Uploaded checkpoint-20000
Browse files- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1795 -5
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 119975656
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:140c22cb100bb7fb3b9f92ae13ec5bb2bfcde7ed82d7e4434fc5a235f98cb24e
|
| 3 |
size 119975656
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 240145026
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f49faa425cc20765775b424c81e8f5599e3725a2dc79226d42d68c4573812cfe
|
| 3 |
size 240145026
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d70b07077c15c8bd127eaf0a24ba45e81ca7ce6ae410b7a625f50c345ec6eb1f
|
| 3 |
size 14244
|
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:e7dc694a733ff91b79c5eaf7bcfe8aa41771c4ef8a47d325d2a9e9f6bc78f946
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 1.
|
| 3 |
-
"best_model_checkpoint": "runs/deepseek_lora_20240422-165831/checkpoint-
|
| 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,
|
|
@@ -12537,6 +12537,1796 @@
|
|
| 12537 |
"eval_samples_per_second": 15.113,
|
| 12538 |
"eval_steps_per_second": 15.113,
|
| 12539 |
"step": 17500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12540 |
}
|
| 12541 |
],
|
| 12542 |
"logging_steps": 10,
|
|
@@ -12544,7 +14334,7 @@
|
|
| 12544 |
"num_input_tokens_seen": 0,
|
| 12545 |
"num_train_epochs": 1,
|
| 12546 |
"save_steps": 2500,
|
| 12547 |
-
"total_flos":
|
| 12548 |
"train_batch_size": 1,
|
| 12549 |
"trial_name": null,
|
| 12550 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.348677158355713,
|
| 3 |
+
"best_model_checkpoint": "runs/deepseek_lora_20240422-165831/checkpoint-20000",
|
| 4 |
+
"epoch": 0.5,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 20000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 12537 |
"eval_samples_per_second": 15.113,
|
| 12538 |
"eval_steps_per_second": 15.113,
|
| 12539 |
"step": 17500
|
| 12540 |
+
},
|
| 12541 |
+
{
|
| 12542 |
+
"epoch": 0.44,
|
| 12543 |
+
"grad_norm": 5.189449310302734,
|
| 12544 |
+
"learning_rate": 8.467796610169492e-06,
|
| 12545 |
+
"loss": 1.352,
|
| 12546 |
+
"step": 17510
|
| 12547 |
+
},
|
| 12548 |
+
{
|
| 12549 |
+
"epoch": 0.44,
|
| 12550 |
+
"grad_norm": 2.765326499938965,
|
| 12551 |
+
"learning_rate": 8.461016949152543e-06,
|
| 12552 |
+
"loss": 1.5275,
|
| 12553 |
+
"step": 17520
|
| 12554 |
+
},
|
| 12555 |
+
{
|
| 12556 |
+
"epoch": 0.44,
|
| 12557 |
+
"grad_norm": 3.60880446434021,
|
| 12558 |
+
"learning_rate": 8.454237288135593e-06,
|
| 12559 |
+
"loss": 1.2793,
|
| 12560 |
+
"step": 17530
|
| 12561 |
+
},
|
| 12562 |
+
{
|
| 12563 |
+
"epoch": 0.44,
|
| 12564 |
+
"grad_norm": 3.072174310684204,
|
| 12565 |
+
"learning_rate": 8.447457627118644e-06,
|
| 12566 |
+
"loss": 1.2746,
|
| 12567 |
+
"step": 17540
|
| 12568 |
+
},
|
| 12569 |
+
{
|
| 12570 |
+
"epoch": 0.44,
|
| 12571 |
+
"grad_norm": 3.3222594261169434,
|
| 12572 |
+
"learning_rate": 8.440677966101696e-06,
|
| 12573 |
+
"loss": 1.3359,
|
| 12574 |
+
"step": 17550
|
| 12575 |
+
},
|
| 12576 |
+
{
|
| 12577 |
+
"epoch": 0.44,
|
| 12578 |
+
"grad_norm": 1.3298100233078003,
|
| 12579 |
+
"learning_rate": 8.433898305084747e-06,
|
| 12580 |
+
"loss": 1.3826,
|
| 12581 |
+
"step": 17560
|
| 12582 |
+
},
|
| 12583 |
+
{
|
| 12584 |
+
"epoch": 0.44,
|
| 12585 |
+
"grad_norm": 4.06503438949585,
|
| 12586 |
+
"learning_rate": 8.427118644067797e-06,
|
| 12587 |
+
"loss": 1.3339,
|
| 12588 |
+
"step": 17570
|
| 12589 |
+
},
|
| 12590 |
+
{
|
| 12591 |
+
"epoch": 0.44,
|
| 12592 |
+
"grad_norm": 6.150786399841309,
|
| 12593 |
+
"learning_rate": 8.420338983050848e-06,
|
| 12594 |
+
"loss": 1.3414,
|
| 12595 |
+
"step": 17580
|
| 12596 |
+
},
|
| 12597 |
+
{
|
| 12598 |
+
"epoch": 0.44,
|
| 12599 |
+
"grad_norm": 9.045748710632324,
|
| 12600 |
+
"learning_rate": 8.413559322033898e-06,
|
| 12601 |
+
"loss": 1.3183,
|
| 12602 |
+
"step": 17590
|
| 12603 |
+
},
|
| 12604 |
+
{
|
| 12605 |
+
"epoch": 0.44,
|
| 12606 |
+
"grad_norm": 4.415981769561768,
|
| 12607 |
+
"learning_rate": 8.40677966101695e-06,
|
| 12608 |
+
"loss": 1.3323,
|
| 12609 |
+
"step": 17600
|
| 12610 |
+
},
|
| 12611 |
+
{
|
| 12612 |
+
"epoch": 0.44,
|
| 12613 |
+
"grad_norm": 8.40713882446289,
|
| 12614 |
+
"learning_rate": 8.400000000000001e-06,
|
| 12615 |
+
"loss": 1.2519,
|
| 12616 |
+
"step": 17610
|
| 12617 |
+
},
|
| 12618 |
+
{
|
| 12619 |
+
"epoch": 0.44,
|
| 12620 |
+
"grad_norm": 6.841925144195557,
|
| 12621 |
+
"learning_rate": 8.39322033898305e-06,
|
| 12622 |
+
"loss": 1.4799,
|
| 12623 |
+
"step": 17620
|
| 12624 |
+
},
|
| 12625 |
+
{
|
| 12626 |
+
"epoch": 0.44,
|
| 12627 |
+
"grad_norm": 10.993582725524902,
|
| 12628 |
+
"learning_rate": 8.386440677966102e-06,
|
| 12629 |
+
"loss": 1.3495,
|
| 12630 |
+
"step": 17630
|
| 12631 |
+
},
|
| 12632 |
+
{
|
| 12633 |
+
"epoch": 0.44,
|
| 12634 |
+
"grad_norm": 2.115006685256958,
|
| 12635 |
+
"learning_rate": 8.379661016949153e-06,
|
| 12636 |
+
"loss": 1.3902,
|
| 12637 |
+
"step": 17640
|
| 12638 |
+
},
|
| 12639 |
+
{
|
| 12640 |
+
"epoch": 0.44,
|
| 12641 |
+
"grad_norm": 3.485889434814453,
|
| 12642 |
+
"learning_rate": 8.372881355932205e-06,
|
| 12643 |
+
"loss": 1.4462,
|
| 12644 |
+
"step": 17650
|
| 12645 |
+
},
|
| 12646 |
+
{
|
| 12647 |
+
"epoch": 0.44,
|
| 12648 |
+
"grad_norm": 5.725930690765381,
|
| 12649 |
+
"learning_rate": 8.366101694915255e-06,
|
| 12650 |
+
"loss": 1.4603,
|
| 12651 |
+
"step": 17660
|
| 12652 |
+
},
|
| 12653 |
+
{
|
| 12654 |
+
"epoch": 0.44,
|
| 12655 |
+
"grad_norm": 6.5530548095703125,
|
| 12656 |
+
"learning_rate": 8.359322033898306e-06,
|
| 12657 |
+
"loss": 1.4954,
|
| 12658 |
+
"step": 17670
|
| 12659 |
+
},
|
| 12660 |
+
{
|
| 12661 |
+
"epoch": 0.44,
|
| 12662 |
+
"grad_norm": 6.006740570068359,
|
| 12663 |
+
"learning_rate": 8.352542372881357e-06,
|
| 12664 |
+
"loss": 1.3556,
|
| 12665 |
+
"step": 17680
|
| 12666 |
+
},
|
| 12667 |
+
{
|
| 12668 |
+
"epoch": 0.44,
|
| 12669 |
+
"grad_norm": 5.5967936515808105,
|
| 12670 |
+
"learning_rate": 8.345762711864409e-06,
|
| 12671 |
+
"loss": 1.0719,
|
| 12672 |
+
"step": 17690
|
| 12673 |
+
},
|
| 12674 |
+
{
|
| 12675 |
+
"epoch": 0.44,
|
| 12676 |
+
"grad_norm": 5.70643949508667,
|
| 12677 |
+
"learning_rate": 8.338983050847458e-06,
|
| 12678 |
+
"loss": 1.2968,
|
| 12679 |
+
"step": 17700
|
| 12680 |
+
},
|
| 12681 |
+
{
|
| 12682 |
+
"epoch": 0.44,
|
| 12683 |
+
"grad_norm": 3.280465841293335,
|
| 12684 |
+
"learning_rate": 8.332203389830508e-06,
|
| 12685 |
+
"loss": 1.2669,
|
| 12686 |
+
"step": 17710
|
| 12687 |
+
},
|
| 12688 |
+
{
|
| 12689 |
+
"epoch": 0.44,
|
| 12690 |
+
"grad_norm": 3.098782777786255,
|
| 12691 |
+
"learning_rate": 8.32542372881356e-06,
|
| 12692 |
+
"loss": 1.4704,
|
| 12693 |
+
"step": 17720
|
| 12694 |
+
},
|
| 12695 |
+
{
|
| 12696 |
+
"epoch": 0.44,
|
| 12697 |
+
"grad_norm": 2.825495958328247,
|
| 12698 |
+
"learning_rate": 8.318644067796611e-06,
|
| 12699 |
+
"loss": 1.2165,
|
| 12700 |
+
"step": 17730
|
| 12701 |
+
},
|
| 12702 |
+
{
|
| 12703 |
+
"epoch": 0.44,
|
| 12704 |
+
"grad_norm": 3.8040249347686768,
|
| 12705 |
+
"learning_rate": 8.311864406779662e-06,
|
| 12706 |
+
"loss": 1.3623,
|
| 12707 |
+
"step": 17740
|
| 12708 |
+
},
|
| 12709 |
+
{
|
| 12710 |
+
"epoch": 0.44,
|
| 12711 |
+
"grad_norm": 3.4438130855560303,
|
| 12712 |
+
"learning_rate": 8.305084745762712e-06,
|
| 12713 |
+
"loss": 1.4255,
|
| 12714 |
+
"step": 17750
|
| 12715 |
+
},
|
| 12716 |
+
{
|
| 12717 |
+
"epoch": 0.44,
|
| 12718 |
+
"grad_norm": 2.187945604324341,
|
| 12719 |
+
"learning_rate": 8.298305084745763e-06,
|
| 12720 |
+
"loss": 1.2181,
|
| 12721 |
+
"step": 17760
|
| 12722 |
+
},
|
| 12723 |
+
{
|
| 12724 |
+
"epoch": 0.44,
|
| 12725 |
+
"grad_norm": 2.0644514560699463,
|
| 12726 |
+
"learning_rate": 8.291525423728815e-06,
|
| 12727 |
+
"loss": 1.4669,
|
| 12728 |
+
"step": 17770
|
| 12729 |
+
},
|
| 12730 |
+
{
|
| 12731 |
+
"epoch": 0.44,
|
| 12732 |
+
"grad_norm": 4.314215660095215,
|
| 12733 |
+
"learning_rate": 8.284745762711866e-06,
|
| 12734 |
+
"loss": 1.3776,
|
| 12735 |
+
"step": 17780
|
| 12736 |
+
},
|
| 12737 |
+
{
|
| 12738 |
+
"epoch": 0.44,
|
| 12739 |
+
"grad_norm": 2.948415517807007,
|
| 12740 |
+
"learning_rate": 8.277966101694916e-06,
|
| 12741 |
+
"loss": 1.3198,
|
| 12742 |
+
"step": 17790
|
| 12743 |
+
},
|
| 12744 |
+
{
|
| 12745 |
+
"epoch": 0.45,
|
| 12746 |
+
"grad_norm": 7.77257776260376,
|
| 12747 |
+
"learning_rate": 8.271186440677966e-06,
|
| 12748 |
+
"loss": 1.3826,
|
| 12749 |
+
"step": 17800
|
| 12750 |
+
},
|
| 12751 |
+
{
|
| 12752 |
+
"epoch": 0.45,
|
| 12753 |
+
"grad_norm": 2.2085230350494385,
|
| 12754 |
+
"learning_rate": 8.264406779661017e-06,
|
| 12755 |
+
"loss": 1.3443,
|
| 12756 |
+
"step": 17810
|
| 12757 |
+
},
|
| 12758 |
+
{
|
| 12759 |
+
"epoch": 0.45,
|
| 12760 |
+
"grad_norm": 2.8318214416503906,
|
| 12761 |
+
"learning_rate": 8.257627118644068e-06,
|
| 12762 |
+
"loss": 1.3126,
|
| 12763 |
+
"step": 17820
|
| 12764 |
+
},
|
| 12765 |
+
{
|
| 12766 |
+
"epoch": 0.45,
|
| 12767 |
+
"grad_norm": 4.544606685638428,
|
| 12768 |
+
"learning_rate": 8.25084745762712e-06,
|
| 12769 |
+
"loss": 1.2955,
|
| 12770 |
+
"step": 17830
|
| 12771 |
+
},
|
| 12772 |
+
{
|
| 12773 |
+
"epoch": 0.45,
|
| 12774 |
+
"grad_norm": 5.595729351043701,
|
| 12775 |
+
"learning_rate": 8.24406779661017e-06,
|
| 12776 |
+
"loss": 1.3814,
|
| 12777 |
+
"step": 17840
|
| 12778 |
+
},
|
| 12779 |
+
{
|
| 12780 |
+
"epoch": 0.45,
|
| 12781 |
+
"grad_norm": 6.201912879943848,
|
| 12782 |
+
"learning_rate": 8.237288135593221e-06,
|
| 12783 |
+
"loss": 1.4243,
|
| 12784 |
+
"step": 17850
|
| 12785 |
+
},
|
| 12786 |
+
{
|
| 12787 |
+
"epoch": 0.45,
|
| 12788 |
+
"grad_norm": 7.697713851928711,
|
| 12789 |
+
"learning_rate": 8.230508474576272e-06,
|
| 12790 |
+
"loss": 1.326,
|
| 12791 |
+
"step": 17860
|
| 12792 |
+
},
|
| 12793 |
+
{
|
| 12794 |
+
"epoch": 0.45,
|
| 12795 |
+
"grad_norm": 3.7588555812835693,
|
| 12796 |
+
"learning_rate": 8.223728813559324e-06,
|
| 12797 |
+
"loss": 1.1265,
|
| 12798 |
+
"step": 17870
|
| 12799 |
+
},
|
| 12800 |
+
{
|
| 12801 |
+
"epoch": 0.45,
|
| 12802 |
+
"grad_norm": 8.472000122070312,
|
| 12803 |
+
"learning_rate": 8.216949152542373e-06,
|
| 12804 |
+
"loss": 1.2678,
|
| 12805 |
+
"step": 17880
|
| 12806 |
+
},
|
| 12807 |
+
{
|
| 12808 |
+
"epoch": 0.45,
|
| 12809 |
+
"grad_norm": 4.802706241607666,
|
| 12810 |
+
"learning_rate": 8.210169491525425e-06,
|
| 12811 |
+
"loss": 1.5116,
|
| 12812 |
+
"step": 17890
|
| 12813 |
+
},
|
| 12814 |
+
{
|
| 12815 |
+
"epoch": 0.45,
|
| 12816 |
+
"grad_norm": 5.357900142669678,
|
| 12817 |
+
"learning_rate": 8.203389830508475e-06,
|
| 12818 |
+
"loss": 1.4435,
|
| 12819 |
+
"step": 17900
|
| 12820 |
+
},
|
| 12821 |
+
{
|
| 12822 |
+
"epoch": 0.45,
|
| 12823 |
+
"grad_norm": 3.5331151485443115,
|
| 12824 |
+
"learning_rate": 8.196610169491526e-06,
|
| 12825 |
+
"loss": 1.3838,
|
| 12826 |
+
"step": 17910
|
| 12827 |
+
},
|
| 12828 |
+
{
|
| 12829 |
+
"epoch": 0.45,
|
| 12830 |
+
"grad_norm": 2.591485023498535,
|
| 12831 |
+
"learning_rate": 8.189830508474577e-06,
|
| 12832 |
+
"loss": 1.3595,
|
| 12833 |
+
"step": 17920
|
| 12834 |
+
},
|
| 12835 |
+
{
|
| 12836 |
+
"epoch": 0.45,
|
| 12837 |
+
"grad_norm": 9.1065034866333,
|
| 12838 |
+
"learning_rate": 8.183050847457627e-06,
|
| 12839 |
+
"loss": 1.4109,
|
| 12840 |
+
"step": 17930
|
| 12841 |
+
},
|
| 12842 |
+
{
|
| 12843 |
+
"epoch": 0.45,
|
| 12844 |
+
"grad_norm": 3.2992680072784424,
|
| 12845 |
+
"learning_rate": 8.176271186440678e-06,
|
| 12846 |
+
"loss": 1.3554,
|
| 12847 |
+
"step": 17940
|
| 12848 |
+
},
|
| 12849 |
+
{
|
| 12850 |
+
"epoch": 0.45,
|
| 12851 |
+
"grad_norm": 10.191650390625,
|
| 12852 |
+
"learning_rate": 8.16949152542373e-06,
|
| 12853 |
+
"loss": 1.3561,
|
| 12854 |
+
"step": 17950
|
| 12855 |
+
},
|
| 12856 |
+
{
|
| 12857 |
+
"epoch": 0.45,
|
| 12858 |
+
"grad_norm": 5.769218444824219,
|
| 12859 |
+
"learning_rate": 8.162711864406781e-06,
|
| 12860 |
+
"loss": 1.4044,
|
| 12861 |
+
"step": 17960
|
| 12862 |
+
},
|
| 12863 |
+
{
|
| 12864 |
+
"epoch": 0.45,
|
| 12865 |
+
"grad_norm": 47.62531661987305,
|
| 12866 |
+
"learning_rate": 8.155932203389831e-06,
|
| 12867 |
+
"loss": 1.6211,
|
| 12868 |
+
"step": 17970
|
| 12869 |
+
},
|
| 12870 |
+
{
|
| 12871 |
+
"epoch": 0.45,
|
| 12872 |
+
"grad_norm": 3.514390230178833,
|
| 12873 |
+
"learning_rate": 8.149152542372882e-06,
|
| 12874 |
+
"loss": 1.288,
|
| 12875 |
+
"step": 17980
|
| 12876 |
+
},
|
| 12877 |
+
{
|
| 12878 |
+
"epoch": 0.45,
|
| 12879 |
+
"grad_norm": 3.9963278770446777,
|
| 12880 |
+
"learning_rate": 8.142372881355934e-06,
|
| 12881 |
+
"loss": 1.3625,
|
| 12882 |
+
"step": 17990
|
| 12883 |
+
},
|
| 12884 |
+
{
|
| 12885 |
+
"epoch": 0.45,
|
| 12886 |
+
"grad_norm": 4.2144856452941895,
|
| 12887 |
+
"learning_rate": 8.135593220338983e-06,
|
| 12888 |
+
"loss": 1.3154,
|
| 12889 |
+
"step": 18000
|
| 12890 |
+
},
|
| 12891 |
+
{
|
| 12892 |
+
"epoch": 0.45,
|
| 12893 |
+
"eval_loss": 1.3663108348846436,
|
| 12894 |
+
"eval_runtime": 66.1644,
|
| 12895 |
+
"eval_samples_per_second": 15.114,
|
| 12896 |
+
"eval_steps_per_second": 15.114,
|
| 12897 |
+
"step": 18000
|
| 12898 |
+
},
|
| 12899 |
+
{
|
| 12900 |
+
"epoch": 0.45,
|
| 12901 |
+
"grad_norm": 3.047025680541992,
|
| 12902 |
+
"learning_rate": 8.128813559322035e-06,
|
| 12903 |
+
"loss": 1.3131,
|
| 12904 |
+
"step": 18010
|
| 12905 |
+
},
|
| 12906 |
+
{
|
| 12907 |
+
"epoch": 0.45,
|
| 12908 |
+
"grad_norm": 5.798911094665527,
|
| 12909 |
+
"learning_rate": 8.122033898305085e-06,
|
| 12910 |
+
"loss": 1.4371,
|
| 12911 |
+
"step": 18020
|
| 12912 |
+
},
|
| 12913 |
+
{
|
| 12914 |
+
"epoch": 0.45,
|
| 12915 |
+
"grad_norm": 3.8710880279541016,
|
| 12916 |
+
"learning_rate": 8.115254237288136e-06,
|
| 12917 |
+
"loss": 1.5147,
|
| 12918 |
+
"step": 18030
|
| 12919 |
+
},
|
| 12920 |
+
{
|
| 12921 |
+
"epoch": 0.45,
|
| 12922 |
+
"grad_norm": 6.445448398590088,
|
| 12923 |
+
"learning_rate": 8.108474576271187e-06,
|
| 12924 |
+
"loss": 1.3332,
|
| 12925 |
+
"step": 18040
|
| 12926 |
+
},
|
| 12927 |
+
{
|
| 12928 |
+
"epoch": 0.45,
|
| 12929 |
+
"grad_norm": 5.0877275466918945,
|
| 12930 |
+
"learning_rate": 8.101694915254237e-06,
|
| 12931 |
+
"loss": 1.5076,
|
| 12932 |
+
"step": 18050
|
| 12933 |
+
},
|
| 12934 |
+
{
|
| 12935 |
+
"epoch": 0.45,
|
| 12936 |
+
"grad_norm": 6.0062737464904785,
|
| 12937 |
+
"learning_rate": 8.094915254237289e-06,
|
| 12938 |
+
"loss": 1.3257,
|
| 12939 |
+
"step": 18060
|
| 12940 |
+
},
|
| 12941 |
+
{
|
| 12942 |
+
"epoch": 0.45,
|
| 12943 |
+
"grad_norm": 3.2079477310180664,
|
| 12944 |
+
"learning_rate": 8.08813559322034e-06,
|
| 12945 |
+
"loss": 1.2162,
|
| 12946 |
+
"step": 18070
|
| 12947 |
+
},
|
| 12948 |
+
{
|
| 12949 |
+
"epoch": 0.45,
|
| 12950 |
+
"grad_norm": 2.974025249481201,
|
| 12951 |
+
"learning_rate": 8.081355932203391e-06,
|
| 12952 |
+
"loss": 1.3433,
|
| 12953 |
+
"step": 18080
|
| 12954 |
+
},
|
| 12955 |
+
{
|
| 12956 |
+
"epoch": 0.45,
|
| 12957 |
+
"grad_norm": 11.321368217468262,
|
| 12958 |
+
"learning_rate": 8.074576271186441e-06,
|
| 12959 |
+
"loss": 1.4845,
|
| 12960 |
+
"step": 18090
|
| 12961 |
+
},
|
| 12962 |
+
{
|
| 12963 |
+
"epoch": 0.45,
|
| 12964 |
+
"grad_norm": 3.768314838409424,
|
| 12965 |
+
"learning_rate": 8.067796610169492e-06,
|
| 12966 |
+
"loss": 1.3288,
|
| 12967 |
+
"step": 18100
|
| 12968 |
+
},
|
| 12969 |
+
{
|
| 12970 |
+
"epoch": 0.45,
|
| 12971 |
+
"grad_norm": 3.9936201572418213,
|
| 12972 |
+
"learning_rate": 8.061016949152542e-06,
|
| 12973 |
+
"loss": 1.2507,
|
| 12974 |
+
"step": 18110
|
| 12975 |
+
},
|
| 12976 |
+
{
|
| 12977 |
+
"epoch": 0.45,
|
| 12978 |
+
"grad_norm": 3.69484543800354,
|
| 12979 |
+
"learning_rate": 8.054237288135594e-06,
|
| 12980 |
+
"loss": 1.2615,
|
| 12981 |
+
"step": 18120
|
| 12982 |
+
},
|
| 12983 |
+
{
|
| 12984 |
+
"epoch": 0.45,
|
| 12985 |
+
"grad_norm": 6.352110862731934,
|
| 12986 |
+
"learning_rate": 8.047457627118645e-06,
|
| 12987 |
+
"loss": 1.2888,
|
| 12988 |
+
"step": 18130
|
| 12989 |
+
},
|
| 12990 |
+
{
|
| 12991 |
+
"epoch": 0.45,
|
| 12992 |
+
"grad_norm": 8.259808540344238,
|
| 12993 |
+
"learning_rate": 8.040677966101695e-06,
|
| 12994 |
+
"loss": 1.3588,
|
| 12995 |
+
"step": 18140
|
| 12996 |
+
},
|
| 12997 |
+
{
|
| 12998 |
+
"epoch": 0.45,
|
| 12999 |
+
"grad_norm": 8.203502655029297,
|
| 13000 |
+
"learning_rate": 8.033898305084746e-06,
|
| 13001 |
+
"loss": 1.2716,
|
| 13002 |
+
"step": 18150
|
| 13003 |
+
},
|
| 13004 |
+
{
|
| 13005 |
+
"epoch": 0.45,
|
| 13006 |
+
"grad_norm": 3.8740904331207275,
|
| 13007 |
+
"learning_rate": 8.027118644067797e-06,
|
| 13008 |
+
"loss": 1.4191,
|
| 13009 |
+
"step": 18160
|
| 13010 |
+
},
|
| 13011 |
+
{
|
| 13012 |
+
"epoch": 0.45,
|
| 13013 |
+
"grad_norm": 3.9104349613189697,
|
| 13014 |
+
"learning_rate": 8.020338983050849e-06,
|
| 13015 |
+
"loss": 1.1126,
|
| 13016 |
+
"step": 18170
|
| 13017 |
+
},
|
| 13018 |
+
{
|
| 13019 |
+
"epoch": 0.45,
|
| 13020 |
+
"grad_norm": 4.973033428192139,
|
| 13021 |
+
"learning_rate": 8.013559322033899e-06,
|
| 13022 |
+
"loss": 1.4034,
|
| 13023 |
+
"step": 18180
|
| 13024 |
+
},
|
| 13025 |
+
{
|
| 13026 |
+
"epoch": 0.45,
|
| 13027 |
+
"grad_norm": 5.2367329597473145,
|
| 13028 |
+
"learning_rate": 8.00677966101695e-06,
|
| 13029 |
+
"loss": 1.4094,
|
| 13030 |
+
"step": 18190
|
| 13031 |
+
},
|
| 13032 |
+
{
|
| 13033 |
+
"epoch": 0.46,
|
| 13034 |
+
"grad_norm": 4.842703819274902,
|
| 13035 |
+
"learning_rate": 8.000000000000001e-06,
|
| 13036 |
+
"loss": 1.4423,
|
| 13037 |
+
"step": 18200
|
| 13038 |
+
},
|
| 13039 |
+
{
|
| 13040 |
+
"epoch": 0.46,
|
| 13041 |
+
"grad_norm": 6.39046573638916,
|
| 13042 |
+
"learning_rate": 7.993220338983053e-06,
|
| 13043 |
+
"loss": 1.3491,
|
| 13044 |
+
"step": 18210
|
| 13045 |
+
},
|
| 13046 |
+
{
|
| 13047 |
+
"epoch": 0.46,
|
| 13048 |
+
"grad_norm": 3.424907922744751,
|
| 13049 |
+
"learning_rate": 7.986440677966102e-06,
|
| 13050 |
+
"loss": 1.268,
|
| 13051 |
+
"step": 18220
|
| 13052 |
+
},
|
| 13053 |
+
{
|
| 13054 |
+
"epoch": 0.46,
|
| 13055 |
+
"grad_norm": 4.363860607147217,
|
| 13056 |
+
"learning_rate": 7.979661016949152e-06,
|
| 13057 |
+
"loss": 1.2921,
|
| 13058 |
+
"step": 18230
|
| 13059 |
+
},
|
| 13060 |
+
{
|
| 13061 |
+
"epoch": 0.46,
|
| 13062 |
+
"grad_norm": 10.35586929321289,
|
| 13063 |
+
"learning_rate": 7.972881355932204e-06,
|
| 13064 |
+
"loss": 1.4456,
|
| 13065 |
+
"step": 18240
|
| 13066 |
+
},
|
| 13067 |
+
{
|
| 13068 |
+
"epoch": 0.46,
|
| 13069 |
+
"grad_norm": 11.094548225402832,
|
| 13070 |
+
"learning_rate": 7.966101694915255e-06,
|
| 13071 |
+
"loss": 1.5682,
|
| 13072 |
+
"step": 18250
|
| 13073 |
+
},
|
| 13074 |
+
{
|
| 13075 |
+
"epoch": 0.46,
|
| 13076 |
+
"grad_norm": 10.801241874694824,
|
| 13077 |
+
"learning_rate": 7.959322033898306e-06,
|
| 13078 |
+
"loss": 1.3475,
|
| 13079 |
+
"step": 18260
|
| 13080 |
+
},
|
| 13081 |
+
{
|
| 13082 |
+
"epoch": 0.46,
|
| 13083 |
+
"grad_norm": 3.893005132675171,
|
| 13084 |
+
"learning_rate": 7.952542372881356e-06,
|
| 13085 |
+
"loss": 1.3858,
|
| 13086 |
+
"step": 18270
|
| 13087 |
+
},
|
| 13088 |
+
{
|
| 13089 |
+
"epoch": 0.46,
|
| 13090 |
+
"grad_norm": 8.01429557800293,
|
| 13091 |
+
"learning_rate": 7.945762711864407e-06,
|
| 13092 |
+
"loss": 1.3695,
|
| 13093 |
+
"step": 18280
|
| 13094 |
+
},
|
| 13095 |
+
{
|
| 13096 |
+
"epoch": 0.46,
|
| 13097 |
+
"grad_norm": 9.507207870483398,
|
| 13098 |
+
"learning_rate": 7.938983050847459e-06,
|
| 13099 |
+
"loss": 1.2442,
|
| 13100 |
+
"step": 18290
|
| 13101 |
+
},
|
| 13102 |
+
{
|
| 13103 |
+
"epoch": 0.46,
|
| 13104 |
+
"grad_norm": 4.206308364868164,
|
| 13105 |
+
"learning_rate": 7.93220338983051e-06,
|
| 13106 |
+
"loss": 1.241,
|
| 13107 |
+
"step": 18300
|
| 13108 |
+
},
|
| 13109 |
+
{
|
| 13110 |
+
"epoch": 0.46,
|
| 13111 |
+
"grad_norm": 7.3320794105529785,
|
| 13112 |
+
"learning_rate": 7.92542372881356e-06,
|
| 13113 |
+
"loss": 1.3338,
|
| 13114 |
+
"step": 18310
|
| 13115 |
+
},
|
| 13116 |
+
{
|
| 13117 |
+
"epoch": 0.46,
|
| 13118 |
+
"grad_norm": 4.259428977966309,
|
| 13119 |
+
"learning_rate": 7.91864406779661e-06,
|
| 13120 |
+
"loss": 1.3902,
|
| 13121 |
+
"step": 18320
|
| 13122 |
+
},
|
| 13123 |
+
{
|
| 13124 |
+
"epoch": 0.46,
|
| 13125 |
+
"grad_norm": 15.748085975646973,
|
| 13126 |
+
"learning_rate": 7.911864406779661e-06,
|
| 13127 |
+
"loss": 1.282,
|
| 13128 |
+
"step": 18330
|
| 13129 |
+
},
|
| 13130 |
+
{
|
| 13131 |
+
"epoch": 0.46,
|
| 13132 |
+
"grad_norm": 7.123409748077393,
|
| 13133 |
+
"learning_rate": 7.905084745762712e-06,
|
| 13134 |
+
"loss": 1.3311,
|
| 13135 |
+
"step": 18340
|
| 13136 |
+
},
|
| 13137 |
+
{
|
| 13138 |
+
"epoch": 0.46,
|
| 13139 |
+
"grad_norm": 7.022697925567627,
|
| 13140 |
+
"learning_rate": 7.898305084745764e-06,
|
| 13141 |
+
"loss": 1.292,
|
| 13142 |
+
"step": 18350
|
| 13143 |
+
},
|
| 13144 |
+
{
|
| 13145 |
+
"epoch": 0.46,
|
| 13146 |
+
"grad_norm": 3.3042855262756348,
|
| 13147 |
+
"learning_rate": 7.891525423728814e-06,
|
| 13148 |
+
"loss": 1.3587,
|
| 13149 |
+
"step": 18360
|
| 13150 |
+
},
|
| 13151 |
+
{
|
| 13152 |
+
"epoch": 0.46,
|
| 13153 |
+
"grad_norm": 3.5053913593292236,
|
| 13154 |
+
"learning_rate": 7.884745762711865e-06,
|
| 13155 |
+
"loss": 1.5188,
|
| 13156 |
+
"step": 18370
|
| 13157 |
+
},
|
| 13158 |
+
{
|
| 13159 |
+
"epoch": 0.46,
|
| 13160 |
+
"grad_norm": 2.6854681968688965,
|
| 13161 |
+
"learning_rate": 7.877966101694916e-06,
|
| 13162 |
+
"loss": 1.3633,
|
| 13163 |
+
"step": 18380
|
| 13164 |
+
},
|
| 13165 |
+
{
|
| 13166 |
+
"epoch": 0.46,
|
| 13167 |
+
"grad_norm": 3.8879082202911377,
|
| 13168 |
+
"learning_rate": 7.871186440677968e-06,
|
| 13169 |
+
"loss": 1.5491,
|
| 13170 |
+
"step": 18390
|
| 13171 |
+
},
|
| 13172 |
+
{
|
| 13173 |
+
"epoch": 0.46,
|
| 13174 |
+
"grad_norm": 9.43989086151123,
|
| 13175 |
+
"learning_rate": 7.864406779661017e-06,
|
| 13176 |
+
"loss": 1.2686,
|
| 13177 |
+
"step": 18400
|
| 13178 |
+
},
|
| 13179 |
+
{
|
| 13180 |
+
"epoch": 0.46,
|
| 13181 |
+
"grad_norm": 4.684770107269287,
|
| 13182 |
+
"learning_rate": 7.857627118644069e-06,
|
| 13183 |
+
"loss": 1.2463,
|
| 13184 |
+
"step": 18410
|
| 13185 |
+
},
|
| 13186 |
+
{
|
| 13187 |
+
"epoch": 0.46,
|
| 13188 |
+
"grad_norm": 2.699023485183716,
|
| 13189 |
+
"learning_rate": 7.850847457627119e-06,
|
| 13190 |
+
"loss": 1.5216,
|
| 13191 |
+
"step": 18420
|
| 13192 |
+
},
|
| 13193 |
+
{
|
| 13194 |
+
"epoch": 0.46,
|
| 13195 |
+
"grad_norm": 3.1254053115844727,
|
| 13196 |
+
"learning_rate": 7.84406779661017e-06,
|
| 13197 |
+
"loss": 1.4231,
|
| 13198 |
+
"step": 18430
|
| 13199 |
+
},
|
| 13200 |
+
{
|
| 13201 |
+
"epoch": 0.46,
|
| 13202 |
+
"grad_norm": 3.4634742736816406,
|
| 13203 |
+
"learning_rate": 7.837288135593221e-06,
|
| 13204 |
+
"loss": 1.4174,
|
| 13205 |
+
"step": 18440
|
| 13206 |
+
},
|
| 13207 |
+
{
|
| 13208 |
+
"epoch": 0.46,
|
| 13209 |
+
"grad_norm": 2.7554078102111816,
|
| 13210 |
+
"learning_rate": 7.830508474576271e-06,
|
| 13211 |
+
"loss": 1.2646,
|
| 13212 |
+
"step": 18450
|
| 13213 |
+
},
|
| 13214 |
+
{
|
| 13215 |
+
"epoch": 0.46,
|
| 13216 |
+
"grad_norm": 7.782464504241943,
|
| 13217 |
+
"learning_rate": 7.823728813559322e-06,
|
| 13218 |
+
"loss": 1.273,
|
| 13219 |
+
"step": 18460
|
| 13220 |
+
},
|
| 13221 |
+
{
|
| 13222 |
+
"epoch": 0.46,
|
| 13223 |
+
"grad_norm": 12.717724800109863,
|
| 13224 |
+
"learning_rate": 7.816949152542374e-06,
|
| 13225 |
+
"loss": 1.3536,
|
| 13226 |
+
"step": 18470
|
| 13227 |
+
},
|
| 13228 |
+
{
|
| 13229 |
+
"epoch": 0.46,
|
| 13230 |
+
"grad_norm": 4.644845008850098,
|
| 13231 |
+
"learning_rate": 7.810169491525425e-06,
|
| 13232 |
+
"loss": 1.567,
|
| 13233 |
+
"step": 18480
|
| 13234 |
+
},
|
| 13235 |
+
{
|
| 13236 |
+
"epoch": 0.46,
|
| 13237 |
+
"grad_norm": 2.032013177871704,
|
| 13238 |
+
"learning_rate": 7.803389830508475e-06,
|
| 13239 |
+
"loss": 1.4428,
|
| 13240 |
+
"step": 18490
|
| 13241 |
+
},
|
| 13242 |
+
{
|
| 13243 |
+
"epoch": 0.46,
|
| 13244 |
+
"grad_norm": 8.91115951538086,
|
| 13245 |
+
"learning_rate": 7.796610169491526e-06,
|
| 13246 |
+
"loss": 1.4071,
|
| 13247 |
+
"step": 18500
|
| 13248 |
+
},
|
| 13249 |
+
{
|
| 13250 |
+
"epoch": 0.46,
|
| 13251 |
+
"eval_loss": 1.3208402395248413,
|
| 13252 |
+
"eval_runtime": 66.1773,
|
| 13253 |
+
"eval_samples_per_second": 15.111,
|
| 13254 |
+
"eval_steps_per_second": 15.111,
|
| 13255 |
+
"step": 18500
|
| 13256 |
+
},
|
| 13257 |
+
{
|
| 13258 |
+
"epoch": 0.46,
|
| 13259 |
+
"grad_norm": 4.1798248291015625,
|
| 13260 |
+
"learning_rate": 7.789830508474578e-06,
|
| 13261 |
+
"loss": 1.3752,
|
| 13262 |
+
"step": 18510
|
| 13263 |
+
},
|
| 13264 |
+
{
|
| 13265 |
+
"epoch": 0.46,
|
| 13266 |
+
"grad_norm": 3.6975016593933105,
|
| 13267 |
+
"learning_rate": 7.783050847457628e-06,
|
| 13268 |
+
"loss": 1.3685,
|
| 13269 |
+
"step": 18520
|
| 13270 |
+
},
|
| 13271 |
+
{
|
| 13272 |
+
"epoch": 0.46,
|
| 13273 |
+
"grad_norm": 4.421247959136963,
|
| 13274 |
+
"learning_rate": 7.776271186440679e-06,
|
| 13275 |
+
"loss": 1.5264,
|
| 13276 |
+
"step": 18530
|
| 13277 |
+
},
|
| 13278 |
+
{
|
| 13279 |
+
"epoch": 0.46,
|
| 13280 |
+
"grad_norm": 8.739845275878906,
|
| 13281 |
+
"learning_rate": 7.769491525423729e-06,
|
| 13282 |
+
"loss": 1.4609,
|
| 13283 |
+
"step": 18540
|
| 13284 |
+
},
|
| 13285 |
+
{
|
| 13286 |
+
"epoch": 0.46,
|
| 13287 |
+
"grad_norm": 3.287424325942993,
|
| 13288 |
+
"learning_rate": 7.76271186440678e-06,
|
| 13289 |
+
"loss": 1.1154,
|
| 13290 |
+
"step": 18550
|
| 13291 |
+
},
|
| 13292 |
+
{
|
| 13293 |
+
"epoch": 0.46,
|
| 13294 |
+
"grad_norm": 3.1192266941070557,
|
| 13295 |
+
"learning_rate": 7.755932203389831e-06,
|
| 13296 |
+
"loss": 1.2787,
|
| 13297 |
+
"step": 18560
|
| 13298 |
+
},
|
| 13299 |
+
{
|
| 13300 |
+
"epoch": 0.46,
|
| 13301 |
+
"grad_norm": 8.368815422058105,
|
| 13302 |
+
"learning_rate": 7.749152542372881e-06,
|
| 13303 |
+
"loss": 1.3339,
|
| 13304 |
+
"step": 18570
|
| 13305 |
+
},
|
| 13306 |
+
{
|
| 13307 |
+
"epoch": 0.46,
|
| 13308 |
+
"grad_norm": 6.241825580596924,
|
| 13309 |
+
"learning_rate": 7.742372881355933e-06,
|
| 13310 |
+
"loss": 1.2293,
|
| 13311 |
+
"step": 18580
|
| 13312 |
+
},
|
| 13313 |
+
{
|
| 13314 |
+
"epoch": 0.46,
|
| 13315 |
+
"grad_norm": 9.267044067382812,
|
| 13316 |
+
"learning_rate": 7.735593220338984e-06,
|
| 13317 |
+
"loss": 1.4354,
|
| 13318 |
+
"step": 18590
|
| 13319 |
+
},
|
| 13320 |
+
{
|
| 13321 |
+
"epoch": 0.47,
|
| 13322 |
+
"grad_norm": 5.569955348968506,
|
| 13323 |
+
"learning_rate": 7.728813559322035e-06,
|
| 13324 |
+
"loss": 1.3941,
|
| 13325 |
+
"step": 18600
|
| 13326 |
+
},
|
| 13327 |
+
{
|
| 13328 |
+
"epoch": 0.47,
|
| 13329 |
+
"grad_norm": 11.273942947387695,
|
| 13330 |
+
"learning_rate": 7.722033898305085e-06,
|
| 13331 |
+
"loss": 1.1127,
|
| 13332 |
+
"step": 18610
|
| 13333 |
+
},
|
| 13334 |
+
{
|
| 13335 |
+
"epoch": 0.47,
|
| 13336 |
+
"grad_norm": 3.7952206134796143,
|
| 13337 |
+
"learning_rate": 7.715254237288136e-06,
|
| 13338 |
+
"loss": 1.3437,
|
| 13339 |
+
"step": 18620
|
| 13340 |
+
},
|
| 13341 |
+
{
|
| 13342 |
+
"epoch": 0.47,
|
| 13343 |
+
"grad_norm": 2.9084484577178955,
|
| 13344 |
+
"learning_rate": 7.708474576271186e-06,
|
| 13345 |
+
"loss": 1.3646,
|
| 13346 |
+
"step": 18630
|
| 13347 |
+
},
|
| 13348 |
+
{
|
| 13349 |
+
"epoch": 0.47,
|
| 13350 |
+
"grad_norm": 4.15964937210083,
|
| 13351 |
+
"learning_rate": 7.701694915254238e-06,
|
| 13352 |
+
"loss": 1.4666,
|
| 13353 |
+
"step": 18640
|
| 13354 |
+
},
|
| 13355 |
+
{
|
| 13356 |
+
"epoch": 0.47,
|
| 13357 |
+
"grad_norm": 4.938930988311768,
|
| 13358 |
+
"learning_rate": 7.694915254237289e-06,
|
| 13359 |
+
"loss": 1.4065,
|
| 13360 |
+
"step": 18650
|
| 13361 |
+
},
|
| 13362 |
+
{
|
| 13363 |
+
"epoch": 0.47,
|
| 13364 |
+
"grad_norm": 8.298666954040527,
|
| 13365 |
+
"learning_rate": 7.688135593220339e-06,
|
| 13366 |
+
"loss": 1.354,
|
| 13367 |
+
"step": 18660
|
| 13368 |
+
},
|
| 13369 |
+
{
|
| 13370 |
+
"epoch": 0.47,
|
| 13371 |
+
"grad_norm": 3.1837921142578125,
|
| 13372 |
+
"learning_rate": 7.68135593220339e-06,
|
| 13373 |
+
"loss": 1.3879,
|
| 13374 |
+
"step": 18670
|
| 13375 |
+
},
|
| 13376 |
+
{
|
| 13377 |
+
"epoch": 0.47,
|
| 13378 |
+
"grad_norm": 2.978053331375122,
|
| 13379 |
+
"learning_rate": 7.674576271186441e-06,
|
| 13380 |
+
"loss": 1.3765,
|
| 13381 |
+
"step": 18680
|
| 13382 |
+
},
|
| 13383 |
+
{
|
| 13384 |
+
"epoch": 0.47,
|
| 13385 |
+
"grad_norm": 5.5602827072143555,
|
| 13386 |
+
"learning_rate": 7.667796610169493e-06,
|
| 13387 |
+
"loss": 1.2135,
|
| 13388 |
+
"step": 18690
|
| 13389 |
+
},
|
| 13390 |
+
{
|
| 13391 |
+
"epoch": 0.47,
|
| 13392 |
+
"grad_norm": 5.481113910675049,
|
| 13393 |
+
"learning_rate": 7.661016949152543e-06,
|
| 13394 |
+
"loss": 1.3841,
|
| 13395 |
+
"step": 18700
|
| 13396 |
+
},
|
| 13397 |
+
{
|
| 13398 |
+
"epoch": 0.47,
|
| 13399 |
+
"grad_norm": 9.419681549072266,
|
| 13400 |
+
"learning_rate": 7.654237288135594e-06,
|
| 13401 |
+
"loss": 1.2349,
|
| 13402 |
+
"step": 18710
|
| 13403 |
+
},
|
| 13404 |
+
{
|
| 13405 |
+
"epoch": 0.47,
|
| 13406 |
+
"grad_norm": 4.955466270446777,
|
| 13407 |
+
"learning_rate": 7.647457627118645e-06,
|
| 13408 |
+
"loss": 1.4208,
|
| 13409 |
+
"step": 18720
|
| 13410 |
+
},
|
| 13411 |
+
{
|
| 13412 |
+
"epoch": 0.47,
|
| 13413 |
+
"grad_norm": 5.841220855712891,
|
| 13414 |
+
"learning_rate": 7.640677966101695e-06,
|
| 13415 |
+
"loss": 1.4526,
|
| 13416 |
+
"step": 18730
|
| 13417 |
+
},
|
| 13418 |
+
{
|
| 13419 |
+
"epoch": 0.47,
|
| 13420 |
+
"grad_norm": 3.1412012577056885,
|
| 13421 |
+
"learning_rate": 7.633898305084746e-06,
|
| 13422 |
+
"loss": 1.4606,
|
| 13423 |
+
"step": 18740
|
| 13424 |
+
},
|
| 13425 |
+
{
|
| 13426 |
+
"epoch": 0.47,
|
| 13427 |
+
"grad_norm": 4.624940395355225,
|
| 13428 |
+
"learning_rate": 7.627118644067797e-06,
|
| 13429 |
+
"loss": 1.3875,
|
| 13430 |
+
"step": 18750
|
| 13431 |
+
},
|
| 13432 |
+
{
|
| 13433 |
+
"epoch": 0.47,
|
| 13434 |
+
"grad_norm": 9.828381538391113,
|
| 13435 |
+
"learning_rate": 7.6203389830508476e-06,
|
| 13436 |
+
"loss": 1.4079,
|
| 13437 |
+
"step": 18760
|
| 13438 |
+
},
|
| 13439 |
+
{
|
| 13440 |
+
"epoch": 0.47,
|
| 13441 |
+
"grad_norm": 5.299017429351807,
|
| 13442 |
+
"learning_rate": 7.613559322033899e-06,
|
| 13443 |
+
"loss": 1.278,
|
| 13444 |
+
"step": 18770
|
| 13445 |
+
},
|
| 13446 |
+
{
|
| 13447 |
+
"epoch": 0.47,
|
| 13448 |
+
"grad_norm": 6.449117183685303,
|
| 13449 |
+
"learning_rate": 7.6067796610169495e-06,
|
| 13450 |
+
"loss": 1.4141,
|
| 13451 |
+
"step": 18780
|
| 13452 |
+
},
|
| 13453 |
+
{
|
| 13454 |
+
"epoch": 0.47,
|
| 13455 |
+
"grad_norm": 6.961145401000977,
|
| 13456 |
+
"learning_rate": 7.600000000000001e-06,
|
| 13457 |
+
"loss": 1.2828,
|
| 13458 |
+
"step": 18790
|
| 13459 |
+
},
|
| 13460 |
+
{
|
| 13461 |
+
"epoch": 0.47,
|
| 13462 |
+
"grad_norm": 8.588834762573242,
|
| 13463 |
+
"learning_rate": 7.5932203389830515e-06,
|
| 13464 |
+
"loss": 1.3544,
|
| 13465 |
+
"step": 18800
|
| 13466 |
+
},
|
| 13467 |
+
{
|
| 13468 |
+
"epoch": 0.47,
|
| 13469 |
+
"grad_norm": 2.5683794021606445,
|
| 13470 |
+
"learning_rate": 7.586440677966103e-06,
|
| 13471 |
+
"loss": 1.4127,
|
| 13472 |
+
"step": 18810
|
| 13473 |
+
},
|
| 13474 |
+
{
|
| 13475 |
+
"epoch": 0.47,
|
| 13476 |
+
"grad_norm": 2.101924180984497,
|
| 13477 |
+
"learning_rate": 7.5796610169491534e-06,
|
| 13478 |
+
"loss": 1.229,
|
| 13479 |
+
"step": 18820
|
| 13480 |
+
},
|
| 13481 |
+
{
|
| 13482 |
+
"epoch": 0.47,
|
| 13483 |
+
"grad_norm": 5.389444351196289,
|
| 13484 |
+
"learning_rate": 7.572881355932205e-06,
|
| 13485 |
+
"loss": 1.3112,
|
| 13486 |
+
"step": 18830
|
| 13487 |
+
},
|
| 13488 |
+
{
|
| 13489 |
+
"epoch": 0.47,
|
| 13490 |
+
"grad_norm": 3.154507875442505,
|
| 13491 |
+
"learning_rate": 7.5661016949152545e-06,
|
| 13492 |
+
"loss": 1.2603,
|
| 13493 |
+
"step": 18840
|
| 13494 |
+
},
|
| 13495 |
+
{
|
| 13496 |
+
"epoch": 0.47,
|
| 13497 |
+
"grad_norm": 4.938506126403809,
|
| 13498 |
+
"learning_rate": 7.559322033898305e-06,
|
| 13499 |
+
"loss": 1.5423,
|
| 13500 |
+
"step": 18850
|
| 13501 |
+
},
|
| 13502 |
+
{
|
| 13503 |
+
"epoch": 0.47,
|
| 13504 |
+
"grad_norm": 5.838810443878174,
|
| 13505 |
+
"learning_rate": 7.5525423728813565e-06,
|
| 13506 |
+
"loss": 1.0651,
|
| 13507 |
+
"step": 18860
|
| 13508 |
+
},
|
| 13509 |
+
{
|
| 13510 |
+
"epoch": 0.47,
|
| 13511 |
+
"grad_norm": 7.946393013000488,
|
| 13512 |
+
"learning_rate": 7.545762711864407e-06,
|
| 13513 |
+
"loss": 1.4706,
|
| 13514 |
+
"step": 18870
|
| 13515 |
+
},
|
| 13516 |
+
{
|
| 13517 |
+
"epoch": 0.47,
|
| 13518 |
+
"grad_norm": 12.6395902633667,
|
| 13519 |
+
"learning_rate": 7.5389830508474584e-06,
|
| 13520 |
+
"loss": 1.4036,
|
| 13521 |
+
"step": 18880
|
| 13522 |
+
},
|
| 13523 |
+
{
|
| 13524 |
+
"epoch": 0.47,
|
| 13525 |
+
"grad_norm": 4.769916534423828,
|
| 13526 |
+
"learning_rate": 7.532203389830509e-06,
|
| 13527 |
+
"loss": 1.33,
|
| 13528 |
+
"step": 18890
|
| 13529 |
+
},
|
| 13530 |
+
{
|
| 13531 |
+
"epoch": 0.47,
|
| 13532 |
+
"grad_norm": 5.928445816040039,
|
| 13533 |
+
"learning_rate": 7.52542372881356e-06,
|
| 13534 |
+
"loss": 1.4844,
|
| 13535 |
+
"step": 18900
|
| 13536 |
+
},
|
| 13537 |
+
{
|
| 13538 |
+
"epoch": 0.47,
|
| 13539 |
+
"grad_norm": 4.121998310089111,
|
| 13540 |
+
"learning_rate": 7.518644067796611e-06,
|
| 13541 |
+
"loss": 1.2672,
|
| 13542 |
+
"step": 18910
|
| 13543 |
+
},
|
| 13544 |
+
{
|
| 13545 |
+
"epoch": 0.47,
|
| 13546 |
+
"grad_norm": 1.256914496421814,
|
| 13547 |
+
"learning_rate": 7.511864406779662e-06,
|
| 13548 |
+
"loss": 1.337,
|
| 13549 |
+
"step": 18920
|
| 13550 |
+
},
|
| 13551 |
+
{
|
| 13552 |
+
"epoch": 0.47,
|
| 13553 |
+
"grad_norm": 9.951703071594238,
|
| 13554 |
+
"learning_rate": 7.505084745762713e-06,
|
| 13555 |
+
"loss": 1.3359,
|
| 13556 |
+
"step": 18930
|
| 13557 |
+
},
|
| 13558 |
+
{
|
| 13559 |
+
"epoch": 0.47,
|
| 13560 |
+
"grad_norm": 5.373715400695801,
|
| 13561 |
+
"learning_rate": 7.498305084745763e-06,
|
| 13562 |
+
"loss": 1.2437,
|
| 13563 |
+
"step": 18940
|
| 13564 |
+
},
|
| 13565 |
+
{
|
| 13566 |
+
"epoch": 0.47,
|
| 13567 |
+
"grad_norm": 7.120504379272461,
|
| 13568 |
+
"learning_rate": 7.491525423728814e-06,
|
| 13569 |
+
"loss": 1.1883,
|
| 13570 |
+
"step": 18950
|
| 13571 |
+
},
|
| 13572 |
+
{
|
| 13573 |
+
"epoch": 0.47,
|
| 13574 |
+
"grad_norm": 7.661159992218018,
|
| 13575 |
+
"learning_rate": 7.4847457627118646e-06,
|
| 13576 |
+
"loss": 1.2648,
|
| 13577 |
+
"step": 18960
|
| 13578 |
+
},
|
| 13579 |
+
{
|
| 13580 |
+
"epoch": 0.47,
|
| 13581 |
+
"grad_norm": 5.859286308288574,
|
| 13582 |
+
"learning_rate": 7.477966101694916e-06,
|
| 13583 |
+
"loss": 1.3106,
|
| 13584 |
+
"step": 18970
|
| 13585 |
+
},
|
| 13586 |
+
{
|
| 13587 |
+
"epoch": 0.47,
|
| 13588 |
+
"grad_norm": 7.276169300079346,
|
| 13589 |
+
"learning_rate": 7.4711864406779665e-06,
|
| 13590 |
+
"loss": 1.393,
|
| 13591 |
+
"step": 18980
|
| 13592 |
+
},
|
| 13593 |
+
{
|
| 13594 |
+
"epoch": 0.47,
|
| 13595 |
+
"grad_norm": 6.31447172164917,
|
| 13596 |
+
"learning_rate": 7.464406779661018e-06,
|
| 13597 |
+
"loss": 1.4051,
|
| 13598 |
+
"step": 18990
|
| 13599 |
+
},
|
| 13600 |
+
{
|
| 13601 |
+
"epoch": 0.47,
|
| 13602 |
+
"grad_norm": 8.36728572845459,
|
| 13603 |
+
"learning_rate": 7.4576271186440685e-06,
|
| 13604 |
+
"loss": 1.4003,
|
| 13605 |
+
"step": 19000
|
| 13606 |
+
},
|
| 13607 |
+
{
|
| 13608 |
+
"epoch": 0.47,
|
| 13609 |
+
"eval_loss": 1.347457766532898,
|
| 13610 |
+
"eval_runtime": 66.1461,
|
| 13611 |
+
"eval_samples_per_second": 15.118,
|
| 13612 |
+
"eval_steps_per_second": 15.118,
|
| 13613 |
+
"step": 19000
|
| 13614 |
+
},
|
| 13615 |
+
{
|
| 13616 |
+
"epoch": 0.48,
|
| 13617 |
+
"grad_norm": 3.911505937576294,
|
| 13618 |
+
"learning_rate": 7.45084745762712e-06,
|
| 13619 |
+
"loss": 1.2646,
|
| 13620 |
+
"step": 19010
|
| 13621 |
+
},
|
| 13622 |
+
{
|
| 13623 |
+
"epoch": 0.48,
|
| 13624 |
+
"grad_norm": 3.6765291690826416,
|
| 13625 |
+
"learning_rate": 7.4440677966101704e-06,
|
| 13626 |
+
"loss": 1.3827,
|
| 13627 |
+
"step": 19020
|
| 13628 |
+
},
|
| 13629 |
+
{
|
| 13630 |
+
"epoch": 0.48,
|
| 13631 |
+
"grad_norm": 3.899599313735962,
|
| 13632 |
+
"learning_rate": 7.437288135593221e-06,
|
| 13633 |
+
"loss": 1.4144,
|
| 13634 |
+
"step": 19030
|
| 13635 |
+
},
|
| 13636 |
+
{
|
| 13637 |
+
"epoch": 0.48,
|
| 13638 |
+
"grad_norm": 6.436791896820068,
|
| 13639 |
+
"learning_rate": 7.430508474576272e-06,
|
| 13640 |
+
"loss": 1.2087,
|
| 13641 |
+
"step": 19040
|
| 13642 |
+
},
|
| 13643 |
+
{
|
| 13644 |
+
"epoch": 0.48,
|
| 13645 |
+
"grad_norm": 4.624211311340332,
|
| 13646 |
+
"learning_rate": 7.423728813559322e-06,
|
| 13647 |
+
"loss": 1.3824,
|
| 13648 |
+
"step": 19050
|
| 13649 |
+
},
|
| 13650 |
+
{
|
| 13651 |
+
"epoch": 0.48,
|
| 13652 |
+
"grad_norm": 6.657593727111816,
|
| 13653 |
+
"learning_rate": 7.4169491525423735e-06,
|
| 13654 |
+
"loss": 1.3856,
|
| 13655 |
+
"step": 19060
|
| 13656 |
+
},
|
| 13657 |
+
{
|
| 13658 |
+
"epoch": 0.48,
|
| 13659 |
+
"grad_norm": 7.132912635803223,
|
| 13660 |
+
"learning_rate": 7.410169491525424e-06,
|
| 13661 |
+
"loss": 1.4319,
|
| 13662 |
+
"step": 19070
|
| 13663 |
+
},
|
| 13664 |
+
{
|
| 13665 |
+
"epoch": 0.48,
|
| 13666 |
+
"grad_norm": 2.8681843280792236,
|
| 13667 |
+
"learning_rate": 7.4033898305084754e-06,
|
| 13668 |
+
"loss": 1.419,
|
| 13669 |
+
"step": 19080
|
| 13670 |
+
},
|
| 13671 |
+
{
|
| 13672 |
+
"epoch": 0.48,
|
| 13673 |
+
"grad_norm": 5.819919109344482,
|
| 13674 |
+
"learning_rate": 7.396610169491526e-06,
|
| 13675 |
+
"loss": 1.3988,
|
| 13676 |
+
"step": 19090
|
| 13677 |
+
},
|
| 13678 |
+
{
|
| 13679 |
+
"epoch": 0.48,
|
| 13680 |
+
"grad_norm": 6.482515811920166,
|
| 13681 |
+
"learning_rate": 7.3898305084745766e-06,
|
| 13682 |
+
"loss": 1.3366,
|
| 13683 |
+
"step": 19100
|
| 13684 |
+
},
|
| 13685 |
+
{
|
| 13686 |
+
"epoch": 0.48,
|
| 13687 |
+
"grad_norm": 10.341208457946777,
|
| 13688 |
+
"learning_rate": 7.383050847457628e-06,
|
| 13689 |
+
"loss": 1.3502,
|
| 13690 |
+
"step": 19110
|
| 13691 |
+
},
|
| 13692 |
+
{
|
| 13693 |
+
"epoch": 0.48,
|
| 13694 |
+
"grad_norm": 2.820133686065674,
|
| 13695 |
+
"learning_rate": 7.3762711864406785e-06,
|
| 13696 |
+
"loss": 1.415,
|
| 13697 |
+
"step": 19120
|
| 13698 |
+
},
|
| 13699 |
+
{
|
| 13700 |
+
"epoch": 0.48,
|
| 13701 |
+
"grad_norm": 7.112204551696777,
|
| 13702 |
+
"learning_rate": 7.36949152542373e-06,
|
| 13703 |
+
"loss": 1.2609,
|
| 13704 |
+
"step": 19130
|
| 13705 |
+
},
|
| 13706 |
+
{
|
| 13707 |
+
"epoch": 0.48,
|
| 13708 |
+
"grad_norm": 6.172835826873779,
|
| 13709 |
+
"learning_rate": 7.3627118644067805e-06,
|
| 13710 |
+
"loss": 1.2456,
|
| 13711 |
+
"step": 19140
|
| 13712 |
+
},
|
| 13713 |
+
{
|
| 13714 |
+
"epoch": 0.48,
|
| 13715 |
+
"grad_norm": 2.9154605865478516,
|
| 13716 |
+
"learning_rate": 7.355932203389831e-06,
|
| 13717 |
+
"loss": 1.3332,
|
| 13718 |
+
"step": 19150
|
| 13719 |
+
},
|
| 13720 |
+
{
|
| 13721 |
+
"epoch": 0.48,
|
| 13722 |
+
"grad_norm": 6.8542256355285645,
|
| 13723 |
+
"learning_rate": 7.3491525423728816e-06,
|
| 13724 |
+
"loss": 1.2728,
|
| 13725 |
+
"step": 19160
|
| 13726 |
+
},
|
| 13727 |
+
{
|
| 13728 |
+
"epoch": 0.48,
|
| 13729 |
+
"grad_norm": 7.636801719665527,
|
| 13730 |
+
"learning_rate": 7.342372881355932e-06,
|
| 13731 |
+
"loss": 1.3744,
|
| 13732 |
+
"step": 19170
|
| 13733 |
+
},
|
| 13734 |
+
{
|
| 13735 |
+
"epoch": 0.48,
|
| 13736 |
+
"grad_norm": 7.556679725646973,
|
| 13737 |
+
"learning_rate": 7.3355932203389835e-06,
|
| 13738 |
+
"loss": 1.446,
|
| 13739 |
+
"step": 19180
|
| 13740 |
+
},
|
| 13741 |
+
{
|
| 13742 |
+
"epoch": 0.48,
|
| 13743 |
+
"grad_norm": 22.831199645996094,
|
| 13744 |
+
"learning_rate": 7.328813559322034e-06,
|
| 13745 |
+
"loss": 1.358,
|
| 13746 |
+
"step": 19190
|
| 13747 |
+
},
|
| 13748 |
+
{
|
| 13749 |
+
"epoch": 0.48,
|
| 13750 |
+
"grad_norm": 10.184800148010254,
|
| 13751 |
+
"learning_rate": 7.3220338983050855e-06,
|
| 13752 |
+
"loss": 1.2379,
|
| 13753 |
+
"step": 19200
|
| 13754 |
+
},
|
| 13755 |
+
{
|
| 13756 |
+
"epoch": 0.48,
|
| 13757 |
+
"grad_norm": 1.9062511920928955,
|
| 13758 |
+
"learning_rate": 7.315254237288136e-06,
|
| 13759 |
+
"loss": 1.3724,
|
| 13760 |
+
"step": 19210
|
| 13761 |
+
},
|
| 13762 |
+
{
|
| 13763 |
+
"epoch": 0.48,
|
| 13764 |
+
"grad_norm": 9.429797172546387,
|
| 13765 |
+
"learning_rate": 7.3084745762711874e-06,
|
| 13766 |
+
"loss": 1.33,
|
| 13767 |
+
"step": 19220
|
| 13768 |
+
},
|
| 13769 |
+
{
|
| 13770 |
+
"epoch": 0.48,
|
| 13771 |
+
"grad_norm": 3.129505157470703,
|
| 13772 |
+
"learning_rate": 7.301694915254238e-06,
|
| 13773 |
+
"loss": 1.1795,
|
| 13774 |
+
"step": 19230
|
| 13775 |
+
},
|
| 13776 |
+
{
|
| 13777 |
+
"epoch": 0.48,
|
| 13778 |
+
"grad_norm": 6.839028835296631,
|
| 13779 |
+
"learning_rate": 7.294915254237289e-06,
|
| 13780 |
+
"loss": 1.2983,
|
| 13781 |
+
"step": 19240
|
| 13782 |
+
},
|
| 13783 |
+
{
|
| 13784 |
+
"epoch": 0.48,
|
| 13785 |
+
"grad_norm": 15.552626609802246,
|
| 13786 |
+
"learning_rate": 7.288135593220339e-06,
|
| 13787 |
+
"loss": 1.4867,
|
| 13788 |
+
"step": 19250
|
| 13789 |
+
},
|
| 13790 |
+
{
|
| 13791 |
+
"epoch": 0.48,
|
| 13792 |
+
"grad_norm": 3.4032084941864014,
|
| 13793 |
+
"learning_rate": 7.28135593220339e-06,
|
| 13794 |
+
"loss": 1.4256,
|
| 13795 |
+
"step": 19260
|
| 13796 |
+
},
|
| 13797 |
+
{
|
| 13798 |
+
"epoch": 0.48,
|
| 13799 |
+
"grad_norm": 6.437047481536865,
|
| 13800 |
+
"learning_rate": 7.274576271186441e-06,
|
| 13801 |
+
"loss": 1.3215,
|
| 13802 |
+
"step": 19270
|
| 13803 |
+
},
|
| 13804 |
+
{
|
| 13805 |
+
"epoch": 0.48,
|
| 13806 |
+
"grad_norm": 5.592834949493408,
|
| 13807 |
+
"learning_rate": 7.267796610169492e-06,
|
| 13808 |
+
"loss": 1.3466,
|
| 13809 |
+
"step": 19280
|
| 13810 |
+
},
|
| 13811 |
+
{
|
| 13812 |
+
"epoch": 0.48,
|
| 13813 |
+
"grad_norm": 10.394824028015137,
|
| 13814 |
+
"learning_rate": 7.261016949152543e-06,
|
| 13815 |
+
"loss": 1.3806,
|
| 13816 |
+
"step": 19290
|
| 13817 |
+
},
|
| 13818 |
+
{
|
| 13819 |
+
"epoch": 0.48,
|
| 13820 |
+
"grad_norm": 6.921483039855957,
|
| 13821 |
+
"learning_rate": 7.2542372881355936e-06,
|
| 13822 |
+
"loss": 1.5624,
|
| 13823 |
+
"step": 19300
|
| 13824 |
+
},
|
| 13825 |
+
{
|
| 13826 |
+
"epoch": 0.48,
|
| 13827 |
+
"grad_norm": 3.078634262084961,
|
| 13828 |
+
"learning_rate": 7.247457627118645e-06,
|
| 13829 |
+
"loss": 1.1216,
|
| 13830 |
+
"step": 19310
|
| 13831 |
+
},
|
| 13832 |
+
{
|
| 13833 |
+
"epoch": 0.48,
|
| 13834 |
+
"grad_norm": 6.003988742828369,
|
| 13835 |
+
"learning_rate": 7.2406779661016955e-06,
|
| 13836 |
+
"loss": 1.2759,
|
| 13837 |
+
"step": 19320
|
| 13838 |
+
},
|
| 13839 |
+
{
|
| 13840 |
+
"epoch": 0.48,
|
| 13841 |
+
"grad_norm": 2.419410467147827,
|
| 13842 |
+
"learning_rate": 7.233898305084747e-06,
|
| 13843 |
+
"loss": 1.2067,
|
| 13844 |
+
"step": 19330
|
| 13845 |
+
},
|
| 13846 |
+
{
|
| 13847 |
+
"epoch": 0.48,
|
| 13848 |
+
"grad_norm": 3.014220952987671,
|
| 13849 |
+
"learning_rate": 7.2271186440677975e-06,
|
| 13850 |
+
"loss": 1.1819,
|
| 13851 |
+
"step": 19340
|
| 13852 |
+
},
|
| 13853 |
+
{
|
| 13854 |
+
"epoch": 0.48,
|
| 13855 |
+
"grad_norm": 5.271297454833984,
|
| 13856 |
+
"learning_rate": 7.220338983050849e-06,
|
| 13857 |
+
"loss": 1.4742,
|
| 13858 |
+
"step": 19350
|
| 13859 |
+
},
|
| 13860 |
+
{
|
| 13861 |
+
"epoch": 0.48,
|
| 13862 |
+
"grad_norm": 2.456838607788086,
|
| 13863 |
+
"learning_rate": 7.2135593220338986e-06,
|
| 13864 |
+
"loss": 1.4656,
|
| 13865 |
+
"step": 19360
|
| 13866 |
+
},
|
| 13867 |
+
{
|
| 13868 |
+
"epoch": 0.48,
|
| 13869 |
+
"grad_norm": 5.350526332855225,
|
| 13870 |
+
"learning_rate": 7.206779661016949e-06,
|
| 13871 |
+
"loss": 1.1238,
|
| 13872 |
+
"step": 19370
|
| 13873 |
+
},
|
| 13874 |
+
{
|
| 13875 |
+
"epoch": 0.48,
|
| 13876 |
+
"grad_norm": 5.42751932144165,
|
| 13877 |
+
"learning_rate": 7.2000000000000005e-06,
|
| 13878 |
+
"loss": 1.4704,
|
| 13879 |
+
"step": 19380
|
| 13880 |
+
},
|
| 13881 |
+
{
|
| 13882 |
+
"epoch": 0.48,
|
| 13883 |
+
"grad_norm": 3.97116756439209,
|
| 13884 |
+
"learning_rate": 7.193220338983051e-06,
|
| 13885 |
+
"loss": 1.4455,
|
| 13886 |
+
"step": 19390
|
| 13887 |
+
},
|
| 13888 |
+
{
|
| 13889 |
+
"epoch": 0.48,
|
| 13890 |
+
"grad_norm": 5.157166481018066,
|
| 13891 |
+
"learning_rate": 7.1864406779661025e-06,
|
| 13892 |
+
"loss": 1.3725,
|
| 13893 |
+
"step": 19400
|
| 13894 |
+
},
|
| 13895 |
+
{
|
| 13896 |
+
"epoch": 0.49,
|
| 13897 |
+
"grad_norm": 4.2230072021484375,
|
| 13898 |
+
"learning_rate": 7.179661016949153e-06,
|
| 13899 |
+
"loss": 1.3768,
|
| 13900 |
+
"step": 19410
|
| 13901 |
+
},
|
| 13902 |
+
{
|
| 13903 |
+
"epoch": 0.49,
|
| 13904 |
+
"grad_norm": 5.513181686401367,
|
| 13905 |
+
"learning_rate": 7.1728813559322044e-06,
|
| 13906 |
+
"loss": 1.3018,
|
| 13907 |
+
"step": 19420
|
| 13908 |
+
},
|
| 13909 |
+
{
|
| 13910 |
+
"epoch": 0.49,
|
| 13911 |
+
"grad_norm": 7.353570461273193,
|
| 13912 |
+
"learning_rate": 7.166101694915255e-06,
|
| 13913 |
+
"loss": 1.513,
|
| 13914 |
+
"step": 19430
|
| 13915 |
+
},
|
| 13916 |
+
{
|
| 13917 |
+
"epoch": 0.49,
|
| 13918 |
+
"grad_norm": 6.263894081115723,
|
| 13919 |
+
"learning_rate": 7.159322033898306e-06,
|
| 13920 |
+
"loss": 1.4521,
|
| 13921 |
+
"step": 19440
|
| 13922 |
+
},
|
| 13923 |
+
{
|
| 13924 |
+
"epoch": 0.49,
|
| 13925 |
+
"grad_norm": 13.368189811706543,
|
| 13926 |
+
"learning_rate": 7.152542372881357e-06,
|
| 13927 |
+
"loss": 1.3158,
|
| 13928 |
+
"step": 19450
|
| 13929 |
+
},
|
| 13930 |
+
{
|
| 13931 |
+
"epoch": 0.49,
|
| 13932 |
+
"grad_norm": 4.644199371337891,
|
| 13933 |
+
"learning_rate": 7.145762711864407e-06,
|
| 13934 |
+
"loss": 1.4184,
|
| 13935 |
+
"step": 19460
|
| 13936 |
+
},
|
| 13937 |
+
{
|
| 13938 |
+
"epoch": 0.49,
|
| 13939 |
+
"grad_norm": 5.206189155578613,
|
| 13940 |
+
"learning_rate": 7.138983050847458e-06,
|
| 13941 |
+
"loss": 1.3848,
|
| 13942 |
+
"step": 19470
|
| 13943 |
+
},
|
| 13944 |
+
{
|
| 13945 |
+
"epoch": 0.49,
|
| 13946 |
+
"grad_norm": 7.3628153800964355,
|
| 13947 |
+
"learning_rate": 7.132203389830509e-06,
|
| 13948 |
+
"loss": 1.2512,
|
| 13949 |
+
"step": 19480
|
| 13950 |
+
},
|
| 13951 |
+
{
|
| 13952 |
+
"epoch": 0.49,
|
| 13953 |
+
"grad_norm": 7.519322395324707,
|
| 13954 |
+
"learning_rate": 7.12542372881356e-06,
|
| 13955 |
+
"loss": 1.261,
|
| 13956 |
+
"step": 19490
|
| 13957 |
+
},
|
| 13958 |
+
{
|
| 13959 |
+
"epoch": 0.49,
|
| 13960 |
+
"grad_norm": 1.5350415706634521,
|
| 13961 |
+
"learning_rate": 7.1186440677966106e-06,
|
| 13962 |
+
"loss": 1.2517,
|
| 13963 |
+
"step": 19500
|
| 13964 |
+
},
|
| 13965 |
+
{
|
| 13966 |
+
"epoch": 0.49,
|
| 13967 |
+
"eval_loss": 1.3194345235824585,
|
| 13968 |
+
"eval_runtime": 66.1305,
|
| 13969 |
+
"eval_samples_per_second": 15.122,
|
| 13970 |
+
"eval_steps_per_second": 15.122,
|
| 13971 |
+
"step": 19500
|
| 13972 |
+
},
|
| 13973 |
+
{
|
| 13974 |
+
"epoch": 0.49,
|
| 13975 |
+
"grad_norm": 3.0648765563964844,
|
| 13976 |
+
"learning_rate": 7.111864406779662e-06,
|
| 13977 |
+
"loss": 1.398,
|
| 13978 |
+
"step": 19510
|
| 13979 |
+
},
|
| 13980 |
+
{
|
| 13981 |
+
"epoch": 0.49,
|
| 13982 |
+
"grad_norm": 5.124095439910889,
|
| 13983 |
+
"learning_rate": 7.1050847457627125e-06,
|
| 13984 |
+
"loss": 1.404,
|
| 13985 |
+
"step": 19520
|
| 13986 |
+
},
|
| 13987 |
+
{
|
| 13988 |
+
"epoch": 0.49,
|
| 13989 |
+
"grad_norm": 2.8961544036865234,
|
| 13990 |
+
"learning_rate": 7.098305084745764e-06,
|
| 13991 |
+
"loss": 1.2651,
|
| 13992 |
+
"step": 19530
|
| 13993 |
+
},
|
| 13994 |
+
{
|
| 13995 |
+
"epoch": 0.49,
|
| 13996 |
+
"grad_norm": 3.0641605854034424,
|
| 13997 |
+
"learning_rate": 7.0915254237288145e-06,
|
| 13998 |
+
"loss": 1.3587,
|
| 13999 |
+
"step": 19540
|
| 14000 |
+
},
|
| 14001 |
+
{
|
| 14002 |
+
"epoch": 0.49,
|
| 14003 |
+
"grad_norm": 3.618454933166504,
|
| 14004 |
+
"learning_rate": 7.084745762711865e-06,
|
| 14005 |
+
"loss": 1.3533,
|
| 14006 |
+
"step": 19550
|
| 14007 |
+
},
|
| 14008 |
+
{
|
| 14009 |
+
"epoch": 0.49,
|
| 14010 |
+
"grad_norm": 3.3083536624908447,
|
| 14011 |
+
"learning_rate": 7.077966101694916e-06,
|
| 14012 |
+
"loss": 1.2859,
|
| 14013 |
+
"step": 19560
|
| 14014 |
+
},
|
| 14015 |
+
{
|
| 14016 |
+
"epoch": 0.49,
|
| 14017 |
+
"grad_norm": 1.260048508644104,
|
| 14018 |
+
"learning_rate": 7.071186440677966e-06,
|
| 14019 |
+
"loss": 1.2224,
|
| 14020 |
+
"step": 19570
|
| 14021 |
+
},
|
| 14022 |
+
{
|
| 14023 |
+
"epoch": 0.49,
|
| 14024 |
+
"grad_norm": 5.410586833953857,
|
| 14025 |
+
"learning_rate": 7.0644067796610175e-06,
|
| 14026 |
+
"loss": 1.2916,
|
| 14027 |
+
"step": 19580
|
| 14028 |
+
},
|
| 14029 |
+
{
|
| 14030 |
+
"epoch": 0.49,
|
| 14031 |
+
"grad_norm": 1.9295905828475952,
|
| 14032 |
+
"learning_rate": 7.057627118644068e-06,
|
| 14033 |
+
"loss": 1.3004,
|
| 14034 |
+
"step": 19590
|
| 14035 |
+
},
|
| 14036 |
+
{
|
| 14037 |
+
"epoch": 0.49,
|
| 14038 |
+
"grad_norm": 10.48406982421875,
|
| 14039 |
+
"learning_rate": 7.0508474576271195e-06,
|
| 14040 |
+
"loss": 1.3105,
|
| 14041 |
+
"step": 19600
|
| 14042 |
+
},
|
| 14043 |
+
{
|
| 14044 |
+
"epoch": 0.49,
|
| 14045 |
+
"grad_norm": 4.99778413772583,
|
| 14046 |
+
"learning_rate": 7.04406779661017e-06,
|
| 14047 |
+
"loss": 1.2481,
|
| 14048 |
+
"step": 19610
|
| 14049 |
+
},
|
| 14050 |
+
{
|
| 14051 |
+
"epoch": 0.49,
|
| 14052 |
+
"grad_norm": 5.476469993591309,
|
| 14053 |
+
"learning_rate": 7.037288135593221e-06,
|
| 14054 |
+
"loss": 1.345,
|
| 14055 |
+
"step": 19620
|
| 14056 |
+
},
|
| 14057 |
+
{
|
| 14058 |
+
"epoch": 0.49,
|
| 14059 |
+
"grad_norm": 7.250665664672852,
|
| 14060 |
+
"learning_rate": 7.030508474576272e-06,
|
| 14061 |
+
"loss": 1.2993,
|
| 14062 |
+
"step": 19630
|
| 14063 |
+
},
|
| 14064 |
+
{
|
| 14065 |
+
"epoch": 0.49,
|
| 14066 |
+
"grad_norm": 7.802820682525635,
|
| 14067 |
+
"learning_rate": 7.0237288135593225e-06,
|
| 14068 |
+
"loss": 1.3106,
|
| 14069 |
+
"step": 19640
|
| 14070 |
+
},
|
| 14071 |
+
{
|
| 14072 |
+
"epoch": 0.49,
|
| 14073 |
+
"grad_norm": 2.5077409744262695,
|
| 14074 |
+
"learning_rate": 7.016949152542374e-06,
|
| 14075 |
+
"loss": 1.1298,
|
| 14076 |
+
"step": 19650
|
| 14077 |
+
},
|
| 14078 |
+
{
|
| 14079 |
+
"epoch": 0.49,
|
| 14080 |
+
"grad_norm": 2.3012547492980957,
|
| 14081 |
+
"learning_rate": 7.0101694915254245e-06,
|
| 14082 |
+
"loss": 1.2944,
|
| 14083 |
+
"step": 19660
|
| 14084 |
+
},
|
| 14085 |
+
{
|
| 14086 |
+
"epoch": 0.49,
|
| 14087 |
+
"grad_norm": 10.731738090515137,
|
| 14088 |
+
"learning_rate": 7.003389830508475e-06,
|
| 14089 |
+
"loss": 1.2545,
|
| 14090 |
+
"step": 19670
|
| 14091 |
+
},
|
| 14092 |
+
{
|
| 14093 |
+
"epoch": 0.49,
|
| 14094 |
+
"grad_norm": 11.851224899291992,
|
| 14095 |
+
"learning_rate": 6.996610169491526e-06,
|
| 14096 |
+
"loss": 1.3364,
|
| 14097 |
+
"step": 19680
|
| 14098 |
+
},
|
| 14099 |
+
{
|
| 14100 |
+
"epoch": 0.49,
|
| 14101 |
+
"grad_norm": 2.060750722885132,
|
| 14102 |
+
"learning_rate": 6.989830508474576e-06,
|
| 14103 |
+
"loss": 1.2909,
|
| 14104 |
+
"step": 19690
|
| 14105 |
+
},
|
| 14106 |
+
{
|
| 14107 |
+
"epoch": 0.49,
|
| 14108 |
+
"grad_norm": 8.549054145812988,
|
| 14109 |
+
"learning_rate": 6.9830508474576275e-06,
|
| 14110 |
+
"loss": 1.405,
|
| 14111 |
+
"step": 19700
|
| 14112 |
+
},
|
| 14113 |
+
{
|
| 14114 |
+
"epoch": 0.49,
|
| 14115 |
+
"grad_norm": 5.37472677230835,
|
| 14116 |
+
"learning_rate": 6.976271186440678e-06,
|
| 14117 |
+
"loss": 1.5615,
|
| 14118 |
+
"step": 19710
|
| 14119 |
+
},
|
| 14120 |
+
{
|
| 14121 |
+
"epoch": 0.49,
|
| 14122 |
+
"grad_norm": 4.753462314605713,
|
| 14123 |
+
"learning_rate": 6.9694915254237295e-06,
|
| 14124 |
+
"loss": 1.362,
|
| 14125 |
+
"step": 19720
|
| 14126 |
+
},
|
| 14127 |
+
{
|
| 14128 |
+
"epoch": 0.49,
|
| 14129 |
+
"grad_norm": 6.330374717712402,
|
| 14130 |
+
"learning_rate": 6.96271186440678e-06,
|
| 14131 |
+
"loss": 1.3563,
|
| 14132 |
+
"step": 19730
|
| 14133 |
+
},
|
| 14134 |
+
{
|
| 14135 |
+
"epoch": 0.49,
|
| 14136 |
+
"grad_norm": 6.962733268737793,
|
| 14137 |
+
"learning_rate": 6.9559322033898315e-06,
|
| 14138 |
+
"loss": 1.3272,
|
| 14139 |
+
"step": 19740
|
| 14140 |
+
},
|
| 14141 |
+
{
|
| 14142 |
+
"epoch": 0.49,
|
| 14143 |
+
"grad_norm": 13.372142791748047,
|
| 14144 |
+
"learning_rate": 6.949152542372882e-06,
|
| 14145 |
+
"loss": 1.4422,
|
| 14146 |
+
"step": 19750
|
| 14147 |
+
},
|
| 14148 |
+
{
|
| 14149 |
+
"epoch": 0.49,
|
| 14150 |
+
"grad_norm": 4.561896800994873,
|
| 14151 |
+
"learning_rate": 6.942372881355933e-06,
|
| 14152 |
+
"loss": 1.3691,
|
| 14153 |
+
"step": 19760
|
| 14154 |
+
},
|
| 14155 |
+
{
|
| 14156 |
+
"epoch": 0.49,
|
| 14157 |
+
"grad_norm": 2.2547922134399414,
|
| 14158 |
+
"learning_rate": 6.935593220338983e-06,
|
| 14159 |
+
"loss": 1.4905,
|
| 14160 |
+
"step": 19770
|
| 14161 |
+
},
|
| 14162 |
+
{
|
| 14163 |
+
"epoch": 0.49,
|
| 14164 |
+
"grad_norm": 4.15011739730835,
|
| 14165 |
+
"learning_rate": 6.928813559322034e-06,
|
| 14166 |
+
"loss": 1.3899,
|
| 14167 |
+
"step": 19780
|
| 14168 |
+
},
|
| 14169 |
+
{
|
| 14170 |
+
"epoch": 0.49,
|
| 14171 |
+
"grad_norm": 11.278037071228027,
|
| 14172 |
+
"learning_rate": 6.922033898305085e-06,
|
| 14173 |
+
"loss": 1.0803,
|
| 14174 |
+
"step": 19790
|
| 14175 |
+
},
|
| 14176 |
+
{
|
| 14177 |
+
"epoch": 0.49,
|
| 14178 |
+
"grad_norm": 9.100043296813965,
|
| 14179 |
+
"learning_rate": 6.915254237288136e-06,
|
| 14180 |
+
"loss": 1.4593,
|
| 14181 |
+
"step": 19800
|
| 14182 |
+
},
|
| 14183 |
+
{
|
| 14184 |
+
"epoch": 0.5,
|
| 14185 |
+
"grad_norm": 10.643383026123047,
|
| 14186 |
+
"learning_rate": 6.908474576271187e-06,
|
| 14187 |
+
"loss": 1.3571,
|
| 14188 |
+
"step": 19810
|
| 14189 |
+
},
|
| 14190 |
+
{
|
| 14191 |
+
"epoch": 0.5,
|
| 14192 |
+
"grad_norm": 11.314017295837402,
|
| 14193 |
+
"learning_rate": 6.901694915254238e-06,
|
| 14194 |
+
"loss": 1.2376,
|
| 14195 |
+
"step": 19820
|
| 14196 |
+
},
|
| 14197 |
+
{
|
| 14198 |
+
"epoch": 0.5,
|
| 14199 |
+
"grad_norm": 3.855220317840576,
|
| 14200 |
+
"learning_rate": 6.894915254237289e-06,
|
| 14201 |
+
"loss": 1.1095,
|
| 14202 |
+
"step": 19830
|
| 14203 |
+
},
|
| 14204 |
+
{
|
| 14205 |
+
"epoch": 0.5,
|
| 14206 |
+
"grad_norm": 5.896849155426025,
|
| 14207 |
+
"learning_rate": 6.8881355932203395e-06,
|
| 14208 |
+
"loss": 1.2078,
|
| 14209 |
+
"step": 19840
|
| 14210 |
+
},
|
| 14211 |
+
{
|
| 14212 |
+
"epoch": 0.5,
|
| 14213 |
+
"grad_norm": 2.1142632961273193,
|
| 14214 |
+
"learning_rate": 6.881355932203391e-06,
|
| 14215 |
+
"loss": 1.3017,
|
| 14216 |
+
"step": 19850
|
| 14217 |
+
},
|
| 14218 |
+
{
|
| 14219 |
+
"epoch": 0.5,
|
| 14220 |
+
"grad_norm": 7.116094589233398,
|
| 14221 |
+
"learning_rate": 6.8745762711864415e-06,
|
| 14222 |
+
"loss": 1.217,
|
| 14223 |
+
"step": 19860
|
| 14224 |
+
},
|
| 14225 |
+
{
|
| 14226 |
+
"epoch": 0.5,
|
| 14227 |
+
"grad_norm": 3.392282724380493,
|
| 14228 |
+
"learning_rate": 6.867796610169493e-06,
|
| 14229 |
+
"loss": 1.4027,
|
| 14230 |
+
"step": 19870
|
| 14231 |
+
},
|
| 14232 |
+
{
|
| 14233 |
+
"epoch": 0.5,
|
| 14234 |
+
"grad_norm": 5.760110855102539,
|
| 14235 |
+
"learning_rate": 6.861016949152543e-06,
|
| 14236 |
+
"loss": 1.2888,
|
| 14237 |
+
"step": 19880
|
| 14238 |
+
},
|
| 14239 |
+
{
|
| 14240 |
+
"epoch": 0.5,
|
| 14241 |
+
"grad_norm": 2.312904119491577,
|
| 14242 |
+
"learning_rate": 6.854237288135593e-06,
|
| 14243 |
+
"loss": 1.39,
|
| 14244 |
+
"step": 19890
|
| 14245 |
+
},
|
| 14246 |
+
{
|
| 14247 |
+
"epoch": 0.5,
|
| 14248 |
+
"grad_norm": 9.748230934143066,
|
| 14249 |
+
"learning_rate": 6.8474576271186445e-06,
|
| 14250 |
+
"loss": 1.408,
|
| 14251 |
+
"step": 19900
|
| 14252 |
+
},
|
| 14253 |
+
{
|
| 14254 |
+
"epoch": 0.5,
|
| 14255 |
+
"grad_norm": 8.029982566833496,
|
| 14256 |
+
"learning_rate": 6.840677966101695e-06,
|
| 14257 |
+
"loss": 1.2631,
|
| 14258 |
+
"step": 19910
|
| 14259 |
+
},
|
| 14260 |
+
{
|
| 14261 |
+
"epoch": 0.5,
|
| 14262 |
+
"grad_norm": 5.408463954925537,
|
| 14263 |
+
"learning_rate": 6.8338983050847465e-06,
|
| 14264 |
+
"loss": 1.3545,
|
| 14265 |
+
"step": 19920
|
| 14266 |
+
},
|
| 14267 |
+
{
|
| 14268 |
+
"epoch": 0.5,
|
| 14269 |
+
"grad_norm": 4.480403900146484,
|
| 14270 |
+
"learning_rate": 6.827118644067797e-06,
|
| 14271 |
+
"loss": 1.4549,
|
| 14272 |
+
"step": 19930
|
| 14273 |
+
},
|
| 14274 |
+
{
|
| 14275 |
+
"epoch": 0.5,
|
| 14276 |
+
"grad_norm": 8.981225967407227,
|
| 14277 |
+
"learning_rate": 6.8203389830508485e-06,
|
| 14278 |
+
"loss": 1.3352,
|
| 14279 |
+
"step": 19940
|
| 14280 |
+
},
|
| 14281 |
+
{
|
| 14282 |
+
"epoch": 0.5,
|
| 14283 |
+
"grad_norm": 1.9335066080093384,
|
| 14284 |
+
"learning_rate": 6.813559322033899e-06,
|
| 14285 |
+
"loss": 1.2825,
|
| 14286 |
+
"step": 19950
|
| 14287 |
+
},
|
| 14288 |
+
{
|
| 14289 |
+
"epoch": 0.5,
|
| 14290 |
+
"grad_norm": 7.6206464767456055,
|
| 14291 |
+
"learning_rate": 6.80677966101695e-06,
|
| 14292 |
+
"loss": 1.4976,
|
| 14293 |
+
"step": 19960
|
| 14294 |
+
},
|
| 14295 |
+
{
|
| 14296 |
+
"epoch": 0.5,
|
| 14297 |
+
"grad_norm": 14.687816619873047,
|
| 14298 |
+
"learning_rate": 6.800000000000001e-06,
|
| 14299 |
+
"loss": 1.4687,
|
| 14300 |
+
"step": 19970
|
| 14301 |
+
},
|
| 14302 |
+
{
|
| 14303 |
+
"epoch": 0.5,
|
| 14304 |
+
"grad_norm": 9.034219741821289,
|
| 14305 |
+
"learning_rate": 6.793220338983051e-06,
|
| 14306 |
+
"loss": 1.4715,
|
| 14307 |
+
"step": 19980
|
| 14308 |
+
},
|
| 14309 |
+
{
|
| 14310 |
+
"epoch": 0.5,
|
| 14311 |
+
"grad_norm": 8.120539665222168,
|
| 14312 |
+
"learning_rate": 6.786440677966102e-06,
|
| 14313 |
+
"loss": 1.4331,
|
| 14314 |
+
"step": 19990
|
| 14315 |
+
},
|
| 14316 |
+
{
|
| 14317 |
+
"epoch": 0.5,
|
| 14318 |
+
"grad_norm": 6.594362258911133,
|
| 14319 |
+
"learning_rate": 6.779661016949153e-06,
|
| 14320 |
+
"loss": 1.4516,
|
| 14321 |
+
"step": 20000
|
| 14322 |
+
},
|
| 14323 |
+
{
|
| 14324 |
+
"epoch": 0.5,
|
| 14325 |
+
"eval_loss": 1.348677158355713,
|
| 14326 |
+
"eval_runtime": 66.1188,
|
| 14327 |
+
"eval_samples_per_second": 15.124,
|
| 14328 |
+
"eval_steps_per_second": 15.124,
|
| 14329 |
+
"step": 20000
|
| 14330 |
}
|
| 14331 |
],
|
| 14332 |
"logging_steps": 10,
|
|
|
|
| 14334 |
"num_input_tokens_seen": 0,
|
| 14335 |
"num_train_epochs": 1,
|
| 14336 |
"save_steps": 2500,
|
| 14337 |
+
"total_flos": 3.2204251987968e+17,
|
| 14338 |
"train_batch_size": 1,
|
| 14339 |
"trial_name": null,
|
| 14340 |
"trial_params": null
|