Upload 10 files
Browse files- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +3045 -3
- training_args.bin +1 -1
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 598635032
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88686ba0b6f9018075e16e833e34f3f004b964bcf7967bf8369b4dbf247c40fb
|
| 3 |
size 598635032
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1197359627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:40aaa4a3ba27ad6081e57d1921e2ebd4f5848dcec7b2f4b3574f9612e4a102fb
|
| 3 |
size 1197359627
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:894d7385052854ff6333edf420f03929d421cabce8f7a3233655bfaec3d667fc
|
| 3 |
size 14645
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2864175f72493ab2957d7cfc2953c788bbc11e28fc49502d93a74ed5d700c286
|
| 3 |
size 1465
|
trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 1000,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -4071,6 +4071,3048 @@
|
|
| 4071 |
"eval_samples_per_second": 202.032,
|
| 4072 |
"eval_steps_per_second": 3.166,
|
| 4073 |
"step": 52000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4074 |
}
|
| 4075 |
],
|
| 4076 |
"logging_steps": 100,
|
|
@@ -4090,7 +7132,7 @@
|
|
| 4090 |
"attributes": {}
|
| 4091 |
}
|
| 4092 |
},
|
| 4093 |
-
"total_flos":
|
| 4094 |
"train_batch_size": 64,
|
| 4095 |
"trial_name": null,
|
| 4096 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.062,
|
| 6 |
"eval_steps": 1000,
|
| 7 |
+
"global_step": 91000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 4071 |
"eval_samples_per_second": 202.032,
|
| 4072 |
"eval_steps_per_second": 3.166,
|
| 4073 |
"step": 52000
|
| 4074 |
+
},
|
| 4075 |
+
{
|
| 4076 |
+
"epoch": 0.0442,
|
| 4077 |
+
"grad_norm": 1.2689367532730103,
|
| 4078 |
+
"learning_rate": 4.889137031740717e-05,
|
| 4079 |
+
"loss": 1.2189,
|
| 4080 |
+
"step": 52100
|
| 4081 |
+
},
|
| 4082 |
+
{
|
| 4083 |
+
"epoch": 0.0444,
|
| 4084 |
+
"grad_norm": 1.0029367208480835,
|
| 4085 |
+
"learning_rate": 4.888669295522137e-05,
|
| 4086 |
+
"loss": 1.1754,
|
| 4087 |
+
"step": 52200
|
| 4088 |
+
},
|
| 4089 |
+
{
|
| 4090 |
+
"epoch": 0.0446,
|
| 4091 |
+
"grad_norm": 0.6958720684051514,
|
| 4092 |
+
"learning_rate": 4.8882005971478504e-05,
|
| 4093 |
+
"loss": 1.1601,
|
| 4094 |
+
"step": 52300
|
| 4095 |
+
},
|
| 4096 |
+
{
|
| 4097 |
+
"epoch": 0.0448,
|
| 4098 |
+
"grad_norm": 1.2337570190429688,
|
| 4099 |
+
"learning_rate": 4.887730936806648e-05,
|
| 4100 |
+
"loss": 1.2244,
|
| 4101 |
+
"step": 52400
|
| 4102 |
+
},
|
| 4103 |
+
{
|
| 4104 |
+
"epoch": 0.045,
|
| 4105 |
+
"grad_norm": 1.2311972379684448,
|
| 4106 |
+
"learning_rate": 4.8872603146877104e-05,
|
| 4107 |
+
"loss": 1.2031,
|
| 4108 |
+
"step": 52500
|
| 4109 |
+
},
|
| 4110 |
+
{
|
| 4111 |
+
"epoch": 0.0452,
|
| 4112 |
+
"grad_norm": 1.145331859588623,
|
| 4113 |
+
"learning_rate": 4.886788730980604e-05,
|
| 4114 |
+
"loss": 1.1947,
|
| 4115 |
+
"step": 52600
|
| 4116 |
+
},
|
| 4117 |
+
{
|
| 4118 |
+
"epoch": 0.0454,
|
| 4119 |
+
"grad_norm": 1.1688799858093262,
|
| 4120 |
+
"learning_rate": 4.886316185875282e-05,
|
| 4121 |
+
"loss": 1.1655,
|
| 4122 |
+
"step": 52700
|
| 4123 |
+
},
|
| 4124 |
+
{
|
| 4125 |
+
"epoch": 0.0456,
|
| 4126 |
+
"grad_norm": 1.2751972675323486,
|
| 4127 |
+
"learning_rate": 4.885842679562085e-05,
|
| 4128 |
+
"loss": 1.2038,
|
| 4129 |
+
"step": 52800
|
| 4130 |
+
},
|
| 4131 |
+
{
|
| 4132 |
+
"epoch": 0.0458,
|
| 4133 |
+
"grad_norm": 0.6860191822052002,
|
| 4134 |
+
"learning_rate": 4.8853682122317426e-05,
|
| 4135 |
+
"loss": 1.1922,
|
| 4136 |
+
"step": 52900
|
| 4137 |
+
},
|
| 4138 |
+
{
|
| 4139 |
+
"epoch": 0.046,
|
| 4140 |
+
"grad_norm": 1.4772953987121582,
|
| 4141 |
+
"learning_rate": 4.8848927840753695e-05,
|
| 4142 |
+
"loss": 1.1856,
|
| 4143 |
+
"step": 53000
|
| 4144 |
+
},
|
| 4145 |
+
{
|
| 4146 |
+
"epoch": 0.046,
|
| 4147 |
+
"eval_loss": 1.0836056470870972,
|
| 4148 |
+
"eval_runtime": 76.3679,
|
| 4149 |
+
"eval_samples_per_second": 202.218,
|
| 4150 |
+
"eval_steps_per_second": 3.169,
|
| 4151 |
+
"step": 53000
|
| 4152 |
+
},
|
| 4153 |
+
{
|
| 4154 |
+
"epoch": 0.0462,
|
| 4155 |
+
"grad_norm": 1.2491508722305298,
|
| 4156 |
+
"learning_rate": 4.884416395284468e-05,
|
| 4157 |
+
"loss": 1.1924,
|
| 4158 |
+
"step": 53100
|
| 4159 |
+
},
|
| 4160 |
+
{
|
| 4161 |
+
"epoch": 0.0464,
|
| 4162 |
+
"grad_norm": 1.1689327955245972,
|
| 4163 |
+
"learning_rate": 4.883939046050928e-05,
|
| 4164 |
+
"loss": 1.1675,
|
| 4165 |
+
"step": 53200
|
| 4166 |
+
},
|
| 4167 |
+
{
|
| 4168 |
+
"epoch": 0.0466,
|
| 4169 |
+
"grad_norm": 1.0528528690338135,
|
| 4170 |
+
"learning_rate": 4.883460736567025e-05,
|
| 4171 |
+
"loss": 1.1879,
|
| 4172 |
+
"step": 53300
|
| 4173 |
+
},
|
| 4174 |
+
{
|
| 4175 |
+
"epoch": 0.0468,
|
| 4176 |
+
"grad_norm": 1.141653060913086,
|
| 4177 |
+
"learning_rate": 4.8829814670254226e-05,
|
| 4178 |
+
"loss": 1.1637,
|
| 4179 |
+
"step": 53400
|
| 4180 |
+
},
|
| 4181 |
+
{
|
| 4182 |
+
"epoch": 0.047,
|
| 4183 |
+
"grad_norm": 0.8094840049743652,
|
| 4184 |
+
"learning_rate": 4.88250123761917e-05,
|
| 4185 |
+
"loss": 1.1924,
|
| 4186 |
+
"step": 53500
|
| 4187 |
+
},
|
| 4188 |
+
{
|
| 4189 |
+
"epoch": 0.0472,
|
| 4190 |
+
"grad_norm": 1.4988161325454712,
|
| 4191 |
+
"learning_rate": 4.8820200485417036e-05,
|
| 4192 |
+
"loss": 1.1962,
|
| 4193 |
+
"step": 53600
|
| 4194 |
+
},
|
| 4195 |
+
{
|
| 4196 |
+
"epoch": 0.0474,
|
| 4197 |
+
"grad_norm": 0.8497682809829712,
|
| 4198 |
+
"learning_rate": 4.881537899986847e-05,
|
| 4199 |
+
"loss": 1.1987,
|
| 4200 |
+
"step": 53700
|
| 4201 |
+
},
|
| 4202 |
+
{
|
| 4203 |
+
"epoch": 0.0476,
|
| 4204 |
+
"grad_norm": 1.0132189989089966,
|
| 4205 |
+
"learning_rate": 4.8810547921488083e-05,
|
| 4206 |
+
"loss": 1.1666,
|
| 4207 |
+
"step": 53800
|
| 4208 |
+
},
|
| 4209 |
+
{
|
| 4210 |
+
"epoch": 0.0478,
|
| 4211 |
+
"grad_norm": 1.275478720664978,
|
| 4212 |
+
"learning_rate": 4.8805707252221846e-05,
|
| 4213 |
+
"loss": 1.2072,
|
| 4214 |
+
"step": 53900
|
| 4215 |
+
},
|
| 4216 |
+
{
|
| 4217 |
+
"epoch": 0.048,
|
| 4218 |
+
"grad_norm": 1.1257511377334595,
|
| 4219 |
+
"learning_rate": 4.880085699401958e-05,
|
| 4220 |
+
"loss": 1.2128,
|
| 4221 |
+
"step": 54000
|
| 4222 |
+
},
|
| 4223 |
+
{
|
| 4224 |
+
"epoch": 0.048,
|
| 4225 |
+
"eval_loss": 1.081576943397522,
|
| 4226 |
+
"eval_runtime": 76.6431,
|
| 4227 |
+
"eval_samples_per_second": 201.492,
|
| 4228 |
+
"eval_steps_per_second": 3.157,
|
| 4229 |
+
"step": 54000
|
| 4230 |
+
},
|
| 4231 |
+
{
|
| 4232 |
+
"epoch": 0.0482,
|
| 4233 |
+
"grad_norm": 1.132750153541565,
|
| 4234 |
+
"learning_rate": 4.879599714883496e-05,
|
| 4235 |
+
"loss": 1.2239,
|
| 4236 |
+
"step": 54100
|
| 4237 |
+
},
|
| 4238 |
+
{
|
| 4239 |
+
"epoch": 0.0484,
|
| 4240 |
+
"grad_norm": 1.3854628801345825,
|
| 4241 |
+
"learning_rate": 4.8791127718625526e-05,
|
| 4242 |
+
"loss": 1.1447,
|
| 4243 |
+
"step": 54200
|
| 4244 |
+
},
|
| 4245 |
+
{
|
| 4246 |
+
"epoch": 0.0486,
|
| 4247 |
+
"grad_norm": 1.32233464717865,
|
| 4248 |
+
"learning_rate": 4.87862487053527e-05,
|
| 4249 |
+
"loss": 1.1765,
|
| 4250 |
+
"step": 54300
|
| 4251 |
+
},
|
| 4252 |
+
{
|
| 4253 |
+
"epoch": 0.0488,
|
| 4254 |
+
"grad_norm": 1.1571578979492188,
|
| 4255 |
+
"learning_rate": 4.8781360110981744e-05,
|
| 4256 |
+
"loss": 1.1844,
|
| 4257 |
+
"step": 54400
|
| 4258 |
+
},
|
| 4259 |
+
{
|
| 4260 |
+
"epoch": 0.049,
|
| 4261 |
+
"grad_norm": 1.552740216255188,
|
| 4262 |
+
"learning_rate": 4.877646193748177e-05,
|
| 4263 |
+
"loss": 1.1336,
|
| 4264 |
+
"step": 54500
|
| 4265 |
+
},
|
| 4266 |
+
{
|
| 4267 |
+
"epoch": 0.0492,
|
| 4268 |
+
"grad_norm": 1.3447420597076416,
|
| 4269 |
+
"learning_rate": 4.8771554186825774e-05,
|
| 4270 |
+
"loss": 1.2401,
|
| 4271 |
+
"step": 54600
|
| 4272 |
+
},
|
| 4273 |
+
{
|
| 4274 |
+
"epoch": 0.0494,
|
| 4275 |
+
"grad_norm": 1.0012767314910889,
|
| 4276 |
+
"learning_rate": 4.87666368609906e-05,
|
| 4277 |
+
"loss": 1.2236,
|
| 4278 |
+
"step": 54700
|
| 4279 |
+
},
|
| 4280 |
+
{
|
| 4281 |
+
"epoch": 0.0496,
|
| 4282 |
+
"grad_norm": 1.1246662139892578,
|
| 4283 |
+
"learning_rate": 4.876170996195693e-05,
|
| 4284 |
+
"loss": 1.2452,
|
| 4285 |
+
"step": 54800
|
| 4286 |
+
},
|
| 4287 |
+
{
|
| 4288 |
+
"epoch": 0.0498,
|
| 4289 |
+
"grad_norm": 0.7534450888633728,
|
| 4290 |
+
"learning_rate": 4.875677349170934e-05,
|
| 4291 |
+
"loss": 1.2333,
|
| 4292 |
+
"step": 54900
|
| 4293 |
+
},
|
| 4294 |
+
{
|
| 4295 |
+
"epoch": 0.05,
|
| 4296 |
+
"grad_norm": 1.2943884134292603,
|
| 4297 |
+
"learning_rate": 4.875182745223622e-05,
|
| 4298 |
+
"loss": 1.1986,
|
| 4299 |
+
"step": 55000
|
| 4300 |
+
},
|
| 4301 |
+
{
|
| 4302 |
+
"epoch": 0.05,
|
| 4303 |
+
"eval_loss": 1.0774849653244019,
|
| 4304 |
+
"eval_runtime": 76.6003,
|
| 4305 |
+
"eval_samples_per_second": 201.605,
|
| 4306 |
+
"eval_steps_per_second": 3.159,
|
| 4307 |
+
"step": 55000
|
| 4308 |
+
},
|
| 4309 |
+
{
|
| 4310 |
+
"epoch": 0.0502,
|
| 4311 |
+
"grad_norm": 1.0771546363830566,
|
| 4312 |
+
"learning_rate": 4.874687184552984e-05,
|
| 4313 |
+
"loss": 1.2022,
|
| 4314 |
+
"step": 55100
|
| 4315 |
+
},
|
| 4316 |
+
{
|
| 4317 |
+
"epoch": 0.0504,
|
| 4318 |
+
"grad_norm": 1.1722393035888672,
|
| 4319 |
+
"learning_rate": 4.8741906673586334e-05,
|
| 4320 |
+
"loss": 1.1856,
|
| 4321 |
+
"step": 55200
|
| 4322 |
+
},
|
| 4323 |
+
{
|
| 4324 |
+
"epoch": 0.0506,
|
| 4325 |
+
"grad_norm": 0.7547242045402527,
|
| 4326 |
+
"learning_rate": 4.873693193840565e-05,
|
| 4327 |
+
"loss": 1.153,
|
| 4328 |
+
"step": 55300
|
| 4329 |
+
},
|
| 4330 |
+
{
|
| 4331 |
+
"epoch": 0.0508,
|
| 4332 |
+
"grad_norm": 0.9694270491600037,
|
| 4333 |
+
"learning_rate": 4.873194764199162e-05,
|
| 4334 |
+
"loss": 1.2389,
|
| 4335 |
+
"step": 55400
|
| 4336 |
+
},
|
| 4337 |
+
{
|
| 4338 |
+
"epoch": 0.051,
|
| 4339 |
+
"grad_norm": 0.6288232803344727,
|
| 4340 |
+
"learning_rate": 4.872695378635192e-05,
|
| 4341 |
+
"loss": 1.195,
|
| 4342 |
+
"step": 55500
|
| 4343 |
+
},
|
| 4344 |
+
{
|
| 4345 |
+
"epoch": 0.0512,
|
| 4346 |
+
"grad_norm": 1.1400961875915527,
|
| 4347 |
+
"learning_rate": 4.872195037349807e-05,
|
| 4348 |
+
"loss": 1.1903,
|
| 4349 |
+
"step": 55600
|
| 4350 |
+
},
|
| 4351 |
+
{
|
| 4352 |
+
"epoch": 0.0514,
|
| 4353 |
+
"grad_norm": 1.0738123655319214,
|
| 4354 |
+
"learning_rate": 4.871693740544545e-05,
|
| 4355 |
+
"loss": 1.1764,
|
| 4356 |
+
"step": 55700
|
| 4357 |
+
},
|
| 4358 |
+
{
|
| 4359 |
+
"epoch": 0.0516,
|
| 4360 |
+
"grad_norm": 1.2298240661621094,
|
| 4361 |
+
"learning_rate": 4.871191488421327e-05,
|
| 4362 |
+
"loss": 1.1701,
|
| 4363 |
+
"step": 55800
|
| 4364 |
+
},
|
| 4365 |
+
{
|
| 4366 |
+
"epoch": 0.0518,
|
| 4367 |
+
"grad_norm": 1.3240865468978882,
|
| 4368 |
+
"learning_rate": 4.8706882811824624e-05,
|
| 4369 |
+
"loss": 1.1828,
|
| 4370 |
+
"step": 55900
|
| 4371 |
+
},
|
| 4372 |
+
{
|
| 4373 |
+
"epoch": 0.052,
|
| 4374 |
+
"grad_norm": 1.4167003631591797,
|
| 4375 |
+
"learning_rate": 4.870184119030641e-05,
|
| 4376 |
+
"loss": 1.204,
|
| 4377 |
+
"step": 56000
|
| 4378 |
+
},
|
| 4379 |
+
{
|
| 4380 |
+
"epoch": 0.052,
|
| 4381 |
+
"eval_loss": 1.0775164365768433,
|
| 4382 |
+
"eval_runtime": 76.8889,
|
| 4383 |
+
"eval_samples_per_second": 200.848,
|
| 4384 |
+
"eval_steps_per_second": 3.147,
|
| 4385 |
+
"step": 56000
|
| 4386 |
+
},
|
| 4387 |
+
{
|
| 4388 |
+
"epoch": 0.0522,
|
| 4389 |
+
"grad_norm": 0.6648851037025452,
|
| 4390 |
+
"learning_rate": 4.86967900216894e-05,
|
| 4391 |
+
"loss": 1.174,
|
| 4392 |
+
"step": 56100
|
| 4393 |
+
},
|
| 4394 |
+
{
|
| 4395 |
+
"epoch": 0.0524,
|
| 4396 |
+
"grad_norm": 1.29317307472229,
|
| 4397 |
+
"learning_rate": 4.8691729308008196e-05,
|
| 4398 |
+
"loss": 1.1695,
|
| 4399 |
+
"step": 56200
|
| 4400 |
+
},
|
| 4401 |
+
{
|
| 4402 |
+
"epoch": 0.0526,
|
| 4403 |
+
"grad_norm": 1.3121986389160156,
|
| 4404 |
+
"learning_rate": 4.868665905130127e-05,
|
| 4405 |
+
"loss": 1.1941,
|
| 4406 |
+
"step": 56300
|
| 4407 |
+
},
|
| 4408 |
+
{
|
| 4409 |
+
"epoch": 0.0528,
|
| 4410 |
+
"grad_norm": 0.6604340672492981,
|
| 4411 |
+
"learning_rate": 4.868157925361091e-05,
|
| 4412 |
+
"loss": 1.1875,
|
| 4413 |
+
"step": 56400
|
| 4414 |
+
},
|
| 4415 |
+
{
|
| 4416 |
+
"epoch": 0.053,
|
| 4417 |
+
"grad_norm": 1.0366885662078857,
|
| 4418 |
+
"learning_rate": 4.867648991698325e-05,
|
| 4419 |
+
"loss": 1.2265,
|
| 4420 |
+
"step": 56500
|
| 4421 |
+
},
|
| 4422 |
+
{
|
| 4423 |
+
"epoch": 0.0532,
|
| 4424 |
+
"grad_norm": 1.382543683052063,
|
| 4425 |
+
"learning_rate": 4.867139104346829e-05,
|
| 4426 |
+
"loss": 1.2122,
|
| 4427 |
+
"step": 56600
|
| 4428 |
+
},
|
| 4429 |
+
{
|
| 4430 |
+
"epoch": 0.0534,
|
| 4431 |
+
"grad_norm": 1.0773979425430298,
|
| 4432 |
+
"learning_rate": 4.866628263511985e-05,
|
| 4433 |
+
"loss": 1.2375,
|
| 4434 |
+
"step": 56700
|
| 4435 |
+
},
|
| 4436 |
+
{
|
| 4437 |
+
"epoch": 0.0536,
|
| 4438 |
+
"grad_norm": 1.178758978843689,
|
| 4439 |
+
"learning_rate": 4.8661164693995584e-05,
|
| 4440 |
+
"loss": 1.1959,
|
| 4441 |
+
"step": 56800
|
| 4442 |
+
},
|
| 4443 |
+
{
|
| 4444 |
+
"epoch": 0.0538,
|
| 4445 |
+
"grad_norm": 0.7048764228820801,
|
| 4446 |
+
"learning_rate": 4.865603722215702e-05,
|
| 4447 |
+
"loss": 1.1841,
|
| 4448 |
+
"step": 56900
|
| 4449 |
+
},
|
| 4450 |
+
{
|
| 4451 |
+
"epoch": 0.054,
|
| 4452 |
+
"grad_norm": 1.3390711545944214,
|
| 4453 |
+
"learning_rate": 4.865090022166949e-05,
|
| 4454 |
+
"loss": 1.2033,
|
| 4455 |
+
"step": 57000
|
| 4456 |
+
},
|
| 4457 |
+
{
|
| 4458 |
+
"epoch": 0.054,
|
| 4459 |
+
"eval_loss": 1.0746017694473267,
|
| 4460 |
+
"eval_runtime": 77.1768,
|
| 4461 |
+
"eval_samples_per_second": 200.099,
|
| 4462 |
+
"eval_steps_per_second": 3.136,
|
| 4463 |
+
"step": 57000
|
| 4464 |
+
},
|
| 4465 |
+
{
|
| 4466 |
+
"epoch": 0.0542,
|
| 4467 |
+
"grad_norm": 1.0639598369598389,
|
| 4468 |
+
"learning_rate": 4.864575369460218e-05,
|
| 4469 |
+
"loss": 1.1948,
|
| 4470 |
+
"step": 57100
|
| 4471 |
+
},
|
| 4472 |
+
{
|
| 4473 |
+
"epoch": 0.0544,
|
| 4474 |
+
"grad_norm": 1.1349152326583862,
|
| 4475 |
+
"learning_rate": 4.86405976430281e-05,
|
| 4476 |
+
"loss": 1.1666,
|
| 4477 |
+
"step": 57200
|
| 4478 |
+
},
|
| 4479 |
+
{
|
| 4480 |
+
"epoch": 0.0546,
|
| 4481 |
+
"grad_norm": 1.0187245607376099,
|
| 4482 |
+
"learning_rate": 4.8635432069024125e-05,
|
| 4483 |
+
"loss": 1.1614,
|
| 4484 |
+
"step": 57300
|
| 4485 |
+
},
|
| 4486 |
+
{
|
| 4487 |
+
"epoch": 0.0548,
|
| 4488 |
+
"grad_norm": 0.6468742489814758,
|
| 4489 |
+
"learning_rate": 4.863025697467094e-05,
|
| 4490 |
+
"loss": 1.2043,
|
| 4491 |
+
"step": 57400
|
| 4492 |
+
},
|
| 4493 |
+
{
|
| 4494 |
+
"epoch": 0.055,
|
| 4495 |
+
"grad_norm": 1.1298869848251343,
|
| 4496 |
+
"learning_rate": 4.862507236205307e-05,
|
| 4497 |
+
"loss": 1.1884,
|
| 4498 |
+
"step": 57500
|
| 4499 |
+
},
|
| 4500 |
+
{
|
| 4501 |
+
"epoch": 0.0552,
|
| 4502 |
+
"grad_norm": 0.7240111827850342,
|
| 4503 |
+
"learning_rate": 4.861987823325887e-05,
|
| 4504 |
+
"loss": 1.186,
|
| 4505 |
+
"step": 57600
|
| 4506 |
+
},
|
| 4507 |
+
{
|
| 4508 |
+
"epoch": 0.0554,
|
| 4509 |
+
"grad_norm": 0.8047366142272949,
|
| 4510 |
+
"learning_rate": 4.861467459038056e-05,
|
| 4511 |
+
"loss": 1.2029,
|
| 4512 |
+
"step": 57700
|
| 4513 |
+
},
|
| 4514 |
+
{
|
| 4515 |
+
"epoch": 0.0556,
|
| 4516 |
+
"grad_norm": 0.8840340375900269,
|
| 4517 |
+
"learning_rate": 4.860946143551413e-05,
|
| 4518 |
+
"loss": 1.19,
|
| 4519 |
+
"step": 57800
|
| 4520 |
+
},
|
| 4521 |
+
{
|
| 4522 |
+
"epoch": 0.0558,
|
| 4523 |
+
"grad_norm": 1.1575409173965454,
|
| 4524 |
+
"learning_rate": 4.860423877075947e-05,
|
| 4525 |
+
"loss": 1.1637,
|
| 4526 |
+
"step": 57900
|
| 4527 |
+
},
|
| 4528 |
+
{
|
| 4529 |
+
"epoch": 0.056,
|
| 4530 |
+
"grad_norm": 0.6591224074363708,
|
| 4531 |
+
"learning_rate": 4.859900659822025e-05,
|
| 4532 |
+
"loss": 1.2203,
|
| 4533 |
+
"step": 58000
|
| 4534 |
+
},
|
| 4535 |
+
{
|
| 4536 |
+
"epoch": 0.056,
|
| 4537 |
+
"eval_loss": 1.0788133144378662,
|
| 4538 |
+
"eval_runtime": 76.7654,
|
| 4539 |
+
"eval_samples_per_second": 201.171,
|
| 4540 |
+
"eval_steps_per_second": 3.152,
|
| 4541 |
+
"step": 58000
|
| 4542 |
+
},
|
| 4543 |
+
{
|
| 4544 |
+
"epoch": 0.0562,
|
| 4545 |
+
"grad_norm": 1.3405015468597412,
|
| 4546 |
+
"learning_rate": 4.859376492000399e-05,
|
| 4547 |
+
"loss": 1.19,
|
| 4548 |
+
"step": 58100
|
| 4549 |
+
},
|
| 4550 |
+
{
|
| 4551 |
+
"epoch": 0.0564,
|
| 4552 |
+
"grad_norm": 1.1912270784378052,
|
| 4553 |
+
"learning_rate": 4.858851373822205e-05,
|
| 4554 |
+
"loss": 1.1521,
|
| 4555 |
+
"step": 58200
|
| 4556 |
+
},
|
| 4557 |
+
{
|
| 4558 |
+
"epoch": 0.0566,
|
| 4559 |
+
"grad_norm": 1.0169751644134521,
|
| 4560 |
+
"learning_rate": 4.85832530549896e-05,
|
| 4561 |
+
"loss": 1.2054,
|
| 4562 |
+
"step": 58300
|
| 4563 |
+
},
|
| 4564 |
+
{
|
| 4565 |
+
"epoch": 0.0568,
|
| 4566 |
+
"grad_norm": 0.6713104248046875,
|
| 4567 |
+
"learning_rate": 4.857798287242563e-05,
|
| 4568 |
+
"loss": 1.2033,
|
| 4569 |
+
"step": 58400
|
| 4570 |
+
},
|
| 4571 |
+
{
|
| 4572 |
+
"epoch": 0.057,
|
| 4573 |
+
"grad_norm": 1.2116252183914185,
|
| 4574 |
+
"learning_rate": 4.857270319265298e-05,
|
| 4575 |
+
"loss": 1.1919,
|
| 4576 |
+
"step": 58500
|
| 4577 |
+
},
|
| 4578 |
+
{
|
| 4579 |
+
"epoch": 0.0572,
|
| 4580 |
+
"grad_norm": 0.9526674151420593,
|
| 4581 |
+
"learning_rate": 4.856741401779831e-05,
|
| 4582 |
+
"loss": 1.1724,
|
| 4583 |
+
"step": 58600
|
| 4584 |
+
},
|
| 4585 |
+
{
|
| 4586 |
+
"epoch": 0.0574,
|
| 4587 |
+
"grad_norm": 1.458253264427185,
|
| 4588 |
+
"learning_rate": 4.856211534999209e-05,
|
| 4589 |
+
"loss": 1.1479,
|
| 4590 |
+
"step": 58700
|
| 4591 |
+
},
|
| 4592 |
+
{
|
| 4593 |
+
"epoch": 0.0576,
|
| 4594 |
+
"grad_norm": 1.173437237739563,
|
| 4595 |
+
"learning_rate": 4.855680719136862e-05,
|
| 4596 |
+
"loss": 1.2005,
|
| 4597 |
+
"step": 58800
|
| 4598 |
+
},
|
| 4599 |
+
{
|
| 4600 |
+
"epoch": 0.0578,
|
| 4601 |
+
"grad_norm": 0.7292013168334961,
|
| 4602 |
+
"learning_rate": 4.8551489544066034e-05,
|
| 4603 |
+
"loss": 1.1292,
|
| 4604 |
+
"step": 58900
|
| 4605 |
+
},
|
| 4606 |
+
{
|
| 4607 |
+
"epoch": 0.058,
|
| 4608 |
+
"grad_norm": 0.6017533540725708,
|
| 4609 |
+
"learning_rate": 4.854616241022627e-05,
|
| 4610 |
+
"loss": 1.1527,
|
| 4611 |
+
"step": 59000
|
| 4612 |
+
},
|
| 4613 |
+
{
|
| 4614 |
+
"epoch": 0.058,
|
| 4615 |
+
"eval_loss": 1.0688042640686035,
|
| 4616 |
+
"eval_runtime": 76.596,
|
| 4617 |
+
"eval_samples_per_second": 201.616,
|
| 4618 |
+
"eval_steps_per_second": 3.159,
|
| 4619 |
+
"step": 59000
|
| 4620 |
+
},
|
| 4621 |
+
{
|
| 4622 |
+
"epoch": 0.0582,
|
| 4623 |
+
"grad_norm": 0.8270254731178284,
|
| 4624 |
+
"learning_rate": 4.8540825791995114e-05,
|
| 4625 |
+
"loss": 1.1517,
|
| 4626 |
+
"step": 59100
|
| 4627 |
+
},
|
| 4628 |
+
{
|
| 4629 |
+
"epoch": 0.0584,
|
| 4630 |
+
"grad_norm": 1.1182663440704346,
|
| 4631 |
+
"learning_rate": 4.8535479691522136e-05,
|
| 4632 |
+
"loss": 1.2282,
|
| 4633 |
+
"step": 59200
|
| 4634 |
+
},
|
| 4635 |
+
{
|
| 4636 |
+
"epoch": 0.0586,
|
| 4637 |
+
"grad_norm": 1.1249291896820068,
|
| 4638 |
+
"learning_rate": 4.853012411096075e-05,
|
| 4639 |
+
"loss": 1.1314,
|
| 4640 |
+
"step": 59300
|
| 4641 |
+
},
|
| 4642 |
+
{
|
| 4643 |
+
"epoch": 0.0588,
|
| 4644 |
+
"grad_norm": 0.6025962233543396,
|
| 4645 |
+
"learning_rate": 4.85247590524682e-05,
|
| 4646 |
+
"loss": 1.1879,
|
| 4647 |
+
"step": 59400
|
| 4648 |
+
},
|
| 4649 |
+
{
|
| 4650 |
+
"epoch": 0.059,
|
| 4651 |
+
"grad_norm": 1.2914466857910156,
|
| 4652 |
+
"learning_rate": 4.85193845182055e-05,
|
| 4653 |
+
"loss": 1.1926,
|
| 4654 |
+
"step": 59500
|
| 4655 |
+
},
|
| 4656 |
+
{
|
| 4657 |
+
"epoch": 0.0592,
|
| 4658 |
+
"grad_norm": 0.7965525388717651,
|
| 4659 |
+
"learning_rate": 4.8514000510337544e-05,
|
| 4660 |
+
"loss": 1.2344,
|
| 4661 |
+
"step": 59600
|
| 4662 |
+
},
|
| 4663 |
+
{
|
| 4664 |
+
"epoch": 0.0594,
|
| 4665 |
+
"grad_norm": 0.6595709323883057,
|
| 4666 |
+
"learning_rate": 4.850860703103298e-05,
|
| 4667 |
+
"loss": 1.2056,
|
| 4668 |
+
"step": 59700
|
| 4669 |
+
},
|
| 4670 |
+
{
|
| 4671 |
+
"epoch": 0.0596,
|
| 4672 |
+
"grad_norm": 0.783892035484314,
|
| 4673 |
+
"learning_rate": 4.850320408246433e-05,
|
| 4674 |
+
"loss": 1.1343,
|
| 4675 |
+
"step": 59800
|
| 4676 |
+
},
|
| 4677 |
+
{
|
| 4678 |
+
"epoch": 0.0598,
|
| 4679 |
+
"grad_norm": 0.948952853679657,
|
| 4680 |
+
"learning_rate": 4.849779166680788e-05,
|
| 4681 |
+
"loss": 1.1607,
|
| 4682 |
+
"step": 59900
|
| 4683 |
+
},
|
| 4684 |
+
{
|
| 4685 |
+
"epoch": 0.06,
|
| 4686 |
+
"grad_norm": 0.725027322769165,
|
| 4687 |
+
"learning_rate": 4.849236978624375e-05,
|
| 4688 |
+
"loss": 1.2125,
|
| 4689 |
+
"step": 60000
|
| 4690 |
+
},
|
| 4691 |
+
{
|
| 4692 |
+
"epoch": 0.06,
|
| 4693 |
+
"eval_loss": 1.0838971138000488,
|
| 4694 |
+
"eval_runtime": 76.8451,
|
| 4695 |
+
"eval_samples_per_second": 200.963,
|
| 4696 |
+
"eval_steps_per_second": 3.149,
|
| 4697 |
+
"step": 60000
|
| 4698 |
+
},
|
| 4699 |
+
{
|
| 4700 |
+
"epoch": 0.0002,
|
| 4701 |
+
"grad_norm": 1.1788556575775146,
|
| 4702 |
+
"learning_rate": 4.848693844295589e-05,
|
| 4703 |
+
"loss": 1.1917,
|
| 4704 |
+
"step": 60100
|
| 4705 |
+
},
|
| 4706 |
+
{
|
| 4707 |
+
"epoch": 0.0004,
|
| 4708 |
+
"grad_norm": 1.3381775617599487,
|
| 4709 |
+
"learning_rate": 4.848149763913202e-05,
|
| 4710 |
+
"loss": 1.2108,
|
| 4711 |
+
"step": 60200
|
| 4712 |
+
},
|
| 4713 |
+
{
|
| 4714 |
+
"epoch": 0.0006,
|
| 4715 |
+
"grad_norm": 0.9748820066452026,
|
| 4716 |
+
"learning_rate": 4.847604737696372e-05,
|
| 4717 |
+
"loss": 1.2054,
|
| 4718 |
+
"step": 60300
|
| 4719 |
+
},
|
| 4720 |
+
{
|
| 4721 |
+
"epoch": 0.0008,
|
| 4722 |
+
"grad_norm": 1.3528317213058472,
|
| 4723 |
+
"learning_rate": 4.847058765864634e-05,
|
| 4724 |
+
"loss": 1.1582,
|
| 4725 |
+
"step": 60400
|
| 4726 |
+
},
|
| 4727 |
+
{
|
| 4728 |
+
"epoch": 0.001,
|
| 4729 |
+
"grad_norm": 1.0475611686706543,
|
| 4730 |
+
"learning_rate": 4.8465118486379065e-05,
|
| 4731 |
+
"loss": 1.1409,
|
| 4732 |
+
"step": 60500
|
| 4733 |
+
},
|
| 4734 |
+
{
|
| 4735 |
+
"epoch": 0.0012,
|
| 4736 |
+
"grad_norm": 0.667515754699707,
|
| 4737 |
+
"learning_rate": 4.8459639862364866e-05,
|
| 4738 |
+
"loss": 1.1548,
|
| 4739 |
+
"step": 60600
|
| 4740 |
+
},
|
| 4741 |
+
{
|
| 4742 |
+
"epoch": 0.0014,
|
| 4743 |
+
"grad_norm": 1.3529212474822998,
|
| 4744 |
+
"learning_rate": 4.845415178881053e-05,
|
| 4745 |
+
"loss": 1.1741,
|
| 4746 |
+
"step": 60700
|
| 4747 |
+
},
|
| 4748 |
+
{
|
| 4749 |
+
"epoch": 0.0016,
|
| 4750 |
+
"grad_norm": 1.2415704727172852,
|
| 4751 |
+
"learning_rate": 4.844865426792666e-05,
|
| 4752 |
+
"loss": 1.1689,
|
| 4753 |
+
"step": 60800
|
| 4754 |
+
},
|
| 4755 |
+
{
|
| 4756 |
+
"epoch": 0.0018,
|
| 4757 |
+
"grad_norm": 0.9598329663276672,
|
| 4758 |
+
"learning_rate": 4.844314730192766e-05,
|
| 4759 |
+
"loss": 1.2138,
|
| 4760 |
+
"step": 60900
|
| 4761 |
+
},
|
| 4762 |
+
{
|
| 4763 |
+
"epoch": 0.002,
|
| 4764 |
+
"grad_norm": 0.660463273525238,
|
| 4765 |
+
"learning_rate": 4.843763089303173e-05,
|
| 4766 |
+
"loss": 1.1897,
|
| 4767 |
+
"step": 61000
|
| 4768 |
+
},
|
| 4769 |
+
{
|
| 4770 |
+
"epoch": 0.002,
|
| 4771 |
+
"eval_loss": 1.0804229974746704,
|
| 4772 |
+
"eval_runtime": 77.9042,
|
| 4773 |
+
"eval_samples_per_second": 198.231,
|
| 4774 |
+
"eval_steps_per_second": 3.106,
|
| 4775 |
+
"step": 61000
|
| 4776 |
+
},
|
| 4777 |
+
{
|
| 4778 |
+
"epoch": 0.0022,
|
| 4779 |
+
"grad_norm": 1.3137476444244385,
|
| 4780 |
+
"learning_rate": 4.843210504346088e-05,
|
| 4781 |
+
"loss": 1.2149,
|
| 4782 |
+
"step": 61100
|
| 4783 |
+
},
|
| 4784 |
+
{
|
| 4785 |
+
"epoch": 0.0024,
|
| 4786 |
+
"grad_norm": 2.466374158859253,
|
| 4787 |
+
"learning_rate": 4.842656975544092e-05,
|
| 4788 |
+
"loss": 1.2294,
|
| 4789 |
+
"step": 61200
|
| 4790 |
+
},
|
| 4791 |
+
{
|
| 4792 |
+
"epoch": 0.0026,
|
| 4793 |
+
"grad_norm": 0.9236732721328735,
|
| 4794 |
+
"learning_rate": 4.842102503120146e-05,
|
| 4795 |
+
"loss": 1.2316,
|
| 4796 |
+
"step": 61300
|
| 4797 |
+
},
|
| 4798 |
+
{
|
| 4799 |
+
"epoch": 0.0028,
|
| 4800 |
+
"grad_norm": 0.9453101754188538,
|
| 4801 |
+
"learning_rate": 4.841547087297592e-05,
|
| 4802 |
+
"loss": 1.1903,
|
| 4803 |
+
"step": 61400
|
| 4804 |
+
},
|
| 4805 |
+
{
|
| 4806 |
+
"epoch": 0.003,
|
| 4807 |
+
"grad_norm": 1.0694693326950073,
|
| 4808 |
+
"learning_rate": 4.840990728300151e-05,
|
| 4809 |
+
"loss": 1.2027,
|
| 4810 |
+
"step": 61500
|
| 4811 |
+
},
|
| 4812 |
+
{
|
| 4813 |
+
"epoch": 0.0032,
|
| 4814 |
+
"grad_norm": 1.0661156177520752,
|
| 4815 |
+
"learning_rate": 4.8404334263519254e-05,
|
| 4816 |
+
"loss": 1.2268,
|
| 4817 |
+
"step": 61600
|
| 4818 |
+
},
|
| 4819 |
+
{
|
| 4820 |
+
"epoch": 0.0034,
|
| 4821 |
+
"grad_norm": 1.3803828954696655,
|
| 4822 |
+
"learning_rate": 4.839875181677394e-05,
|
| 4823 |
+
"loss": 1.2084,
|
| 4824 |
+
"step": 61700
|
| 4825 |
+
},
|
| 4826 |
+
{
|
| 4827 |
+
"epoch": 0.0036,
|
| 4828 |
+
"grad_norm": 0.896979033946991,
|
| 4829 |
+
"learning_rate": 4.839315994501421e-05,
|
| 4830 |
+
"loss": 1.1818,
|
| 4831 |
+
"step": 61800
|
| 4832 |
+
},
|
| 4833 |
+
{
|
| 4834 |
+
"epoch": 0.0038,
|
| 4835 |
+
"grad_norm": 1.1509560346603394,
|
| 4836 |
+
"learning_rate": 4.8387558650492446e-05,
|
| 4837 |
+
"loss": 1.226,
|
| 4838 |
+
"step": 61900
|
| 4839 |
+
},
|
| 4840 |
+
{
|
| 4841 |
+
"epoch": 0.004,
|
| 4842 |
+
"grad_norm": 1.2490339279174805,
|
| 4843 |
+
"learning_rate": 4.8381947935464854e-05,
|
| 4844 |
+
"loss": 1.2283,
|
| 4845 |
+
"step": 62000
|
| 4846 |
+
},
|
| 4847 |
+
{
|
| 4848 |
+
"epoch": 0.004,
|
| 4849 |
+
"eval_loss": 1.086965560913086,
|
| 4850 |
+
"eval_runtime": 75.4991,
|
| 4851 |
+
"eval_samples_per_second": 204.545,
|
| 4852 |
+
"eval_steps_per_second": 3.205,
|
| 4853 |
+
"step": 62000
|
| 4854 |
+
},
|
| 4855 |
+
{
|
| 4856 |
+
"epoch": 0.0042,
|
| 4857 |
+
"grad_norm": 1.0047966241836548,
|
| 4858 |
+
"learning_rate": 4.837632780219142e-05,
|
| 4859 |
+
"loss": 1.2006,
|
| 4860 |
+
"step": 62100
|
| 4861 |
+
},
|
| 4862 |
+
{
|
| 4863 |
+
"epoch": 0.0044,
|
| 4864 |
+
"grad_norm": 1.3791793584823608,
|
| 4865 |
+
"learning_rate": 4.837069825293596e-05,
|
| 4866 |
+
"loss": 1.2191,
|
| 4867 |
+
"step": 62200
|
| 4868 |
+
},
|
| 4869 |
+
{
|
| 4870 |
+
"epoch": 0.0046,
|
| 4871 |
+
"grad_norm": 1.4083282947540283,
|
| 4872 |
+
"learning_rate": 4.836505928996603e-05,
|
| 4873 |
+
"loss": 1.2232,
|
| 4874 |
+
"step": 62300
|
| 4875 |
+
},
|
| 4876 |
+
{
|
| 4877 |
+
"epoch": 0.0048,
|
| 4878 |
+
"grad_norm": 1.5420063734054565,
|
| 4879 |
+
"learning_rate": 4.835941091555301e-05,
|
| 4880 |
+
"loss": 1.2281,
|
| 4881 |
+
"step": 62400
|
| 4882 |
+
},
|
| 4883 |
+
{
|
| 4884 |
+
"epoch": 0.005,
|
| 4885 |
+
"grad_norm": 0.7661809921264648,
|
| 4886 |
+
"learning_rate": 4.8353753131972066e-05,
|
| 4887 |
+
"loss": 1.2262,
|
| 4888 |
+
"step": 62500
|
| 4889 |
+
},
|
| 4890 |
+
{
|
| 4891 |
+
"epoch": 0.0052,
|
| 4892 |
+
"grad_norm": 0.5983784198760986,
|
| 4893 |
+
"learning_rate": 4.8348085941502164e-05,
|
| 4894 |
+
"loss": 1.2203,
|
| 4895 |
+
"step": 62600
|
| 4896 |
+
},
|
| 4897 |
+
{
|
| 4898 |
+
"epoch": 0.0054,
|
| 4899 |
+
"grad_norm": 0.8108716011047363,
|
| 4900 |
+
"learning_rate": 4.8342409346426024e-05,
|
| 4901 |
+
"loss": 1.1536,
|
| 4902 |
+
"step": 62700
|
| 4903 |
+
},
|
| 4904 |
+
{
|
| 4905 |
+
"epoch": 0.0056,
|
| 4906 |
+
"grad_norm": 0.9011421203613281,
|
| 4907 |
+
"learning_rate": 4.83367233490302e-05,
|
| 4908 |
+
"loss": 1.2214,
|
| 4909 |
+
"step": 62800
|
| 4910 |
+
},
|
| 4911 |
+
{
|
| 4912 |
+
"epoch": 0.0058,
|
| 4913 |
+
"grad_norm": 0.6667259335517883,
|
| 4914 |
+
"learning_rate": 4.8331027951604995e-05,
|
| 4915 |
+
"loss": 1.1932,
|
| 4916 |
+
"step": 62900
|
| 4917 |
+
},
|
| 4918 |
+
{
|
| 4919 |
+
"epoch": 0.006,
|
| 4920 |
+
"grad_norm": 1.2265853881835938,
|
| 4921 |
+
"learning_rate": 4.8325323156444525e-05,
|
| 4922 |
+
"loss": 1.235,
|
| 4923 |
+
"step": 63000
|
| 4924 |
+
},
|
| 4925 |
+
{
|
| 4926 |
+
"epoch": 0.006,
|
| 4927 |
+
"eval_loss": 1.0849037170410156,
|
| 4928 |
+
"eval_runtime": 76.5492,
|
| 4929 |
+
"eval_samples_per_second": 201.74,
|
| 4930 |
+
"eval_steps_per_second": 3.161,
|
| 4931 |
+
"step": 63000
|
| 4932 |
+
},
|
| 4933 |
+
{
|
| 4934 |
+
"epoch": 0.0062,
|
| 4935 |
+
"grad_norm": 1.468518614768982,
|
| 4936 |
+
"learning_rate": 4.831960896584667e-05,
|
| 4937 |
+
"loss": 1.1886,
|
| 4938 |
+
"step": 63100
|
| 4939 |
+
},
|
| 4940 |
+
{
|
| 4941 |
+
"epoch": 0.0064,
|
| 4942 |
+
"grad_norm": 1.2378790378570557,
|
| 4943 |
+
"learning_rate": 4.831388538211312e-05,
|
| 4944 |
+
"loss": 1.1983,
|
| 4945 |
+
"step": 63200
|
| 4946 |
+
},
|
| 4947 |
+
{
|
| 4948 |
+
"epoch": 0.0066,
|
| 4949 |
+
"grad_norm": 1.2989089488983154,
|
| 4950 |
+
"learning_rate": 4.830815240754933e-05,
|
| 4951 |
+
"loss": 1.1894,
|
| 4952 |
+
"step": 63300
|
| 4953 |
+
},
|
| 4954 |
+
{
|
| 4955 |
+
"epoch": 0.0068,
|
| 4956 |
+
"grad_norm": 1.3696600198745728,
|
| 4957 |
+
"learning_rate": 4.830241004446453e-05,
|
| 4958 |
+
"loss": 1.1798,
|
| 4959 |
+
"step": 63400
|
| 4960 |
+
},
|
| 4961 |
+
{
|
| 4962 |
+
"epoch": 0.007,
|
| 4963 |
+
"grad_norm": 1.3715136051177979,
|
| 4964 |
+
"learning_rate": 4.829665829517175e-05,
|
| 4965 |
+
"loss": 1.2323,
|
| 4966 |
+
"step": 63500
|
| 4967 |
+
},
|
| 4968 |
+
{
|
| 4969 |
+
"epoch": 0.0072,
|
| 4970 |
+
"grad_norm": 0.7888614535331726,
|
| 4971 |
+
"learning_rate": 4.82908971619878e-05,
|
| 4972 |
+
"loss": 1.2098,
|
| 4973 |
+
"step": 63600
|
| 4974 |
+
},
|
| 4975 |
+
{
|
| 4976 |
+
"epoch": 0.0074,
|
| 4977 |
+
"grad_norm": 1.0456979274749756,
|
| 4978 |
+
"learning_rate": 4.828512664723326e-05,
|
| 4979 |
+
"loss": 1.21,
|
| 4980 |
+
"step": 63700
|
| 4981 |
+
},
|
| 4982 |
+
{
|
| 4983 |
+
"epoch": 0.0076,
|
| 4984 |
+
"grad_norm": 1.4525970220565796,
|
| 4985 |
+
"learning_rate": 4.827934675323248e-05,
|
| 4986 |
+
"loss": 1.191,
|
| 4987 |
+
"step": 63800
|
| 4988 |
+
},
|
| 4989 |
+
{
|
| 4990 |
+
"epoch": 0.0078,
|
| 4991 |
+
"grad_norm": 1.6751583814620972,
|
| 4992 |
+
"learning_rate": 4.8273557482313625e-05,
|
| 4993 |
+
"loss": 1.2084,
|
| 4994 |
+
"step": 63900
|
| 4995 |
+
},
|
| 4996 |
+
{
|
| 4997 |
+
"epoch": 0.008,
|
| 4998 |
+
"grad_norm": 0.7282904982566833,
|
| 4999 |
+
"learning_rate": 4.826775883680859e-05,
|
| 5000 |
+
"loss": 1.2376,
|
| 5001 |
+
"step": 64000
|
| 5002 |
+
},
|
| 5003 |
+
{
|
| 5004 |
+
"epoch": 0.008,
|
| 5005 |
+
"eval_loss": 1.0806148052215576,
|
| 5006 |
+
"eval_runtime": 75.7629,
|
| 5007 |
+
"eval_samples_per_second": 203.833,
|
| 5008 |
+
"eval_steps_per_second": 3.194,
|
| 5009 |
+
"step": 64000
|
| 5010 |
+
},
|
| 5011 |
+
{
|
| 5012 |
+
"epoch": 0.0082,
|
| 5013 |
+
"grad_norm": 1.0859407186508179,
|
| 5014 |
+
"learning_rate": 4.826195081905308e-05,
|
| 5015 |
+
"loss": 1.1807,
|
| 5016 |
+
"step": 64100
|
| 5017 |
+
},
|
| 5018 |
+
{
|
| 5019 |
+
"epoch": 0.0084,
|
| 5020 |
+
"grad_norm": 1.3917006254196167,
|
| 5021 |
+
"learning_rate": 4.8256133431386566e-05,
|
| 5022 |
+
"loss": 1.2012,
|
| 5023 |
+
"step": 64200
|
| 5024 |
+
},
|
| 5025 |
+
{
|
| 5026 |
+
"epoch": 0.0086,
|
| 5027 |
+
"grad_norm": 1.4448059797286987,
|
| 5028 |
+
"learning_rate": 4.825030667615228e-05,
|
| 5029 |
+
"loss": 1.2305,
|
| 5030 |
+
"step": 64300
|
| 5031 |
+
},
|
| 5032 |
+
{
|
| 5033 |
+
"epoch": 0.0088,
|
| 5034 |
+
"grad_norm": 1.0721293687820435,
|
| 5035 |
+
"learning_rate": 4.824447055569725e-05,
|
| 5036 |
+
"loss": 1.2332,
|
| 5037 |
+
"step": 64400
|
| 5038 |
+
},
|
| 5039 |
+
{
|
| 5040 |
+
"epoch": 0.009,
|
| 5041 |
+
"grad_norm": 0.9940403699874878,
|
| 5042 |
+
"learning_rate": 4.823862507237226e-05,
|
| 5043 |
+
"loss": 1.2096,
|
| 5044 |
+
"step": 64500
|
| 5045 |
+
},
|
| 5046 |
+
{
|
| 5047 |
+
"epoch": 0.0092,
|
| 5048 |
+
"grad_norm": 1.5473828315734863,
|
| 5049 |
+
"learning_rate": 4.823277022853187e-05,
|
| 5050 |
+
"loss": 1.1706,
|
| 5051 |
+
"step": 64600
|
| 5052 |
+
},
|
| 5053 |
+
{
|
| 5054 |
+
"epoch": 0.0094,
|
| 5055 |
+
"grad_norm": 1.3127409219741821,
|
| 5056 |
+
"learning_rate": 4.822690602653441e-05,
|
| 5057 |
+
"loss": 1.2051,
|
| 5058 |
+
"step": 64700
|
| 5059 |
+
},
|
| 5060 |
+
{
|
| 5061 |
+
"epoch": 0.0096,
|
| 5062 |
+
"grad_norm": 1.7532451152801514,
|
| 5063 |
+
"learning_rate": 4.822103246874198e-05,
|
| 5064 |
+
"loss": 1.196,
|
| 5065 |
+
"step": 64800
|
| 5066 |
+
},
|
| 5067 |
+
{
|
| 5068 |
+
"epoch": 0.0098,
|
| 5069 |
+
"grad_norm": 0.8706884980201721,
|
| 5070 |
+
"learning_rate": 4.8215149557520446e-05,
|
| 5071 |
+
"loss": 1.1862,
|
| 5072 |
+
"step": 64900
|
| 5073 |
+
},
|
| 5074 |
+
{
|
| 5075 |
+
"epoch": 0.01,
|
| 5076 |
+
"grad_norm": 1.5764431953430176,
|
| 5077 |
+
"learning_rate": 4.8209257295239455e-05,
|
| 5078 |
+
"loss": 1.2257,
|
| 5079 |
+
"step": 65000
|
| 5080 |
+
},
|
| 5081 |
+
{
|
| 5082 |
+
"epoch": 0.01,
|
| 5083 |
+
"eval_loss": 1.0817573070526123,
|
| 5084 |
+
"eval_runtime": 75.771,
|
| 5085 |
+
"eval_samples_per_second": 203.811,
|
| 5086 |
+
"eval_steps_per_second": 3.194,
|
| 5087 |
+
"step": 65000
|
| 5088 |
+
},
|
| 5089 |
+
{
|
| 5090 |
+
"epoch": 0.0102,
|
| 5091 |
+
"grad_norm": 1.467939019203186,
|
| 5092 |
+
"learning_rate": 4.820335568427239e-05,
|
| 5093 |
+
"loss": 1.2317,
|
| 5094 |
+
"step": 65100
|
| 5095 |
+
},
|
| 5096 |
+
{
|
| 5097 |
+
"epoch": 0.0104,
|
| 5098 |
+
"grad_norm": 1.270477533340454,
|
| 5099 |
+
"learning_rate": 4.819744472699643e-05,
|
| 5100 |
+
"loss": 1.2308,
|
| 5101 |
+
"step": 65200
|
| 5102 |
+
},
|
| 5103 |
+
{
|
| 5104 |
+
"epoch": 0.0106,
|
| 5105 |
+
"grad_norm": 1.073867917060852,
|
| 5106 |
+
"learning_rate": 4.8191524425792526e-05,
|
| 5107 |
+
"loss": 1.1991,
|
| 5108 |
+
"step": 65300
|
| 5109 |
+
},
|
| 5110 |
+
{
|
| 5111 |
+
"epoch": 0.0108,
|
| 5112 |
+
"grad_norm": 1.0844908952713013,
|
| 5113 |
+
"learning_rate": 4.818559478304534e-05,
|
| 5114 |
+
"loss": 1.1914,
|
| 5115 |
+
"step": 65400
|
| 5116 |
+
},
|
| 5117 |
+
{
|
| 5118 |
+
"epoch": 0.011,
|
| 5119 |
+
"grad_norm": 1.282365322113037,
|
| 5120 |
+
"learning_rate": 4.817965580114335e-05,
|
| 5121 |
+
"loss": 1.2035,
|
| 5122 |
+
"step": 65500
|
| 5123 |
+
},
|
| 5124 |
+
{
|
| 5125 |
+
"epoch": 0.0112,
|
| 5126 |
+
"grad_norm": 1.3751475811004639,
|
| 5127 |
+
"learning_rate": 4.817370748247878e-05,
|
| 5128 |
+
"loss": 1.215,
|
| 5129 |
+
"step": 65600
|
| 5130 |
+
},
|
| 5131 |
+
{
|
| 5132 |
+
"epoch": 0.0114,
|
| 5133 |
+
"grad_norm": 1.484107255935669,
|
| 5134 |
+
"learning_rate": 4.81677498294476e-05,
|
| 5135 |
+
"loss": 1.2298,
|
| 5136 |
+
"step": 65700
|
| 5137 |
+
},
|
| 5138 |
+
{
|
| 5139 |
+
"epoch": 0.0116,
|
| 5140 |
+
"grad_norm": 1.326803207397461,
|
| 5141 |
+
"learning_rate": 4.8161782844449566e-05,
|
| 5142 |
+
"loss": 1.1794,
|
| 5143 |
+
"step": 65800
|
| 5144 |
+
},
|
| 5145 |
+
{
|
| 5146 |
+
"epoch": 0.0118,
|
| 5147 |
+
"grad_norm": 1.6823039054870605,
|
| 5148 |
+
"learning_rate": 4.815580652988817e-05,
|
| 5149 |
+
"loss": 1.1896,
|
| 5150 |
+
"step": 65900
|
| 5151 |
+
},
|
| 5152 |
+
{
|
| 5153 |
+
"epoch": 0.012,
|
| 5154 |
+
"grad_norm": 1.1735076904296875,
|
| 5155 |
+
"learning_rate": 4.8149820888170673e-05,
|
| 5156 |
+
"loss": 1.2089,
|
| 5157 |
+
"step": 66000
|
| 5158 |
+
},
|
| 5159 |
+
{
|
| 5160 |
+
"epoch": 0.012,
|
| 5161 |
+
"eval_loss": 1.081894874572754,
|
| 5162 |
+
"eval_runtime": 75.5115,
|
| 5163 |
+
"eval_samples_per_second": 204.512,
|
| 5164 |
+
"eval_steps_per_second": 3.205,
|
| 5165 |
+
"step": 66000
|
| 5166 |
+
},
|
| 5167 |
+
{
|
| 5168 |
+
"epoch": 0.0122,
|
| 5169 |
+
"grad_norm": 1.0032376050949097,
|
| 5170 |
+
"learning_rate": 4.814382592170808e-05,
|
| 5171 |
+
"loss": 1.2197,
|
| 5172 |
+
"step": 66100
|
| 5173 |
+
},
|
| 5174 |
+
{
|
| 5175 |
+
"epoch": 0.0124,
|
| 5176 |
+
"grad_norm": 1.2638306617736816,
|
| 5177 |
+
"learning_rate": 4.813782163291519e-05,
|
| 5178 |
+
"loss": 1.2009,
|
| 5179 |
+
"step": 66200
|
| 5180 |
+
},
|
| 5181 |
+
{
|
| 5182 |
+
"epoch": 0.0126,
|
| 5183 |
+
"grad_norm": 1.2233041524887085,
|
| 5184 |
+
"learning_rate": 4.813180802421051e-05,
|
| 5185 |
+
"loss": 1.2069,
|
| 5186 |
+
"step": 66300
|
| 5187 |
+
},
|
| 5188 |
+
{
|
| 5189 |
+
"epoch": 0.0128,
|
| 5190 |
+
"grad_norm": 0.857825756072998,
|
| 5191 |
+
"learning_rate": 4.812578509801632e-05,
|
| 5192 |
+
"loss": 1.1942,
|
| 5193 |
+
"step": 66400
|
| 5194 |
+
},
|
| 5195 |
+
{
|
| 5196 |
+
"epoch": 0.013,
|
| 5197 |
+
"grad_norm": 0.8879494667053223,
|
| 5198 |
+
"learning_rate": 4.811975285675866e-05,
|
| 5199 |
+
"loss": 1.1689,
|
| 5200 |
+
"step": 66500
|
| 5201 |
+
},
|
| 5202 |
+
{
|
| 5203 |
+
"epoch": 0.0132,
|
| 5204 |
+
"grad_norm": 1.3842177391052246,
|
| 5205 |
+
"learning_rate": 4.811371130286731e-05,
|
| 5206 |
+
"loss": 1.1941,
|
| 5207 |
+
"step": 66600
|
| 5208 |
+
},
|
| 5209 |
+
{
|
| 5210 |
+
"epoch": 0.0134,
|
| 5211 |
+
"grad_norm": 1.303063988685608,
|
| 5212 |
+
"learning_rate": 4.810766043877582e-05,
|
| 5213 |
+
"loss": 1.194,
|
| 5214 |
+
"step": 66700
|
| 5215 |
+
},
|
| 5216 |
+
{
|
| 5217 |
+
"epoch": 0.0136,
|
| 5218 |
+
"grad_norm": 1.3135032653808594,
|
| 5219 |
+
"learning_rate": 4.810160026692147e-05,
|
| 5220 |
+
"loss": 1.1536,
|
| 5221 |
+
"step": 66800
|
| 5222 |
+
},
|
| 5223 |
+
{
|
| 5224 |
+
"epoch": 0.0138,
|
| 5225 |
+
"grad_norm": 0.8059789538383484,
|
| 5226 |
+
"learning_rate": 4.809553078974528e-05,
|
| 5227 |
+
"loss": 1.2083,
|
| 5228 |
+
"step": 66900
|
| 5229 |
+
},
|
| 5230 |
+
{
|
| 5231 |
+
"epoch": 0.014,
|
| 5232 |
+
"grad_norm": 1.493458867073059,
|
| 5233 |
+
"learning_rate": 4.808945200969206e-05,
|
| 5234 |
+
"loss": 1.2031,
|
| 5235 |
+
"step": 67000
|
| 5236 |
+
},
|
| 5237 |
+
{
|
| 5238 |
+
"epoch": 0.014,
|
| 5239 |
+
"eval_loss": 1.0807029008865356,
|
| 5240 |
+
"eval_runtime": 76.4746,
|
| 5241 |
+
"eval_samples_per_second": 201.936,
|
| 5242 |
+
"eval_steps_per_second": 3.164,
|
| 5243 |
+
"step": 67000
|
| 5244 |
+
},
|
| 5245 |
+
{
|
| 5246 |
+
"epoch": 0.0142,
|
| 5247 |
+
"grad_norm": 0.9932582378387451,
|
| 5248 |
+
"learning_rate": 4.808336392921033e-05,
|
| 5249 |
+
"loss": 1.1932,
|
| 5250 |
+
"step": 67100
|
| 5251 |
+
},
|
| 5252 |
+
{
|
| 5253 |
+
"epoch": 0.0144,
|
| 5254 |
+
"grad_norm": 1.1588648557662964,
|
| 5255 |
+
"learning_rate": 4.807726655075237e-05,
|
| 5256 |
+
"loss": 1.2004,
|
| 5257 |
+
"step": 67200
|
| 5258 |
+
},
|
| 5259 |
+
{
|
| 5260 |
+
"epoch": 0.0146,
|
| 5261 |
+
"grad_norm": 0.713295042514801,
|
| 5262 |
+
"learning_rate": 4.80711598767742e-05,
|
| 5263 |
+
"loss": 1.1336,
|
| 5264 |
+
"step": 67300
|
| 5265 |
+
},
|
| 5266 |
+
{
|
| 5267 |
+
"epoch": 0.0148,
|
| 5268 |
+
"grad_norm": 1.474853277206421,
|
| 5269 |
+
"learning_rate": 4.80650439097356e-05,
|
| 5270 |
+
"loss": 1.1909,
|
| 5271 |
+
"step": 67400
|
| 5272 |
+
},
|
| 5273 |
+
{
|
| 5274 |
+
"epoch": 0.015,
|
| 5275 |
+
"grad_norm": 1.0433249473571777,
|
| 5276 |
+
"learning_rate": 4.805891865210006e-05,
|
| 5277 |
+
"loss": 1.1868,
|
| 5278 |
+
"step": 67500
|
| 5279 |
+
},
|
| 5280 |
+
{
|
| 5281 |
+
"epoch": 0.0152,
|
| 5282 |
+
"grad_norm": 0.9942545294761658,
|
| 5283 |
+
"learning_rate": 4.8052784106334854e-05,
|
| 5284 |
+
"loss": 1.1896,
|
| 5285 |
+
"step": 67600
|
| 5286 |
+
},
|
| 5287 |
+
{
|
| 5288 |
+
"epoch": 0.0154,
|
| 5289 |
+
"grad_norm": 0.9021309018135071,
|
| 5290 |
+
"learning_rate": 4.804664027491096e-05,
|
| 5291 |
+
"loss": 1.2265,
|
| 5292 |
+
"step": 67700
|
| 5293 |
+
},
|
| 5294 |
+
{
|
| 5295 |
+
"epoch": 0.0156,
|
| 5296 |
+
"grad_norm": 1.4818402528762817,
|
| 5297 |
+
"learning_rate": 4.8040487160303126e-05,
|
| 5298 |
+
"loss": 1.2149,
|
| 5299 |
+
"step": 67800
|
| 5300 |
+
},
|
| 5301 |
+
{
|
| 5302 |
+
"epoch": 0.0158,
|
| 5303 |
+
"grad_norm": 0.74870365858078,
|
| 5304 |
+
"learning_rate": 4.803432476498981e-05,
|
| 5305 |
+
"loss": 1.1928,
|
| 5306 |
+
"step": 67900
|
| 5307 |
+
},
|
| 5308 |
+
{
|
| 5309 |
+
"epoch": 0.016,
|
| 5310 |
+
"grad_norm": 0.7827754020690918,
|
| 5311 |
+
"learning_rate": 4.8028153091453246e-05,
|
| 5312 |
+
"loss": 1.2062,
|
| 5313 |
+
"step": 68000
|
| 5314 |
+
},
|
| 5315 |
+
{
|
| 5316 |
+
"epoch": 0.016,
|
| 5317 |
+
"eval_loss": 1.0748348236083984,
|
| 5318 |
+
"eval_runtime": 75.9274,
|
| 5319 |
+
"eval_samples_per_second": 203.392,
|
| 5320 |
+
"eval_steps_per_second": 3.187,
|
| 5321 |
+
"step": 68000
|
| 5322 |
+
},
|
| 5323 |
+
{
|
| 5324 |
+
"epoch": 0.0162,
|
| 5325 |
+
"grad_norm": 1.2250913381576538,
|
| 5326 |
+
"learning_rate": 4.802197214217936e-05,
|
| 5327 |
+
"loss": 1.1412,
|
| 5328 |
+
"step": 68100
|
| 5329 |
+
},
|
| 5330 |
+
{
|
| 5331 |
+
"epoch": 0.0164,
|
| 5332 |
+
"grad_norm": 1.4763202667236328,
|
| 5333 |
+
"learning_rate": 4.801578191965785e-05,
|
| 5334 |
+
"loss": 1.173,
|
| 5335 |
+
"step": 68200
|
| 5336 |
+
},
|
| 5337 |
+
{
|
| 5338 |
+
"epoch": 0.0166,
|
| 5339 |
+
"grad_norm": 0.8980317115783691,
|
| 5340 |
+
"learning_rate": 4.800958242638214e-05,
|
| 5341 |
+
"loss": 1.1801,
|
| 5342 |
+
"step": 68300
|
| 5343 |
+
},
|
| 5344 |
+
{
|
| 5345 |
+
"epoch": 0.0168,
|
| 5346 |
+
"grad_norm": 1.2781926393508911,
|
| 5347 |
+
"learning_rate": 4.800337366484937e-05,
|
| 5348 |
+
"loss": 1.2012,
|
| 5349 |
+
"step": 68400
|
| 5350 |
+
},
|
| 5351 |
+
{
|
| 5352 |
+
"epoch": 0.017,
|
| 5353 |
+
"grad_norm": 0.8269230723381042,
|
| 5354 |
+
"learning_rate": 4.799715563756045e-05,
|
| 5355 |
+
"loss": 1.2319,
|
| 5356 |
+
"step": 68500
|
| 5357 |
+
},
|
| 5358 |
+
{
|
| 5359 |
+
"epoch": 0.0172,
|
| 5360 |
+
"grad_norm": 0.633537232875824,
|
| 5361 |
+
"learning_rate": 4.7990928347019984e-05,
|
| 5362 |
+
"loss": 1.2058,
|
| 5363 |
+
"step": 68600
|
| 5364 |
+
},
|
| 5365 |
+
{
|
| 5366 |
+
"epoch": 0.0174,
|
| 5367 |
+
"grad_norm": 1.39164400100708,
|
| 5368 |
+
"learning_rate": 4.7984691795736324e-05,
|
| 5369 |
+
"loss": 1.2066,
|
| 5370 |
+
"step": 68700
|
| 5371 |
+
},
|
| 5372 |
+
{
|
| 5373 |
+
"epoch": 0.0176,
|
| 5374 |
+
"grad_norm": 1.5555399656295776,
|
| 5375 |
+
"learning_rate": 4.7978445986221566e-05,
|
| 5376 |
+
"loss": 1.2088,
|
| 5377 |
+
"step": 68800
|
| 5378 |
+
},
|
| 5379 |
+
{
|
| 5380 |
+
"epoch": 0.0178,
|
| 5381 |
+
"grad_norm": 1.2505526542663574,
|
| 5382 |
+
"learning_rate": 4.7972190920991514e-05,
|
| 5383 |
+
"loss": 1.203,
|
| 5384 |
+
"step": 68900
|
| 5385 |
+
},
|
| 5386 |
+
{
|
| 5387 |
+
"epoch": 0.018,
|
| 5388 |
+
"grad_norm": 1.5910965204238892,
|
| 5389 |
+
"learning_rate": 4.7965926602565705e-05,
|
| 5390 |
+
"loss": 1.1877,
|
| 5391 |
+
"step": 69000
|
| 5392 |
+
},
|
| 5393 |
+
{
|
| 5394 |
+
"epoch": 0.018,
|
| 5395 |
+
"eval_loss": 1.0717748403549194,
|
| 5396 |
+
"eval_runtime": 75.7519,
|
| 5397 |
+
"eval_samples_per_second": 203.863,
|
| 5398 |
+
"eval_steps_per_second": 3.195,
|
| 5399 |
+
"step": 69000
|
| 5400 |
+
},
|
| 5401 |
+
{
|
| 5402 |
+
"epoch": 0.0182,
|
| 5403 |
+
"grad_norm": 0.7755507826805115,
|
| 5404 |
+
"learning_rate": 4.79596530334674e-05,
|
| 5405 |
+
"loss": 1.1864,
|
| 5406 |
+
"step": 69100
|
| 5407 |
+
},
|
| 5408 |
+
{
|
| 5409 |
+
"epoch": 0.0184,
|
| 5410 |
+
"grad_norm": 1.2141857147216797,
|
| 5411 |
+
"learning_rate": 4.79533702162236e-05,
|
| 5412 |
+
"loss": 1.1849,
|
| 5413 |
+
"step": 69200
|
| 5414 |
+
},
|
| 5415 |
+
{
|
| 5416 |
+
"epoch": 0.0186,
|
| 5417 |
+
"grad_norm": 1.399149775505066,
|
| 5418 |
+
"learning_rate": 4.794707815336503e-05,
|
| 5419 |
+
"loss": 1.1702,
|
| 5420 |
+
"step": 69300
|
| 5421 |
+
},
|
| 5422 |
+
{
|
| 5423 |
+
"epoch": 0.0188,
|
| 5424 |
+
"grad_norm": 1.3381379842758179,
|
| 5425 |
+
"learning_rate": 4.7940776847426114e-05,
|
| 5426 |
+
"loss": 1.2052,
|
| 5427 |
+
"step": 69400
|
| 5428 |
+
},
|
| 5429 |
+
{
|
| 5430 |
+
"epoch": 0.019,
|
| 5431 |
+
"grad_norm": 1.347264051437378,
|
| 5432 |
+
"learning_rate": 4.793446630094503e-05,
|
| 5433 |
+
"loss": 1.1998,
|
| 5434 |
+
"step": 69500
|
| 5435 |
+
},
|
| 5436 |
+
{
|
| 5437 |
+
"epoch": 0.0192,
|
| 5438 |
+
"grad_norm": 1.2072675228118896,
|
| 5439 |
+
"learning_rate": 4.792814651646367e-05,
|
| 5440 |
+
"loss": 1.2127,
|
| 5441 |
+
"step": 69600
|
| 5442 |
+
},
|
| 5443 |
+
{
|
| 5444 |
+
"epoch": 0.0194,
|
| 5445 |
+
"grad_norm": 0.7959086894989014,
|
| 5446 |
+
"learning_rate": 4.792181749652763e-05,
|
| 5447 |
+
"loss": 1.1474,
|
| 5448 |
+
"step": 69700
|
| 5449 |
+
},
|
| 5450 |
+
{
|
| 5451 |
+
"epoch": 0.0196,
|
| 5452 |
+
"grad_norm": 1.0272786617279053,
|
| 5453 |
+
"learning_rate": 4.7915479243686244e-05,
|
| 5454 |
+
"loss": 1.2033,
|
| 5455 |
+
"step": 69800
|
| 5456 |
+
},
|
| 5457 |
+
{
|
| 5458 |
+
"epoch": 0.0198,
|
| 5459 |
+
"grad_norm": 0.8985835909843445,
|
| 5460 |
+
"learning_rate": 4.790913176049256e-05,
|
| 5461 |
+
"loss": 1.1942,
|
| 5462 |
+
"step": 69900
|
| 5463 |
+
},
|
| 5464 |
+
{
|
| 5465 |
+
"epoch": 0.02,
|
| 5466 |
+
"grad_norm": 0.676177442073822,
|
| 5467 |
+
"learning_rate": 4.7902775049503346e-05,
|
| 5468 |
+
"loss": 1.1883,
|
| 5469 |
+
"step": 70000
|
| 5470 |
+
},
|
| 5471 |
+
{
|
| 5472 |
+
"epoch": 0.02,
|
| 5473 |
+
"eval_loss": 1.0733578205108643,
|
| 5474 |
+
"eval_runtime": 75.8186,
|
| 5475 |
+
"eval_samples_per_second": 203.684,
|
| 5476 |
+
"eval_steps_per_second": 3.192,
|
| 5477 |
+
"step": 70000
|
| 5478 |
+
},
|
| 5479 |
+
{
|
| 5480 |
+
"epoch": 0.0202,
|
| 5481 |
+
"grad_norm": 0.7747570872306824,
|
| 5482 |
+
"learning_rate": 4.789640911327907e-05,
|
| 5483 |
+
"loss": 1.1883,
|
| 5484 |
+
"step": 70100
|
| 5485 |
+
},
|
| 5486 |
+
{
|
| 5487 |
+
"epoch": 0.0204,
|
| 5488 |
+
"grad_norm": 1.1808815002441406,
|
| 5489 |
+
"learning_rate": 4.789003395438395e-05,
|
| 5490 |
+
"loss": 1.1932,
|
| 5491 |
+
"step": 70200
|
| 5492 |
+
},
|
| 5493 |
+
{
|
| 5494 |
+
"epoch": 0.0206,
|
| 5495 |
+
"grad_norm": 1.29102623462677,
|
| 5496 |
+
"learning_rate": 4.7883649575385894e-05,
|
| 5497 |
+
"loss": 1.1654,
|
| 5498 |
+
"step": 70300
|
| 5499 |
+
},
|
| 5500 |
+
{
|
| 5501 |
+
"epoch": 0.0208,
|
| 5502 |
+
"grad_norm": 0.8418052792549133,
|
| 5503 |
+
"learning_rate": 4.7877255978856516e-05,
|
| 5504 |
+
"loss": 1.1702,
|
| 5505 |
+
"step": 70400
|
| 5506 |
+
},
|
| 5507 |
+
{
|
| 5508 |
+
"epoch": 0.021,
|
| 5509 |
+
"grad_norm": 1.1825124025344849,
|
| 5510 |
+
"learning_rate": 4.787085316737116e-05,
|
| 5511 |
+
"loss": 1.1654,
|
| 5512 |
+
"step": 70500
|
| 5513 |
+
},
|
| 5514 |
+
{
|
| 5515 |
+
"epoch": 0.0212,
|
| 5516 |
+
"grad_norm": 1.301255702972412,
|
| 5517 |
+
"learning_rate": 4.78644411435089e-05,
|
| 5518 |
+
"loss": 1.2505,
|
| 5519 |
+
"step": 70600
|
| 5520 |
+
},
|
| 5521 |
+
{
|
| 5522 |
+
"epoch": 0.0214,
|
| 5523 |
+
"grad_norm": 1.2461885213851929,
|
| 5524 |
+
"learning_rate": 4.785801990985247e-05,
|
| 5525 |
+
"loss": 1.1907,
|
| 5526 |
+
"step": 70700
|
| 5527 |
+
},
|
| 5528 |
+
{
|
| 5529 |
+
"epoch": 0.0216,
|
| 5530 |
+
"grad_norm": 1.2957687377929688,
|
| 5531 |
+
"learning_rate": 4.7851589468988364e-05,
|
| 5532 |
+
"loss": 1.2244,
|
| 5533 |
+
"step": 70800
|
| 5534 |
+
},
|
| 5535 |
+
{
|
| 5536 |
+
"epoch": 0.0218,
|
| 5537 |
+
"grad_norm": 1.9566733837127686,
|
| 5538 |
+
"learning_rate": 4.7845149823506744e-05,
|
| 5539 |
+
"loss": 1.1688,
|
| 5540 |
+
"step": 70900
|
| 5541 |
+
},
|
| 5542 |
+
{
|
| 5543 |
+
"epoch": 0.022,
|
| 5544 |
+
"grad_norm": 0.9749345779418945,
|
| 5545 |
+
"learning_rate": 4.783870097600151e-05,
|
| 5546 |
+
"loss": 1.2178,
|
| 5547 |
+
"step": 71000
|
| 5548 |
+
},
|
| 5549 |
+
{
|
| 5550 |
+
"epoch": 0.022,
|
| 5551 |
+
"eval_loss": 1.076163649559021,
|
| 5552 |
+
"eval_runtime": 75.78,
|
| 5553 |
+
"eval_samples_per_second": 203.787,
|
| 5554 |
+
"eval_steps_per_second": 3.193,
|
| 5555 |
+
"step": 71000
|
| 5556 |
+
},
|
| 5557 |
+
{
|
| 5558 |
+
"epoch": 0.0222,
|
| 5559 |
+
"grad_norm": 1.1278064250946045,
|
| 5560 |
+
"learning_rate": 4.783224292907025e-05,
|
| 5561 |
+
"loss": 1.1899,
|
| 5562 |
+
"step": 71100
|
| 5563 |
+
},
|
| 5564 |
+
{
|
| 5565 |
+
"epoch": 0.0224,
|
| 5566 |
+
"grad_norm": 1.023586392402649,
|
| 5567 |
+
"learning_rate": 4.7825775685314277e-05,
|
| 5568 |
+
"loss": 1.1967,
|
| 5569 |
+
"step": 71200
|
| 5570 |
+
},
|
| 5571 |
+
{
|
| 5572 |
+
"epoch": 0.0226,
|
| 5573 |
+
"grad_norm": 1.2925764322280884,
|
| 5574 |
+
"learning_rate": 4.781929924733858e-05,
|
| 5575 |
+
"loss": 1.2154,
|
| 5576 |
+
"step": 71300
|
| 5577 |
+
},
|
| 5578 |
+
{
|
| 5579 |
+
"epoch": 0.0228,
|
| 5580 |
+
"grad_norm": 0.8185212016105652,
|
| 5581 |
+
"learning_rate": 4.781281361775188e-05,
|
| 5582 |
+
"loss": 1.195,
|
| 5583 |
+
"step": 71400
|
| 5584 |
+
},
|
| 5585 |
+
{
|
| 5586 |
+
"epoch": 0.023,
|
| 5587 |
+
"grad_norm": 0.8742319941520691,
|
| 5588 |
+
"learning_rate": 4.7806318799166586e-05,
|
| 5589 |
+
"loss": 1.1746,
|
| 5590 |
+
"step": 71500
|
| 5591 |
+
},
|
| 5592 |
+
{
|
| 5593 |
+
"epoch": 0.0232,
|
| 5594 |
+
"grad_norm": 1.2598085403442383,
|
| 5595 |
+
"learning_rate": 4.77998147941988e-05,
|
| 5596 |
+
"loss": 1.1781,
|
| 5597 |
+
"step": 71600
|
| 5598 |
+
},
|
| 5599 |
+
{
|
| 5600 |
+
"epoch": 0.0234,
|
| 5601 |
+
"grad_norm": 1.2358424663543701,
|
| 5602 |
+
"learning_rate": 4.7793301605468344e-05,
|
| 5603 |
+
"loss": 1.2345,
|
| 5604 |
+
"step": 71700
|
| 5605 |
+
},
|
| 5606 |
+
{
|
| 5607 |
+
"epoch": 0.0236,
|
| 5608 |
+
"grad_norm": 1.2528828382492065,
|
| 5609 |
+
"learning_rate": 4.778677923559872e-05,
|
| 5610 |
+
"loss": 1.2109,
|
| 5611 |
+
"step": 71800
|
| 5612 |
+
},
|
| 5613 |
+
{
|
| 5614 |
+
"epoch": 0.0238,
|
| 5615 |
+
"grad_norm": 0.5741105675697327,
|
| 5616 |
+
"learning_rate": 4.778024768721716e-05,
|
| 5617 |
+
"loss": 1.2076,
|
| 5618 |
+
"step": 71900
|
| 5619 |
+
},
|
| 5620 |
+
{
|
| 5621 |
+
"epoch": 0.024,
|
| 5622 |
+
"grad_norm": 1.3200185298919678,
|
| 5623 |
+
"learning_rate": 4.7773706962954545e-05,
|
| 5624 |
+
"loss": 1.2124,
|
| 5625 |
+
"step": 72000
|
| 5626 |
+
},
|
| 5627 |
+
{
|
| 5628 |
+
"epoch": 0.024,
|
| 5629 |
+
"eval_loss": 1.0720120668411255,
|
| 5630 |
+
"eval_runtime": 76.4471,
|
| 5631 |
+
"eval_samples_per_second": 202.009,
|
| 5632 |
+
"eval_steps_per_second": 3.166,
|
| 5633 |
+
"step": 72000
|
| 5634 |
+
},
|
| 5635 |
+
{
|
| 5636 |
+
"epoch": 0.0242,
|
| 5637 |
+
"grad_norm": 1.4096635580062866,
|
| 5638 |
+
"learning_rate": 4.776715706544549e-05,
|
| 5639 |
+
"loss": 1.2283,
|
| 5640 |
+
"step": 72100
|
| 5641 |
+
},
|
| 5642 |
+
{
|
| 5643 |
+
"epoch": 0.0244,
|
| 5644 |
+
"grad_norm": 1.5862853527069092,
|
| 5645 |
+
"learning_rate": 4.7760597997328295e-05,
|
| 5646 |
+
"loss": 1.1927,
|
| 5647 |
+
"step": 72200
|
| 5648 |
+
},
|
| 5649 |
+
{
|
| 5650 |
+
"epoch": 0.0246,
|
| 5651 |
+
"grad_norm": 1.3406593799591064,
|
| 5652 |
+
"learning_rate": 4.7754029761244945e-05,
|
| 5653 |
+
"loss": 1.1709,
|
| 5654 |
+
"step": 72300
|
| 5655 |
+
},
|
| 5656 |
+
{
|
| 5657 |
+
"epoch": 0.0248,
|
| 5658 |
+
"grad_norm": 1.189676284790039,
|
| 5659 |
+
"learning_rate": 4.774745235984113e-05,
|
| 5660 |
+
"loss": 1.2176,
|
| 5661 |
+
"step": 72400
|
| 5662 |
+
},
|
| 5663 |
+
{
|
| 5664 |
+
"epoch": 0.025,
|
| 5665 |
+
"grad_norm": 1.4424960613250732,
|
| 5666 |
+
"learning_rate": 4.7740865795766224e-05,
|
| 5667 |
+
"loss": 1.2212,
|
| 5668 |
+
"step": 72500
|
| 5669 |
+
},
|
| 5670 |
+
{
|
| 5671 |
+
"epoch": 0.0252,
|
| 5672 |
+
"grad_norm": 0.7654275298118591,
|
| 5673 |
+
"learning_rate": 4.77342700716733e-05,
|
| 5674 |
+
"loss": 1.2196,
|
| 5675 |
+
"step": 72600
|
| 5676 |
+
},
|
| 5677 |
+
{
|
| 5678 |
+
"epoch": 0.0254,
|
| 5679 |
+
"grad_norm": 1.1389504671096802,
|
| 5680 |
+
"learning_rate": 4.772766519021911e-05,
|
| 5681 |
+
"loss": 1.1937,
|
| 5682 |
+
"step": 72700
|
| 5683 |
+
},
|
| 5684 |
+
{
|
| 5685 |
+
"epoch": 0.0256,
|
| 5686 |
+
"grad_norm": 1.1204986572265625,
|
| 5687 |
+
"learning_rate": 4.772105115406409e-05,
|
| 5688 |
+
"loss": 1.1623,
|
| 5689 |
+
"step": 72800
|
| 5690 |
+
},
|
| 5691 |
+
{
|
| 5692 |
+
"epoch": 0.0258,
|
| 5693 |
+
"grad_norm": 1.2594044208526611,
|
| 5694 |
+
"learning_rate": 4.771442796587239e-05,
|
| 5695 |
+
"loss": 1.2127,
|
| 5696 |
+
"step": 72900
|
| 5697 |
+
},
|
| 5698 |
+
{
|
| 5699 |
+
"epoch": 0.026,
|
| 5700 |
+
"grad_norm": 1.3245586156845093,
|
| 5701 |
+
"learning_rate": 4.770779562831181e-05,
|
| 5702 |
+
"loss": 1.1919,
|
| 5703 |
+
"step": 73000
|
| 5704 |
+
},
|
| 5705 |
+
{
|
| 5706 |
+
"epoch": 0.026,
|
| 5707 |
+
"eval_loss": 1.0672369003295898,
|
| 5708 |
+
"eval_runtime": 76.1554,
|
| 5709 |
+
"eval_samples_per_second": 202.783,
|
| 5710 |
+
"eval_steps_per_second": 3.178,
|
| 5711 |
+
"step": 73000
|
| 5712 |
+
},
|
| 5713 |
+
{
|
| 5714 |
+
"epoch": 0.0262,
|
| 5715 |
+
"grad_norm": 0.813410222530365,
|
| 5716 |
+
"learning_rate": 4.770115414405388e-05,
|
| 5717 |
+
"loss": 1.224,
|
| 5718 |
+
"step": 73100
|
| 5719 |
+
},
|
| 5720 |
+
{
|
| 5721 |
+
"epoch": 0.0264,
|
| 5722 |
+
"grad_norm": 1.3278921842575073,
|
| 5723 |
+
"learning_rate": 4.769450351577377e-05,
|
| 5724 |
+
"loss": 1.2304,
|
| 5725 |
+
"step": 73200
|
| 5726 |
+
},
|
| 5727 |
+
{
|
| 5728 |
+
"epoch": 0.0266,
|
| 5729 |
+
"grad_norm": 1.1676868200302124,
|
| 5730 |
+
"learning_rate": 4.768784374615036e-05,
|
| 5731 |
+
"loss": 1.2144,
|
| 5732 |
+
"step": 73300
|
| 5733 |
+
},
|
| 5734 |
+
{
|
| 5735 |
+
"epoch": 0.0268,
|
| 5736 |
+
"grad_norm": 1.2270694971084595,
|
| 5737 |
+
"learning_rate": 4.7681174837866196e-05,
|
| 5738 |
+
"loss": 1.2584,
|
| 5739 |
+
"step": 73400
|
| 5740 |
+
},
|
| 5741 |
+
{
|
| 5742 |
+
"epoch": 0.027,
|
| 5743 |
+
"grad_norm": 1.5095762014389038,
|
| 5744 |
+
"learning_rate": 4.7674496793607525e-05,
|
| 5745 |
+
"loss": 1.1892,
|
| 5746 |
+
"step": 73500
|
| 5747 |
+
},
|
| 5748 |
+
{
|
| 5749 |
+
"epoch": 0.0272,
|
| 5750 |
+
"grad_norm": 1.0437262058258057,
|
| 5751 |
+
"learning_rate": 4.766780961606426e-05,
|
| 5752 |
+
"loss": 1.2003,
|
| 5753 |
+
"step": 73600
|
| 5754 |
+
},
|
| 5755 |
+
{
|
| 5756 |
+
"epoch": 0.0274,
|
| 5757 |
+
"grad_norm": 0.6719204187393188,
|
| 5758 |
+
"learning_rate": 4.766111330793e-05,
|
| 5759 |
+
"loss": 1.2145,
|
| 5760 |
+
"step": 73700
|
| 5761 |
+
},
|
| 5762 |
+
{
|
| 5763 |
+
"epoch": 0.0276,
|
| 5764 |
+
"grad_norm": 0.7166513204574585,
|
| 5765 |
+
"learning_rate": 4.765440787190199e-05,
|
| 5766 |
+
"loss": 1.2463,
|
| 5767 |
+
"step": 73800
|
| 5768 |
+
},
|
| 5769 |
+
{
|
| 5770 |
+
"epoch": 0.0278,
|
| 5771 |
+
"grad_norm": 0.9765319228172302,
|
| 5772 |
+
"learning_rate": 4.7647693310681204e-05,
|
| 5773 |
+
"loss": 1.2095,
|
| 5774 |
+
"step": 73900
|
| 5775 |
+
},
|
| 5776 |
+
{
|
| 5777 |
+
"epoch": 0.028,
|
| 5778 |
+
"grad_norm": 1.298134446144104,
|
| 5779 |
+
"learning_rate": 4.7640969626972265e-05,
|
| 5780 |
+
"loss": 1.2089,
|
| 5781 |
+
"step": 74000
|
| 5782 |
+
},
|
| 5783 |
+
{
|
| 5784 |
+
"epoch": 0.028,
|
| 5785 |
+
"eval_loss": 1.0727263689041138,
|
| 5786 |
+
"eval_runtime": 76.0016,
|
| 5787 |
+
"eval_samples_per_second": 203.193,
|
| 5788 |
+
"eval_steps_per_second": 3.184,
|
| 5789 |
+
"step": 74000
|
| 5790 |
+
},
|
| 5791 |
+
{
|
| 5792 |
+
"epoch": 0.0282,
|
| 5793 |
+
"grad_norm": 1.1968761682510376,
|
| 5794 |
+
"learning_rate": 4.763423682348347e-05,
|
| 5795 |
+
"loss": 1.1719,
|
| 5796 |
+
"step": 74100
|
| 5797 |
+
},
|
| 5798 |
+
{
|
| 5799 |
+
"epoch": 0.0284,
|
| 5800 |
+
"grad_norm": 1.1887174844741821,
|
| 5801 |
+
"learning_rate": 4.762749490292678e-05,
|
| 5802 |
+
"loss": 1.1961,
|
| 5803 |
+
"step": 74200
|
| 5804 |
+
},
|
| 5805 |
+
{
|
| 5806 |
+
"epoch": 0.0286,
|
| 5807 |
+
"grad_norm": 1.4029371738433838,
|
| 5808 |
+
"learning_rate": 4.762074386801786e-05,
|
| 5809 |
+
"loss": 1.1609,
|
| 5810 |
+
"step": 74300
|
| 5811 |
+
},
|
| 5812 |
+
{
|
| 5813 |
+
"epoch": 0.0288,
|
| 5814 |
+
"grad_norm": 1.3785122632980347,
|
| 5815 |
+
"learning_rate": 4.761398372147601e-05,
|
| 5816 |
+
"loss": 1.1931,
|
| 5817 |
+
"step": 74400
|
| 5818 |
+
},
|
| 5819 |
+
{
|
| 5820 |
+
"epoch": 0.029,
|
| 5821 |
+
"grad_norm": 1.1329565048217773,
|
| 5822 |
+
"learning_rate": 4.760721446602422e-05,
|
| 5823 |
+
"loss": 1.2107,
|
| 5824 |
+
"step": 74500
|
| 5825 |
+
},
|
| 5826 |
+
{
|
| 5827 |
+
"epoch": 0.0292,
|
| 5828 |
+
"grad_norm": 1.2266113758087158,
|
| 5829 |
+
"learning_rate": 4.760043610438915e-05,
|
| 5830 |
+
"loss": 1.1708,
|
| 5831 |
+
"step": 74600
|
| 5832 |
+
},
|
| 5833 |
+
{
|
| 5834 |
+
"epoch": 0.0294,
|
| 5835 |
+
"grad_norm": 1.2526196241378784,
|
| 5836 |
+
"learning_rate": 4.759364863930112e-05,
|
| 5837 |
+
"loss": 1.2073,
|
| 5838 |
+
"step": 74700
|
| 5839 |
+
},
|
| 5840 |
+
{
|
| 5841 |
+
"epoch": 0.0296,
|
| 5842 |
+
"grad_norm": 1.3959336280822754,
|
| 5843 |
+
"learning_rate": 4.7586852073494113e-05,
|
| 5844 |
+
"loss": 1.1995,
|
| 5845 |
+
"step": 74800
|
| 5846 |
+
},
|
| 5847 |
+
{
|
| 5848 |
+
"epoch": 0.0298,
|
| 5849 |
+
"grad_norm": 1.2470852136611938,
|
| 5850 |
+
"learning_rate": 4.7580046409705806e-05,
|
| 5851 |
+
"loss": 1.2227,
|
| 5852 |
+
"step": 74900
|
| 5853 |
+
},
|
| 5854 |
+
{
|
| 5855 |
+
"epoch": 0.03,
|
| 5856 |
+
"grad_norm": 1.0915220975875854,
|
| 5857 |
+
"learning_rate": 4.7573231650677495e-05,
|
| 5858 |
+
"loss": 1.1955,
|
| 5859 |
+
"step": 75000
|
| 5860 |
+
},
|
| 5861 |
+
{
|
| 5862 |
+
"epoch": 0.03,
|
| 5863 |
+
"eval_loss": 1.0732471942901611,
|
| 5864 |
+
"eval_runtime": 75.8455,
|
| 5865 |
+
"eval_samples_per_second": 203.611,
|
| 5866 |
+
"eval_steps_per_second": 3.191,
|
| 5867 |
+
"step": 75000
|
| 5868 |
+
},
|
| 5869 |
+
{
|
| 5870 |
+
"epoch": 0.0302,
|
| 5871 |
+
"grad_norm": 1.4608689546585083,
|
| 5872 |
+
"learning_rate": 4.756640779915418e-05,
|
| 5873 |
+
"loss": 1.1588,
|
| 5874 |
+
"step": 75100
|
| 5875 |
+
},
|
| 5876 |
+
{
|
| 5877 |
+
"epoch": 0.0304,
|
| 5878 |
+
"grad_norm": 1.2811450958251953,
|
| 5879 |
+
"learning_rate": 4.755957485788449e-05,
|
| 5880 |
+
"loss": 1.1722,
|
| 5881 |
+
"step": 75200
|
| 5882 |
+
},
|
| 5883 |
+
{
|
| 5884 |
+
"epoch": 0.0306,
|
| 5885 |
+
"grad_norm": 1.3260635137557983,
|
| 5886 |
+
"learning_rate": 4.755273282962075e-05,
|
| 5887 |
+
"loss": 1.2238,
|
| 5888 |
+
"step": 75300
|
| 5889 |
+
},
|
| 5890 |
+
{
|
| 5891 |
+
"epoch": 0.0308,
|
| 5892 |
+
"grad_norm": 1.219567060470581,
|
| 5893 |
+
"learning_rate": 4.754588171711893e-05,
|
| 5894 |
+
"loss": 1.2718,
|
| 5895 |
+
"step": 75400
|
| 5896 |
+
},
|
| 5897 |
+
{
|
| 5898 |
+
"epoch": 0.031,
|
| 5899 |
+
"grad_norm": 1.368947982788086,
|
| 5900 |
+
"learning_rate": 4.753902152313865e-05,
|
| 5901 |
+
"loss": 1.1998,
|
| 5902 |
+
"step": 75500
|
| 5903 |
+
},
|
| 5904 |
+
{
|
| 5905 |
+
"epoch": 0.0312,
|
| 5906 |
+
"grad_norm": 1.3364487886428833,
|
| 5907 |
+
"learning_rate": 4.7532152250443194e-05,
|
| 5908 |
+
"loss": 1.2043,
|
| 5909 |
+
"step": 75600
|
| 5910 |
+
},
|
| 5911 |
+
{
|
| 5912 |
+
"epoch": 0.0314,
|
| 5913 |
+
"grad_norm": 1.348130702972412,
|
| 5914 |
+
"learning_rate": 4.7525273901799506e-05,
|
| 5915 |
+
"loss": 1.1834,
|
| 5916 |
+
"step": 75700
|
| 5917 |
+
},
|
| 5918 |
+
{
|
| 5919 |
+
"epoch": 0.0316,
|
| 5920 |
+
"grad_norm": 1.1862202882766724,
|
| 5921 |
+
"learning_rate": 4.751838647997818e-05,
|
| 5922 |
+
"loss": 1.2061,
|
| 5923 |
+
"step": 75800
|
| 5924 |
+
},
|
| 5925 |
+
{
|
| 5926 |
+
"epoch": 0.0318,
|
| 5927 |
+
"grad_norm": 0.7471460103988647,
|
| 5928 |
+
"learning_rate": 4.7511489987753476e-05,
|
| 5929 |
+
"loss": 1.1866,
|
| 5930 |
+
"step": 75900
|
| 5931 |
+
},
|
| 5932 |
+
{
|
| 5933 |
+
"epoch": 0.032,
|
| 5934 |
+
"grad_norm": 1.4090434312820435,
|
| 5935 |
+
"learning_rate": 4.75045844279033e-05,
|
| 5936 |
+
"loss": 1.1878,
|
| 5937 |
+
"step": 76000
|
| 5938 |
+
},
|
| 5939 |
+
{
|
| 5940 |
+
"epoch": 0.032,
|
| 5941 |
+
"eval_loss": 1.0745600461959839,
|
| 5942 |
+
"eval_runtime": 76.306,
|
| 5943 |
+
"eval_samples_per_second": 202.382,
|
| 5944 |
+
"eval_steps_per_second": 3.171,
|
| 5945 |
+
"step": 76000
|
| 5946 |
+
},
|
| 5947 |
+
{
|
| 5948 |
+
"epoch": 0.0322,
|
| 5949 |
+
"grad_norm": 1.591199278831482,
|
| 5950 |
+
"learning_rate": 4.7497669803209204e-05,
|
| 5951 |
+
"loss": 1.1824,
|
| 5952 |
+
"step": 76100
|
| 5953 |
+
},
|
| 5954 |
+
{
|
| 5955 |
+
"epoch": 0.0324,
|
| 5956 |
+
"grad_norm": 0.8325656652450562,
|
| 5957 |
+
"learning_rate": 4.749074611645641e-05,
|
| 5958 |
+
"loss": 1.1723,
|
| 5959 |
+
"step": 76200
|
| 5960 |
+
},
|
| 5961 |
+
{
|
| 5962 |
+
"epoch": 0.0326,
|
| 5963 |
+
"grad_norm": 0.8313425779342651,
|
| 5964 |
+
"learning_rate": 4.748381337043376e-05,
|
| 5965 |
+
"loss": 1.2033,
|
| 5966 |
+
"step": 76300
|
| 5967 |
+
},
|
| 5968 |
+
{
|
| 5969 |
+
"epoch": 0.0328,
|
| 5970 |
+
"grad_norm": 1.4721826314926147,
|
| 5971 |
+
"learning_rate": 4.7476871567933775e-05,
|
| 5972 |
+
"loss": 1.1988,
|
| 5973 |
+
"step": 76400
|
| 5974 |
+
},
|
| 5975 |
+
{
|
| 5976 |
+
"epoch": 0.033,
|
| 5977 |
+
"grad_norm": 0.9206506013870239,
|
| 5978 |
+
"learning_rate": 4.746992071175261e-05,
|
| 5979 |
+
"loss": 1.1844,
|
| 5980 |
+
"step": 76500
|
| 5981 |
+
},
|
| 5982 |
+
{
|
| 5983 |
+
"epoch": 0.0332,
|
| 5984 |
+
"grad_norm": 1.0820422172546387,
|
| 5985 |
+
"learning_rate": 4.746296080469007e-05,
|
| 5986 |
+
"loss": 1.1902,
|
| 5987 |
+
"step": 76600
|
| 5988 |
+
},
|
| 5989 |
+
{
|
| 5990 |
+
"epoch": 0.0334,
|
| 5991 |
+
"grad_norm": 0.9319769144058228,
|
| 5992 |
+
"learning_rate": 4.745599184954961e-05,
|
| 5993 |
+
"loss": 1.2031,
|
| 5994 |
+
"step": 76700
|
| 5995 |
+
},
|
| 5996 |
+
{
|
| 5997 |
+
"epoch": 0.0336,
|
| 5998 |
+
"grad_norm": 1.1914819478988647,
|
| 5999 |
+
"learning_rate": 4.744901384913831e-05,
|
| 6000 |
+
"loss": 1.166,
|
| 6001 |
+
"step": 76800
|
| 6002 |
+
},
|
| 6003 |
+
{
|
| 6004 |
+
"epoch": 0.0338,
|
| 6005 |
+
"grad_norm": 0.8440219163894653,
|
| 6006 |
+
"learning_rate": 4.7442026806266914e-05,
|
| 6007 |
+
"loss": 1.1493,
|
| 6008 |
+
"step": 76900
|
| 6009 |
+
},
|
| 6010 |
+
{
|
| 6011 |
+
"epoch": 0.034,
|
| 6012 |
+
"grad_norm": 1.001518726348877,
|
| 6013 |
+
"learning_rate": 4.7435030723749813e-05,
|
| 6014 |
+
"loss": 1.1835,
|
| 6015 |
+
"step": 77000
|
| 6016 |
+
},
|
| 6017 |
+
{
|
| 6018 |
+
"epoch": 0.034,
|
| 6019 |
+
"eval_loss": 1.0681182146072388,
|
| 6020 |
+
"eval_runtime": 76.1301,
|
| 6021 |
+
"eval_samples_per_second": 202.85,
|
| 6022 |
+
"eval_steps_per_second": 3.179,
|
| 6023 |
+
"step": 77000
|
| 6024 |
+
},
|
| 6025 |
+
{
|
| 6026 |
+
"epoch": 0.0342,
|
| 6027 |
+
"grad_norm": 1.347307562828064,
|
| 6028 |
+
"learning_rate": 4.742802560440501e-05,
|
| 6029 |
+
"loss": 1.2213,
|
| 6030 |
+
"step": 77100
|
| 6031 |
+
},
|
| 6032 |
+
{
|
| 6033 |
+
"epoch": 0.0344,
|
| 6034 |
+
"grad_norm": 1.1187894344329834,
|
| 6035 |
+
"learning_rate": 4.742101145105419e-05,
|
| 6036 |
+
"loss": 1.1949,
|
| 6037 |
+
"step": 77200
|
| 6038 |
+
},
|
| 6039 |
+
{
|
| 6040 |
+
"epoch": 0.0346,
|
| 6041 |
+
"grad_norm": 0.8066337704658508,
|
| 6042 |
+
"learning_rate": 4.741398826652262e-05,
|
| 6043 |
+
"loss": 1.2008,
|
| 6044 |
+
"step": 77300
|
| 6045 |
+
},
|
| 6046 |
+
{
|
| 6047 |
+
"epoch": 0.0348,
|
| 6048 |
+
"grad_norm": 1.0704104900360107,
|
| 6049 |
+
"learning_rate": 4.740695605363927e-05,
|
| 6050 |
+
"loss": 1.1804,
|
| 6051 |
+
"step": 77400
|
| 6052 |
+
},
|
| 6053 |
+
{
|
| 6054 |
+
"epoch": 0.035,
|
| 6055 |
+
"grad_norm": 1.104546070098877,
|
| 6056 |
+
"learning_rate": 4.7399914815236704e-05,
|
| 6057 |
+
"loss": 1.2232,
|
| 6058 |
+
"step": 77500
|
| 6059 |
+
},
|
| 6060 |
+
{
|
| 6061 |
+
"epoch": 0.0352,
|
| 6062 |
+
"grad_norm": 1.1818023920059204,
|
| 6063 |
+
"learning_rate": 4.7392864554151126e-05,
|
| 6064 |
+
"loss": 1.2062,
|
| 6065 |
+
"step": 77600
|
| 6066 |
+
},
|
| 6067 |
+
{
|
| 6068 |
+
"epoch": 0.0354,
|
| 6069 |
+
"grad_norm": 1.3036936521530151,
|
| 6070 |
+
"learning_rate": 4.738580527322238e-05,
|
| 6071 |
+
"loss": 1.1905,
|
| 6072 |
+
"step": 77700
|
| 6073 |
+
},
|
| 6074 |
+
{
|
| 6075 |
+
"epoch": 0.0356,
|
| 6076 |
+
"grad_norm": 1.1169214248657227,
|
| 6077 |
+
"learning_rate": 4.737873697529395e-05,
|
| 6078 |
+
"loss": 1.1759,
|
| 6079 |
+
"step": 77800
|
| 6080 |
+
},
|
| 6081 |
+
{
|
| 6082 |
+
"epoch": 0.0358,
|
| 6083 |
+
"grad_norm": 0.8993995189666748,
|
| 6084 |
+
"learning_rate": 4.7371659663212934e-05,
|
| 6085 |
+
"loss": 1.1677,
|
| 6086 |
+
"step": 77900
|
| 6087 |
+
},
|
| 6088 |
+
{
|
| 6089 |
+
"epoch": 0.036,
|
| 6090 |
+
"grad_norm": 1.258747935295105,
|
| 6091 |
+
"learning_rate": 4.736457333983009e-05,
|
| 6092 |
+
"loss": 1.2166,
|
| 6093 |
+
"step": 78000
|
| 6094 |
+
},
|
| 6095 |
+
{
|
| 6096 |
+
"epoch": 0.036,
|
| 6097 |
+
"eval_loss": 1.0701075792312622,
|
| 6098 |
+
"eval_runtime": 75.9209,
|
| 6099 |
+
"eval_samples_per_second": 203.409,
|
| 6100 |
+
"eval_steps_per_second": 3.188,
|
| 6101 |
+
"step": 78000
|
| 6102 |
+
},
|
| 6103 |
+
{
|
| 6104 |
+
"epoch": 0.0362,
|
| 6105 |
+
"grad_norm": 1.269551396369934,
|
| 6106 |
+
"learning_rate": 4.735747800799978e-05,
|
| 6107 |
+
"loss": 1.2185,
|
| 6108 |
+
"step": 78100
|
| 6109 |
+
},
|
| 6110 |
+
{
|
| 6111 |
+
"epoch": 0.0364,
|
| 6112 |
+
"grad_norm": 1.3016119003295898,
|
| 6113 |
+
"learning_rate": 4.735037367057999e-05,
|
| 6114 |
+
"loss": 1.182,
|
| 6115 |
+
"step": 78200
|
| 6116 |
+
},
|
| 6117 |
+
{
|
| 6118 |
+
"epoch": 0.0366,
|
| 6119 |
+
"grad_norm": 1.1407994031906128,
|
| 6120 |
+
"learning_rate": 4.734326033043238e-05,
|
| 6121 |
+
"loss": 1.2102,
|
| 6122 |
+
"step": 78300
|
| 6123 |
+
},
|
| 6124 |
+
{
|
| 6125 |
+
"epoch": 0.0368,
|
| 6126 |
+
"grad_norm": 1.1673243045806885,
|
| 6127 |
+
"learning_rate": 4.7336137990422164e-05,
|
| 6128 |
+
"loss": 1.1902,
|
| 6129 |
+
"step": 78400
|
| 6130 |
+
},
|
| 6131 |
+
{
|
| 6132 |
+
"epoch": 0.037,
|
| 6133 |
+
"grad_norm": 0.9958565831184387,
|
| 6134 |
+
"learning_rate": 4.732900665341824e-05,
|
| 6135 |
+
"loss": 1.2112,
|
| 6136 |
+
"step": 78500
|
| 6137 |
+
},
|
| 6138 |
+
{
|
| 6139 |
+
"epoch": 0.0372,
|
| 6140 |
+
"grad_norm": 0.6769017577171326,
|
| 6141 |
+
"learning_rate": 4.732186632229311e-05,
|
| 6142 |
+
"loss": 1.1933,
|
| 6143 |
+
"step": 78600
|
| 6144 |
+
},
|
| 6145 |
+
{
|
| 6146 |
+
"epoch": 0.0374,
|
| 6147 |
+
"grad_norm": 0.6430754661560059,
|
| 6148 |
+
"learning_rate": 4.7314716999922896e-05,
|
| 6149 |
+
"loss": 1.1851,
|
| 6150 |
+
"step": 78700
|
| 6151 |
+
},
|
| 6152 |
+
{
|
| 6153 |
+
"epoch": 0.0376,
|
| 6154 |
+
"grad_norm": 1.103901982307434,
|
| 6155 |
+
"learning_rate": 4.7307558689187334e-05,
|
| 6156 |
+
"loss": 1.2234,
|
| 6157 |
+
"step": 78800
|
| 6158 |
+
},
|
| 6159 |
+
{
|
| 6160 |
+
"epoch": 0.0378,
|
| 6161 |
+
"grad_norm": 1.143268346786499,
|
| 6162 |
+
"learning_rate": 4.73003913929698e-05,
|
| 6163 |
+
"loss": 1.1609,
|
| 6164 |
+
"step": 78900
|
| 6165 |
+
},
|
| 6166 |
+
{
|
| 6167 |
+
"epoch": 0.038,
|
| 6168 |
+
"grad_norm": 1.2543673515319824,
|
| 6169 |
+
"learning_rate": 4.7293215114157284e-05,
|
| 6170 |
+
"loss": 1.1862,
|
| 6171 |
+
"step": 79000
|
| 6172 |
+
},
|
| 6173 |
+
{
|
| 6174 |
+
"epoch": 0.038,
|
| 6175 |
+
"eval_loss": 1.075058937072754,
|
| 6176 |
+
"eval_runtime": 77.0151,
|
| 6177 |
+
"eval_samples_per_second": 200.519,
|
| 6178 |
+
"eval_steps_per_second": 3.142,
|
| 6179 |
+
"step": 79000
|
| 6180 |
+
},
|
| 6181 |
+
{
|
| 6182 |
+
"epoch": 0.0382,
|
| 6183 |
+
"grad_norm": 1.0687370300292969,
|
| 6184 |
+
"learning_rate": 4.728602985564039e-05,
|
| 6185 |
+
"loss": 1.1878,
|
| 6186 |
+
"step": 79100
|
| 6187 |
+
},
|
| 6188 |
+
{
|
| 6189 |
+
"epoch": 0.0384,
|
| 6190 |
+
"grad_norm": 1.230892539024353,
|
| 6191 |
+
"learning_rate": 4.727883562031333e-05,
|
| 6192 |
+
"loss": 1.1561,
|
| 6193 |
+
"step": 79200
|
| 6194 |
+
},
|
| 6195 |
+
{
|
| 6196 |
+
"epoch": 0.0386,
|
| 6197 |
+
"grad_norm": 1.0465742349624634,
|
| 6198 |
+
"learning_rate": 4.727163241107395e-05,
|
| 6199 |
+
"loss": 1.1677,
|
| 6200 |
+
"step": 79300
|
| 6201 |
+
},
|
| 6202 |
+
{
|
| 6203 |
+
"epoch": 0.0388,
|
| 6204 |
+
"grad_norm": 0.6553373336791992,
|
| 6205 |
+
"learning_rate": 4.726442023082369e-05,
|
| 6206 |
+
"loss": 1.2035,
|
| 6207 |
+
"step": 79400
|
| 6208 |
+
},
|
| 6209 |
+
{
|
| 6210 |
+
"epoch": 0.039,
|
| 6211 |
+
"grad_norm": 0.9347487688064575,
|
| 6212 |
+
"learning_rate": 4.725719908246763e-05,
|
| 6213 |
+
"loss": 1.2116,
|
| 6214 |
+
"step": 79500
|
| 6215 |
+
},
|
| 6216 |
+
{
|
| 6217 |
+
"epoch": 0.0392,
|
| 6218 |
+
"grad_norm": 1.0414602756500244,
|
| 6219 |
+
"learning_rate": 4.724996896891445e-05,
|
| 6220 |
+
"loss": 1.2237,
|
| 6221 |
+
"step": 79600
|
| 6222 |
+
},
|
| 6223 |
+
{
|
| 6224 |
+
"epoch": 0.0394,
|
| 6225 |
+
"grad_norm": 1.1857577562332153,
|
| 6226 |
+
"learning_rate": 4.724272989307642e-05,
|
| 6227 |
+
"loss": 1.1653,
|
| 6228 |
+
"step": 79700
|
| 6229 |
+
},
|
| 6230 |
+
{
|
| 6231 |
+
"epoch": 0.0396,
|
| 6232 |
+
"grad_norm": 1.3574703931808472,
|
| 6233 |
+
"learning_rate": 4.7235481857869446e-05,
|
| 6234 |
+
"loss": 1.2176,
|
| 6235 |
+
"step": 79800
|
| 6236 |
+
},
|
| 6237 |
+
{
|
| 6238 |
+
"epoch": 0.0398,
|
| 6239 |
+
"grad_norm": 1.3188464641571045,
|
| 6240 |
+
"learning_rate": 4.722822486621304e-05,
|
| 6241 |
+
"loss": 1.1872,
|
| 6242 |
+
"step": 79900
|
| 6243 |
+
},
|
| 6244 |
+
{
|
| 6245 |
+
"epoch": 0.04,
|
| 6246 |
+
"grad_norm": 1.1241661310195923,
|
| 6247 |
+
"learning_rate": 4.722095892103032e-05,
|
| 6248 |
+
"loss": 1.1926,
|
| 6249 |
+
"step": 80000
|
| 6250 |
+
},
|
| 6251 |
+
{
|
| 6252 |
+
"epoch": 0.04,
|
| 6253 |
+
"eval_loss": 1.0716365575790405,
|
| 6254 |
+
"eval_runtime": 76.5906,
|
| 6255 |
+
"eval_samples_per_second": 201.63,
|
| 6256 |
+
"eval_steps_per_second": 3.16,
|
| 6257 |
+
"step": 80000
|
| 6258 |
+
},
|
| 6259 |
+
{
|
| 6260 |
+
"epoch": 0.0402,
|
| 6261 |
+
"grad_norm": 0.9855309724807739,
|
| 6262 |
+
"learning_rate": 4.721368402524801e-05,
|
| 6263 |
+
"loss": 1.1427,
|
| 6264 |
+
"step": 80100
|
| 6265 |
+
},
|
| 6266 |
+
{
|
| 6267 |
+
"epoch": 0.0404,
|
| 6268 |
+
"grad_norm": 0.6458451151847839,
|
| 6269 |
+
"learning_rate": 4.720640018179642e-05,
|
| 6270 |
+
"loss": 1.2032,
|
| 6271 |
+
"step": 80200
|
| 6272 |
+
},
|
| 6273 |
+
{
|
| 6274 |
+
"epoch": 0.0406,
|
| 6275 |
+
"grad_norm": 1.1878600120544434,
|
| 6276 |
+
"learning_rate": 4.71991073936095e-05,
|
| 6277 |
+
"loss": 1.1879,
|
| 6278 |
+
"step": 80300
|
| 6279 |
+
},
|
| 6280 |
+
{
|
| 6281 |
+
"epoch": 0.0408,
|
| 6282 |
+
"grad_norm": 0.8349748253822327,
|
| 6283 |
+
"learning_rate": 4.719180566362477e-05,
|
| 6284 |
+
"loss": 1.1355,
|
| 6285 |
+
"step": 80400
|
| 6286 |
+
},
|
| 6287 |
+
{
|
| 6288 |
+
"epoch": 0.041,
|
| 6289 |
+
"grad_norm": 1.1937662363052368,
|
| 6290 |
+
"learning_rate": 4.7184494994783376e-05,
|
| 6291 |
+
"loss": 1.2018,
|
| 6292 |
+
"step": 80500
|
| 6293 |
+
},
|
| 6294 |
+
{
|
| 6295 |
+
"epoch": 0.0412,
|
| 6296 |
+
"grad_norm": 1.3011997938156128,
|
| 6297 |
+
"learning_rate": 4.7177175390030054e-05,
|
| 6298 |
+
"loss": 1.1697,
|
| 6299 |
+
"step": 80600
|
| 6300 |
+
},
|
| 6301 |
+
{
|
| 6302 |
+
"epoch": 0.0414,
|
| 6303 |
+
"grad_norm": 1.1909871101379395,
|
| 6304 |
+
"learning_rate": 4.7169846852313137e-05,
|
| 6305 |
+
"loss": 1.2126,
|
| 6306 |
+
"step": 80700
|
| 6307 |
+
},
|
| 6308 |
+
{
|
| 6309 |
+
"epoch": 0.0416,
|
| 6310 |
+
"grad_norm": 1.5078299045562744,
|
| 6311 |
+
"learning_rate": 4.7162509384584555e-05,
|
| 6312 |
+
"loss": 1.1983,
|
| 6313 |
+
"step": 80800
|
| 6314 |
+
},
|
| 6315 |
+
{
|
| 6316 |
+
"epoch": 0.0418,
|
| 6317 |
+
"grad_norm": 1.3141160011291504,
|
| 6318 |
+
"learning_rate": 4.715516298979984e-05,
|
| 6319 |
+
"loss": 1.2118,
|
| 6320 |
+
"step": 80900
|
| 6321 |
+
},
|
| 6322 |
+
{
|
| 6323 |
+
"epoch": 0.042,
|
| 6324 |
+
"grad_norm": 1.3565207719802856,
|
| 6325 |
+
"learning_rate": 4.714780767091813e-05,
|
| 6326 |
+
"loss": 1.2054,
|
| 6327 |
+
"step": 81000
|
| 6328 |
+
},
|
| 6329 |
+
{
|
| 6330 |
+
"epoch": 0.042,
|
| 6331 |
+
"eval_loss": 1.0669591426849365,
|
| 6332 |
+
"eval_runtime": 75.959,
|
| 6333 |
+
"eval_samples_per_second": 203.307,
|
| 6334 |
+
"eval_steps_per_second": 3.186,
|
| 6335 |
+
"step": 81000
|
| 6336 |
+
},
|
| 6337 |
+
{
|
| 6338 |
+
"epoch": 0.0422,
|
| 6339 |
+
"grad_norm": 1.3890715837478638,
|
| 6340 |
+
"learning_rate": 4.714044343090214e-05,
|
| 6341 |
+
"loss": 1.1917,
|
| 6342 |
+
"step": 81100
|
| 6343 |
+
},
|
| 6344 |
+
{
|
| 6345 |
+
"epoch": 0.0424,
|
| 6346 |
+
"grad_norm": 0.9992968440055847,
|
| 6347 |
+
"learning_rate": 4.713307027271817e-05,
|
| 6348 |
+
"loss": 1.1869,
|
| 6349 |
+
"step": 81200
|
| 6350 |
+
},
|
| 6351 |
+
{
|
| 6352 |
+
"epoch": 0.0426,
|
| 6353 |
+
"grad_norm": 0.8716880679130554,
|
| 6354 |
+
"learning_rate": 4.712568819933615e-05,
|
| 6355 |
+
"loss": 1.1941,
|
| 6356 |
+
"step": 81300
|
| 6357 |
+
},
|
| 6358 |
+
{
|
| 6359 |
+
"epoch": 0.0428,
|
| 6360 |
+
"grad_norm": 1.243594765663147,
|
| 6361 |
+
"learning_rate": 4.711829721372957e-05,
|
| 6362 |
+
"loss": 1.1667,
|
| 6363 |
+
"step": 81400
|
| 6364 |
+
},
|
| 6365 |
+
{
|
| 6366 |
+
"epoch": 0.043,
|
| 6367 |
+
"grad_norm": 0.6567416191101074,
|
| 6368 |
+
"learning_rate": 4.7110897318875516e-05,
|
| 6369 |
+
"loss": 1.2105,
|
| 6370 |
+
"step": 81500
|
| 6371 |
+
},
|
| 6372 |
+
{
|
| 6373 |
+
"epoch": 0.0432,
|
| 6374 |
+
"grad_norm": 0.5886017084121704,
|
| 6375 |
+
"learning_rate": 4.710348851775467e-05,
|
| 6376 |
+
"loss": 1.1867,
|
| 6377 |
+
"step": 81600
|
| 6378 |
+
},
|
| 6379 |
+
{
|
| 6380 |
+
"epoch": 0.0434,
|
| 6381 |
+
"grad_norm": 0.6296970248222351,
|
| 6382 |
+
"learning_rate": 4.709607081335129e-05,
|
| 6383 |
+
"loss": 1.1702,
|
| 6384 |
+
"step": 81700
|
| 6385 |
+
},
|
| 6386 |
+
{
|
| 6387 |
+
"epoch": 0.0436,
|
| 6388 |
+
"grad_norm": 0.9896938800811768,
|
| 6389 |
+
"learning_rate": 4.7088644208653226e-05,
|
| 6390 |
+
"loss": 1.1628,
|
| 6391 |
+
"step": 81800
|
| 6392 |
+
},
|
| 6393 |
+
{
|
| 6394 |
+
"epoch": 0.0438,
|
| 6395 |
+
"grad_norm": 0.7199723720550537,
|
| 6396 |
+
"learning_rate": 4.708120870665192e-05,
|
| 6397 |
+
"loss": 1.1792,
|
| 6398 |
+
"step": 81900
|
| 6399 |
+
},
|
| 6400 |
+
{
|
| 6401 |
+
"epoch": 0.044,
|
| 6402 |
+
"grad_norm": 1.3148512840270996,
|
| 6403 |
+
"learning_rate": 4.707376431034238e-05,
|
| 6404 |
+
"loss": 1.185,
|
| 6405 |
+
"step": 82000
|
| 6406 |
+
},
|
| 6407 |
+
{
|
| 6408 |
+
"epoch": 0.044,
|
| 6409 |
+
"eval_loss": 1.0709099769592285,
|
| 6410 |
+
"eval_runtime": 75.8635,
|
| 6411 |
+
"eval_samples_per_second": 203.563,
|
| 6412 |
+
"eval_steps_per_second": 3.19,
|
| 6413 |
+
"step": 82000
|
| 6414 |
+
},
|
| 6415 |
+
{
|
| 6416 |
+
"epoch": 0.0442,
|
| 6417 |
+
"grad_norm": 0.6634069681167603,
|
| 6418 |
+
"learning_rate": 4.706631102272323e-05,
|
| 6419 |
+
"loss": 1.1633,
|
| 6420 |
+
"step": 82100
|
| 6421 |
+
},
|
| 6422 |
+
{
|
| 6423 |
+
"epoch": 0.0444,
|
| 6424 |
+
"grad_norm": 1.3700015544891357,
|
| 6425 |
+
"learning_rate": 4.705884884679663e-05,
|
| 6426 |
+
"loss": 1.1712,
|
| 6427 |
+
"step": 82200
|
| 6428 |
+
},
|
| 6429 |
+
{
|
| 6430 |
+
"epoch": 0.0446,
|
| 6431 |
+
"grad_norm": 1.1697111129760742,
|
| 6432 |
+
"learning_rate": 4.705137778556835e-05,
|
| 6433 |
+
"loss": 1.1902,
|
| 6434 |
+
"step": 82300
|
| 6435 |
+
},
|
| 6436 |
+
{
|
| 6437 |
+
"epoch": 0.0448,
|
| 6438 |
+
"grad_norm": 1.4012552499771118,
|
| 6439 |
+
"learning_rate": 4.7043897842047735e-05,
|
| 6440 |
+
"loss": 1.216,
|
| 6441 |
+
"step": 82400
|
| 6442 |
+
},
|
| 6443 |
+
{
|
| 6444 |
+
"epoch": 0.045,
|
| 6445 |
+
"grad_norm": 1.2128801345825195,
|
| 6446 |
+
"learning_rate": 4.7036409019247706e-05,
|
| 6447 |
+
"loss": 1.2169,
|
| 6448 |
+
"step": 82500
|
| 6449 |
+
},
|
| 6450 |
+
{
|
| 6451 |
+
"epoch": 0.0452,
|
| 6452 |
+
"grad_norm": 1.435831904411316,
|
| 6453 |
+
"learning_rate": 4.7028911320184766e-05,
|
| 6454 |
+
"loss": 1.1839,
|
| 6455 |
+
"step": 82600
|
| 6456 |
+
},
|
| 6457 |
+
{
|
| 6458 |
+
"epoch": 0.0454,
|
| 6459 |
+
"grad_norm": 0.8126788139343262,
|
| 6460 |
+
"learning_rate": 4.702140474787898e-05,
|
| 6461 |
+
"loss": 1.1652,
|
| 6462 |
+
"step": 82700
|
| 6463 |
+
},
|
| 6464 |
+
{
|
| 6465 |
+
"epoch": 0.0456,
|
| 6466 |
+
"grad_norm": 1.1946730613708496,
|
| 6467 |
+
"learning_rate": 4.7013889305353985e-05,
|
| 6468 |
+
"loss": 1.2277,
|
| 6469 |
+
"step": 82800
|
| 6470 |
+
},
|
| 6471 |
+
{
|
| 6472 |
+
"epoch": 0.0458,
|
| 6473 |
+
"grad_norm": 0.6007882952690125,
|
| 6474 |
+
"learning_rate": 4.700636499563702e-05,
|
| 6475 |
+
"loss": 1.1586,
|
| 6476 |
+
"step": 82900
|
| 6477 |
+
},
|
| 6478 |
+
{
|
| 6479 |
+
"epoch": 0.046,
|
| 6480 |
+
"grad_norm": 0.6656979322433472,
|
| 6481 |
+
"learning_rate": 4.699883182175886e-05,
|
| 6482 |
+
"loss": 1.1902,
|
| 6483 |
+
"step": 83000
|
| 6484 |
+
},
|
| 6485 |
+
{
|
| 6486 |
+
"epoch": 0.046,
|
| 6487 |
+
"eval_loss": 1.072899580001831,
|
| 6488 |
+
"eval_runtime": 77.2342,
|
| 6489 |
+
"eval_samples_per_second": 199.95,
|
| 6490 |
+
"eval_steps_per_second": 3.133,
|
| 6491 |
+
"step": 83000
|
| 6492 |
+
},
|
| 6493 |
+
{
|
| 6494 |
+
"epoch": 0.0462,
|
| 6495 |
+
"grad_norm": 1.5463351011276245,
|
| 6496 |
+
"learning_rate": 4.6991289786753876e-05,
|
| 6497 |
+
"loss": 1.1988,
|
| 6498 |
+
"step": 83100
|
| 6499 |
+
},
|
| 6500 |
+
{
|
| 6501 |
+
"epoch": 0.0464,
|
| 6502 |
+
"grad_norm": 1.202536940574646,
|
| 6503 |
+
"learning_rate": 4.698373889366e-05,
|
| 6504 |
+
"loss": 1.1983,
|
| 6505 |
+
"step": 83200
|
| 6506 |
+
},
|
| 6507 |
+
{
|
| 6508 |
+
"epoch": 0.0466,
|
| 6509 |
+
"grad_norm": 0.7186087369918823,
|
| 6510 |
+
"learning_rate": 4.6976179145518724e-05,
|
| 6511 |
+
"loss": 1.15,
|
| 6512 |
+
"step": 83300
|
| 6513 |
+
},
|
| 6514 |
+
{
|
| 6515 |
+
"epoch": 0.0468,
|
| 6516 |
+
"grad_norm": 1.3059759140014648,
|
| 6517 |
+
"learning_rate": 4.6968610545375116e-05,
|
| 6518 |
+
"loss": 1.1896,
|
| 6519 |
+
"step": 83400
|
| 6520 |
+
},
|
| 6521 |
+
{
|
| 6522 |
+
"epoch": 0.047,
|
| 6523 |
+
"grad_norm": 0.8425590991973877,
|
| 6524 |
+
"learning_rate": 4.696103309627781e-05,
|
| 6525 |
+
"loss": 1.1747,
|
| 6526 |
+
"step": 83500
|
| 6527 |
+
},
|
| 6528 |
+
{
|
| 6529 |
+
"epoch": 0.0472,
|
| 6530 |
+
"grad_norm": 1.1745330095291138,
|
| 6531 |
+
"learning_rate": 4.695344680127899e-05,
|
| 6532 |
+
"loss": 1.1591,
|
| 6533 |
+
"step": 83600
|
| 6534 |
+
},
|
| 6535 |
+
{
|
| 6536 |
+
"epoch": 0.0474,
|
| 6537 |
+
"grad_norm": 0.6429449915885925,
|
| 6538 |
+
"learning_rate": 4.694585166343443e-05,
|
| 6539 |
+
"loss": 1.1893,
|
| 6540 |
+
"step": 83700
|
| 6541 |
+
},
|
| 6542 |
+
{
|
| 6543 |
+
"epoch": 0.0476,
|
| 6544 |
+
"grad_norm": 1.5323892831802368,
|
| 6545 |
+
"learning_rate": 4.693824768580344e-05,
|
| 6546 |
+
"loss": 1.2037,
|
| 6547 |
+
"step": 83800
|
| 6548 |
+
},
|
| 6549 |
+
{
|
| 6550 |
+
"epoch": 0.0478,
|
| 6551 |
+
"grad_norm": 1.2719688415527344,
|
| 6552 |
+
"learning_rate": 4.693063487144891e-05,
|
| 6553 |
+
"loss": 1.191,
|
| 6554 |
+
"step": 83900
|
| 6555 |
+
},
|
| 6556 |
+
{
|
| 6557 |
+
"epoch": 0.048,
|
| 6558 |
+
"grad_norm": 1.1735507249832153,
|
| 6559 |
+
"learning_rate": 4.6923013223437276e-05,
|
| 6560 |
+
"loss": 1.1904,
|
| 6561 |
+
"step": 84000
|
| 6562 |
+
},
|
| 6563 |
+
{
|
| 6564 |
+
"epoch": 0.048,
|
| 6565 |
+
"eval_loss": 1.0721956491470337,
|
| 6566 |
+
"eval_runtime": 76.3531,
|
| 6567 |
+
"eval_samples_per_second": 202.258,
|
| 6568 |
+
"eval_steps_per_second": 3.169,
|
| 6569 |
+
"step": 84000
|
| 6570 |
+
},
|
| 6571 |
+
{
|
| 6572 |
+
"epoch": 0.0482,
|
| 6573 |
+
"grad_norm": 1.1949045658111572,
|
| 6574 |
+
"learning_rate": 4.6915382744838536e-05,
|
| 6575 |
+
"loss": 1.1507,
|
| 6576 |
+
"step": 84100
|
| 6577 |
+
},
|
| 6578 |
+
{
|
| 6579 |
+
"epoch": 0.0484,
|
| 6580 |
+
"grad_norm": 1.074385404586792,
|
| 6581 |
+
"learning_rate": 4.690774343872625e-05,
|
| 6582 |
+
"loss": 1.1504,
|
| 6583 |
+
"step": 84200
|
| 6584 |
+
},
|
| 6585 |
+
{
|
| 6586 |
+
"epoch": 0.0486,
|
| 6587 |
+
"grad_norm": 1.0720473527908325,
|
| 6588 |
+
"learning_rate": 4.690009530817753e-05,
|
| 6589 |
+
"loss": 1.1758,
|
| 6590 |
+
"step": 84300
|
| 6591 |
+
},
|
| 6592 |
+
{
|
| 6593 |
+
"epoch": 0.0488,
|
| 6594 |
+
"grad_norm": 1.0596733093261719,
|
| 6595 |
+
"learning_rate": 4.6892438356273024e-05,
|
| 6596 |
+
"loss": 1.1778,
|
| 6597 |
+
"step": 84400
|
| 6598 |
+
},
|
| 6599 |
+
{
|
| 6600 |
+
"epoch": 0.049,
|
| 6601 |
+
"grad_norm": 1.2753647565841675,
|
| 6602 |
+
"learning_rate": 4.688477258609698e-05,
|
| 6603 |
+
"loss": 1.1827,
|
| 6604 |
+
"step": 84500
|
| 6605 |
+
},
|
| 6606 |
+
{
|
| 6607 |
+
"epoch": 0.0492,
|
| 6608 |
+
"grad_norm": 1.2803727388381958,
|
| 6609 |
+
"learning_rate": 4.687709800073715e-05,
|
| 6610 |
+
"loss": 1.164,
|
| 6611 |
+
"step": 84600
|
| 6612 |
+
},
|
| 6613 |
+
{
|
| 6614 |
+
"epoch": 0.0494,
|
| 6615 |
+
"grad_norm": 1.4797301292419434,
|
| 6616 |
+
"learning_rate": 4.6869414603284865e-05,
|
| 6617 |
+
"loss": 1.1748,
|
| 6618 |
+
"step": 84700
|
| 6619 |
+
},
|
| 6620 |
+
{
|
| 6621 |
+
"epoch": 0.0496,
|
| 6622 |
+
"grad_norm": 1.1455540657043457,
|
| 6623 |
+
"learning_rate": 4.6861722396834996e-05,
|
| 6624 |
+
"loss": 1.1918,
|
| 6625 |
+
"step": 84800
|
| 6626 |
+
},
|
| 6627 |
+
{
|
| 6628 |
+
"epoch": 0.0498,
|
| 6629 |
+
"grad_norm": 1.1636658906936646,
|
| 6630 |
+
"learning_rate": 4.6854021384485954e-05,
|
| 6631 |
+
"loss": 1.208,
|
| 6632 |
+
"step": 84900
|
| 6633 |
+
},
|
| 6634 |
+
{
|
| 6635 |
+
"epoch": 0.05,
|
| 6636 |
+
"grad_norm": 1.267817735671997,
|
| 6637 |
+
"learning_rate": 4.684631156933971e-05,
|
| 6638 |
+
"loss": 1.1679,
|
| 6639 |
+
"step": 85000
|
| 6640 |
+
},
|
| 6641 |
+
{
|
| 6642 |
+
"epoch": 0.05,
|
| 6643 |
+
"eval_loss": 1.0709259510040283,
|
| 6644 |
+
"eval_runtime": 76.3729,
|
| 6645 |
+
"eval_samples_per_second": 202.205,
|
| 6646 |
+
"eval_steps_per_second": 3.169,
|
| 6647 |
+
"step": 85000
|
| 6648 |
+
},
|
| 6649 |
+
{
|
| 6650 |
+
"epoch": 0.0502,
|
| 6651 |
+
"grad_norm": 1.5029271841049194,
|
| 6652 |
+
"learning_rate": 4.683859295450178e-05,
|
| 6653 |
+
"loss": 1.1459,
|
| 6654 |
+
"step": 85100
|
| 6655 |
+
},
|
| 6656 |
+
{
|
| 6657 |
+
"epoch": 0.0504,
|
| 6658 |
+
"grad_norm": 0.7328454256057739,
|
| 6659 |
+
"learning_rate": 4.683086554308123e-05,
|
| 6660 |
+
"loss": 1.1861,
|
| 6661 |
+
"step": 85200
|
| 6662 |
+
},
|
| 6663 |
+
{
|
| 6664 |
+
"epoch": 0.0506,
|
| 6665 |
+
"grad_norm": 1.114625334739685,
|
| 6666 |
+
"learning_rate": 4.682312933819063e-05,
|
| 6667 |
+
"loss": 1.1609,
|
| 6668 |
+
"step": 85300
|
| 6669 |
+
},
|
| 6670 |
+
{
|
| 6671 |
+
"epoch": 0.0508,
|
| 6672 |
+
"grad_norm": 1.4052484035491943,
|
| 6673 |
+
"learning_rate": 4.681538434294615e-05,
|
| 6674 |
+
"loss": 1.1534,
|
| 6675 |
+
"step": 85400
|
| 6676 |
+
},
|
| 6677 |
+
{
|
| 6678 |
+
"epoch": 0.051,
|
| 6679 |
+
"grad_norm": 0.7364799976348877,
|
| 6680 |
+
"learning_rate": 4.6807630560467475e-05,
|
| 6681 |
+
"loss": 1.1973,
|
| 6682 |
+
"step": 85500
|
| 6683 |
+
},
|
| 6684 |
+
{
|
| 6685 |
+
"epoch": 0.0512,
|
| 6686 |
+
"grad_norm": 0.701787531375885,
|
| 6687 |
+
"learning_rate": 4.679986799387781e-05,
|
| 6688 |
+
"loss": 1.1743,
|
| 6689 |
+
"step": 85600
|
| 6690 |
+
},
|
| 6691 |
+
{
|
| 6692 |
+
"epoch": 0.0514,
|
| 6693 |
+
"grad_norm": 1.331763744354248,
|
| 6694 |
+
"learning_rate": 4.679209664630393e-05,
|
| 6695 |
+
"loss": 1.1516,
|
| 6696 |
+
"step": 85700
|
| 6697 |
+
},
|
| 6698 |
+
{
|
| 6699 |
+
"epoch": 0.0516,
|
| 6700 |
+
"grad_norm": 0.9733197689056396,
|
| 6701 |
+
"learning_rate": 4.6784316520876124e-05,
|
| 6702 |
+
"loss": 1.1646,
|
| 6703 |
+
"step": 85800
|
| 6704 |
+
},
|
| 6705 |
+
{
|
| 6706 |
+
"epoch": 0.0518,
|
| 6707 |
+
"grad_norm": 0.7415294051170349,
|
| 6708 |
+
"learning_rate": 4.677652762072823e-05,
|
| 6709 |
+
"loss": 1.2005,
|
| 6710 |
+
"step": 85900
|
| 6711 |
+
},
|
| 6712 |
+
{
|
| 6713 |
+
"epoch": 0.052,
|
| 6714 |
+
"grad_norm": 1.1027395725250244,
|
| 6715 |
+
"learning_rate": 4.6768729948997606e-05,
|
| 6716 |
+
"loss": 1.1601,
|
| 6717 |
+
"step": 86000
|
| 6718 |
+
},
|
| 6719 |
+
{
|
| 6720 |
+
"epoch": 0.052,
|
| 6721 |
+
"eval_loss": 1.0681675672531128,
|
| 6722 |
+
"eval_runtime": 76.2441,
|
| 6723 |
+
"eval_samples_per_second": 202.547,
|
| 6724 |
+
"eval_steps_per_second": 3.174,
|
| 6725 |
+
"step": 86000
|
| 6726 |
+
},
|
| 6727 |
+
{
|
| 6728 |
+
"epoch": 0.0522,
|
| 6729 |
+
"grad_norm": 0.7156331539154053,
|
| 6730 |
+
"learning_rate": 4.676092350882517e-05,
|
| 6731 |
+
"loss": 1.1854,
|
| 6732 |
+
"step": 86100
|
| 6733 |
+
},
|
| 6734 |
+
{
|
| 6735 |
+
"epoch": 0.0524,
|
| 6736 |
+
"grad_norm": 1.3423713445663452,
|
| 6737 |
+
"learning_rate": 4.675310830335534e-05,
|
| 6738 |
+
"loss": 1.2135,
|
| 6739 |
+
"step": 86200
|
| 6740 |
+
},
|
| 6741 |
+
{
|
| 6742 |
+
"epoch": 0.0526,
|
| 6743 |
+
"grad_norm": 1.1925442218780518,
|
| 6744 |
+
"learning_rate": 4.6745284335736095e-05,
|
| 6745 |
+
"loss": 1.1414,
|
| 6746 |
+
"step": 86300
|
| 6747 |
+
},
|
| 6748 |
+
{
|
| 6749 |
+
"epoch": 0.0528,
|
| 6750 |
+
"grad_norm": 1.1717417240142822,
|
| 6751 |
+
"learning_rate": 4.673745160911891e-05,
|
| 6752 |
+
"loss": 1.184,
|
| 6753 |
+
"step": 86400
|
| 6754 |
+
},
|
| 6755 |
+
{
|
| 6756 |
+
"epoch": 0.053,
|
| 6757 |
+
"grad_norm": 1.0722715854644775,
|
| 6758 |
+
"learning_rate": 4.672961012665883e-05,
|
| 6759 |
+
"loss": 1.1685,
|
| 6760 |
+
"step": 86500
|
| 6761 |
+
},
|
| 6762 |
+
{
|
| 6763 |
+
"epoch": 0.0532,
|
| 6764 |
+
"grad_norm": 1.293058156967163,
|
| 6765 |
+
"learning_rate": 4.6721759891514386e-05,
|
| 6766 |
+
"loss": 1.1639,
|
| 6767 |
+
"step": 86600
|
| 6768 |
+
},
|
| 6769 |
+
{
|
| 6770 |
+
"epoch": 0.0534,
|
| 6771 |
+
"grad_norm": 1.1121761798858643,
|
| 6772 |
+
"learning_rate": 4.671390090684765e-05,
|
| 6773 |
+
"loss": 1.1693,
|
| 6774 |
+
"step": 86700
|
| 6775 |
+
},
|
| 6776 |
+
{
|
| 6777 |
+
"epoch": 0.0536,
|
| 6778 |
+
"grad_norm": 1.1979039907455444,
|
| 6779 |
+
"learning_rate": 4.6706033175824226e-05,
|
| 6780 |
+
"loss": 1.2123,
|
| 6781 |
+
"step": 86800
|
| 6782 |
+
},
|
| 6783 |
+
{
|
| 6784 |
+
"epoch": 0.0538,
|
| 6785 |
+
"grad_norm": 1.3003602027893066,
|
| 6786 |
+
"learning_rate": 4.669815670161324e-05,
|
| 6787 |
+
"loss": 1.1529,
|
| 6788 |
+
"step": 86900
|
| 6789 |
+
},
|
| 6790 |
+
{
|
| 6791 |
+
"epoch": 0.054,
|
| 6792 |
+
"grad_norm": 0.627068817615509,
|
| 6793 |
+
"learning_rate": 4.669027148738732e-05,
|
| 6794 |
+
"loss": 1.1901,
|
| 6795 |
+
"step": 87000
|
| 6796 |
+
},
|
| 6797 |
+
{
|
| 6798 |
+
"epoch": 0.054,
|
| 6799 |
+
"eval_loss": 1.0730499029159546,
|
| 6800 |
+
"eval_runtime": 76.271,
|
| 6801 |
+
"eval_samples_per_second": 202.475,
|
| 6802 |
+
"eval_steps_per_second": 3.173,
|
| 6803 |
+
"step": 87000
|
| 6804 |
+
},
|
| 6805 |
+
{
|
| 6806 |
+
"epoch": 0.0542,
|
| 6807 |
+
"grad_norm": 1.0153006315231323,
|
| 6808 |
+
"learning_rate": 4.6682377536322644e-05,
|
| 6809 |
+
"loss": 1.1295,
|
| 6810 |
+
"step": 87100
|
| 6811 |
+
},
|
| 6812 |
+
{
|
| 6813 |
+
"epoch": 0.0544,
|
| 6814 |
+
"grad_norm": 1.3619033098220825,
|
| 6815 |
+
"learning_rate": 4.667447485159889e-05,
|
| 6816 |
+
"loss": 1.1759,
|
| 6817 |
+
"step": 87200
|
| 6818 |
+
},
|
| 6819 |
+
{
|
| 6820 |
+
"epoch": 0.0546,
|
| 6821 |
+
"grad_norm": 0.8665468692779541,
|
| 6822 |
+
"learning_rate": 4.666656343639926e-05,
|
| 6823 |
+
"loss": 1.1602,
|
| 6824 |
+
"step": 87300
|
| 6825 |
+
},
|
| 6826 |
+
{
|
| 6827 |
+
"epoch": 0.0548,
|
| 6828 |
+
"grad_norm": 0.7338219285011292,
|
| 6829 |
+
"learning_rate": 4.665864329391046e-05,
|
| 6830 |
+
"loss": 1.1856,
|
| 6831 |
+
"step": 87400
|
| 6832 |
+
},
|
| 6833 |
+
{
|
| 6834 |
+
"epoch": 0.055,
|
| 6835 |
+
"grad_norm": 0.7363407611846924,
|
| 6836 |
+
"learning_rate": 4.665071442732274e-05,
|
| 6837 |
+
"loss": 1.1629,
|
| 6838 |
+
"step": 87500
|
| 6839 |
+
},
|
| 6840 |
+
{
|
| 6841 |
+
"epoch": 0.0552,
|
| 6842 |
+
"grad_norm": 0.9836055636405945,
|
| 6843 |
+
"learning_rate": 4.664277683982984e-05,
|
| 6844 |
+
"loss": 1.1755,
|
| 6845 |
+
"step": 87600
|
| 6846 |
+
},
|
| 6847 |
+
{
|
| 6848 |
+
"epoch": 0.0554,
|
| 6849 |
+
"grad_norm": 1.0638995170593262,
|
| 6850 |
+
"learning_rate": 4.663483053462901e-05,
|
| 6851 |
+
"loss": 1.186,
|
| 6852 |
+
"step": 87700
|
| 6853 |
+
},
|
| 6854 |
+
{
|
| 6855 |
+
"epoch": 0.0556,
|
| 6856 |
+
"grad_norm": 0.9050219058990479,
|
| 6857 |
+
"learning_rate": 4.662687551492103e-05,
|
| 6858 |
+
"loss": 1.2357,
|
| 6859 |
+
"step": 87800
|
| 6860 |
+
},
|
| 6861 |
+
{
|
| 6862 |
+
"epoch": 0.0558,
|
| 6863 |
+
"grad_norm": 0.917178213596344,
|
| 6864 |
+
"learning_rate": 4.661891178391018e-05,
|
| 6865 |
+
"loss": 1.1573,
|
| 6866 |
+
"step": 87900
|
| 6867 |
+
},
|
| 6868 |
+
{
|
| 6869 |
+
"epoch": 0.056,
|
| 6870 |
+
"grad_norm": 1.2023630142211914,
|
| 6871 |
+
"learning_rate": 4.661093934480425e-05,
|
| 6872 |
+
"loss": 1.1795,
|
| 6873 |
+
"step": 88000
|
| 6874 |
+
},
|
| 6875 |
+
{
|
| 6876 |
+
"epoch": 0.056,
|
| 6877 |
+
"eval_loss": 1.0689297914505005,
|
| 6878 |
+
"eval_runtime": 77.0471,
|
| 6879 |
+
"eval_samples_per_second": 200.436,
|
| 6880 |
+
"eval_steps_per_second": 3.141,
|
| 6881 |
+
"step": 88000
|
| 6882 |
+
},
|
| 6883 |
+
{
|
| 6884 |
+
"epoch": 0.0562,
|
| 6885 |
+
"grad_norm": 1.2633955478668213,
|
| 6886 |
+
"learning_rate": 4.660295820081453e-05,
|
| 6887 |
+
"loss": 1.1501,
|
| 6888 |
+
"step": 88100
|
| 6889 |
+
},
|
| 6890 |
+
{
|
| 6891 |
+
"epoch": 0.0564,
|
| 6892 |
+
"grad_norm": 0.5867215991020203,
|
| 6893 |
+
"learning_rate": 4.6594968355155835e-05,
|
| 6894 |
+
"loss": 1.2096,
|
| 6895 |
+
"step": 88200
|
| 6896 |
+
},
|
| 6897 |
+
{
|
| 6898 |
+
"epoch": 0.0566,
|
| 6899 |
+
"grad_norm": 1.3425019979476929,
|
| 6900 |
+
"learning_rate": 4.658696981104646e-05,
|
| 6901 |
+
"loss": 1.2016,
|
| 6902 |
+
"step": 88300
|
| 6903 |
+
},
|
| 6904 |
+
{
|
| 6905 |
+
"epoch": 0.0568,
|
| 6906 |
+
"grad_norm": 0.8101886510848999,
|
| 6907 |
+
"learning_rate": 4.657896257170825e-05,
|
| 6908 |
+
"loss": 1.1512,
|
| 6909 |
+
"step": 88400
|
| 6910 |
+
},
|
| 6911 |
+
{
|
| 6912 |
+
"epoch": 0.057,
|
| 6913 |
+
"grad_norm": 1.43784761428833,
|
| 6914 |
+
"learning_rate": 4.6570946640366474e-05,
|
| 6915 |
+
"loss": 1.1536,
|
| 6916 |
+
"step": 88500
|
| 6917 |
+
},
|
| 6918 |
+
{
|
| 6919 |
+
"epoch": 0.0572,
|
| 6920 |
+
"grad_norm": 0.766494870185852,
|
| 6921 |
+
"learning_rate": 4.6562922020249984e-05,
|
| 6922 |
+
"loss": 1.1521,
|
| 6923 |
+
"step": 88600
|
| 6924 |
+
},
|
| 6925 |
+
{
|
| 6926 |
+
"epoch": 0.0574,
|
| 6927 |
+
"grad_norm": 1.5485390424728394,
|
| 6928 |
+
"learning_rate": 4.6554888714591076e-05,
|
| 6929 |
+
"loss": 1.176,
|
| 6930 |
+
"step": 88700
|
| 6931 |
+
},
|
| 6932 |
+
{
|
| 6933 |
+
"epoch": 0.0576,
|
| 6934 |
+
"grad_norm": 0.8266467452049255,
|
| 6935 |
+
"learning_rate": 4.654684672662557e-05,
|
| 6936 |
+
"loss": 1.1514,
|
| 6937 |
+
"step": 88800
|
| 6938 |
+
},
|
| 6939 |
+
{
|
| 6940 |
+
"epoch": 0.0578,
|
| 6941 |
+
"grad_norm": 1.2086583375930786,
|
| 6942 |
+
"learning_rate": 4.6538796059592784e-05,
|
| 6943 |
+
"loss": 1.177,
|
| 6944 |
+
"step": 88900
|
| 6945 |
+
},
|
| 6946 |
+
{
|
| 6947 |
+
"epoch": 0.058,
|
| 6948 |
+
"grad_norm": 1.4609780311584473,
|
| 6949 |
+
"learning_rate": 4.6530736716735526e-05,
|
| 6950 |
+
"loss": 1.1447,
|
| 6951 |
+
"step": 89000
|
| 6952 |
+
},
|
| 6953 |
+
{
|
| 6954 |
+
"epoch": 0.058,
|
| 6955 |
+
"eval_loss": 1.0664150714874268,
|
| 6956 |
+
"eval_runtime": 76.3509,
|
| 6957 |
+
"eval_samples_per_second": 202.264,
|
| 6958 |
+
"eval_steps_per_second": 3.17,
|
| 6959 |
+
"step": 89000
|
| 6960 |
+
},
|
| 6961 |
+
{
|
| 6962 |
+
"epoch": 0.0582,
|
| 6963 |
+
"grad_norm": 1.0640435218811035,
|
| 6964 |
+
"learning_rate": 4.652266870130008e-05,
|
| 6965 |
+
"loss": 1.1392,
|
| 6966 |
+
"step": 89100
|
| 6967 |
+
},
|
| 6968 |
+
{
|
| 6969 |
+
"epoch": 0.0584,
|
| 6970 |
+
"grad_norm": 1.3286436796188354,
|
| 6971 |
+
"learning_rate": 4.651459201653626e-05,
|
| 6972 |
+
"loss": 1.222,
|
| 6973 |
+
"step": 89200
|
| 6974 |
+
},
|
| 6975 |
+
{
|
| 6976 |
+
"epoch": 0.0586,
|
| 6977 |
+
"grad_norm": 0.7577000260353088,
|
| 6978 |
+
"learning_rate": 4.650650666569736e-05,
|
| 6979 |
+
"loss": 1.1842,
|
| 6980 |
+
"step": 89300
|
| 6981 |
+
},
|
| 6982 |
+
{
|
| 6983 |
+
"epoch": 0.0588,
|
| 6984 |
+
"grad_norm": 1.0623698234558105,
|
| 6985 |
+
"learning_rate": 4.6498412652040137e-05,
|
| 6986 |
+
"loss": 1.2071,
|
| 6987 |
+
"step": 89400
|
| 6988 |
+
},
|
| 6989 |
+
{
|
| 6990 |
+
"epoch": 0.059,
|
| 6991 |
+
"grad_norm": 0.9597827792167664,
|
| 6992 |
+
"learning_rate": 4.6490309978824866e-05,
|
| 6993 |
+
"loss": 1.1781,
|
| 6994 |
+
"step": 89500
|
| 6995 |
+
},
|
| 6996 |
+
{
|
| 6997 |
+
"epoch": 0.0592,
|
| 6998 |
+
"grad_norm": 1.126639485359192,
|
| 6999 |
+
"learning_rate": 4.6482198649315306e-05,
|
| 7000 |
+
"loss": 1.1897,
|
| 7001 |
+
"step": 89600
|
| 7002 |
+
},
|
| 7003 |
+
{
|
| 7004 |
+
"epoch": 0.0594,
|
| 7005 |
+
"grad_norm": 1.1724388599395752,
|
| 7006 |
+
"learning_rate": 4.64740786667787e-05,
|
| 7007 |
+
"loss": 1.1567,
|
| 7008 |
+
"step": 89700
|
| 7009 |
+
},
|
| 7010 |
+
{
|
| 7011 |
+
"epoch": 0.0596,
|
| 7012 |
+
"grad_norm": 1.14126718044281,
|
| 7013 |
+
"learning_rate": 4.6465950034485776e-05,
|
| 7014 |
+
"loss": 1.1819,
|
| 7015 |
+
"step": 89800
|
| 7016 |
+
},
|
| 7017 |
+
{
|
| 7018 |
+
"epoch": 0.0598,
|
| 7019 |
+
"grad_norm": 0.8016234040260315,
|
| 7020 |
+
"learning_rate": 4.645781275571075e-05,
|
| 7021 |
+
"loss": 1.1906,
|
| 7022 |
+
"step": 89900
|
| 7023 |
+
},
|
| 7024 |
+
{
|
| 7025 |
+
"epoch": 0.06,
|
| 7026 |
+
"grad_norm": 1.3095015287399292,
|
| 7027 |
+
"learning_rate": 4.644966683373131e-05,
|
| 7028 |
+
"loss": 1.1976,
|
| 7029 |
+
"step": 90000
|
| 7030 |
+
},
|
| 7031 |
+
{
|
| 7032 |
+
"epoch": 0.06,
|
| 7033 |
+
"eval_loss": 1.0730445384979248,
|
| 7034 |
+
"eval_runtime": 76.1401,
|
| 7035 |
+
"eval_samples_per_second": 202.823,
|
| 7036 |
+
"eval_steps_per_second": 3.178,
|
| 7037 |
+
"step": 90000
|
| 7038 |
+
},
|
| 7039 |
+
{
|
| 7040 |
+
"epoch": 0.0602,
|
| 7041 |
+
"grad_norm": 0.5794508457183838,
|
| 7042 |
+
"learning_rate": 4.6441512271828626e-05,
|
| 7043 |
+
"loss": 1.1478,
|
| 7044 |
+
"step": 90100
|
| 7045 |
+
},
|
| 7046 |
+
{
|
| 7047 |
+
"epoch": 0.0604,
|
| 7048 |
+
"grad_norm": 0.9965047240257263,
|
| 7049 |
+
"learning_rate": 4.6433349073287366e-05,
|
| 7050 |
+
"loss": 1.201,
|
| 7051 |
+
"step": 90200
|
| 7052 |
+
},
|
| 7053 |
+
{
|
| 7054 |
+
"epoch": 0.0606,
|
| 7055 |
+
"grad_norm": 1.280166506767273,
|
| 7056 |
+
"learning_rate": 4.642517724139567e-05,
|
| 7057 |
+
"loss": 1.1542,
|
| 7058 |
+
"step": 90300
|
| 7059 |
+
},
|
| 7060 |
+
{
|
| 7061 |
+
"epoch": 0.0608,
|
| 7062 |
+
"grad_norm": 0.7828945517539978,
|
| 7063 |
+
"learning_rate": 4.641699677944514e-05,
|
| 7064 |
+
"loss": 1.186,
|
| 7065 |
+
"step": 90400
|
| 7066 |
+
},
|
| 7067 |
+
{
|
| 7068 |
+
"epoch": 0.061,
|
| 7069 |
+
"grad_norm": 1.096155047416687,
|
| 7070 |
+
"learning_rate": 4.640880769073087e-05,
|
| 7071 |
+
"loss": 1.1969,
|
| 7072 |
+
"step": 90500
|
| 7073 |
+
},
|
| 7074 |
+
{
|
| 7075 |
+
"epoch": 0.0612,
|
| 7076 |
+
"grad_norm": 0.7447170615196228,
|
| 7077 |
+
"learning_rate": 4.6400609978551416e-05,
|
| 7078 |
+
"loss": 1.1482,
|
| 7079 |
+
"step": 90600
|
| 7080 |
+
},
|
| 7081 |
+
{
|
| 7082 |
+
"epoch": 0.0614,
|
| 7083 |
+
"grad_norm": 0.8162779808044434,
|
| 7084 |
+
"learning_rate": 4.639240364620882e-05,
|
| 7085 |
+
"loss": 1.2072,
|
| 7086 |
+
"step": 90700
|
| 7087 |
+
},
|
| 7088 |
+
{
|
| 7089 |
+
"epoch": 0.0616,
|
| 7090 |
+
"grad_norm": 1.2612018585205078,
|
| 7091 |
+
"learning_rate": 4.638418869700861e-05,
|
| 7092 |
+
"loss": 1.1402,
|
| 7093 |
+
"step": 90800
|
| 7094 |
+
},
|
| 7095 |
+
{
|
| 7096 |
+
"epoch": 0.0618,
|
| 7097 |
+
"grad_norm": 0.8543398380279541,
|
| 7098 |
+
"learning_rate": 4.637596513425974e-05,
|
| 7099 |
+
"loss": 1.1718,
|
| 7100 |
+
"step": 90900
|
| 7101 |
+
},
|
| 7102 |
+
{
|
| 7103 |
+
"epoch": 0.062,
|
| 7104 |
+
"grad_norm": 1.2375905513763428,
|
| 7105 |
+
"learning_rate": 4.636773296127467e-05,
|
| 7106 |
+
"loss": 1.1587,
|
| 7107 |
+
"step": 91000
|
| 7108 |
+
},
|
| 7109 |
+
{
|
| 7110 |
+
"epoch": 0.062,
|
| 7111 |
+
"eval_loss": 1.0713858604431152,
|
| 7112 |
+
"eval_runtime": 76.3385,
|
| 7113 |
+
"eval_samples_per_second": 202.296,
|
| 7114 |
+
"eval_steps_per_second": 3.17,
|
| 7115 |
+
"step": 91000
|
| 7116 |
}
|
| 7117 |
],
|
| 7118 |
"logging_steps": 100,
|
|
|
|
| 7132 |
"attributes": {}
|
| 7133 |
}
|
| 7134 |
},
|
| 7135 |
+
"total_flos": 3.970894657486848e+18,
|
| 7136 |
"train_batch_size": 64,
|
| 7137 |
"trial_name": null,
|
| 7138 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5777
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dd3ce0d8d8d7eebf502ef39400e4cf1a772a55755bed316d453e5ed6e3921245
|
| 3 |
size 5777
|