": 0.0001732690499815454,
|
| 1346 |
+
"loss": 0.6052,
|
| 1347 |
+
"step": 22300
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.36,
|
| 1351 |
+
"learning_rate": 0.00017232786696023492,
|
| 1352 |
+
"loss": 0.6015,
|
| 1353 |
+
"step": 22400
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.36,
|
| 1357 |
+
"learning_rate": 0.00017138578410312162,
|
| 1358 |
+
"loss": 0.6002,
|
| 1359 |
+
"step": 22500
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 1.37,
|
| 1363 |
+
"learning_rate": 0.0001704428393771051,
|
| 1364 |
+
"loss": 0.6011,
|
| 1365 |
+
"step": 22600
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 1.38,
|
| 1369 |
+
"learning_rate": 0.00016949907078381927,
|
| 1370 |
+
"loss": 0.6045,
|
| 1371 |
+
"step": 22700
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 1.38,
|
| 1375 |
+
"learning_rate": 0.00016855451635810058,
|
| 1376 |
+
"loss": 0.607,
|
| 1377 |
+
"step": 22800
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 1.39,
|
| 1381 |
+
"learning_rate": 0.00016760921416645544,
|
| 1382 |
+
"loss": 0.6062,
|
| 1383 |
+
"step": 22900
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 1.39,
|
| 1387 |
+
"learning_rate": 0.00016666320230552593,
|
| 1388 |
+
"loss": 0.5966,
|
| 1389 |
+
"step": 23000
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.4,
|
| 1393 |
+
"learning_rate": 0.00016571651890055452,
|
| 1394 |
+
"loss": 0.6026,
|
| 1395 |
+
"step": 23100
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.41,
|
| 1399 |
+
"learning_rate": 0.0001647692021038477,
|
| 1400 |
+
"loss": 0.5941,
|
| 1401 |
+
"step": 23200
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 1.41,
|
| 1405 |
+
"learning_rate": 0.00016382129009323817,
|
| 1406 |
+
"loss": 0.5919,
|
| 1407 |
+
"step": 23300
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 1.42,
|
| 1411 |
+
"learning_rate": 0.00016287282107054643,
|
| 1412 |
+
"loss": 0.6035,
|
| 1413 |
+
"step": 23400
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 1.42,
|
| 1417 |
+
"learning_rate": 0.00016192383326004106,
|
| 1418 |
+
"loss": 0.6033,
|
| 1419 |
+
"step": 23500
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 1.43,
|
| 1423 |
+
"learning_rate": 0.00016097436490689838,
|
| 1424 |
+
"loss": 0.5961,
|
| 1425 |
+
"step": 23600
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 1.44,
|
| 1429 |
+
"learning_rate": 0.00016002445427566107,
|
| 1430 |
+
"loss": 0.589,
|
| 1431 |
+
"step": 23700
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.44,
|
| 1435 |
+
"learning_rate": 0.000159074139648696,
|
| 1436 |
+
"loss": 0.6,
|
| 1437 |
+
"step": 23800
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.45,
|
| 1441 |
+
"learning_rate": 0.0001581234593246516,
|
| 1442 |
+
"loss": 0.5951,
|
| 1443 |
+
"step": 23900
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 1.45,
|
| 1447 |
+
"learning_rate": 0.0001571724516169141,
|
| 1448 |
+
"loss": 0.6027,
|
| 1449 |
+
"step": 24000
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 1.46,
|
| 1453 |
+
"learning_rate": 0.00015622115485206385,
|
| 1454 |
+
"loss": 0.5993,
|
| 1455 |
+
"step": 24100
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 1.47,
|
| 1459 |
+
"learning_rate": 0.00015526960736833025,
|
| 1460 |
+
"loss": 0.5915,
|
| 1461 |
+
"step": 24200
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 1.47,
|
| 1465 |
+
"learning_rate": 0.00015431784751404707,
|
| 1466 |
+
"loss": 0.5977,
|
| 1467 |
+
"step": 24300
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 1.48,
|
| 1471 |
+
"learning_rate": 0.00015336591364610686,
|
| 1472 |
+
"loss": 0.6022,
|
| 1473 |
+
"step": 24400
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.48,
|
| 1477 |
+
"learning_rate": 0.00015241384412841493,
|
| 1478 |
+
"loss": 0.6019,
|
| 1479 |
+
"step": 24500
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.49,
|
| 1483 |
+
"learning_rate": 0.00015146167733034367,
|
| 1484 |
+
"loss": 0.5921,
|
| 1485 |
+
"step": 24600
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 1.5,
|
| 1489 |
+
"learning_rate": 0.00015050945162518574,
|
| 1490 |
+
"loss": 0.6011,
|
| 1491 |
+
"step": 24700
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 1.5,
|
| 1495 |
+
"learning_rate": 0.0001495572053886079,
|
| 1496 |
+
"loss": 0.5934,
|
| 1497 |
+
"step": 24800
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 1.51,
|
| 1501 |
+
"learning_rate": 0.00014860497699710433,
|
| 1502 |
+
"loss": 0.5893,
|
| 1503 |
+
"step": 24900
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 1.52,
|
| 1507 |
+
"learning_rate": 0.00014765280482645005,
|
| 1508 |
+
"loss": 0.5951,
|
| 1509 |
+
"step": 25000
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 1.52,
|
| 1513 |
+
"learning_rate": 0.00014670072725015437,
|
| 1514 |
+
"loss": 0.59,
|
| 1515 |
+
"step": 25100
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.53,
|
| 1519 |
+
"learning_rate": 0.00014574878263791426,
|
| 1520 |
+
"loss": 0.6008,
|
| 1521 |
+
"step": 25200
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.53,
|
| 1525 |
+
"learning_rate": 0.00014479700935406817,
|
| 1526 |
+
"loss": 0.5907,
|
| 1527 |
+
"step": 25300
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 1.54,
|
| 1531 |
+
"learning_rate": 0.0001438454457560498,
|
| 1532 |
+
"loss": 0.5977,
|
| 1533 |
+
"step": 25400
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 1.55,
|
| 1537 |
+
"learning_rate": 0.00014289413019284236,
|
| 1538 |
+
"loss": 0.5999,
|
| 1539 |
+
"step": 25500
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 1.55,
|
| 1543 |
+
"learning_rate": 0.00014194310100343292,
|
| 1544 |
+
"loss": 0.5979,
|
| 1545 |
+
"step": 25600
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 1.56,
|
| 1549 |
+
"learning_rate": 0.00014099239651526742,
|
| 1550 |
+
"loss": 0.5889,
|
| 1551 |
+
"step": 25700
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 1.56,
|
| 1555 |
+
"learning_rate": 0.0001400420550427061,
|
| 1556 |
+
"loss": 0.5994,
|
| 1557 |
+
"step": 25800
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.57,
|
| 1561 |
+
"learning_rate": 0.0001390921148854791,
|
| 1562 |
+
"loss": 0.5881,
|
| 1563 |
+
"step": 25900
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.58,
|
| 1567 |
+
"learning_rate": 0.00013814261432714336,
|
| 1568 |
+
"loss": 0.5947,
|
| 1569 |
+
"step": 26000
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 1.58,
|
| 1573 |
+
"learning_rate": 0.00013719359163353944,
|
| 1574 |
+
"loss": 0.5939,
|
| 1575 |
+
"step": 26100
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 1.59,
|
| 1579 |
+
"learning_rate": 0.0001362450850512494,
|
| 1580 |
+
"loss": 0.5905,
|
| 1581 |
+
"step": 26200
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 1.59,
|
| 1585 |
+
"learning_rate": 0.00013529713280605567,
|
| 1586 |
+
"loss": 0.5866,
|
| 1587 |
+
"step": 26300
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 1.6,
|
| 1591 |
+
"learning_rate": 0.00013434977310140012,
|
| 1592 |
+
"loss": 0.5942,
|
| 1593 |
+
"step": 26400
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 1.61,
|
| 1597 |
+
"learning_rate": 0.0001334030441168447,
|
| 1598 |
+
"loss": 0.584,
|
| 1599 |
+
"step": 26500
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.61,
|
| 1603 |
+
"learning_rate": 0.0001324569840065328,
|
| 1604 |
+
"loss": 0.5932,
|
| 1605 |
+
"step": 26600
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.62,
|
| 1609 |
+
"learning_rate": 0.0001315116308976514,
|
| 1610 |
+
"loss": 0.5896,
|
| 1611 |
+
"step": 26700
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 1.62,
|
| 1615 |
+
"learning_rate": 0.00013056702288889458,
|
| 1616 |
+
"loss": 0.5899,
|
| 1617 |
+
"step": 26800
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 1.63,
|
| 1621 |
+
"learning_rate": 0.00012962319804892827,
|
| 1622 |
+
"loss": 0.583,
|
| 1623 |
+
"step": 26900
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 1.64,
|
| 1627 |
+
"learning_rate": 0.00012868019441485568,
|
| 1628 |
+
"loss": 0.5895,
|
| 1629 |
+
"step": 27000
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 1.64,
|
| 1633 |
+
"learning_rate": 0.00012773804999068473,
|
| 1634 |
+
"loss": 0.5857,
|
| 1635 |
+
"step": 27100
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 1.65,
|
| 1639 |
+
"learning_rate": 0.00012679680274579636,
|
| 1640 |
+
"loss": 0.5929,
|
| 1641 |
+
"step": 27200
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.65,
|
| 1645 |
+
"learning_rate": 0.00012585649061341405,
|
| 1646 |
+
"loss": 0.5899,
|
| 1647 |
+
"step": 27300
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.66,
|
| 1651 |
+
"learning_rate": 0.00012491715148907554,
|
| 1652 |
+
"loss": 0.5856,
|
| 1653 |
+
"step": 27400
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 1.67,
|
| 1657 |
+
"learning_rate": 0.0001239788232291052,
|
| 1658 |
+
"loss": 0.5907,
|
| 1659 |
+
"step": 27500
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 1.67,
|
| 1663 |
+
"learning_rate": 0.00012304154364908856,
|
| 1664 |
+
"loss": 0.591,
|
| 1665 |
+
"step": 27600
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 1.68,
|
| 1669 |
+
"learning_rate": 0.00012210535052234835,
|
| 1670 |
+
"loss": 0.593,
|
| 1671 |
+
"step": 27700
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 1.68,
|
| 1675 |
+
"learning_rate": 0.00012117028157842202,
|
| 1676 |
+
"loss": 0.5879,
|
| 1677 |
+
"step": 27800
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 1.69,
|
| 1681 |
+
"learning_rate": 0.00012023637450154138,
|
| 1682 |
+
"loss": 0.5859,
|
| 1683 |
+
"step": 27900
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.7,
|
| 1687 |
+
"learning_rate": 0.00011930366692911378,
|
| 1688 |
+
"loss": 0.5866,
|
| 1689 |
+
"step": 28000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.7,
|
| 1693 |
+
"learning_rate": 0.00011837219645020536,
|
| 1694 |
+
"loss": 0.587,
|
| 1695 |
+
"step": 28100
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 1.71,
|
| 1699 |
+
"learning_rate": 0.00011744200060402608,
|
| 1700 |
+
"loss": 0.5844,
|
| 1701 |
+
"step": 28200
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 1.72,
|
| 1705 |
+
"learning_rate": 0.00011651311687841697,
|
| 1706 |
+
"loss": 0.5848,
|
| 1707 |
+
"step": 28300
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 1.72,
|
| 1711 |
+
"learning_rate": 0.00011558558270833906,
|
| 1712 |
+
"loss": 0.5925,
|
| 1713 |
+
"step": 28400
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 1.73,
|
| 1717 |
+
"learning_rate": 0.00011465943547436524,
|
| 1718 |
+
"loss": 0.5885,
|
| 1719 |
+
"step": 28500
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 1.73,
|
| 1723 |
+
"learning_rate": 0.00011373471250117322,
|
| 1724 |
+
"loss": 0.5904,
|
| 1725 |
+
"step": 28600
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.74,
|
| 1729 |
+
"learning_rate": 0.0001128114510560416,
|
| 1730 |
+
"loss": 0.581,
|
| 1731 |
+
"step": 28700
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.75,
|
| 1735 |
+
"learning_rate": 0.00011188968834734798,
|
| 1736 |
+
"loss": 0.5822,
|
| 1737 |
+
"step": 28800
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 1.75,
|
| 1741 |
+
"learning_rate": 0.00011096946152306923,
|
| 1742 |
+
"loss": 0.5816,
|
| 1743 |
+
"step": 28900
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 1.76,
|
| 1747 |
+
"learning_rate": 0.00011005080766928467,
|
| 1748 |
+
"loss": 0.5848,
|
| 1749 |
+
"step": 29000
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 1.76,
|
| 1753 |
+
"learning_rate": 0.00010913376380868118,
|
| 1754 |
+
"loss": 0.5783,
|
| 1755 |
+
"step": 29100
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 1.77,
|
| 1759 |
+
"learning_rate": 0.00010821836689906128,
|
| 1760 |
+
"loss": 0.5801,
|
| 1761 |
+
"step": 29200
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 1.78,
|
| 1765 |
+
"learning_rate": 0.00010730465383185379,
|
| 1766 |
+
"loss": 0.5799,
|
| 1767 |
+
"step": 29300
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.78,
|
| 1771 |
+
"learning_rate": 0.00010639266143062683,
|
| 1772 |
+
"loss": 0.5884,
|
| 1773 |
+
"step": 29400
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 1.79,
|
| 1777 |
+
"learning_rate": 0.00010548242644960404,
|
| 1778 |
+
"loss": 0.5763,
|
| 1779 |
+
"step": 29500
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 1.79,
|
| 1783 |
+
"learning_rate": 0.00010457398557218315,
|
| 1784 |
+
"loss": 0.5836,
|
| 1785 |
+
"step": 29600
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 1.8,
|
| 1789 |
+
"learning_rate": 0.00010366737540945772,
|
| 1790 |
+
"loss": 0.5814,
|
| 1791 |
+
"step": 29700
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 1.81,
|
| 1795 |
+
"learning_rate": 0.00010276263249874166,
|
| 1796 |
+
"loss": 0.5764,
|
| 1797 |
+
"step": 29800
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 1.81,
|
| 1801 |
+
"learning_rate": 0.00010185979330209668,
|
| 1802 |
+
"loss": 0.5776,
|
| 1803 |
+
"step": 29900
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 1.82,
|
| 1807 |
+
"learning_rate": 0.00010095889420486292,
|
| 1808 |
+
"loss": 0.5788,
|
| 1809 |
+
"step": 30000
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 1.82,
|
| 1813 |
+
"learning_rate": 0.0001000599715141925,
|
| 1814 |
+
"loss": 0.5809,
|
| 1815 |
+
"step": 30100
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.83,
|
| 1819 |
+
"learning_rate": 9.916306145758637e-05,
|
| 1820 |
+
"loss": 0.5828,
|
| 1821 |
+
"step": 30200
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 1.84,
|
| 1825 |
+
"learning_rate": 9.826820018143417e-05,
|
| 1826 |
+
"loss": 0.5809,
|
| 1827 |
+
"step": 30300
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 1.84,
|
| 1831 |
+
"learning_rate": 9.737542374955779e-05,
|
| 1832 |
+
"loss": 0.5855,
|
| 1833 |
+
"step": 30400
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 1.85,
|
| 1837 |
+
"learning_rate": 9.648476814175755e-05,
|
| 1838 |
+
"loss": 0.5814,
|
| 1839 |
+
"step": 30500
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 1.85,
|
| 1843 |
+
"learning_rate": 9.559626925236263e-05,
|
| 1844 |
+
"loss": 0.5905,
|
| 1845 |
+
"step": 30600
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 1.86,
|
| 1849 |
+
"learning_rate": 9.470996288878409e-05,
|
| 1850 |
+
"loss": 0.5781,
|
| 1851 |
+
"step": 30700
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.87,
|
| 1855 |
+
"learning_rate": 9.382588477007196e-05,
|
| 1856 |
+
"loss": 0.5846,
|
| 1857 |
+
"step": 30800
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.87,
|
| 1861 |
+
"learning_rate": 9.294407052547586e-05,
|
| 1862 |
+
"loss": 0.5803,
|
| 1863 |
+
"step": 30900
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 1.88,
|
| 1867 |
+
"learning_rate": 9.206455569300888e-05,
|
| 1868 |
+
"loss": 0.5849,
|
| 1869 |
+
"step": 31000
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 1.88,
|
| 1873 |
+
"learning_rate": 9.118737571801549e-05,
|
| 1874 |
+
"loss": 0.5753,
|
| 1875 |
+
"step": 31100
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 1.89,
|
| 1879 |
+
"learning_rate": 9.031256595174299e-05,
|
| 1880 |
+
"loss": 0.5806,
|
| 1881 |
+
"step": 31200
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 1.9,
|
| 1885 |
+
"learning_rate": 8.944016164991682e-05,
|
| 1886 |
+
"loss": 0.5783,
|
| 1887 |
+
"step": 31300
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 1.9,
|
| 1891 |
+
"learning_rate": 8.857019797131991e-05,
|
| 1892 |
+
"loss": 0.5853,
|
| 1893 |
+
"step": 31400
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 1.91,
|
| 1897 |
+
"learning_rate": 8.77027099763754e-05,
|
| 1898 |
+
"loss": 0.5874,
|
| 1899 |
+
"step": 31500
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.92,
|
| 1903 |
+
"learning_rate": 8.683773262573396e-05,
|
| 1904 |
+
"loss": 0.5801,
|
| 1905 |
+
"step": 31600
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 1.92,
|
| 1909 |
+
"learning_rate": 8.597530077886474e-05,
|
| 1910 |
+
"loss": 0.5749,
|
| 1911 |
+
"step": 31700
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 1.93,
|
| 1915 |
+
"learning_rate": 8.511544919265039e-05,
|
| 1916 |
+
"loss": 0.5751,
|
| 1917 |
+
"step": 31800
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 1.93,
|
| 1921 |
+
"learning_rate": 8.425821251998646e-05,
|
| 1922 |
+
"loss": 0.585,
|
| 1923 |
+
"step": 31900
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 1.94,
|
| 1927 |
+
"learning_rate": 8.340362530838499e-05,
|
| 1928 |
+
"loss": 0.5758,
|
| 1929 |
+
"step": 32000
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 1.95,
|
| 1933 |
+
"learning_rate": 8.255172199858192e-05,
|
| 1934 |
+
"loss": 0.5839,
|
| 1935 |
+
"step": 32100
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 1.95,
|
| 1939 |
+
"learning_rate": 8.1702536923149e-05,
|
| 1940 |
+
"loss": 0.5723,
|
| 1941 |
+
"step": 32200
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 1.96,
|
| 1945 |
+
"learning_rate": 8.085610430511064e-05,
|
| 1946 |
+
"loss": 0.5811,
|
| 1947 |
+
"step": 32300
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 1.96,
|
| 1951 |
+
"learning_rate": 8.001245825656439e-05,
|
| 1952 |
+
"loss": 0.5797,
|
| 1953 |
+
"step": 32400
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 1.97,
|
| 1957 |
+
"learning_rate": 7.917163277730609e-05,
|
| 1958 |
+
"loss": 0.5683,
|
| 1959 |
+
"step": 32500
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 1.98,
|
| 1963 |
+
"learning_rate": 7.833366175345985e-05,
|
| 1964 |
+
"loss": 0.5804,
|
| 1965 |
+
"step": 32600
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 1.98,
|
| 1969 |
+
"learning_rate": 7.749857895611223e-05,
|
| 1970 |
+
"loss": 0.5676,
|
| 1971 |
+
"step": 32700
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 1.99,
|
| 1975 |
+
"learning_rate": 7.666641803995134e-05,
|
| 1976 |
+
"loss": 0.5692,
|
| 1977 |
+
"step": 32800
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 1.99,
|
| 1981 |
+
"learning_rate": 7.583721254191065e-05,
|
| 1982 |
+
"loss": 0.5822,
|
| 1983 |
+
"step": 32900
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 2.0,
|
| 1987 |
+
"learning_rate": 7.50109958798171e-05,
|
| 1988 |
+
"loss": 0.5727,
|
| 1989 |
+
"step": 33000
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 2.01,
|
| 1993 |
+
"learning_rate": 7.418780135104454e-05,
|
| 1994 |
+
"loss": 0.5524,
|
| 1995 |
+
"step": 33100
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 2.01,
|
| 1999 |
+
"learning_rate": 7.336766213117173e-05,
|
| 2000 |
+
"loss": 0.5589,
|
| 2001 |
+
"step": 33200
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 2.02,
|
| 2005 |
+
"learning_rate": 7.255061127264536e-05,
|
| 2006 |
+
"loss": 0.5523,
|
| 2007 |
+
"step": 33300
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 2.02,
|
| 2011 |
+
"learning_rate": 7.173668170344819e-05,
|
| 2012 |
+
"loss": 0.5576,
|
| 2013 |
+
"step": 33400
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 2.03,
|
| 2017 |
+
"learning_rate": 7.092590622577162e-05,
|
| 2018 |
+
"loss": 0.5579,
|
| 2019 |
+
"step": 33500
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 2.04,
|
| 2023 |
+
"learning_rate": 7.011831751469404e-05,
|
| 2024 |
+
"loss": 0.5557,
|
| 2025 |
+
"step": 33600
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 2.04,
|
| 2029 |
+
"learning_rate": 6.931394811686386e-05,
|
| 2030 |
+
"loss": 0.5579,
|
| 2031 |
+
"step": 33700
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 2.05,
|
| 2035 |
+
"learning_rate": 6.851283044918787e-05,
|
| 2036 |
+
"loss": 0.5561,
|
| 2037 |
+
"step": 33800
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 2.05,
|
| 2041 |
+
"learning_rate": 6.771499679752496e-05,
|
| 2042 |
+
"loss": 0.5587,
|
| 2043 |
+
"step": 33900
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 2.06,
|
| 2047 |
+
"learning_rate": 6.692047931538474e-05,
|
| 2048 |
+
"loss": 0.557,
|
| 2049 |
+
"step": 34000
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 2.07,
|
| 2053 |
+
"learning_rate": 6.612931002263158e-05,
|
| 2054 |
+
"loss": 0.5597,
|
| 2055 |
+
"step": 34100
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 2.07,
|
| 2059 |
+
"learning_rate": 6.534152080419484e-05,
|
| 2060 |
+
"loss": 0.5546,
|
| 2061 |
+
"step": 34200
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 2.08,
|
| 2065 |
+
"learning_rate": 6.455714340878308e-05,
|
| 2066 |
+
"loss": 0.5592,
|
| 2067 |
+
"step": 34300
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 2.08,
|
| 2071 |
+
"learning_rate": 6.377620944760513e-05,
|
| 2072 |
+
"loss": 0.5576,
|
| 2073 |
+
"step": 34400
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 2.09,
|
| 2077 |
+
"learning_rate": 6.299875039309576e-05,
|
| 2078 |
+
"loss": 0.562,
|
| 2079 |
+
"step": 34500
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 2.1,
|
| 2083 |
+
"learning_rate": 6.22247975776475e-05,
|
| 2084 |
+
"loss": 0.5596,
|
| 2085 |
+
"step": 34600
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 2.1,
|
| 2089 |
+
"learning_rate": 6.1454382192348e-05,
|
| 2090 |
+
"loss": 0.5537,
|
| 2091 |
+
"step": 34700
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 2.11,
|
| 2095 |
+
"learning_rate": 6.068753528572271e-05,
|
| 2096 |
+
"loss": 0.5592,
|
| 2097 |
+
"step": 34800
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 2.12,
|
| 2101 |
+
"learning_rate": 5.992428776248381e-05,
|
| 2102 |
+
"loss": 0.555,
|
| 2103 |
+
"step": 34900
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 2.12,
|
| 2107 |
+
"learning_rate": 5.9164670382284635e-05,
|
| 2108 |
+
"loss": 0.5572,
|
| 2109 |
+
"step": 35000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 2.13,
|
| 2113 |
+
"learning_rate": 5.840871375848003e-05,
|
| 2114 |
+
"loss": 0.5531,
|
| 2115 |
+
"step": 35100
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 2.13,
|
| 2119 |
+
"learning_rate": 5.7656448356892776e-05,
|
| 2120 |
+
"loss": 0.5583,
|
| 2121 |
+
"step": 35200
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 2.14,
|
| 2125 |
+
"learning_rate": 5.690790449458548e-05,
|
| 2126 |
+
"loss": 0.5529,
|
| 2127 |
+
"step": 35300
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 2.15,
|
| 2131 |
+
"learning_rate": 5.6163112338638965e-05,
|
| 2132 |
+
"loss": 0.5566,
|
| 2133 |
+
"step": 35400
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 2.15,
|
| 2137 |
+
"learning_rate": 5.5422101904936474e-05,
|
| 2138 |
+
"loss": 0.5539,
|
| 2139 |
+
"step": 35500
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 2.16,
|
| 2143 |
+
"learning_rate": 5.468490305695393e-05,
|
| 2144 |
+
"loss": 0.5507,
|
| 2145 |
+
"step": 35600
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 2.16,
|
| 2149 |
+
"learning_rate": 5.39515455045566e-05,
|
| 2150 |
+
"loss": 0.557,
|
| 2151 |
+
"step": 35700
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.17,
|
| 2155 |
+
"learning_rate": 5.32220588028015e-05,
|
| 2156 |
+
"loss": 0.5498,
|
| 2157 |
+
"step": 35800
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 2.18,
|
| 2161 |
+
"learning_rate": 5.249647235074647e-05,
|
| 2162 |
+
"loss": 0.5547,
|
| 2163 |
+
"step": 35900
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 2.18,
|
| 2167 |
+
"learning_rate": 5.177481539026529e-05,
|
| 2168 |
+
"loss": 0.5542,
|
| 2169 |
+
"step": 36000
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 2.19,
|
| 2173 |
+
"learning_rate": 5.105711700486922e-05,
|
| 2174 |
+
"loss": 0.5597,
|
| 2175 |
+
"step": 36100
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 2.19,
|
| 2179 |
+
"learning_rate": 5.0343406118534935e-05,
|
| 2180 |
+
"loss": 0.5583,
|
| 2181 |
+
"step": 36200
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 2.2,
|
| 2185 |
+
"learning_rate": 4.963371149453881e-05,
|
| 2186 |
+
"loss": 0.5565,
|
| 2187 |
+
"step": 36300
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 2.21,
|
| 2191 |
+
"learning_rate": 4.8928061734297765e-05,
|
| 2192 |
+
"loss": 0.5534,
|
| 2193 |
+
"step": 36400
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.21,
|
| 2197 |
+
"learning_rate": 4.822648527621653e-05,
|
| 2198 |
+
"loss": 0.5549,
|
| 2199 |
+
"step": 36500
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 2.22,
|
| 2203 |
+
"learning_rate": 4.7529010394541746e-05,
|
| 2204 |
+
"loss": 0.5552,
|
| 2205 |
+
"step": 36600
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 2.22,
|
| 2209 |
+
"learning_rate": 4.6835665198222224e-05,
|
| 2210 |
+
"loss": 0.5515,
|
| 2211 |
+
"step": 36700
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 2.23,
|
| 2215 |
+
"learning_rate": 4.614647762977626e-05,
|
| 2216 |
+
"loss": 0.5538,
|
| 2217 |
+
"step": 36800
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 2.24,
|
| 2221 |
+
"learning_rate": 4.5461475464165534e-05,
|
| 2222 |
+
"loss": 0.5562,
|
| 2223 |
+
"step": 36900
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 2.24,
|
| 2227 |
+
"learning_rate": 4.478068630767565e-05,
|
| 2228 |
+
"loss": 0.5553,
|
| 2229 |
+
"step": 37000
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 2.25,
|
| 2233 |
+
"learning_rate": 4.410413759680383e-05,
|
| 2234 |
+
"loss": 0.5539,
|
| 2235 |
+
"step": 37100
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 2.25,
|
| 2239 |
+
"learning_rate": 4.343185659715283e-05,
|
| 2240 |
+
"loss": 0.557,
|
| 2241 |
+
"step": 37200
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 2.26,
|
| 2245 |
+
"learning_rate": 4.276387040233236e-05,
|
| 2246 |
+
"loss": 0.5542,
|
| 2247 |
+
"step": 37300
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 2.27,
|
| 2251 |
+
"learning_rate": 4.210020593286711e-05,
|
| 2252 |
+
"loss": 0.5518,
|
| 2253 |
+
"step": 37400
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 2.27,
|
| 2257 |
+
"learning_rate": 4.144088993511181e-05,
|
| 2258 |
+
"loss": 0.5555,
|
| 2259 |
+
"step": 37500
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 2.28,
|
| 2263 |
+
"learning_rate": 4.078594898017346e-05,
|
| 2264 |
+
"loss": 0.5565,
|
| 2265 |
+
"step": 37600
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 2.28,
|
| 2269 |
+
"learning_rate": 4.013540946284024e-05,
|
| 2270 |
+
"loss": 0.5544,
|
| 2271 |
+
"step": 37700
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 2.29,
|
| 2275 |
+
"learning_rate": 3.948929760051797e-05,
|
| 2276 |
+
"loss": 0.5555,
|
| 2277 |
+
"step": 37800
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 2.3,
|
| 2281 |
+
"learning_rate": 3.8847639432173405e-05,
|
| 2282 |
+
"loss": 0.5563,
|
| 2283 |
+
"step": 37900
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 2.3,
|
| 2287 |
+
"learning_rate": 3.821046081728497e-05,
|
| 2288 |
+
"loss": 0.5525,
|
| 2289 |
+
"step": 38000
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 2.31,
|
| 2293 |
+
"learning_rate": 3.757778743480045e-05,
|
| 2294 |
+
"loss": 0.5541,
|
| 2295 |
+
"step": 38100
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 2.32,
|
| 2299 |
+
"learning_rate": 3.694964478210221e-05,
|
| 2300 |
+
"loss": 0.5564,
|
| 2301 |
+
"step": 38200
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 2.32,
|
| 2305 |
+
"learning_rate": 3.63260581739796e-05,
|
| 2306 |
+
"loss": 0.5551,
|
| 2307 |
+
"step": 38300
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 2.33,
|
| 2311 |
+
"learning_rate": 3.5707052741608636e-05,
|
| 2312 |
+
"loss": 0.5506,
|
| 2313 |
+
"step": 38400
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 2.33,
|
| 2317 |
+
"learning_rate": 3.5092653431539436e-05,
|
| 2318 |
+
"loss": 0.5483,
|
| 2319 |
+
"step": 38500
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 2.34,
|
| 2323 |
+
"learning_rate": 3.448288500469058e-05,
|
| 2324 |
+
"loss": 0.556,
|
| 2325 |
+
"step": 38600
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 2.35,
|
| 2329 |
+
"learning_rate": 3.3877772035351326e-05,
|
| 2330 |
+
"loss": 0.5509,
|
| 2331 |
+
"step": 38700
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 2.35,
|
| 2335 |
+
"learning_rate": 3.327733891019132e-05,
|
| 2336 |
+
"loss": 0.5574,
|
| 2337 |
+
"step": 38800
|
| 2338 |
+
},
|
| 2339 |
+
{
|
| 2340 |
+
"epoch": 2.36,
|
| 2341 |
+
"learning_rate": 3.268160982727759e-05,
|
| 2342 |
+
"loss": 0.5525,
|
| 2343 |
+
"step": 38900
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 2.36,
|
| 2347 |
+
"learning_rate": 3.209060879509968e-05,
|
| 2348 |
+
"loss": 0.5502,
|
| 2349 |
+
"step": 39000
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 2.37,
|
| 2353 |
+
"learning_rate": 3.150435963160168e-05,
|
| 2354 |
+
"loss": 0.5523,
|
| 2355 |
+
"step": 39100
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 2.38,
|
| 2359 |
+
"learning_rate": 3.0922885963222585e-05,
|
| 2360 |
+
"loss": 0.5483,
|
| 2361 |
+
"step": 39200
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.38,
|
| 2365 |
+
"learning_rate": 3.0346211223944077e-05,
|
| 2366 |
+
"loss": 0.547,
|
| 2367 |
+
"step": 39300
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 2.39,
|
| 2371 |
+
"learning_rate": 2.9774358654346046e-05,
|
| 2372 |
+
"loss": 0.55,
|
| 2373 |
+
"step": 39400
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 2.39,
|
| 2377 |
+
"learning_rate": 2.9207351300670178e-05,
|
| 2378 |
+
"loss": 0.549,
|
| 2379 |
+
"step": 39500
|
| 2380 |
+
},
|
| 2381 |
+
{
|
| 2382 |
+
"epoch": 2.4,
|
| 2383 |
+
"learning_rate": 2.864521201389085e-05,
|
| 2384 |
+
"loss": 0.5536,
|
| 2385 |
+
"step": 39600
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 2.41,
|
| 2389 |
+
"learning_rate": 2.8087963448794476e-05,
|
| 2390 |
+
"loss": 0.546,
|
| 2391 |
+
"step": 39700
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 2.41,
|
| 2395 |
+
"learning_rate": 2.7535628063066368e-05,
|
| 2396 |
+
"loss": 0.551,
|
| 2397 |
+
"step": 39800
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 2.42,
|
| 2401 |
+
"learning_rate": 2.698822811638569e-05,
|
| 2402 |
+
"loss": 0.5547,
|
| 2403 |
+
"step": 39900
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.42,
|
| 2407 |
+
"learning_rate": 2.6445785669528386e-05,
|
| 2408 |
+
"loss": 0.547,
|
| 2409 |
+
"step": 40000
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 2.43,
|
| 2413 |
+
"learning_rate": 2.590832258347814e-05,
|
| 2414 |
+
"loss": 0.5556,
|
| 2415 |
+
"step": 40100
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 2.44,
|
| 2419 |
+
"learning_rate": 2.537586051854522e-05,
|
| 2420 |
+
"loss": 0.5505,
|
| 2421 |
+
"step": 40200
|
| 2422 |
+
},
|
| 2423 |
+
{
|
| 2424 |
+
"epoch": 2.44,
|
| 2425 |
+
"learning_rate": 2.4848420933493824e-05,
|
| 2426 |
+
"loss": 0.547,
|
| 2427 |
+
"step": 40300
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 2.45,
|
| 2431 |
+
"learning_rate": 2.432602508467691e-05,
|
| 2432 |
+
"loss": 0.5483,
|
| 2433 |
+
"step": 40400
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 2.45,
|
| 2437 |
+
"learning_rate": 2.3808694025179804e-05,
|
| 2438 |
+
"loss": 0.5534,
|
| 2439 |
+
"step": 40500
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 2.46,
|
| 2443 |
+
"learning_rate": 2.3296448603971657e-05,
|
| 2444 |
+
"loss": 0.5479,
|
| 2445 |
+
"step": 40600
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 2.47,
|
| 2449 |
+
"learning_rate": 2.2789309465065154e-05,
|
| 2450 |
+
"loss": 0.5593,
|
| 2451 |
+
"step": 40700
|
| 2452 |
+
},
|
| 2453 |
+
{
|
| 2454 |
+
"epoch": 2.47,
|
| 2455 |
+
"learning_rate": 2.2287297046684737e-05,
|
| 2456 |
+
"loss": 0.5478,
|
| 2457 |
+
"step": 40800
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"epoch": 2.48,
|
| 2461 |
+
"learning_rate": 2.179043158044263e-05,
|
| 2462 |
+
"loss": 0.5562,
|
| 2463 |
+
"step": 40900
|
| 2464 |
+
},
|
| 2465 |
+
{
|
| 2466 |
+
"epoch": 2.48,
|
| 2467 |
+
"learning_rate": 2.1298733090523722e-05,
|
| 2468 |
+
"loss": 0.5513,
|
| 2469 |
+
"step": 41000
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 2.49,
|
| 2473 |
+
"learning_rate": 2.0812221392878463e-05,
|
| 2474 |
+
"loss": 0.5507,
|
| 2475 |
+
"step": 41100
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 2.5,
|
| 2479 |
+
"learning_rate": 2.0330916094424244e-05,
|
| 2480 |
+
"loss": 0.5465,
|
| 2481 |
+
"step": 41200
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 2.5,
|
| 2485 |
+
"learning_rate": 1.985483659225539e-05,
|
| 2486 |
+
"loss": 0.5504,
|
| 2487 |
+
"step": 41300
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 2.51,
|
| 2491 |
+
"learning_rate": 1.9384002072861186e-05,
|
| 2492 |
+
"loss": 0.5444,
|
| 2493 |
+
"step": 41400
|
| 2494 |
+
},
|
| 2495 |
+
{
|
| 2496 |
+
"epoch": 2.52,
|
| 2497 |
+
"learning_rate": 1.89184315113528e-05,
|
| 2498 |
+
"loss": 0.5514,
|
| 2499 |
+
"step": 41500
|
| 2500 |
+
},
|
| 2501 |
+
{
|
| 2502 |
+
"epoch": 2.52,
|
| 2503 |
+
"learning_rate": 1.8458143670698522e-05,
|
| 2504 |
+
"loss": 0.5448,
|
| 2505 |
+
"step": 41600
|
| 2506 |
+
},
|
| 2507 |
+
{
|
| 2508 |
+
"epoch": 2.53,
|
| 2509 |
+
"learning_rate": 1.80031571009676e-05,
|
| 2510 |
+
"loss": 0.5419,
|
| 2511 |
+
"step": 41700
|
| 2512 |
+
},
|
| 2513 |
+
{
|
| 2514 |
+
"epoch": 2.53,
|
| 2515 |
+
"learning_rate": 1.7553490138582786e-05,
|
| 2516 |
+
"loss": 0.5537,
|
| 2517 |
+
"step": 41800
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 2.54,
|
| 2521 |
+
"learning_rate": 1.7109160905580982e-05,
|
| 2522 |
+
"loss": 0.5493,
|
| 2523 |
+
"step": 41900
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 2.55,
|
| 2527 |
+
"learning_rate": 1.6670187308883364e-05,
|
| 2528 |
+
"loss": 0.5496,
|
| 2529 |
+
"step": 42000
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 2.55,
|
| 2533 |
+
"learning_rate": 1.6236587039573383e-05,
|
| 2534 |
+
"loss": 0.5491,
|
| 2535 |
+
"step": 42100
|
| 2536 |
+
},
|
| 2537 |
+
{
|
| 2538 |
+
"epoch": 2.56,
|
| 2539 |
+
"learning_rate": 1.5808377572184044e-05,
|
| 2540 |
+
"loss": 0.5502,
|
| 2541 |
+
"step": 42200
|
| 2542 |
+
},
|
| 2543 |
+
{
|
| 2544 |
+
"epoch": 2.56,
|
| 2545 |
+
"learning_rate": 1.5385576163993417e-05,
|
| 2546 |
+
"loss": 0.5539,
|
| 2547 |
+
"step": 42300
|
| 2548 |
+
},
|
| 2549 |
+
{
|
| 2550 |
+
"epoch": 2.57,
|
| 2551 |
+
"learning_rate": 1.4968199854329322e-05,
|
| 2552 |
+
"loss": 0.5473,
|
| 2553 |
+
"step": 42400
|
| 2554 |
+
},
|
| 2555 |
+
{
|
| 2556 |
+
"epoch": 2.58,
|
| 2557 |
+
"learning_rate": 1.4556265463882594e-05,
|
| 2558 |
+
"loss": 0.5523,
|
| 2559 |
+
"step": 42500
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 2.58,
|
| 2563 |
+
"learning_rate": 1.4149789594029093e-05,
|
| 2564 |
+
"loss": 0.554,
|
| 2565 |
+
"step": 42600
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 2.59,
|
| 2569 |
+
"learning_rate": 1.3748788626160878e-05,
|
| 2570 |
+
"loss": 0.5487,
|
| 2571 |
+
"step": 42700
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 2.59,
|
| 2575 |
+
"learning_rate": 1.3353278721025756e-05,
|
| 2576 |
+
"loss": 0.5558,
|
| 2577 |
+
"step": 42800
|
| 2578 |
+
},
|
| 2579 |
+
{
|
| 2580 |
+
"epoch": 2.6,
|
| 2581 |
+
"learning_rate": 1.2963275818076152e-05,
|
| 2582 |
+
"loss": 0.5501,
|
| 2583 |
+
"step": 42900
|
| 2584 |
+
},
|
| 2585 |
+
{
|
| 2586 |
+
"epoch": 2.61,
|
| 2587 |
+
"learning_rate": 1.2578795634826671e-05,
|
| 2588 |
+
"loss": 0.5561,
|
| 2589 |
+
"step": 43000
|
| 2590 |
+
},
|
| 2591 |
+
{
|
| 2592 |
+
"epoch": 2.61,
|
| 2593 |
+
"learning_rate": 1.2199853666220678e-05,
|
| 2594 |
+
"loss": 0.5486,
|
| 2595 |
+
"step": 43100
|
| 2596 |
+
},
|
| 2597 |
+
{
|
| 2598 |
+
"epoch": 2.62,
|
| 2599 |
+
"learning_rate": 1.18264651840059e-05,
|
| 2600 |
+
"loss": 0.5514,
|
| 2601 |
+
"step": 43200
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 2.62,
|
| 2605 |
+
"learning_rate": 1.145864523611884e-05,
|
| 2606 |
+
"loss": 0.5537,
|
| 2607 |
+
"step": 43300
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 2.63,
|
| 2611 |
+
"learning_rate": 1.1096408646078409e-05,
|
| 2612 |
+
"loss": 0.548,
|
| 2613 |
+
"step": 43400
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 2.64,
|
| 2617 |
+
"learning_rate": 1.073977001238851e-05,
|
| 2618 |
+
"loss": 0.5509,
|
| 2619 |
+
"step": 43500
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 2.64,
|
| 2623 |
+
"learning_rate": 1.0388743707949648e-05,
|
| 2624 |
+
"loss": 0.549,
|
| 2625 |
+
"step": 43600
|
| 2626 |
}
|
| 2627 |
],
|
| 2628 |
"logging_steps": 100,
|
|
|
|
| 2630 |
"num_input_tokens_seen": 0,
|
| 2631 |
"num_train_epochs": 3,
|
| 2632 |
"save_steps": 100,
|
| 2633 |
+
"total_flos": 3.4286253848387387e+19,
|
| 2634 |
"train_batch_size": 3,
|
| 2635 |
"trial_name": null,
|
| 2636 |
"trial_params": null
|