Training in progress, step 10000
Browse files- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state_0.pth +1 -1
- rng_state_1.pth +1 -1
- runs/Jun07_20-36-29_829f6f605e43/events.out.tfevents.1717792718.829f6f605e43.85.0 +2 -2
- scheduler.pt +1 -1
- trainer_state.json +1763 -4
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 44644496
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7ba74cf25c786c33ab51f49853610332c7c6fac2a629cf396b37bb2a5daf3218
|
| 3 |
size 44644496
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 11230198
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:307f93c818fe9e997251c586a4bbef0f98546328ebc4260c26d059dbb40e19bc
|
| 3 |
size 11230198
|
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
|
runs/Jun07_20-36-29_829f6f605e43/events.out.tfevents.1717792718.829f6f605e43.85.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:91a8415078fc8a593e4dafbb625bc6e914226237ff615cf5f31057d766f26011
|
| 3 |
+
size 110924
|
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 sha256:751d576a2eea45359933cc7817354c80776613355e39eb96453c641164eefa3f
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 5000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1766,6 +1766,1765 @@
|
|
| 1766 |
"eval_samples_per_second": 3517.479,
|
| 1767 |
"eval_steps_per_second": 3.437,
|
| 1768 |
"step": 5000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1769 |
}
|
| 1770 |
],
|
| 1771 |
"logging_steps": 20,
|
|
@@ -1779,13 +3538,13 @@
|
|
| 1779 |
"should_epoch_stop": false,
|
| 1780 |
"should_evaluate": false,
|
| 1781 |
"should_log": false,
|
| 1782 |
-
"should_save":
|
| 1783 |
"should_training_stop": false
|
| 1784 |
},
|
| 1785 |
"attributes": {}
|
| 1786 |
}
|
| 1787 |
},
|
| 1788 |
-
"total_flos":
|
| 1789 |
"train_batch_size": 512,
|
| 1790 |
"trial_name": null,
|
| 1791 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.1809790969143064,
|
| 5 |
"eval_steps": 5000,
|
| 6 |
+
"global_step": 10000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1766 |
"eval_samples_per_second": 3517.479,
|
| 1767 |
"eval_steps_per_second": 3.437,
|
| 1768 |
"step": 5000
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.09085150665098181,
|
| 1772 |
+
"grad_norm": 4.571134567260742,
|
| 1773 |
+
"learning_rate": 0.00027227483863183924,
|
| 1774 |
+
"loss": 8.0411,
|
| 1775 |
+
"step": 5020
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 0.09121346484481042,
|
| 1779 |
+
"grad_norm": 5.403674602508545,
|
| 1780 |
+
"learning_rate": 0.0002733606804608795,
|
| 1781 |
+
"loss": 8.0345,
|
| 1782 |
+
"step": 5040
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 0.09157542303863904,
|
| 1786 |
+
"grad_norm": 4.911813259124756,
|
| 1787 |
+
"learning_rate": 0.0002744465222899197,
|
| 1788 |
+
"loss": 8.0946,
|
| 1789 |
+
"step": 5060
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 0.09193738123246765,
|
| 1793 |
+
"grad_norm": 5.051859378814697,
|
| 1794 |
+
"learning_rate": 0.00027553236411895996,
|
| 1795 |
+
"loss": 8.0794,
|
| 1796 |
+
"step": 5080
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.09229933942629626,
|
| 1800 |
+
"grad_norm": 5.029412746429443,
|
| 1801 |
+
"learning_rate": 0.0002766182059480002,
|
| 1802 |
+
"loss": 8.0873,
|
| 1803 |
+
"step": 5100
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.09266129762012487,
|
| 1807 |
+
"grad_norm": 4.249423503875732,
|
| 1808 |
+
"learning_rate": 0.00027770404777704044,
|
| 1809 |
+
"loss": 8.098,
|
| 1810 |
+
"step": 5120
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 0.09302325581395349,
|
| 1814 |
+
"grad_norm": 4.305202960968018,
|
| 1815 |
+
"learning_rate": 0.0002787898896060807,
|
| 1816 |
+
"loss": 8.0734,
|
| 1817 |
+
"step": 5140
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 0.0933852140077821,
|
| 1821 |
+
"grad_norm": 5.8397345542907715,
|
| 1822 |
+
"learning_rate": 0.0002798757314351209,
|
| 1823 |
+
"loss": 8.12,
|
| 1824 |
+
"step": 5160
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 0.09374717220161072,
|
| 1828 |
+
"grad_norm": 4.833122253417969,
|
| 1829 |
+
"learning_rate": 0.00028096157326416117,
|
| 1830 |
+
"loss": 8.0778,
|
| 1831 |
+
"step": 5180
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 0.09410913039543933,
|
| 1835 |
+
"grad_norm": 3.883317232131958,
|
| 1836 |
+
"learning_rate": 0.0002820474150932014,
|
| 1837 |
+
"loss": 8.0997,
|
| 1838 |
+
"step": 5200
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 0.09447108858926793,
|
| 1842 |
+
"grad_norm": 4.111331939697266,
|
| 1843 |
+
"learning_rate": 0.00028313325692224165,
|
| 1844 |
+
"loss": 8.1039,
|
| 1845 |
+
"step": 5220
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 0.09483304678309655,
|
| 1849 |
+
"grad_norm": 3.988339424133301,
|
| 1850 |
+
"learning_rate": 0.0002842190987512819,
|
| 1851 |
+
"loss": 8.0941,
|
| 1852 |
+
"step": 5240
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 0.09519500497692517,
|
| 1856 |
+
"grad_norm": 3.9682798385620117,
|
| 1857 |
+
"learning_rate": 0.0002853049405803221,
|
| 1858 |
+
"loss": 8.0477,
|
| 1859 |
+
"step": 5260
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.09555696317075378,
|
| 1863 |
+
"grad_norm": 5.8604655265808105,
|
| 1864 |
+
"learning_rate": 0.0002863907824093623,
|
| 1865 |
+
"loss": 8.077,
|
| 1866 |
+
"step": 5280
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 0.0959189213645824,
|
| 1870 |
+
"grad_norm": 4.823431015014648,
|
| 1871 |
+
"learning_rate": 0.00028747662423840256,
|
| 1872 |
+
"loss": 8.0723,
|
| 1873 |
+
"step": 5300
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 0.096280879558411,
|
| 1877 |
+
"grad_norm": 8.868093490600586,
|
| 1878 |
+
"learning_rate": 0.0002885624660674428,
|
| 1879 |
+
"loss": 8.312,
|
| 1880 |
+
"step": 5320
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 0.09664283775223961,
|
| 1884 |
+
"grad_norm": 3.8635661602020264,
|
| 1885 |
+
"learning_rate": 0.00028964830789648304,
|
| 1886 |
+
"loss": 8.1751,
|
| 1887 |
+
"step": 5340
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 0.09700479594606823,
|
| 1891 |
+
"grad_norm": 4.368812084197998,
|
| 1892 |
+
"learning_rate": 0.0002907341497255233,
|
| 1893 |
+
"loss": 8.0755,
|
| 1894 |
+
"step": 5360
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 0.09736675413989684,
|
| 1898 |
+
"grad_norm": 4.956539154052734,
|
| 1899 |
+
"learning_rate": 0.0002918199915545635,
|
| 1900 |
+
"loss": 8.0912,
|
| 1901 |
+
"step": 5380
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 0.09772871233372546,
|
| 1905 |
+
"grad_norm": 4.059233665466309,
|
| 1906 |
+
"learning_rate": 0.00029290583338360376,
|
| 1907 |
+
"loss": 8.0815,
|
| 1908 |
+
"step": 5400
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 0.09809067052755406,
|
| 1912 |
+
"grad_norm": 4.747849464416504,
|
| 1913 |
+
"learning_rate": 0.000293991675212644,
|
| 1914 |
+
"loss": 8.0874,
|
| 1915 |
+
"step": 5420
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 0.09845262872138268,
|
| 1919 |
+
"grad_norm": 4.266099452972412,
|
| 1920 |
+
"learning_rate": 0.00029507751704168424,
|
| 1921 |
+
"loss": 8.0662,
|
| 1922 |
+
"step": 5440
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 0.09881458691521129,
|
| 1926 |
+
"grad_norm": 4.489388465881348,
|
| 1927 |
+
"learning_rate": 0.0002961633588707245,
|
| 1928 |
+
"loss": 8.0817,
|
| 1929 |
+
"step": 5460
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 0.09917654510903991,
|
| 1933 |
+
"grad_norm": 4.930331230163574,
|
| 1934 |
+
"learning_rate": 0.0002972492006997647,
|
| 1935 |
+
"loss": 8.0799,
|
| 1936 |
+
"step": 5480
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 0.09953850330286852,
|
| 1940 |
+
"grad_norm": 5.424180030822754,
|
| 1941 |
+
"learning_rate": 0.00029833504252880496,
|
| 1942 |
+
"loss": 8.065,
|
| 1943 |
+
"step": 5500
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 0.09990046149669714,
|
| 1947 |
+
"grad_norm": 4.461760520935059,
|
| 1948 |
+
"learning_rate": 0.0002994208843578452,
|
| 1949 |
+
"loss": 8.0738,
|
| 1950 |
+
"step": 5520
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 0.10026241969052574,
|
| 1954 |
+
"grad_norm": 3.7566046714782715,
|
| 1955 |
+
"learning_rate": 0.0003005067261868854,
|
| 1956 |
+
"loss": 8.0848,
|
| 1957 |
+
"step": 5540
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 0.10062437788435435,
|
| 1961 |
+
"grad_norm": 6.27786111831665,
|
| 1962 |
+
"learning_rate": 0.00030159256801592563,
|
| 1963 |
+
"loss": 8.088,
|
| 1964 |
+
"step": 5560
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 0.10098633607818297,
|
| 1968 |
+
"grad_norm": 4.437178134918213,
|
| 1969 |
+
"learning_rate": 0.00030267840984496587,
|
| 1970 |
+
"loss": 8.068,
|
| 1971 |
+
"step": 5580
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 0.10134829427201159,
|
| 1975 |
+
"grad_norm": 4.757803440093994,
|
| 1976 |
+
"learning_rate": 0.0003037642516740061,
|
| 1977 |
+
"loss": 8.0712,
|
| 1978 |
+
"step": 5600
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 0.1017102524658402,
|
| 1982 |
+
"grad_norm": 4.162949085235596,
|
| 1983 |
+
"learning_rate": 0.00030485009350304635,
|
| 1984 |
+
"loss": 8.0941,
|
| 1985 |
+
"step": 5620
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 0.1020722106596688,
|
| 1989 |
+
"grad_norm": 5.036799430847168,
|
| 1990 |
+
"learning_rate": 0.0003059359353320866,
|
| 1991 |
+
"loss": 8.1277,
|
| 1992 |
+
"step": 5640
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 0.10243416885349742,
|
| 1996 |
+
"grad_norm": 6.508123397827148,
|
| 1997 |
+
"learning_rate": 0.00030702177716112683,
|
| 1998 |
+
"loss": 8.1014,
|
| 1999 |
+
"step": 5660
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 0.10279612704732603,
|
| 2003 |
+
"grad_norm": 4.099546432495117,
|
| 2004 |
+
"learning_rate": 0.0003081076189901671,
|
| 2005 |
+
"loss": 8.1013,
|
| 2006 |
+
"step": 5680
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 0.10315808524115465,
|
| 2010 |
+
"grad_norm": 5.094971656799316,
|
| 2011 |
+
"learning_rate": 0.0003091934608192073,
|
| 2012 |
+
"loss": 8.0943,
|
| 2013 |
+
"step": 5700
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 0.10352004343498326,
|
| 2017 |
+
"grad_norm": 3.8019566535949707,
|
| 2018 |
+
"learning_rate": 0.00031027930264824756,
|
| 2019 |
+
"loss": 8.0894,
|
| 2020 |
+
"step": 5720
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 0.10388200162881188,
|
| 2024 |
+
"grad_norm": 5.287724494934082,
|
| 2025 |
+
"learning_rate": 0.0003113651444772878,
|
| 2026 |
+
"loss": 8.0937,
|
| 2027 |
+
"step": 5740
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 0.10424395982264048,
|
| 2031 |
+
"grad_norm": 4.729523658752441,
|
| 2032 |
+
"learning_rate": 0.00031245098630632804,
|
| 2033 |
+
"loss": 8.1002,
|
| 2034 |
+
"step": 5760
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 0.1046059180164691,
|
| 2038 |
+
"grad_norm": 4.7321319580078125,
|
| 2039 |
+
"learning_rate": 0.0003135368281353683,
|
| 2040 |
+
"loss": 8.0607,
|
| 2041 |
+
"step": 5780
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 0.10496787621029771,
|
| 2045 |
+
"grad_norm": 5.564339637756348,
|
| 2046 |
+
"learning_rate": 0.0003146226699644085,
|
| 2047 |
+
"loss": 8.1031,
|
| 2048 |
+
"step": 5800
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 0.10532983440412633,
|
| 2052 |
+
"grad_norm": 3.883352041244507,
|
| 2053 |
+
"learning_rate": 0.00031570851179344876,
|
| 2054 |
+
"loss": 8.0864,
|
| 2055 |
+
"step": 5820
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 0.10569179259795494,
|
| 2059 |
+
"grad_norm": 6.192608833312988,
|
| 2060 |
+
"learning_rate": 0.000316794353622489,
|
| 2061 |
+
"loss": 8.0805,
|
| 2062 |
+
"step": 5840
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 0.10605375079178354,
|
| 2066 |
+
"grad_norm": 3.4210565090179443,
|
| 2067 |
+
"learning_rate": 0.00031788019545152924,
|
| 2068 |
+
"loss": 8.104,
|
| 2069 |
+
"step": 5860
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 0.10641570898561216,
|
| 2073 |
+
"grad_norm": 5.195882797241211,
|
| 2074 |
+
"learning_rate": 0.0003189660372805695,
|
| 2075 |
+
"loss": 8.0621,
|
| 2076 |
+
"step": 5880
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 0.10677766717944077,
|
| 2080 |
+
"grad_norm": 4.863924980163574,
|
| 2081 |
+
"learning_rate": 0.0003200518791096097,
|
| 2082 |
+
"loss": 8.2223,
|
| 2083 |
+
"step": 5900
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 0.10713962537326939,
|
| 2087 |
+
"grad_norm": 6.122769355773926,
|
| 2088 |
+
"learning_rate": 0.00032113772093864996,
|
| 2089 |
+
"loss": 8.1023,
|
| 2090 |
+
"step": 5920
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 0.107501583567098,
|
| 2094 |
+
"grad_norm": 10.902450561523438,
|
| 2095 |
+
"learning_rate": 0.00032222356276769015,
|
| 2096 |
+
"loss": 8.1196,
|
| 2097 |
+
"step": 5940
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 0.1078635417609266,
|
| 2101 |
+
"grad_norm": 35.497840881347656,
|
| 2102 |
+
"learning_rate": 0.0003233094045967304,
|
| 2103 |
+
"loss": 8.1074,
|
| 2104 |
+
"step": 5960
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 0.10822549995475522,
|
| 2108 |
+
"grad_norm": 4.4585862159729,
|
| 2109 |
+
"learning_rate": 0.00032417807805996257,
|
| 2110 |
+
"loss": 8.9866,
|
| 2111 |
+
"step": 5980
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 0.10858745814858384,
|
| 2115 |
+
"grad_norm": 7.580636024475098,
|
| 2116 |
+
"learning_rate": 0.0003252639198890028,
|
| 2117 |
+
"loss": 9.1972,
|
| 2118 |
+
"step": 6000
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 0.10894941634241245,
|
| 2122 |
+
"grad_norm": 4.390789031982422,
|
| 2123 |
+
"learning_rate": 0.00032634976171804305,
|
| 2124 |
+
"loss": 10.38,
|
| 2125 |
+
"step": 6020
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 0.10931137453624107,
|
| 2129 |
+
"grad_norm": 3.9264190196990967,
|
| 2130 |
+
"learning_rate": 0.0003274356035470833,
|
| 2131 |
+
"loss": 9.9821,
|
| 2132 |
+
"step": 6040
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 0.10967333273006968,
|
| 2136 |
+
"grad_norm": 4.207955837249756,
|
| 2137 |
+
"learning_rate": 0.00032852144537612354,
|
| 2138 |
+
"loss": 9.7519,
|
| 2139 |
+
"step": 6060
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 0.11003529092389829,
|
| 2143 |
+
"grad_norm": 3.7314298152923584,
|
| 2144 |
+
"learning_rate": 0.0003292815346564517,
|
| 2145 |
+
"loss": 9.4857,
|
| 2146 |
+
"step": 6080
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 0.1103972491177269,
|
| 2150 |
+
"grad_norm": NaN,
|
| 2151 |
+
"learning_rate": 0.0003296072872051638,
|
| 2152 |
+
"loss": 8.6937,
|
| 2153 |
+
"step": 6100
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 0.11075920731155552,
|
| 2157 |
+
"grad_norm": NaN,
|
| 2158 |
+
"learning_rate": 0.0003296072872051638,
|
| 2159 |
+
"loss": 4.8707,
|
| 2160 |
+
"step": 6120
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 0.11112116550538413,
|
| 2164 |
+
"grad_norm": NaN,
|
| 2165 |
+
"learning_rate": 0.0003296072872051638,
|
| 2166 |
+
"loss": 4.4629,
|
| 2167 |
+
"step": 6140
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"epoch": 0.11148312369921275,
|
| 2171 |
+
"grad_norm": NaN,
|
| 2172 |
+
"learning_rate": 0.0003297158713880678,
|
| 2173 |
+
"loss": 4.7899,
|
| 2174 |
+
"step": 6160
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 0.11184508189304135,
|
| 2178 |
+
"grad_norm": NaN,
|
| 2179 |
+
"learning_rate": 0.0003297158713880678,
|
| 2180 |
+
"loss": 4.5487,
|
| 2181 |
+
"step": 6180
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 0.11220704008686996,
|
| 2185 |
+
"grad_norm": NaN,
|
| 2186 |
+
"learning_rate": 0.0003297158713880678,
|
| 2187 |
+
"loss": 0.3068,
|
| 2188 |
+
"step": 6200
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"epoch": 0.11256899828069858,
|
| 2192 |
+
"grad_norm": NaN,
|
| 2193 |
+
"learning_rate": 0.0003297701634795198,
|
| 2194 |
+
"loss": 3.6855,
|
| 2195 |
+
"step": 6220
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 0.1129309564745272,
|
| 2199 |
+
"grad_norm": NaN,
|
| 2200 |
+
"learning_rate": 0.0003297701634795198,
|
| 2201 |
+
"loss": 1.1705,
|
| 2202 |
+
"step": 6240
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 0.11329291466835581,
|
| 2206 |
+
"grad_norm": NaN,
|
| 2207 |
+
"learning_rate": 0.0003297701634795198,
|
| 2208 |
+
"loss": 4.775,
|
| 2209 |
+
"step": 6260
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 0.11365487286218441,
|
| 2213 |
+
"grad_norm": NaN,
|
| 2214 |
+
"learning_rate": 0.0003297701634795198,
|
| 2215 |
+
"loss": 1.8774,
|
| 2216 |
+
"step": 6280
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 0.11401683105601303,
|
| 2220 |
+
"grad_norm": NaN,
|
| 2221 |
+
"learning_rate": 0.0003297701634795198,
|
| 2222 |
+
"loss": 1.6138,
|
| 2223 |
+
"step": 6300
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 0.11437878924984164,
|
| 2227 |
+
"grad_norm": NaN,
|
| 2228 |
+
"learning_rate": 0.0003297701634795198,
|
| 2229 |
+
"loss": 4.092,
|
| 2230 |
+
"step": 6320
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 0.11474074744367026,
|
| 2234 |
+
"grad_norm": NaN,
|
| 2235 |
+
"learning_rate": 0.0003297701634795198,
|
| 2236 |
+
"loss": 1.877,
|
| 2237 |
+
"step": 6340
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 0.11510270563749887,
|
| 2241 |
+
"grad_norm": NaN,
|
| 2242 |
+
"learning_rate": 0.0003297701634795198,
|
| 2243 |
+
"loss": 0.0,
|
| 2244 |
+
"step": 6360
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 0.11546466383132749,
|
| 2248 |
+
"grad_norm": NaN,
|
| 2249 |
+
"learning_rate": 0.0003297701634795198,
|
| 2250 |
+
"loss": 1.4496,
|
| 2251 |
+
"step": 6380
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 0.11582662202515609,
|
| 2255 |
+
"grad_norm": NaN,
|
| 2256 |
+
"learning_rate": 0.0003297701634795198,
|
| 2257 |
+
"loss": 0.8283,
|
| 2258 |
+
"step": 6400
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 0.1161885802189847,
|
| 2262 |
+
"grad_norm": NaN,
|
| 2263 |
+
"learning_rate": 0.0003297701634795198,
|
| 2264 |
+
"loss": 1.0648,
|
| 2265 |
+
"step": 6420
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 0.11655053841281332,
|
| 2269 |
+
"grad_norm": NaN,
|
| 2270 |
+
"learning_rate": 0.0003297701634795198,
|
| 2271 |
+
"loss": 1.5593,
|
| 2272 |
+
"step": 6440
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 0.11691249660664194,
|
| 2276 |
+
"grad_norm": NaN,
|
| 2277 |
+
"learning_rate": 0.0003297701634795198,
|
| 2278 |
+
"loss": 0.3318,
|
| 2279 |
+
"step": 6460
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 0.11727445480047055,
|
| 2283 |
+
"grad_norm": NaN,
|
| 2284 |
+
"learning_rate": 0.0003297701634795198,
|
| 2285 |
+
"loss": 0.2933,
|
| 2286 |
+
"step": 6480
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 0.11763641299429915,
|
| 2290 |
+
"grad_norm": NaN,
|
| 2291 |
+
"learning_rate": 0.0003297701634795198,
|
| 2292 |
+
"loss": 0.8112,
|
| 2293 |
+
"step": 6500
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 0.11799837118812777,
|
| 2297 |
+
"grad_norm": NaN,
|
| 2298 |
+
"learning_rate": 0.0003297701634795198,
|
| 2299 |
+
"loss": 0.9949,
|
| 2300 |
+
"step": 6520
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 0.11836032938195638,
|
| 2304 |
+
"grad_norm": NaN,
|
| 2305 |
+
"learning_rate": 0.0003297701634795198,
|
| 2306 |
+
"loss": 1.0499,
|
| 2307 |
+
"step": 6540
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 0.118722287575785,
|
| 2311 |
+
"grad_norm": NaN,
|
| 2312 |
+
"learning_rate": 0.0003297701634795198,
|
| 2313 |
+
"loss": 1.1902,
|
| 2314 |
+
"step": 6560
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 0.11908424576961361,
|
| 2318 |
+
"grad_norm": NaN,
|
| 2319 |
+
"learning_rate": 0.0003297701634795198,
|
| 2320 |
+
"loss": 4.6519,
|
| 2321 |
+
"step": 6580
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 0.11944620396344222,
|
| 2325 |
+
"grad_norm": NaN,
|
| 2326 |
+
"learning_rate": 0.0003297701634795198,
|
| 2327 |
+
"loss": 3.6229,
|
| 2328 |
+
"step": 6600
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 0.11980816215727083,
|
| 2332 |
+
"grad_norm": NaN,
|
| 2333 |
+
"learning_rate": 0.0003297701634795198,
|
| 2334 |
+
"loss": 0.0,
|
| 2335 |
+
"step": 6620
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 0.12017012035109945,
|
| 2339 |
+
"grad_norm": NaN,
|
| 2340 |
+
"learning_rate": 0.00032982445557097183,
|
| 2341 |
+
"loss": 1.116,
|
| 2342 |
+
"step": 6640
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 0.12053207854492806,
|
| 2346 |
+
"grad_norm": NaN,
|
| 2347 |
+
"learning_rate": 0.00032987874766242386,
|
| 2348 |
+
"loss": 2.6337,
|
| 2349 |
+
"step": 6660
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 0.12089403673875668,
|
| 2353 |
+
"grad_norm": NaN,
|
| 2354 |
+
"learning_rate": 0.00032987874766242386,
|
| 2355 |
+
"loss": 2.4189,
|
| 2356 |
+
"step": 6680
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 0.12125599493258529,
|
| 2360 |
+
"grad_norm": NaN,
|
| 2361 |
+
"learning_rate": 0.00032987874766242386,
|
| 2362 |
+
"loss": 0.0,
|
| 2363 |
+
"step": 6700
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 0.1216179531264139,
|
| 2367 |
+
"grad_norm": NaN,
|
| 2368 |
+
"learning_rate": 0.00032993303975387584,
|
| 2369 |
+
"loss": 5.334,
|
| 2370 |
+
"step": 6720
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 0.12197991132024251,
|
| 2374 |
+
"grad_norm": NaN,
|
| 2375 |
+
"learning_rate": 0.00032993303975387584,
|
| 2376 |
+
"loss": 1.2443,
|
| 2377 |
+
"step": 6740
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 0.12234186951407112,
|
| 2381 |
+
"grad_norm": NaN,
|
| 2382 |
+
"learning_rate": 0.00032993303975387584,
|
| 2383 |
+
"loss": 4.497,
|
| 2384 |
+
"step": 6760
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 0.12270382770789974,
|
| 2388 |
+
"grad_norm": NaN,
|
| 2389 |
+
"learning_rate": 0.00032993303975387584,
|
| 2390 |
+
"loss": 1.03,
|
| 2391 |
+
"step": 6780
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 0.12306578590172836,
|
| 2395 |
+
"grad_norm": NaN,
|
| 2396 |
+
"learning_rate": 0.00032993303975387584,
|
| 2397 |
+
"loss": 0.5876,
|
| 2398 |
+
"step": 6800
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 0.12342774409555696,
|
| 2402 |
+
"grad_norm": NaN,
|
| 2403 |
+
"learning_rate": 0.00032993303975387584,
|
| 2404 |
+
"loss": 3.1744,
|
| 2405 |
+
"step": 6820
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 0.12378970228938557,
|
| 2409 |
+
"grad_norm": NaN,
|
| 2410 |
+
"learning_rate": 0.00032993303975387584,
|
| 2411 |
+
"loss": 1.7883,
|
| 2412 |
+
"step": 6840
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 0.12415166048321419,
|
| 2416 |
+
"grad_norm": NaN,
|
| 2417 |
+
"learning_rate": 0.00032993303975387584,
|
| 2418 |
+
"loss": 0.6097,
|
| 2419 |
+
"step": 6860
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 0.1245136186770428,
|
| 2423 |
+
"grad_norm": NaN,
|
| 2424 |
+
"learning_rate": 0.00032993303975387584,
|
| 2425 |
+
"loss": 2.116,
|
| 2426 |
+
"step": 6880
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 0.12487557687087142,
|
| 2430 |
+
"grad_norm": NaN,
|
| 2431 |
+
"learning_rate": 0.00032993303975387584,
|
| 2432 |
+
"loss": 1.1183,
|
| 2433 |
+
"step": 6900
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 0.12523753506470003,
|
| 2437 |
+
"grad_norm": NaN,
|
| 2438 |
+
"learning_rate": 0.00032998733184532787,
|
| 2439 |
+
"loss": 2.0884,
|
| 2440 |
+
"step": 6920
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 0.12559949325852865,
|
| 2444 |
+
"grad_norm": NaN,
|
| 2445 |
+
"learning_rate": 0.00032998733184532787,
|
| 2446 |
+
"loss": 0.0,
|
| 2447 |
+
"step": 6940
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 0.12596145145235726,
|
| 2451 |
+
"grad_norm": NaN,
|
| 2452 |
+
"learning_rate": 0.00032998733184532787,
|
| 2453 |
+
"loss": 3.2805,
|
| 2454 |
+
"step": 6960
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 0.12632340964618585,
|
| 2458 |
+
"grad_norm": NaN,
|
| 2459 |
+
"learning_rate": 0.0003300416239367799,
|
| 2460 |
+
"loss": 2.9308,
|
| 2461 |
+
"step": 6980
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 0.12668536784001447,
|
| 2465 |
+
"grad_norm": NaN,
|
| 2466 |
+
"learning_rate": 0.0003300416239367799,
|
| 2467 |
+
"loss": 0.2905,
|
| 2468 |
+
"step": 7000
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 0.12704732603384308,
|
| 2472 |
+
"grad_norm": NaN,
|
| 2473 |
+
"learning_rate": 0.0003300416239367799,
|
| 2474 |
+
"loss": 1.5342,
|
| 2475 |
+
"step": 7020
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 0.1274092842276717,
|
| 2479 |
+
"grad_norm": NaN,
|
| 2480 |
+
"learning_rate": 0.0003300416239367799,
|
| 2481 |
+
"loss": 2.1439,
|
| 2482 |
+
"step": 7040
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 0.12777124242150031,
|
| 2486 |
+
"grad_norm": NaN,
|
| 2487 |
+
"learning_rate": 0.0003300416239367799,
|
| 2488 |
+
"loss": 1.8486,
|
| 2489 |
+
"step": 7060
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 0.12813320061532893,
|
| 2493 |
+
"grad_norm": NaN,
|
| 2494 |
+
"learning_rate": 0.0003300416239367799,
|
| 2495 |
+
"loss": 2.2983,
|
| 2496 |
+
"step": 7080
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 0.12849515880915754,
|
| 2500 |
+
"grad_norm": NaN,
|
| 2501 |
+
"learning_rate": 0.00033009591602823187,
|
| 2502 |
+
"loss": 5.8454,
|
| 2503 |
+
"step": 7100
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 0.12885711700298616,
|
| 2507 |
+
"grad_norm": NaN,
|
| 2508 |
+
"learning_rate": 0.00033009591602823187,
|
| 2509 |
+
"loss": 0.0,
|
| 2510 |
+
"step": 7120
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 0.12921907519681478,
|
| 2514 |
+
"grad_norm": NaN,
|
| 2515 |
+
"learning_rate": 0.00033009591602823187,
|
| 2516 |
+
"loss": 3.4919,
|
| 2517 |
+
"step": 7140
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 0.1295810333906434,
|
| 2521 |
+
"grad_norm": NaN,
|
| 2522 |
+
"learning_rate": 0.00033009591602823187,
|
| 2523 |
+
"loss": 0.0,
|
| 2524 |
+
"step": 7160
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 0.129942991584472,
|
| 2528 |
+
"grad_norm": NaN,
|
| 2529 |
+
"learning_rate": 0.0003301502081196839,
|
| 2530 |
+
"loss": 1.6828,
|
| 2531 |
+
"step": 7180
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 0.1303049497783006,
|
| 2535 |
+
"grad_norm": NaN,
|
| 2536 |
+
"learning_rate": 0.0003302045002111359,
|
| 2537 |
+
"loss": 2.8762,
|
| 2538 |
+
"step": 7200
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 0.1306669079721292,
|
| 2542 |
+
"grad_norm": NaN,
|
| 2543 |
+
"learning_rate": 0.0003302045002111359,
|
| 2544 |
+
"loss": 0.995,
|
| 2545 |
+
"step": 7220
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 0.13102886616595782,
|
| 2549 |
+
"grad_norm": NaN,
|
| 2550 |
+
"learning_rate": 0.0003302045002111359,
|
| 2551 |
+
"loss": 3.0309,
|
| 2552 |
+
"step": 7240
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 0.13139082435978644,
|
| 2556 |
+
"grad_norm": NaN,
|
| 2557 |
+
"learning_rate": 0.0003302045002111359,
|
| 2558 |
+
"loss": 2.1536,
|
| 2559 |
+
"step": 7260
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 0.13175278255361506,
|
| 2563 |
+
"grad_norm": NaN,
|
| 2564 |
+
"learning_rate": 0.0003302045002111359,
|
| 2565 |
+
"loss": 1.7328,
|
| 2566 |
+
"step": 7280
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 0.13211474074744367,
|
| 2570 |
+
"grad_norm": NaN,
|
| 2571 |
+
"learning_rate": 0.0003302045002111359,
|
| 2572 |
+
"loss": 1.4516,
|
| 2573 |
+
"step": 7300
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 0.1324766989412723,
|
| 2577 |
+
"grad_norm": NaN,
|
| 2578 |
+
"learning_rate": 0.0003302045002111359,
|
| 2579 |
+
"loss": 2.3328,
|
| 2580 |
+
"step": 7320
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 0.1328386571351009,
|
| 2584 |
+
"grad_norm": NaN,
|
| 2585 |
+
"learning_rate": 0.0003302045002111359,
|
| 2586 |
+
"loss": 5.0458,
|
| 2587 |
+
"step": 7340
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 0.13320061532892952,
|
| 2591 |
+
"grad_norm": NaN,
|
| 2592 |
+
"learning_rate": 0.0003302045002111359,
|
| 2593 |
+
"loss": 0.6658,
|
| 2594 |
+
"step": 7360
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 0.13356257352275813,
|
| 2598 |
+
"grad_norm": NaN,
|
| 2599 |
+
"learning_rate": 0.0003302587923025879,
|
| 2600 |
+
"loss": 3.0189,
|
| 2601 |
+
"step": 7380
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 0.13392453171658675,
|
| 2605 |
+
"grad_norm": NaN,
|
| 2606 |
+
"learning_rate": 0.0003302587923025879,
|
| 2607 |
+
"loss": 1.0563,
|
| 2608 |
+
"step": 7400
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 0.13428648991041534,
|
| 2612 |
+
"grad_norm": NaN,
|
| 2613 |
+
"learning_rate": 0.0003302587923025879,
|
| 2614 |
+
"loss": 0.2771,
|
| 2615 |
+
"step": 7420
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 0.13464844810424395,
|
| 2619 |
+
"grad_norm": NaN,
|
| 2620 |
+
"learning_rate": 0.0003302587923025879,
|
| 2621 |
+
"loss": 2.2785,
|
| 2622 |
+
"step": 7440
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 0.13501040629807257,
|
| 2626 |
+
"grad_norm": NaN,
|
| 2627 |
+
"learning_rate": 0.0003302587923025879,
|
| 2628 |
+
"loss": 3.2136,
|
| 2629 |
+
"step": 7460
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 0.13537236449190118,
|
| 2633 |
+
"grad_norm": NaN,
|
| 2634 |
+
"learning_rate": 0.0003302587923025879,
|
| 2635 |
+
"loss": 0.4388,
|
| 2636 |
+
"step": 7480
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 0.1357343226857298,
|
| 2640 |
+
"grad_norm": NaN,
|
| 2641 |
+
"learning_rate": 0.0003302587923025879,
|
| 2642 |
+
"loss": 1.5316,
|
| 2643 |
+
"step": 7500
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 0.1360962808795584,
|
| 2647 |
+
"grad_norm": NaN,
|
| 2648 |
+
"learning_rate": 0.0003302587923025879,
|
| 2649 |
+
"loss": 4.91,
|
| 2650 |
+
"step": 7520
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 0.13645823907338703,
|
| 2654 |
+
"grad_norm": NaN,
|
| 2655 |
+
"learning_rate": 0.0003302587923025879,
|
| 2656 |
+
"loss": 0.6,
|
| 2657 |
+
"step": 7540
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 0.13682019726721564,
|
| 2661 |
+
"grad_norm": NaN,
|
| 2662 |
+
"learning_rate": 0.0003302587923025879,
|
| 2663 |
+
"loss": 1.3436,
|
| 2664 |
+
"step": 7560
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 0.13718215546104426,
|
| 2668 |
+
"grad_norm": NaN,
|
| 2669 |
+
"learning_rate": 0.0003302587923025879,
|
| 2670 |
+
"loss": 0.971,
|
| 2671 |
+
"step": 7580
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 0.13754411365487287,
|
| 2675 |
+
"grad_norm": NaN,
|
| 2676 |
+
"learning_rate": 0.0003302587923025879,
|
| 2677 |
+
"loss": 0.9318,
|
| 2678 |
+
"step": 7600
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 0.13790607184870146,
|
| 2682 |
+
"grad_norm": NaN,
|
| 2683 |
+
"learning_rate": 0.0003302587923025879,
|
| 2684 |
+
"loss": 1.9039,
|
| 2685 |
+
"step": 7620
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 0.13826803004253008,
|
| 2689 |
+
"grad_norm": NaN,
|
| 2690 |
+
"learning_rate": 0.0003302587923025879,
|
| 2691 |
+
"loss": 1.7444,
|
| 2692 |
+
"step": 7640
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 0.1386299882363587,
|
| 2696 |
+
"grad_norm": NaN,
|
| 2697 |
+
"learning_rate": 0.0003302587923025879,
|
| 2698 |
+
"loss": 1.1124,
|
| 2699 |
+
"step": 7660
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 0.1389919464301873,
|
| 2703 |
+
"grad_norm": NaN,
|
| 2704 |
+
"learning_rate": 0.0003302587923025879,
|
| 2705 |
+
"loss": 2.8302,
|
| 2706 |
+
"step": 7680
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 0.13935390462401592,
|
| 2710 |
+
"grad_norm": NaN,
|
| 2711 |
+
"learning_rate": 0.0003302587923025879,
|
| 2712 |
+
"loss": 4.0674,
|
| 2713 |
+
"step": 7700
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 0.13971586281784454,
|
| 2717 |
+
"grad_norm": NaN,
|
| 2718 |
+
"learning_rate": 0.0003302587923025879,
|
| 2719 |
+
"loss": 0.275,
|
| 2720 |
+
"step": 7720
|
| 2721 |
+
},
|
| 2722 |
+
{
|
| 2723 |
+
"epoch": 0.14007782101167315,
|
| 2724 |
+
"grad_norm": NaN,
|
| 2725 |
+
"learning_rate": 0.0003302587923025879,
|
| 2726 |
+
"loss": 2.7992,
|
| 2727 |
+
"step": 7740
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 0.14043977920550177,
|
| 2731 |
+
"grad_norm": NaN,
|
| 2732 |
+
"learning_rate": 0.0003302587923025879,
|
| 2733 |
+
"loss": 3.474,
|
| 2734 |
+
"step": 7760
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 0.14080173739933038,
|
| 2738 |
+
"grad_norm": NaN,
|
| 2739 |
+
"learning_rate": 0.0003302587923025879,
|
| 2740 |
+
"loss": 2.9148,
|
| 2741 |
+
"step": 7780
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 0.141163695593159,
|
| 2745 |
+
"grad_norm": NaN,
|
| 2746 |
+
"learning_rate": 0.0003302587923025879,
|
| 2747 |
+
"loss": 0.2909,
|
| 2748 |
+
"step": 7800
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 0.14152565378698762,
|
| 2752 |
+
"grad_norm": NaN,
|
| 2753 |
+
"learning_rate": 0.0003302587923025879,
|
| 2754 |
+
"loss": 2.1936,
|
| 2755 |
+
"step": 7820
|
| 2756 |
+
},
|
| 2757 |
+
{
|
| 2758 |
+
"epoch": 0.1418876119808162,
|
| 2759 |
+
"grad_norm": NaN,
|
| 2760 |
+
"learning_rate": 0.0003302587923025879,
|
| 2761 |
+
"loss": 4.1645,
|
| 2762 |
+
"step": 7840
|
| 2763 |
+
},
|
| 2764 |
+
{
|
| 2765 |
+
"epoch": 0.14224957017464482,
|
| 2766 |
+
"grad_norm": NaN,
|
| 2767 |
+
"learning_rate": 0.0003303130843940399,
|
| 2768 |
+
"loss": 4.7352,
|
| 2769 |
+
"step": 7860
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 0.14261152836847343,
|
| 2773 |
+
"grad_norm": NaN,
|
| 2774 |
+
"learning_rate": 0.0003303130843940399,
|
| 2775 |
+
"loss": 0.9052,
|
| 2776 |
+
"step": 7880
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 0.14297348656230205,
|
| 2780 |
+
"grad_norm": NaN,
|
| 2781 |
+
"learning_rate": 0.0003303130843940399,
|
| 2782 |
+
"loss": 3.5735,
|
| 2783 |
+
"step": 7900
|
| 2784 |
+
},
|
| 2785 |
+
{
|
| 2786 |
+
"epoch": 0.14333544475613066,
|
| 2787 |
+
"grad_norm": NaN,
|
| 2788 |
+
"learning_rate": 0.0003303130843940399,
|
| 2789 |
+
"loss": 0.4788,
|
| 2790 |
+
"step": 7920
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 0.14369740294995928,
|
| 2794 |
+
"grad_norm": NaN,
|
| 2795 |
+
"learning_rate": 0.00033036737648549196,
|
| 2796 |
+
"loss": 2.5942,
|
| 2797 |
+
"step": 7940
|
| 2798 |
+
},
|
| 2799 |
+
{
|
| 2800 |
+
"epoch": 0.1440593611437879,
|
| 2801 |
+
"grad_norm": NaN,
|
| 2802 |
+
"learning_rate": 0.00033036737648549196,
|
| 2803 |
+
"loss": 0.5562,
|
| 2804 |
+
"step": 7960
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 0.1444213193376165,
|
| 2808 |
+
"grad_norm": NaN,
|
| 2809 |
+
"learning_rate": 0.00033036737648549196,
|
| 2810 |
+
"loss": 1.268,
|
| 2811 |
+
"step": 7980
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 0.14478327753144513,
|
| 2815 |
+
"grad_norm": NaN,
|
| 2816 |
+
"learning_rate": 0.00033036737648549196,
|
| 2817 |
+
"loss": 2.5866,
|
| 2818 |
+
"step": 8000
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 0.14514523572527374,
|
| 2822 |
+
"grad_norm": NaN,
|
| 2823 |
+
"learning_rate": 0.00033036737648549196,
|
| 2824 |
+
"loss": 0.0,
|
| 2825 |
+
"step": 8020
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 0.14550719391910236,
|
| 2829 |
+
"grad_norm": NaN,
|
| 2830 |
+
"learning_rate": 0.00033036737648549196,
|
| 2831 |
+
"loss": 1.2675,
|
| 2832 |
+
"step": 8040
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 0.14586915211293094,
|
| 2836 |
+
"grad_norm": NaN,
|
| 2837 |
+
"learning_rate": 0.00033036737648549196,
|
| 2838 |
+
"loss": 0.407,
|
| 2839 |
+
"step": 8060
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 0.14623111030675956,
|
| 2843 |
+
"grad_norm": NaN,
|
| 2844 |
+
"learning_rate": 0.00033036737648549196,
|
| 2845 |
+
"loss": 0.0,
|
| 2846 |
+
"step": 8080
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 0.14659306850058818,
|
| 2850 |
+
"grad_norm": NaN,
|
| 2851 |
+
"learning_rate": 0.00033036737648549196,
|
| 2852 |
+
"loss": 1.6142,
|
| 2853 |
+
"step": 8100
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 0.1469550266944168,
|
| 2857 |
+
"grad_norm": NaN,
|
| 2858 |
+
"learning_rate": 0.00033036737648549196,
|
| 2859 |
+
"loss": 0.6237,
|
| 2860 |
+
"step": 8120
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 0.1473169848882454,
|
| 2864 |
+
"grad_norm": NaN,
|
| 2865 |
+
"learning_rate": 0.00033036737648549196,
|
| 2866 |
+
"loss": 1.5398,
|
| 2867 |
+
"step": 8140
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 0.14767894308207402,
|
| 2871 |
+
"grad_norm": NaN,
|
| 2872 |
+
"learning_rate": 0.00033036737648549196,
|
| 2873 |
+
"loss": 0.4752,
|
| 2874 |
+
"step": 8160
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 0.14804090127590264,
|
| 2878 |
+
"grad_norm": NaN,
|
| 2879 |
+
"learning_rate": 0.00033036737648549196,
|
| 2880 |
+
"loss": 2.9843,
|
| 2881 |
+
"step": 8180
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 0.14840285946973125,
|
| 2885 |
+
"grad_norm": NaN,
|
| 2886 |
+
"learning_rate": 0.00033036737648549196,
|
| 2887 |
+
"loss": 1.2889,
|
| 2888 |
+
"step": 8200
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 0.14876481766355987,
|
| 2892 |
+
"grad_norm": NaN,
|
| 2893 |
+
"learning_rate": 0.00033036737648549196,
|
| 2894 |
+
"loss": 0.0,
|
| 2895 |
+
"step": 8220
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"epoch": 0.14912677585738848,
|
| 2899 |
+
"grad_norm": NaN,
|
| 2900 |
+
"learning_rate": 0.00033036737648549196,
|
| 2901 |
+
"loss": 1.5344,
|
| 2902 |
+
"step": 8240
|
| 2903 |
+
},
|
| 2904 |
+
{
|
| 2905 |
+
"epoch": 0.1494887340512171,
|
| 2906 |
+
"grad_norm": NaN,
|
| 2907 |
+
"learning_rate": 0.00033036737648549196,
|
| 2908 |
+
"loss": 4.309,
|
| 2909 |
+
"step": 8260
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 0.14985069224504569,
|
| 2913 |
+
"grad_norm": NaN,
|
| 2914 |
+
"learning_rate": 0.00033036737648549196,
|
| 2915 |
+
"loss": 4.7959,
|
| 2916 |
+
"step": 8280
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 0.1502126504388743,
|
| 2920 |
+
"grad_norm": NaN,
|
| 2921 |
+
"learning_rate": 0.00033036737648549196,
|
| 2922 |
+
"loss": 0.2643,
|
| 2923 |
+
"step": 8300
|
| 2924 |
+
},
|
| 2925 |
+
{
|
| 2926 |
+
"epoch": 0.15057460863270292,
|
| 2927 |
+
"grad_norm": NaN,
|
| 2928 |
+
"learning_rate": 0.00033036737648549196,
|
| 2929 |
+
"loss": 1.5931,
|
| 2930 |
+
"step": 8320
|
| 2931 |
+
},
|
| 2932 |
+
{
|
| 2933 |
+
"epoch": 0.15093656682653153,
|
| 2934 |
+
"grad_norm": NaN,
|
| 2935 |
+
"learning_rate": 0.00033036737648549196,
|
| 2936 |
+
"loss": 3.252,
|
| 2937 |
+
"step": 8340
|
| 2938 |
+
},
|
| 2939 |
+
{
|
| 2940 |
+
"epoch": 0.15129852502036015,
|
| 2941 |
+
"grad_norm": NaN,
|
| 2942 |
+
"learning_rate": 0.00033036737648549196,
|
| 2943 |
+
"loss": 0.4066,
|
| 2944 |
+
"step": 8360
|
| 2945 |
+
},
|
| 2946 |
+
{
|
| 2947 |
+
"epoch": 0.15166048321418876,
|
| 2948 |
+
"grad_norm": NaN,
|
| 2949 |
+
"learning_rate": 0.00033036737648549196,
|
| 2950 |
+
"loss": 1.5006,
|
| 2951 |
+
"step": 8380
|
| 2952 |
+
},
|
| 2953 |
+
{
|
| 2954 |
+
"epoch": 0.15202244140801738,
|
| 2955 |
+
"grad_norm": NaN,
|
| 2956 |
+
"learning_rate": 0.00033036737648549196,
|
| 2957 |
+
"loss": 1.7442,
|
| 2958 |
+
"step": 8400
|
| 2959 |
+
},
|
| 2960 |
+
{
|
| 2961 |
+
"epoch": 0.152384399601846,
|
| 2962 |
+
"grad_norm": NaN,
|
| 2963 |
+
"learning_rate": 0.00033036737648549196,
|
| 2964 |
+
"loss": 0.4045,
|
| 2965 |
+
"step": 8420
|
| 2966 |
+
},
|
| 2967 |
+
{
|
| 2968 |
+
"epoch": 0.1527463577956746,
|
| 2969 |
+
"grad_norm": NaN,
|
| 2970 |
+
"learning_rate": 0.00033036737648549196,
|
| 2971 |
+
"loss": 1.252,
|
| 2972 |
+
"step": 8440
|
| 2973 |
+
},
|
| 2974 |
+
{
|
| 2975 |
+
"epoch": 0.15310831598950322,
|
| 2976 |
+
"grad_norm": NaN,
|
| 2977 |
+
"learning_rate": 0.00033036737648549196,
|
| 2978 |
+
"loss": 1.0349,
|
| 2979 |
+
"step": 8460
|
| 2980 |
+
},
|
| 2981 |
+
{
|
| 2982 |
+
"epoch": 0.1534702741833318,
|
| 2983 |
+
"grad_norm": NaN,
|
| 2984 |
+
"learning_rate": 0.00033042166857694393,
|
| 2985 |
+
"loss": 0.5888,
|
| 2986 |
+
"step": 8480
|
| 2987 |
+
},
|
| 2988 |
+
{
|
| 2989 |
+
"epoch": 0.15383223237716043,
|
| 2990 |
+
"grad_norm": NaN,
|
| 2991 |
+
"learning_rate": 0.00033047596066839596,
|
| 2992 |
+
"loss": 1.9747,
|
| 2993 |
+
"step": 8500
|
| 2994 |
+
},
|
| 2995 |
+
{
|
| 2996 |
+
"epoch": 0.15419419057098904,
|
| 2997 |
+
"grad_norm": NaN,
|
| 2998 |
+
"learning_rate": 0.00033047596066839596,
|
| 2999 |
+
"loss": 1.1524,
|
| 3000 |
+
"step": 8520
|
| 3001 |
+
},
|
| 3002 |
+
{
|
| 3003 |
+
"epoch": 0.15455614876481766,
|
| 3004 |
+
"grad_norm": NaN,
|
| 3005 |
+
"learning_rate": 0.00033047596066839596,
|
| 3006 |
+
"loss": 3.9692,
|
| 3007 |
+
"step": 8540
|
| 3008 |
+
},
|
| 3009 |
+
{
|
| 3010 |
+
"epoch": 0.15491810695864627,
|
| 3011 |
+
"grad_norm": NaN,
|
| 3012 |
+
"learning_rate": 0.00033047596066839596,
|
| 3013 |
+
"loss": 2.6825,
|
| 3014 |
+
"step": 8560
|
| 3015 |
+
},
|
| 3016 |
+
{
|
| 3017 |
+
"epoch": 0.1552800651524749,
|
| 3018 |
+
"grad_norm": NaN,
|
| 3019 |
+
"learning_rate": 0.00033047596066839596,
|
| 3020 |
+
"loss": 2.5432,
|
| 3021 |
+
"step": 8580
|
| 3022 |
+
},
|
| 3023 |
+
{
|
| 3024 |
+
"epoch": 0.1556420233463035,
|
| 3025 |
+
"grad_norm": NaN,
|
| 3026 |
+
"learning_rate": 0.000330530252759848,
|
| 3027 |
+
"loss": 3.0545,
|
| 3028 |
+
"step": 8600
|
| 3029 |
+
},
|
| 3030 |
+
{
|
| 3031 |
+
"epoch": 0.15600398154013212,
|
| 3032 |
+
"grad_norm": NaN,
|
| 3033 |
+
"learning_rate": 0.00033058454485129996,
|
| 3034 |
+
"loss": 1.709,
|
| 3035 |
+
"step": 8620
|
| 3036 |
+
},
|
| 3037 |
+
{
|
| 3038 |
+
"epoch": 0.15636593973396073,
|
| 3039 |
+
"grad_norm": NaN,
|
| 3040 |
+
"learning_rate": 0.00033058454485129996,
|
| 3041 |
+
"loss": 1.1557,
|
| 3042 |
+
"step": 8640
|
| 3043 |
+
},
|
| 3044 |
+
{
|
| 3045 |
+
"epoch": 0.15672789792778935,
|
| 3046 |
+
"grad_norm": NaN,
|
| 3047 |
+
"learning_rate": 0.00033058454485129996,
|
| 3048 |
+
"loss": 1.9568,
|
| 3049 |
+
"step": 8660
|
| 3050 |
+
},
|
| 3051 |
+
{
|
| 3052 |
+
"epoch": 0.15708985612161797,
|
| 3053 |
+
"grad_norm": NaN,
|
| 3054 |
+
"learning_rate": 0.00033058454485129996,
|
| 3055 |
+
"loss": 0.4101,
|
| 3056 |
+
"step": 8680
|
| 3057 |
+
},
|
| 3058 |
+
{
|
| 3059 |
+
"epoch": 0.15745181431544655,
|
| 3060 |
+
"grad_norm": NaN,
|
| 3061 |
+
"learning_rate": 0.00033058454485129996,
|
| 3062 |
+
"loss": 1.0798,
|
| 3063 |
+
"step": 8700
|
| 3064 |
+
},
|
| 3065 |
+
{
|
| 3066 |
+
"epoch": 0.15781377250927517,
|
| 3067 |
+
"grad_norm": NaN,
|
| 3068 |
+
"learning_rate": 0.00033058454485129996,
|
| 3069 |
+
"loss": 0.9417,
|
| 3070 |
+
"step": 8720
|
| 3071 |
+
},
|
| 3072 |
+
{
|
| 3073 |
+
"epoch": 0.15817573070310378,
|
| 3074 |
+
"grad_norm": NaN,
|
| 3075 |
+
"learning_rate": 0.00033058454485129996,
|
| 3076 |
+
"loss": 0.3268,
|
| 3077 |
+
"step": 8740
|
| 3078 |
+
},
|
| 3079 |
+
{
|
| 3080 |
+
"epoch": 0.1585376888969324,
|
| 3081 |
+
"grad_norm": NaN,
|
| 3082 |
+
"learning_rate": 0.00033058454485129996,
|
| 3083 |
+
"loss": 2.7045,
|
| 3084 |
+
"step": 8760
|
| 3085 |
+
},
|
| 3086 |
+
{
|
| 3087 |
+
"epoch": 0.15889964709076101,
|
| 3088 |
+
"grad_norm": NaN,
|
| 3089 |
+
"learning_rate": 0.00033058454485129996,
|
| 3090 |
+
"loss": 3.1654,
|
| 3091 |
+
"step": 8780
|
| 3092 |
+
},
|
| 3093 |
+
{
|
| 3094 |
+
"epoch": 0.15926160528458963,
|
| 3095 |
+
"grad_norm": NaN,
|
| 3096 |
+
"learning_rate": 0.00033058454485129996,
|
| 3097 |
+
"loss": 3.7107,
|
| 3098 |
+
"step": 8800
|
| 3099 |
+
},
|
| 3100 |
+
{
|
| 3101 |
+
"epoch": 0.15962356347841825,
|
| 3102 |
+
"grad_norm": NaN,
|
| 3103 |
+
"learning_rate": 0.00033058454485129996,
|
| 3104 |
+
"loss": 1.1776,
|
| 3105 |
+
"step": 8820
|
| 3106 |
+
},
|
| 3107 |
+
{
|
| 3108 |
+
"epoch": 0.15998552167224686,
|
| 3109 |
+
"grad_norm": NaN,
|
| 3110 |
+
"learning_rate": 0.00033058454485129996,
|
| 3111 |
+
"loss": 0.7538,
|
| 3112 |
+
"step": 8840
|
| 3113 |
+
},
|
| 3114 |
+
{
|
| 3115 |
+
"epoch": 0.16034747986607548,
|
| 3116 |
+
"grad_norm": NaN,
|
| 3117 |
+
"learning_rate": 0.000330638836942752,
|
| 3118 |
+
"loss": 5.9555,
|
| 3119 |
+
"step": 8860
|
| 3120 |
+
},
|
| 3121 |
+
{
|
| 3122 |
+
"epoch": 0.1607094380599041,
|
| 3123 |
+
"grad_norm": NaN,
|
| 3124 |
+
"learning_rate": 0.000330693129034204,
|
| 3125 |
+
"loss": 2.4652,
|
| 3126 |
+
"step": 8880
|
| 3127 |
+
},
|
| 3128 |
+
{
|
| 3129 |
+
"epoch": 0.1610713962537327,
|
| 3130 |
+
"grad_norm": NaN,
|
| 3131 |
+
"learning_rate": 0.000330693129034204,
|
| 3132 |
+
"loss": 0.6042,
|
| 3133 |
+
"step": 8900
|
| 3134 |
+
},
|
| 3135 |
+
{
|
| 3136 |
+
"epoch": 0.1614333544475613,
|
| 3137 |
+
"grad_norm": NaN,
|
| 3138 |
+
"learning_rate": 0.000330693129034204,
|
| 3139 |
+
"loss": 2.164,
|
| 3140 |
+
"step": 8920
|
| 3141 |
+
},
|
| 3142 |
+
{
|
| 3143 |
+
"epoch": 0.1617953126413899,
|
| 3144 |
+
"grad_norm": NaN,
|
| 3145 |
+
"learning_rate": 0.000330693129034204,
|
| 3146 |
+
"loss": 1.5732,
|
| 3147 |
+
"step": 8940
|
| 3148 |
+
},
|
| 3149 |
+
{
|
| 3150 |
+
"epoch": 0.16215727083521853,
|
| 3151 |
+
"grad_norm": NaN,
|
| 3152 |
+
"learning_rate": 0.000330693129034204,
|
| 3153 |
+
"loss": 0.7035,
|
| 3154 |
+
"step": 8960
|
| 3155 |
+
},
|
| 3156 |
+
{
|
| 3157 |
+
"epoch": 0.16251922902904714,
|
| 3158 |
+
"grad_norm": NaN,
|
| 3159 |
+
"learning_rate": 0.000330693129034204,
|
| 3160 |
+
"loss": 1.3199,
|
| 3161 |
+
"step": 8980
|
| 3162 |
+
},
|
| 3163 |
+
{
|
| 3164 |
+
"epoch": 0.16288118722287576,
|
| 3165 |
+
"grad_norm": NaN,
|
| 3166 |
+
"learning_rate": 0.000330693129034204,
|
| 3167 |
+
"loss": 2.9008,
|
| 3168 |
+
"step": 9000
|
| 3169 |
+
},
|
| 3170 |
+
{
|
| 3171 |
+
"epoch": 0.16324314541670437,
|
| 3172 |
+
"grad_norm": NaN,
|
| 3173 |
+
"learning_rate": 0.000330693129034204,
|
| 3174 |
+
"loss": 2.8915,
|
| 3175 |
+
"step": 9020
|
| 3176 |
+
},
|
| 3177 |
+
{
|
| 3178 |
+
"epoch": 0.163605103610533,
|
| 3179 |
+
"grad_norm": NaN,
|
| 3180 |
+
"learning_rate": 0.000330693129034204,
|
| 3181 |
+
"loss": 0.6013,
|
| 3182 |
+
"step": 9040
|
| 3183 |
+
},
|
| 3184 |
+
{
|
| 3185 |
+
"epoch": 0.1639670618043616,
|
| 3186 |
+
"grad_norm": NaN,
|
| 3187 |
+
"learning_rate": 0.000330693129034204,
|
| 3188 |
+
"loss": 0.7603,
|
| 3189 |
+
"step": 9060
|
| 3190 |
+
},
|
| 3191 |
+
{
|
| 3192 |
+
"epoch": 0.16432901999819022,
|
| 3193 |
+
"grad_norm": NaN,
|
| 3194 |
+
"learning_rate": 0.000330693129034204,
|
| 3195 |
+
"loss": 2.1826,
|
| 3196 |
+
"step": 9080
|
| 3197 |
+
},
|
| 3198 |
+
{
|
| 3199 |
+
"epoch": 0.16469097819201883,
|
| 3200 |
+
"grad_norm": NaN,
|
| 3201 |
+
"learning_rate": 0.000330693129034204,
|
| 3202 |
+
"loss": 3.9643,
|
| 3203 |
+
"step": 9100
|
| 3204 |
+
},
|
| 3205 |
+
{
|
| 3206 |
+
"epoch": 0.16505293638584742,
|
| 3207 |
+
"grad_norm": NaN,
|
| 3208 |
+
"learning_rate": 0.000330693129034204,
|
| 3209 |
+
"loss": 1.6171,
|
| 3210 |
+
"step": 9120
|
| 3211 |
+
},
|
| 3212 |
+
{
|
| 3213 |
+
"epoch": 0.16541489457967604,
|
| 3214 |
+
"grad_norm": NaN,
|
| 3215 |
+
"learning_rate": 0.000330693129034204,
|
| 3216 |
+
"loss": 0.5245,
|
| 3217 |
+
"step": 9140
|
| 3218 |
+
},
|
| 3219 |
+
{
|
| 3220 |
+
"epoch": 0.16577685277350465,
|
| 3221 |
+
"grad_norm": NaN,
|
| 3222 |
+
"learning_rate": 0.000330693129034204,
|
| 3223 |
+
"loss": 2.243,
|
| 3224 |
+
"step": 9160
|
| 3225 |
+
},
|
| 3226 |
+
{
|
| 3227 |
+
"epoch": 0.16613881096733327,
|
| 3228 |
+
"grad_norm": NaN,
|
| 3229 |
+
"learning_rate": 0.000330693129034204,
|
| 3230 |
+
"loss": 2.6442,
|
| 3231 |
+
"step": 9180
|
| 3232 |
+
},
|
| 3233 |
+
{
|
| 3234 |
+
"epoch": 0.16650076916116188,
|
| 3235 |
+
"grad_norm": NaN,
|
| 3236 |
+
"learning_rate": 0.000330747421125656,
|
| 3237 |
+
"loss": 1.4556,
|
| 3238 |
+
"step": 9200
|
| 3239 |
+
},
|
| 3240 |
+
{
|
| 3241 |
+
"epoch": 0.1668627273549905,
|
| 3242 |
+
"grad_norm": NaN,
|
| 3243 |
+
"learning_rate": 0.000330747421125656,
|
| 3244 |
+
"loss": 2.888,
|
| 3245 |
+
"step": 9220
|
| 3246 |
+
},
|
| 3247 |
+
{
|
| 3248 |
+
"epoch": 0.1672246855488191,
|
| 3249 |
+
"grad_norm": NaN,
|
| 3250 |
+
"learning_rate": 0.000330747421125656,
|
| 3251 |
+
"loss": 0.8837,
|
| 3252 |
+
"step": 9240
|
| 3253 |
+
},
|
| 3254 |
+
{
|
| 3255 |
+
"epoch": 0.16758664374264773,
|
| 3256 |
+
"grad_norm": NaN,
|
| 3257 |
+
"learning_rate": 0.000330747421125656,
|
| 3258 |
+
"loss": 3.352,
|
| 3259 |
+
"step": 9260
|
| 3260 |
+
},
|
| 3261 |
+
{
|
| 3262 |
+
"epoch": 0.16794860193647634,
|
| 3263 |
+
"grad_norm": NaN,
|
| 3264 |
+
"learning_rate": 0.000330747421125656,
|
| 3265 |
+
"loss": 1.1593,
|
| 3266 |
+
"step": 9280
|
| 3267 |
+
},
|
| 3268 |
+
{
|
| 3269 |
+
"epoch": 0.16831056013030496,
|
| 3270 |
+
"grad_norm": NaN,
|
| 3271 |
+
"learning_rate": 0.000330801713217108,
|
| 3272 |
+
"loss": 1.2725,
|
| 3273 |
+
"step": 9300
|
| 3274 |
+
},
|
| 3275 |
+
{
|
| 3276 |
+
"epoch": 0.16867251832413357,
|
| 3277 |
+
"grad_norm": NaN,
|
| 3278 |
+
"learning_rate": 0.000330801713217108,
|
| 3279 |
+
"loss": 2.86,
|
| 3280 |
+
"step": 9320
|
| 3281 |
+
},
|
| 3282 |
+
{
|
| 3283 |
+
"epoch": 0.16903447651796216,
|
| 3284 |
+
"grad_norm": NaN,
|
| 3285 |
+
"learning_rate": 0.000330801713217108,
|
| 3286 |
+
"loss": 1.3259,
|
| 3287 |
+
"step": 9340
|
| 3288 |
+
},
|
| 3289 |
+
{
|
| 3290 |
+
"epoch": 0.16939643471179078,
|
| 3291 |
+
"grad_norm": NaN,
|
| 3292 |
+
"learning_rate": 0.000330801713217108,
|
| 3293 |
+
"loss": 1.9639,
|
| 3294 |
+
"step": 9360
|
| 3295 |
+
},
|
| 3296 |
+
{
|
| 3297 |
+
"epoch": 0.1697583929056194,
|
| 3298 |
+
"grad_norm": NaN,
|
| 3299 |
+
"learning_rate": 0.000330801713217108,
|
| 3300 |
+
"loss": 1.0361,
|
| 3301 |
+
"step": 9380
|
| 3302 |
+
},
|
| 3303 |
+
{
|
| 3304 |
+
"epoch": 0.170120351099448,
|
| 3305 |
+
"grad_norm": NaN,
|
| 3306 |
+
"learning_rate": 0.000330801713217108,
|
| 3307 |
+
"loss": 0.8044,
|
| 3308 |
+
"step": 9400
|
| 3309 |
+
},
|
| 3310 |
+
{
|
| 3311 |
+
"epoch": 0.17048230929327662,
|
| 3312 |
+
"grad_norm": NaN,
|
| 3313 |
+
"learning_rate": 0.00033085600530856005,
|
| 3314 |
+
"loss": 2.3303,
|
| 3315 |
+
"step": 9420
|
| 3316 |
+
},
|
| 3317 |
+
{
|
| 3318 |
+
"epoch": 0.17084426748710524,
|
| 3319 |
+
"grad_norm": NaN,
|
| 3320 |
+
"learning_rate": 0.00033085600530856005,
|
| 3321 |
+
"loss": 2.9078,
|
| 3322 |
+
"step": 9440
|
| 3323 |
+
},
|
| 3324 |
+
{
|
| 3325 |
+
"epoch": 0.17120622568093385,
|
| 3326 |
+
"grad_norm": NaN,
|
| 3327 |
+
"learning_rate": 0.00033085600530856005,
|
| 3328 |
+
"loss": 0.4421,
|
| 3329 |
+
"step": 9460
|
| 3330 |
+
},
|
| 3331 |
+
{
|
| 3332 |
+
"epoch": 0.17156818387476247,
|
| 3333 |
+
"grad_norm": NaN,
|
| 3334 |
+
"learning_rate": 0.00033085600530856005,
|
| 3335 |
+
"loss": 5.0288,
|
| 3336 |
+
"step": 9480
|
| 3337 |
+
},
|
| 3338 |
+
{
|
| 3339 |
+
"epoch": 0.17193014206859109,
|
| 3340 |
+
"grad_norm": NaN,
|
| 3341 |
+
"learning_rate": 0.00033085600530856005,
|
| 3342 |
+
"loss": 0.8455,
|
| 3343 |
+
"step": 9500
|
| 3344 |
+
},
|
| 3345 |
+
{
|
| 3346 |
+
"epoch": 0.1722921002624197,
|
| 3347 |
+
"grad_norm": NaN,
|
| 3348 |
+
"learning_rate": 0.00033085600530856005,
|
| 3349 |
+
"loss": 3.2879,
|
| 3350 |
+
"step": 9520
|
| 3351 |
+
},
|
| 3352 |
+
{
|
| 3353 |
+
"epoch": 0.17265405845624832,
|
| 3354 |
+
"grad_norm": NaN,
|
| 3355 |
+
"learning_rate": 0.00033085600530856005,
|
| 3356 |
+
"loss": 3.2948,
|
| 3357 |
+
"step": 9540
|
| 3358 |
+
},
|
| 3359 |
+
{
|
| 3360 |
+
"epoch": 0.1730160166500769,
|
| 3361 |
+
"grad_norm": NaN,
|
| 3362 |
+
"learning_rate": 0.00033085600530856005,
|
| 3363 |
+
"loss": 1.6391,
|
| 3364 |
+
"step": 9560
|
| 3365 |
+
},
|
| 3366 |
+
{
|
| 3367 |
+
"epoch": 0.17337797484390552,
|
| 3368 |
+
"grad_norm": NaN,
|
| 3369 |
+
"learning_rate": 0.00033085600530856005,
|
| 3370 |
+
"loss": 0.8699,
|
| 3371 |
+
"step": 9580
|
| 3372 |
+
},
|
| 3373 |
+
{
|
| 3374 |
+
"epoch": 0.17373993303773413,
|
| 3375 |
+
"grad_norm": NaN,
|
| 3376 |
+
"learning_rate": 0.00033085600530856005,
|
| 3377 |
+
"loss": 0.3308,
|
| 3378 |
+
"step": 9600
|
| 3379 |
+
},
|
| 3380 |
+
{
|
| 3381 |
+
"epoch": 0.17410189123156275,
|
| 3382 |
+
"grad_norm": NaN,
|
| 3383 |
+
"learning_rate": 0.00033085600530856005,
|
| 3384 |
+
"loss": 0.5138,
|
| 3385 |
+
"step": 9620
|
| 3386 |
+
},
|
| 3387 |
+
{
|
| 3388 |
+
"epoch": 0.17446384942539137,
|
| 3389 |
+
"grad_norm": NaN,
|
| 3390 |
+
"learning_rate": 0.00033085600530856005,
|
| 3391 |
+
"loss": 2.5083,
|
| 3392 |
+
"step": 9640
|
| 3393 |
+
},
|
| 3394 |
+
{
|
| 3395 |
+
"epoch": 0.17482580761921998,
|
| 3396 |
+
"grad_norm": NaN,
|
| 3397 |
+
"learning_rate": 0.00033085600530856005,
|
| 3398 |
+
"loss": 2.8029,
|
| 3399 |
+
"step": 9660
|
| 3400 |
+
},
|
| 3401 |
+
{
|
| 3402 |
+
"epoch": 0.1751877658130486,
|
| 3403 |
+
"grad_norm": NaN,
|
| 3404 |
+
"learning_rate": 0.0003309102974000121,
|
| 3405 |
+
"loss": 2.2681,
|
| 3406 |
+
"step": 9680
|
| 3407 |
+
},
|
| 3408 |
+
{
|
| 3409 |
+
"epoch": 0.1755497240068772,
|
| 3410 |
+
"grad_norm": NaN,
|
| 3411 |
+
"learning_rate": 0.0003309102974000121,
|
| 3412 |
+
"loss": 0.0,
|
| 3413 |
+
"step": 9700
|
| 3414 |
+
},
|
| 3415 |
+
{
|
| 3416 |
+
"epoch": 0.17591168220070583,
|
| 3417 |
+
"grad_norm": NaN,
|
| 3418 |
+
"learning_rate": 0.0003309102974000121,
|
| 3419 |
+
"loss": 0.7887,
|
| 3420 |
+
"step": 9720
|
| 3421 |
+
},
|
| 3422 |
+
{
|
| 3423 |
+
"epoch": 0.17627364039453444,
|
| 3424 |
+
"grad_norm": NaN,
|
| 3425 |
+
"learning_rate": 0.0003309102974000121,
|
| 3426 |
+
"loss": 3.1913,
|
| 3427 |
+
"step": 9740
|
| 3428 |
+
},
|
| 3429 |
+
{
|
| 3430 |
+
"epoch": 0.17663559858836306,
|
| 3431 |
+
"grad_norm": NaN,
|
| 3432 |
+
"learning_rate": 0.0003309102974000121,
|
| 3433 |
+
"loss": 0.0,
|
| 3434 |
+
"step": 9760
|
| 3435 |
+
},
|
| 3436 |
+
{
|
| 3437 |
+
"epoch": 0.17699755678219165,
|
| 3438 |
+
"grad_norm": NaN,
|
| 3439 |
+
"learning_rate": 0.0003309102974000121,
|
| 3440 |
+
"loss": 0.4424,
|
| 3441 |
+
"step": 9780
|
| 3442 |
+
},
|
| 3443 |
+
{
|
| 3444 |
+
"epoch": 0.17735951497602026,
|
| 3445 |
+
"grad_norm": NaN,
|
| 3446 |
+
"learning_rate": 0.0003309102974000121,
|
| 3447 |
+
"loss": 2.4798,
|
| 3448 |
+
"step": 9800
|
| 3449 |
+
},
|
| 3450 |
+
{
|
| 3451 |
+
"epoch": 0.17772147316984888,
|
| 3452 |
+
"grad_norm": NaN,
|
| 3453 |
+
"learning_rate": 0.0003309645894914641,
|
| 3454 |
+
"loss": 2.4174,
|
| 3455 |
+
"step": 9820
|
| 3456 |
+
},
|
| 3457 |
+
{
|
| 3458 |
+
"epoch": 0.1780834313636775,
|
| 3459 |
+
"grad_norm": NaN,
|
| 3460 |
+
"learning_rate": 0.0003309645894914641,
|
| 3461 |
+
"loss": 1.3177,
|
| 3462 |
+
"step": 9840
|
| 3463 |
+
},
|
| 3464 |
+
{
|
| 3465 |
+
"epoch": 0.1784453895575061,
|
| 3466 |
+
"grad_norm": NaN,
|
| 3467 |
+
"learning_rate": 0.0003309645894914641,
|
| 3468 |
+
"loss": 3.8794,
|
| 3469 |
+
"step": 9860
|
| 3470 |
+
},
|
| 3471 |
+
{
|
| 3472 |
+
"epoch": 0.17880734775133472,
|
| 3473 |
+
"grad_norm": NaN,
|
| 3474 |
+
"learning_rate": 0.0003309645894914641,
|
| 3475 |
+
"loss": 1.0543,
|
| 3476 |
+
"step": 9880
|
| 3477 |
+
},
|
| 3478 |
+
{
|
| 3479 |
+
"epoch": 0.17916930594516334,
|
| 3480 |
+
"grad_norm": NaN,
|
| 3481 |
+
"learning_rate": 0.0003309645894914641,
|
| 3482 |
+
"loss": 1.4625,
|
| 3483 |
+
"step": 9900
|
| 3484 |
+
},
|
| 3485 |
+
{
|
| 3486 |
+
"epoch": 0.17953126413899195,
|
| 3487 |
+
"grad_norm": NaN,
|
| 3488 |
+
"learning_rate": 0.0003309645894914641,
|
| 3489 |
+
"loss": 3.5846,
|
| 3490 |
+
"step": 9920
|
| 3491 |
+
},
|
| 3492 |
+
{
|
| 3493 |
+
"epoch": 0.17989322233282057,
|
| 3494 |
+
"grad_norm": NaN,
|
| 3495 |
+
"learning_rate": 0.0003309645894914641,
|
| 3496 |
+
"loss": 1.6589,
|
| 3497 |
+
"step": 9940
|
| 3498 |
+
},
|
| 3499 |
+
{
|
| 3500 |
+
"epoch": 0.18025518052664918,
|
| 3501 |
+
"grad_norm": NaN,
|
| 3502 |
+
"learning_rate": 0.0003309645894914641,
|
| 3503 |
+
"loss": 3.6713,
|
| 3504 |
+
"step": 9960
|
| 3505 |
+
},
|
| 3506 |
+
{
|
| 3507 |
+
"epoch": 0.18061713872047777,
|
| 3508 |
+
"grad_norm": NaN,
|
| 3509 |
+
"learning_rate": 0.0003309645894914641,
|
| 3510 |
+
"loss": 0.0,
|
| 3511 |
+
"step": 9980
|
| 3512 |
+
},
|
| 3513 |
+
{
|
| 3514 |
+
"epoch": 0.1809790969143064,
|
| 3515 |
+
"grad_norm": NaN,
|
| 3516 |
+
"learning_rate": 0.0003309645894914641,
|
| 3517 |
+
"loss": 0.404,
|
| 3518 |
+
"step": 10000
|
| 3519 |
+
},
|
| 3520 |
+
{
|
| 3521 |
+
"epoch": 0.1809790969143064,
|
| 3522 |
+
"eval_accuracy": 4.7270488930998444e-05,
|
| 3523 |
+
"eval_loss": NaN,
|
| 3524 |
+
"eval_runtime": 168.7884,
|
| 3525 |
+
"eval_samples_per_second": 3601.195,
|
| 3526 |
+
"eval_steps_per_second": 3.519,
|
| 3527 |
+
"step": 10000
|
| 3528 |
}
|
| 3529 |
],
|
| 3530 |
"logging_steps": 20,
|
|
|
|
| 3538 |
"should_epoch_stop": false,
|
| 3539 |
"should_evaluate": false,
|
| 3540 |
"should_log": false,
|
| 3541 |
+
"should_save": true,
|
| 3542 |
"should_training_stop": false
|
| 3543 |
},
|
| 3544 |
"attributes": {}
|
| 3545 |
}
|
| 3546 |
},
|
| 3547 |
+
"total_flos": 1.438362107904e+16,
|
| 3548 |
"train_batch_size": 512,
|
| 3549 |
"trial_name": null,
|
| 3550 |
"trial_params": null
|