Upload folder using huggingface_hub
Browse files- adapter_model.safetensors +1 -1
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1783 -0
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3158328
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e66e6b1c90ce6eb3fff7922af0316575c1c793bde5b3a21462828ed18cf91882
|
3 |
size 3158328
|
optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ccad07c0ed8a21da7b5ff443f024fe4090b9a1614db782964db12ec45a7e4bdd
|
3 |
+
size 1687034
|
rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:81f35d95292e4cf3d431d6701e989072077ed20b36d5f7d55b935fcc4ad16a06
|
3 |
+
size 14244
|
scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e02038e00bb5fbedaa0b48cd78170283d24119039429a860c32a8fd1d229193d
|
3 |
+
size 1064
|
trainer_state.json
ADDED
@@ -0,0 +1,1783 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 24.405125076266017,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 5000,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.09762050030506407,
|
13 |
+
"grad_norm": 0.5707364082336426,
|
14 |
+
"learning_rate": 4e-05,
|
15 |
+
"loss": 2.2566,
|
16 |
+
"step": 20
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.19524100061012814,
|
20 |
+
"grad_norm": 0.7393651008605957,
|
21 |
+
"learning_rate": 8e-05,
|
22 |
+
"loss": 2.203,
|
23 |
+
"step": 40
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.2928615009151922,
|
27 |
+
"grad_norm": 0.8320680856704712,
|
28 |
+
"learning_rate": 0.00012,
|
29 |
+
"loss": 1.9533,
|
30 |
+
"step": 60
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.3904820012202563,
|
34 |
+
"grad_norm": 1.2113606929779053,
|
35 |
+
"learning_rate": 0.00016,
|
36 |
+
"loss": 1.8878,
|
37 |
+
"step": 80
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.4881025015253203,
|
41 |
+
"grad_norm": 1.4215683937072754,
|
42 |
+
"learning_rate": 0.0002,
|
43 |
+
"loss": 1.7688,
|
44 |
+
"step": 100
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.5857230018303844,
|
48 |
+
"grad_norm": 1.8242409229278564,
|
49 |
+
"learning_rate": 0.0001998688524590164,
|
50 |
+
"loss": 1.7308,
|
51 |
+
"step": 120
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.6833435021354485,
|
55 |
+
"grad_norm": 1.6889506578445435,
|
56 |
+
"learning_rate": 0.0001997377049180328,
|
57 |
+
"loss": 1.6553,
|
58 |
+
"step": 140
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.7809640024405126,
|
62 |
+
"grad_norm": 2.1998350620269775,
|
63 |
+
"learning_rate": 0.00019960655737704918,
|
64 |
+
"loss": 1.5954,
|
65 |
+
"step": 160
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.8785845027455765,
|
69 |
+
"grad_norm": 1.5626238584518433,
|
70 |
+
"learning_rate": 0.0001994754098360656,
|
71 |
+
"loss": 1.5732,
|
72 |
+
"step": 180
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.9762050030506406,
|
76 |
+
"grad_norm": 2.350252389907837,
|
77 |
+
"learning_rate": 0.00019934426229508198,
|
78 |
+
"loss": 1.5001,
|
79 |
+
"step": 200
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 1.0738255033557047,
|
83 |
+
"grad_norm": 1.7499576807022095,
|
84 |
+
"learning_rate": 0.00019921311475409837,
|
85 |
+
"loss": 1.5056,
|
86 |
+
"step": 220
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 1.1714460036607688,
|
90 |
+
"grad_norm": 2.5555903911590576,
|
91 |
+
"learning_rate": 0.00019908196721311476,
|
92 |
+
"loss": 1.3946,
|
93 |
+
"step": 240
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 1.2690665039658329,
|
97 |
+
"grad_norm": 2.44691801071167,
|
98 |
+
"learning_rate": 0.00019895081967213115,
|
99 |
+
"loss": 1.4672,
|
100 |
+
"step": 260
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 1.366687004270897,
|
104 |
+
"grad_norm": 1.75604248046875,
|
105 |
+
"learning_rate": 0.00019881967213114757,
|
106 |
+
"loss": 1.4255,
|
107 |
+
"step": 280
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 1.4643075045759608,
|
111 |
+
"grad_norm": 2.170323610305786,
|
112 |
+
"learning_rate": 0.00019868852459016393,
|
113 |
+
"loss": 1.3473,
|
114 |
+
"step": 300
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 1.561928004881025,
|
118 |
+
"grad_norm": 2.3095695972442627,
|
119 |
+
"learning_rate": 0.00019855737704918035,
|
120 |
+
"loss": 1.3471,
|
121 |
+
"step": 320
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 1.659548505186089,
|
125 |
+
"grad_norm": 2.668290138244629,
|
126 |
+
"learning_rate": 0.00019842622950819674,
|
127 |
+
"loss": 1.2861,
|
128 |
+
"step": 340
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 1.757169005491153,
|
132 |
+
"grad_norm": 2.395897150039673,
|
133 |
+
"learning_rate": 0.00019829508196721313,
|
134 |
+
"loss": 1.2914,
|
135 |
+
"step": 360
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.8547895057962172,
|
139 |
+
"grad_norm": 2.7138261795043945,
|
140 |
+
"learning_rate": 0.00019816393442622951,
|
141 |
+
"loss": 1.3455,
|
142 |
+
"step": 380
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 1.9524100061012812,
|
146 |
+
"grad_norm": 2.28128981590271,
|
147 |
+
"learning_rate": 0.0001980327868852459,
|
148 |
+
"loss": 1.2907,
|
149 |
+
"step": 400
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 2.0500305064063453,
|
153 |
+
"grad_norm": 2.284174919128418,
|
154 |
+
"learning_rate": 0.00019790163934426232,
|
155 |
+
"loss": 1.2074,
|
156 |
+
"step": 420
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 2.1476510067114094,
|
160 |
+
"grad_norm": 2.7497994899749756,
|
161 |
+
"learning_rate": 0.00019777049180327868,
|
162 |
+
"loss": 1.2135,
|
163 |
+
"step": 440
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 2.2452715070164735,
|
167 |
+
"grad_norm": 3.0772790908813477,
|
168 |
+
"learning_rate": 0.0001976393442622951,
|
169 |
+
"loss": 1.1622,
|
170 |
+
"step": 460
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 2.3428920073215376,
|
174 |
+
"grad_norm": 2.3680825233459473,
|
175 |
+
"learning_rate": 0.0001975081967213115,
|
176 |
+
"loss": 1.2219,
|
177 |
+
"step": 480
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 2.4405125076266017,
|
181 |
+
"grad_norm": 3.132049798965454,
|
182 |
+
"learning_rate": 0.00019737704918032788,
|
183 |
+
"loss": 1.2523,
|
184 |
+
"step": 500
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 2.5381330079316657,
|
188 |
+
"grad_norm": 2.8542871475219727,
|
189 |
+
"learning_rate": 0.00019724590163934427,
|
190 |
+
"loss": 1.1163,
|
191 |
+
"step": 520
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 2.63575350823673,
|
195 |
+
"grad_norm": 2.5308499336242676,
|
196 |
+
"learning_rate": 0.00019711475409836066,
|
197 |
+
"loss": 1.1624,
|
198 |
+
"step": 540
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 2.733374008541794,
|
202 |
+
"grad_norm": 2.3678996562957764,
|
203 |
+
"learning_rate": 0.00019698360655737707,
|
204 |
+
"loss": 1.1856,
|
205 |
+
"step": 560
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 2.830994508846858,
|
209 |
+
"grad_norm": 3.0295815467834473,
|
210 |
+
"learning_rate": 0.00019685245901639344,
|
211 |
+
"loss": 1.1193,
|
212 |
+
"step": 580
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 2.9286150091519216,
|
216 |
+
"grad_norm": 3.7163286209106445,
|
217 |
+
"learning_rate": 0.00019672131147540985,
|
218 |
+
"loss": 1.1479,
|
219 |
+
"step": 600
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 3.026235509456986,
|
223 |
+
"grad_norm": 2.840359926223755,
|
224 |
+
"learning_rate": 0.00019659016393442624,
|
225 |
+
"loss": 1.1554,
|
226 |
+
"step": 620
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 3.1238560097620502,
|
230 |
+
"grad_norm": 2.579162836074829,
|
231 |
+
"learning_rate": 0.00019645901639344263,
|
232 |
+
"loss": 1.0088,
|
233 |
+
"step": 640
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 3.221476510067114,
|
237 |
+
"grad_norm": 2.772268295288086,
|
238 |
+
"learning_rate": 0.00019632786885245902,
|
239 |
+
"loss": 1.0412,
|
240 |
+
"step": 660
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 3.319097010372178,
|
244 |
+
"grad_norm": 3.2339484691619873,
|
245 |
+
"learning_rate": 0.0001961967213114754,
|
246 |
+
"loss": 1.043,
|
247 |
+
"step": 680
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 3.416717510677242,
|
251 |
+
"grad_norm": 3.7858259677886963,
|
252 |
+
"learning_rate": 0.00019606557377049183,
|
253 |
+
"loss": 1.1187,
|
254 |
+
"step": 700
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 3.514338010982306,
|
258 |
+
"grad_norm": 3.17150616645813,
|
259 |
+
"learning_rate": 0.00019593442622950822,
|
260 |
+
"loss": 1.0731,
|
261 |
+
"step": 720
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 3.61195851128737,
|
265 |
+
"grad_norm": 2.9254567623138428,
|
266 |
+
"learning_rate": 0.0001958032786885246,
|
267 |
+
"loss": 1.0242,
|
268 |
+
"step": 740
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 3.7095790115924343,
|
272 |
+
"grad_norm": 2.4400126934051514,
|
273 |
+
"learning_rate": 0.000195672131147541,
|
274 |
+
"loss": 1.0911,
|
275 |
+
"step": 760
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 3.8071995118974984,
|
279 |
+
"grad_norm": 2.7202625274658203,
|
280 |
+
"learning_rate": 0.00019554098360655738,
|
281 |
+
"loss": 1.0451,
|
282 |
+
"step": 780
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 3.9048200122025625,
|
286 |
+
"grad_norm": 3.0309741497039795,
|
287 |
+
"learning_rate": 0.00019540983606557377,
|
288 |
+
"loss": 1.0188,
|
289 |
+
"step": 800
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 4.002440512507627,
|
293 |
+
"grad_norm": 4.291023254394531,
|
294 |
+
"learning_rate": 0.00019527868852459016,
|
295 |
+
"loss": 1.0858,
|
296 |
+
"step": 820
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 4.100061012812691,
|
300 |
+
"grad_norm": 3.4548487663269043,
|
301 |
+
"learning_rate": 0.00019514754098360658,
|
302 |
+
"loss": 0.9553,
|
303 |
+
"step": 840
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 4.197681513117755,
|
307 |
+
"grad_norm": 2.882427453994751,
|
308 |
+
"learning_rate": 0.00019501639344262297,
|
309 |
+
"loss": 0.9491,
|
310 |
+
"step": 860
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 4.295302013422819,
|
314 |
+
"grad_norm": 3.2026467323303223,
|
315 |
+
"learning_rate": 0.00019488524590163936,
|
316 |
+
"loss": 0.9529,
|
317 |
+
"step": 880
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 4.392922513727883,
|
321 |
+
"grad_norm": 2.676392078399658,
|
322 |
+
"learning_rate": 0.00019475409836065575,
|
323 |
+
"loss": 0.9296,
|
324 |
+
"step": 900
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 4.490543014032947,
|
328 |
+
"grad_norm": 2.8527650833129883,
|
329 |
+
"learning_rate": 0.00019462295081967214,
|
330 |
+
"loss": 0.9826,
|
331 |
+
"step": 920
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 4.588163514338011,
|
335 |
+
"grad_norm": 2.832995891571045,
|
336 |
+
"learning_rate": 0.00019449180327868855,
|
337 |
+
"loss": 0.9337,
|
338 |
+
"step": 940
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 4.685784014643075,
|
342 |
+
"grad_norm": 2.9931800365448,
|
343 |
+
"learning_rate": 0.00019436065573770491,
|
344 |
+
"loss": 0.9455,
|
345 |
+
"step": 960
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 4.783404514948139,
|
349 |
+
"grad_norm": 3.4157402515411377,
|
350 |
+
"learning_rate": 0.00019422950819672133,
|
351 |
+
"loss": 1.0286,
|
352 |
+
"step": 980
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 4.881025015253203,
|
356 |
+
"grad_norm": 2.846972942352295,
|
357 |
+
"learning_rate": 0.00019409836065573772,
|
358 |
+
"loss": 0.9536,
|
359 |
+
"step": 1000
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 4.978645515558267,
|
363 |
+
"grad_norm": 3.4363763332366943,
|
364 |
+
"learning_rate": 0.0001939672131147541,
|
365 |
+
"loss": 0.967,
|
366 |
+
"step": 1020
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 5.0762660158633315,
|
370 |
+
"grad_norm": 3.347905397415161,
|
371 |
+
"learning_rate": 0.0001938360655737705,
|
372 |
+
"loss": 0.9291,
|
373 |
+
"step": 1040
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 5.173886516168396,
|
377 |
+
"grad_norm": 3.016284704208374,
|
378 |
+
"learning_rate": 0.0001937049180327869,
|
379 |
+
"loss": 0.8722,
|
380 |
+
"step": 1060
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 5.27150701647346,
|
384 |
+
"grad_norm": 3.5366976261138916,
|
385 |
+
"learning_rate": 0.0001935737704918033,
|
386 |
+
"loss": 0.8839,
|
387 |
+
"step": 1080
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 5.369127516778524,
|
391 |
+
"grad_norm": 3.550899028778076,
|
392 |
+
"learning_rate": 0.00019344262295081967,
|
393 |
+
"loss": 0.8386,
|
394 |
+
"step": 1100
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 5.466748017083588,
|
398 |
+
"grad_norm": 3.651362180709839,
|
399 |
+
"learning_rate": 0.00019331147540983608,
|
400 |
+
"loss": 0.8998,
|
401 |
+
"step": 1120
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 5.564368517388652,
|
405 |
+
"grad_norm": 2.9918806552886963,
|
406 |
+
"learning_rate": 0.00019318032786885247,
|
407 |
+
"loss": 0.8472,
|
408 |
+
"step": 1140
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 5.661989017693716,
|
412 |
+
"grad_norm": 3.7066636085510254,
|
413 |
+
"learning_rate": 0.00019304918032786886,
|
414 |
+
"loss": 0.8966,
|
415 |
+
"step": 1160
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 5.75960951799878,
|
419 |
+
"grad_norm": 3.1231799125671387,
|
420 |
+
"learning_rate": 0.00019291803278688525,
|
421 |
+
"loss": 0.8989,
|
422 |
+
"step": 1180
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 5.857230018303844,
|
426 |
+
"grad_norm": 3.0318100452423096,
|
427 |
+
"learning_rate": 0.00019278688524590164,
|
428 |
+
"loss": 0.9222,
|
429 |
+
"step": 1200
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 5.954850518608908,
|
433 |
+
"grad_norm": 3.3124337196350098,
|
434 |
+
"learning_rate": 0.00019265573770491806,
|
435 |
+
"loss": 0.8827,
|
436 |
+
"step": 1220
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 6.052471018913972,
|
440 |
+
"grad_norm": 3.1748123168945312,
|
441 |
+
"learning_rate": 0.00019252459016393442,
|
442 |
+
"loss": 0.845,
|
443 |
+
"step": 1240
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 6.150091519219036,
|
447 |
+
"grad_norm": 3.320768356323242,
|
448 |
+
"learning_rate": 0.00019239344262295084,
|
449 |
+
"loss": 0.7315,
|
450 |
+
"step": 1260
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 6.2477120195241005,
|
454 |
+
"grad_norm": 4.226164817810059,
|
455 |
+
"learning_rate": 0.00019226229508196723,
|
456 |
+
"loss": 0.7999,
|
457 |
+
"step": 1280
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 6.345332519829164,
|
461 |
+
"grad_norm": 3.12298583984375,
|
462 |
+
"learning_rate": 0.00019213114754098362,
|
463 |
+
"loss": 0.8337,
|
464 |
+
"step": 1300
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 6.442953020134228,
|
468 |
+
"grad_norm": 3.4633209705352783,
|
469 |
+
"learning_rate": 0.000192,
|
470 |
+
"loss": 0.7834,
|
471 |
+
"step": 1320
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 6.540573520439292,
|
475 |
+
"grad_norm": 3.116576671600342,
|
476 |
+
"learning_rate": 0.0001918688524590164,
|
477 |
+
"loss": 0.8386,
|
478 |
+
"step": 1340
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 6.638194020744356,
|
482 |
+
"grad_norm": 3.255690336227417,
|
483 |
+
"learning_rate": 0.0001917377049180328,
|
484 |
+
"loss": 0.902,
|
485 |
+
"step": 1360
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 6.73581452104942,
|
489 |
+
"grad_norm": 4.117364406585693,
|
490 |
+
"learning_rate": 0.00019160655737704917,
|
491 |
+
"loss": 0.7897,
|
492 |
+
"step": 1380
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 6.833435021354484,
|
496 |
+
"grad_norm": 3.813042163848877,
|
497 |
+
"learning_rate": 0.0001914754098360656,
|
498 |
+
"loss": 0.7975,
|
499 |
+
"step": 1400
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 6.931055521659548,
|
503 |
+
"grad_norm": 3.2562716007232666,
|
504 |
+
"learning_rate": 0.00019134426229508198,
|
505 |
+
"loss": 0.885,
|
506 |
+
"step": 1420
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 7.028676021964612,
|
510 |
+
"grad_norm": 3.0337297916412354,
|
511 |
+
"learning_rate": 0.00019121311475409837,
|
512 |
+
"loss": 0.7571,
|
513 |
+
"step": 1440
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 7.126296522269676,
|
517 |
+
"grad_norm": 3.4187958240509033,
|
518 |
+
"learning_rate": 0.00019108196721311476,
|
519 |
+
"loss": 0.7317,
|
520 |
+
"step": 1460
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 7.22391702257474,
|
524 |
+
"grad_norm": 3.2497494220733643,
|
525 |
+
"learning_rate": 0.00019095081967213115,
|
526 |
+
"loss": 0.6953,
|
527 |
+
"step": 1480
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 7.3215375228798045,
|
531 |
+
"grad_norm": 3.452319383621216,
|
532 |
+
"learning_rate": 0.00019081967213114756,
|
533 |
+
"loss": 0.7026,
|
534 |
+
"step": 1500
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 7.419158023184869,
|
538 |
+
"grad_norm": 4.487509250640869,
|
539 |
+
"learning_rate": 0.00019068852459016395,
|
540 |
+
"loss": 0.7552,
|
541 |
+
"step": 1520
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 7.516778523489933,
|
545 |
+
"grad_norm": 3.4719717502593994,
|
546 |
+
"learning_rate": 0.00019055737704918034,
|
547 |
+
"loss": 0.7426,
|
548 |
+
"step": 1540
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 7.614399023794997,
|
552 |
+
"grad_norm": 3.9266843795776367,
|
553 |
+
"learning_rate": 0.00019042622950819673,
|
554 |
+
"loss": 0.762,
|
555 |
+
"step": 1560
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 7.712019524100061,
|
559 |
+
"grad_norm": 3.0717546939849854,
|
560 |
+
"learning_rate": 0.00019029508196721312,
|
561 |
+
"loss": 0.7574,
|
562 |
+
"step": 1580
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 7.809640024405125,
|
566 |
+
"grad_norm": 3.793149948120117,
|
567 |
+
"learning_rate": 0.0001901639344262295,
|
568 |
+
"loss": 0.7817,
|
569 |
+
"step": 1600
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 7.907260524710189,
|
573 |
+
"grad_norm": 3.288499593734741,
|
574 |
+
"learning_rate": 0.0001900327868852459,
|
575 |
+
"loss": 0.7759,
|
576 |
+
"step": 1620
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 8.004881025015253,
|
580 |
+
"grad_norm": 2.6891720294952393,
|
581 |
+
"learning_rate": 0.00018990163934426232,
|
582 |
+
"loss": 0.8212,
|
583 |
+
"step": 1640
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 8.102501525320317,
|
587 |
+
"grad_norm": 3.331376075744629,
|
588 |
+
"learning_rate": 0.0001897704918032787,
|
589 |
+
"loss": 0.6659,
|
590 |
+
"step": 1660
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 8.200122025625381,
|
594 |
+
"grad_norm": 3.3468194007873535,
|
595 |
+
"learning_rate": 0.0001896393442622951,
|
596 |
+
"loss": 0.6458,
|
597 |
+
"step": 1680
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 8.297742525930445,
|
601 |
+
"grad_norm": 4.916980266571045,
|
602 |
+
"learning_rate": 0.00018950819672131148,
|
603 |
+
"loss": 0.6748,
|
604 |
+
"step": 1700
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 8.39536302623551,
|
608 |
+
"grad_norm": 3.4678823947906494,
|
609 |
+
"learning_rate": 0.00018937704918032787,
|
610 |
+
"loss": 0.6349,
|
611 |
+
"step": 1720
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 8.492983526540574,
|
615 |
+
"grad_norm": 4.540040493011475,
|
616 |
+
"learning_rate": 0.0001892459016393443,
|
617 |
+
"loss": 0.7256,
|
618 |
+
"step": 1740
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 8.590604026845638,
|
622 |
+
"grad_norm": 3.6092071533203125,
|
623 |
+
"learning_rate": 0.00018911475409836065,
|
624 |
+
"loss": 0.6708,
|
625 |
+
"step": 1760
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 8.688224527150702,
|
629 |
+
"grad_norm": 4.159832954406738,
|
630 |
+
"learning_rate": 0.00018898360655737707,
|
631 |
+
"loss": 0.6974,
|
632 |
+
"step": 1780
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 8.785845027455766,
|
636 |
+
"grad_norm": 3.980943202972412,
|
637 |
+
"learning_rate": 0.00018885245901639346,
|
638 |
+
"loss": 0.7268,
|
639 |
+
"step": 1800
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 8.88346552776083,
|
643 |
+
"grad_norm": 3.106861114501953,
|
644 |
+
"learning_rate": 0.00018872131147540985,
|
645 |
+
"loss": 0.7583,
|
646 |
+
"step": 1820
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 8.981086028065894,
|
650 |
+
"grad_norm": 2.672461748123169,
|
651 |
+
"learning_rate": 0.00018859016393442624,
|
652 |
+
"loss": 0.7199,
|
653 |
+
"step": 1840
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 9.078706528370958,
|
657 |
+
"grad_norm": 3.1480796337127686,
|
658 |
+
"learning_rate": 0.00018845901639344263,
|
659 |
+
"loss": 0.6583,
|
660 |
+
"step": 1860
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 9.176327028676022,
|
664 |
+
"grad_norm": 3.3770973682403564,
|
665 |
+
"learning_rate": 0.00018832786885245904,
|
666 |
+
"loss": 0.6317,
|
667 |
+
"step": 1880
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 9.273947528981086,
|
671 |
+
"grad_norm": 3.452083110809326,
|
672 |
+
"learning_rate": 0.0001881967213114754,
|
673 |
+
"loss": 0.6166,
|
674 |
+
"step": 1900
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 9.37156802928615,
|
678 |
+
"grad_norm": 3.6544485092163086,
|
679 |
+
"learning_rate": 0.00018806557377049182,
|
680 |
+
"loss": 0.6128,
|
681 |
+
"step": 1920
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 9.469188529591214,
|
685 |
+
"grad_norm": 3.462320566177368,
|
686 |
+
"learning_rate": 0.0001879344262295082,
|
687 |
+
"loss": 0.6442,
|
688 |
+
"step": 1940
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 9.566809029896278,
|
692 |
+
"grad_norm": 3.0532991886138916,
|
693 |
+
"learning_rate": 0.0001878032786885246,
|
694 |
+
"loss": 0.6631,
|
695 |
+
"step": 1960
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 9.664429530201343,
|
699 |
+
"grad_norm": 4.109907627105713,
|
700 |
+
"learning_rate": 0.000187672131147541,
|
701 |
+
"loss": 0.6634,
|
702 |
+
"step": 1980
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 9.762050030506407,
|
706 |
+
"grad_norm": 3.7715537548065186,
|
707 |
+
"learning_rate": 0.00018754098360655738,
|
708 |
+
"loss": 0.6004,
|
709 |
+
"step": 2000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 9.85967053081147,
|
713 |
+
"grad_norm": 4.588443756103516,
|
714 |
+
"learning_rate": 0.0001874098360655738,
|
715 |
+
"loss": 0.6816,
|
716 |
+
"step": 2020
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 9.957291031116535,
|
720 |
+
"grad_norm": 3.2902603149414062,
|
721 |
+
"learning_rate": 0.00018727868852459016,
|
722 |
+
"loss": 0.682,
|
723 |
+
"step": 2040
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 10.054911531421599,
|
727 |
+
"grad_norm": 3.070920705795288,
|
728 |
+
"learning_rate": 0.00018714754098360657,
|
729 |
+
"loss": 0.6007,
|
730 |
+
"step": 2060
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 10.152532031726663,
|
734 |
+
"grad_norm": 3.761850357055664,
|
735 |
+
"learning_rate": 0.00018701639344262296,
|
736 |
+
"loss": 0.5478,
|
737 |
+
"step": 2080
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 10.250152532031727,
|
741 |
+
"grad_norm": 2.9650540351867676,
|
742 |
+
"learning_rate": 0.00018688524590163935,
|
743 |
+
"loss": 0.5792,
|
744 |
+
"step": 2100
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 10.347773032336791,
|
748 |
+
"grad_norm": 3.3296761512756348,
|
749 |
+
"learning_rate": 0.00018675409836065574,
|
750 |
+
"loss": 0.5849,
|
751 |
+
"step": 2120
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 10.445393532641855,
|
755 |
+
"grad_norm": 4.055027008056641,
|
756 |
+
"learning_rate": 0.00018662295081967213,
|
757 |
+
"loss": 0.5417,
|
758 |
+
"step": 2140
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 10.54301403294692,
|
762 |
+
"grad_norm": 3.581878662109375,
|
763 |
+
"learning_rate": 0.00018649180327868855,
|
764 |
+
"loss": 0.6285,
|
765 |
+
"step": 2160
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 10.640634533251983,
|
769 |
+
"grad_norm": 3.667356491088867,
|
770 |
+
"learning_rate": 0.00018636065573770494,
|
771 |
+
"loss": 0.5767,
|
772 |
+
"step": 2180
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 10.738255033557047,
|
776 |
+
"grad_norm": 4.004351615905762,
|
777 |
+
"learning_rate": 0.00018622950819672133,
|
778 |
+
"loss": 0.6339,
|
779 |
+
"step": 2200
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 10.835875533862112,
|
783 |
+
"grad_norm": 3.4327890872955322,
|
784 |
+
"learning_rate": 0.00018609836065573772,
|
785 |
+
"loss": 0.6258,
|
786 |
+
"step": 2220
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 10.933496034167176,
|
790 |
+
"grad_norm": 3.3382508754730225,
|
791 |
+
"learning_rate": 0.0001859672131147541,
|
792 |
+
"loss": 0.641,
|
793 |
+
"step": 2240
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 11.03111653447224,
|
797 |
+
"grad_norm": 2.9254095554351807,
|
798 |
+
"learning_rate": 0.0001858360655737705,
|
799 |
+
"loss": 0.5932,
|
800 |
+
"step": 2260
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 11.128737034777304,
|
804 |
+
"grad_norm": 4.12520694732666,
|
805 |
+
"learning_rate": 0.00018570491803278688,
|
806 |
+
"loss": 0.5007,
|
807 |
+
"step": 2280
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 11.226357535082368,
|
811 |
+
"grad_norm": 3.348695993423462,
|
812 |
+
"learning_rate": 0.0001855737704918033,
|
813 |
+
"loss": 0.4967,
|
814 |
+
"step": 2300
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 11.323978035387432,
|
818 |
+
"grad_norm": 3.62343430519104,
|
819 |
+
"learning_rate": 0.0001854426229508197,
|
820 |
+
"loss": 0.5312,
|
821 |
+
"step": 2320
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 11.421598535692496,
|
825 |
+
"grad_norm": 3.4885001182556152,
|
826 |
+
"learning_rate": 0.00018531147540983608,
|
827 |
+
"loss": 0.5675,
|
828 |
+
"step": 2340
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 11.51921903599756,
|
832 |
+
"grad_norm": 4.001424789428711,
|
833 |
+
"learning_rate": 0.00018518032786885247,
|
834 |
+
"loss": 0.5654,
|
835 |
+
"step": 2360
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 11.616839536302624,
|
839 |
+
"grad_norm": 3.180741310119629,
|
840 |
+
"learning_rate": 0.00018504918032786886,
|
841 |
+
"loss": 0.5473,
|
842 |
+
"step": 2380
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 11.714460036607688,
|
846 |
+
"grad_norm": 3.3385472297668457,
|
847 |
+
"learning_rate": 0.00018491803278688527,
|
848 |
+
"loss": 0.601,
|
849 |
+
"step": 2400
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 11.812080536912752,
|
853 |
+
"grad_norm": 3.1975746154785156,
|
854 |
+
"learning_rate": 0.00018478688524590164,
|
855 |
+
"loss": 0.559,
|
856 |
+
"step": 2420
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 11.909701037217816,
|
860 |
+
"grad_norm": 4.126131057739258,
|
861 |
+
"learning_rate": 0.00018465573770491805,
|
862 |
+
"loss": 0.517,
|
863 |
+
"step": 2440
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 12.00732153752288,
|
867 |
+
"grad_norm": 3.1047425270080566,
|
868 |
+
"learning_rate": 0.00018452459016393444,
|
869 |
+
"loss": 0.5853,
|
870 |
+
"step": 2460
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 12.104942037827945,
|
874 |
+
"grad_norm": 4.153341293334961,
|
875 |
+
"learning_rate": 0.0001843934426229508,
|
876 |
+
"loss": 0.4683,
|
877 |
+
"step": 2480
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 12.202562538133009,
|
881 |
+
"grad_norm": 3.550233840942383,
|
882 |
+
"learning_rate": 0.00018426229508196722,
|
883 |
+
"loss": 0.4802,
|
884 |
+
"step": 2500
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 12.300183038438073,
|
888 |
+
"grad_norm": 3.3238728046417236,
|
889 |
+
"learning_rate": 0.0001841311475409836,
|
890 |
+
"loss": 0.5219,
|
891 |
+
"step": 2520
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 12.397803538743137,
|
895 |
+
"grad_norm": 4.092185974121094,
|
896 |
+
"learning_rate": 0.00018400000000000003,
|
897 |
+
"loss": 0.4783,
|
898 |
+
"step": 2540
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 12.495424039048201,
|
902 |
+
"grad_norm": 3.4403531551361084,
|
903 |
+
"learning_rate": 0.0001838688524590164,
|
904 |
+
"loss": 0.5247,
|
905 |
+
"step": 2560
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 12.593044539353265,
|
909 |
+
"grad_norm": 4.026747226715088,
|
910 |
+
"learning_rate": 0.0001837377049180328,
|
911 |
+
"loss": 0.4838,
|
912 |
+
"step": 2580
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 12.690665039658327,
|
916 |
+
"grad_norm": 2.7140889167785645,
|
917 |
+
"learning_rate": 0.0001836065573770492,
|
918 |
+
"loss": 0.5417,
|
919 |
+
"step": 2600
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 12.788285539963393,
|
923 |
+
"grad_norm": 4.172866344451904,
|
924 |
+
"learning_rate": 0.00018347540983606558,
|
925 |
+
"loss": 0.5133,
|
926 |
+
"step": 2620
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 12.885906040268456,
|
930 |
+
"grad_norm": 4.674078464508057,
|
931 |
+
"learning_rate": 0.00018334426229508197,
|
932 |
+
"loss": 0.5093,
|
933 |
+
"step": 2640
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 12.98352654057352,
|
937 |
+
"grad_norm": 3.6748011112213135,
|
938 |
+
"learning_rate": 0.00018321311475409836,
|
939 |
+
"loss": 0.5365,
|
940 |
+
"step": 2660
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 13.081147040878584,
|
944 |
+
"grad_norm": 5.05755615234375,
|
945 |
+
"learning_rate": 0.00018308196721311478,
|
946 |
+
"loss": 0.4356,
|
947 |
+
"step": 2680
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 13.178767541183648,
|
951 |
+
"grad_norm": 3.9897286891937256,
|
952 |
+
"learning_rate": 0.00018295081967213114,
|
953 |
+
"loss": 0.4228,
|
954 |
+
"step": 2700
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 13.276388041488712,
|
958 |
+
"grad_norm": 3.8581581115722656,
|
959 |
+
"learning_rate": 0.00018281967213114756,
|
960 |
+
"loss": 0.4019,
|
961 |
+
"step": 2720
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 13.374008541793776,
|
965 |
+
"grad_norm": 3.3325164318084717,
|
966 |
+
"learning_rate": 0.00018268852459016395,
|
967 |
+
"loss": 0.4715,
|
968 |
+
"step": 2740
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 13.47162904209884,
|
972 |
+
"grad_norm": 3.3978488445281982,
|
973 |
+
"learning_rate": 0.00018255737704918034,
|
974 |
+
"loss": 0.4634,
|
975 |
+
"step": 2760
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 13.569249542403904,
|
979 |
+
"grad_norm": 3.491314172744751,
|
980 |
+
"learning_rate": 0.00018242622950819673,
|
981 |
+
"loss": 0.4942,
|
982 |
+
"step": 2780
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 13.666870042708968,
|
986 |
+
"grad_norm": 3.275783061981201,
|
987 |
+
"learning_rate": 0.00018229508196721312,
|
988 |
+
"loss": 0.4807,
|
989 |
+
"step": 2800
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 13.764490543014032,
|
993 |
+
"grad_norm": 4.8541717529296875,
|
994 |
+
"learning_rate": 0.00018216393442622953,
|
995 |
+
"loss": 0.4684,
|
996 |
+
"step": 2820
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 13.862111043319096,
|
1000 |
+
"grad_norm": 4.203775882720947,
|
1001 |
+
"learning_rate": 0.00018203278688524592,
|
1002 |
+
"loss": 0.515,
|
1003 |
+
"step": 2840
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 13.95973154362416,
|
1007 |
+
"grad_norm": 3.885730504989624,
|
1008 |
+
"learning_rate": 0.0001819016393442623,
|
1009 |
+
"loss": 0.5263,
|
1010 |
+
"step": 2860
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 14.057352043929225,
|
1014 |
+
"grad_norm": 4.247331619262695,
|
1015 |
+
"learning_rate": 0.0001817704918032787,
|
1016 |
+
"loss": 0.4553,
|
1017 |
+
"step": 2880
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 14.154972544234289,
|
1021 |
+
"grad_norm": 4.659804344177246,
|
1022 |
+
"learning_rate": 0.0001816393442622951,
|
1023 |
+
"loss": 0.403,
|
1024 |
+
"step": 2900
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 14.252593044539353,
|
1028 |
+
"grad_norm": 3.525756359100342,
|
1029 |
+
"learning_rate": 0.00018150819672131148,
|
1030 |
+
"loss": 0.4128,
|
1031 |
+
"step": 2920
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 14.350213544844417,
|
1035 |
+
"grad_norm": 3.7535457611083984,
|
1036 |
+
"learning_rate": 0.00018137704918032787,
|
1037 |
+
"loss": 0.3969,
|
1038 |
+
"step": 2940
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 14.44783404514948,
|
1042 |
+
"grad_norm": 3.828665256500244,
|
1043 |
+
"learning_rate": 0.00018124590163934429,
|
1044 |
+
"loss": 0.4321,
|
1045 |
+
"step": 2960
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 14.545454545454545,
|
1049 |
+
"grad_norm": 3.7969393730163574,
|
1050 |
+
"learning_rate": 0.00018111475409836067,
|
1051 |
+
"loss": 0.4382,
|
1052 |
+
"step": 2980
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 14.643075045759609,
|
1056 |
+
"grad_norm": 4.983769416809082,
|
1057 |
+
"learning_rate": 0.00018098360655737704,
|
1058 |
+
"loss": 0.4656,
|
1059 |
+
"step": 3000
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 14.740695546064673,
|
1063 |
+
"grad_norm": 3.585157632827759,
|
1064 |
+
"learning_rate": 0.00018085245901639345,
|
1065 |
+
"loss": 0.4576,
|
1066 |
+
"step": 3020
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 14.838316046369737,
|
1070 |
+
"grad_norm": 3.859992265701294,
|
1071 |
+
"learning_rate": 0.00018072131147540984,
|
1072 |
+
"loss": 0.4712,
|
1073 |
+
"step": 3040
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 14.935936546674801,
|
1077 |
+
"grad_norm": 3.2431323528289795,
|
1078 |
+
"learning_rate": 0.00018059016393442626,
|
1079 |
+
"loss": 0.4526,
|
1080 |
+
"step": 3060
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 15.033557046979865,
|
1084 |
+
"grad_norm": 3.1578800678253174,
|
1085 |
+
"learning_rate": 0.00018045901639344262,
|
1086 |
+
"loss": 0.4129,
|
1087 |
+
"step": 3080
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 15.13117754728493,
|
1091 |
+
"grad_norm": 4.049219608306885,
|
1092 |
+
"learning_rate": 0.00018032786885245904,
|
1093 |
+
"loss": 0.3485,
|
1094 |
+
"step": 3100
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 15.228798047589994,
|
1098 |
+
"grad_norm": 3.642855167388916,
|
1099 |
+
"learning_rate": 0.00018019672131147543,
|
1100 |
+
"loss": 0.3607,
|
1101 |
+
"step": 3120
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 15.326418547895058,
|
1105 |
+
"grad_norm": 3.5239570140838623,
|
1106 |
+
"learning_rate": 0.0001800655737704918,
|
1107 |
+
"loss": 0.3844,
|
1108 |
+
"step": 3140
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 15.424039048200122,
|
1112 |
+
"grad_norm": 2.894887685775757,
|
1113 |
+
"learning_rate": 0.0001799344262295082,
|
1114 |
+
"loss": 0.4111,
|
1115 |
+
"step": 3160
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 15.521659548505186,
|
1119 |
+
"grad_norm": 3.1862661838531494,
|
1120 |
+
"learning_rate": 0.0001798032786885246,
|
1121 |
+
"loss": 0.4,
|
1122 |
+
"step": 3180
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 15.61928004881025,
|
1126 |
+
"grad_norm": 3.169221878051758,
|
1127 |
+
"learning_rate": 0.000179672131147541,
|
1128 |
+
"loss": 0.3914,
|
1129 |
+
"step": 3200
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 15.716900549115314,
|
1133 |
+
"grad_norm": 4.413252830505371,
|
1134 |
+
"learning_rate": 0.00017954098360655737,
|
1135 |
+
"loss": 0.4268,
|
1136 |
+
"step": 3220
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 15.814521049420378,
|
1140 |
+
"grad_norm": 4.1161208152771,
|
1141 |
+
"learning_rate": 0.0001794098360655738,
|
1142 |
+
"loss": 0.4261,
|
1143 |
+
"step": 3240
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 15.912141549725442,
|
1147 |
+
"grad_norm": 3.044705867767334,
|
1148 |
+
"learning_rate": 0.00017927868852459018,
|
1149 |
+
"loss": 0.4628,
|
1150 |
+
"step": 3260
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 16.009762050030506,
|
1154 |
+
"grad_norm": 3.2134783267974854,
|
1155 |
+
"learning_rate": 0.00017914754098360657,
|
1156 |
+
"loss": 0.4368,
|
1157 |
+
"step": 3280
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 16.107382550335572,
|
1161 |
+
"grad_norm": 3.8068933486938477,
|
1162 |
+
"learning_rate": 0.00017901639344262296,
|
1163 |
+
"loss": 0.2993,
|
1164 |
+
"step": 3300
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 16.205003050640634,
|
1168 |
+
"grad_norm": 4.097984313964844,
|
1169 |
+
"learning_rate": 0.00017888524590163935,
|
1170 |
+
"loss": 0.3458,
|
1171 |
+
"step": 3320
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 16.3026235509457,
|
1175 |
+
"grad_norm": 3.5351762771606445,
|
1176 |
+
"learning_rate": 0.00017875409836065576,
|
1177 |
+
"loss": 0.3449,
|
1178 |
+
"step": 3340
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 16.400244051250763,
|
1182 |
+
"grad_norm": 3.3672287464141846,
|
1183 |
+
"learning_rate": 0.00017862295081967213,
|
1184 |
+
"loss": 0.3708,
|
1185 |
+
"step": 3360
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 16.49786455155583,
|
1189 |
+
"grad_norm": 3.6430435180664062,
|
1190 |
+
"learning_rate": 0.00017849180327868852,
|
1191 |
+
"loss": 0.369,
|
1192 |
+
"step": 3380
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 16.59548505186089,
|
1196 |
+
"grad_norm": 4.193872928619385,
|
1197 |
+
"learning_rate": 0.00017836065573770493,
|
1198 |
+
"loss": 0.4063,
|
1199 |
+
"step": 3400
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 16.693105552165953,
|
1203 |
+
"grad_norm": 3.818737506866455,
|
1204 |
+
"learning_rate": 0.00017822950819672132,
|
1205 |
+
"loss": 0.3777,
|
1206 |
+
"step": 3420
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 16.79072605247102,
|
1210 |
+
"grad_norm": 3.7047250270843506,
|
1211 |
+
"learning_rate": 0.0001780983606557377,
|
1212 |
+
"loss": 0.3939,
|
1213 |
+
"step": 3440
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 16.888346552776085,
|
1217 |
+
"grad_norm": 4.005129337310791,
|
1218 |
+
"learning_rate": 0.0001779672131147541,
|
1219 |
+
"loss": 0.4125,
|
1220 |
+
"step": 3460
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 16.985967053081147,
|
1224 |
+
"grad_norm": 4.845798015594482,
|
1225 |
+
"learning_rate": 0.00017783606557377052,
|
1226 |
+
"loss": 0.428,
|
1227 |
+
"step": 3480
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 17.08358755338621,
|
1231 |
+
"grad_norm": 4.9272894859313965,
|
1232 |
+
"learning_rate": 0.0001777049180327869,
|
1233 |
+
"loss": 0.3169,
|
1234 |
+
"step": 3500
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 17.181208053691275,
|
1238 |
+
"grad_norm": 3.334076166152954,
|
1239 |
+
"learning_rate": 0.00017757377049180327,
|
1240 |
+
"loss": 0.336,
|
1241 |
+
"step": 3520
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 17.278828553996338,
|
1245 |
+
"grad_norm": 3.2157669067382812,
|
1246 |
+
"learning_rate": 0.00017744262295081969,
|
1247 |
+
"loss": 0.3612,
|
1248 |
+
"step": 3540
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 17.376449054301403,
|
1252 |
+
"grad_norm": 3.595738410949707,
|
1253 |
+
"learning_rate": 0.00017731147540983607,
|
1254 |
+
"loss": 0.3627,
|
1255 |
+
"step": 3560
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 17.474069554606466,
|
1259 |
+
"grad_norm": 4.483953475952148,
|
1260 |
+
"learning_rate": 0.00017718032786885246,
|
1261 |
+
"loss": 0.316,
|
1262 |
+
"step": 3580
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 17.57169005491153,
|
1266 |
+
"grad_norm": 3.636593818664551,
|
1267 |
+
"learning_rate": 0.00017704918032786885,
|
1268 |
+
"loss": 0.3751,
|
1269 |
+
"step": 3600
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 17.669310555216594,
|
1273 |
+
"grad_norm": 3.6285219192504883,
|
1274 |
+
"learning_rate": 0.00017691803278688527,
|
1275 |
+
"loss": 0.3695,
|
1276 |
+
"step": 3620
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 17.76693105552166,
|
1280 |
+
"grad_norm": 3.8501439094543457,
|
1281 |
+
"learning_rate": 0.00017678688524590166,
|
1282 |
+
"loss": 0.3525,
|
1283 |
+
"step": 3640
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 17.864551555826722,
|
1287 |
+
"grad_norm": 3.410341739654541,
|
1288 |
+
"learning_rate": 0.00017665573770491802,
|
1289 |
+
"loss": 0.3672,
|
1290 |
+
"step": 3660
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 17.962172056131788,
|
1294 |
+
"grad_norm": 3.972015380859375,
|
1295 |
+
"learning_rate": 0.00017652459016393444,
|
1296 |
+
"loss": 0.3551,
|
1297 |
+
"step": 3680
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 18.05979255643685,
|
1301 |
+
"grad_norm": 4.821686744689941,
|
1302 |
+
"learning_rate": 0.00017639344262295083,
|
1303 |
+
"loss": 0.334,
|
1304 |
+
"step": 3700
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 18.157413056741916,
|
1308 |
+
"grad_norm": 3.1545865535736084,
|
1309 |
+
"learning_rate": 0.00017626229508196724,
|
1310 |
+
"loss": 0.307,
|
1311 |
+
"step": 3720
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 18.25503355704698,
|
1315 |
+
"grad_norm": 3.6867642402648926,
|
1316 |
+
"learning_rate": 0.0001761311475409836,
|
1317 |
+
"loss": 0.2859,
|
1318 |
+
"step": 3740
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 18.352654057352044,
|
1322 |
+
"grad_norm": 3.3141942024230957,
|
1323 |
+
"learning_rate": 0.00017600000000000002,
|
1324 |
+
"loss": 0.341,
|
1325 |
+
"step": 3760
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 18.450274557657107,
|
1329 |
+
"grad_norm": 3.5395824909210205,
|
1330 |
+
"learning_rate": 0.0001758688524590164,
|
1331 |
+
"loss": 0.3412,
|
1332 |
+
"step": 3780
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 18.547895057962172,
|
1336 |
+
"grad_norm": 3.7669973373413086,
|
1337 |
+
"learning_rate": 0.00017573770491803277,
|
1338 |
+
"loss": 0.3216,
|
1339 |
+
"step": 3800
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 18.645515558267235,
|
1343 |
+
"grad_norm": 3.6290788650512695,
|
1344 |
+
"learning_rate": 0.0001756065573770492,
|
1345 |
+
"loss": 0.3222,
|
1346 |
+
"step": 3820
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 18.7431360585723,
|
1350 |
+
"grad_norm": 3.3772177696228027,
|
1351 |
+
"learning_rate": 0.00017547540983606558,
|
1352 |
+
"loss": 0.3301,
|
1353 |
+
"step": 3840
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 18.840756558877363,
|
1357 |
+
"grad_norm": 4.57468843460083,
|
1358 |
+
"learning_rate": 0.000175344262295082,
|
1359 |
+
"loss": 0.3289,
|
1360 |
+
"step": 3860
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 18.93837705918243,
|
1364 |
+
"grad_norm": 4.921297073364258,
|
1365 |
+
"learning_rate": 0.00017521311475409836,
|
1366 |
+
"loss": 0.3456,
|
1367 |
+
"step": 3880
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 19.03599755948749,
|
1371 |
+
"grad_norm": 3.4319562911987305,
|
1372 |
+
"learning_rate": 0.00017508196721311475,
|
1373 |
+
"loss": 0.3273,
|
1374 |
+
"step": 3900
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 19.133618059792557,
|
1378 |
+
"grad_norm": 3.177849054336548,
|
1379 |
+
"learning_rate": 0.00017495081967213116,
|
1380 |
+
"loss": 0.2475,
|
1381 |
+
"step": 3920
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 19.23123856009762,
|
1385 |
+
"grad_norm": 3.6201579570770264,
|
1386 |
+
"learning_rate": 0.00017481967213114753,
|
1387 |
+
"loss": 0.2825,
|
1388 |
+
"step": 3940
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 19.328859060402685,
|
1392 |
+
"grad_norm": 3.9348137378692627,
|
1393 |
+
"learning_rate": 0.00017468852459016394,
|
1394 |
+
"loss": 0.2999,
|
1395 |
+
"step": 3960
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 19.426479560707747,
|
1399 |
+
"grad_norm": 3.8892483711242676,
|
1400 |
+
"learning_rate": 0.00017455737704918033,
|
1401 |
+
"loss": 0.2946,
|
1402 |
+
"step": 3980
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 19.524100061012813,
|
1406 |
+
"grad_norm": 3.8046462535858154,
|
1407 |
+
"learning_rate": 0.00017442622950819675,
|
1408 |
+
"loss": 0.3273,
|
1409 |
+
"step": 4000
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 19.621720561317876,
|
1413 |
+
"grad_norm": 3.8215136528015137,
|
1414 |
+
"learning_rate": 0.0001742950819672131,
|
1415 |
+
"loss": 0.3057,
|
1416 |
+
"step": 4020
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 19.71934106162294,
|
1420 |
+
"grad_norm": 4.213250637054443,
|
1421 |
+
"learning_rate": 0.0001741639344262295,
|
1422 |
+
"loss": 0.3168,
|
1423 |
+
"step": 4040
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 19.816961561928004,
|
1427 |
+
"grad_norm": 4.45851469039917,
|
1428 |
+
"learning_rate": 0.00017403278688524592,
|
1429 |
+
"loss": 0.3483,
|
1430 |
+
"step": 4060
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 19.91458206223307,
|
1434 |
+
"grad_norm": 3.3840818405151367,
|
1435 |
+
"learning_rate": 0.0001739016393442623,
|
1436 |
+
"loss": 0.316,
|
1437 |
+
"step": 4080
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 20.012202562538132,
|
1441 |
+
"grad_norm": 2.586810350418091,
|
1442 |
+
"learning_rate": 0.0001737704918032787,
|
1443 |
+
"loss": 0.3167,
|
1444 |
+
"step": 4100
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 20.109823062843198,
|
1448 |
+
"grad_norm": 2.4695241451263428,
|
1449 |
+
"learning_rate": 0.00017363934426229509,
|
1450 |
+
"loss": 0.2693,
|
1451 |
+
"step": 4120
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 20.20744356314826,
|
1455 |
+
"grad_norm": 3.0270965099334717,
|
1456 |
+
"learning_rate": 0.0001735081967213115,
|
1457 |
+
"loss": 0.2699,
|
1458 |
+
"step": 4140
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 20.305064063453326,
|
1462 |
+
"grad_norm": 3.3233821392059326,
|
1463 |
+
"learning_rate": 0.00017337704918032786,
|
1464 |
+
"loss": 0.261,
|
1465 |
+
"step": 4160
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 20.40268456375839,
|
1469 |
+
"grad_norm": 3.30206561088562,
|
1470 |
+
"learning_rate": 0.00017324590163934425,
|
1471 |
+
"loss": 0.2843,
|
1472 |
+
"step": 4180
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 20.500305064063454,
|
1476 |
+
"grad_norm": 3.157876968383789,
|
1477 |
+
"learning_rate": 0.00017311475409836067,
|
1478 |
+
"loss": 0.2794,
|
1479 |
+
"step": 4200
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 20.597925564368516,
|
1483 |
+
"grad_norm": 2.9424281120300293,
|
1484 |
+
"learning_rate": 0.00017298360655737706,
|
1485 |
+
"loss": 0.3006,
|
1486 |
+
"step": 4220
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 20.695546064673582,
|
1490 |
+
"grad_norm": 4.470980167388916,
|
1491 |
+
"learning_rate": 0.00017285245901639345,
|
1492 |
+
"loss": 0.3001,
|
1493 |
+
"step": 4240
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 20.793166564978645,
|
1497 |
+
"grad_norm": 5.169465065002441,
|
1498 |
+
"learning_rate": 0.00017272131147540984,
|
1499 |
+
"loss": 0.2855,
|
1500 |
+
"step": 4260
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 20.89078706528371,
|
1504 |
+
"grad_norm": 4.266078472137451,
|
1505 |
+
"learning_rate": 0.00017259016393442625,
|
1506 |
+
"loss": 0.3039,
|
1507 |
+
"step": 4280
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 20.988407565588773,
|
1511 |
+
"grad_norm": 3.450538158416748,
|
1512 |
+
"learning_rate": 0.00017245901639344264,
|
1513 |
+
"loss": 0.3151,
|
1514 |
+
"step": 4300
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 21.08602806589384,
|
1518 |
+
"grad_norm": 3.0616092681884766,
|
1519 |
+
"learning_rate": 0.000172327868852459,
|
1520 |
+
"loss": 0.2414,
|
1521 |
+
"step": 4320
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 21.1836485661989,
|
1525 |
+
"grad_norm": 2.623687267303467,
|
1526 |
+
"learning_rate": 0.00017219672131147542,
|
1527 |
+
"loss": 0.2257,
|
1528 |
+
"step": 4340
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 21.281269066503967,
|
1532 |
+
"grad_norm": 3.4922776222229004,
|
1533 |
+
"learning_rate": 0.0001720655737704918,
|
1534 |
+
"loss": 0.2405,
|
1535 |
+
"step": 4360
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 21.37888956680903,
|
1539 |
+
"grad_norm": 3.7959911823272705,
|
1540 |
+
"learning_rate": 0.0001719344262295082,
|
1541 |
+
"loss": 0.2732,
|
1542 |
+
"step": 4380
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 21.476510067114095,
|
1546 |
+
"grad_norm": 3.0746493339538574,
|
1547 |
+
"learning_rate": 0.0001718032786885246,
|
1548 |
+
"loss": 0.2592,
|
1549 |
+
"step": 4400
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 21.574130567419157,
|
1553 |
+
"grad_norm": 4.570422649383545,
|
1554 |
+
"learning_rate": 0.00017167213114754098,
|
1555 |
+
"loss": 0.2742,
|
1556 |
+
"step": 4420
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 21.671751067724223,
|
1560 |
+
"grad_norm": 4.301755905151367,
|
1561 |
+
"learning_rate": 0.0001715409836065574,
|
1562 |
+
"loss": 0.2843,
|
1563 |
+
"step": 4440
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 21.769371568029285,
|
1567 |
+
"grad_norm": 3.110278367996216,
|
1568 |
+
"learning_rate": 0.00017140983606557376,
|
1569 |
+
"loss": 0.2863,
|
1570 |
+
"step": 4460
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 21.86699206833435,
|
1574 |
+
"grad_norm": 4.253392219543457,
|
1575 |
+
"learning_rate": 0.00017127868852459018,
|
1576 |
+
"loss": 0.2968,
|
1577 |
+
"step": 4480
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 21.964612568639414,
|
1581 |
+
"grad_norm": 3.6689677238464355,
|
1582 |
+
"learning_rate": 0.00017114754098360656,
|
1583 |
+
"loss": 0.2918,
|
1584 |
+
"step": 4500
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 22.06223306894448,
|
1588 |
+
"grad_norm": 3.50288462638855,
|
1589 |
+
"learning_rate": 0.00017101639344262298,
|
1590 |
+
"loss": 0.2567,
|
1591 |
+
"step": 4520
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 22.15985356924954,
|
1595 |
+
"grad_norm": 3.2581892013549805,
|
1596 |
+
"learning_rate": 0.00017088524590163934,
|
1597 |
+
"loss": 0.2029,
|
1598 |
+
"step": 4540
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 22.257474069554608,
|
1602 |
+
"grad_norm": 3.4691686630249023,
|
1603 |
+
"learning_rate": 0.00017075409836065573,
|
1604 |
+
"loss": 0.2299,
|
1605 |
+
"step": 4560
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 22.35509456985967,
|
1609 |
+
"grad_norm": 4.3092427253723145,
|
1610 |
+
"learning_rate": 0.00017062295081967215,
|
1611 |
+
"loss": 0.2578,
|
1612 |
+
"step": 4580
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 22.452715070164736,
|
1616 |
+
"grad_norm": 3.8213038444519043,
|
1617 |
+
"learning_rate": 0.0001704918032786885,
|
1618 |
+
"loss": 0.2766,
|
1619 |
+
"step": 4600
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 22.550335570469798,
|
1623 |
+
"grad_norm": 3.121110200881958,
|
1624 |
+
"learning_rate": 0.00017036065573770493,
|
1625 |
+
"loss": 0.2612,
|
1626 |
+
"step": 4620
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 22.647956070774864,
|
1630 |
+
"grad_norm": 2.7411513328552246,
|
1631 |
+
"learning_rate": 0.00017022950819672132,
|
1632 |
+
"loss": 0.2459,
|
1633 |
+
"step": 4640
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 22.745576571079926,
|
1637 |
+
"grad_norm": 3.85884428024292,
|
1638 |
+
"learning_rate": 0.00017009836065573773,
|
1639 |
+
"loss": 0.2757,
|
1640 |
+
"step": 4660
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 22.843197071384992,
|
1644 |
+
"grad_norm": 3.128300189971924,
|
1645 |
+
"learning_rate": 0.0001699672131147541,
|
1646 |
+
"loss": 0.2498,
|
1647 |
+
"step": 4680
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 22.940817571690054,
|
1651 |
+
"grad_norm": 2.7686431407928467,
|
1652 |
+
"learning_rate": 0.00016983606557377049,
|
1653 |
+
"loss": 0.2608,
|
1654 |
+
"step": 4700
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 23.03843807199512,
|
1658 |
+
"grad_norm": 3.427732229232788,
|
1659 |
+
"learning_rate": 0.0001697049180327869,
|
1660 |
+
"loss": 0.2449,
|
1661 |
+
"step": 4720
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 23.136058572300183,
|
1665 |
+
"grad_norm": 3.1296660900115967,
|
1666 |
+
"learning_rate": 0.0001695737704918033,
|
1667 |
+
"loss": 0.2177,
|
1668 |
+
"step": 4740
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 23.23367907260525,
|
1672 |
+
"grad_norm": 3.503549575805664,
|
1673 |
+
"learning_rate": 0.00016944262295081968,
|
1674 |
+
"loss": 0.2174,
|
1675 |
+
"step": 4760
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 23.33129957291031,
|
1679 |
+
"grad_norm": 3.4126474857330322,
|
1680 |
+
"learning_rate": 0.00016931147540983607,
|
1681 |
+
"loss": 0.2354,
|
1682 |
+
"step": 4780
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 23.428920073215377,
|
1686 |
+
"grad_norm": 3.743910074234009,
|
1687 |
+
"learning_rate": 0.00016918032786885249,
|
1688 |
+
"loss": 0.2471,
|
1689 |
+
"step": 4800
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 23.52654057352044,
|
1693 |
+
"grad_norm": 3.852505683898926,
|
1694 |
+
"learning_rate": 0.00016904918032786885,
|
1695 |
+
"loss": 0.2558,
|
1696 |
+
"step": 4820
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 23.624161073825505,
|
1700 |
+
"grad_norm": 3.120332956314087,
|
1701 |
+
"learning_rate": 0.00016891803278688524,
|
1702 |
+
"loss": 0.2273,
|
1703 |
+
"step": 4840
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 23.721781574130567,
|
1707 |
+
"grad_norm": 3.634793758392334,
|
1708 |
+
"learning_rate": 0.00016878688524590165,
|
1709 |
+
"loss": 0.2452,
|
1710 |
+
"step": 4860
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 23.819402074435633,
|
1714 |
+
"grad_norm": 3.3393948078155518,
|
1715 |
+
"learning_rate": 0.00016865573770491804,
|
1716 |
+
"loss": 0.2503,
|
1717 |
+
"step": 4880
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 23.917022574740695,
|
1721 |
+
"grad_norm": 2.6347219944000244,
|
1722 |
+
"learning_rate": 0.00016852459016393443,
|
1723 |
+
"loss": 0.2315,
|
1724 |
+
"step": 4900
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 24.01464307504576,
|
1728 |
+
"grad_norm": 2.7050418853759766,
|
1729 |
+
"learning_rate": 0.00016839344262295082,
|
1730 |
+
"loss": 0.2638,
|
1731 |
+
"step": 4920
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 24.112263575350823,
|
1735 |
+
"grad_norm": 2.6256654262542725,
|
1736 |
+
"learning_rate": 0.0001682622950819672,
|
1737 |
+
"loss": 0.2156,
|
1738 |
+
"step": 4940
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 24.20988407565589,
|
1742 |
+
"grad_norm": 3.150125741958618,
|
1743 |
+
"learning_rate": 0.00016813114754098363,
|
1744 |
+
"loss": 0.2109,
|
1745 |
+
"step": 4960
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 24.30750457596095,
|
1749 |
+
"grad_norm": 2.8489086627960205,
|
1750 |
+
"learning_rate": 0.000168,
|
1751 |
+
"loss": 0.2092,
|
1752 |
+
"step": 4980
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 24.405125076266017,
|
1756 |
+
"grad_norm": 2.6634905338287354,
|
1757 |
+
"learning_rate": 0.0001678688524590164,
|
1758 |
+
"loss": 0.2152,
|
1759 |
+
"step": 5000
|
1760 |
+
}
|
1761 |
+
],
|
1762 |
+
"logging_steps": 20,
|
1763 |
+
"max_steps": 30600,
|
1764 |
+
"num_input_tokens_seen": 0,
|
1765 |
+
"num_train_epochs": 150,
|
1766 |
+
"save_steps": 500,
|
1767 |
+
"stateful_callbacks": {
|
1768 |
+
"TrainerControl": {
|
1769 |
+
"args": {
|
1770 |
+
"should_epoch_stop": false,
|
1771 |
+
"should_evaluate": false,
|
1772 |
+
"should_log": false,
|
1773 |
+
"should_save": true,
|
1774 |
+
"should_training_stop": false
|
1775 |
+
},
|
1776 |
+
"attributes": {}
|
1777 |
+
}
|
1778 |
+
},
|
1779 |
+
"total_flos": 5.480289203268403e+16,
|
1780 |
+
"train_batch_size": 1,
|
1781 |
+
"trial_name": null,
|
1782 |
+
"trial_params": null
|
1783 |
+
}
|