Training in progress, step 408, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 50503544
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5cc49cd2123962dccbe325a2bceb54ce95be21d864954094864c3a779fd90cf9
|
| 3 |
size 50503544
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 101184122
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:773b86455e5f22c5da806f865a8d957c30131f2ceaac157c9d1818f781c9d876
|
| 3 |
size 101184122
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:20768a83e56b6fa769e574419f07c86e79c791425ce3347059f718b93fcccfcd
|
| 3 |
size 14512
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:844d07a2596435d36c66584eee737f312e21cc8a87613c0830623fb4b3e18b1d
|
| 3 |
size 14512
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a830cc09aea0d73a752aadc4db39e884877eb126ff93c26e59113e94cd6ce260
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 204,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1451,6 +1451,1442 @@
|
|
| 1451 |
"eval_samples_per_second": 10.473,
|
| 1452 |
"eval_steps_per_second": 5.244,
|
| 1453 |
"step": 204
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1454 |
}
|
| 1455 |
],
|
| 1456 |
"logging_steps": 1,
|
|
@@ -1470,7 +2906,7 @@
|
|
| 1470 |
"attributes": {}
|
| 1471 |
}
|
| 1472 |
},
|
| 1473 |
-
"total_flos":
|
| 1474 |
"train_batch_size": 1,
|
| 1475 |
"trial_name": null,
|
| 1476 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.5015365703749232,
|
| 5 |
"eval_steps": 204,
|
| 6 |
+
"global_step": 408,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1451 |
"eval_samples_per_second": 10.473,
|
| 1452 |
"eval_steps_per_second": 5.244,
|
| 1453 |
"step": 204
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 0.2519975414874001,
|
| 1457 |
+
"grad_norm": 0.3105545938014984,
|
| 1458 |
+
"learning_rate": 9.015846028376462e-05,
|
| 1459 |
+
"loss": 1.2827,
|
| 1460 |
+
"step": 205
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 0.25322679778733864,
|
| 1464 |
+
"grad_norm": 0.2826372981071472,
|
| 1465 |
+
"learning_rate": 9.00354723283191e-05,
|
| 1466 |
+
"loss": 1.1159,
|
| 1467 |
+
"step": 206
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 0.2544560540872772,
|
| 1471 |
+
"grad_norm": 0.2823708951473236,
|
| 1472 |
+
"learning_rate": 8.991180564605086e-05,
|
| 1473 |
+
"loss": 1.0368,
|
| 1474 |
+
"step": 207
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 0.25568531038721576,
|
| 1478 |
+
"grad_norm": 0.28265297412872314,
|
| 1479 |
+
"learning_rate": 8.978746233349802e-05,
|
| 1480 |
+
"loss": 1.1583,
|
| 1481 |
+
"step": 208
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 0.2569145666871543,
|
| 1485 |
+
"grad_norm": 0.3202212452888489,
|
| 1486 |
+
"learning_rate": 8.966244449866973e-05,
|
| 1487 |
+
"loss": 1.2069,
|
| 1488 |
+
"step": 209
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 0.2581438229870928,
|
| 1492 |
+
"grad_norm": 0.30576291680336,
|
| 1493 |
+
"learning_rate": 8.953675426101038e-05,
|
| 1494 |
+
"loss": 1.1588,
|
| 1495 |
+
"step": 210
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 0.25937307928703135,
|
| 1499 |
+
"grad_norm": 0.3853960633277893,
|
| 1500 |
+
"learning_rate": 8.941039375136371e-05,
|
| 1501 |
+
"loss": 1.1947,
|
| 1502 |
+
"step": 211
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 0.2606023355869699,
|
| 1506 |
+
"grad_norm": 0.4404067099094391,
|
| 1507 |
+
"learning_rate": 8.928336511193669e-05,
|
| 1508 |
+
"loss": 1.0786,
|
| 1509 |
+
"step": 212
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 0.2618315918869084,
|
| 1513 |
+
"grad_norm": 0.422333300113678,
|
| 1514 |
+
"learning_rate": 8.915567049626315e-05,
|
| 1515 |
+
"loss": 1.1454,
|
| 1516 |
+
"step": 213
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 0.26306084818684694,
|
| 1520 |
+
"grad_norm": 0.5277565121650696,
|
| 1521 |
+
"learning_rate": 8.902731206916734e-05,
|
| 1522 |
+
"loss": 0.7775,
|
| 1523 |
+
"step": 214
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 0.26429010448678547,
|
| 1527 |
+
"grad_norm": 0.7032243609428406,
|
| 1528 |
+
"learning_rate": 8.889829200672719e-05,
|
| 1529 |
+
"loss": 0.5771,
|
| 1530 |
+
"step": 215
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 0.26551936078672406,
|
| 1534 |
+
"grad_norm": 0.6663339734077454,
|
| 1535 |
+
"learning_rate": 8.876861249623739e-05,
|
| 1536 |
+
"loss": 0.616,
|
| 1537 |
+
"step": 216
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 0.2667486170866626,
|
| 1541 |
+
"grad_norm": 0.8129518628120422,
|
| 1542 |
+
"learning_rate": 8.863827573617238e-05,
|
| 1543 |
+
"loss": 1.1483,
|
| 1544 |
+
"step": 217
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 0.2679778733866011,
|
| 1548 |
+
"grad_norm": 1.0273211002349854,
|
| 1549 |
+
"learning_rate": 8.850728393614902e-05,
|
| 1550 |
+
"loss": 1.2066,
|
| 1551 |
+
"step": 218
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 0.26920712968653965,
|
| 1555 |
+
"grad_norm": 1.5424954891204834,
|
| 1556 |
+
"learning_rate": 8.837563931688919e-05,
|
| 1557 |
+
"loss": 1.247,
|
| 1558 |
+
"step": 219
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 0.2704363859864782,
|
| 1562 |
+
"grad_norm": 2.9167752265930176,
|
| 1563 |
+
"learning_rate": 8.824334411018204e-05,
|
| 1564 |
+
"loss": 1.3413,
|
| 1565 |
+
"step": 220
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 0.2716656422864167,
|
| 1569 |
+
"grad_norm": 5.498292446136475,
|
| 1570 |
+
"learning_rate": 8.811040055884629e-05,
|
| 1571 |
+
"loss": 1.0072,
|
| 1572 |
+
"step": 221
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 0.27289489858635524,
|
| 1576 |
+
"grad_norm": 3.1687686443328857,
|
| 1577 |
+
"learning_rate": 8.797681091669206e-05,
|
| 1578 |
+
"loss": 1.3309,
|
| 1579 |
+
"step": 222
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 0.27412415488629377,
|
| 1583 |
+
"grad_norm": 2.760160446166992,
|
| 1584 |
+
"learning_rate": 8.784257744848279e-05,
|
| 1585 |
+
"loss": 1.5268,
|
| 1586 |
+
"step": 223
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.27535341118623236,
|
| 1590 |
+
"grad_norm": 2.3323326110839844,
|
| 1591 |
+
"learning_rate": 8.770770242989679e-05,
|
| 1592 |
+
"loss": 1.27,
|
| 1593 |
+
"step": 224
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 0.2765826674861709,
|
| 1597 |
+
"grad_norm": 2.150510549545288,
|
| 1598 |
+
"learning_rate": 8.75721881474886e-05,
|
| 1599 |
+
"loss": 1.0602,
|
| 1600 |
+
"step": 225
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 0.2778119237861094,
|
| 1604 |
+
"grad_norm": 0.23049846291542053,
|
| 1605 |
+
"learning_rate": 8.743603689865039e-05,
|
| 1606 |
+
"loss": 1.0067,
|
| 1607 |
+
"step": 226
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 0.27904118008604795,
|
| 1611 |
+
"grad_norm": 0.2650708556175232,
|
| 1612 |
+
"learning_rate": 8.729925099157281e-05,
|
| 1613 |
+
"loss": 1.1932,
|
| 1614 |
+
"step": 227
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 0.2802704363859865,
|
| 1618 |
+
"grad_norm": 0.2723963260650635,
|
| 1619 |
+
"learning_rate": 8.7161832745206e-05,
|
| 1620 |
+
"loss": 1.2495,
|
| 1621 |
+
"step": 228
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 0.281499692685925,
|
| 1625 |
+
"grad_norm": 0.26627010107040405,
|
| 1626 |
+
"learning_rate": 8.702378448922026e-05,
|
| 1627 |
+
"loss": 1.2837,
|
| 1628 |
+
"step": 229
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 0.28272894898586354,
|
| 1632 |
+
"grad_norm": 0.2728361189365387,
|
| 1633 |
+
"learning_rate": 8.688510856396648e-05,
|
| 1634 |
+
"loss": 1.2969,
|
| 1635 |
+
"step": 230
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 0.28395820528580207,
|
| 1639 |
+
"grad_norm": 0.26788559556007385,
|
| 1640 |
+
"learning_rate": 8.674580732043656e-05,
|
| 1641 |
+
"loss": 1.0944,
|
| 1642 |
+
"step": 231
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 0.28518746158574065,
|
| 1646 |
+
"grad_norm": 0.3129604160785675,
|
| 1647 |
+
"learning_rate": 8.660588312022344e-05,
|
| 1648 |
+
"loss": 1.3591,
|
| 1649 |
+
"step": 232
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 0.2864167178856792,
|
| 1653 |
+
"grad_norm": 0.32250627875328064,
|
| 1654 |
+
"learning_rate": 8.646533833548119e-05,
|
| 1655 |
+
"loss": 1.1469,
|
| 1656 |
+
"step": 233
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 0.2876459741856177,
|
| 1660 |
+
"grad_norm": 0.32614386081695557,
|
| 1661 |
+
"learning_rate": 8.632417534888473e-05,
|
| 1662 |
+
"loss": 1.3551,
|
| 1663 |
+
"step": 234
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 0.28887523048555624,
|
| 1667 |
+
"grad_norm": 0.3620636463165283,
|
| 1668 |
+
"learning_rate": 8.61823965535894e-05,
|
| 1669 |
+
"loss": 1.1427,
|
| 1670 |
+
"step": 235
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 0.2901044867854948,
|
| 1674 |
+
"grad_norm": 0.39082473516464233,
|
| 1675 |
+
"learning_rate": 8.604000435319047e-05,
|
| 1676 |
+
"loss": 1.0041,
|
| 1677 |
+
"step": 236
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 0.2913337430854333,
|
| 1681 |
+
"grad_norm": 0.3823097050189972,
|
| 1682 |
+
"learning_rate": 8.589700116168232e-05,
|
| 1683 |
+
"loss": 1.1756,
|
| 1684 |
+
"step": 237
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 0.29256299938537184,
|
| 1688 |
+
"grad_norm": 0.5359341502189636,
|
| 1689 |
+
"learning_rate": 8.575338940341757e-05,
|
| 1690 |
+
"loss": 1.1814,
|
| 1691 |
+
"step": 238
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 0.29379225568531037,
|
| 1695 |
+
"grad_norm": 0.6902546286582947,
|
| 1696 |
+
"learning_rate": 8.560917151306593e-05,
|
| 1697 |
+
"loss": 0.9253,
|
| 1698 |
+
"step": 239
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 0.29502151198524895,
|
| 1702 |
+
"grad_norm": 0.7236252427101135,
|
| 1703 |
+
"learning_rate": 8.5464349935573e-05,
|
| 1704 |
+
"loss": 0.6398,
|
| 1705 |
+
"step": 240
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 0.2962507682851875,
|
| 1709 |
+
"grad_norm": 0.7172759175300598,
|
| 1710 |
+
"learning_rate": 8.53189271261187e-05,
|
| 1711 |
+
"loss": 0.9061,
|
| 1712 |
+
"step": 241
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 0.297480024585126,
|
| 1716 |
+
"grad_norm": 0.7999723553657532,
|
| 1717 |
+
"learning_rate": 8.517290555007578e-05,
|
| 1718 |
+
"loss": 1.0691,
|
| 1719 |
+
"step": 242
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 0.29870928088506454,
|
| 1723 |
+
"grad_norm": 1.235872745513916,
|
| 1724 |
+
"learning_rate": 8.502628768296788e-05,
|
| 1725 |
+
"loss": 1.5235,
|
| 1726 |
+
"step": 243
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 0.2999385371850031,
|
| 1730 |
+
"grad_norm": 1.9676207304000854,
|
| 1731 |
+
"learning_rate": 8.487907601042777e-05,
|
| 1732 |
+
"loss": 1.5859,
|
| 1733 |
+
"step": 244
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.3011677934849416,
|
| 1737 |
+
"grad_norm": 3.5035860538482666,
|
| 1738 |
+
"learning_rate": 8.473127302815496e-05,
|
| 1739 |
+
"loss": 1.1743,
|
| 1740 |
+
"step": 245
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 0.30239704978488013,
|
| 1744 |
+
"grad_norm": 4.519472599029541,
|
| 1745 |
+
"learning_rate": 8.458288124187359e-05,
|
| 1746 |
+
"loss": 0.7165,
|
| 1747 |
+
"step": 246
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 0.30362630608481866,
|
| 1751 |
+
"grad_norm": 2.3718838691711426,
|
| 1752 |
+
"learning_rate": 8.443390316728987e-05,
|
| 1753 |
+
"loss": 1.1449,
|
| 1754 |
+
"step": 247
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 0.3048555623847572,
|
| 1758 |
+
"grad_norm": 2.1668829917907715,
|
| 1759 |
+
"learning_rate": 8.428434133004937e-05,
|
| 1760 |
+
"loss": 1.0383,
|
| 1761 |
+
"step": 248
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 0.3060848186846958,
|
| 1765 |
+
"grad_norm": 3.2350733280181885,
|
| 1766 |
+
"learning_rate": 8.413419826569435e-05,
|
| 1767 |
+
"loss": 1.2341,
|
| 1768 |
+
"step": 249
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.3073140749846343,
|
| 1772 |
+
"grad_norm": 2.3541886806488037,
|
| 1773 |
+
"learning_rate": 8.398347651962064e-05,
|
| 1774 |
+
"loss": 1.0355,
|
| 1775 |
+
"step": 250
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 0.30854333128457284,
|
| 1779 |
+
"grad_norm": 0.2730487883090973,
|
| 1780 |
+
"learning_rate": 8.383217864703456e-05,
|
| 1781 |
+
"loss": 1.2813,
|
| 1782 |
+
"step": 251
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 0.30977258758451137,
|
| 1786 |
+
"grad_norm": 0.2517383098602295,
|
| 1787 |
+
"learning_rate": 8.36803072129096e-05,
|
| 1788 |
+
"loss": 1.1793,
|
| 1789 |
+
"step": 252
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 0.3110018438844499,
|
| 1793 |
+
"grad_norm": 0.28486472368240356,
|
| 1794 |
+
"learning_rate": 8.352786479194288e-05,
|
| 1795 |
+
"loss": 1.4065,
|
| 1796 |
+
"step": 253
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.31223110018438843,
|
| 1800 |
+
"grad_norm": 0.3247184455394745,
|
| 1801 |
+
"learning_rate": 8.337485396851155e-05,
|
| 1802 |
+
"loss": 1.4863,
|
| 1803 |
+
"step": 254
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.31346035648432696,
|
| 1807 |
+
"grad_norm": 0.26896461844444275,
|
| 1808 |
+
"learning_rate": 8.322127733662897e-05,
|
| 1809 |
+
"loss": 1.1373,
|
| 1810 |
+
"step": 255
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 0.3146896127842655,
|
| 1814 |
+
"grad_norm": 0.29333245754241943,
|
| 1815 |
+
"learning_rate": 8.306713749990072e-05,
|
| 1816 |
+
"loss": 1.0615,
|
| 1817 |
+
"step": 256
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 0.3159188690842041,
|
| 1821 |
+
"grad_norm": 0.2958793640136719,
|
| 1822 |
+
"learning_rate": 8.291243707148048e-05,
|
| 1823 |
+
"loss": 0.9392,
|
| 1824 |
+
"step": 257
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 0.3171481253841426,
|
| 1828 |
+
"grad_norm": 0.3320540487766266,
|
| 1829 |
+
"learning_rate": 8.275717867402575e-05,
|
| 1830 |
+
"loss": 1.2935,
|
| 1831 |
+
"step": 258
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 0.31837738168408114,
|
| 1835 |
+
"grad_norm": 0.3567339777946472,
|
| 1836 |
+
"learning_rate": 8.260136493965326e-05,
|
| 1837 |
+
"loss": 1.0954,
|
| 1838 |
+
"step": 259
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 0.31960663798401967,
|
| 1842 |
+
"grad_norm": 0.38393881916999817,
|
| 1843 |
+
"learning_rate": 8.244499850989452e-05,
|
| 1844 |
+
"loss": 1.045,
|
| 1845 |
+
"step": 260
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 0.3208358942839582,
|
| 1849 |
+
"grad_norm": 0.41993001103401184,
|
| 1850 |
+
"learning_rate": 8.228808203565095e-05,
|
| 1851 |
+
"loss": 1.2225,
|
| 1852 |
+
"step": 261
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 0.32206515058389673,
|
| 1856 |
+
"grad_norm": 0.6547941565513611,
|
| 1857 |
+
"learning_rate": 8.213061817714893e-05,
|
| 1858 |
+
"loss": 0.9286,
|
| 1859 |
+
"step": 262
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.32329440688383526,
|
| 1863 |
+
"grad_norm": 0.7117279767990112,
|
| 1864 |
+
"learning_rate": 8.197260960389474e-05,
|
| 1865 |
+
"loss": 0.5088,
|
| 1866 |
+
"step": 263
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 0.3245236631837738,
|
| 1870 |
+
"grad_norm": 0.7041743993759155,
|
| 1871 |
+
"learning_rate": 8.181405899462926e-05,
|
| 1872 |
+
"loss": 0.8899,
|
| 1873 |
+
"step": 264
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 0.3257529194837124,
|
| 1877 |
+
"grad_norm": 0.7142787575721741,
|
| 1878 |
+
"learning_rate": 8.16549690372826e-05,
|
| 1879 |
+
"loss": 0.7447,
|
| 1880 |
+
"step": 265
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 0.3269821757836509,
|
| 1884 |
+
"grad_norm": 0.8879908323287964,
|
| 1885 |
+
"learning_rate": 8.14953424289285e-05,
|
| 1886 |
+
"loss": 1.2607,
|
| 1887 |
+
"step": 266
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 0.32821143208358944,
|
| 1891 |
+
"grad_norm": 0.9387282133102417,
|
| 1892 |
+
"learning_rate": 8.133518187573862e-05,
|
| 1893 |
+
"loss": 1.1611,
|
| 1894 |
+
"step": 267
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 0.32944068838352797,
|
| 1898 |
+
"grad_norm": 1.4039078950881958,
|
| 1899 |
+
"learning_rate": 8.117449009293668e-05,
|
| 1900 |
+
"loss": 0.9947,
|
| 1901 |
+
"step": 268
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 0.3306699446834665,
|
| 1905 |
+
"grad_norm": 3.3686740398406982,
|
| 1906 |
+
"learning_rate": 8.101326980475237e-05,
|
| 1907 |
+
"loss": 1.0783,
|
| 1908 |
+
"step": 269
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 0.33189920098340503,
|
| 1912 |
+
"grad_norm": 2.8384785652160645,
|
| 1913 |
+
"learning_rate": 8.085152374437525e-05,
|
| 1914 |
+
"loss": 0.9008,
|
| 1915 |
+
"step": 270
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 0.33312845728334356,
|
| 1919 |
+
"grad_norm": 2.453441619873047,
|
| 1920 |
+
"learning_rate": 8.06892546539083e-05,
|
| 1921 |
+
"loss": 0.5504,
|
| 1922 |
+
"step": 271
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 0.3343577135832821,
|
| 1926 |
+
"grad_norm": 2.592667579650879,
|
| 1927 |
+
"learning_rate": 8.052646528432158e-05,
|
| 1928 |
+
"loss": 0.7489,
|
| 1929 |
+
"step": 272
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 0.3355869698832207,
|
| 1933 |
+
"grad_norm": 1.9753395318984985,
|
| 1934 |
+
"learning_rate": 8.036315839540545e-05,
|
| 1935 |
+
"loss": 0.9747,
|
| 1936 |
+
"step": 273
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 0.3368162261831592,
|
| 1940 |
+
"grad_norm": 3.042698860168457,
|
| 1941 |
+
"learning_rate": 8.019933675572389e-05,
|
| 1942 |
+
"loss": 1.6841,
|
| 1943 |
+
"step": 274
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 0.33804548248309774,
|
| 1947 |
+
"grad_norm": 2.4343316555023193,
|
| 1948 |
+
"learning_rate": 8.00350031425675e-05,
|
| 1949 |
+
"loss": 0.869,
|
| 1950 |
+
"step": 275
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 0.33927473878303627,
|
| 1954 |
+
"grad_norm": 0.2026144415140152,
|
| 1955 |
+
"learning_rate": 7.98701603419064e-05,
|
| 1956 |
+
"loss": 0.8867,
|
| 1957 |
+
"step": 276
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 0.3405039950829748,
|
| 1961 |
+
"grad_norm": 0.24370141327381134,
|
| 1962 |
+
"learning_rate": 7.970481114834312e-05,
|
| 1963 |
+
"loss": 1.3135,
|
| 1964 |
+
"step": 277
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 0.34173325138291333,
|
| 1968 |
+
"grad_norm": 0.22894087433815002,
|
| 1969 |
+
"learning_rate": 7.953895836506508e-05,
|
| 1970 |
+
"loss": 1.0986,
|
| 1971 |
+
"step": 278
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 0.34296250768285186,
|
| 1975 |
+
"grad_norm": 0.2533970773220062,
|
| 1976 |
+
"learning_rate": 7.937260480379712e-05,
|
| 1977 |
+
"loss": 1.1821,
|
| 1978 |
+
"step": 279
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 0.3441917639827904,
|
| 1982 |
+
"grad_norm": 0.25789350271224976,
|
| 1983 |
+
"learning_rate": 7.920575328475385e-05,
|
| 1984 |
+
"loss": 1.1414,
|
| 1985 |
+
"step": 280
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 0.345421020282729,
|
| 1989 |
+
"grad_norm": 0.28820541501045227,
|
| 1990 |
+
"learning_rate": 7.903840663659186e-05,
|
| 1991 |
+
"loss": 1.3332,
|
| 1992 |
+
"step": 281
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 0.3466502765826675,
|
| 1996 |
+
"grad_norm": 0.28611505031585693,
|
| 1997 |
+
"learning_rate": 7.887056769636165e-05,
|
| 1998 |
+
"loss": 1.0901,
|
| 1999 |
+
"step": 282
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 0.34787953288260604,
|
| 2003 |
+
"grad_norm": 0.28022873401641846,
|
| 2004 |
+
"learning_rate": 7.870223930945972e-05,
|
| 2005 |
+
"loss": 0.8461,
|
| 2006 |
+
"step": 283
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 0.34910878918254457,
|
| 2010 |
+
"grad_norm": 0.3246136009693146,
|
| 2011 |
+
"learning_rate": 7.853342432958013e-05,
|
| 2012 |
+
"loss": 0.9325,
|
| 2013 |
+
"step": 284
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 0.3503380454824831,
|
| 2017 |
+
"grad_norm": 0.3149406611919403,
|
| 2018 |
+
"learning_rate": 7.836412561866629e-05,
|
| 2019 |
+
"loss": 1.013,
|
| 2020 |
+
"step": 285
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 0.3515673017824216,
|
| 2024 |
+
"grad_norm": 0.3745490610599518,
|
| 2025 |
+
"learning_rate": 7.819434604686228e-05,
|
| 2026 |
+
"loss": 1.2624,
|
| 2027 |
+
"step": 286
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 0.35279655808236016,
|
| 2031 |
+
"grad_norm": 0.4822925329208374,
|
| 2032 |
+
"learning_rate": 7.802408849246442e-05,
|
| 2033 |
+
"loss": 1.2424,
|
| 2034 |
+
"step": 287
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 0.3540258143822987,
|
| 2038 |
+
"grad_norm": 0.6210641264915466,
|
| 2039 |
+
"learning_rate": 7.785335584187219e-05,
|
| 2040 |
+
"loss": 1.2527,
|
| 2041 |
+
"step": 288
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 0.3552550706822373,
|
| 2045 |
+
"grad_norm": 0.6488444805145264,
|
| 2046 |
+
"learning_rate": 7.768215098953952e-05,
|
| 2047 |
+
"loss": 0.7986,
|
| 2048 |
+
"step": 289
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 0.3564843269821758,
|
| 2052 |
+
"grad_norm": 0.760388195514679,
|
| 2053 |
+
"learning_rate": 7.751047683792561e-05,
|
| 2054 |
+
"loss": 1.0136,
|
| 2055 |
+
"step": 290
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 0.35771358328211433,
|
| 2059 |
+
"grad_norm": 0.7666548490524292,
|
| 2060 |
+
"learning_rate": 7.73383362974458e-05,
|
| 2061 |
+
"loss": 0.8205,
|
| 2062 |
+
"step": 291
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 0.35894283958205286,
|
| 2066 |
+
"grad_norm": 0.7492078542709351,
|
| 2067 |
+
"learning_rate": 7.71657322864221e-05,
|
| 2068 |
+
"loss": 0.9254,
|
| 2069 |
+
"step": 292
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 0.3601720958819914,
|
| 2073 |
+
"grad_norm": 0.9061193466186523,
|
| 2074 |
+
"learning_rate": 7.699266773103389e-05,
|
| 2075 |
+
"loss": 1.3013,
|
| 2076 |
+
"step": 293
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 0.3614013521819299,
|
| 2080 |
+
"grad_norm": 2.1404013633728027,
|
| 2081 |
+
"learning_rate": 7.681914556526817e-05,
|
| 2082 |
+
"loss": 1.5957,
|
| 2083 |
+
"step": 294
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 0.36263060848186845,
|
| 2087 |
+
"grad_norm": 2.647864580154419,
|
| 2088 |
+
"learning_rate": 7.664516873086987e-05,
|
| 2089 |
+
"loss": 1.1658,
|
| 2090 |
+
"step": 295
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 0.363859864781807,
|
| 2094 |
+
"grad_norm": 3.0906460285186768,
|
| 2095 |
+
"learning_rate": 7.647074017729202e-05,
|
| 2096 |
+
"loss": 1.1344,
|
| 2097 |
+
"step": 296
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 0.36508912108174557,
|
| 2101 |
+
"grad_norm": 2.2348814010620117,
|
| 2102 |
+
"learning_rate": 7.629586286164565e-05,
|
| 2103 |
+
"loss": 0.8813,
|
| 2104 |
+
"step": 297
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 0.3663183773816841,
|
| 2108 |
+
"grad_norm": 2.937446117401123,
|
| 2109 |
+
"learning_rate": 7.612053974864976e-05,
|
| 2110 |
+
"loss": 1.0414,
|
| 2111 |
+
"step": 298
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 0.36754763368162263,
|
| 2115 |
+
"grad_norm": 2.5343546867370605,
|
| 2116 |
+
"learning_rate": 7.594477381058098e-05,
|
| 2117 |
+
"loss": 1.1847,
|
| 2118 |
+
"step": 299
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 0.36877688998156116,
|
| 2122 |
+
"grad_norm": 2.8971638679504395,
|
| 2123 |
+
"learning_rate": 7.576856802722325e-05,
|
| 2124 |
+
"loss": 0.9029,
|
| 2125 |
+
"step": 300
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 0.3700061462814997,
|
| 2129 |
+
"grad_norm": 0.1982557773590088,
|
| 2130 |
+
"learning_rate": 7.559192538581722e-05,
|
| 2131 |
+
"loss": 0.9314,
|
| 2132 |
+
"step": 301
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 0.3712354025814382,
|
| 2136 |
+
"grad_norm": 0.24721381068229675,
|
| 2137 |
+
"learning_rate": 7.541484888100974e-05,
|
| 2138 |
+
"loss": 1.2432,
|
| 2139 |
+
"step": 302
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 0.37246465888137675,
|
| 2143 |
+
"grad_norm": 0.24999506771564484,
|
| 2144 |
+
"learning_rate": 7.523734151480289e-05,
|
| 2145 |
+
"loss": 1.285,
|
| 2146 |
+
"step": 303
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 0.3736939151813153,
|
| 2150 |
+
"grad_norm": 0.267764151096344,
|
| 2151 |
+
"learning_rate": 7.505940629650326e-05,
|
| 2152 |
+
"loss": 1.198,
|
| 2153 |
+
"step": 304
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 0.3749231714812538,
|
| 2157 |
+
"grad_norm": 0.26003679633140564,
|
| 2158 |
+
"learning_rate": 7.488104624267091e-05,
|
| 2159 |
+
"loss": 1.2001,
|
| 2160 |
+
"step": 305
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 0.3761524277811924,
|
| 2164 |
+
"grad_norm": 0.28197526931762695,
|
| 2165 |
+
"learning_rate": 7.470226437706813e-05,
|
| 2166 |
+
"loss": 1.1687,
|
| 2167 |
+
"step": 306
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"epoch": 0.37738168408113093,
|
| 2171 |
+
"grad_norm": 0.29367661476135254,
|
| 2172 |
+
"learning_rate": 7.452306373060829e-05,
|
| 2173 |
+
"loss": 1.211,
|
| 2174 |
+
"step": 307
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 0.37861094038106946,
|
| 2178 |
+
"grad_norm": 0.2982727885246277,
|
| 2179 |
+
"learning_rate": 7.434344734130437e-05,
|
| 2180 |
+
"loss": 1.151,
|
| 2181 |
+
"step": 308
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 0.379840196681008,
|
| 2185 |
+
"grad_norm": 0.3283758759498596,
|
| 2186 |
+
"learning_rate": 7.416341825421754e-05,
|
| 2187 |
+
"loss": 0.9875,
|
| 2188 |
+
"step": 309
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"epoch": 0.3810694529809465,
|
| 2192 |
+
"grad_norm": 0.32420334219932556,
|
| 2193 |
+
"learning_rate": 7.398297952140544e-05,
|
| 2194 |
+
"loss": 1.0796,
|
| 2195 |
+
"step": 310
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 0.38229870928088505,
|
| 2199 |
+
"grad_norm": 0.4046980142593384,
|
| 2200 |
+
"learning_rate": 7.380213420187055e-05,
|
| 2201 |
+
"loss": 1.1158,
|
| 2202 |
+
"step": 311
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 0.3835279655808236,
|
| 2206 |
+
"grad_norm": 0.391736775636673,
|
| 2207 |
+
"learning_rate": 7.36208853615082e-05,
|
| 2208 |
+
"loss": 1.1682,
|
| 2209 |
+
"step": 312
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 0.3847572218807621,
|
| 2213 |
+
"grad_norm": 0.6027556657791138,
|
| 2214 |
+
"learning_rate": 7.343923607305471e-05,
|
| 2215 |
+
"loss": 1.0696,
|
| 2216 |
+
"step": 313
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 0.3859864781807007,
|
| 2220 |
+
"grad_norm": 0.6483603119850159,
|
| 2221 |
+
"learning_rate": 7.325718941603527e-05,
|
| 2222 |
+
"loss": 0.7843,
|
| 2223 |
+
"step": 314
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 0.38721573448063923,
|
| 2227 |
+
"grad_norm": 0.6711483001708984,
|
| 2228 |
+
"learning_rate": 7.307474847671168e-05,
|
| 2229 |
+
"loss": 0.7247,
|
| 2230 |
+
"step": 315
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 0.38844499078057776,
|
| 2234 |
+
"grad_norm": 0.7372632026672363,
|
| 2235 |
+
"learning_rate": 7.289191634803003e-05,
|
| 2236 |
+
"loss": 1.0535,
|
| 2237 |
+
"step": 316
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 0.3896742470805163,
|
| 2241 |
+
"grad_norm": 0.7427420020103455,
|
| 2242 |
+
"learning_rate": 7.270869612956835e-05,
|
| 2243 |
+
"loss": 1.0563,
|
| 2244 |
+
"step": 317
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 0.3909035033804548,
|
| 2248 |
+
"grad_norm": 2.6449501514434814,
|
| 2249 |
+
"learning_rate": 7.252509092748401e-05,
|
| 2250 |
+
"loss": 1.3099,
|
| 2251 |
+
"step": 318
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 0.39213275968039335,
|
| 2255 |
+
"grad_norm": 3.1938464641571045,
|
| 2256 |
+
"learning_rate": 7.234110385446103e-05,
|
| 2257 |
+
"loss": 1.1728,
|
| 2258 |
+
"step": 319
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 0.3933620159803319,
|
| 2262 |
+
"grad_norm": 2.584103584289551,
|
| 2263 |
+
"learning_rate": 7.215673802965734e-05,
|
| 2264 |
+
"loss": 0.792,
|
| 2265 |
+
"step": 320
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 0.3945912722802704,
|
| 2269 |
+
"grad_norm": 2.358025074005127,
|
| 2270 |
+
"learning_rate": 7.197199657865195e-05,
|
| 2271 |
+
"loss": 1.0462,
|
| 2272 |
+
"step": 321
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 0.395820528580209,
|
| 2276 |
+
"grad_norm": 2.9621617794036865,
|
| 2277 |
+
"learning_rate": 7.178688263339184e-05,
|
| 2278 |
+
"loss": 1.4222,
|
| 2279 |
+
"step": 322
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 0.3970497848801475,
|
| 2283 |
+
"grad_norm": 2.5362660884857178,
|
| 2284 |
+
"learning_rate": 7.160139933213898e-05,
|
| 2285 |
+
"loss": 1.1527,
|
| 2286 |
+
"step": 323
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 0.39827904118008606,
|
| 2290 |
+
"grad_norm": 2.4901375770568848,
|
| 2291 |
+
"learning_rate": 7.141554981941709e-05,
|
| 2292 |
+
"loss": 1.1712,
|
| 2293 |
+
"step": 324
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 0.3995082974800246,
|
| 2297 |
+
"grad_norm": 2.9214236736297607,
|
| 2298 |
+
"learning_rate": 7.12293372459583e-05,
|
| 2299 |
+
"loss": 1.1977,
|
| 2300 |
+
"step": 325
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 0.4007375537799631,
|
| 2304 |
+
"grad_norm": 0.24753543734550476,
|
| 2305 |
+
"learning_rate": 7.104276476864974e-05,
|
| 2306 |
+
"loss": 1.2176,
|
| 2307 |
+
"step": 326
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 0.40196681007990165,
|
| 2311 |
+
"grad_norm": 0.25986090302467346,
|
| 2312 |
+
"learning_rate": 7.085583555048008e-05,
|
| 2313 |
+
"loss": 1.2854,
|
| 2314 |
+
"step": 327
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 0.4031960663798402,
|
| 2318 |
+
"grad_norm": 0.2640175521373749,
|
| 2319 |
+
"learning_rate": 7.066855276048587e-05,
|
| 2320 |
+
"loss": 1.2204,
|
| 2321 |
+
"step": 328
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 0.4044253226797787,
|
| 2325 |
+
"grad_norm": 0.2603614330291748,
|
| 2326 |
+
"learning_rate": 7.048091957369776e-05,
|
| 2327 |
+
"loss": 1.2621,
|
| 2328 |
+
"step": 329
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 0.4056545789797173,
|
| 2332 |
+
"grad_norm": 0.2921195924282074,
|
| 2333 |
+
"learning_rate": 7.029293917108678e-05,
|
| 2334 |
+
"loss": 1.281,
|
| 2335 |
+
"step": 330
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 0.4068838352796558,
|
| 2339 |
+
"grad_norm": 0.2984941899776459,
|
| 2340 |
+
"learning_rate": 7.010461473951033e-05,
|
| 2341 |
+
"loss": 1.071,
|
| 2342 |
+
"step": 331
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 0.40811309157959436,
|
| 2346 |
+
"grad_norm": 0.31219175457954407,
|
| 2347 |
+
"learning_rate": 6.991594947165818e-05,
|
| 2348 |
+
"loss": 1.3161,
|
| 2349 |
+
"step": 332
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 0.4093423478795329,
|
| 2353 |
+
"grad_norm": 0.31329602003097534,
|
| 2354 |
+
"learning_rate": 6.972694656599834e-05,
|
| 2355 |
+
"loss": 0.9854,
|
| 2356 |
+
"step": 333
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 0.4105716041794714,
|
| 2360 |
+
"grad_norm": 0.3356671929359436,
|
| 2361 |
+
"learning_rate": 6.953760922672286e-05,
|
| 2362 |
+
"loss": 1.02,
|
| 2363 |
+
"step": 334
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 0.41180086047940995,
|
| 2367 |
+
"grad_norm": 0.3843994438648224,
|
| 2368 |
+
"learning_rate": 6.934794066369348e-05,
|
| 2369 |
+
"loss": 1.2173,
|
| 2370 |
+
"step": 335
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 0.4130301167793485,
|
| 2374 |
+
"grad_norm": 0.45338544249534607,
|
| 2375 |
+
"learning_rate": 6.915794409238718e-05,
|
| 2376 |
+
"loss": 1.3614,
|
| 2377 |
+
"step": 336
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 0.414259373079287,
|
| 2381 |
+
"grad_norm": 0.4857298731803894,
|
| 2382 |
+
"learning_rate": 6.896762273384178e-05,
|
| 2383 |
+
"loss": 1.0175,
|
| 2384 |
+
"step": 337
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 0.4154886293792256,
|
| 2388 |
+
"grad_norm": 0.6512896418571472,
|
| 2389 |
+
"learning_rate": 6.877697981460125e-05,
|
| 2390 |
+
"loss": 0.6555,
|
| 2391 |
+
"step": 338
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 0.4167178856791641,
|
| 2395 |
+
"grad_norm": 0.6744720935821533,
|
| 2396 |
+
"learning_rate": 6.858601856666094e-05,
|
| 2397 |
+
"loss": 0.6057,
|
| 2398 |
+
"step": 339
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 0.41794714197910265,
|
| 2402 |
+
"grad_norm": 0.6527014374732971,
|
| 2403 |
+
"learning_rate": 6.839474222741299e-05,
|
| 2404 |
+
"loss": 0.9116,
|
| 2405 |
+
"step": 340
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 0.4191763982790412,
|
| 2409 |
+
"grad_norm": 0.6935631036758423,
|
| 2410 |
+
"learning_rate": 6.820315403959123e-05,
|
| 2411 |
+
"loss": 0.9876,
|
| 2412 |
+
"step": 341
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 0.4204056545789797,
|
| 2416 |
+
"grad_norm": 0.6856899261474609,
|
| 2417 |
+
"learning_rate": 6.801125725121636e-05,
|
| 2418 |
+
"loss": 0.9591,
|
| 2419 |
+
"step": 342
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 0.42163491087891825,
|
| 2423 |
+
"grad_norm": 1.2577812671661377,
|
| 2424 |
+
"learning_rate": 6.781905511554079e-05,
|
| 2425 |
+
"loss": 1.3174,
|
| 2426 |
+
"step": 343
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 0.4228641671788568,
|
| 2430 |
+
"grad_norm": 2.421950578689575,
|
| 2431 |
+
"learning_rate": 6.762655089099353e-05,
|
| 2432 |
+
"loss": 1.6442,
|
| 2433 |
+
"step": 344
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 0.4240934234787953,
|
| 2437 |
+
"grad_norm": 2.6432454586029053,
|
| 2438 |
+
"learning_rate": 6.743374784112501e-05,
|
| 2439 |
+
"loss": 1.0468,
|
| 2440 |
+
"step": 345
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 0.4253226797787339,
|
| 2444 |
+
"grad_norm": 2.7061827182769775,
|
| 2445 |
+
"learning_rate": 6.724064923455155e-05,
|
| 2446 |
+
"loss": 1.1526,
|
| 2447 |
+
"step": 346
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 0.4265519360786724,
|
| 2451 |
+
"grad_norm": 2.466057777404785,
|
| 2452 |
+
"learning_rate": 6.704725834490024e-05,
|
| 2453 |
+
"loss": 1.1463,
|
| 2454 |
+
"step": 347
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 0.42778119237861095,
|
| 2458 |
+
"grad_norm": 2.753512144088745,
|
| 2459 |
+
"learning_rate": 6.685357845075315e-05,
|
| 2460 |
+
"loss": 0.9492,
|
| 2461 |
+
"step": 348
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 0.4290104486785495,
|
| 2465 |
+
"grad_norm": 2.76118803024292,
|
| 2466 |
+
"learning_rate": 6.665961283559197e-05,
|
| 2467 |
+
"loss": 0.8543,
|
| 2468 |
+
"step": 349
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 0.430239704978488,
|
| 2472 |
+
"grad_norm": 2.295574426651001,
|
| 2473 |
+
"learning_rate": 6.646536478774222e-05,
|
| 2474 |
+
"loss": 0.9564,
|
| 2475 |
+
"step": 350
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 0.43146896127842654,
|
| 2479 |
+
"grad_norm": 0.22731368243694305,
|
| 2480 |
+
"learning_rate": 6.627083760031754e-05,
|
| 2481 |
+
"loss": 0.9719,
|
| 2482 |
+
"step": 351
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 0.4326982175783651,
|
| 2486 |
+
"grad_norm": 0.20097078382968903,
|
| 2487 |
+
"learning_rate": 6.60760345711639e-05,
|
| 2488 |
+
"loss": 1.0094,
|
| 2489 |
+
"step": 352
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 0.4339274738783036,
|
| 2493 |
+
"grad_norm": 0.23321934044361115,
|
| 2494 |
+
"learning_rate": 6.58809590028036e-05,
|
| 2495 |
+
"loss": 1.101,
|
| 2496 |
+
"step": 353
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 0.4351567301782422,
|
| 2500 |
+
"grad_norm": 0.27995625138282776,
|
| 2501 |
+
"learning_rate": 6.568561420237935e-05,
|
| 2502 |
+
"loss": 1.3545,
|
| 2503 |
+
"step": 354
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 0.4363859864781807,
|
| 2507 |
+
"grad_norm": 0.259082168340683,
|
| 2508 |
+
"learning_rate": 6.54900034815982e-05,
|
| 2509 |
+
"loss": 1.1598,
|
| 2510 |
+
"step": 355
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 0.43761524277811925,
|
| 2514 |
+
"grad_norm": 0.2688703238964081,
|
| 2515 |
+
"learning_rate": 6.52941301566754e-05,
|
| 2516 |
+
"loss": 1.1141,
|
| 2517 |
+
"step": 356
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 0.4388444990780578,
|
| 2521 |
+
"grad_norm": 0.34018442034721375,
|
| 2522 |
+
"learning_rate": 6.50979975482781e-05,
|
| 2523 |
+
"loss": 1.2811,
|
| 2524 |
+
"step": 357
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 0.4400737553779963,
|
| 2528 |
+
"grad_norm": 0.2925175130367279,
|
| 2529 |
+
"learning_rate": 6.490160898146918e-05,
|
| 2530 |
+
"loss": 0.9025,
|
| 2531 |
+
"step": 358
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 0.44130301167793484,
|
| 2535 |
+
"grad_norm": 0.30208972096443176,
|
| 2536 |
+
"learning_rate": 6.470496778565082e-05,
|
| 2537 |
+
"loss": 1.0301,
|
| 2538 |
+
"step": 359
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 0.4425322679778734,
|
| 2542 |
+
"grad_norm": 0.3110770285129547,
|
| 2543 |
+
"learning_rate": 6.4508077294508e-05,
|
| 2544 |
+
"loss": 1.0911,
|
| 2545 |
+
"step": 360
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 0.4437615242778119,
|
| 2549 |
+
"grad_norm": 0.426252543926239,
|
| 2550 |
+
"learning_rate": 6.431094084595209e-05,
|
| 2551 |
+
"loss": 1.1214,
|
| 2552 |
+
"step": 361
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 0.44499078057775043,
|
| 2556 |
+
"grad_norm": 0.4019356966018677,
|
| 2557 |
+
"learning_rate": 6.411356178206419e-05,
|
| 2558 |
+
"loss": 1.3063,
|
| 2559 |
+
"step": 362
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 0.446220036877689,
|
| 2563 |
+
"grad_norm": 0.4622703194618225,
|
| 2564 |
+
"learning_rate": 6.391594344903848e-05,
|
| 2565 |
+
"loss": 1.1208,
|
| 2566 |
+
"step": 363
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 0.44744929317762755,
|
| 2570 |
+
"grad_norm": 0.5752270817756653,
|
| 2571 |
+
"learning_rate": 6.371808919712549e-05,
|
| 2572 |
+
"loss": 0.9653,
|
| 2573 |
+
"step": 364
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 0.4486785494775661,
|
| 2577 |
+
"grad_norm": 0.6309720277786255,
|
| 2578 |
+
"learning_rate": 6.35200023805754e-05,
|
| 2579 |
+
"loss": 0.5664,
|
| 2580 |
+
"step": 365
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 0.4499078057775046,
|
| 2584 |
+
"grad_norm": 0.612684965133667,
|
| 2585 |
+
"learning_rate": 6.332168635758097e-05,
|
| 2586 |
+
"loss": 1.0443,
|
| 2587 |
+
"step": 366
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 0.45113706207744314,
|
| 2591 |
+
"grad_norm": 0.6797056794166565,
|
| 2592 |
+
"learning_rate": 6.31231444902208e-05,
|
| 2593 |
+
"loss": 0.8389,
|
| 2594 |
+
"step": 367
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 0.45236631837738167,
|
| 2598 |
+
"grad_norm": 1.222960352897644,
|
| 2599 |
+
"learning_rate": 6.292438014440227e-05,
|
| 2600 |
+
"loss": 1.4688,
|
| 2601 |
+
"step": 368
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 0.4535955746773202,
|
| 2605 |
+
"grad_norm": 2.9443516731262207,
|
| 2606 |
+
"learning_rate": 6.272539668980441e-05,
|
| 2607 |
+
"loss": 1.0079,
|
| 2608 |
+
"step": 369
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 0.45482483097725873,
|
| 2612 |
+
"grad_norm": 3.0168612003326416,
|
| 2613 |
+
"learning_rate": 6.252619749982089e-05,
|
| 2614 |
+
"loss": 0.9232,
|
| 2615 |
+
"step": 370
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 0.4560540872771973,
|
| 2619 |
+
"grad_norm": 1.9470983743667603,
|
| 2620 |
+
"learning_rate": 6.232678595150275e-05,
|
| 2621 |
+
"loss": 0.8126,
|
| 2622 |
+
"step": 371
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 0.45728334357713585,
|
| 2626 |
+
"grad_norm": 2.4769911766052246,
|
| 2627 |
+
"learning_rate": 6.212716542550112e-05,
|
| 2628 |
+
"loss": 0.7786,
|
| 2629 |
+
"step": 372
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 0.4585125998770744,
|
| 2633 |
+
"grad_norm": 2.849158525466919,
|
| 2634 |
+
"learning_rate": 6.192733930601005e-05,
|
| 2635 |
+
"loss": 1.1914,
|
| 2636 |
+
"step": 373
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 0.4597418561770129,
|
| 2640 |
+
"grad_norm": 2.6154119968414307,
|
| 2641 |
+
"learning_rate": 6.172731098070899e-05,
|
| 2642 |
+
"loss": 0.9171,
|
| 2643 |
+
"step": 374
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 0.46097111247695144,
|
| 2647 |
+
"grad_norm": 3.5901479721069336,
|
| 2648 |
+
"learning_rate": 6.152708384070541e-05,
|
| 2649 |
+
"loss": 1.1269,
|
| 2650 |
+
"step": 375
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 0.46220036877688997,
|
| 2654 |
+
"grad_norm": 0.23536159098148346,
|
| 2655 |
+
"learning_rate": 6.132666128047732e-05,
|
| 2656 |
+
"loss": 0.8768,
|
| 2657 |
+
"step": 376
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 0.4634296250768285,
|
| 2661 |
+
"grad_norm": 0.24834086000919342,
|
| 2662 |
+
"learning_rate": 6.112604669781572e-05,
|
| 2663 |
+
"loss": 1.0644,
|
| 2664 |
+
"step": 377
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 0.46465888137676703,
|
| 2668 |
+
"grad_norm": 0.3041445016860962,
|
| 2669 |
+
"learning_rate": 6.0925243493767016e-05,
|
| 2670 |
+
"loss": 1.2779,
|
| 2671 |
+
"step": 378
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 0.4658881376767056,
|
| 2675 |
+
"grad_norm": 0.3158765137195587,
|
| 2676 |
+
"learning_rate": 6.0724255072575275e-05,
|
| 2677 |
+
"loss": 1.352,
|
| 2678 |
+
"step": 379
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 0.46711739397664415,
|
| 2682 |
+
"grad_norm": 0.2845201790332794,
|
| 2683 |
+
"learning_rate": 6.0523084841624635e-05,
|
| 2684 |
+
"loss": 1.2567,
|
| 2685 |
+
"step": 380
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 0.4683466502765827,
|
| 2689 |
+
"grad_norm": 0.2909673750400543,
|
| 2690 |
+
"learning_rate": 6.0321736211381464e-05,
|
| 2691 |
+
"loss": 1.1735,
|
| 2692 |
+
"step": 381
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 0.4695759065765212,
|
| 2696 |
+
"grad_norm": 0.2946690022945404,
|
| 2697 |
+
"learning_rate": 6.0120212595336545e-05,
|
| 2698 |
+
"loss": 1.1514,
|
| 2699 |
+
"step": 382
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 0.47080516287645974,
|
| 2703 |
+
"grad_norm": 0.302846223115921,
|
| 2704 |
+
"learning_rate": 5.9918517409947215e-05,
|
| 2705 |
+
"loss": 1.0621,
|
| 2706 |
+
"step": 383
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 0.47203441917639827,
|
| 2710 |
+
"grad_norm": 0.3197241425514221,
|
| 2711 |
+
"learning_rate": 5.971665407457948e-05,
|
| 2712 |
+
"loss": 1.0299,
|
| 2713 |
+
"step": 384
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 0.4732636754763368,
|
| 2717 |
+
"grad_norm": 0.342777281999588,
|
| 2718 |
+
"learning_rate": 5.951462601144998e-05,
|
| 2719 |
+
"loss": 1.0858,
|
| 2720 |
+
"step": 385
|
| 2721 |
+
},
|
| 2722 |
+
{
|
| 2723 |
+
"epoch": 0.47449293177627533,
|
| 2724 |
+
"grad_norm": 0.3554008901119232,
|
| 2725 |
+
"learning_rate": 5.931243664556803e-05,
|
| 2726 |
+
"loss": 1.1441,
|
| 2727 |
+
"step": 386
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 0.4757221880762139,
|
| 2731 |
+
"grad_norm": 0.36057665944099426,
|
| 2732 |
+
"learning_rate": 5.9110089404677524e-05,
|
| 2733 |
+
"loss": 1.1836,
|
| 2734 |
+
"step": 387
|
| 2735 |
+
},
|
| 2736 |
+
{
|
| 2737 |
+
"epoch": 0.47695144437615244,
|
| 2738 |
+
"grad_norm": 0.5004509091377258,
|
| 2739 |
+
"learning_rate": 5.890758771919884e-05,
|
| 2740 |
+
"loss": 1.4109,
|
| 2741 |
+
"step": 388
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 0.478180700676091,
|
| 2745 |
+
"grad_norm": 0.5211744904518127,
|
| 2746 |
+
"learning_rate": 5.8704935022170684e-05,
|
| 2747 |
+
"loss": 1.0097,
|
| 2748 |
+
"step": 389
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 0.4794099569760295,
|
| 2752 |
+
"grad_norm": 0.7474620938301086,
|
| 2753 |
+
"learning_rate": 5.8502134749191816e-05,
|
| 2754 |
+
"loss": 0.8777,
|
| 2755 |
+
"step": 390
|
| 2756 |
+
},
|
| 2757 |
+
{
|
| 2758 |
+
"epoch": 0.48063921327596804,
|
| 2759 |
+
"grad_norm": 0.7044636011123657,
|
| 2760 |
+
"learning_rate": 5.8299190338362996e-05,
|
| 2761 |
+
"loss": 0.9007,
|
| 2762 |
+
"step": 391
|
| 2763 |
+
},
|
| 2764 |
+
{
|
| 2765 |
+
"epoch": 0.48186846957590657,
|
| 2766 |
+
"grad_norm": 0.6484948992729187,
|
| 2767 |
+
"learning_rate": 5.8096105230228435e-05,
|
| 2768 |
+
"loss": 0.8261,
|
| 2769 |
+
"step": 392
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 0.4830977258758451,
|
| 2773 |
+
"grad_norm": 0.672816812992096,
|
| 2774 |
+
"learning_rate": 5.78928828677177e-05,
|
| 2775 |
+
"loss": 1.0531,
|
| 2776 |
+
"step": 393
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 0.4843269821757836,
|
| 2780 |
+
"grad_norm": 1.1637803316116333,
|
| 2781 |
+
"learning_rate": 5.768952669608724e-05,
|
| 2782 |
+
"loss": 1.1586,
|
| 2783 |
+
"step": 394
|
| 2784 |
+
},
|
| 2785 |
+
{
|
| 2786 |
+
"epoch": 0.4855562384757222,
|
| 2787 |
+
"grad_norm": 3.1862003803253174,
|
| 2788 |
+
"learning_rate": 5.748604016286192e-05,
|
| 2789 |
+
"loss": 1.6232,
|
| 2790 |
+
"step": 395
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 0.48678549477566074,
|
| 2794 |
+
"grad_norm": 3.3833253383636475,
|
| 2795 |
+
"learning_rate": 5.728242671777672e-05,
|
| 2796 |
+
"loss": 1.0918,
|
| 2797 |
+
"step": 396
|
| 2798 |
+
},
|
| 2799 |
+
{
|
| 2800 |
+
"epoch": 0.4880147510755993,
|
| 2801 |
+
"grad_norm": 3.116319417953491,
|
| 2802 |
+
"learning_rate": 5.707868981271815e-05,
|
| 2803 |
+
"loss": 0.8615,
|
| 2804 |
+
"step": 397
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 0.4892440073755378,
|
| 2808 |
+
"grad_norm": 2.5967965126037598,
|
| 2809 |
+
"learning_rate": 5.687483290166573e-05,
|
| 2810 |
+
"loss": 0.8579,
|
| 2811 |
+
"step": 398
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 0.49047326367547633,
|
| 2815 |
+
"grad_norm": 3.7683048248291016,
|
| 2816 |
+
"learning_rate": 5.6670859440633486e-05,
|
| 2817 |
+
"loss": 1.0777,
|
| 2818 |
+
"step": 399
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 0.49170251997541486,
|
| 2822 |
+
"grad_norm": 3.182317018508911,
|
| 2823 |
+
"learning_rate": 5.646677288761132e-05,
|
| 2824 |
+
"loss": 0.8932,
|
| 2825 |
+
"step": 400
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 0.4929317762753534,
|
| 2829 |
+
"grad_norm": 0.19372360408306122,
|
| 2830 |
+
"learning_rate": 5.6262576702506406e-05,
|
| 2831 |
+
"loss": 0.8516,
|
| 2832 |
+
"step": 401
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 0.4941610325752919,
|
| 2836 |
+
"grad_norm": 0.223338320851326,
|
| 2837 |
+
"learning_rate": 5.6058274347084504e-05,
|
| 2838 |
+
"loss": 1.1287,
|
| 2839 |
+
"step": 402
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 0.4953902888752305,
|
| 2843 |
+
"grad_norm": 0.23367400467395782,
|
| 2844 |
+
"learning_rate": 5.585386928491134e-05,
|
| 2845 |
+
"loss": 1.1128,
|
| 2846 |
+
"step": 403
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 0.49661954517516904,
|
| 2850 |
+
"grad_norm": 0.2717371881008148,
|
| 2851 |
+
"learning_rate": 5.5649364981293786e-05,
|
| 2852 |
+
"loss": 1.2813,
|
| 2853 |
+
"step": 404
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 0.49784880147510757,
|
| 2857 |
+
"grad_norm": 0.25909724831581116,
|
| 2858 |
+
"learning_rate": 5.54447649032212e-05,
|
| 2859 |
+
"loss": 1.2149,
|
| 2860 |
+
"step": 405
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 0.4990780577750461,
|
| 2864 |
+
"grad_norm": 0.25411197543144226,
|
| 2865 |
+
"learning_rate": 5.5240072519306606e-05,
|
| 2866 |
+
"loss": 1.0679,
|
| 2867 |
+
"step": 406
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 0.5003073140749846,
|
| 2871 |
+
"grad_norm": 0.2962128520011902,
|
| 2872 |
+
"learning_rate": 5.503529129972792e-05,
|
| 2873 |
+
"loss": 1.1156,
|
| 2874 |
+
"step": 407
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 0.5015365703749232,
|
| 2878 |
+
"grad_norm": 0.29252949357032776,
|
| 2879 |
+
"learning_rate": 5.483042471616908e-05,
|
| 2880 |
+
"loss": 1.125,
|
| 2881 |
+
"step": 408
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 0.5015365703749232,
|
| 2885 |
+
"eval_loss": 0.9695894122123718,
|
| 2886 |
+
"eval_runtime": 65.3254,
|
| 2887 |
+
"eval_samples_per_second": 10.486,
|
| 2888 |
+
"eval_steps_per_second": 5.251,
|
| 2889 |
+
"step": 408
|
| 2890 |
}
|
| 2891 |
],
|
| 2892 |
"logging_steps": 1,
|
|
|
|
| 2906 |
"attributes": {}
|
| 2907 |
}
|
| 2908 |
},
|
| 2909 |
+
"total_flos": 8.411229584359424e+16,
|
| 2910 |
"train_batch_size": 1,
|
| 2911 |
"trial_name": null,
|
| 2912 |
"trial_params": null
|