Upload 8 files
Browse files- model.safetensors +1 -1
- optimizer.pt +1 -1
- scheduler.pt +1 -1
- trainer_state.json +2964 -4
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 18494040
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ebc653e18935ef6e6d593d943659324be81a89c736dc6b33141d9a72bc9696c
|
| 3 |
size 18494040
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 37035002
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8c16f2324ca710c73fffdab8ac0b51ca0937adca9d3ea12a4fcf7b6b75c642b3
|
| 3 |
size 37035002
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:caeb79da12fade882c795419ac73c6806820c6ccef19831ac9e9b66b6ca1212b
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1391,11 +1391,2971 @@
|
|
| 1391 |
"loss": 4.5421,
|
| 1392 |
"num_input_tokens_seen": 151536512,
|
| 1393 |
"step": 25950
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1394 |
}
|
| 1395 |
],
|
| 1396 |
"logging_steps": 150,
|
| 1397 |
"max_steps": 272232,
|
| 1398 |
-
"num_input_tokens_seen":
|
| 1399 |
"num_train_epochs": 3,
|
| 1400 |
"save_steps": 500,
|
| 1401 |
"stateful_callbacks": {
|
|
@@ -1410,7 +4370,7 @@
|
|
| 1410 |
"attributes": {}
|
| 1411 |
}
|
| 1412 |
},
|
| 1413 |
-
"total_flos":
|
| 1414 |
"train_batch_size": 32,
|
| 1415 |
"trial_name": null,
|
| 1416 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.8981260572266088,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 81500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1391 |
"loss": 4.5421,
|
| 1392 |
"num_input_tokens_seen": 151536512,
|
| 1393 |
"step": 25950
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 0.2876207373449631,
|
| 1397 |
+
"grad_norm": 1.851283073425293,
|
| 1398 |
+
"learning_rate": 0.0001446669703336909,
|
| 1399 |
+
"loss": 4.5469,
|
| 1400 |
+
"num_input_tokens_seen": 152395904,
|
| 1401 |
+
"step": 26100
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 0.2892737300883249,
|
| 1405 |
+
"grad_norm": 1.8593121767044067,
|
| 1406 |
+
"learning_rate": 0.00014457880893956537,
|
| 1407 |
+
"loss": 4.5536,
|
| 1408 |
+
"num_input_tokens_seen": 153263904,
|
| 1409 |
+
"step": 26250
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.29092672283168675,
|
| 1413 |
+
"grad_norm": 1.8973345756530762,
|
| 1414 |
+
"learning_rate": 0.00014449064754543987,
|
| 1415 |
+
"loss": 4.5613,
|
| 1416 |
+
"num_input_tokens_seen": 154143616,
|
| 1417 |
+
"step": 26400
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.29257971557504864,
|
| 1421 |
+
"grad_norm": 1.9224932193756104,
|
| 1422 |
+
"learning_rate": 0.00014440307389394186,
|
| 1423 |
+
"loss": 4.5513,
|
| 1424 |
+
"num_input_tokens_seen": 155013568,
|
| 1425 |
+
"step": 26550
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 0.2942327083184105,
|
| 1429 |
+
"grad_norm": 1.8901547193527222,
|
| 1430 |
+
"learning_rate": 0.00014431491249981634,
|
| 1431 |
+
"loss": 4.5552,
|
| 1432 |
+
"num_input_tokens_seen": 155889408,
|
| 1433 |
+
"step": 26700
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 0.29588570106177237,
|
| 1437 |
+
"grad_norm": 1.8522729873657227,
|
| 1438 |
+
"learning_rate": 0.0001442267511056908,
|
| 1439 |
+
"loss": 4.546,
|
| 1440 |
+
"num_input_tokens_seen": 156769184,
|
| 1441 |
+
"step": 26850
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.2975386938051342,
|
| 1445 |
+
"grad_norm": 1.8729475736618042,
|
| 1446 |
+
"learning_rate": 0.00014413858971156531,
|
| 1447 |
+
"loss": 4.553,
|
| 1448 |
+
"num_input_tokens_seen": 157648416,
|
| 1449 |
+
"step": 27000
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.29919168654849604,
|
| 1453 |
+
"grad_norm": 1.8859171867370605,
|
| 1454 |
+
"learning_rate": 0.0001440504283174398,
|
| 1455 |
+
"loss": 4.5468,
|
| 1456 |
+
"num_input_tokens_seen": 158528480,
|
| 1457 |
+
"step": 27150
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 0.30084467929185793,
|
| 1461 |
+
"grad_norm": 1.8305902481079102,
|
| 1462 |
+
"learning_rate": 0.0001439628546659418,
|
| 1463 |
+
"loss": 4.554,
|
| 1464 |
+
"num_input_tokens_seen": 159401376,
|
| 1465 |
+
"step": 27300
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.30249767203521977,
|
| 1469 |
+
"grad_norm": 1.7858612537384033,
|
| 1470 |
+
"learning_rate": 0.00014387469327181628,
|
| 1471 |
+
"loss": 4.5459,
|
| 1472 |
+
"num_input_tokens_seen": 160297888,
|
| 1473 |
+
"step": 27450
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.3041506647785816,
|
| 1477 |
+
"grad_norm": 1.9333041906356812,
|
| 1478 |
+
"learning_rate": 0.00014378653187769078,
|
| 1479 |
+
"loss": 4.5488,
|
| 1480 |
+
"num_input_tokens_seen": 161189632,
|
| 1481 |
+
"step": 27600
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 0.3058036575219435,
|
| 1485 |
+
"grad_norm": 1.904167652130127,
|
| 1486 |
+
"learning_rate": 0.00014369837048356526,
|
| 1487 |
+
"loss": 4.5457,
|
| 1488 |
+
"num_input_tokens_seen": 162069728,
|
| 1489 |
+
"step": 27750
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 0.30745665026530533,
|
| 1493 |
+
"grad_norm": 1.8620493412017822,
|
| 1494 |
+
"learning_rate": 0.00014361020908943976,
|
| 1495 |
+
"loss": 4.543,
|
| 1496 |
+
"num_input_tokens_seen": 162956000,
|
| 1497 |
+
"step": 27900
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 0.30910964300866717,
|
| 1501 |
+
"grad_norm": 1.9523295164108276,
|
| 1502 |
+
"learning_rate": 0.00014352204769531424,
|
| 1503 |
+
"loss": 4.5368,
|
| 1504 |
+
"num_input_tokens_seen": 163850880,
|
| 1505 |
+
"step": 28050
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 0.31076263575202906,
|
| 1509 |
+
"grad_norm": 1.842017412185669,
|
| 1510 |
+
"learning_rate": 0.00014343388630118874,
|
| 1511 |
+
"loss": 4.5423,
|
| 1512 |
+
"num_input_tokens_seen": 164723840,
|
| 1513 |
+
"step": 28200
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 0.3124156284953909,
|
| 1517 |
+
"grad_norm": 2.015977621078491,
|
| 1518 |
+
"learning_rate": 0.0001433457249070632,
|
| 1519 |
+
"loss": 4.5306,
|
| 1520 |
+
"num_input_tokens_seen": 165582464,
|
| 1521 |
+
"step": 28350
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.3140686212387528,
|
| 1525 |
+
"grad_norm": 1.7622820138931274,
|
| 1526 |
+
"learning_rate": 0.00014325756351293772,
|
| 1527 |
+
"loss": 4.5336,
|
| 1528 |
+
"num_input_tokens_seen": 166442496,
|
| 1529 |
+
"step": 28500
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.3157216139821146,
|
| 1533 |
+
"grad_norm": 1.8762463331222534,
|
| 1534 |
+
"learning_rate": 0.0001431694021188122,
|
| 1535 |
+
"loss": 4.5363,
|
| 1536 |
+
"num_input_tokens_seen": 167318048,
|
| 1537 |
+
"step": 28650
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 0.31737460672547646,
|
| 1541 |
+
"grad_norm": 1.9524190425872803,
|
| 1542 |
+
"learning_rate": 0.00014308124072468667,
|
| 1543 |
+
"loss": 4.5498,
|
| 1544 |
+
"num_input_tokens_seen": 168209312,
|
| 1545 |
+
"step": 28800
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 0.31902759946883835,
|
| 1549 |
+
"grad_norm": 1.8535164594650269,
|
| 1550 |
+
"learning_rate": 0.00014299307933056117,
|
| 1551 |
+
"loss": 4.5481,
|
| 1552 |
+
"num_input_tokens_seen": 169096896,
|
| 1553 |
+
"step": 28950
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 0.3206805922122002,
|
| 1557 |
+
"grad_norm": 1.9309056997299194,
|
| 1558 |
+
"learning_rate": 0.00014290491793643564,
|
| 1559 |
+
"loss": 4.5416,
|
| 1560 |
+
"num_input_tokens_seen": 169963008,
|
| 1561 |
+
"step": 29100
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 0.322333584955562,
|
| 1565 |
+
"grad_norm": 1.8341314792633057,
|
| 1566 |
+
"learning_rate": 0.00014281675654231014,
|
| 1567 |
+
"loss": 4.5439,
|
| 1568 |
+
"num_input_tokens_seen": 170836736,
|
| 1569 |
+
"step": 29250
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 0.3239865776989239,
|
| 1573 |
+
"grad_norm": 1.8341432809829712,
|
| 1574 |
+
"learning_rate": 0.00014272859514818462,
|
| 1575 |
+
"loss": 4.5365,
|
| 1576 |
+
"num_input_tokens_seen": 171728896,
|
| 1577 |
+
"step": 29400
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.32563957044228575,
|
| 1581 |
+
"grad_norm": 1.9161962270736694,
|
| 1582 |
+
"learning_rate": 0.00014264043375405912,
|
| 1583 |
+
"loss": 4.5428,
|
| 1584 |
+
"num_input_tokens_seen": 172603008,
|
| 1585 |
+
"step": 29550
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.3272925631856476,
|
| 1589 |
+
"grad_norm": 1.8521162271499634,
|
| 1590 |
+
"learning_rate": 0.0001425522723599336,
|
| 1591 |
+
"loss": 4.5283,
|
| 1592 |
+
"num_input_tokens_seen": 173475392,
|
| 1593 |
+
"step": 29700
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 0.3289455559290095,
|
| 1597 |
+
"grad_norm": 1.9026546478271484,
|
| 1598 |
+
"learning_rate": 0.0001424641109658081,
|
| 1599 |
+
"loss": 4.5306,
|
| 1600 |
+
"num_input_tokens_seen": 174360832,
|
| 1601 |
+
"step": 29850
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 0.3305985486723713,
|
| 1605 |
+
"grad_norm": 1.9297667741775513,
|
| 1606 |
+
"learning_rate": 0.00014237594957168257,
|
| 1607 |
+
"loss": 4.5301,
|
| 1608 |
+
"num_input_tokens_seen": 175244576,
|
| 1609 |
+
"step": 30000
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 0.3322515414157332,
|
| 1613 |
+
"grad_norm": 1.9747087955474854,
|
| 1614 |
+
"learning_rate": 0.00014228778817755705,
|
| 1615 |
+
"loss": 4.54,
|
| 1616 |
+
"num_input_tokens_seen": 176108352,
|
| 1617 |
+
"step": 30150
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 0.33390453415909505,
|
| 1621 |
+
"grad_norm": 1.9235451221466064,
|
| 1622 |
+
"learning_rate": 0.00014219962678343155,
|
| 1623 |
+
"loss": 4.5438,
|
| 1624 |
+
"num_input_tokens_seen": 176987808,
|
| 1625 |
+
"step": 30300
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 0.3355575269024569,
|
| 1629 |
+
"grad_norm": 1.8416277170181274,
|
| 1630 |
+
"learning_rate": 0.00014211146538930603,
|
| 1631 |
+
"loss": 4.5476,
|
| 1632 |
+
"num_input_tokens_seen": 177875584,
|
| 1633 |
+
"step": 30450
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.3372105196458188,
|
| 1637 |
+
"grad_norm": 1.899798035621643,
|
| 1638 |
+
"learning_rate": 0.00014202330399518053,
|
| 1639 |
+
"loss": 4.5413,
|
| 1640 |
+
"num_input_tokens_seen": 178752480,
|
| 1641 |
+
"step": 30600
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.3388635123891806,
|
| 1645 |
+
"grad_norm": 1.8849375247955322,
|
| 1646 |
+
"learning_rate": 0.000141935142601055,
|
| 1647 |
+
"loss": 4.5199,
|
| 1648 |
+
"num_input_tokens_seen": 179634080,
|
| 1649 |
+
"step": 30750
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 0.34051650513254245,
|
| 1653 |
+
"grad_norm": 1.7944519519805908,
|
| 1654 |
+
"learning_rate": 0.0001418469812069295,
|
| 1655 |
+
"loss": 4.5268,
|
| 1656 |
+
"num_input_tokens_seen": 180501472,
|
| 1657 |
+
"step": 30900
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 0.34216949787590434,
|
| 1661 |
+
"grad_norm": 1.8572932481765747,
|
| 1662 |
+
"learning_rate": 0.00014175881981280398,
|
| 1663 |
+
"loss": 4.5268,
|
| 1664 |
+
"num_input_tokens_seen": 181388384,
|
| 1665 |
+
"step": 31050
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 0.3438224906192662,
|
| 1669 |
+
"grad_norm": 1.8637559413909912,
|
| 1670 |
+
"learning_rate": 0.00014167065841867846,
|
| 1671 |
+
"loss": 4.5245,
|
| 1672 |
+
"num_input_tokens_seen": 182271040,
|
| 1673 |
+
"step": 31200
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 0.345475483362628,
|
| 1677 |
+
"grad_norm": 1.902794361114502,
|
| 1678 |
+
"learning_rate": 0.00014158249702455296,
|
| 1679 |
+
"loss": 4.5138,
|
| 1680 |
+
"num_input_tokens_seen": 183160032,
|
| 1681 |
+
"step": 31350
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 0.3471284761059899,
|
| 1685 |
+
"grad_norm": 1.8915212154388428,
|
| 1686 |
+
"learning_rate": 0.00014149433563042743,
|
| 1687 |
+
"loss": 4.5296,
|
| 1688 |
+
"num_input_tokens_seen": 184045856,
|
| 1689 |
+
"step": 31500
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.34878146884935174,
|
| 1693 |
+
"grad_norm": 1.9054772853851318,
|
| 1694 |
+
"learning_rate": 0.00014140617423630193,
|
| 1695 |
+
"loss": 4.5343,
|
| 1696 |
+
"num_input_tokens_seen": 184935872,
|
| 1697 |
+
"step": 31650
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.35043446159271363,
|
| 1701 |
+
"grad_norm": 1.8381603956222534,
|
| 1702 |
+
"learning_rate": 0.0001413180128421764,
|
| 1703 |
+
"loss": 4.5241,
|
| 1704 |
+
"num_input_tokens_seen": 185812160,
|
| 1705 |
+
"step": 31800
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 0.35208745433607547,
|
| 1709 |
+
"grad_norm": 1.8929849863052368,
|
| 1710 |
+
"learning_rate": 0.0001412298514480509,
|
| 1711 |
+
"loss": 4.5311,
|
| 1712 |
+
"num_input_tokens_seen": 186698304,
|
| 1713 |
+
"step": 31950
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 0.3537404470794373,
|
| 1717 |
+
"grad_norm": 1.8554471731185913,
|
| 1718 |
+
"learning_rate": 0.0001411422777965529,
|
| 1719 |
+
"loss": 4.52,
|
| 1720 |
+
"num_input_tokens_seen": 187570560,
|
| 1721 |
+
"step": 32100
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 0.3553934398227992,
|
| 1725 |
+
"grad_norm": 1.8524342775344849,
|
| 1726 |
+
"learning_rate": 0.00014105411640242738,
|
| 1727 |
+
"loss": 4.5231,
|
| 1728 |
+
"num_input_tokens_seen": 188446880,
|
| 1729 |
+
"step": 32250
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 0.35704643256616103,
|
| 1733 |
+
"grad_norm": 1.8730753660202026,
|
| 1734 |
+
"learning_rate": 0.00014096595500830188,
|
| 1735 |
+
"loss": 4.5315,
|
| 1736 |
+
"num_input_tokens_seen": 189329856,
|
| 1737 |
+
"step": 32400
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 0.35869942530952287,
|
| 1741 |
+
"grad_norm": 1.8252793550491333,
|
| 1742 |
+
"learning_rate": 0.00014087779361417636,
|
| 1743 |
+
"loss": 4.5165,
|
| 1744 |
+
"num_input_tokens_seen": 190203872,
|
| 1745 |
+
"step": 32550
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.36035241805288476,
|
| 1749 |
+
"grad_norm": 1.8541933298110962,
|
| 1750 |
+
"learning_rate": 0.00014078963222005086,
|
| 1751 |
+
"loss": 4.5318,
|
| 1752 |
+
"num_input_tokens_seen": 191086784,
|
| 1753 |
+
"step": 32700
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.3620054107962466,
|
| 1757 |
+
"grad_norm": 1.9790152311325073,
|
| 1758 |
+
"learning_rate": 0.00014070147082592533,
|
| 1759 |
+
"loss": 4.5202,
|
| 1760 |
+
"num_input_tokens_seen": 191964288,
|
| 1761 |
+
"step": 32850
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 0.3636584035396085,
|
| 1765 |
+
"grad_norm": 1.980690836906433,
|
| 1766 |
+
"learning_rate": 0.00014061330943179983,
|
| 1767 |
+
"loss": 4.5282,
|
| 1768 |
+
"num_input_tokens_seen": 192848672,
|
| 1769 |
+
"step": 33000
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 0.3653113962829703,
|
| 1773 |
+
"grad_norm": 1.8499431610107422,
|
| 1774 |
+
"learning_rate": 0.0001405251480376743,
|
| 1775 |
+
"loss": 4.5094,
|
| 1776 |
+
"num_input_tokens_seen": 193723936,
|
| 1777 |
+
"step": 33150
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 0.36696438902633216,
|
| 1781 |
+
"grad_norm": 1.7975043058395386,
|
| 1782 |
+
"learning_rate": 0.0001404369866435488,
|
| 1783 |
+
"loss": 4.5296,
|
| 1784 |
+
"num_input_tokens_seen": 194603072,
|
| 1785 |
+
"step": 33300
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 0.36861738176969405,
|
| 1789 |
+
"grad_norm": 1.8439886569976807,
|
| 1790 |
+
"learning_rate": 0.0001403488252494233,
|
| 1791 |
+
"loss": 4.5274,
|
| 1792 |
+
"num_input_tokens_seen": 195468512,
|
| 1793 |
+
"step": 33450
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 0.3702703745130559,
|
| 1797 |
+
"grad_norm": 1.8969649076461792,
|
| 1798 |
+
"learning_rate": 0.0001402606638552978,
|
| 1799 |
+
"loss": 4.5195,
|
| 1800 |
+
"num_input_tokens_seen": 196345888,
|
| 1801 |
+
"step": 33600
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.3719233672564177,
|
| 1805 |
+
"grad_norm": 1.8763043880462646,
|
| 1806 |
+
"learning_rate": 0.00014017250246117226,
|
| 1807 |
+
"loss": 4.5071,
|
| 1808 |
+
"num_input_tokens_seen": 197239776,
|
| 1809 |
+
"step": 33750
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 0.3735763599997796,
|
| 1813 |
+
"grad_norm": 1.8754463195800781,
|
| 1814 |
+
"learning_rate": 0.00014008434106704677,
|
| 1815 |
+
"loss": 4.513,
|
| 1816 |
+
"num_input_tokens_seen": 198122528,
|
| 1817 |
+
"step": 33900
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 0.37522935274314145,
|
| 1821 |
+
"grad_norm": 2.011179208755493,
|
| 1822 |
+
"learning_rate": 0.00013999617967292124,
|
| 1823 |
+
"loss": 4.5146,
|
| 1824 |
+
"num_input_tokens_seen": 198998208,
|
| 1825 |
+
"step": 34050
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 0.3768823454865033,
|
| 1829 |
+
"grad_norm": 1.8127686977386475,
|
| 1830 |
+
"learning_rate": 0.00013990801827879574,
|
| 1831 |
+
"loss": 4.5103,
|
| 1832 |
+
"num_input_tokens_seen": 199877440,
|
| 1833 |
+
"step": 34200
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 0.3785353382298652,
|
| 1837 |
+
"grad_norm": 1.792159080505371,
|
| 1838 |
+
"learning_rate": 0.00013981985688467022,
|
| 1839 |
+
"loss": 4.5139,
|
| 1840 |
+
"num_input_tokens_seen": 200738880,
|
| 1841 |
+
"step": 34350
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 0.380188330973227,
|
| 1845 |
+
"grad_norm": 1.7812391519546509,
|
| 1846 |
+
"learning_rate": 0.0001397322832331722,
|
| 1847 |
+
"loss": 4.5151,
|
| 1848 |
+
"num_input_tokens_seen": 201618880,
|
| 1849 |
+
"step": 34500
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 0.3818413237165889,
|
| 1853 |
+
"grad_norm": 1.8573509454727173,
|
| 1854 |
+
"learning_rate": 0.00013964412183904668,
|
| 1855 |
+
"loss": 4.5113,
|
| 1856 |
+
"num_input_tokens_seen": 202481888,
|
| 1857 |
+
"step": 34650
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.38349431645995075,
|
| 1861 |
+
"grad_norm": 1.9190624952316284,
|
| 1862 |
+
"learning_rate": 0.0001395559604449212,
|
| 1863 |
+
"loss": 4.5118,
|
| 1864 |
+
"num_input_tokens_seen": 203344320,
|
| 1865 |
+
"step": 34800
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 0.3851473092033126,
|
| 1869 |
+
"grad_norm": 1.8508821725845337,
|
| 1870 |
+
"learning_rate": 0.00013946779905079566,
|
| 1871 |
+
"loss": 4.5152,
|
| 1872 |
+
"num_input_tokens_seen": 204227936,
|
| 1873 |
+
"step": 34950
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 0.3868003019466745,
|
| 1877 |
+
"grad_norm": 1.8276251554489136,
|
| 1878 |
+
"learning_rate": 0.00013937963765667016,
|
| 1879 |
+
"loss": 4.5182,
|
| 1880 |
+
"num_input_tokens_seen": 205119008,
|
| 1881 |
+
"step": 35100
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 0.3884532946900363,
|
| 1885 |
+
"grad_norm": 1.773224949836731,
|
| 1886 |
+
"learning_rate": 0.00013929147626254464,
|
| 1887 |
+
"loss": 4.5091,
|
| 1888 |
+
"num_input_tokens_seen": 205989696,
|
| 1889 |
+
"step": 35250
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 0.39010628743339815,
|
| 1893 |
+
"grad_norm": 1.889553189277649,
|
| 1894 |
+
"learning_rate": 0.00013920331486841914,
|
| 1895 |
+
"loss": 4.5111,
|
| 1896 |
+
"num_input_tokens_seen": 206865088,
|
| 1897 |
+
"step": 35400
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 0.39175928017676004,
|
| 1901 |
+
"grad_norm": 1.844975233078003,
|
| 1902 |
+
"learning_rate": 0.00013911515347429362,
|
| 1903 |
+
"loss": 4.4982,
|
| 1904 |
+
"num_input_tokens_seen": 207738880,
|
| 1905 |
+
"step": 35550
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 0.3934122729201219,
|
| 1909 |
+
"grad_norm": 1.8811978101730347,
|
| 1910 |
+
"learning_rate": 0.0001390269920801681,
|
| 1911 |
+
"loss": 4.5168,
|
| 1912 |
+
"num_input_tokens_seen": 208626688,
|
| 1913 |
+
"step": 35700
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.3950652656634837,
|
| 1917 |
+
"grad_norm": 1.8564339876174927,
|
| 1918 |
+
"learning_rate": 0.0001389388306860426,
|
| 1919 |
+
"loss": 4.5122,
|
| 1920 |
+
"num_input_tokens_seen": 209484800,
|
| 1921 |
+
"step": 35850
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 0.3967182584068456,
|
| 1925 |
+
"grad_norm": 1.8316396474838257,
|
| 1926 |
+
"learning_rate": 0.00013885066929191707,
|
| 1927 |
+
"loss": 4.5046,
|
| 1928 |
+
"num_input_tokens_seen": 210369792,
|
| 1929 |
+
"step": 36000
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 0.39837125115020744,
|
| 1933 |
+
"grad_norm": 1.925075650215149,
|
| 1934 |
+
"learning_rate": 0.00013876250789779157,
|
| 1935 |
+
"loss": 4.5116,
|
| 1936 |
+
"num_input_tokens_seen": 211250176,
|
| 1937 |
+
"step": 36150
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 0.40002424389356933,
|
| 1941 |
+
"grad_norm": 1.860660195350647,
|
| 1942 |
+
"learning_rate": 0.00013867434650366605,
|
| 1943 |
+
"loss": 4.4998,
|
| 1944 |
+
"num_input_tokens_seen": 212136000,
|
| 1945 |
+
"step": 36300
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 0.40167723663693117,
|
| 1949 |
+
"grad_norm": 1.8505064249038696,
|
| 1950 |
+
"learning_rate": 0.00013858618510954055,
|
| 1951 |
+
"loss": 4.5124,
|
| 1952 |
+
"num_input_tokens_seen": 213009056,
|
| 1953 |
+
"step": 36450
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 0.403330229380293,
|
| 1957 |
+
"grad_norm": 1.8910654783248901,
|
| 1958 |
+
"learning_rate": 0.00013849802371541502,
|
| 1959 |
+
"loss": 4.4966,
|
| 1960 |
+
"num_input_tokens_seen": 213871744,
|
| 1961 |
+
"step": 36600
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 0.4049832221236549,
|
| 1965 |
+
"grad_norm": 1.883748173713684,
|
| 1966 |
+
"learning_rate": 0.0001384098623212895,
|
| 1967 |
+
"loss": 4.5074,
|
| 1968 |
+
"num_input_tokens_seen": 214739520,
|
| 1969 |
+
"step": 36750
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.40663621486701673,
|
| 1973 |
+
"grad_norm": 1.8115665912628174,
|
| 1974 |
+
"learning_rate": 0.000138321700927164,
|
| 1975 |
+
"loss": 4.5187,
|
| 1976 |
+
"num_input_tokens_seen": 215629472,
|
| 1977 |
+
"step": 36900
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 0.40828920761037857,
|
| 1981 |
+
"grad_norm": 1.9036102294921875,
|
| 1982 |
+
"learning_rate": 0.000138234127275666,
|
| 1983 |
+
"loss": 4.4892,
|
| 1984 |
+
"num_input_tokens_seen": 216497184,
|
| 1985 |
+
"step": 37050
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 0.40994220035374046,
|
| 1989 |
+
"grad_norm": 1.8916597366333008,
|
| 1990 |
+
"learning_rate": 0.0001381459658815405,
|
| 1991 |
+
"loss": 4.5019,
|
| 1992 |
+
"num_input_tokens_seen": 217381152,
|
| 1993 |
+
"step": 37200
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 0.4115951930971023,
|
| 1997 |
+
"grad_norm": 1.8847101926803589,
|
| 1998 |
+
"learning_rate": 0.00013805780448741497,
|
| 1999 |
+
"loss": 4.5064,
|
| 2000 |
+
"num_input_tokens_seen": 218251456,
|
| 2001 |
+
"step": 37350
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 0.41324818584046413,
|
| 2005 |
+
"grad_norm": 1.730322241783142,
|
| 2006 |
+
"learning_rate": 0.00013796964309328947,
|
| 2007 |
+
"loss": 4.4988,
|
| 2008 |
+
"num_input_tokens_seen": 219141056,
|
| 2009 |
+
"step": 37500
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 0.414901178583826,
|
| 2013 |
+
"grad_norm": 1.833764672279358,
|
| 2014 |
+
"learning_rate": 0.00013788148169916395,
|
| 2015 |
+
"loss": 4.5155,
|
| 2016 |
+
"num_input_tokens_seen": 220036992,
|
| 2017 |
+
"step": 37650
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 0.41655417132718786,
|
| 2021 |
+
"grad_norm": 1.8188276290893555,
|
| 2022 |
+
"learning_rate": 0.00013779332030503845,
|
| 2023 |
+
"loss": 4.4984,
|
| 2024 |
+
"num_input_tokens_seen": 220910208,
|
| 2025 |
+
"step": 37800
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.41820716407054975,
|
| 2029 |
+
"grad_norm": 1.7197022438049316,
|
| 2030 |
+
"learning_rate": 0.00013770515891091292,
|
| 2031 |
+
"loss": 4.5018,
|
| 2032 |
+
"num_input_tokens_seen": 221803520,
|
| 2033 |
+
"step": 37950
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 0.4198601568139116,
|
| 2037 |
+
"grad_norm": 1.879470944404602,
|
| 2038 |
+
"learning_rate": 0.00013761699751678742,
|
| 2039 |
+
"loss": 4.497,
|
| 2040 |
+
"num_input_tokens_seen": 222679392,
|
| 2041 |
+
"step": 38100
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 0.4215131495572734,
|
| 2045 |
+
"grad_norm": 1.716430902481079,
|
| 2046 |
+
"learning_rate": 0.0001375288361226619,
|
| 2047 |
+
"loss": 4.4979,
|
| 2048 |
+
"num_input_tokens_seen": 223568224,
|
| 2049 |
+
"step": 38250
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 0.4231661423006353,
|
| 2053 |
+
"grad_norm": 1.8879536390304565,
|
| 2054 |
+
"learning_rate": 0.0001374406747285364,
|
| 2055 |
+
"loss": 4.4972,
|
| 2056 |
+
"num_input_tokens_seen": 224460192,
|
| 2057 |
+
"step": 38400
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 0.42481913504399715,
|
| 2061 |
+
"grad_norm": 1.9361037015914917,
|
| 2062 |
+
"learning_rate": 0.00013735251333441088,
|
| 2063 |
+
"loss": 4.4968,
|
| 2064 |
+
"num_input_tokens_seen": 225318816,
|
| 2065 |
+
"step": 38550
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 0.426472127787359,
|
| 2069 |
+
"grad_norm": 1.8587005138397217,
|
| 2070 |
+
"learning_rate": 0.00013726435194028535,
|
| 2071 |
+
"loss": 4.4982,
|
| 2072 |
+
"num_input_tokens_seen": 226176704,
|
| 2073 |
+
"step": 38700
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 0.4281251205307209,
|
| 2077 |
+
"grad_norm": 1.8819633722305298,
|
| 2078 |
+
"learning_rate": 0.00013717619054615985,
|
| 2079 |
+
"loss": 4.4929,
|
| 2080 |
+
"num_input_tokens_seen": 227049984,
|
| 2081 |
+
"step": 38850
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.4297781132740827,
|
| 2085 |
+
"grad_norm": 1.8268516063690186,
|
| 2086 |
+
"learning_rate": 0.00013708802915203433,
|
| 2087 |
+
"loss": 4.5063,
|
| 2088 |
+
"num_input_tokens_seen": 227926784,
|
| 2089 |
+
"step": 39000
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 0.4314311060174446,
|
| 2093 |
+
"grad_norm": 1.8819466829299927,
|
| 2094 |
+
"learning_rate": 0.00013699986775790883,
|
| 2095 |
+
"loss": 4.5024,
|
| 2096 |
+
"num_input_tokens_seen": 228823584,
|
| 2097 |
+
"step": 39150
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 0.43308409876080645,
|
| 2101 |
+
"grad_norm": 1.8801319599151611,
|
| 2102 |
+
"learning_rate": 0.0001369117063637833,
|
| 2103 |
+
"loss": 4.5043,
|
| 2104 |
+
"num_input_tokens_seen": 229691904,
|
| 2105 |
+
"step": 39300
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 0.4347370915041683,
|
| 2109 |
+
"grad_norm": 1.8677760362625122,
|
| 2110 |
+
"learning_rate": 0.0001368241327122853,
|
| 2111 |
+
"loss": 4.5102,
|
| 2112 |
+
"num_input_tokens_seen": 230569056,
|
| 2113 |
+
"step": 39450
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 0.4363900842475302,
|
| 2117 |
+
"grad_norm": 1.8280800580978394,
|
| 2118 |
+
"learning_rate": 0.0001367359713181598,
|
| 2119 |
+
"loss": 4.5014,
|
| 2120 |
+
"num_input_tokens_seen": 231437792,
|
| 2121 |
+
"step": 39600
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 0.438043076990892,
|
| 2125 |
+
"grad_norm": 1.7932913303375244,
|
| 2126 |
+
"learning_rate": 0.00013664780992403427,
|
| 2127 |
+
"loss": 4.5082,
|
| 2128 |
+
"num_input_tokens_seen": 232310848,
|
| 2129 |
+
"step": 39750
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 0.43969606973425385,
|
| 2133 |
+
"grad_norm": 1.7872642278671265,
|
| 2134 |
+
"learning_rate": 0.00013655964852990878,
|
| 2135 |
+
"loss": 4.5091,
|
| 2136 |
+
"num_input_tokens_seen": 233190464,
|
| 2137 |
+
"step": 39900
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.44134906247761574,
|
| 2141 |
+
"grad_norm": 1.9207595586776733,
|
| 2142 |
+
"learning_rate": 0.00013647148713578325,
|
| 2143 |
+
"loss": 4.4932,
|
| 2144 |
+
"num_input_tokens_seen": 234064256,
|
| 2145 |
+
"step": 40050
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 0.4430020552209776,
|
| 2149 |
+
"grad_norm": 1.9176338911056519,
|
| 2150 |
+
"learning_rate": 0.00013638332574165773,
|
| 2151 |
+
"loss": 4.4982,
|
| 2152 |
+
"num_input_tokens_seen": 234931104,
|
| 2153 |
+
"step": 40200
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 0.4446550479643394,
|
| 2157 |
+
"grad_norm": 1.839328646659851,
|
| 2158 |
+
"learning_rate": 0.00013629516434753223,
|
| 2159 |
+
"loss": 4.5,
|
| 2160 |
+
"num_input_tokens_seen": 235790176,
|
| 2161 |
+
"step": 40350
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 0.4463080407077013,
|
| 2165 |
+
"grad_norm": 1.8120410442352295,
|
| 2166 |
+
"learning_rate": 0.0001362070029534067,
|
| 2167 |
+
"loss": 4.4866,
|
| 2168 |
+
"num_input_tokens_seen": 236663488,
|
| 2169 |
+
"step": 40500
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 0.44796103345106314,
|
| 2173 |
+
"grad_norm": 1.7365509271621704,
|
| 2174 |
+
"learning_rate": 0.0001361188415592812,
|
| 2175 |
+
"loss": 4.484,
|
| 2176 |
+
"num_input_tokens_seen": 237550016,
|
| 2177 |
+
"step": 40650
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 0.44961402619442503,
|
| 2181 |
+
"grad_norm": 1.8573724031448364,
|
| 2182 |
+
"learning_rate": 0.00013603068016515568,
|
| 2183 |
+
"loss": 4.4949,
|
| 2184 |
+
"num_input_tokens_seen": 238428800,
|
| 2185 |
+
"step": 40800
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 0.45126701893778687,
|
| 2189 |
+
"grad_norm": 1.8714196681976318,
|
| 2190 |
+
"learning_rate": 0.00013594251877103018,
|
| 2191 |
+
"loss": 4.4861,
|
| 2192 |
+
"num_input_tokens_seen": 239301216,
|
| 2193 |
+
"step": 40950
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.4529200116811487,
|
| 2197 |
+
"grad_norm": 1.813636064529419,
|
| 2198 |
+
"learning_rate": 0.00013585435737690466,
|
| 2199 |
+
"loss": 4.4939,
|
| 2200 |
+
"num_input_tokens_seen": 240172352,
|
| 2201 |
+
"step": 41100
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.4545730044245106,
|
| 2205 |
+
"grad_norm": 1.7828494310379028,
|
| 2206 |
+
"learning_rate": 0.00013576619598277913,
|
| 2207 |
+
"loss": 4.5073,
|
| 2208 |
+
"num_input_tokens_seen": 241058368,
|
| 2209 |
+
"step": 41250
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 0.45622599716787243,
|
| 2213 |
+
"grad_norm": 1.9278478622436523,
|
| 2214 |
+
"learning_rate": 0.00013567803458865363,
|
| 2215 |
+
"loss": 4.4831,
|
| 2216 |
+
"num_input_tokens_seen": 241944320,
|
| 2217 |
+
"step": 41400
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 0.45787898991123427,
|
| 2221 |
+
"grad_norm": 1.8244856595993042,
|
| 2222 |
+
"learning_rate": 0.0001355898731945281,
|
| 2223 |
+
"loss": 4.4889,
|
| 2224 |
+
"num_input_tokens_seen": 242827808,
|
| 2225 |
+
"step": 41550
|
| 2226 |
+
},
|
| 2227 |
+
{
|
| 2228 |
+
"epoch": 0.45953198265459616,
|
| 2229 |
+
"grad_norm": 1.9578852653503418,
|
| 2230 |
+
"learning_rate": 0.0001355017118004026,
|
| 2231 |
+
"loss": 4.4938,
|
| 2232 |
+
"num_input_tokens_seen": 243697248,
|
| 2233 |
+
"step": 41700
|
| 2234 |
+
},
|
| 2235 |
+
{
|
| 2236 |
+
"epoch": 0.461184975397958,
|
| 2237 |
+
"grad_norm": 1.8681639432907104,
|
| 2238 |
+
"learning_rate": 0.0001354135504062771,
|
| 2239 |
+
"loss": 4.493,
|
| 2240 |
+
"num_input_tokens_seen": 244549632,
|
| 2241 |
+
"step": 41850
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 0.46283796814131983,
|
| 2245 |
+
"grad_norm": 1.9330469369888306,
|
| 2246 |
+
"learning_rate": 0.0001353253890121516,
|
| 2247 |
+
"loss": 4.4823,
|
| 2248 |
+
"num_input_tokens_seen": 245429600,
|
| 2249 |
+
"step": 42000
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.4644909608846817,
|
| 2253 |
+
"grad_norm": 1.775819182395935,
|
| 2254 |
+
"learning_rate": 0.00013523722761802606,
|
| 2255 |
+
"loss": 4.4887,
|
| 2256 |
+
"num_input_tokens_seen": 246307872,
|
| 2257 |
+
"step": 42150
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.46614395362804356,
|
| 2261 |
+
"grad_norm": 1.9359744787216187,
|
| 2262 |
+
"learning_rate": 0.00013514906622390057,
|
| 2263 |
+
"loss": 4.4813,
|
| 2264 |
+
"num_input_tokens_seen": 247194432,
|
| 2265 |
+
"step": 42300
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 0.46779694637140545,
|
| 2269 |
+
"grad_norm": 1.8034732341766357,
|
| 2270 |
+
"learning_rate": 0.00013506090482977504,
|
| 2271 |
+
"loss": 4.4957,
|
| 2272 |
+
"num_input_tokens_seen": 248049248,
|
| 2273 |
+
"step": 42450
|
| 2274 |
+
},
|
| 2275 |
+
{
|
| 2276 |
+
"epoch": 0.4694499391147673,
|
| 2277 |
+
"grad_norm": 1.8069413900375366,
|
| 2278 |
+
"learning_rate": 0.00013497274343564954,
|
| 2279 |
+
"loss": 4.484,
|
| 2280 |
+
"num_input_tokens_seen": 248930624,
|
| 2281 |
+
"step": 42600
|
| 2282 |
+
},
|
| 2283 |
+
{
|
| 2284 |
+
"epoch": 0.4711029318581291,
|
| 2285 |
+
"grad_norm": 1.8796617984771729,
|
| 2286 |
+
"learning_rate": 0.00013488458204152402,
|
| 2287 |
+
"loss": 4.4887,
|
| 2288 |
+
"num_input_tokens_seen": 249808512,
|
| 2289 |
+
"step": 42750
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 0.472755924601491,
|
| 2293 |
+
"grad_norm": 1.8367630243301392,
|
| 2294 |
+
"learning_rate": 0.00013479642064739852,
|
| 2295 |
+
"loss": 4.4779,
|
| 2296 |
+
"num_input_tokens_seen": 250669088,
|
| 2297 |
+
"step": 42900
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 0.47440891734485285,
|
| 2301 |
+
"grad_norm": 1.9085215330123901,
|
| 2302 |
+
"learning_rate": 0.0001347088469959005,
|
| 2303 |
+
"loss": 4.489,
|
| 2304 |
+
"num_input_tokens_seen": 251553792,
|
| 2305 |
+
"step": 43050
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.4760619100882147,
|
| 2309 |
+
"grad_norm": 1.8323450088500977,
|
| 2310 |
+
"learning_rate": 0.000134620685601775,
|
| 2311 |
+
"loss": 4.482,
|
| 2312 |
+
"num_input_tokens_seen": 252440032,
|
| 2313 |
+
"step": 43200
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.4777149028315766,
|
| 2317 |
+
"grad_norm": 2.1364452838897705,
|
| 2318 |
+
"learning_rate": 0.0001345325242076495,
|
| 2319 |
+
"loss": 4.4873,
|
| 2320 |
+
"num_input_tokens_seen": 253320576,
|
| 2321 |
+
"step": 43350
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 0.4793678955749384,
|
| 2325 |
+
"grad_norm": 1.8431365489959717,
|
| 2326 |
+
"learning_rate": 0.00013444436281352396,
|
| 2327 |
+
"loss": 4.4844,
|
| 2328 |
+
"num_input_tokens_seen": 254203552,
|
| 2329 |
+
"step": 43500
|
| 2330 |
+
},
|
| 2331 |
+
{
|
| 2332 |
+
"epoch": 0.4810208883183003,
|
| 2333 |
+
"grad_norm": 1.853715419769287,
|
| 2334 |
+
"learning_rate": 0.00013435620141939847,
|
| 2335 |
+
"loss": 4.4905,
|
| 2336 |
+
"num_input_tokens_seen": 255078656,
|
| 2337 |
+
"step": 43650
|
| 2338 |
+
},
|
| 2339 |
+
{
|
| 2340 |
+
"epoch": 0.48267388106166215,
|
| 2341 |
+
"grad_norm": 2.0351803302764893,
|
| 2342 |
+
"learning_rate": 0.00013426804002527294,
|
| 2343 |
+
"loss": 4.4854,
|
| 2344 |
+
"num_input_tokens_seen": 255961792,
|
| 2345 |
+
"step": 43800
|
| 2346 |
+
},
|
| 2347 |
+
{
|
| 2348 |
+
"epoch": 0.484326873805024,
|
| 2349 |
+
"grad_norm": 1.867595911026001,
|
| 2350 |
+
"learning_rate": 0.00013417987863114744,
|
| 2351 |
+
"loss": 4.4734,
|
| 2352 |
+
"num_input_tokens_seen": 256823744,
|
| 2353 |
+
"step": 43950
|
| 2354 |
+
},
|
| 2355 |
+
{
|
| 2356 |
+
"epoch": 0.4859798665483859,
|
| 2357 |
+
"grad_norm": 1.9231115579605103,
|
| 2358 |
+
"learning_rate": 0.00013409171723702192,
|
| 2359 |
+
"loss": 4.4982,
|
| 2360 |
+
"num_input_tokens_seen": 257687488,
|
| 2361 |
+
"step": 44100
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.4876328592917477,
|
| 2365 |
+
"grad_norm": 1.9620647430419922,
|
| 2366 |
+
"learning_rate": 0.00013400355584289642,
|
| 2367 |
+
"loss": 4.4815,
|
| 2368 |
+
"num_input_tokens_seen": 258561568,
|
| 2369 |
+
"step": 44250
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 0.48928585203510955,
|
| 2373 |
+
"grad_norm": 1.741739273071289,
|
| 2374 |
+
"learning_rate": 0.0001339153944487709,
|
| 2375 |
+
"loss": 4.4863,
|
| 2376 |
+
"num_input_tokens_seen": 259449344,
|
| 2377 |
+
"step": 44400
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 0.49093884477847144,
|
| 2381 |
+
"grad_norm": 1.9869145154953003,
|
| 2382 |
+
"learning_rate": 0.0001338272330546454,
|
| 2383 |
+
"loss": 4.48,
|
| 2384 |
+
"num_input_tokens_seen": 260317888,
|
| 2385 |
+
"step": 44550
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 0.4925918375218333,
|
| 2389 |
+
"grad_norm": 1.9364657402038574,
|
| 2390 |
+
"learning_rate": 0.00013373907166051987,
|
| 2391 |
+
"loss": 4.477,
|
| 2392 |
+
"num_input_tokens_seen": 261198656,
|
| 2393 |
+
"step": 44700
|
| 2394 |
+
},
|
| 2395 |
+
{
|
| 2396 |
+
"epoch": 0.4942448302651951,
|
| 2397 |
+
"grad_norm": 1.946509599685669,
|
| 2398 |
+
"learning_rate": 0.00013365091026639437,
|
| 2399 |
+
"loss": 4.4762,
|
| 2400 |
+
"num_input_tokens_seen": 262082432,
|
| 2401 |
+
"step": 44850
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 0.495897823008557,
|
| 2405 |
+
"grad_norm": 1.7882816791534424,
|
| 2406 |
+
"learning_rate": 0.00013356274887226885,
|
| 2407 |
+
"loss": 4.478,
|
| 2408 |
+
"num_input_tokens_seen": 262961504,
|
| 2409 |
+
"step": 45000
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 0.49755081575191884,
|
| 2413 |
+
"grad_norm": 1.8857409954071045,
|
| 2414 |
+
"learning_rate": 0.00013347458747814332,
|
| 2415 |
+
"loss": 4.48,
|
| 2416 |
+
"num_input_tokens_seen": 263810336,
|
| 2417 |
+
"step": 45150
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.49920380849528073,
|
| 2421 |
+
"grad_norm": 1.8983428478240967,
|
| 2422 |
+
"learning_rate": 0.00013338642608401783,
|
| 2423 |
+
"loss": 4.4792,
|
| 2424 |
+
"num_input_tokens_seen": 264684512,
|
| 2425 |
+
"step": 45300
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 0.5008568012386425,
|
| 2429 |
+
"grad_norm": 1.8649096488952637,
|
| 2430 |
+
"learning_rate": 0.0001332982646898923,
|
| 2431 |
+
"loss": 4.473,
|
| 2432 |
+
"num_input_tokens_seen": 265553056,
|
| 2433 |
+
"step": 45450
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 0.5025097939820045,
|
| 2437 |
+
"grad_norm": 1.8233070373535156,
|
| 2438 |
+
"learning_rate": 0.0001332101032957668,
|
| 2439 |
+
"loss": 4.4854,
|
| 2440 |
+
"num_input_tokens_seen": 266438976,
|
| 2441 |
+
"step": 45600
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 0.5041627867253663,
|
| 2445 |
+
"grad_norm": 1.8780871629714966,
|
| 2446 |
+
"learning_rate": 0.00013312194190164128,
|
| 2447 |
+
"loss": 4.4806,
|
| 2448 |
+
"num_input_tokens_seen": 267305760,
|
| 2449 |
+
"step": 45750
|
| 2450 |
+
},
|
| 2451 |
+
{
|
| 2452 |
+
"epoch": 0.5058157794687281,
|
| 2453 |
+
"grad_norm": 1.9368178844451904,
|
| 2454 |
+
"learning_rate": 0.00013303378050751578,
|
| 2455 |
+
"loss": 4.4805,
|
| 2456 |
+
"num_input_tokens_seen": 268185792,
|
| 2457 |
+
"step": 45900
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"epoch": 0.50746877221209,
|
| 2461 |
+
"grad_norm": 1.8207215070724487,
|
| 2462 |
+
"learning_rate": 0.00013294561911339026,
|
| 2463 |
+
"loss": 4.4839,
|
| 2464 |
+
"num_input_tokens_seen": 269074432,
|
| 2465 |
+
"step": 46050
|
| 2466 |
+
},
|
| 2467 |
+
{
|
| 2468 |
+
"epoch": 0.5091217649554518,
|
| 2469 |
+
"grad_norm": 1.9221956729888916,
|
| 2470 |
+
"learning_rate": 0.00013285745771926473,
|
| 2471 |
+
"loss": 4.4724,
|
| 2472 |
+
"num_input_tokens_seen": 269945728,
|
| 2473 |
+
"step": 46200
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.5107747576988138,
|
| 2477 |
+
"grad_norm": 1.8592416048049927,
|
| 2478 |
+
"learning_rate": 0.00013276929632513923,
|
| 2479 |
+
"loss": 4.4801,
|
| 2480 |
+
"num_input_tokens_seen": 270826912,
|
| 2481 |
+
"step": 46350
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 0.5124277504421756,
|
| 2485 |
+
"grad_norm": 1.8793606758117676,
|
| 2486 |
+
"learning_rate": 0.0001326811349310137,
|
| 2487 |
+
"loss": 4.4832,
|
| 2488 |
+
"num_input_tokens_seen": 271696640,
|
| 2489 |
+
"step": 46500
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 0.5140807431855374,
|
| 2493 |
+
"grad_norm": 1.9293113946914673,
|
| 2494 |
+
"learning_rate": 0.0001325929735368882,
|
| 2495 |
+
"loss": 4.4703,
|
| 2496 |
+
"num_input_tokens_seen": 272577280,
|
| 2497 |
+
"step": 46650
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 0.5157337359288993,
|
| 2501 |
+
"grad_norm": 1.8582684993743896,
|
| 2502 |
+
"learning_rate": 0.00013250481214276269,
|
| 2503 |
+
"loss": 4.4785,
|
| 2504 |
+
"num_input_tokens_seen": 273448128,
|
| 2505 |
+
"step": 46800
|
| 2506 |
+
},
|
| 2507 |
+
{
|
| 2508 |
+
"epoch": 0.5173867286722611,
|
| 2509 |
+
"grad_norm": 1.8803818225860596,
|
| 2510 |
+
"learning_rate": 0.0001324166507486372,
|
| 2511 |
+
"loss": 4.4851,
|
| 2512 |
+
"num_input_tokens_seen": 274313472,
|
| 2513 |
+
"step": 46950
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 0.5190397214156229,
|
| 2517 |
+
"grad_norm": 1.989717721939087,
|
| 2518 |
+
"learning_rate": 0.00013232907709713918,
|
| 2519 |
+
"loss": 4.4814,
|
| 2520 |
+
"num_input_tokens_seen": 275168416,
|
| 2521 |
+
"step": 47100
|
| 2522 |
+
},
|
| 2523 |
+
{
|
| 2524 |
+
"epoch": 0.5206927141589849,
|
| 2525 |
+
"grad_norm": 1.9770921468734741,
|
| 2526 |
+
"learning_rate": 0.00013224091570301365,
|
| 2527 |
+
"loss": 4.4825,
|
| 2528 |
+
"num_input_tokens_seen": 276035776,
|
| 2529 |
+
"step": 47250
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.5223457069023467,
|
| 2533 |
+
"grad_norm": 1.8764408826828003,
|
| 2534 |
+
"learning_rate": 0.00013215275430888816,
|
| 2535 |
+
"loss": 4.4788,
|
| 2536 |
+
"num_input_tokens_seen": 276909952,
|
| 2537 |
+
"step": 47400
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 0.5239986996457086,
|
| 2541 |
+
"grad_norm": 1.860009789466858,
|
| 2542 |
+
"learning_rate": 0.00013206459291476263,
|
| 2543 |
+
"loss": 4.4834,
|
| 2544 |
+
"num_input_tokens_seen": 277795232,
|
| 2545 |
+
"step": 47550
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 0.5256516923890704,
|
| 2549 |
+
"grad_norm": 1.949278473854065,
|
| 2550 |
+
"learning_rate": 0.00013197643152063713,
|
| 2551 |
+
"loss": 4.4752,
|
| 2552 |
+
"num_input_tokens_seen": 278690720,
|
| 2553 |
+
"step": 47700
|
| 2554 |
+
},
|
| 2555 |
+
{
|
| 2556 |
+
"epoch": 0.5273046851324322,
|
| 2557 |
+
"grad_norm": 1.868780255317688,
|
| 2558 |
+
"learning_rate": 0.0001318882701265116,
|
| 2559 |
+
"loss": 4.4797,
|
| 2560 |
+
"num_input_tokens_seen": 279567232,
|
| 2561 |
+
"step": 47850
|
| 2562 |
+
},
|
| 2563 |
+
{
|
| 2564 |
+
"epoch": 0.5289576778757942,
|
| 2565 |
+
"grad_norm": 1.7434320449829102,
|
| 2566 |
+
"learning_rate": 0.0001318001087323861,
|
| 2567 |
+
"loss": 4.4726,
|
| 2568 |
+
"num_input_tokens_seen": 280424832,
|
| 2569 |
+
"step": 48000
|
| 2570 |
+
},
|
| 2571 |
+
{
|
| 2572 |
+
"epoch": 0.530610670619156,
|
| 2573 |
+
"grad_norm": 1.8644661903381348,
|
| 2574 |
+
"learning_rate": 0.00013171194733826059,
|
| 2575 |
+
"loss": 4.4648,
|
| 2576 |
+
"num_input_tokens_seen": 281294752,
|
| 2577 |
+
"step": 48150
|
| 2578 |
+
},
|
| 2579 |
+
{
|
| 2580 |
+
"epoch": 0.5322636633625178,
|
| 2581 |
+
"grad_norm": 1.9029775857925415,
|
| 2582 |
+
"learning_rate": 0.00013162378594413506,
|
| 2583 |
+
"loss": 4.4791,
|
| 2584 |
+
"num_input_tokens_seen": 282156544,
|
| 2585 |
+
"step": 48300
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.5339166561058797,
|
| 2589 |
+
"grad_norm": 1.7862669229507446,
|
| 2590 |
+
"learning_rate": 0.00013153562455000956,
|
| 2591 |
+
"loss": 4.4688,
|
| 2592 |
+
"num_input_tokens_seen": 283033248,
|
| 2593 |
+
"step": 48450
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 0.5355696488492415,
|
| 2597 |
+
"grad_norm": 1.7603411674499512,
|
| 2598 |
+
"learning_rate": 0.00013144746315588404,
|
| 2599 |
+
"loss": 4.4807,
|
| 2600 |
+
"num_input_tokens_seen": 283922048,
|
| 2601 |
+
"step": 48600
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 0.5372226415926034,
|
| 2605 |
+
"grad_norm": 1.868235468864441,
|
| 2606 |
+
"learning_rate": 0.00013135930176175854,
|
| 2607 |
+
"loss": 4.4702,
|
| 2608 |
+
"num_input_tokens_seen": 284800416,
|
| 2609 |
+
"step": 48750
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 0.5388756343359653,
|
| 2613 |
+
"grad_norm": 1.864640474319458,
|
| 2614 |
+
"learning_rate": 0.00013127114036763301,
|
| 2615 |
+
"loss": 4.479,
|
| 2616 |
+
"num_input_tokens_seen": 285673888,
|
| 2617 |
+
"step": 48900
|
| 2618 |
+
},
|
| 2619 |
+
{
|
| 2620 |
+
"epoch": 0.5405286270793271,
|
| 2621 |
+
"grad_norm": 1.7705098390579224,
|
| 2622 |
+
"learning_rate": 0.00013118297897350752,
|
| 2623 |
+
"loss": 4.4782,
|
| 2624 |
+
"num_input_tokens_seen": 286553856,
|
| 2625 |
+
"step": 49050
|
| 2626 |
+
},
|
| 2627 |
+
{
|
| 2628 |
+
"epoch": 0.542181619822689,
|
| 2629 |
+
"grad_norm": 1.9763901233673096,
|
| 2630 |
+
"learning_rate": 0.000131094817579382,
|
| 2631 |
+
"loss": 4.4654,
|
| 2632 |
+
"num_input_tokens_seen": 287436736,
|
| 2633 |
+
"step": 49200
|
| 2634 |
+
},
|
| 2635 |
+
{
|
| 2636 |
+
"epoch": 0.5438346125660508,
|
| 2637 |
+
"grad_norm": 1.905661702156067,
|
| 2638 |
+
"learning_rate": 0.0001310066561852565,
|
| 2639 |
+
"loss": 4.4718,
|
| 2640 |
+
"num_input_tokens_seen": 288306976,
|
| 2641 |
+
"step": 49350
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.5454876053094126,
|
| 2645 |
+
"grad_norm": 1.8861178159713745,
|
| 2646 |
+
"learning_rate": 0.00013091849479113097,
|
| 2647 |
+
"loss": 4.4618,
|
| 2648 |
+
"num_input_tokens_seen": 289176352,
|
| 2649 |
+
"step": 49500
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 0.5471405980527746,
|
| 2653 |
+
"grad_norm": 1.8697651624679565,
|
| 2654 |
+
"learning_rate": 0.00013083033339700547,
|
| 2655 |
+
"loss": 4.4725,
|
| 2656 |
+
"num_input_tokens_seen": 290061760,
|
| 2657 |
+
"step": 49650
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 0.5487935907961364,
|
| 2661 |
+
"grad_norm": 1.8051494359970093,
|
| 2662 |
+
"learning_rate": 0.00013074217200287995,
|
| 2663 |
+
"loss": 4.459,
|
| 2664 |
+
"num_input_tokens_seen": 290924640,
|
| 2665 |
+
"step": 49800
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 0.5504465835394983,
|
| 2669 |
+
"grad_norm": 1.7766984701156616,
|
| 2670 |
+
"learning_rate": 0.00013065401060875445,
|
| 2671 |
+
"loss": 4.4729,
|
| 2672 |
+
"num_input_tokens_seen": 291801280,
|
| 2673 |
+
"step": 49950
|
| 2674 |
+
},
|
| 2675 |
+
{
|
| 2676 |
+
"epoch": 0.5520995762828601,
|
| 2677 |
+
"grad_norm": 1.7969352006912231,
|
| 2678 |
+
"learning_rate": 0.00013056584921462892,
|
| 2679 |
+
"loss": 4.4789,
|
| 2680 |
+
"num_input_tokens_seen": 292673632,
|
| 2681 |
+
"step": 50100
|
| 2682 |
+
},
|
| 2683 |
+
{
|
| 2684 |
+
"epoch": 0.5537525690262219,
|
| 2685 |
+
"grad_norm": 1.8939694166183472,
|
| 2686 |
+
"learning_rate": 0.00013047768782050343,
|
| 2687 |
+
"loss": 4.4674,
|
| 2688 |
+
"num_input_tokens_seen": 293542176,
|
| 2689 |
+
"step": 50250
|
| 2690 |
+
},
|
| 2691 |
+
{
|
| 2692 |
+
"epoch": 0.5554055617695838,
|
| 2693 |
+
"grad_norm": 1.8721749782562256,
|
| 2694 |
+
"learning_rate": 0.0001303895264263779,
|
| 2695 |
+
"loss": 4.4591,
|
| 2696 |
+
"num_input_tokens_seen": 294415424,
|
| 2697 |
+
"step": 50400
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.5570585545129457,
|
| 2701 |
+
"grad_norm": 1.7810068130493164,
|
| 2702 |
+
"learning_rate": 0.0001303013650322524,
|
| 2703 |
+
"loss": 4.475,
|
| 2704 |
+
"num_input_tokens_seen": 295292384,
|
| 2705 |
+
"step": 50550
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 0.5587115472563076,
|
| 2709 |
+
"grad_norm": 1.8184670209884644,
|
| 2710 |
+
"learning_rate": 0.00013021320363812688,
|
| 2711 |
+
"loss": 4.473,
|
| 2712 |
+
"num_input_tokens_seen": 296168640,
|
| 2713 |
+
"step": 50700
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"epoch": 0.5603645399996694,
|
| 2717 |
+
"grad_norm": 1.8167736530303955,
|
| 2718 |
+
"learning_rate": 0.00013012504224400138,
|
| 2719 |
+
"loss": 4.4604,
|
| 2720 |
+
"num_input_tokens_seen": 297035904,
|
| 2721 |
+
"step": 50850
|
| 2722 |
+
},
|
| 2723 |
+
{
|
| 2724 |
+
"epoch": 0.5620175327430312,
|
| 2725 |
+
"grad_norm": 1.902001142501831,
|
| 2726 |
+
"learning_rate": 0.00013003688084987585,
|
| 2727 |
+
"loss": 4.4653,
|
| 2728 |
+
"num_input_tokens_seen": 297893248,
|
| 2729 |
+
"step": 51000
|
| 2730 |
+
},
|
| 2731 |
+
{
|
| 2732 |
+
"epoch": 0.5636705254863931,
|
| 2733 |
+
"grad_norm": 1.7989624738693237,
|
| 2734 |
+
"learning_rate": 0.00012994930719837785,
|
| 2735 |
+
"loss": 4.4581,
|
| 2736 |
+
"num_input_tokens_seen": 298773088,
|
| 2737 |
+
"step": 51150
|
| 2738 |
+
},
|
| 2739 |
+
{
|
| 2740 |
+
"epoch": 0.565323518229755,
|
| 2741 |
+
"grad_norm": 1.9608672857284546,
|
| 2742 |
+
"learning_rate": 0.00012986114580425232,
|
| 2743 |
+
"loss": 4.4553,
|
| 2744 |
+
"num_input_tokens_seen": 299639712,
|
| 2745 |
+
"step": 51300
|
| 2746 |
+
},
|
| 2747 |
+
{
|
| 2748 |
+
"epoch": 0.5669765109731169,
|
| 2749 |
+
"grad_norm": 1.8063366413116455,
|
| 2750 |
+
"learning_rate": 0.00012977298441012682,
|
| 2751 |
+
"loss": 4.4748,
|
| 2752 |
+
"num_input_tokens_seen": 300504576,
|
| 2753 |
+
"step": 51450
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 0.5686295037164787,
|
| 2757 |
+
"grad_norm": 1.7892522811889648,
|
| 2758 |
+
"learning_rate": 0.0001296848230160013,
|
| 2759 |
+
"loss": 4.4759,
|
| 2760 |
+
"num_input_tokens_seen": 301381600,
|
| 2761 |
+
"step": 51600
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.5702824964598405,
|
| 2765 |
+
"grad_norm": 1.9009445905685425,
|
| 2766 |
+
"learning_rate": 0.00012959666162187577,
|
| 2767 |
+
"loss": 4.4643,
|
| 2768 |
+
"num_input_tokens_seen": 302264064,
|
| 2769 |
+
"step": 51750
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 0.5719354892032024,
|
| 2773 |
+
"grad_norm": 1.8855173587799072,
|
| 2774 |
+
"learning_rate": 0.00012950850022775027,
|
| 2775 |
+
"loss": 4.4569,
|
| 2776 |
+
"num_input_tokens_seen": 303131648,
|
| 2777 |
+
"step": 51900
|
| 2778 |
+
},
|
| 2779 |
+
{
|
| 2780 |
+
"epoch": 0.5735884819465642,
|
| 2781 |
+
"grad_norm": 1.8198268413543701,
|
| 2782 |
+
"learning_rate": 0.00012942033883362475,
|
| 2783 |
+
"loss": 4.465,
|
| 2784 |
+
"num_input_tokens_seen": 304010848,
|
| 2785 |
+
"step": 52050
|
| 2786 |
+
},
|
| 2787 |
+
{
|
| 2788 |
+
"epoch": 0.5752414746899261,
|
| 2789 |
+
"grad_norm": 1.8278980255126953,
|
| 2790 |
+
"learning_rate": 0.00012933217743949925,
|
| 2791 |
+
"loss": 4.4674,
|
| 2792 |
+
"num_input_tokens_seen": 304901952,
|
| 2793 |
+
"step": 52200
|
| 2794 |
+
},
|
| 2795 |
+
{
|
| 2796 |
+
"epoch": 0.576894467433288,
|
| 2797 |
+
"grad_norm": 1.8281151056289673,
|
| 2798 |
+
"learning_rate": 0.00012924401604537373,
|
| 2799 |
+
"loss": 4.4718,
|
| 2800 |
+
"num_input_tokens_seen": 305763424,
|
| 2801 |
+
"step": 52350
|
| 2802 |
+
},
|
| 2803 |
+
{
|
| 2804 |
+
"epoch": 0.5785474601766498,
|
| 2805 |
+
"grad_norm": 1.8816980123519897,
|
| 2806 |
+
"learning_rate": 0.00012915585465124823,
|
| 2807 |
+
"loss": 4.4726,
|
| 2808 |
+
"num_input_tokens_seen": 306641856,
|
| 2809 |
+
"step": 52500
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 0.5802004529200117,
|
| 2813 |
+
"grad_norm": 1.8755218982696533,
|
| 2814 |
+
"learning_rate": 0.0001290676932571227,
|
| 2815 |
+
"loss": 4.4744,
|
| 2816 |
+
"num_input_tokens_seen": 307520288,
|
| 2817 |
+
"step": 52650
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 0.5818534456633735,
|
| 2821 |
+
"grad_norm": 1.8329659700393677,
|
| 2822 |
+
"learning_rate": 0.0001289795318629972,
|
| 2823 |
+
"loss": 4.4572,
|
| 2824 |
+
"num_input_tokens_seen": 308410624,
|
| 2825 |
+
"step": 52800
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 0.5835064384067354,
|
| 2829 |
+
"grad_norm": 1.9654128551483154,
|
| 2830 |
+
"learning_rate": 0.00012889137046887168,
|
| 2831 |
+
"loss": 4.4671,
|
| 2832 |
+
"num_input_tokens_seen": 309288736,
|
| 2833 |
+
"step": 52950
|
| 2834 |
+
},
|
| 2835 |
+
{
|
| 2836 |
+
"epoch": 0.5851594311500973,
|
| 2837 |
+
"grad_norm": 1.8310860395431519,
|
| 2838 |
+
"learning_rate": 0.00012880320907474618,
|
| 2839 |
+
"loss": 4.4684,
|
| 2840 |
+
"num_input_tokens_seen": 310165440,
|
| 2841 |
+
"step": 53100
|
| 2842 |
+
},
|
| 2843 |
+
{
|
| 2844 |
+
"epoch": 0.5868124238934591,
|
| 2845 |
+
"grad_norm": 1.8104560375213623,
|
| 2846 |
+
"learning_rate": 0.00012871563542324817,
|
| 2847 |
+
"loss": 4.4546,
|
| 2848 |
+
"num_input_tokens_seen": 311044352,
|
| 2849 |
+
"step": 53250
|
| 2850 |
+
},
|
| 2851 |
+
{
|
| 2852 |
+
"epoch": 0.588465416636821,
|
| 2853 |
+
"grad_norm": 1.8414585590362549,
|
| 2854 |
+
"learning_rate": 0.00012862747402912265,
|
| 2855 |
+
"loss": 4.4625,
|
| 2856 |
+
"num_input_tokens_seen": 311919424,
|
| 2857 |
+
"step": 53400
|
| 2858 |
+
},
|
| 2859 |
+
{
|
| 2860 |
+
"epoch": 0.5901184093801828,
|
| 2861 |
+
"grad_norm": 1.724381685256958,
|
| 2862 |
+
"learning_rate": 0.00012853931263499715,
|
| 2863 |
+
"loss": 4.4632,
|
| 2864 |
+
"num_input_tokens_seen": 312803712,
|
| 2865 |
+
"step": 53550
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 0.5917714021235447,
|
| 2869 |
+
"grad_norm": 1.7701301574707031,
|
| 2870 |
+
"learning_rate": 0.00012845115124087163,
|
| 2871 |
+
"loss": 4.4609,
|
| 2872 |
+
"num_input_tokens_seen": 313672800,
|
| 2873 |
+
"step": 53700
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 0.5934243948669066,
|
| 2877 |
+
"grad_norm": 1.8755768537521362,
|
| 2878 |
+
"learning_rate": 0.00012836298984674613,
|
| 2879 |
+
"loss": 4.461,
|
| 2880 |
+
"num_input_tokens_seen": 314543776,
|
| 2881 |
+
"step": 53850
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 0.5950773876102684,
|
| 2885 |
+
"grad_norm": 1.8842816352844238,
|
| 2886 |
+
"learning_rate": 0.0001282748284526206,
|
| 2887 |
+
"loss": 4.462,
|
| 2888 |
+
"num_input_tokens_seen": 315413216,
|
| 2889 |
+
"step": 54000
|
| 2890 |
+
},
|
| 2891 |
+
{
|
| 2892 |
+
"epoch": 0.5967303803536302,
|
| 2893 |
+
"grad_norm": 1.8173580169677734,
|
| 2894 |
+
"learning_rate": 0.0001281866670584951,
|
| 2895 |
+
"loss": 4.4595,
|
| 2896 |
+
"num_input_tokens_seen": 316286592,
|
| 2897 |
+
"step": 54150
|
| 2898 |
+
},
|
| 2899 |
+
{
|
| 2900 |
+
"epoch": 0.5983833730969921,
|
| 2901 |
+
"grad_norm": 1.8613582849502563,
|
| 2902 |
+
"learning_rate": 0.00012809850566436958,
|
| 2903 |
+
"loss": 4.4729,
|
| 2904 |
+
"num_input_tokens_seen": 317171968,
|
| 2905 |
+
"step": 54300
|
| 2906 |
+
},
|
| 2907 |
+
{
|
| 2908 |
+
"epoch": 0.6000363658403539,
|
| 2909 |
+
"grad_norm": 1.8345390558242798,
|
| 2910 |
+
"learning_rate": 0.00012801034427024408,
|
| 2911 |
+
"loss": 4.4558,
|
| 2912 |
+
"num_input_tokens_seen": 318069504,
|
| 2913 |
+
"step": 54450
|
| 2914 |
+
},
|
| 2915 |
+
{
|
| 2916 |
+
"epoch": 0.6016893585837159,
|
| 2917 |
+
"grad_norm": 1.9001188278198242,
|
| 2918 |
+
"learning_rate": 0.00012792218287611856,
|
| 2919 |
+
"loss": 4.4493,
|
| 2920 |
+
"num_input_tokens_seen": 318929088,
|
| 2921 |
+
"step": 54600
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 0.6033423513270777,
|
| 2925 |
+
"grad_norm": 1.7820019721984863,
|
| 2926 |
+
"learning_rate": 0.00012783402148199306,
|
| 2927 |
+
"loss": 4.4529,
|
| 2928 |
+
"num_input_tokens_seen": 319802400,
|
| 2929 |
+
"step": 54750
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 0.6049953440704395,
|
| 2933 |
+
"grad_norm": 1.8836514949798584,
|
| 2934 |
+
"learning_rate": 0.00012774586008786754,
|
| 2935 |
+
"loss": 4.4601,
|
| 2936 |
+
"num_input_tokens_seen": 320667168,
|
| 2937 |
+
"step": 54900
|
| 2938 |
+
},
|
| 2939 |
+
{
|
| 2940 |
+
"epoch": 0.6066483368138014,
|
| 2941 |
+
"grad_norm": 1.820078730583191,
|
| 2942 |
+
"learning_rate": 0.00012765828643636953,
|
| 2943 |
+
"loss": 4.4583,
|
| 2944 |
+
"num_input_tokens_seen": 321544160,
|
| 2945 |
+
"step": 55050
|
| 2946 |
+
},
|
| 2947 |
+
{
|
| 2948 |
+
"epoch": 0.6083013295571632,
|
| 2949 |
+
"grad_norm": 1.7549668550491333,
|
| 2950 |
+
"learning_rate": 0.000127570125042244,
|
| 2951 |
+
"loss": 4.4507,
|
| 2952 |
+
"num_input_tokens_seen": 322413504,
|
| 2953 |
+
"step": 55200
|
| 2954 |
+
},
|
| 2955 |
+
{
|
| 2956 |
+
"epoch": 0.6099543223005252,
|
| 2957 |
+
"grad_norm": 1.819643497467041,
|
| 2958 |
+
"learning_rate": 0.0001274819636481185,
|
| 2959 |
+
"loss": 4.4572,
|
| 2960 |
+
"num_input_tokens_seen": 323301120,
|
| 2961 |
+
"step": 55350
|
| 2962 |
+
},
|
| 2963 |
+
{
|
| 2964 |
+
"epoch": 0.611607315043887,
|
| 2965 |
+
"grad_norm": 1.8832948207855225,
|
| 2966 |
+
"learning_rate": 0.00012739380225399298,
|
| 2967 |
+
"loss": 4.4499,
|
| 2968 |
+
"num_input_tokens_seen": 324179040,
|
| 2969 |
+
"step": 55500
|
| 2970 |
+
},
|
| 2971 |
+
{
|
| 2972 |
+
"epoch": 0.6132603077872488,
|
| 2973 |
+
"grad_norm": 1.9329428672790527,
|
| 2974 |
+
"learning_rate": 0.00012730564085986748,
|
| 2975 |
+
"loss": 4.4598,
|
| 2976 |
+
"num_input_tokens_seen": 325044704,
|
| 2977 |
+
"step": 55650
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 0.6149133005306107,
|
| 2981 |
+
"grad_norm": 1.8368948698043823,
|
| 2982 |
+
"learning_rate": 0.00012721747946574196,
|
| 2983 |
+
"loss": 4.4636,
|
| 2984 |
+
"num_input_tokens_seen": 325903360,
|
| 2985 |
+
"step": 55800
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 0.6165662932739725,
|
| 2989 |
+
"grad_norm": 1.8610767126083374,
|
| 2990 |
+
"learning_rate": 0.00012712931807161646,
|
| 2991 |
+
"loss": 4.4602,
|
| 2992 |
+
"num_input_tokens_seen": 326791648,
|
| 2993 |
+
"step": 55950
|
| 2994 |
+
},
|
| 2995 |
+
{
|
| 2996 |
+
"epoch": 0.6182192860173343,
|
| 2997 |
+
"grad_norm": 1.853129506111145,
|
| 2998 |
+
"learning_rate": 0.00012704115667749093,
|
| 2999 |
+
"loss": 4.4581,
|
| 3000 |
+
"num_input_tokens_seen": 327669216,
|
| 3001 |
+
"step": 56100
|
| 3002 |
+
},
|
| 3003 |
+
{
|
| 3004 |
+
"epoch": 0.6198722787606963,
|
| 3005 |
+
"grad_norm": 1.8630894422531128,
|
| 3006 |
+
"learning_rate": 0.0001269529952833654,
|
| 3007 |
+
"loss": 4.4584,
|
| 3008 |
+
"num_input_tokens_seen": 328547424,
|
| 3009 |
+
"step": 56250
|
| 3010 |
+
},
|
| 3011 |
+
{
|
| 3012 |
+
"epoch": 0.6215252715040581,
|
| 3013 |
+
"grad_norm": 1.8581258058547974,
|
| 3014 |
+
"learning_rate": 0.0001268648338892399,
|
| 3015 |
+
"loss": 4.4632,
|
| 3016 |
+
"num_input_tokens_seen": 329414400,
|
| 3017 |
+
"step": 56400
|
| 3018 |
+
},
|
| 3019 |
+
{
|
| 3020 |
+
"epoch": 0.62317826424742,
|
| 3021 |
+
"grad_norm": 1.8294817209243774,
|
| 3022 |
+
"learning_rate": 0.00012677667249511439,
|
| 3023 |
+
"loss": 4.4634,
|
| 3024 |
+
"num_input_tokens_seen": 330297376,
|
| 3025 |
+
"step": 56550
|
| 3026 |
+
},
|
| 3027 |
+
{
|
| 3028 |
+
"epoch": 0.6248312569907818,
|
| 3029 |
+
"grad_norm": 1.9625203609466553,
|
| 3030 |
+
"learning_rate": 0.0001266885111009889,
|
| 3031 |
+
"loss": 4.4598,
|
| 3032 |
+
"num_input_tokens_seen": 331181792,
|
| 3033 |
+
"step": 56700
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 0.6264842497341436,
|
| 3037 |
+
"grad_norm": 1.821718454360962,
|
| 3038 |
+
"learning_rate": 0.00012660034970686336,
|
| 3039 |
+
"loss": 4.4549,
|
| 3040 |
+
"num_input_tokens_seen": 332056416,
|
| 3041 |
+
"step": 56850
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 0.6281372424775056,
|
| 3045 |
+
"grad_norm": 1.8366010189056396,
|
| 3046 |
+
"learning_rate": 0.00012651218831273786,
|
| 3047 |
+
"loss": 4.4418,
|
| 3048 |
+
"num_input_tokens_seen": 332908864,
|
| 3049 |
+
"step": 57000
|
| 3050 |
+
},
|
| 3051 |
+
{
|
| 3052 |
+
"epoch": 0.6297902352208674,
|
| 3053 |
+
"grad_norm": 1.858789086341858,
|
| 3054 |
+
"learning_rate": 0.00012642402691861234,
|
| 3055 |
+
"loss": 4.4592,
|
| 3056 |
+
"num_input_tokens_seen": 333791520,
|
| 3057 |
+
"step": 57150
|
| 3058 |
+
},
|
| 3059 |
+
{
|
| 3060 |
+
"epoch": 0.6314432279642292,
|
| 3061 |
+
"grad_norm": 1.9188382625579834,
|
| 3062 |
+
"learning_rate": 0.00012633586552448684,
|
| 3063 |
+
"loss": 4.4531,
|
| 3064 |
+
"num_input_tokens_seen": 334675488,
|
| 3065 |
+
"step": 57300
|
| 3066 |
+
},
|
| 3067 |
+
{
|
| 3068 |
+
"epoch": 0.6330962207075911,
|
| 3069 |
+
"grad_norm": 1.8480638265609741,
|
| 3070 |
+
"learning_rate": 0.00012624770413036132,
|
| 3071 |
+
"loss": 4.4557,
|
| 3072 |
+
"num_input_tokens_seen": 335569664,
|
| 3073 |
+
"step": 57450
|
| 3074 |
+
},
|
| 3075 |
+
{
|
| 3076 |
+
"epoch": 0.6347492134509529,
|
| 3077 |
+
"grad_norm": 1.8409630060195923,
|
| 3078 |
+
"learning_rate": 0.00012615954273623582,
|
| 3079 |
+
"loss": 4.454,
|
| 3080 |
+
"num_input_tokens_seen": 336438752,
|
| 3081 |
+
"step": 57600
|
| 3082 |
+
},
|
| 3083 |
+
{
|
| 3084 |
+
"epoch": 0.6364022061943148,
|
| 3085 |
+
"grad_norm": 1.7734564542770386,
|
| 3086 |
+
"learning_rate": 0.0001260713813421103,
|
| 3087 |
+
"loss": 4.4587,
|
| 3088 |
+
"num_input_tokens_seen": 337318464,
|
| 3089 |
+
"step": 57750
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 0.6380551989376767,
|
| 3093 |
+
"grad_norm": 1.8258734941482544,
|
| 3094 |
+
"learning_rate": 0.0001259832199479848,
|
| 3095 |
+
"loss": 4.4501,
|
| 3096 |
+
"num_input_tokens_seen": 338196416,
|
| 3097 |
+
"step": 57900
|
| 3098 |
+
},
|
| 3099 |
+
{
|
| 3100 |
+
"epoch": 0.6397081916810385,
|
| 3101 |
+
"grad_norm": 1.8730100393295288,
|
| 3102 |
+
"learning_rate": 0.00012589505855385927,
|
| 3103 |
+
"loss": 4.4508,
|
| 3104 |
+
"num_input_tokens_seen": 339066272,
|
| 3105 |
+
"step": 58050
|
| 3106 |
+
},
|
| 3107 |
+
{
|
| 3108 |
+
"epoch": 0.6413611844244004,
|
| 3109 |
+
"grad_norm": 1.7968626022338867,
|
| 3110 |
+
"learning_rate": 0.00012580689715973375,
|
| 3111 |
+
"loss": 4.4465,
|
| 3112 |
+
"num_input_tokens_seen": 339940576,
|
| 3113 |
+
"step": 58200
|
| 3114 |
+
},
|
| 3115 |
+
{
|
| 3116 |
+
"epoch": 0.6430141771677622,
|
| 3117 |
+
"grad_norm": 1.8305721282958984,
|
| 3118 |
+
"learning_rate": 0.00012571873576560825,
|
| 3119 |
+
"loss": 4.4452,
|
| 3120 |
+
"num_input_tokens_seen": 340827872,
|
| 3121 |
+
"step": 58350
|
| 3122 |
+
},
|
| 3123 |
+
{
|
| 3124 |
+
"epoch": 0.644667169911124,
|
| 3125 |
+
"grad_norm": 1.8106398582458496,
|
| 3126 |
+
"learning_rate": 0.00012563057437148272,
|
| 3127 |
+
"loss": 4.4436,
|
| 3128 |
+
"num_input_tokens_seen": 341710720,
|
| 3129 |
+
"step": 58500
|
| 3130 |
+
},
|
| 3131 |
+
{
|
| 3132 |
+
"epoch": 0.646320162654486,
|
| 3133 |
+
"grad_norm": 1.8428856134414673,
|
| 3134 |
+
"learning_rate": 0.00012554300071998474,
|
| 3135 |
+
"loss": 4.4607,
|
| 3136 |
+
"num_input_tokens_seen": 342592992,
|
| 3137 |
+
"step": 58650
|
| 3138 |
+
},
|
| 3139 |
+
{
|
| 3140 |
+
"epoch": 0.6479731553978478,
|
| 3141 |
+
"grad_norm": 1.89970064163208,
|
| 3142 |
+
"learning_rate": 0.00012545483932585922,
|
| 3143 |
+
"loss": 4.446,
|
| 3144 |
+
"num_input_tokens_seen": 343458496,
|
| 3145 |
+
"step": 58800
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 0.6496261481412097,
|
| 3149 |
+
"grad_norm": 1.887024998664856,
|
| 3150 |
+
"learning_rate": 0.00012536667793173372,
|
| 3151 |
+
"loss": 4.4438,
|
| 3152 |
+
"num_input_tokens_seen": 344324192,
|
| 3153 |
+
"step": 58950
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 0.6512791408845715,
|
| 3157 |
+
"grad_norm": 1.751080870628357,
|
| 3158 |
+
"learning_rate": 0.0001252785165376082,
|
| 3159 |
+
"loss": 4.4439,
|
| 3160 |
+
"num_input_tokens_seen": 345204320,
|
| 3161 |
+
"step": 59100
|
| 3162 |
+
},
|
| 3163 |
+
{
|
| 3164 |
+
"epoch": 0.6529321336279333,
|
| 3165 |
+
"grad_norm": 1.8455328941345215,
|
| 3166 |
+
"learning_rate": 0.0001251903551434827,
|
| 3167 |
+
"loss": 4.4446,
|
| 3168 |
+
"num_input_tokens_seen": 346100960,
|
| 3169 |
+
"step": 59250
|
| 3170 |
+
},
|
| 3171 |
+
{
|
| 3172 |
+
"epoch": 0.6545851263712952,
|
| 3173 |
+
"grad_norm": 1.9079509973526,
|
| 3174 |
+
"learning_rate": 0.00012510219374935717,
|
| 3175 |
+
"loss": 4.4441,
|
| 3176 |
+
"num_input_tokens_seen": 346981376,
|
| 3177 |
+
"step": 59400
|
| 3178 |
+
},
|
| 3179 |
+
{
|
| 3180 |
+
"epoch": 0.6562381191146571,
|
| 3181 |
+
"grad_norm": 1.8034120798110962,
|
| 3182 |
+
"learning_rate": 0.00012501403235523167,
|
| 3183 |
+
"loss": 4.4555,
|
| 3184 |
+
"num_input_tokens_seen": 347855584,
|
| 3185 |
+
"step": 59550
|
| 3186 |
+
},
|
| 3187 |
+
{
|
| 3188 |
+
"epoch": 0.657891111858019,
|
| 3189 |
+
"grad_norm": 1.7936707735061646,
|
| 3190 |
+
"learning_rate": 0.00012492587096110615,
|
| 3191 |
+
"loss": 4.4493,
|
| 3192 |
+
"num_input_tokens_seen": 348733664,
|
| 3193 |
+
"step": 59700
|
| 3194 |
+
},
|
| 3195 |
+
{
|
| 3196 |
+
"epoch": 0.6595441046013808,
|
| 3197 |
+
"grad_norm": 1.80596923828125,
|
| 3198 |
+
"learning_rate": 0.00012483770956698065,
|
| 3199 |
+
"loss": 4.4445,
|
| 3200 |
+
"num_input_tokens_seen": 349592128,
|
| 3201 |
+
"step": 59850
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 0.6611970973447426,
|
| 3205 |
+
"grad_norm": 1.7837984561920166,
|
| 3206 |
+
"learning_rate": 0.00012474954817285512,
|
| 3207 |
+
"loss": 4.4472,
|
| 3208 |
+
"num_input_tokens_seen": 350464672,
|
| 3209 |
+
"step": 60000
|
| 3210 |
+
},
|
| 3211 |
+
{
|
| 3212 |
+
"epoch": 0.6628500900881045,
|
| 3213 |
+
"grad_norm": 1.8550629615783691,
|
| 3214 |
+
"learning_rate": 0.0001246613867787296,
|
| 3215 |
+
"loss": 4.4436,
|
| 3216 |
+
"num_input_tokens_seen": 351360576,
|
| 3217 |
+
"step": 60150
|
| 3218 |
+
},
|
| 3219 |
+
{
|
| 3220 |
+
"epoch": 0.6645030828314664,
|
| 3221 |
+
"grad_norm": 1.8099464178085327,
|
| 3222 |
+
"learning_rate": 0.0001245738131272316,
|
| 3223 |
+
"loss": 4.4439,
|
| 3224 |
+
"num_input_tokens_seen": 352250944,
|
| 3225 |
+
"step": 60300
|
| 3226 |
+
},
|
| 3227 |
+
{
|
| 3228 |
+
"epoch": 0.6661560755748283,
|
| 3229 |
+
"grad_norm": 1.869233250617981,
|
| 3230 |
+
"learning_rate": 0.0001244856517331061,
|
| 3231 |
+
"loss": 4.45,
|
| 3232 |
+
"num_input_tokens_seen": 353120608,
|
| 3233 |
+
"step": 60450
|
| 3234 |
+
},
|
| 3235 |
+
{
|
| 3236 |
+
"epoch": 0.6678090683181901,
|
| 3237 |
+
"grad_norm": 1.8628960847854614,
|
| 3238 |
+
"learning_rate": 0.00012439749033898057,
|
| 3239 |
+
"loss": 4.4446,
|
| 3240 |
+
"num_input_tokens_seen": 353986944,
|
| 3241 |
+
"step": 60600
|
| 3242 |
+
},
|
| 3243 |
+
{
|
| 3244 |
+
"epoch": 0.6694620610615519,
|
| 3245 |
+
"grad_norm": 1.7935791015625,
|
| 3246 |
+
"learning_rate": 0.00012430932894485504,
|
| 3247 |
+
"loss": 4.4481,
|
| 3248 |
+
"num_input_tokens_seen": 354851456,
|
| 3249 |
+
"step": 60750
|
| 3250 |
+
},
|
| 3251 |
+
{
|
| 3252 |
+
"epoch": 0.6711150538049138,
|
| 3253 |
+
"grad_norm": 1.919735074043274,
|
| 3254 |
+
"learning_rate": 0.00012422116755072955,
|
| 3255 |
+
"loss": 4.4491,
|
| 3256 |
+
"num_input_tokens_seen": 355735616,
|
| 3257 |
+
"step": 60900
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 0.6727680465482756,
|
| 3261 |
+
"grad_norm": 1.9296785593032837,
|
| 3262 |
+
"learning_rate": 0.00012413300615660402,
|
| 3263 |
+
"loss": 4.4384,
|
| 3264 |
+
"num_input_tokens_seen": 356613408,
|
| 3265 |
+
"step": 61050
|
| 3266 |
+
},
|
| 3267 |
+
{
|
| 3268 |
+
"epoch": 0.6744210392916375,
|
| 3269 |
+
"grad_norm": 1.8167061805725098,
|
| 3270 |
+
"learning_rate": 0.00012404484476247852,
|
| 3271 |
+
"loss": 4.4326,
|
| 3272 |
+
"num_input_tokens_seen": 357463840,
|
| 3273 |
+
"step": 61200
|
| 3274 |
+
},
|
| 3275 |
+
{
|
| 3276 |
+
"epoch": 0.6760740320349994,
|
| 3277 |
+
"grad_norm": 1.86695396900177,
|
| 3278 |
+
"learning_rate": 0.000123956683368353,
|
| 3279 |
+
"loss": 4.4501,
|
| 3280 |
+
"num_input_tokens_seen": 358354464,
|
| 3281 |
+
"step": 61350
|
| 3282 |
+
},
|
| 3283 |
+
{
|
| 3284 |
+
"epoch": 0.6777270247783612,
|
| 3285 |
+
"grad_norm": 1.8627629280090332,
|
| 3286 |
+
"learning_rate": 0.0001238685219742275,
|
| 3287 |
+
"loss": 4.4496,
|
| 3288 |
+
"num_input_tokens_seen": 359222016,
|
| 3289 |
+
"step": 61500
|
| 3290 |
+
},
|
| 3291 |
+
{
|
| 3292 |
+
"epoch": 0.6793800175217231,
|
| 3293 |
+
"grad_norm": 1.8496758937835693,
|
| 3294 |
+
"learning_rate": 0.00012378036058010197,
|
| 3295 |
+
"loss": 4.4505,
|
| 3296 |
+
"num_input_tokens_seen": 360112096,
|
| 3297 |
+
"step": 61650
|
| 3298 |
+
},
|
| 3299 |
+
{
|
| 3300 |
+
"epoch": 0.6810330102650849,
|
| 3301 |
+
"grad_norm": 1.8193156719207764,
|
| 3302 |
+
"learning_rate": 0.00012369219918597645,
|
| 3303 |
+
"loss": 4.452,
|
| 3304 |
+
"num_input_tokens_seen": 360995520,
|
| 3305 |
+
"step": 61800
|
| 3306 |
+
},
|
| 3307 |
+
{
|
| 3308 |
+
"epoch": 0.6826860030084468,
|
| 3309 |
+
"grad_norm": 1.7519707679748535,
|
| 3310 |
+
"learning_rate": 0.00012360403779185095,
|
| 3311 |
+
"loss": 4.4439,
|
| 3312 |
+
"num_input_tokens_seen": 361873184,
|
| 3313 |
+
"step": 61950
|
| 3314 |
+
},
|
| 3315 |
+
{
|
| 3316 |
+
"epoch": 0.6843389957518087,
|
| 3317 |
+
"grad_norm": 1.9227124452590942,
|
| 3318 |
+
"learning_rate": 0.00012351587639772543,
|
| 3319 |
+
"loss": 4.4416,
|
| 3320 |
+
"num_input_tokens_seen": 362749312,
|
| 3321 |
+
"step": 62100
|
| 3322 |
+
},
|
| 3323 |
+
{
|
| 3324 |
+
"epoch": 0.6859919884951705,
|
| 3325 |
+
"grad_norm": 1.8492848873138428,
|
| 3326 |
+
"learning_rate": 0.00012342771500359993,
|
| 3327 |
+
"loss": 4.4541,
|
| 3328 |
+
"num_input_tokens_seen": 363635936,
|
| 3329 |
+
"step": 62250
|
| 3330 |
+
},
|
| 3331 |
+
{
|
| 3332 |
+
"epoch": 0.6876449812385323,
|
| 3333 |
+
"grad_norm": 1.946057677268982,
|
| 3334 |
+
"learning_rate": 0.0001233395536094744,
|
| 3335 |
+
"loss": 4.435,
|
| 3336 |
+
"num_input_tokens_seen": 364500576,
|
| 3337 |
+
"step": 62400
|
| 3338 |
+
},
|
| 3339 |
+
{
|
| 3340 |
+
"epoch": 0.6892979739818942,
|
| 3341 |
+
"grad_norm": 1.8880736827850342,
|
| 3342 |
+
"learning_rate": 0.0001232513922153489,
|
| 3343 |
+
"loss": 4.4442,
|
| 3344 |
+
"num_input_tokens_seen": 365363744,
|
| 3345 |
+
"step": 62550
|
| 3346 |
+
},
|
| 3347 |
+
{
|
| 3348 |
+
"epoch": 0.690950966725256,
|
| 3349 |
+
"grad_norm": 1.864534854888916,
|
| 3350 |
+
"learning_rate": 0.00012316323082122338,
|
| 3351 |
+
"loss": 4.4398,
|
| 3352 |
+
"num_input_tokens_seen": 366253600,
|
| 3353 |
+
"step": 62700
|
| 3354 |
+
},
|
| 3355 |
+
{
|
| 3356 |
+
"epoch": 0.692603959468618,
|
| 3357 |
+
"grad_norm": 1.8077435493469238,
|
| 3358 |
+
"learning_rate": 0.00012307506942709788,
|
| 3359 |
+
"loss": 4.4462,
|
| 3360 |
+
"num_input_tokens_seen": 367119136,
|
| 3361 |
+
"step": 62850
|
| 3362 |
+
},
|
| 3363 |
+
{
|
| 3364 |
+
"epoch": 0.6942569522119798,
|
| 3365 |
+
"grad_norm": 1.8797168731689453,
|
| 3366 |
+
"learning_rate": 0.00012298690803297236,
|
| 3367 |
+
"loss": 4.4535,
|
| 3368 |
+
"num_input_tokens_seen": 367998656,
|
| 3369 |
+
"step": 63000
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 0.6959099449553416,
|
| 3373 |
+
"grad_norm": 1.9124201536178589,
|
| 3374 |
+
"learning_rate": 0.00012289874663884686,
|
| 3375 |
+
"loss": 4.4314,
|
| 3376 |
+
"num_input_tokens_seen": 368873888,
|
| 3377 |
+
"step": 63150
|
| 3378 |
+
},
|
| 3379 |
+
{
|
| 3380 |
+
"epoch": 0.6975629376987035,
|
| 3381 |
+
"grad_norm": 1.919708013534546,
|
| 3382 |
+
"learning_rate": 0.00012281058524472134,
|
| 3383 |
+
"loss": 4.4524,
|
| 3384 |
+
"num_input_tokens_seen": 369761216,
|
| 3385 |
+
"step": 63300
|
| 3386 |
+
},
|
| 3387 |
+
{
|
| 3388 |
+
"epoch": 0.6992159304420653,
|
| 3389 |
+
"grad_norm": 1.8248168230056763,
|
| 3390 |
+
"learning_rate": 0.00012272242385059584,
|
| 3391 |
+
"loss": 4.4422,
|
| 3392 |
+
"num_input_tokens_seen": 370638688,
|
| 3393 |
+
"step": 63450
|
| 3394 |
+
},
|
| 3395 |
+
{
|
| 3396 |
+
"epoch": 0.7008689231854273,
|
| 3397 |
+
"grad_norm": 1.810051441192627,
|
| 3398 |
+
"learning_rate": 0.0001226342624564703,
|
| 3399 |
+
"loss": 4.4313,
|
| 3400 |
+
"num_input_tokens_seen": 371541344,
|
| 3401 |
+
"step": 63600
|
| 3402 |
+
},
|
| 3403 |
+
{
|
| 3404 |
+
"epoch": 0.7025219159287891,
|
| 3405 |
+
"grad_norm": 1.8361635208129883,
|
| 3406 |
+
"learning_rate": 0.00012254610106234481,
|
| 3407 |
+
"loss": 4.436,
|
| 3408 |
+
"num_input_tokens_seen": 372415168,
|
| 3409 |
+
"step": 63750
|
| 3410 |
+
},
|
| 3411 |
+
{
|
| 3412 |
+
"epoch": 0.7041749086721509,
|
| 3413 |
+
"grad_norm": 1.8005433082580566,
|
| 3414 |
+
"learning_rate": 0.0001224579396682193,
|
| 3415 |
+
"loss": 4.4353,
|
| 3416 |
+
"num_input_tokens_seen": 373283872,
|
| 3417 |
+
"step": 63900
|
| 3418 |
+
},
|
| 3419 |
+
{
|
| 3420 |
+
"epoch": 0.7058279014155128,
|
| 3421 |
+
"grad_norm": 1.8291569948196411,
|
| 3422 |
+
"learning_rate": 0.0001223697782740938,
|
| 3423 |
+
"loss": 4.4486,
|
| 3424 |
+
"num_input_tokens_seen": 374156960,
|
| 3425 |
+
"step": 64050
|
| 3426 |
+
},
|
| 3427 |
+
{
|
| 3428 |
+
"epoch": 0.7074808941588746,
|
| 3429 |
+
"grad_norm": 1.6987590789794922,
|
| 3430 |
+
"learning_rate": 0.00012228161687996827,
|
| 3431 |
+
"loss": 4.4402,
|
| 3432 |
+
"num_input_tokens_seen": 375045632,
|
| 3433 |
+
"step": 64200
|
| 3434 |
+
},
|
| 3435 |
+
{
|
| 3436 |
+
"epoch": 0.7091338869022366,
|
| 3437 |
+
"grad_norm": 1.8456915616989136,
|
| 3438 |
+
"learning_rate": 0.00012219345548584277,
|
| 3439 |
+
"loss": 4.4502,
|
| 3440 |
+
"num_input_tokens_seen": 375936576,
|
| 3441 |
+
"step": 64350
|
| 3442 |
+
},
|
| 3443 |
+
{
|
| 3444 |
+
"epoch": 0.7107868796455984,
|
| 3445 |
+
"grad_norm": 1.9141839742660522,
|
| 3446 |
+
"learning_rate": 0.00012210529409171724,
|
| 3447 |
+
"loss": 4.4521,
|
| 3448 |
+
"num_input_tokens_seen": 376816512,
|
| 3449 |
+
"step": 64500
|
| 3450 |
+
},
|
| 3451 |
+
{
|
| 3452 |
+
"epoch": 0.7124398723889602,
|
| 3453 |
+
"grad_norm": 1.8822457790374756,
|
| 3454 |
+
"learning_rate": 0.00012201713269759175,
|
| 3455 |
+
"loss": 4.4347,
|
| 3456 |
+
"num_input_tokens_seen": 377684448,
|
| 3457 |
+
"step": 64650
|
| 3458 |
+
},
|
| 3459 |
+
{
|
| 3460 |
+
"epoch": 0.7140928651323221,
|
| 3461 |
+
"grad_norm": 1.8143234252929688,
|
| 3462 |
+
"learning_rate": 0.00012192897130346622,
|
| 3463 |
+
"loss": 4.4336,
|
| 3464 |
+
"num_input_tokens_seen": 378553120,
|
| 3465 |
+
"step": 64800
|
| 3466 |
+
},
|
| 3467 |
+
{
|
| 3468 |
+
"epoch": 0.7157458578756839,
|
| 3469 |
+
"grad_norm": 1.8877683877944946,
|
| 3470 |
+
"learning_rate": 0.00012184080990934071,
|
| 3471 |
+
"loss": 4.4425,
|
| 3472 |
+
"num_input_tokens_seen": 379413888,
|
| 3473 |
+
"step": 64950
|
| 3474 |
+
},
|
| 3475 |
+
{
|
| 3476 |
+
"epoch": 0.7173988506190457,
|
| 3477 |
+
"grad_norm": 1.8746610879898071,
|
| 3478 |
+
"learning_rate": 0.0001217526485152152,
|
| 3479 |
+
"loss": 4.4417,
|
| 3480 |
+
"num_input_tokens_seen": 380290304,
|
| 3481 |
+
"step": 65100
|
| 3482 |
+
},
|
| 3483 |
+
{
|
| 3484 |
+
"epoch": 0.7190518433624077,
|
| 3485 |
+
"grad_norm": 2.0395500659942627,
|
| 3486 |
+
"learning_rate": 0.00012166448712108969,
|
| 3487 |
+
"loss": 4.4423,
|
| 3488 |
+
"num_input_tokens_seen": 381185920,
|
| 3489 |
+
"step": 65250
|
| 3490 |
+
},
|
| 3491 |
+
{
|
| 3492 |
+
"epoch": 0.7207048361057695,
|
| 3493 |
+
"grad_norm": 1.992492914199829,
|
| 3494 |
+
"learning_rate": 0.00012157632572696418,
|
| 3495 |
+
"loss": 4.4484,
|
| 3496 |
+
"num_input_tokens_seen": 382075808,
|
| 3497 |
+
"step": 65400
|
| 3498 |
+
},
|
| 3499 |
+
{
|
| 3500 |
+
"epoch": 0.7223578288491314,
|
| 3501 |
+
"grad_norm": 1.8621459007263184,
|
| 3502 |
+
"learning_rate": 0.00012148816433283866,
|
| 3503 |
+
"loss": 4.4274,
|
| 3504 |
+
"num_input_tokens_seen": 382955232,
|
| 3505 |
+
"step": 65550
|
| 3506 |
+
},
|
| 3507 |
+
{
|
| 3508 |
+
"epoch": 0.7240108215924932,
|
| 3509 |
+
"grad_norm": 1.8787345886230469,
|
| 3510 |
+
"learning_rate": 0.00012140000293871315,
|
| 3511 |
+
"loss": 4.4378,
|
| 3512 |
+
"num_input_tokens_seen": 383834592,
|
| 3513 |
+
"step": 65700
|
| 3514 |
+
},
|
| 3515 |
+
{
|
| 3516 |
+
"epoch": 0.725663814335855,
|
| 3517 |
+
"grad_norm": 1.8640894889831543,
|
| 3518 |
+
"learning_rate": 0.00012131184154458764,
|
| 3519 |
+
"loss": 4.4557,
|
| 3520 |
+
"num_input_tokens_seen": 384710016,
|
| 3521 |
+
"step": 65850
|
| 3522 |
+
},
|
| 3523 |
+
{
|
| 3524 |
+
"epoch": 0.727316807079217,
|
| 3525 |
+
"grad_norm": 1.918143630027771,
|
| 3526 |
+
"learning_rate": 0.00012122368015046212,
|
| 3527 |
+
"loss": 4.4467,
|
| 3528 |
+
"num_input_tokens_seen": 385593120,
|
| 3529 |
+
"step": 66000
|
| 3530 |
+
},
|
| 3531 |
+
{
|
| 3532 |
+
"epoch": 0.7289697998225788,
|
| 3533 |
+
"grad_norm": 1.8295505046844482,
|
| 3534 |
+
"learning_rate": 0.00012113551875633662,
|
| 3535 |
+
"loss": 4.4257,
|
| 3536 |
+
"num_input_tokens_seen": 386460160,
|
| 3537 |
+
"step": 66150
|
| 3538 |
+
},
|
| 3539 |
+
{
|
| 3540 |
+
"epoch": 0.7306227925659406,
|
| 3541 |
+
"grad_norm": 1.880216360092163,
|
| 3542 |
+
"learning_rate": 0.00012104794510483861,
|
| 3543 |
+
"loss": 4.4312,
|
| 3544 |
+
"num_input_tokens_seen": 387328000,
|
| 3545 |
+
"step": 66300
|
| 3546 |
+
},
|
| 3547 |
+
{
|
| 3548 |
+
"epoch": 0.7322757853093025,
|
| 3549 |
+
"grad_norm": 1.818788766860962,
|
| 3550 |
+
"learning_rate": 0.00012095978371071308,
|
| 3551 |
+
"loss": 4.4402,
|
| 3552 |
+
"num_input_tokens_seen": 388200192,
|
| 3553 |
+
"step": 66450
|
| 3554 |
+
},
|
| 3555 |
+
{
|
| 3556 |
+
"epoch": 0.7339287780526643,
|
| 3557 |
+
"grad_norm": 1.82147216796875,
|
| 3558 |
+
"learning_rate": 0.00012087162231658759,
|
| 3559 |
+
"loss": 4.4468,
|
| 3560 |
+
"num_input_tokens_seen": 389079264,
|
| 3561 |
+
"step": 66600
|
| 3562 |
+
},
|
| 3563 |
+
{
|
| 3564 |
+
"epoch": 0.7355817707960262,
|
| 3565 |
+
"grad_norm": 1.8930702209472656,
|
| 3566 |
+
"learning_rate": 0.00012078346092246206,
|
| 3567 |
+
"loss": 4.4362,
|
| 3568 |
+
"num_input_tokens_seen": 389974112,
|
| 3569 |
+
"step": 66750
|
| 3570 |
+
},
|
| 3571 |
+
{
|
| 3572 |
+
"epoch": 0.7372347635393881,
|
| 3573 |
+
"grad_norm": 1.8484946489334106,
|
| 3574 |
+
"learning_rate": 0.00012069529952833656,
|
| 3575 |
+
"loss": 4.4409,
|
| 3576 |
+
"num_input_tokens_seen": 390875168,
|
| 3577 |
+
"step": 66900
|
| 3578 |
+
},
|
| 3579 |
+
{
|
| 3580 |
+
"epoch": 0.7388877562827499,
|
| 3581 |
+
"grad_norm": 1.894093632698059,
|
| 3582 |
+
"learning_rate": 0.00012060713813421104,
|
| 3583 |
+
"loss": 4.4197,
|
| 3584 |
+
"num_input_tokens_seen": 391736384,
|
| 3585 |
+
"step": 67050
|
| 3586 |
+
},
|
| 3587 |
+
{
|
| 3588 |
+
"epoch": 0.7405407490261118,
|
| 3589 |
+
"grad_norm": 1.918149471282959,
|
| 3590 |
+
"learning_rate": 0.00012051897674008553,
|
| 3591 |
+
"loss": 4.4487,
|
| 3592 |
+
"num_input_tokens_seen": 392626688,
|
| 3593 |
+
"step": 67200
|
| 3594 |
+
},
|
| 3595 |
+
{
|
| 3596 |
+
"epoch": 0.7421937417694736,
|
| 3597 |
+
"grad_norm": 1.8563427925109863,
|
| 3598 |
+
"learning_rate": 0.00012043081534596002,
|
| 3599 |
+
"loss": 4.4375,
|
| 3600 |
+
"num_input_tokens_seen": 393508640,
|
| 3601 |
+
"step": 67350
|
| 3602 |
+
},
|
| 3603 |
+
{
|
| 3604 |
+
"epoch": 0.7438467345128355,
|
| 3605 |
+
"grad_norm": 1.8275529146194458,
|
| 3606 |
+
"learning_rate": 0.0001203426539518345,
|
| 3607 |
+
"loss": 4.4333,
|
| 3608 |
+
"num_input_tokens_seen": 394397120,
|
| 3609 |
+
"step": 67500
|
| 3610 |
+
},
|
| 3611 |
+
{
|
| 3612 |
+
"epoch": 0.7454997272561974,
|
| 3613 |
+
"grad_norm": 1.824823260307312,
|
| 3614 |
+
"learning_rate": 0.00012025449255770899,
|
| 3615 |
+
"loss": 4.4213,
|
| 3616 |
+
"num_input_tokens_seen": 395273408,
|
| 3617 |
+
"step": 67650
|
| 3618 |
+
},
|
| 3619 |
+
{
|
| 3620 |
+
"epoch": 0.7471527199995592,
|
| 3621 |
+
"grad_norm": 1.7815489768981934,
|
| 3622 |
+
"learning_rate": 0.00012016633116358348,
|
| 3623 |
+
"loss": 4.4295,
|
| 3624 |
+
"num_input_tokens_seen": 396151584,
|
| 3625 |
+
"step": 67800
|
| 3626 |
+
},
|
| 3627 |
+
{
|
| 3628 |
+
"epoch": 0.7488057127429211,
|
| 3629 |
+
"grad_norm": 1.9288073778152466,
|
| 3630 |
+
"learning_rate": 0.00012007816976945797,
|
| 3631 |
+
"loss": 4.4348,
|
| 3632 |
+
"num_input_tokens_seen": 397006880,
|
| 3633 |
+
"step": 67950
|
| 3634 |
+
},
|
| 3635 |
+
{
|
| 3636 |
+
"epoch": 0.7504587054862829,
|
| 3637 |
+
"grad_norm": 1.866746425628662,
|
| 3638 |
+
"learning_rate": 0.00011999000837533245,
|
| 3639 |
+
"loss": 4.4306,
|
| 3640 |
+
"num_input_tokens_seen": 397879072,
|
| 3641 |
+
"step": 68100
|
| 3642 |
+
},
|
| 3643 |
+
{
|
| 3644 |
+
"epoch": 0.7521116982296447,
|
| 3645 |
+
"grad_norm": 1.8168858289718628,
|
| 3646 |
+
"learning_rate": 0.00011990184698120693,
|
| 3647 |
+
"loss": 4.4321,
|
| 3648 |
+
"num_input_tokens_seen": 398772672,
|
| 3649 |
+
"step": 68250
|
| 3650 |
+
},
|
| 3651 |
+
{
|
| 3652 |
+
"epoch": 0.7537646909730066,
|
| 3653 |
+
"grad_norm": 1.7801350355148315,
|
| 3654 |
+
"learning_rate": 0.00011981368558708142,
|
| 3655 |
+
"loss": 4.4358,
|
| 3656 |
+
"num_input_tokens_seen": 399663136,
|
| 3657 |
+
"step": 68400
|
| 3658 |
+
},
|
| 3659 |
+
{
|
| 3660 |
+
"epoch": 0.7554176837163685,
|
| 3661 |
+
"grad_norm": 1.9442716836929321,
|
| 3662 |
+
"learning_rate": 0.00011972611193558343,
|
| 3663 |
+
"loss": 4.4357,
|
| 3664 |
+
"num_input_tokens_seen": 400543936,
|
| 3665 |
+
"step": 68550
|
| 3666 |
+
},
|
| 3667 |
+
{
|
| 3668 |
+
"epoch": 0.7570706764597304,
|
| 3669 |
+
"grad_norm": 1.8754234313964844,
|
| 3670 |
+
"learning_rate": 0.0001196379505414579,
|
| 3671 |
+
"loss": 4.4279,
|
| 3672 |
+
"num_input_tokens_seen": 401411136,
|
| 3673 |
+
"step": 68700
|
| 3674 |
+
},
|
| 3675 |
+
{
|
| 3676 |
+
"epoch": 0.7587236692030922,
|
| 3677 |
+
"grad_norm": 1.8986996412277222,
|
| 3678 |
+
"learning_rate": 0.0001195497891473324,
|
| 3679 |
+
"loss": 4.4345,
|
| 3680 |
+
"num_input_tokens_seen": 402290464,
|
| 3681 |
+
"step": 68850
|
| 3682 |
+
},
|
| 3683 |
+
{
|
| 3684 |
+
"epoch": 0.760376661946454,
|
| 3685 |
+
"grad_norm": 1.8807158470153809,
|
| 3686 |
+
"learning_rate": 0.00011946162775320688,
|
| 3687 |
+
"loss": 4.4329,
|
| 3688 |
+
"num_input_tokens_seen": 403176768,
|
| 3689 |
+
"step": 69000
|
| 3690 |
+
},
|
| 3691 |
+
{
|
| 3692 |
+
"epoch": 0.7620296546898159,
|
| 3693 |
+
"grad_norm": 1.8661843538284302,
|
| 3694 |
+
"learning_rate": 0.00011937346635908138,
|
| 3695 |
+
"loss": 4.4327,
|
| 3696 |
+
"num_input_tokens_seen": 404053888,
|
| 3697 |
+
"step": 69150
|
| 3698 |
+
},
|
| 3699 |
+
{
|
| 3700 |
+
"epoch": 0.7636826474331778,
|
| 3701 |
+
"grad_norm": 1.9022386074066162,
|
| 3702 |
+
"learning_rate": 0.00011928530496495586,
|
| 3703 |
+
"loss": 4.4304,
|
| 3704 |
+
"num_input_tokens_seen": 404951328,
|
| 3705 |
+
"step": 69300
|
| 3706 |
+
},
|
| 3707 |
+
{
|
| 3708 |
+
"epoch": 0.7653356401765397,
|
| 3709 |
+
"grad_norm": 1.9497708082199097,
|
| 3710 |
+
"learning_rate": 0.00011919714357083035,
|
| 3711 |
+
"loss": 4.4319,
|
| 3712 |
+
"num_input_tokens_seen": 405824128,
|
| 3713 |
+
"step": 69450
|
| 3714 |
+
},
|
| 3715 |
+
{
|
| 3716 |
+
"epoch": 0.7669886329199015,
|
| 3717 |
+
"grad_norm": 1.7283419370651245,
|
| 3718 |
+
"learning_rate": 0.00011910898217670483,
|
| 3719 |
+
"loss": 4.4222,
|
| 3720 |
+
"num_input_tokens_seen": 406694592,
|
| 3721 |
+
"step": 69600
|
| 3722 |
+
},
|
| 3723 |
+
{
|
| 3724 |
+
"epoch": 0.7686416256632633,
|
| 3725 |
+
"grad_norm": 1.8692352771759033,
|
| 3726 |
+
"learning_rate": 0.00011902082078257931,
|
| 3727 |
+
"loss": 4.4257,
|
| 3728 |
+
"num_input_tokens_seen": 407577856,
|
| 3729 |
+
"step": 69750
|
| 3730 |
+
},
|
| 3731 |
+
{
|
| 3732 |
+
"epoch": 0.7702946184066252,
|
| 3733 |
+
"grad_norm": 1.918215036392212,
|
| 3734 |
+
"learning_rate": 0.00011893265938845381,
|
| 3735 |
+
"loss": 4.4275,
|
| 3736 |
+
"num_input_tokens_seen": 408455424,
|
| 3737 |
+
"step": 69900
|
| 3738 |
+
},
|
| 3739 |
+
{
|
| 3740 |
+
"epoch": 0.771947611149987,
|
| 3741 |
+
"grad_norm": 1.8184279203414917,
|
| 3742 |
+
"learning_rate": 0.00011884449799432829,
|
| 3743 |
+
"loss": 4.4287,
|
| 3744 |
+
"num_input_tokens_seen": 409330752,
|
| 3745 |
+
"step": 70050
|
| 3746 |
+
},
|
| 3747 |
+
{
|
| 3748 |
+
"epoch": 0.773600603893349,
|
| 3749 |
+
"grad_norm": 1.846740961074829,
|
| 3750 |
+
"learning_rate": 0.00011875633660020279,
|
| 3751 |
+
"loss": 4.4244,
|
| 3752 |
+
"num_input_tokens_seen": 410215104,
|
| 3753 |
+
"step": 70200
|
| 3754 |
+
},
|
| 3755 |
+
{
|
| 3756 |
+
"epoch": 0.7752535966367108,
|
| 3757 |
+
"grad_norm": 1.9468152523040771,
|
| 3758 |
+
"learning_rate": 0.00011866817520607726,
|
| 3759 |
+
"loss": 4.4223,
|
| 3760 |
+
"num_input_tokens_seen": 411089696,
|
| 3761 |
+
"step": 70350
|
| 3762 |
+
},
|
| 3763 |
+
{
|
| 3764 |
+
"epoch": 0.7769065893800726,
|
| 3765 |
+
"grad_norm": 1.87180495262146,
|
| 3766 |
+
"learning_rate": 0.00011858001381195175,
|
| 3767 |
+
"loss": 4.4511,
|
| 3768 |
+
"num_input_tokens_seen": 411988000,
|
| 3769 |
+
"step": 70500
|
| 3770 |
+
},
|
| 3771 |
+
{
|
| 3772 |
+
"epoch": 0.7785595821234345,
|
| 3773 |
+
"grad_norm": 1.8375773429870605,
|
| 3774 |
+
"learning_rate": 0.00011849185241782624,
|
| 3775 |
+
"loss": 4.4361,
|
| 3776 |
+
"num_input_tokens_seen": 412876544,
|
| 3777 |
+
"step": 70650
|
| 3778 |
+
},
|
| 3779 |
+
{
|
| 3780 |
+
"epoch": 0.7802125748667963,
|
| 3781 |
+
"grad_norm": 1.7592004537582397,
|
| 3782 |
+
"learning_rate": 0.00011840369102370073,
|
| 3783 |
+
"loss": 4.4272,
|
| 3784 |
+
"num_input_tokens_seen": 413737824,
|
| 3785 |
+
"step": 70800
|
| 3786 |
+
},
|
| 3787 |
+
{
|
| 3788 |
+
"epoch": 0.7818655676101582,
|
| 3789 |
+
"grad_norm": 1.9243676662445068,
|
| 3790 |
+
"learning_rate": 0.00011831552962957522,
|
| 3791 |
+
"loss": 4.4343,
|
| 3792 |
+
"num_input_tokens_seen": 414613184,
|
| 3793 |
+
"step": 70950
|
| 3794 |
+
},
|
| 3795 |
+
{
|
| 3796 |
+
"epoch": 0.7835185603535201,
|
| 3797 |
+
"grad_norm": 1.9014674425125122,
|
| 3798 |
+
"learning_rate": 0.0001182273682354497,
|
| 3799 |
+
"loss": 4.4217,
|
| 3800 |
+
"num_input_tokens_seen": 415494144,
|
| 3801 |
+
"step": 71100
|
| 3802 |
+
},
|
| 3803 |
+
{
|
| 3804 |
+
"epoch": 0.7851715530968819,
|
| 3805 |
+
"grad_norm": 1.8528156280517578,
|
| 3806 |
+
"learning_rate": 0.0001181397945839517,
|
| 3807 |
+
"loss": 4.4323,
|
| 3808 |
+
"num_input_tokens_seen": 416354240,
|
| 3809 |
+
"step": 71250
|
| 3810 |
+
},
|
| 3811 |
+
{
|
| 3812 |
+
"epoch": 0.7868245458402437,
|
| 3813 |
+
"grad_norm": 1.7702356576919556,
|
| 3814 |
+
"learning_rate": 0.0001180516331898262,
|
| 3815 |
+
"loss": 4.4335,
|
| 3816 |
+
"num_input_tokens_seen": 417214944,
|
| 3817 |
+
"step": 71400
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 0.7884775385836056,
|
| 3821 |
+
"grad_norm": 1.893778920173645,
|
| 3822 |
+
"learning_rate": 0.00011796347179570067,
|
| 3823 |
+
"loss": 4.4288,
|
| 3824 |
+
"num_input_tokens_seen": 418073984,
|
| 3825 |
+
"step": 71550
|
| 3826 |
+
},
|
| 3827 |
+
{
|
| 3828 |
+
"epoch": 0.7901305313269674,
|
| 3829 |
+
"grad_norm": 1.8179432153701782,
|
| 3830 |
+
"learning_rate": 0.00011787531040157515,
|
| 3831 |
+
"loss": 4.4161,
|
| 3832 |
+
"num_input_tokens_seen": 418956256,
|
| 3833 |
+
"step": 71700
|
| 3834 |
+
},
|
| 3835 |
+
{
|
| 3836 |
+
"epoch": 0.7917835240703294,
|
| 3837 |
+
"grad_norm": 1.8786159753799438,
|
| 3838 |
+
"learning_rate": 0.00011778714900744965,
|
| 3839 |
+
"loss": 4.4224,
|
| 3840 |
+
"num_input_tokens_seen": 419832544,
|
| 3841 |
+
"step": 71850
|
| 3842 |
+
},
|
| 3843 |
+
{
|
| 3844 |
+
"epoch": 0.7934365168136912,
|
| 3845 |
+
"grad_norm": 1.864493727684021,
|
| 3846 |
+
"learning_rate": 0.00011769898761332413,
|
| 3847 |
+
"loss": 4.4294,
|
| 3848 |
+
"num_input_tokens_seen": 420706176,
|
| 3849 |
+
"step": 72000
|
| 3850 |
+
},
|
| 3851 |
+
{
|
| 3852 |
+
"epoch": 0.795089509557053,
|
| 3853 |
+
"grad_norm": 1.7827798128128052,
|
| 3854 |
+
"learning_rate": 0.00011761082621919863,
|
| 3855 |
+
"loss": 4.4265,
|
| 3856 |
+
"num_input_tokens_seen": 421594880,
|
| 3857 |
+
"step": 72150
|
| 3858 |
+
},
|
| 3859 |
+
{
|
| 3860 |
+
"epoch": 0.7967425023004149,
|
| 3861 |
+
"grad_norm": 1.8714325428009033,
|
| 3862 |
+
"learning_rate": 0.0001175226648250731,
|
| 3863 |
+
"loss": 4.4428,
|
| 3864 |
+
"num_input_tokens_seen": 422456256,
|
| 3865 |
+
"step": 72300
|
| 3866 |
+
},
|
| 3867 |
+
{
|
| 3868 |
+
"epoch": 0.7983954950437767,
|
| 3869 |
+
"grad_norm": 1.8954764604568481,
|
| 3870 |
+
"learning_rate": 0.0001174345034309476,
|
| 3871 |
+
"loss": 4.4198,
|
| 3872 |
+
"num_input_tokens_seen": 423334208,
|
| 3873 |
+
"step": 72450
|
| 3874 |
+
},
|
| 3875 |
+
{
|
| 3876 |
+
"epoch": 0.8000484877871387,
|
| 3877 |
+
"grad_norm": 1.9334732294082642,
|
| 3878 |
+
"learning_rate": 0.00011734634203682208,
|
| 3879 |
+
"loss": 4.4285,
|
| 3880 |
+
"num_input_tokens_seen": 424227104,
|
| 3881 |
+
"step": 72600
|
| 3882 |
+
},
|
| 3883 |
+
{
|
| 3884 |
+
"epoch": 0.8017014805305005,
|
| 3885 |
+
"grad_norm": 1.8234983682632446,
|
| 3886 |
+
"learning_rate": 0.00011725818064269657,
|
| 3887 |
+
"loss": 4.438,
|
| 3888 |
+
"num_input_tokens_seen": 425093536,
|
| 3889 |
+
"step": 72750
|
| 3890 |
+
},
|
| 3891 |
+
{
|
| 3892 |
+
"epoch": 0.8033544732738623,
|
| 3893 |
+
"grad_norm": 1.8719639778137207,
|
| 3894 |
+
"learning_rate": 0.00011717001924857106,
|
| 3895 |
+
"loss": 4.432,
|
| 3896 |
+
"num_input_tokens_seen": 425967904,
|
| 3897 |
+
"step": 72900
|
| 3898 |
+
},
|
| 3899 |
+
{
|
| 3900 |
+
"epoch": 0.8050074660172242,
|
| 3901 |
+
"grad_norm": 1.879062533378601,
|
| 3902 |
+
"learning_rate": 0.00011708185785444555,
|
| 3903 |
+
"loss": 4.4157,
|
| 3904 |
+
"num_input_tokens_seen": 426848192,
|
| 3905 |
+
"step": 73050
|
| 3906 |
+
},
|
| 3907 |
+
{
|
| 3908 |
+
"epoch": 0.806660458760586,
|
| 3909 |
+
"grad_norm": 1.8409887552261353,
|
| 3910 |
+
"learning_rate": 0.00011699369646032003,
|
| 3911 |
+
"loss": 4.4173,
|
| 3912 |
+
"num_input_tokens_seen": 427730528,
|
| 3913 |
+
"step": 73200
|
| 3914 |
+
},
|
| 3915 |
+
{
|
| 3916 |
+
"epoch": 0.8083134515039478,
|
| 3917 |
+
"grad_norm": 1.9242078065872192,
|
| 3918 |
+
"learning_rate": 0.00011690553506619452,
|
| 3919 |
+
"loss": 4.4296,
|
| 3920 |
+
"num_input_tokens_seen": 428621792,
|
| 3921 |
+
"step": 73350
|
| 3922 |
+
},
|
| 3923 |
+
{
|
| 3924 |
+
"epoch": 0.8099664442473098,
|
| 3925 |
+
"grad_norm": 1.8767496347427368,
|
| 3926 |
+
"learning_rate": 0.00011681737367206901,
|
| 3927 |
+
"loss": 4.4245,
|
| 3928 |
+
"num_input_tokens_seen": 429494624,
|
| 3929 |
+
"step": 73500
|
| 3930 |
+
},
|
| 3931 |
+
{
|
| 3932 |
+
"epoch": 0.8116194369906716,
|
| 3933 |
+
"grad_norm": 1.8519647121429443,
|
| 3934 |
+
"learning_rate": 0.0001167292122779435,
|
| 3935 |
+
"loss": 4.4254,
|
| 3936 |
+
"num_input_tokens_seen": 430360704,
|
| 3937 |
+
"step": 73650
|
| 3938 |
+
},
|
| 3939 |
+
{
|
| 3940 |
+
"epoch": 0.8132724297340335,
|
| 3941 |
+
"grad_norm": 1.9251487255096436,
|
| 3942 |
+
"learning_rate": 0.00011664105088381798,
|
| 3943 |
+
"loss": 4.4362,
|
| 3944 |
+
"num_input_tokens_seen": 431239616,
|
| 3945 |
+
"step": 73800
|
| 3946 |
+
},
|
| 3947 |
+
{
|
| 3948 |
+
"epoch": 0.8149254224773953,
|
| 3949 |
+
"grad_norm": 1.8970694541931152,
|
| 3950 |
+
"learning_rate": 0.00011655288948969248,
|
| 3951 |
+
"loss": 4.4261,
|
| 3952 |
+
"num_input_tokens_seen": 432109120,
|
| 3953 |
+
"step": 73950
|
| 3954 |
+
},
|
| 3955 |
+
{
|
| 3956 |
+
"epoch": 0.8165784152207571,
|
| 3957 |
+
"grad_norm": 1.8284028768539429,
|
| 3958 |
+
"learning_rate": 0.00011646472809556695,
|
| 3959 |
+
"loss": 4.4406,
|
| 3960 |
+
"num_input_tokens_seen": 433005440,
|
| 3961 |
+
"step": 74100
|
| 3962 |
+
},
|
| 3963 |
+
{
|
| 3964 |
+
"epoch": 0.8182314079641191,
|
| 3965 |
+
"grad_norm": 1.7933986186981201,
|
| 3966 |
+
"learning_rate": 0.00011637656670144145,
|
| 3967 |
+
"loss": 4.4333,
|
| 3968 |
+
"num_input_tokens_seen": 433891456,
|
| 3969 |
+
"step": 74250
|
| 3970 |
+
},
|
| 3971 |
+
{
|
| 3972 |
+
"epoch": 0.8198844007074809,
|
| 3973 |
+
"grad_norm": 1.802509069442749,
|
| 3974 |
+
"learning_rate": 0.00011628840530731593,
|
| 3975 |
+
"loss": 4.4201,
|
| 3976 |
+
"num_input_tokens_seen": 434769856,
|
| 3977 |
+
"step": 74400
|
| 3978 |
+
},
|
| 3979 |
+
{
|
| 3980 |
+
"epoch": 0.8215373934508428,
|
| 3981 |
+
"grad_norm": 1.7515144348144531,
|
| 3982 |
+
"learning_rate": 0.00011620024391319043,
|
| 3983 |
+
"loss": 4.4225,
|
| 3984 |
+
"num_input_tokens_seen": 435665920,
|
| 3985 |
+
"step": 74550
|
| 3986 |
+
},
|
| 3987 |
+
{
|
| 3988 |
+
"epoch": 0.8231903861942046,
|
| 3989 |
+
"grad_norm": 1.8373006582260132,
|
| 3990 |
+
"learning_rate": 0.00011611208251906491,
|
| 3991 |
+
"loss": 4.4265,
|
| 3992 |
+
"num_input_tokens_seen": 436549984,
|
| 3993 |
+
"step": 74700
|
| 3994 |
+
},
|
| 3995 |
+
{
|
| 3996 |
+
"epoch": 0.8248433789375664,
|
| 3997 |
+
"grad_norm": 1.8570173978805542,
|
| 3998 |
+
"learning_rate": 0.00011602392112493941,
|
| 3999 |
+
"loss": 4.4196,
|
| 4000 |
+
"num_input_tokens_seen": 437427456,
|
| 4001 |
+
"step": 74850
|
| 4002 |
+
},
|
| 4003 |
+
{
|
| 4004 |
+
"epoch": 0.8264963716809283,
|
| 4005 |
+
"grad_norm": 1.9485052824020386,
|
| 4006 |
+
"learning_rate": 0.00011593575973081388,
|
| 4007 |
+
"loss": 4.4235,
|
| 4008 |
+
"num_input_tokens_seen": 438316576,
|
| 4009 |
+
"step": 75000
|
| 4010 |
+
},
|
| 4011 |
+
{
|
| 4012 |
+
"epoch": 0.8281493644242902,
|
| 4013 |
+
"grad_norm": 1.8972394466400146,
|
| 4014 |
+
"learning_rate": 0.00011584818607931588,
|
| 4015 |
+
"loss": 4.4231,
|
| 4016 |
+
"num_input_tokens_seen": 439211648,
|
| 4017 |
+
"step": 75150
|
| 4018 |
+
},
|
| 4019 |
+
{
|
| 4020 |
+
"epoch": 0.829802357167652,
|
| 4021 |
+
"grad_norm": 1.778745412826538,
|
| 4022 |
+
"learning_rate": 0.00011576002468519036,
|
| 4023 |
+
"loss": 4.423,
|
| 4024 |
+
"num_input_tokens_seen": 440086912,
|
| 4025 |
+
"step": 75300
|
| 4026 |
+
},
|
| 4027 |
+
{
|
| 4028 |
+
"epoch": 0.8314553499110139,
|
| 4029 |
+
"grad_norm": 1.923743486404419,
|
| 4030 |
+
"learning_rate": 0.00011567245103369236,
|
| 4031 |
+
"loss": 4.4112,
|
| 4032 |
+
"num_input_tokens_seen": 440976736,
|
| 4033 |
+
"step": 75450
|
| 4034 |
+
},
|
| 4035 |
+
{
|
| 4036 |
+
"epoch": 0.8331083426543757,
|
| 4037 |
+
"grad_norm": 1.8902959823608398,
|
| 4038 |
+
"learning_rate": 0.00011558428963956683,
|
| 4039 |
+
"loss": 4.4171,
|
| 4040 |
+
"num_input_tokens_seen": 441827264,
|
| 4041 |
+
"step": 75600
|
| 4042 |
+
},
|
| 4043 |
+
{
|
| 4044 |
+
"epoch": 0.8347613353977376,
|
| 4045 |
+
"grad_norm": 1.882279396057129,
|
| 4046 |
+
"learning_rate": 0.00011549612824544133,
|
| 4047 |
+
"loss": 4.4386,
|
| 4048 |
+
"num_input_tokens_seen": 442687328,
|
| 4049 |
+
"step": 75750
|
| 4050 |
+
},
|
| 4051 |
+
{
|
| 4052 |
+
"epoch": 0.8364143281410995,
|
| 4053 |
+
"grad_norm": 1.8508954048156738,
|
| 4054 |
+
"learning_rate": 0.00011540796685131581,
|
| 4055 |
+
"loss": 4.4261,
|
| 4056 |
+
"num_input_tokens_seen": 443558176,
|
| 4057 |
+
"step": 75900
|
| 4058 |
+
},
|
| 4059 |
+
{
|
| 4060 |
+
"epoch": 0.8380673208844613,
|
| 4061 |
+
"grad_norm": 1.8582794666290283,
|
| 4062 |
+
"learning_rate": 0.00011531980545719031,
|
| 4063 |
+
"loss": 4.4214,
|
| 4064 |
+
"num_input_tokens_seen": 444448512,
|
| 4065 |
+
"step": 76050
|
| 4066 |
+
},
|
| 4067 |
+
{
|
| 4068 |
+
"epoch": 0.8397203136278232,
|
| 4069 |
+
"grad_norm": 1.8337671756744385,
|
| 4070 |
+
"learning_rate": 0.00011523164406306478,
|
| 4071 |
+
"loss": 4.4257,
|
| 4072 |
+
"num_input_tokens_seen": 445311808,
|
| 4073 |
+
"step": 76200
|
| 4074 |
+
},
|
| 4075 |
+
{
|
| 4076 |
+
"epoch": 0.841373306371185,
|
| 4077 |
+
"grad_norm": 1.8980998992919922,
|
| 4078 |
+
"learning_rate": 0.00011514348266893929,
|
| 4079 |
+
"loss": 4.4294,
|
| 4080 |
+
"num_input_tokens_seen": 446200832,
|
| 4081 |
+
"step": 76350
|
| 4082 |
+
},
|
| 4083 |
+
{
|
| 4084 |
+
"epoch": 0.8430262991145469,
|
| 4085 |
+
"grad_norm": 1.8506239652633667,
|
| 4086 |
+
"learning_rate": 0.00011505532127481376,
|
| 4087 |
+
"loss": 4.4205,
|
| 4088 |
+
"num_input_tokens_seen": 447081376,
|
| 4089 |
+
"step": 76500
|
| 4090 |
+
},
|
| 4091 |
+
{
|
| 4092 |
+
"epoch": 0.8446792918579088,
|
| 4093 |
+
"grad_norm": 1.8824795484542847,
|
| 4094 |
+
"learning_rate": 0.00011496715988068826,
|
| 4095 |
+
"loss": 4.4213,
|
| 4096 |
+
"num_input_tokens_seen": 447960352,
|
| 4097 |
+
"step": 76650
|
| 4098 |
+
},
|
| 4099 |
+
{
|
| 4100 |
+
"epoch": 0.8463322846012706,
|
| 4101 |
+
"grad_norm": 1.8223339319229126,
|
| 4102 |
+
"learning_rate": 0.00011487899848656274,
|
| 4103 |
+
"loss": 4.4267,
|
| 4104 |
+
"num_input_tokens_seen": 448832096,
|
| 4105 |
+
"step": 76800
|
| 4106 |
+
},
|
| 4107 |
+
{
|
| 4108 |
+
"epoch": 0.8479852773446325,
|
| 4109 |
+
"grad_norm": 1.8224749565124512,
|
| 4110 |
+
"learning_rate": 0.00011479083709243724,
|
| 4111 |
+
"loss": 4.4181,
|
| 4112 |
+
"num_input_tokens_seen": 449706720,
|
| 4113 |
+
"step": 76950
|
| 4114 |
+
},
|
| 4115 |
+
{
|
| 4116 |
+
"epoch": 0.8496382700879943,
|
| 4117 |
+
"grad_norm": 1.903432011604309,
|
| 4118 |
+
"learning_rate": 0.00011470267569831172,
|
| 4119 |
+
"loss": 4.4309,
|
| 4120 |
+
"num_input_tokens_seen": 450578752,
|
| 4121 |
+
"step": 77100
|
| 4122 |
+
},
|
| 4123 |
+
{
|
| 4124 |
+
"epoch": 0.8512912628313561,
|
| 4125 |
+
"grad_norm": 1.8261497020721436,
|
| 4126 |
+
"learning_rate": 0.0001146145143041862,
|
| 4127 |
+
"loss": 4.4195,
|
| 4128 |
+
"num_input_tokens_seen": 451441120,
|
| 4129 |
+
"step": 77250
|
| 4130 |
+
},
|
| 4131 |
+
{
|
| 4132 |
+
"epoch": 0.852944255574718,
|
| 4133 |
+
"grad_norm": 1.8583135604858398,
|
| 4134 |
+
"learning_rate": 0.00011452635291006069,
|
| 4135 |
+
"loss": 4.4198,
|
| 4136 |
+
"num_input_tokens_seen": 452318560,
|
| 4137 |
+
"step": 77400
|
| 4138 |
+
},
|
| 4139 |
+
{
|
| 4140 |
+
"epoch": 0.8545972483180799,
|
| 4141 |
+
"grad_norm": 1.8936694860458374,
|
| 4142 |
+
"learning_rate": 0.00011443819151593518,
|
| 4143 |
+
"loss": 4.4287,
|
| 4144 |
+
"num_input_tokens_seen": 453195136,
|
| 4145 |
+
"step": 77550
|
| 4146 |
+
},
|
| 4147 |
+
{
|
| 4148 |
+
"epoch": 0.8562502410614418,
|
| 4149 |
+
"grad_norm": 1.9256082773208618,
|
| 4150 |
+
"learning_rate": 0.00011435003012180967,
|
| 4151 |
+
"loss": 4.4156,
|
| 4152 |
+
"num_input_tokens_seen": 454055456,
|
| 4153 |
+
"step": 77700
|
| 4154 |
+
},
|
| 4155 |
+
{
|
| 4156 |
+
"epoch": 0.8579032338048036,
|
| 4157 |
+
"grad_norm": 1.8237937688827515,
|
| 4158 |
+
"learning_rate": 0.00011426186872768416,
|
| 4159 |
+
"loss": 4.4348,
|
| 4160 |
+
"num_input_tokens_seen": 454932800,
|
| 4161 |
+
"step": 77850
|
| 4162 |
+
},
|
| 4163 |
+
{
|
| 4164 |
+
"epoch": 0.8595562265481654,
|
| 4165 |
+
"grad_norm": 1.8298827409744263,
|
| 4166 |
+
"learning_rate": 0.00011417370733355865,
|
| 4167 |
+
"loss": 4.4178,
|
| 4168 |
+
"num_input_tokens_seen": 455826208,
|
| 4169 |
+
"step": 78000
|
| 4170 |
+
},
|
| 4171 |
+
{
|
| 4172 |
+
"epoch": 0.8612092192915273,
|
| 4173 |
+
"grad_norm": 1.895670771598816,
|
| 4174 |
+
"learning_rate": 0.00011408554593943314,
|
| 4175 |
+
"loss": 4.4164,
|
| 4176 |
+
"num_input_tokens_seen": 456726848,
|
| 4177 |
+
"step": 78150
|
| 4178 |
+
},
|
| 4179 |
+
{
|
| 4180 |
+
"epoch": 0.8628622120348892,
|
| 4181 |
+
"grad_norm": 1.750807523727417,
|
| 4182 |
+
"learning_rate": 0.00011399738454530761,
|
| 4183 |
+
"loss": 4.4051,
|
| 4184 |
+
"num_input_tokens_seen": 457606464,
|
| 4185 |
+
"step": 78300
|
| 4186 |
+
},
|
| 4187 |
+
{
|
| 4188 |
+
"epoch": 0.8645152047782511,
|
| 4189 |
+
"grad_norm": 1.8419345617294312,
|
| 4190 |
+
"learning_rate": 0.00011390922315118211,
|
| 4191 |
+
"loss": 4.4249,
|
| 4192 |
+
"num_input_tokens_seen": 458484896,
|
| 4193 |
+
"step": 78450
|
| 4194 |
+
},
|
| 4195 |
+
{
|
| 4196 |
+
"epoch": 0.8661681975216129,
|
| 4197 |
+
"grad_norm": 2.033911943435669,
|
| 4198 |
+
"learning_rate": 0.00011382106175705659,
|
| 4199 |
+
"loss": 4.421,
|
| 4200 |
+
"num_input_tokens_seen": 459353824,
|
| 4201 |
+
"step": 78600
|
| 4202 |
+
},
|
| 4203 |
+
{
|
| 4204 |
+
"epoch": 0.8678211902649747,
|
| 4205 |
+
"grad_norm": 1.9020805358886719,
|
| 4206 |
+
"learning_rate": 0.00011373290036293109,
|
| 4207 |
+
"loss": 4.4016,
|
| 4208 |
+
"num_input_tokens_seen": 460221184,
|
| 4209 |
+
"step": 78750
|
| 4210 |
+
},
|
| 4211 |
+
{
|
| 4212 |
+
"epoch": 0.8694741830083366,
|
| 4213 |
+
"grad_norm": 1.91862952709198,
|
| 4214 |
+
"learning_rate": 0.00011364473896880557,
|
| 4215 |
+
"loss": 4.4093,
|
| 4216 |
+
"num_input_tokens_seen": 461087744,
|
| 4217 |
+
"step": 78900
|
| 4218 |
+
},
|
| 4219 |
+
{
|
| 4220 |
+
"epoch": 0.8711271757516984,
|
| 4221 |
+
"grad_norm": 1.7994396686553955,
|
| 4222 |
+
"learning_rate": 0.00011355657757468007,
|
| 4223 |
+
"loss": 4.4281,
|
| 4224 |
+
"num_input_tokens_seen": 461960160,
|
| 4225 |
+
"step": 79050
|
| 4226 |
+
},
|
| 4227 |
+
{
|
| 4228 |
+
"epoch": 0.8727801684950603,
|
| 4229 |
+
"grad_norm": 1.7911181449890137,
|
| 4230 |
+
"learning_rate": 0.00011346841618055454,
|
| 4231 |
+
"loss": 4.4229,
|
| 4232 |
+
"num_input_tokens_seen": 462838144,
|
| 4233 |
+
"step": 79200
|
| 4234 |
+
},
|
| 4235 |
+
{
|
| 4236 |
+
"epoch": 0.8744331612384222,
|
| 4237 |
+
"grad_norm": 1.923474907875061,
|
| 4238 |
+
"learning_rate": 0.00011338025478642904,
|
| 4239 |
+
"loss": 4.4103,
|
| 4240 |
+
"num_input_tokens_seen": 463703328,
|
| 4241 |
+
"step": 79350
|
| 4242 |
+
},
|
| 4243 |
+
{
|
| 4244 |
+
"epoch": 0.876086153981784,
|
| 4245 |
+
"grad_norm": 1.994814157485962,
|
| 4246 |
+
"learning_rate": 0.00011329268113493102,
|
| 4247 |
+
"loss": 4.4128,
|
| 4248 |
+
"num_input_tokens_seen": 464568896,
|
| 4249 |
+
"step": 79500
|
| 4250 |
+
},
|
| 4251 |
+
{
|
| 4252 |
+
"epoch": 0.8777391467251459,
|
| 4253 |
+
"grad_norm": 1.875200867652893,
|
| 4254 |
+
"learning_rate": 0.00011320451974080551,
|
| 4255 |
+
"loss": 4.4224,
|
| 4256 |
+
"num_input_tokens_seen": 465434144,
|
| 4257 |
+
"step": 79650
|
| 4258 |
+
},
|
| 4259 |
+
{
|
| 4260 |
+
"epoch": 0.8793921394685077,
|
| 4261 |
+
"grad_norm": 1.8729829788208008,
|
| 4262 |
+
"learning_rate": 0.00011311635834668,
|
| 4263 |
+
"loss": 4.4274,
|
| 4264 |
+
"num_input_tokens_seen": 466293984,
|
| 4265 |
+
"step": 79800
|
| 4266 |
+
},
|
| 4267 |
+
{
|
| 4268 |
+
"epoch": 0.8810451322118696,
|
| 4269 |
+
"grad_norm": 1.772687315940857,
|
| 4270 |
+
"learning_rate": 0.00011302819695255449,
|
| 4271 |
+
"loss": 4.4178,
|
| 4272 |
+
"num_input_tokens_seen": 467169280,
|
| 4273 |
+
"step": 79950
|
| 4274 |
+
},
|
| 4275 |
+
{
|
| 4276 |
+
"epoch": 0.8826981249552315,
|
| 4277 |
+
"grad_norm": 1.8293451070785522,
|
| 4278 |
+
"learning_rate": 0.00011294003555842898,
|
| 4279 |
+
"loss": 4.412,
|
| 4280 |
+
"num_input_tokens_seen": 468023552,
|
| 4281 |
+
"step": 80100
|
| 4282 |
+
},
|
| 4283 |
+
{
|
| 4284 |
+
"epoch": 0.8843511176985933,
|
| 4285 |
+
"grad_norm": 1.9000316858291626,
|
| 4286 |
+
"learning_rate": 0.00011285187416430346,
|
| 4287 |
+
"loss": 4.4114,
|
| 4288 |
+
"num_input_tokens_seen": 468883168,
|
| 4289 |
+
"step": 80250
|
| 4290 |
+
},
|
| 4291 |
+
{
|
| 4292 |
+
"epoch": 0.8860041104419552,
|
| 4293 |
+
"grad_norm": 1.8056668043136597,
|
| 4294 |
+
"learning_rate": 0.00011276371277017795,
|
| 4295 |
+
"loss": 4.4225,
|
| 4296 |
+
"num_input_tokens_seen": 469761120,
|
| 4297 |
+
"step": 80400
|
| 4298 |
+
},
|
| 4299 |
+
{
|
| 4300 |
+
"epoch": 0.887657103185317,
|
| 4301 |
+
"grad_norm": 1.7813293933868408,
|
| 4302 |
+
"learning_rate": 0.00011267555137605243,
|
| 4303 |
+
"loss": 4.4176,
|
| 4304 |
+
"num_input_tokens_seen": 470629344,
|
| 4305 |
+
"step": 80550
|
| 4306 |
+
},
|
| 4307 |
+
{
|
| 4308 |
+
"epoch": 0.8893100959286788,
|
| 4309 |
+
"grad_norm": 1.8244847059249878,
|
| 4310 |
+
"learning_rate": 0.00011258738998192693,
|
| 4311 |
+
"loss": 4.4082,
|
| 4312 |
+
"num_input_tokens_seen": 471489280,
|
| 4313 |
+
"step": 80700
|
| 4314 |
+
},
|
| 4315 |
+
{
|
| 4316 |
+
"epoch": 0.8909630886720408,
|
| 4317 |
+
"grad_norm": 1.8946529626846313,
|
| 4318 |
+
"learning_rate": 0.00011249981633042892,
|
| 4319 |
+
"loss": 4.4162,
|
| 4320 |
+
"num_input_tokens_seen": 472365504,
|
| 4321 |
+
"step": 80850
|
| 4322 |
+
},
|
| 4323 |
+
{
|
| 4324 |
+
"epoch": 0.8926160814154026,
|
| 4325 |
+
"grad_norm": 1.870685338973999,
|
| 4326 |
+
"learning_rate": 0.0001124116549363034,
|
| 4327 |
+
"loss": 4.4009,
|
| 4328 |
+
"num_input_tokens_seen": 473238752,
|
| 4329 |
+
"step": 81000
|
| 4330 |
+
},
|
| 4331 |
+
{
|
| 4332 |
+
"epoch": 0.8942690741587644,
|
| 4333 |
+
"grad_norm": 1.9169375896453857,
|
| 4334 |
+
"learning_rate": 0.0001123234935421779,
|
| 4335 |
+
"loss": 4.4176,
|
| 4336 |
+
"num_input_tokens_seen": 474122880,
|
| 4337 |
+
"step": 81150
|
| 4338 |
+
},
|
| 4339 |
+
{
|
| 4340 |
+
"epoch": 0.8959220669021263,
|
| 4341 |
+
"grad_norm": 1.9780856370925903,
|
| 4342 |
+
"learning_rate": 0.00011223533214805237,
|
| 4343 |
+
"loss": 4.401,
|
| 4344 |
+
"num_input_tokens_seen": 475002240,
|
| 4345 |
+
"step": 81300
|
| 4346 |
+
},
|
| 4347 |
+
{
|
| 4348 |
+
"epoch": 0.8975750596454881,
|
| 4349 |
+
"grad_norm": 1.8493030071258545,
|
| 4350 |
+
"learning_rate": 0.00011214717075392688,
|
| 4351 |
+
"loss": 4.4091,
|
| 4352 |
+
"num_input_tokens_seen": 475882624,
|
| 4353 |
+
"step": 81450
|
| 4354 |
}
|
| 4355 |
],
|
| 4356 |
"logging_steps": 150,
|
| 4357 |
"max_steps": 272232,
|
| 4358 |
+
"num_input_tokens_seen": 476177504,
|
| 4359 |
"num_train_epochs": 3,
|
| 4360 |
"save_steps": 500,
|
| 4361 |
"stateful_callbacks": {
|
|
|
|
| 4370 |
"attributes": {}
|
| 4371 |
}
|
| 4372 |
},
|
| 4373 |
+
"total_flos": 7352485415362560.0,
|
| 4374 |
"train_batch_size": 32,
|
| 4375 |
"trial_name": null,
|
| 4376 |
"trial_params": null
|