Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1729 -0
trainer_state.json
ADDED
@@ -0,0 +1,1729 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 19.958677685950413,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 2415,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.08264462809917356,
|
13 |
+
"grad_norm": 4.703310012817383,
|
14 |
+
"learning_rate": 1.652892561983471e-05,
|
15 |
+
"loss": 0.9413,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.1652892561983471,
|
20 |
+
"grad_norm": 2.9132559299468994,
|
21 |
+
"learning_rate": 3.305785123966942e-05,
|
22 |
+
"loss": 0.4897,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.24793388429752067,
|
27 |
+
"grad_norm": 1.5209449529647827,
|
28 |
+
"learning_rate": 4.958677685950414e-05,
|
29 |
+
"loss": 0.2702,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.3305785123966942,
|
34 |
+
"grad_norm": 0.6627697944641113,
|
35 |
+
"learning_rate": 6.611570247933885e-05,
|
36 |
+
"loss": 0.1735,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.4132231404958678,
|
41 |
+
"grad_norm": 1.2193719148635864,
|
42 |
+
"learning_rate": 8.264462809917356e-05,
|
43 |
+
"loss": 0.138,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.49586776859504134,
|
48 |
+
"grad_norm": 1.7384006977081299,
|
49 |
+
"learning_rate": 9.917355371900827e-05,
|
50 |
+
"loss": 0.1263,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.5785123966942148,
|
55 |
+
"grad_norm": 1.2912031412124634,
|
56 |
+
"learning_rate": 0.00011570247933884298,
|
57 |
+
"loss": 0.1205,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.6611570247933884,
|
62 |
+
"grad_norm": 0.9840922355651855,
|
63 |
+
"learning_rate": 0.0001322314049586777,
|
64 |
+
"loss": 0.1048,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.743801652892562,
|
69 |
+
"grad_norm": 0.4457073509693146,
|
70 |
+
"learning_rate": 0.0001487603305785124,
|
71 |
+
"loss": 0.0894,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.8264462809917356,
|
76 |
+
"grad_norm": 0.7870272994041443,
|
77 |
+
"learning_rate": 0.00016528925619834712,
|
78 |
+
"loss": 0.075,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.9090909090909091,
|
83 |
+
"grad_norm": 0.5919432640075684,
|
84 |
+
"learning_rate": 0.00018181818181818183,
|
85 |
+
"loss": 0.0839,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.9917355371900827,
|
90 |
+
"grad_norm": 0.5343506932258606,
|
91 |
+
"learning_rate": 0.00019834710743801655,
|
92 |
+
"loss": 0.0772,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 1.0743801652892562,
|
97 |
+
"grad_norm": 0.6658118963241577,
|
98 |
+
"learning_rate": 0.00019999240439374418,
|
99 |
+
"loss": 0.0697,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 1.1570247933884297,
|
104 |
+
"grad_norm": 0.8337640762329102,
|
105 |
+
"learning_rate": 0.0001999661494584419,
|
106 |
+
"loss": 0.067,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 1.2396694214876034,
|
111 |
+
"grad_norm": 0.4294738471508026,
|
112 |
+
"learning_rate": 0.00019992114634399312,
|
113 |
+
"loss": 0.07,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 1.322314049586777,
|
118 |
+
"grad_norm": 0.5833399891853333,
|
119 |
+
"learning_rate": 0.00019985740349051935,
|
120 |
+
"loss": 0.0684,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 1.4049586776859504,
|
125 |
+
"grad_norm": 0.5372803211212158,
|
126 |
+
"learning_rate": 0.00019977493285269171,
|
127 |
+
"loss": 0.0576,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 1.487603305785124,
|
132 |
+
"grad_norm": 0.4818974435329437,
|
133 |
+
"learning_rate": 0.00019967374989748876,
|
134 |
+
"loss": 0.0519,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.5702479338842976,
|
139 |
+
"grad_norm": 0.4126225709915161,
|
140 |
+
"learning_rate": 0.0001995538736012958,
|
141 |
+
"loss": 0.0542,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 1.6528925619834711,
|
146 |
+
"grad_norm": 0.4318453073501587,
|
147 |
+
"learning_rate": 0.00019941532644634596,
|
148 |
+
"loss": 0.0516,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 1.7355371900826446,
|
153 |
+
"grad_norm": 0.6815105676651001,
|
154 |
+
"learning_rate": 0.0001992581344165037,
|
155 |
+
"loss": 0.0559,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 1.8181818181818183,
|
160 |
+
"grad_norm": 0.5503532290458679,
|
161 |
+
"learning_rate": 0.00019908232699239182,
|
162 |
+
"loss": 0.0531,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 1.9008264462809916,
|
167 |
+
"grad_norm": 0.5689036846160889,
|
168 |
+
"learning_rate": 0.0001988879371458622,
|
169 |
+
"loss": 0.0528,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 1.9834710743801653,
|
174 |
+
"grad_norm": 0.5625288486480713,
|
175 |
+
"learning_rate": 0.00019867500133381242,
|
176 |
+
"loss": 0.0541,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 2.0661157024793386,
|
181 |
+
"grad_norm": 0.4508824646472931,
|
182 |
+
"learning_rate": 0.00019844355949134824,
|
183 |
+
"loss": 0.0472,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 2.1487603305785123,
|
188 |
+
"grad_norm": 0.5423370599746704,
|
189 |
+
"learning_rate": 0.000198193655024294,
|
190 |
+
"loss": 0.0455,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 2.231404958677686,
|
195 |
+
"grad_norm": 0.5978049039840698,
|
196 |
+
"learning_rate": 0.00019792533480105203,
|
197 |
+
"loss": 0.0453,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 2.3140495867768593,
|
202 |
+
"grad_norm": 0.301248162984848,
|
203 |
+
"learning_rate": 0.0001976386491438129,
|
204 |
+
"loss": 0.0401,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 2.396694214876033,
|
209 |
+
"grad_norm": 0.2878156006336212,
|
210 |
+
"learning_rate": 0.00019733365181911748,
|
211 |
+
"loss": 0.041,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 2.479338842975207,
|
216 |
+
"grad_norm": 0.3870035707950592,
|
217 |
+
"learning_rate": 0.00019701040002777332,
|
218 |
+
"loss": 0.0434,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 2.56198347107438,
|
223 |
+
"grad_norm": 0.5360531210899353,
|
224 |
+
"learning_rate": 0.000196668954394127,
|
225 |
+
"loss": 0.0496,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 2.644628099173554,
|
230 |
+
"grad_norm": 0.5107744336128235,
|
231 |
+
"learning_rate": 0.0001963093789546944,
|
232 |
+
"loss": 0.0418,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 2.7272727272727275,
|
237 |
+
"grad_norm": 0.46189603209495544,
|
238 |
+
"learning_rate": 0.00019593174114615068,
|
239 |
+
"loss": 0.0388,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 2.809917355371901,
|
244 |
+
"grad_norm": 0.3216039836406708,
|
245 |
+
"learning_rate": 0.0001955361117926831,
|
246 |
+
"loss": 0.0443,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 2.8925619834710745,
|
251 |
+
"grad_norm": 0.5578595399856567,
|
252 |
+
"learning_rate": 0.00019512256509270822,
|
253 |
+
"loss": 0.0443,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 2.975206611570248,
|
258 |
+
"grad_norm": 0.33241087198257446,
|
259 |
+
"learning_rate": 0.0001946911786049562,
|
260 |
+
"loss": 0.0446,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 3.0578512396694215,
|
265 |
+
"grad_norm": 0.3146415948867798,
|
266 |
+
"learning_rate": 0.0001942420332339252,
|
267 |
+
"loss": 0.0388,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 3.1404958677685952,
|
272 |
+
"grad_norm": 0.30948349833488464,
|
273 |
+
"learning_rate": 0.00019377521321470805,
|
274 |
+
"loss": 0.0349,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 3.2231404958677685,
|
279 |
+
"grad_norm": 0.3389824330806732,
|
280 |
+
"learning_rate": 0.00019329080609719435,
|
281 |
+
"loss": 0.0387,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 3.3057851239669422,
|
286 |
+
"grad_norm": 0.27903950214385986,
|
287 |
+
"learning_rate": 0.00019278890272965096,
|
288 |
+
"loss": 0.0365,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 3.3884297520661155,
|
293 |
+
"grad_norm": 0.3887731730937958,
|
294 |
+
"learning_rate": 0.00019226959724168367,
|
295 |
+
"loss": 0.0398,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 3.4710743801652892,
|
300 |
+
"grad_norm": 0.23304741084575653,
|
301 |
+
"learning_rate": 0.0001917329870265839,
|
302 |
+
"loss": 0.035,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 3.553719008264463,
|
307 |
+
"grad_norm": 0.4025675654411316,
|
308 |
+
"learning_rate": 0.00019117917272306283,
|
309 |
+
"loss": 0.0373,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 3.6363636363636362,
|
314 |
+
"grad_norm": 0.4401100277900696,
|
315 |
+
"learning_rate": 0.00019060825819637716,
|
316 |
+
"loss": 0.0374,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 3.71900826446281,
|
321 |
+
"grad_norm": 0.3880658447742462,
|
322 |
+
"learning_rate": 0.00019002035051884973,
|
323 |
+
"loss": 0.0396,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 3.8016528925619832,
|
328 |
+
"grad_norm": 0.3345029950141907,
|
329 |
+
"learning_rate": 0.00018941555994978848,
|
330 |
+
"loss": 0.0332,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 3.884297520661157,
|
335 |
+
"grad_norm": 0.297284334897995,
|
336 |
+
"learning_rate": 0.00018879399991480792,
|
337 |
+
"loss": 0.0321,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 3.9669421487603307,
|
342 |
+
"grad_norm": 0.27577996253967285,
|
343 |
+
"learning_rate": 0.00018815578698455675,
|
344 |
+
"loss": 0.0344,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 4.049586776859504,
|
349 |
+
"grad_norm": 0.2191019505262375,
|
350 |
+
"learning_rate": 0.00018750104085285546,
|
351 |
+
"loss": 0.0308,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 4.132231404958677,
|
356 |
+
"grad_norm": 0.2145395576953888,
|
357 |
+
"learning_rate": 0.0001868298843142484,
|
358 |
+
"loss": 0.0319,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 4.214876033057851,
|
363 |
+
"grad_norm": 0.2143179327249527,
|
364 |
+
"learning_rate": 0.0001861424432409742,
|
365 |
+
"loss": 0.031,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 4.297520661157025,
|
370 |
+
"grad_norm": 0.6055622100830078,
|
371 |
+
"learning_rate": 0.0001854388465593591,
|
372 |
+
"loss": 0.0355,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 4.380165289256198,
|
377 |
+
"grad_norm": 0.5569704174995422,
|
378 |
+
"learning_rate": 0.00018471922622563728,
|
379 |
+
"loss": 0.0319,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 4.462809917355372,
|
384 |
+
"grad_norm": 0.2579587399959564,
|
385 |
+
"learning_rate": 0.0001839837172012033,
|
386 |
+
"loss": 0.0331,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 4.545454545454545,
|
391 |
+
"grad_norm": 0.22480279207229614,
|
392 |
+
"learning_rate": 0.00018323245742730054,
|
393 |
+
"loss": 0.0306,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 4.628099173553719,
|
398 |
+
"grad_norm": 0.260975182056427,
|
399 |
+
"learning_rate": 0.00018246558779915105,
|
400 |
+
"loss": 0.0342,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 4.710743801652892,
|
405 |
+
"grad_norm": 0.35224688053131104,
|
406 |
+
"learning_rate": 0.0001816832521395313,
|
407 |
+
"loss": 0.0407,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 4.793388429752066,
|
412 |
+
"grad_norm": 0.265455961227417,
|
413 |
+
"learning_rate": 0.000180885597171799,
|
414 |
+
"loss": 0.0332,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 4.87603305785124,
|
419 |
+
"grad_norm": 0.2791174352169037,
|
420 |
+
"learning_rate": 0.00018007277249237563,
|
421 |
+
"loss": 0.0342,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 4.958677685950414,
|
426 |
+
"grad_norm": 0.39626914262771606,
|
427 |
+
"learning_rate": 0.00017924493054269046,
|
428 |
+
"loss": 0.0342,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 5.041322314049586,
|
433 |
+
"grad_norm": 0.3018089532852173,
|
434 |
+
"learning_rate": 0.00017840222658059096,
|
435 |
+
"loss": 0.0306,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 5.12396694214876,
|
440 |
+
"grad_norm": 0.4229370653629303,
|
441 |
+
"learning_rate": 0.00017754481865122474,
|
442 |
+
"loss": 0.0293,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 5.206611570247934,
|
447 |
+
"grad_norm": 0.35095998644828796,
|
448 |
+
"learning_rate": 0.00017667286755739895,
|
449 |
+
"loss": 0.0334,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 5.289256198347108,
|
454 |
+
"grad_norm": 0.3846903145313263,
|
455 |
+
"learning_rate": 0.00017578653682942275,
|
456 |
+
"loss": 0.0346,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 5.371900826446281,
|
461 |
+
"grad_norm": 0.3931705057621002,
|
462 |
+
"learning_rate": 0.00017488599269443753,
|
463 |
+
"loss": 0.0363,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 5.454545454545454,
|
468 |
+
"grad_norm": 0.39143308997154236,
|
469 |
+
"learning_rate": 0.0001739714040452422,
|
470 |
+
"loss": 0.0274,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 5.537190082644628,
|
475 |
+
"grad_norm": 0.26642906665802,
|
476 |
+
"learning_rate": 0.0001730429424086179,
|
477 |
+
"loss": 0.0294,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 5.619834710743802,
|
482 |
+
"grad_norm": 0.27915745973587036,
|
483 |
+
"learning_rate": 0.00017210078191315904,
|
484 |
+
"loss": 0.0275,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 5.702479338842975,
|
489 |
+
"grad_norm": 0.3474787771701813,
|
490 |
+
"learning_rate": 0.0001711450992566163,
|
491 |
+
"loss": 0.0249,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 5.785123966942149,
|
496 |
+
"grad_norm": 0.4248145818710327,
|
497 |
+
"learning_rate": 0.00017017607367275777,
|
498 |
+
"loss": 0.0266,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 5.867768595041323,
|
503 |
+
"grad_norm": 0.35364291071891785,
|
504 |
+
"learning_rate": 0.00016919388689775464,
|
505 |
+
"loss": 0.0295,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 5.950413223140496,
|
510 |
+
"grad_norm": 0.32174554467201233,
|
511 |
+
"learning_rate": 0.00016819872313609716,
|
512 |
+
"loss": 0.033,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 6.033057851239669,
|
517 |
+
"grad_norm": 0.3670883774757385,
|
518 |
+
"learning_rate": 0.00016719076902604836,
|
519 |
+
"loss": 0.0325,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 6.115702479338843,
|
524 |
+
"grad_norm": 0.2812848687171936,
|
525 |
+
"learning_rate": 0.00016617021360464045,
|
526 |
+
"loss": 0.0324,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 6.198347107438017,
|
531 |
+
"grad_norm": 0.2678919732570648,
|
532 |
+
"learning_rate": 0.00016513724827222227,
|
533 |
+
"loss": 0.0258,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 6.2809917355371905,
|
538 |
+
"grad_norm": 0.33413299918174744,
|
539 |
+
"learning_rate": 0.00016409206675656264,
|
540 |
+
"loss": 0.0295,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 6.363636363636363,
|
545 |
+
"grad_norm": 0.31874868273735046,
|
546 |
+
"learning_rate": 0.00016303486507651807,
|
547 |
+
"loss": 0.0267,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 6.446280991735537,
|
552 |
+
"grad_norm": 0.2576686441898346,
|
553 |
+
"learning_rate": 0.00016196584150527,
|
554 |
+
"loss": 0.033,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 6.528925619834711,
|
559 |
+
"grad_norm": 0.16514086723327637,
|
560 |
+
"learning_rate": 0.0001608851965331397,
|
561 |
+
"loss": 0.0305,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 6.6115702479338845,
|
566 |
+
"grad_norm": 0.2784286141395569,
|
567 |
+
"learning_rate": 0.0001597931328299874,
|
568 |
+
"loss": 0.0288,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 6.694214876033058,
|
573 |
+
"grad_norm": 0.21867933869361877,
|
574 |
+
"learning_rate": 0.0001586898552072023,
|
575 |
+
"loss": 0.0257,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 6.776859504132231,
|
580 |
+
"grad_norm": 0.13699640333652496,
|
581 |
+
"learning_rate": 0.00015757557057929127,
|
582 |
+
"loss": 0.0243,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 6.859504132231405,
|
587 |
+
"grad_norm": 0.2195553332567215,
|
588 |
+
"learning_rate": 0.00015645048792507297,
|
589 |
+
"loss": 0.0298,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 6.9421487603305785,
|
594 |
+
"grad_norm": 0.21387651562690735,
|
595 |
+
"learning_rate": 0.00015531481824848504,
|
596 |
+
"loss": 0.0285,
|
597 |
+
"step": 840
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 7.024793388429752,
|
601 |
+
"grad_norm": 0.1597180813550949,
|
602 |
+
"learning_rate": 0.00015416877453901096,
|
603 |
+
"loss": 0.0259,
|
604 |
+
"step": 850
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 7.107438016528926,
|
608 |
+
"grad_norm": 0.2567303776741028,
|
609 |
+
"learning_rate": 0.00015301257173173523,
|
610 |
+
"loss": 0.0258,
|
611 |
+
"step": 860
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 7.190082644628099,
|
615 |
+
"grad_norm": 0.18878251314163208,
|
616 |
+
"learning_rate": 0.00015184642666703308,
|
617 |
+
"loss": 0.0293,
|
618 |
+
"step": 870
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 7.2727272727272725,
|
622 |
+
"grad_norm": 0.24876275658607483,
|
623 |
+
"learning_rate": 0.00015067055804990318,
|
624 |
+
"loss": 0.0283,
|
625 |
+
"step": 880
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 7.355371900826446,
|
629 |
+
"grad_norm": 0.29571253061294556,
|
630 |
+
"learning_rate": 0.0001494851864089505,
|
631 |
+
"loss": 0.0288,
|
632 |
+
"step": 890
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 7.43801652892562,
|
636 |
+
"grad_norm": 0.2919931709766388,
|
637 |
+
"learning_rate": 0.00014829053405502723,
|
638 |
+
"loss": 0.0282,
|
639 |
+
"step": 900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 7.520661157024794,
|
643 |
+
"grad_norm": 0.25039032101631165,
|
644 |
+
"learning_rate": 0.00014708682503953956,
|
645 |
+
"loss": 0.0258,
|
646 |
+
"step": 910
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 7.6033057851239665,
|
650 |
+
"grad_norm": 0.2013220340013504,
|
651 |
+
"learning_rate": 0.00014587428511242766,
|
652 |
+
"loss": 0.0247,
|
653 |
+
"step": 920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 7.68595041322314,
|
657 |
+
"grad_norm": 0.3657705783843994,
|
658 |
+
"learning_rate": 0.0001446531416798275,
|
659 |
+
"loss": 0.0287,
|
660 |
+
"step": 930
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 7.768595041322314,
|
664 |
+
"grad_norm": 0.36600229144096375,
|
665 |
+
"learning_rate": 0.00014342362376142185,
|
666 |
+
"loss": 0.0271,
|
667 |
+
"step": 940
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 7.851239669421488,
|
671 |
+
"grad_norm": 0.20832963287830353,
|
672 |
+
"learning_rate": 0.00014218596194748875,
|
673 |
+
"loss": 0.0253,
|
674 |
+
"step": 950
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 7.933884297520661,
|
678 |
+
"grad_norm": 0.15197999775409698,
|
679 |
+
"learning_rate": 0.0001409403883556554,
|
680 |
+
"loss": 0.0271,
|
681 |
+
"step": 960
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 8.016528925619834,
|
685 |
+
"grad_norm": 0.16558091342449188,
|
686 |
+
"learning_rate": 0.00013968713658736552,
|
687 |
+
"loss": 0.0265,
|
688 |
+
"step": 970
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 8.099173553719009,
|
692 |
+
"grad_norm": 0.1627083718776703,
|
693 |
+
"learning_rate": 0.00013842644168406864,
|
694 |
+
"loss": 0.0234,
|
695 |
+
"step": 980
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 8.181818181818182,
|
699 |
+
"grad_norm": 0.33982279896736145,
|
700 |
+
"learning_rate": 0.000137158540083139,
|
701 |
+
"loss": 0.025,
|
702 |
+
"step": 990
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 8.264462809917354,
|
706 |
+
"grad_norm": 0.415449857711792,
|
707 |
+
"learning_rate": 0.000135883669573533,
|
708 |
+
"loss": 0.0279,
|
709 |
+
"step": 1000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 8.347107438016529,
|
713 |
+
"grad_norm": 0.3493754267692566,
|
714 |
+
"learning_rate": 0.00013460206925119297,
|
715 |
+
"loss": 0.0242,
|
716 |
+
"step": 1010
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 8.429752066115702,
|
720 |
+
"grad_norm": 0.1715659201145172,
|
721 |
+
"learning_rate": 0.00013331397947420576,
|
722 |
+
"loss": 0.031,
|
723 |
+
"step": 1020
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 8.512396694214877,
|
727 |
+
"grad_norm": 0.27542078495025635,
|
728 |
+
"learning_rate": 0.00013201964181772485,
|
729 |
+
"loss": 0.0247,
|
730 |
+
"step": 1030
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 8.59504132231405,
|
734 |
+
"grad_norm": 0.25437498092651367,
|
735 |
+
"learning_rate": 0.000130719299028664,
|
736 |
+
"loss": 0.0239,
|
737 |
+
"step": 1040
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 8.677685950413224,
|
741 |
+
"grad_norm": 0.15187709033489227,
|
742 |
+
"learning_rate": 0.00012941319498017116,
|
743 |
+
"loss": 0.0219,
|
744 |
+
"step": 1050
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 8.760330578512397,
|
748 |
+
"grad_norm": 0.2681483328342438,
|
749 |
+
"learning_rate": 0.00012810157462589137,
|
750 |
+
"loss": 0.0229,
|
751 |
+
"step": 1060
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 8.84297520661157,
|
755 |
+
"grad_norm": 0.1943976879119873,
|
756 |
+
"learning_rate": 0.00012678468395402662,
|
757 |
+
"loss": 0.0224,
|
758 |
+
"step": 1070
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 8.925619834710744,
|
762 |
+
"grad_norm": 0.25881528854370117,
|
763 |
+
"learning_rate": 0.00012546276994120228,
|
764 |
+
"loss": 0.026,
|
765 |
+
"step": 1080
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 9.008264462809917,
|
769 |
+
"grad_norm": 0.21106192469596863,
|
770 |
+
"learning_rate": 0.0001241360805061474,
|
771 |
+
"loss": 0.0208,
|
772 |
+
"step": 1090
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 9.090909090909092,
|
776 |
+
"grad_norm": 0.2531886398792267,
|
777 |
+
"learning_rate": 0.00012280486446319895,
|
778 |
+
"loss": 0.0257,
|
779 |
+
"step": 1100
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 9.173553719008265,
|
783 |
+
"grad_norm": 0.22699125111103058,
|
784 |
+
"learning_rate": 0.00012146937147563793,
|
785 |
+
"loss": 0.0172,
|
786 |
+
"step": 1110
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 9.256198347107437,
|
790 |
+
"grad_norm": 0.1889602094888687,
|
791 |
+
"learning_rate": 0.00012012985200886602,
|
792 |
+
"loss": 0.022,
|
793 |
+
"step": 1120
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 9.338842975206612,
|
797 |
+
"grad_norm": 0.24796439707279205,
|
798 |
+
"learning_rate": 0.00011878655728343204,
|
799 |
+
"loss": 0.0217,
|
800 |
+
"step": 1130
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 9.421487603305785,
|
804 |
+
"grad_norm": 0.2704314589500427,
|
805 |
+
"learning_rate": 0.00011743973922791678,
|
806 |
+
"loss": 0.0229,
|
807 |
+
"step": 1140
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 9.50413223140496,
|
811 |
+
"grad_norm": 0.26962810754776,
|
812 |
+
"learning_rate": 0.00011608965043168496,
|
813 |
+
"loss": 0.0201,
|
814 |
+
"step": 1150
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 9.586776859504132,
|
818 |
+
"grad_norm": 0.1735532283782959,
|
819 |
+
"learning_rate": 0.0001147365440975132,
|
820 |
+
"loss": 0.0215,
|
821 |
+
"step": 1160
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 9.669421487603305,
|
825 |
+
"grad_norm": 0.15696899592876434,
|
826 |
+
"learning_rate": 0.00011338067399410319,
|
827 |
+
"loss": 0.0241,
|
828 |
+
"step": 1170
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 9.75206611570248,
|
832 |
+
"grad_norm": 0.23969663679599762,
|
833 |
+
"learning_rate": 0.00011202229440848844,
|
834 |
+
"loss": 0.0238,
|
835 |
+
"step": 1180
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 9.834710743801653,
|
839 |
+
"grad_norm": 0.23791910707950592,
|
840 |
+
"learning_rate": 0.000110661660098344,
|
841 |
+
"loss": 0.0207,
|
842 |
+
"step": 1190
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 9.917355371900827,
|
846 |
+
"grad_norm": 0.24661310017108917,
|
847 |
+
"learning_rate": 0.00010929902624420787,
|
848 |
+
"loss": 0.025,
|
849 |
+
"step": 1200
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 10.0,
|
853 |
+
"grad_norm": 0.3029937446117401,
|
854 |
+
"learning_rate": 0.00010793464840162302,
|
855 |
+
"loss": 0.0198,
|
856 |
+
"step": 1210
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 10.082644628099173,
|
860 |
+
"grad_norm": 0.22985076904296875,
|
861 |
+
"learning_rate": 0.00010656878245320933,
|
862 |
+
"loss": 0.0186,
|
863 |
+
"step": 1220
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 10.165289256198347,
|
867 |
+
"grad_norm": 0.23791086673736572,
|
868 |
+
"learning_rate": 0.00010520168456067381,
|
869 |
+
"loss": 0.0218,
|
870 |
+
"step": 1230
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 10.24793388429752,
|
874 |
+
"grad_norm": 0.23052431643009186,
|
875 |
+
"learning_rate": 0.0001038336111167688,
|
876 |
+
"loss": 0.0207,
|
877 |
+
"step": 1240
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 10.330578512396695,
|
881 |
+
"grad_norm": 0.18408218026161194,
|
882 |
+
"learning_rate": 0.00010246481869720665,
|
883 |
+
"loss": 0.0189,
|
884 |
+
"step": 1250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 10.413223140495868,
|
888 |
+
"grad_norm": 0.1350499838590622,
|
889 |
+
"learning_rate": 0.00010109556401254032,
|
890 |
+
"loss": 0.0214,
|
891 |
+
"step": 1260
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 10.49586776859504,
|
895 |
+
"grad_norm": 0.21356913447380066,
|
896 |
+
"learning_rate": 9.972610386001824e-05,
|
897 |
+
"loss": 0.0204,
|
898 |
+
"step": 1270
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 10.578512396694215,
|
902 |
+
"grad_norm": 0.19436530768871307,
|
903 |
+
"learning_rate": 9.83566950754235e-05,
|
904 |
+
"loss": 0.0234,
|
905 |
+
"step": 1280
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 10.661157024793388,
|
909 |
+
"grad_norm": 0.22124770283699036,
|
910 |
+
"learning_rate": 9.698759448490528e-05,
|
911 |
+
"loss": 0.0204,
|
912 |
+
"step": 1290
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 10.743801652892563,
|
916 |
+
"grad_norm": 0.23618550598621368,
|
917 |
+
"learning_rate": 9.561905885681252e-05,
|
918 |
+
"loss": 0.02,
|
919 |
+
"step": 1300
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 10.826446280991735,
|
923 |
+
"grad_norm": 0.22095705568790436,
|
924 |
+
"learning_rate": 9.425134485353804e-05,
|
925 |
+
"loss": 0.0191,
|
926 |
+
"step": 1310
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 10.909090909090908,
|
930 |
+
"grad_norm": 0.1535860002040863,
|
931 |
+
"learning_rate": 9.288470898338283e-05,
|
932 |
+
"loss": 0.0205,
|
933 |
+
"step": 1320
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 10.991735537190083,
|
937 |
+
"grad_norm": 0.23291295766830444,
|
938 |
+
"learning_rate": 9.151940755244912e-05,
|
939 |
+
"loss": 0.0228,
|
940 |
+
"step": 1330
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 11.074380165289256,
|
944 |
+
"grad_norm": 0.22103387117385864,
|
945 |
+
"learning_rate": 9.01556966165713e-05,
|
946 |
+
"loss": 0.021,
|
947 |
+
"step": 1340
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 11.15702479338843,
|
951 |
+
"grad_norm": 0.2173365205526352,
|
952 |
+
"learning_rate": 8.879383193329409e-05,
|
953 |
+
"loss": 0.0224,
|
954 |
+
"step": 1350
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 11.239669421487603,
|
958 |
+
"grad_norm": 0.18998365104198456,
|
959 |
+
"learning_rate": 8.743406891390607e-05,
|
960 |
+
"loss": 0.0191,
|
961 |
+
"step": 1360
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 11.322314049586776,
|
965 |
+
"grad_norm": 0.22178007662296295,
|
966 |
+
"learning_rate": 8.607666257553881e-05,
|
967 |
+
"loss": 0.0182,
|
968 |
+
"step": 1370
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 11.40495867768595,
|
972 |
+
"grad_norm": 0.17217198014259338,
|
973 |
+
"learning_rate": 8.472186749333941e-05,
|
974 |
+
"loss": 0.017,
|
975 |
+
"step": 1380
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 11.487603305785123,
|
979 |
+
"grad_norm": 0.2341386377811432,
|
980 |
+
"learning_rate": 8.336993775272624e-05,
|
981 |
+
"loss": 0.017,
|
982 |
+
"step": 1390
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 11.570247933884298,
|
986 |
+
"grad_norm": 0.2643473744392395,
|
987 |
+
"learning_rate": 8.202112690173643e-05,
|
988 |
+
"loss": 0.0182,
|
989 |
+
"step": 1400
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 11.652892561983471,
|
993 |
+
"grad_norm": 0.20025375485420227,
|
994 |
+
"learning_rate": 8.067568790347394e-05,
|
995 |
+
"loss": 0.0184,
|
996 |
+
"step": 1410
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 11.735537190082646,
|
1000 |
+
"grad_norm": 0.3023671507835388,
|
1001 |
+
"learning_rate": 7.933387308866793e-05,
|
1002 |
+
"loss": 0.0196,
|
1003 |
+
"step": 1420
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 11.818181818181818,
|
1007 |
+
"grad_norm": 0.3398241400718689,
|
1008 |
+
"learning_rate": 7.799593410834882e-05,
|
1009 |
+
"loss": 0.0219,
|
1010 |
+
"step": 1430
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 11.900826446280991,
|
1014 |
+
"grad_norm": 0.26556509733200073,
|
1015 |
+
"learning_rate": 7.666212188665278e-05,
|
1016 |
+
"loss": 0.0167,
|
1017 |
+
"step": 1440
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 11.983471074380166,
|
1021 |
+
"grad_norm": 0.16627757251262665,
|
1022 |
+
"learning_rate": 7.533268657376175e-05,
|
1023 |
+
"loss": 0.0173,
|
1024 |
+
"step": 1450
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 12.066115702479339,
|
1028 |
+
"grad_norm": 0.1822843700647354,
|
1029 |
+
"learning_rate": 7.400787749898919e-05,
|
1030 |
+
"loss": 0.0211,
|
1031 |
+
"step": 1460
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 12.148760330578513,
|
1035 |
+
"grad_norm": 0.20193220674991608,
|
1036 |
+
"learning_rate": 7.268794312401934e-05,
|
1037 |
+
"loss": 0.0158,
|
1038 |
+
"step": 1470
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 12.231404958677686,
|
1042 |
+
"grad_norm": 0.16035108268260956,
|
1043 |
+
"learning_rate": 7.137313099630965e-05,
|
1044 |
+
"loss": 0.0181,
|
1045 |
+
"step": 1480
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 12.314049586776859,
|
1049 |
+
"grad_norm": 0.14358443021774292,
|
1050 |
+
"learning_rate": 7.006368770266421e-05,
|
1051 |
+
"loss": 0.0176,
|
1052 |
+
"step": 1490
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 12.396694214876034,
|
1056 |
+
"grad_norm": 0.2652245759963989,
|
1057 |
+
"learning_rate": 6.875985882298767e-05,
|
1058 |
+
"loss": 0.0151,
|
1059 |
+
"step": 1500
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 12.479338842975206,
|
1063 |
+
"grad_norm": 0.16228072345256805,
|
1064 |
+
"learning_rate": 6.746188888422788e-05,
|
1065 |
+
"loss": 0.0189,
|
1066 |
+
"step": 1510
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 12.561983471074381,
|
1070 |
+
"grad_norm": 0.2025652527809143,
|
1071 |
+
"learning_rate": 6.617002131451595e-05,
|
1072 |
+
"loss": 0.0175,
|
1073 |
+
"step": 1520
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 12.644628099173554,
|
1077 |
+
"grad_norm": 0.19407929480075836,
|
1078 |
+
"learning_rate": 6.488449839751248e-05,
|
1079 |
+
"loss": 0.0195,
|
1080 |
+
"step": 1530
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 12.727272727272727,
|
1084 |
+
"grad_norm": 0.19878722727298737,
|
1085 |
+
"learning_rate": 6.360556122696847e-05,
|
1086 |
+
"loss": 0.0179,
|
1087 |
+
"step": 1540
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 12.809917355371901,
|
1091 |
+
"grad_norm": 0.14901714026927948,
|
1092 |
+
"learning_rate": 6.233344966150931e-05,
|
1093 |
+
"loss": 0.0174,
|
1094 |
+
"step": 1550
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 12.892561983471074,
|
1098 |
+
"grad_norm": 0.19965632259845734,
|
1099 |
+
"learning_rate": 6.106840227965031e-05,
|
1100 |
+
"loss": 0.0202,
|
1101 |
+
"step": 1560
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 12.975206611570249,
|
1105 |
+
"grad_norm": 0.19168788194656372,
|
1106 |
+
"learning_rate": 5.98106563350526e-05,
|
1107 |
+
"loss": 0.0184,
|
1108 |
+
"step": 1570
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 13.057851239669422,
|
1112 |
+
"grad_norm": 0.21455532312393188,
|
1113 |
+
"learning_rate": 5.8560447712027176e-05,
|
1114 |
+
"loss": 0.0185,
|
1115 |
+
"step": 1580
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 13.140495867768594,
|
1119 |
+
"grad_norm": 0.14643830060958862,
|
1120 |
+
"learning_rate": 5.7318010881296e-05,
|
1121 |
+
"loss": 0.0156,
|
1122 |
+
"step": 1590
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 13.223140495867769,
|
1126 |
+
"grad_norm": 0.15653955936431885,
|
1127 |
+
"learning_rate": 5.6083578856018116e-05,
|
1128 |
+
"loss": 0.0172,
|
1129 |
+
"step": 1600
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 13.305785123966942,
|
1133 |
+
"grad_norm": 0.19662214815616608,
|
1134 |
+
"learning_rate": 5.485738314808905e-05,
|
1135 |
+
"loss": 0.0176,
|
1136 |
+
"step": 1610
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 13.388429752066116,
|
1140 |
+
"grad_norm": 0.17977331578731537,
|
1141 |
+
"learning_rate": 5.363965372472195e-05,
|
1142 |
+
"loss": 0.0149,
|
1143 |
+
"step": 1620
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 13.47107438016529,
|
1147 |
+
"grad_norm": 0.16984397172927856,
|
1148 |
+
"learning_rate": 5.2430618965318356e-05,
|
1149 |
+
"loss": 0.0155,
|
1150 |
+
"step": 1630
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 13.553719008264462,
|
1154 |
+
"grad_norm": 0.23604920506477356,
|
1155 |
+
"learning_rate": 5.123050561863657e-05,
|
1156 |
+
"loss": 0.0172,
|
1157 |
+
"step": 1640
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 13.636363636363637,
|
1161 |
+
"grad_norm": 0.20416028797626495,
|
1162 |
+
"learning_rate": 5.0039538760266034e-05,
|
1163 |
+
"loss": 0.015,
|
1164 |
+
"step": 1650
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 13.71900826446281,
|
1168 |
+
"grad_norm": 0.17371866106987,
|
1169 |
+
"learning_rate": 4.885794175041567e-05,
|
1170 |
+
"loss": 0.0145,
|
1171 |
+
"step": 1660
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 13.801652892561984,
|
1175 |
+
"grad_norm": 0.18007588386535645,
|
1176 |
+
"learning_rate": 4.7685936192023415e-05,
|
1177 |
+
"loss": 0.0159,
|
1178 |
+
"step": 1670
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 13.884297520661157,
|
1182 |
+
"grad_norm": 0.1721491813659668,
|
1183 |
+
"learning_rate": 4.652374188919584e-05,
|
1184 |
+
"loss": 0.0167,
|
1185 |
+
"step": 1680
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 13.96694214876033,
|
1189 |
+
"grad_norm": 0.24759605526924133,
|
1190 |
+
"learning_rate": 4.537157680598486e-05,
|
1191 |
+
"loss": 0.0191,
|
1192 |
+
"step": 1690
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 14.049586776859504,
|
1196 |
+
"grad_norm": 0.147015780210495,
|
1197 |
+
"learning_rate": 4.422965702550971e-05,
|
1198 |
+
"loss": 0.018,
|
1199 |
+
"step": 1700
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 14.132231404958677,
|
1203 |
+
"grad_norm": 0.2071155458688736,
|
1204 |
+
"learning_rate": 4.309819670943139e-05,
|
1205 |
+
"loss": 0.0174,
|
1206 |
+
"step": 1710
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 14.214876033057852,
|
1210 |
+
"grad_norm": 0.19617584347724915,
|
1211 |
+
"learning_rate": 4.197740805778779e-05,
|
1212 |
+
"loss": 0.0147,
|
1213 |
+
"step": 1720
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 14.297520661157025,
|
1217 |
+
"grad_norm": 0.20283041894435883,
|
1218 |
+
"learning_rate": 4.086750126919655e-05,
|
1219 |
+
"loss": 0.0149,
|
1220 |
+
"step": 1730
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 14.380165289256198,
|
1224 |
+
"grad_norm": 0.23828528821468353,
|
1225 |
+
"learning_rate": 3.976868450143329e-05,
|
1226 |
+
"loss": 0.0166,
|
1227 |
+
"step": 1740
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 14.462809917355372,
|
1231 |
+
"grad_norm": 0.2454332709312439,
|
1232 |
+
"learning_rate": 3.868116383239259e-05,
|
1233 |
+
"loss": 0.016,
|
1234 |
+
"step": 1750
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 14.545454545454545,
|
1238 |
+
"grad_norm": 0.21714092791080475,
|
1239 |
+
"learning_rate": 3.7605143221439066e-05,
|
1240 |
+
"loss": 0.0154,
|
1241 |
+
"step": 1760
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 14.62809917355372,
|
1245 |
+
"grad_norm": 0.20886965095996857,
|
1246 |
+
"learning_rate": 3.6540824471155856e-05,
|
1247 |
+
"loss": 0.0176,
|
1248 |
+
"step": 1770
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 14.710743801652892,
|
1252 |
+
"grad_norm": 0.20230558514595032,
|
1253 |
+
"learning_rate": 3.548840718949724e-05,
|
1254 |
+
"loss": 0.0164,
|
1255 |
+
"step": 1780
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 14.793388429752067,
|
1259 |
+
"grad_norm": 0.18747635185718536,
|
1260 |
+
"learning_rate": 3.4448088752353434e-05,
|
1261 |
+
"loss": 0.0159,
|
1262 |
+
"step": 1790
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 14.87603305785124,
|
1266 |
+
"grad_norm": 0.15160124003887177,
|
1267 |
+
"learning_rate": 3.3420064266533415e-05,
|
1268 |
+
"loss": 0.0153,
|
1269 |
+
"step": 1800
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 14.958677685950413,
|
1273 |
+
"grad_norm": 0.21707528829574585,
|
1274 |
+
"learning_rate": 3.240452653317367e-05,
|
1275 |
+
"loss": 0.0163,
|
1276 |
+
"step": 1810
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 15.041322314049587,
|
1280 |
+
"grad_norm": 0.2630583345890045,
|
1281 |
+
"learning_rate": 3.140166601157928e-05,
|
1282 |
+
"loss": 0.0153,
|
1283 |
+
"step": 1820
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 15.12396694214876,
|
1287 |
+
"grad_norm": 0.2211904525756836,
|
1288 |
+
"learning_rate": 3.0411670783504053e-05,
|
1289 |
+
"loss": 0.0159,
|
1290 |
+
"step": 1830
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 15.206611570247935,
|
1294 |
+
"grad_norm": 0.27094215154647827,
|
1295 |
+
"learning_rate": 2.94347265178769e-05,
|
1296 |
+
"loss": 0.0143,
|
1297 |
+
"step": 1840
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 15.289256198347108,
|
1301 |
+
"grad_norm": 0.2170286327600479,
|
1302 |
+
"learning_rate": 2.847101643598028e-05,
|
1303 |
+
"loss": 0.0154,
|
1304 |
+
"step": 1850
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 15.37190082644628,
|
1308 |
+
"grad_norm": 0.15928302705287933,
|
1309 |
+
"learning_rate": 2.7520721277088024e-05,
|
1310 |
+
"loss": 0.0162,
|
1311 |
+
"step": 1860
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 15.454545454545455,
|
1315 |
+
"grad_norm": 0.20816448330879211,
|
1316 |
+
"learning_rate": 2.6584019264568495e-05,
|
1317 |
+
"loss": 0.0127,
|
1318 |
+
"step": 1870
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 15.537190082644628,
|
1322 |
+
"grad_norm": 0.14063189923763275,
|
1323 |
+
"learning_rate": 2.5661086072459707e-05,
|
1324 |
+
"loss": 0.0168,
|
1325 |
+
"step": 1880
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 15.619834710743802,
|
1329 |
+
"grad_norm": 0.2043224722146988,
|
1330 |
+
"learning_rate": 2.4752094792522372e-05,
|
1331 |
+
"loss": 0.0152,
|
1332 |
+
"step": 1890
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 15.702479338842975,
|
1336 |
+
"grad_norm": 0.17056132853031158,
|
1337 |
+
"learning_rate": 2.385721590177754e-05,
|
1338 |
+
"loss": 0.0154,
|
1339 |
+
"step": 1900
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 15.785123966942148,
|
1343 |
+
"grad_norm": 0.20709876716136932,
|
1344 |
+
"learning_rate": 2.2976617230534202e-05,
|
1345 |
+
"loss": 0.0132,
|
1346 |
+
"step": 1910
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 15.867768595041323,
|
1350 |
+
"grad_norm": 0.14429618418216705,
|
1351 |
+
"learning_rate": 2.211046393091374e-05,
|
1352 |
+
"loss": 0.0165,
|
1353 |
+
"step": 1920
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 15.950413223140496,
|
1357 |
+
"grad_norm": 0.15559087693691254,
|
1358 |
+
"learning_rate": 2.1258918445876263e-05,
|
1359 |
+
"loss": 0.0147,
|
1360 |
+
"step": 1930
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 16.03305785123967,
|
1364 |
+
"grad_norm": 0.17711517214775085,
|
1365 |
+
"learning_rate": 2.0422140478755248e-05,
|
1366 |
+
"loss": 0.016,
|
1367 |
+
"step": 1940
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 16.115702479338843,
|
1371 |
+
"grad_norm": 0.14507338404655457,
|
1372 |
+
"learning_rate": 1.9600286963305957e-05,
|
1373 |
+
"loss": 0.0126,
|
1374 |
+
"step": 1950
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 16.198347107438018,
|
1378 |
+
"grad_norm": 0.12779353559017181,
|
1379 |
+
"learning_rate": 1.8793512034273242e-05,
|
1380 |
+
"loss": 0.0133,
|
1381 |
+
"step": 1960
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 16.28099173553719,
|
1385 |
+
"grad_norm": 0.1657179296016693,
|
1386 |
+
"learning_rate": 1.8001966998484253e-05,
|
1387 |
+
"loss": 0.0134,
|
1388 |
+
"step": 1970
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 16.363636363636363,
|
1392 |
+
"grad_norm": 0.17538750171661377,
|
1393 |
+
"learning_rate": 1.7225800306471696e-05,
|
1394 |
+
"loss": 0.0135,
|
1395 |
+
"step": 1980
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 16.446280991735538,
|
1399 |
+
"grad_norm": 0.18412277102470398,
|
1400 |
+
"learning_rate": 1.6465157524632436e-05,
|
1401 |
+
"loss": 0.0138,
|
1402 |
+
"step": 1990
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 16.52892561983471,
|
1406 |
+
"grad_norm": 0.152567520737648,
|
1407 |
+
"learning_rate": 1.5720181307927373e-05,
|
1408 |
+
"loss": 0.012,
|
1409 |
+
"step": 2000
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 16.611570247933884,
|
1413 |
+
"grad_norm": 0.2471626251935959,
|
1414 |
+
"learning_rate": 1.4991011373127084e-05,
|
1415 |
+
"loss": 0.0149,
|
1416 |
+
"step": 2010
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 16.694214876033058,
|
1420 |
+
"grad_norm": 0.22137291729450226,
|
1421 |
+
"learning_rate": 1.4277784472608635e-05,
|
1422 |
+
"loss": 0.0126,
|
1423 |
+
"step": 2020
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 16.776859504132233,
|
1427 |
+
"grad_norm": 0.16841760277748108,
|
1428 |
+
"learning_rate": 1.3580634368708366e-05,
|
1429 |
+
"loss": 0.0151,
|
1430 |
+
"step": 2030
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 16.859504132231404,
|
1434 |
+
"grad_norm": 0.17957067489624023,
|
1435 |
+
"learning_rate": 1.2899691808635283e-05,
|
1436 |
+
"loss": 0.015,
|
1437 |
+
"step": 2040
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 16.94214876033058,
|
1441 |
+
"grad_norm": 0.17319756746292114,
|
1442 |
+
"learning_rate": 1.2235084499950178e-05,
|
1443 |
+
"loss": 0.0128,
|
1444 |
+
"step": 2050
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 17.024793388429753,
|
1448 |
+
"grad_norm": 0.14869403839111328,
|
1449 |
+
"learning_rate": 1.1586937086614525e-05,
|
1450 |
+
"loss": 0.0133,
|
1451 |
+
"step": 2060
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 17.107438016528924,
|
1455 |
+
"grad_norm": 0.15348997712135315,
|
1456 |
+
"learning_rate": 1.0955371125614223e-05,
|
1457 |
+
"loss": 0.0137,
|
1458 |
+
"step": 2070
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 17.1900826446281,
|
1462 |
+
"grad_norm": 0.2330683022737503,
|
1463 |
+
"learning_rate": 1.0340505064162076e-05,
|
1464 |
+
"loss": 0.0143,
|
1465 |
+
"step": 2080
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 17.272727272727273,
|
1469 |
+
"grad_norm": 0.17106673121452332,
|
1470 |
+
"learning_rate": 9.742454217483632e-06,
|
1471 |
+
"loss": 0.014,
|
1472 |
+
"step": 2090
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 17.355371900826448,
|
1476 |
+
"grad_norm": 0.15164901316165924,
|
1477 |
+
"learning_rate": 9.161330747190321e-06,
|
1478 |
+
"loss": 0.0151,
|
1479 |
+
"step": 2100
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 17.43801652892562,
|
1483 |
+
"grad_norm": 0.18551285564899445,
|
1484 |
+
"learning_rate": 8.59724364024409e-06,
|
1485 |
+
"loss": 0.011,
|
1486 |
+
"step": 2110
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 17.520661157024794,
|
1490 |
+
"grad_norm": 0.13972045481204987,
|
1491 |
+
"learning_rate": 8.050298688517444e-06,
|
1492 |
+
"loss": 0.0104,
|
1493 |
+
"step": 2120
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 17.60330578512397,
|
1497 |
+
"grad_norm": 0.17032235860824585,
|
1498 |
+
"learning_rate": 7.520598468952578e-06,
|
1499 |
+
"loss": 0.0136,
|
1500 |
+
"step": 2130
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 17.68595041322314,
|
1504 |
+
"grad_norm": 0.14614145457744598,
|
1505 |
+
"learning_rate": 7.0082423243236995e-06,
|
1506 |
+
"loss": 0.0134,
|
1507 |
+
"step": 2140
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 17.768595041322314,
|
1511 |
+
"grad_norm": 0.19540473818778992,
|
1512 |
+
"learning_rate": 6.513326344605608e-06,
|
1513 |
+
"loss": 0.013,
|
1514 |
+
"step": 2150
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 17.85123966942149,
|
1518 |
+
"grad_norm": 0.16889305412769318,
|
1519 |
+
"learning_rate": 6.035943348952611e-06,
|
1520 |
+
"loss": 0.0128,
|
1521 |
+
"step": 2160
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 17.93388429752066,
|
1525 |
+
"grad_norm": 0.13315258920192719,
|
1526 |
+
"learning_rate": 5.576182868290669e-06,
|
1527 |
+
"loss": 0.0102,
|
1528 |
+
"step": 2170
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 18.016528925619834,
|
1532 |
+
"grad_norm": 0.197654590010643,
|
1533 |
+
"learning_rate": 5.134131128526243e-06,
|
1534 |
+
"loss": 0.012,
|
1535 |
+
"step": 2180
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 18.09917355371901,
|
1539 |
+
"grad_norm": 0.16169218719005585,
|
1540 |
+
"learning_rate": 4.709871034375146e-06,
|
1541 |
+
"loss": 0.0103,
|
1542 |
+
"step": 2190
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 18.181818181818183,
|
1546 |
+
"grad_norm": 0.1659018099308014,
|
1547 |
+
"learning_rate": 4.303482153814064e-06,
|
1548 |
+
"loss": 0.0125,
|
1549 |
+
"step": 2200
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 18.264462809917354,
|
1553 |
+
"grad_norm": 0.3515506982803345,
|
1554 |
+
"learning_rate": 3.915040703157991e-06,
|
1555 |
+
"loss": 0.0141,
|
1556 |
+
"step": 2210
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 18.34710743801653,
|
1560 |
+
"grad_norm": 0.1042797714471817,
|
1561 |
+
"learning_rate": 3.5446195327662133e-06,
|
1562 |
+
"loss": 0.0128,
|
1563 |
+
"step": 2220
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 18.429752066115704,
|
1567 |
+
"grad_norm": 0.19790691137313843,
|
1568 |
+
"learning_rate": 3.1922881133795825e-06,
|
1569 |
+
"loss": 0.015,
|
1570 |
+
"step": 2230
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 18.512396694214875,
|
1574 |
+
"grad_norm": 0.21495257318019867,
|
1575 |
+
"learning_rate": 2.8581125230915605e-06,
|
1576 |
+
"loss": 0.011,
|
1577 |
+
"step": 2240
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 18.59504132231405,
|
1581 |
+
"grad_norm": 0.35978537797927856,
|
1582 |
+
"learning_rate": 2.5421554349556733e-06,
|
1583 |
+
"loss": 0.0119,
|
1584 |
+
"step": 2250
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 18.677685950413224,
|
1588 |
+
"grad_norm": 0.1598854809999466,
|
1589 |
+
"learning_rate": 2.2444761052313856e-06,
|
1590 |
+
"loss": 0.0136,
|
1591 |
+
"step": 2260
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 18.760330578512395,
|
1595 |
+
"grad_norm": 0.1521085500717163,
|
1596 |
+
"learning_rate": 1.965130362270917e-06,
|
1597 |
+
"loss": 0.0115,
|
1598 |
+
"step": 2270
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 18.84297520661157,
|
1602 |
+
"grad_norm": 0.185346782207489,
|
1603 |
+
"learning_rate": 1.704170596048893e-06,
|
1604 |
+
"loss": 0.0134,
|
1605 |
+
"step": 2280
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 18.925619834710744,
|
1609 |
+
"grad_norm": 0.23003143072128296,
|
1610 |
+
"learning_rate": 1.4616457483368263e-06,
|
1611 |
+
"loss": 0.0102,
|
1612 |
+
"step": 2290
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 19.00826446280992,
|
1616 |
+
"grad_norm": 0.23142951726913452,
|
1617 |
+
"learning_rate": 1.2376013035243606e-06,
|
1618 |
+
"loss": 0.0116,
|
1619 |
+
"step": 2300
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 19.09090909090909,
|
1623 |
+
"grad_norm": 0.12257048487663269,
|
1624 |
+
"learning_rate": 1.0320792800888602e-06,
|
1625 |
+
"loss": 0.0132,
|
1626 |
+
"step": 2310
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 19.173553719008265,
|
1630 |
+
"grad_norm": 0.11083350330591202,
|
1631 |
+
"learning_rate": 8.451182227150467e-07,
|
1632 |
+
"loss": 0.0128,
|
1633 |
+
"step": 2320
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 19.25619834710744,
|
1637 |
+
"grad_norm": 0.17381551861763,
|
1638 |
+
"learning_rate": 6.767531950661377e-07,
|
1639 |
+
"loss": 0.0137,
|
1640 |
+
"step": 2330
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 19.33884297520661,
|
1644 |
+
"grad_norm": 0.20176014304161072,
|
1645 |
+
"learning_rate": 5.270157732077951e-07,
|
1646 |
+
"loss": 0.0151,
|
1647 |
+
"step": 2340
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 19.421487603305785,
|
1651 |
+
"grad_norm": 0.17078755795955658,
|
1652 |
+
"learning_rate": 3.959340396862188e-07,
|
1653 |
+
"loss": 0.0125,
|
1654 |
+
"step": 2350
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 19.50413223140496,
|
1658 |
+
"grad_norm": 0.23273628950119019,
|
1659 |
+
"learning_rate": 2.8353257826136957e-07,
|
1660 |
+
"loss": 0.0148,
|
1661 |
+
"step": 2360
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 19.586776859504134,
|
1665 |
+
"grad_norm": 0.11116746068000793,
|
1666 |
+
"learning_rate": 1.8983246929642485e-07,
|
1667 |
+
"loss": 0.0119,
|
1668 |
+
"step": 2370
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 19.669421487603305,
|
1672 |
+
"grad_norm": 0.16333059966564178,
|
1673 |
+
"learning_rate": 1.1485128580422944e-07,
|
1674 |
+
"loss": 0.0108,
|
1675 |
+
"step": 2380
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 19.75206611570248,
|
1679 |
+
"grad_norm": 0.12590940296649933,
|
1680 |
+
"learning_rate": 5.860309015156551e-08,
|
1681 |
+
"loss": 0.0109,
|
1682 |
+
"step": 2390
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 19.834710743801654,
|
1686 |
+
"grad_norm": 0.12979190051555634,
|
1687 |
+
"learning_rate": 2.1098431421862076e-08,
|
1688 |
+
"loss": 0.0131,
|
1689 |
+
"step": 2400
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 19.917355371900825,
|
1693 |
+
"grad_norm": 0.18272912502288818,
|
1694 |
+
"learning_rate": 2.3443434366887673e-09,
|
1695 |
+
"loss": 0.0123,
|
1696 |
+
"step": 2410
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 19.958677685950413,
|
1700 |
+
"step": 2415,
|
1701 |
+
"total_flos": 3.496526197912458e+17,
|
1702 |
+
"train_loss": 0.034255825163599866,
|
1703 |
+
"train_runtime": 2767.1495,
|
1704 |
+
"train_samples_per_second": 55.855,
|
1705 |
+
"train_steps_per_second": 0.873
|
1706 |
+
}
|
1707 |
+
],
|
1708 |
+
"logging_steps": 10,
|
1709 |
+
"max_steps": 2415,
|
1710 |
+
"num_input_tokens_seen": 0,
|
1711 |
+
"num_train_epochs": 20,
|
1712 |
+
"save_steps": 10000,
|
1713 |
+
"stateful_callbacks": {
|
1714 |
+
"TrainerControl": {
|
1715 |
+
"args": {
|
1716 |
+
"should_epoch_stop": false,
|
1717 |
+
"should_evaluate": false,
|
1718 |
+
"should_log": false,
|
1719 |
+
"should_save": true,
|
1720 |
+
"should_training_stop": true
|
1721 |
+
},
|
1722 |
+
"attributes": {}
|
1723 |
+
}
|
1724 |
+
},
|
1725 |
+
"total_flos": 3.496526197912458e+17,
|
1726 |
+
"train_batch_size": 64,
|
1727 |
+
"trial_name": null,
|
1728 |
+
"trial_params": null
|
1729 |
+
}
|