Commit
•
27bd754
1
Parent(s):
ca868d2
Upload trainer_state.json
Browse files- trainer_state.json +1058 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1058 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.997580899624864,
|
| 5 |
+
"global_step": 85500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.02,
|
| 12 |
+
"learning_rate": 0.0019883135247577978,
|
| 13 |
+
"loss": 1.8357,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.04,
|
| 18 |
+
"learning_rate": 0.0019766270495155955,
|
| 19 |
+
"loss": 0.5446,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.05,
|
| 24 |
+
"learning_rate": 0.0019649405742733933,
|
| 25 |
+
"loss": 0.4222,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.07,
|
| 30 |
+
"learning_rate": 0.0019532540990311914,
|
| 31 |
+
"loss": 0.3959,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.09,
|
| 36 |
+
"learning_rate": 0.001941567623788989,
|
| 37 |
+
"loss": 0.3926,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.11,
|
| 42 |
+
"learning_rate": 0.001929881148546787,
|
| 43 |
+
"loss": 0.3792,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.12,
|
| 48 |
+
"learning_rate": 0.0019181946733045848,
|
| 49 |
+
"loss": 0.3692,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.14,
|
| 54 |
+
"learning_rate": 0.0019065081980623824,
|
| 55 |
+
"loss": 0.3622,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.16,
|
| 60 |
+
"learning_rate": 0.0018948217228201803,
|
| 61 |
+
"loss": 0.3643,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.18,
|
| 66 |
+
"learning_rate": 0.0018831352475779778,
|
| 67 |
+
"loss": 0.3535,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.19,
|
| 72 |
+
"learning_rate": 0.0018714487723357758,
|
| 73 |
+
"loss": 0.3642,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.21,
|
| 78 |
+
"learning_rate": 0.0018597622970935737,
|
| 79 |
+
"loss": 0.3462,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.23,
|
| 84 |
+
"learning_rate": 0.0018480758218513713,
|
| 85 |
+
"loss": 0.3508,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.25,
|
| 90 |
+
"learning_rate": 0.0018363893466091692,
|
| 91 |
+
"loss": 0.3387,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.26,
|
| 96 |
+
"learning_rate": 0.0018247028713669672,
|
| 97 |
+
"loss": 0.332,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.28,
|
| 102 |
+
"learning_rate": 0.0018130163961247647,
|
| 103 |
+
"loss": 0.3429,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.3,
|
| 108 |
+
"learning_rate": 0.0018013299208825626,
|
| 109 |
+
"loss": 0.3331,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.32,
|
| 114 |
+
"learning_rate": 0.0017896434456403606,
|
| 115 |
+
"loss": 0.3313,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.33,
|
| 120 |
+
"learning_rate": 0.0017779569703981581,
|
| 121 |
+
"loss": 0.3336,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.35,
|
| 126 |
+
"learning_rate": 0.001766270495155956,
|
| 127 |
+
"loss": 0.327,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.37,
|
| 132 |
+
"learning_rate": 0.001754584019913754,
|
| 133 |
+
"loss": 0.3274,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.39,
|
| 138 |
+
"learning_rate": 0.0017428975446715515,
|
| 139 |
+
"loss": 0.3202,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.4,
|
| 144 |
+
"learning_rate": 0.0017312110694293495,
|
| 145 |
+
"loss": 0.323,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.42,
|
| 150 |
+
"learning_rate": 0.0017195245941871475,
|
| 151 |
+
"loss": 0.3146,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.44,
|
| 156 |
+
"learning_rate": 0.001707838118944945,
|
| 157 |
+
"loss": 0.3192,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.46,
|
| 162 |
+
"learning_rate": 0.001696151643702743,
|
| 163 |
+
"loss": 0.3235,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.47,
|
| 168 |
+
"learning_rate": 0.0016844651684605404,
|
| 169 |
+
"loss": 0.3212,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.49,
|
| 174 |
+
"learning_rate": 0.0016727786932183384,
|
| 175 |
+
"loss": 0.3048,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.51,
|
| 180 |
+
"learning_rate": 0.0016610922179761364,
|
| 181 |
+
"loss": 0.3092,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.53,
|
| 186 |
+
"learning_rate": 0.0016494057427339339,
|
| 187 |
+
"loss": 0.3054,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.54,
|
| 192 |
+
"learning_rate": 0.0016377192674917318,
|
| 193 |
+
"loss": 0.3075,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.56,
|
| 198 |
+
"learning_rate": 0.0016260327922495298,
|
| 199 |
+
"loss": 0.3143,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.58,
|
| 204 |
+
"learning_rate": 0.0016143463170073273,
|
| 205 |
+
"loss": 0.2979,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.6,
|
| 210 |
+
"learning_rate": 0.0016026598417651253,
|
| 211 |
+
"loss": 0.3036,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.61,
|
| 216 |
+
"learning_rate": 0.0015909733665229232,
|
| 217 |
+
"loss": 0.2952,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.63,
|
| 222 |
+
"learning_rate": 0.0015792868912807207,
|
| 223 |
+
"loss": 0.3052,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.65,
|
| 228 |
+
"learning_rate": 0.0015676004160385187,
|
| 229 |
+
"loss": 0.3079,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.67,
|
| 234 |
+
"learning_rate": 0.0015559139407963166,
|
| 235 |
+
"loss": 0.3024,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.68,
|
| 240 |
+
"learning_rate": 0.0015442274655541142,
|
| 241 |
+
"loss": 0.3012,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.7,
|
| 246 |
+
"learning_rate": 0.001532540990311912,
|
| 247 |
+
"loss": 0.2966,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.72,
|
| 252 |
+
"learning_rate": 0.00152085451506971,
|
| 253 |
+
"loss": 0.3007,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.74,
|
| 258 |
+
"learning_rate": 0.0015091680398275076,
|
| 259 |
+
"loss": 0.2952,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.75,
|
| 264 |
+
"learning_rate": 0.0014974815645853055,
|
| 265 |
+
"loss": 0.2933,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.77,
|
| 270 |
+
"learning_rate": 0.0014857950893431033,
|
| 271 |
+
"loss": 0.2942,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.79,
|
| 276 |
+
"learning_rate": 0.001474108614100901,
|
| 277 |
+
"loss": 0.298,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.81,
|
| 282 |
+
"learning_rate": 0.001462422138858699,
|
| 283 |
+
"loss": 0.2969,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.82,
|
| 288 |
+
"learning_rate": 0.0014507356636164965,
|
| 289 |
+
"loss": 0.2905,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.84,
|
| 294 |
+
"learning_rate": 0.0014390491883742944,
|
| 295 |
+
"loss": 0.2926,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.86,
|
| 300 |
+
"learning_rate": 0.0014273627131320924,
|
| 301 |
+
"loss": 0.2925,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.88,
|
| 306 |
+
"learning_rate": 0.00141567623788989,
|
| 307 |
+
"loss": 0.2911,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.89,
|
| 312 |
+
"learning_rate": 0.0014039897626476879,
|
| 313 |
+
"loss": 0.292,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.91,
|
| 318 |
+
"learning_rate": 0.0013923032874054858,
|
| 319 |
+
"loss": 0.2962,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.93,
|
| 324 |
+
"learning_rate": 0.0013806168121632833,
|
| 325 |
+
"loss": 0.289,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.95,
|
| 330 |
+
"learning_rate": 0.0013689303369210813,
|
| 331 |
+
"loss": 0.2966,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.96,
|
| 336 |
+
"learning_rate": 0.0013572438616788792,
|
| 337 |
+
"loss": 0.2871,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.98,
|
| 342 |
+
"learning_rate": 0.0013455573864366768,
|
| 343 |
+
"loss": 0.2855,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 1.0,
|
| 348 |
+
"learning_rate": 0.0013338709111944747,
|
| 349 |
+
"loss": 0.2823,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.0,
|
| 354 |
+
"eval_loss": 0.2679596245288849,
|
| 355 |
+
"eval_runtime": 1712.4301,
|
| 356 |
+
"eval_samples_per_second": 148.053,
|
| 357 |
+
"eval_steps_per_second": 6.169,
|
| 358 |
+
"step": 28523
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 1.02,
|
| 362 |
+
"learning_rate": 0.0013221844359522727,
|
| 363 |
+
"loss": 0.2827,
|
| 364 |
+
"step": 29000
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.03,
|
| 368 |
+
"learning_rate": 0.0013104979607100702,
|
| 369 |
+
"loss": 0.2822,
|
| 370 |
+
"step": 29500
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 1.05,
|
| 374 |
+
"learning_rate": 0.0012988114854678681,
|
| 375 |
+
"loss": 0.2789,
|
| 376 |
+
"step": 30000
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 1.07,
|
| 380 |
+
"learning_rate": 0.0012871250102256659,
|
| 381 |
+
"loss": 0.274,
|
| 382 |
+
"step": 30500
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.09,
|
| 386 |
+
"learning_rate": 0.0012754385349834636,
|
| 387 |
+
"loss": 0.2786,
|
| 388 |
+
"step": 31000
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.1,
|
| 392 |
+
"learning_rate": 0.0012637520597412616,
|
| 393 |
+
"loss": 0.2728,
|
| 394 |
+
"step": 31500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.12,
|
| 398 |
+
"learning_rate": 0.001252065584499059,
|
| 399 |
+
"loss": 0.2812,
|
| 400 |
+
"step": 32000
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 1.14,
|
| 404 |
+
"learning_rate": 0.001240379109256857,
|
| 405 |
+
"loss": 0.2763,
|
| 406 |
+
"step": 32500
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.16,
|
| 410 |
+
"learning_rate": 0.001228692634014655,
|
| 411 |
+
"loss": 0.2783,
|
| 412 |
+
"step": 33000
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 1.17,
|
| 416 |
+
"learning_rate": 0.0012170061587724525,
|
| 417 |
+
"loss": 0.2767,
|
| 418 |
+
"step": 33500
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 1.19,
|
| 422 |
+
"learning_rate": 0.0012053196835302505,
|
| 423 |
+
"loss": 0.2828,
|
| 424 |
+
"step": 34000
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.21,
|
| 428 |
+
"learning_rate": 0.0011936332082880484,
|
| 429 |
+
"loss": 0.2765,
|
| 430 |
+
"step": 34500
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.23,
|
| 434 |
+
"learning_rate": 0.001181946733045846,
|
| 435 |
+
"loss": 0.2723,
|
| 436 |
+
"step": 35000
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.24,
|
| 440 |
+
"learning_rate": 0.0011702602578036439,
|
| 441 |
+
"loss": 0.2768,
|
| 442 |
+
"step": 35500
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 1.26,
|
| 446 |
+
"learning_rate": 0.0011585737825614418,
|
| 447 |
+
"loss": 0.2729,
|
| 448 |
+
"step": 36000
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 1.28,
|
| 452 |
+
"learning_rate": 0.0011468873073192394,
|
| 453 |
+
"loss": 0.2785,
|
| 454 |
+
"step": 36500
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 1.3,
|
| 458 |
+
"learning_rate": 0.0011352008320770373,
|
| 459 |
+
"loss": 0.2765,
|
| 460 |
+
"step": 37000
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 1.31,
|
| 464 |
+
"learning_rate": 0.001123514356834835,
|
| 465 |
+
"loss": 0.2693,
|
| 466 |
+
"step": 37500
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.33,
|
| 470 |
+
"learning_rate": 0.0011118278815926328,
|
| 471 |
+
"loss": 0.266,
|
| 472 |
+
"step": 38000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.35,
|
| 476 |
+
"learning_rate": 0.0011001414063504307,
|
| 477 |
+
"loss": 0.2737,
|
| 478 |
+
"step": 38500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.37,
|
| 482 |
+
"learning_rate": 0.0010884549311082285,
|
| 483 |
+
"loss": 0.2774,
|
| 484 |
+
"step": 39000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 1.38,
|
| 488 |
+
"learning_rate": 0.0010767684558660262,
|
| 489 |
+
"loss": 0.2702,
|
| 490 |
+
"step": 39500
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 1.4,
|
| 494 |
+
"learning_rate": 0.0010650819806238242,
|
| 495 |
+
"loss": 0.2772,
|
| 496 |
+
"step": 40000
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 1.42,
|
| 500 |
+
"learning_rate": 0.0010533955053816217,
|
| 501 |
+
"loss": 0.2718,
|
| 502 |
+
"step": 40500
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 1.44,
|
| 506 |
+
"learning_rate": 0.0010417090301394196,
|
| 507 |
+
"loss": 0.2777,
|
| 508 |
+
"step": 41000
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.45,
|
| 512 |
+
"learning_rate": 0.0010300225548972176,
|
| 513 |
+
"loss": 0.2732,
|
| 514 |
+
"step": 41500
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.47,
|
| 518 |
+
"learning_rate": 0.0010183360796550151,
|
| 519 |
+
"loss": 0.2716,
|
| 520 |
+
"step": 42000
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.49,
|
| 524 |
+
"learning_rate": 0.001006649604412813,
|
| 525 |
+
"loss": 0.2728,
|
| 526 |
+
"step": 42500
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 1.51,
|
| 530 |
+
"learning_rate": 0.0009949631291706108,
|
| 531 |
+
"loss": 0.2792,
|
| 532 |
+
"step": 43000
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 1.53,
|
| 536 |
+
"learning_rate": 0.0009832766539284088,
|
| 537 |
+
"loss": 0.2709,
|
| 538 |
+
"step": 43500
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.54,
|
| 542 |
+
"learning_rate": 0.0009715901786862065,
|
| 543 |
+
"loss": 0.2664,
|
| 544 |
+
"step": 44000
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 1.56,
|
| 548 |
+
"learning_rate": 0.0009599037034440042,
|
| 549 |
+
"loss": 0.2728,
|
| 550 |
+
"step": 44500
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.58,
|
| 554 |
+
"learning_rate": 0.0009482172282018021,
|
| 555 |
+
"loss": 0.2701,
|
| 556 |
+
"step": 45000
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.6,
|
| 560 |
+
"learning_rate": 0.0009365307529595999,
|
| 561 |
+
"loss": 0.2707,
|
| 562 |
+
"step": 45500
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.61,
|
| 566 |
+
"learning_rate": 0.0009248442777173977,
|
| 567 |
+
"loss": 0.27,
|
| 568 |
+
"step": 46000
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 1.63,
|
| 572 |
+
"learning_rate": 0.0009131578024751954,
|
| 573 |
+
"loss": 0.2719,
|
| 574 |
+
"step": 46500
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 1.65,
|
| 578 |
+
"learning_rate": 0.0009014713272329933,
|
| 579 |
+
"loss": 0.2682,
|
| 580 |
+
"step": 47000
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 1.67,
|
| 584 |
+
"learning_rate": 0.0008897848519907911,
|
| 585 |
+
"loss": 0.2675,
|
| 586 |
+
"step": 47500
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.68,
|
| 590 |
+
"learning_rate": 0.0008780983767485888,
|
| 591 |
+
"loss": 0.2689,
|
| 592 |
+
"step": 48000
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.7,
|
| 596 |
+
"learning_rate": 0.0008664119015063867,
|
| 597 |
+
"loss": 0.2629,
|
| 598 |
+
"step": 48500
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.72,
|
| 602 |
+
"learning_rate": 0.0008547254262641845,
|
| 603 |
+
"loss": 0.2636,
|
| 604 |
+
"step": 49000
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.74,
|
| 608 |
+
"learning_rate": 0.0008430389510219822,
|
| 609 |
+
"loss": 0.2635,
|
| 610 |
+
"step": 49500
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 1.75,
|
| 614 |
+
"learning_rate": 0.0008313524757797801,
|
| 615 |
+
"loss": 0.2645,
|
| 616 |
+
"step": 50000
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 1.77,
|
| 620 |
+
"learning_rate": 0.0008196660005375779,
|
| 621 |
+
"loss": 0.2622,
|
| 622 |
+
"step": 50500
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.79,
|
| 626 |
+
"learning_rate": 0.0008079795252953757,
|
| 627 |
+
"loss": 0.2627,
|
| 628 |
+
"step": 51000
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 1.81,
|
| 632 |
+
"learning_rate": 0.0007962930500531734,
|
| 633 |
+
"loss": 0.2576,
|
| 634 |
+
"step": 51500
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.82,
|
| 638 |
+
"learning_rate": 0.0007846065748109713,
|
| 639 |
+
"loss": 0.2662,
|
| 640 |
+
"step": 52000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.84,
|
| 644 |
+
"learning_rate": 0.0007729200995687691,
|
| 645 |
+
"loss": 0.2574,
|
| 646 |
+
"step": 52500
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.86,
|
| 650 |
+
"learning_rate": 0.0007612336243265668,
|
| 651 |
+
"loss": 0.2657,
|
| 652 |
+
"step": 53000
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 1.88,
|
| 656 |
+
"learning_rate": 0.0007495471490843647,
|
| 657 |
+
"loss": 0.2597,
|
| 658 |
+
"step": 53500
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 1.89,
|
| 662 |
+
"learning_rate": 0.0007378606738421625,
|
| 663 |
+
"loss": 0.2529,
|
| 664 |
+
"step": 54000
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 1.91,
|
| 668 |
+
"learning_rate": 0.0007261741985999603,
|
| 669 |
+
"loss": 0.2514,
|
| 670 |
+
"step": 54500
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.93,
|
| 674 |
+
"learning_rate": 0.0007144877233577581,
|
| 675 |
+
"loss": 0.2649,
|
| 676 |
+
"step": 55000
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.95,
|
| 680 |
+
"learning_rate": 0.000702801248115556,
|
| 681 |
+
"loss": 0.2621,
|
| 682 |
+
"step": 55500
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.96,
|
| 686 |
+
"learning_rate": 0.0006911147728733537,
|
| 687 |
+
"loss": 0.2591,
|
| 688 |
+
"step": 56000
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.98,
|
| 692 |
+
"learning_rate": 0.0006794282976311514,
|
| 693 |
+
"loss": 0.2655,
|
| 694 |
+
"step": 56500
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 2.0,
|
| 698 |
+
"learning_rate": 0.0006677418223889493,
|
| 699 |
+
"loss": 0.253,
|
| 700 |
+
"step": 57000
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 2.0,
|
| 704 |
+
"eval_loss": 0.24398696422576904,
|
| 705 |
+
"eval_runtime": 1707.3448,
|
| 706 |
+
"eval_samples_per_second": 148.494,
|
| 707 |
+
"eval_steps_per_second": 6.187,
|
| 708 |
+
"step": 57046
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 2.02,
|
| 712 |
+
"learning_rate": 0.0006560553471467471,
|
| 713 |
+
"loss": 0.2498,
|
| 714 |
+
"step": 57500
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 2.03,
|
| 718 |
+
"learning_rate": 0.0006443688719045449,
|
| 719 |
+
"loss": 0.2548,
|
| 720 |
+
"step": 58000
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 2.05,
|
| 724 |
+
"learning_rate": 0.0006326823966623427,
|
| 725 |
+
"loss": 0.2552,
|
| 726 |
+
"step": 58500
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 2.07,
|
| 730 |
+
"learning_rate": 0.0006209959214201405,
|
| 731 |
+
"loss": 0.2537,
|
| 732 |
+
"step": 59000
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 2.09,
|
| 736 |
+
"learning_rate": 0.0006093094461779383,
|
| 737 |
+
"loss": 0.2466,
|
| 738 |
+
"step": 59500
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 2.1,
|
| 742 |
+
"learning_rate": 0.000597622970935736,
|
| 743 |
+
"loss": 0.2544,
|
| 744 |
+
"step": 60000
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.12,
|
| 748 |
+
"learning_rate": 0.0005859364956935339,
|
| 749 |
+
"loss": 0.2527,
|
| 750 |
+
"step": 60500
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 2.14,
|
| 754 |
+
"learning_rate": 0.0005742500204513317,
|
| 755 |
+
"loss": 0.2471,
|
| 756 |
+
"step": 61000
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 2.16,
|
| 760 |
+
"learning_rate": 0.0005625635452091294,
|
| 761 |
+
"loss": 0.2549,
|
| 762 |
+
"step": 61500
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 2.17,
|
| 766 |
+
"learning_rate": 0.0005508770699669273,
|
| 767 |
+
"loss": 0.2455,
|
| 768 |
+
"step": 62000
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 2.19,
|
| 772 |
+
"learning_rate": 0.0005391905947247251,
|
| 773 |
+
"loss": 0.2487,
|
| 774 |
+
"step": 62500
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 2.21,
|
| 778 |
+
"learning_rate": 0.0005275041194825229,
|
| 779 |
+
"loss": 0.2438,
|
| 780 |
+
"step": 63000
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 2.23,
|
| 784 |
+
"learning_rate": 0.0005158176442403207,
|
| 785 |
+
"loss": 0.2501,
|
| 786 |
+
"step": 63500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.24,
|
| 790 |
+
"learning_rate": 0.0005041311689981185,
|
| 791 |
+
"loss": 0.256,
|
| 792 |
+
"step": 64000
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 2.26,
|
| 796 |
+
"learning_rate": 0.0004924446937559163,
|
| 797 |
+
"loss": 0.2507,
|
| 798 |
+
"step": 64500
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 2.28,
|
| 802 |
+
"learning_rate": 0.0004807582185137141,
|
| 803 |
+
"loss": 0.2445,
|
| 804 |
+
"step": 65000
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 2.3,
|
| 808 |
+
"learning_rate": 0.0004690717432715119,
|
| 809 |
+
"loss": 0.2506,
|
| 810 |
+
"step": 65500
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 2.31,
|
| 814 |
+
"learning_rate": 0.00045738526802930967,
|
| 815 |
+
"loss": 0.2516,
|
| 816 |
+
"step": 66000
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 2.33,
|
| 820 |
+
"learning_rate": 0.0004456987927871075,
|
| 821 |
+
"loss": 0.2439,
|
| 822 |
+
"step": 66500
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 2.35,
|
| 826 |
+
"learning_rate": 0.00043401231754490525,
|
| 827 |
+
"loss": 0.2501,
|
| 828 |
+
"step": 67000
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.37,
|
| 832 |
+
"learning_rate": 0.0004223258423027031,
|
| 833 |
+
"loss": 0.2412,
|
| 834 |
+
"step": 67500
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 2.38,
|
| 838 |
+
"learning_rate": 0.0004106393670605009,
|
| 839 |
+
"loss": 0.2455,
|
| 840 |
+
"step": 68000
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 2.4,
|
| 844 |
+
"learning_rate": 0.0003989528918182987,
|
| 845 |
+
"loss": 0.2464,
|
| 846 |
+
"step": 68500
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.42,
|
| 850 |
+
"learning_rate": 0.0003872664165760965,
|
| 851 |
+
"loss": 0.2471,
|
| 852 |
+
"step": 69000
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 2.44,
|
| 856 |
+
"learning_rate": 0.00037557994133389426,
|
| 857 |
+
"loss": 0.2453,
|
| 858 |
+
"step": 69500
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 2.45,
|
| 862 |
+
"learning_rate": 0.0003638934660916921,
|
| 863 |
+
"loss": 0.2461,
|
| 864 |
+
"step": 70000
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 2.47,
|
| 868 |
+
"learning_rate": 0.0003522069908494899,
|
| 869 |
+
"loss": 0.2433,
|
| 870 |
+
"step": 70500
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.49,
|
| 874 |
+
"learning_rate": 0.0003405205156072877,
|
| 875 |
+
"loss": 0.253,
|
| 876 |
+
"step": 71000
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 2.51,
|
| 880 |
+
"learning_rate": 0.00032883404036508553,
|
| 881 |
+
"loss": 0.2444,
|
| 882 |
+
"step": 71500
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 2.52,
|
| 886 |
+
"learning_rate": 0.00031714756512288327,
|
| 887 |
+
"loss": 0.2446,
|
| 888 |
+
"step": 72000
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 2.54,
|
| 892 |
+
"learning_rate": 0.0003054610898806811,
|
| 893 |
+
"loss": 0.2454,
|
| 894 |
+
"step": 72500
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 2.56,
|
| 898 |
+
"learning_rate": 0.0002937746146384789,
|
| 899 |
+
"loss": 0.2487,
|
| 900 |
+
"step": 73000
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 2.58,
|
| 904 |
+
"learning_rate": 0.0002820881393962767,
|
| 905 |
+
"loss": 0.2407,
|
| 906 |
+
"step": 73500
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 2.59,
|
| 910 |
+
"learning_rate": 0.00027040166415407454,
|
| 911 |
+
"loss": 0.241,
|
| 912 |
+
"step": 74000
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.61,
|
| 916 |
+
"learning_rate": 0.0002587151889118723,
|
| 917 |
+
"loss": 0.2432,
|
| 918 |
+
"step": 74500
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 2.63,
|
| 922 |
+
"learning_rate": 0.00024702871366967007,
|
| 923 |
+
"loss": 0.2425,
|
| 924 |
+
"step": 75000
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 2.65,
|
| 928 |
+
"learning_rate": 0.0002353422384274679,
|
| 929 |
+
"loss": 0.2413,
|
| 930 |
+
"step": 75500
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 2.66,
|
| 934 |
+
"learning_rate": 0.0002236557631852657,
|
| 935 |
+
"loss": 0.2468,
|
| 936 |
+
"step": 76000
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 2.68,
|
| 940 |
+
"learning_rate": 0.0002119692879430635,
|
| 941 |
+
"loss": 0.2391,
|
| 942 |
+
"step": 76500
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 2.7,
|
| 946 |
+
"learning_rate": 0.00020028281270086128,
|
| 947 |
+
"loss": 0.238,
|
| 948 |
+
"step": 77000
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.72,
|
| 952 |
+
"learning_rate": 0.0001885963374586591,
|
| 953 |
+
"loss": 0.2399,
|
| 954 |
+
"step": 77500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.73,
|
| 958 |
+
"learning_rate": 0.0001769098622164569,
|
| 959 |
+
"loss": 0.2449,
|
| 960 |
+
"step": 78000
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 2.75,
|
| 964 |
+
"learning_rate": 0.0001652233869742547,
|
| 965 |
+
"loss": 0.2415,
|
| 966 |
+
"step": 78500
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 2.77,
|
| 970 |
+
"learning_rate": 0.0001535369117320525,
|
| 971 |
+
"loss": 0.2413,
|
| 972 |
+
"step": 79000
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 2.79,
|
| 976 |
+
"learning_rate": 0.0001418504364898503,
|
| 977 |
+
"loss": 0.2433,
|
| 978 |
+
"step": 79500
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 2.8,
|
| 982 |
+
"learning_rate": 0.00013016396124764808,
|
| 983 |
+
"loss": 0.2426,
|
| 984 |
+
"step": 80000
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 2.82,
|
| 988 |
+
"learning_rate": 0.0001184774860054459,
|
| 989 |
+
"loss": 0.2379,
|
| 990 |
+
"step": 80500
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.84,
|
| 994 |
+
"learning_rate": 0.0001067910107632437,
|
| 995 |
+
"loss": 0.2358,
|
| 996 |
+
"step": 81000
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.86,
|
| 1000 |
+
"learning_rate": 9.510453552104151e-05,
|
| 1001 |
+
"loss": 0.2455,
|
| 1002 |
+
"step": 81500
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 2.87,
|
| 1006 |
+
"learning_rate": 8.34180602788393e-05,
|
| 1007 |
+
"loss": 0.2449,
|
| 1008 |
+
"step": 82000
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 2.89,
|
| 1012 |
+
"learning_rate": 7.17315850366371e-05,
|
| 1013 |
+
"loss": 0.2399,
|
| 1014 |
+
"step": 82500
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 2.91,
|
| 1018 |
+
"learning_rate": 6.00451097944349e-05,
|
| 1019 |
+
"loss": 0.2366,
|
| 1020 |
+
"step": 83000
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 2.93,
|
| 1024 |
+
"learning_rate": 4.8358634552232706e-05,
|
| 1025 |
+
"loss": 0.2434,
|
| 1026 |
+
"step": 83500
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 2.94,
|
| 1030 |
+
"learning_rate": 3.6672159310030504e-05,
|
| 1031 |
+
"loss": 0.2407,
|
| 1032 |
+
"step": 84000
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.96,
|
| 1036 |
+
"learning_rate": 2.4985684067828304e-05,
|
| 1037 |
+
"loss": 0.2394,
|
| 1038 |
+
"step": 84500
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.98,
|
| 1042 |
+
"learning_rate": 1.3299208825626102e-05,
|
| 1043 |
+
"loss": 0.2439,
|
| 1044 |
+
"step": 85000
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 3.0,
|
| 1048 |
+
"learning_rate": 1.6127335834239036e-06,
|
| 1049 |
+
"loss": 0.2351,
|
| 1050 |
+
"step": 85500
|
| 1051 |
+
}
|
| 1052 |
+
],
|
| 1053 |
+
"max_steps": 85569,
|
| 1054 |
+
"num_train_epochs": 3,
|
| 1055 |
+
"total_flos": 2.7771623363208806e+17,
|
| 1056 |
+
"trial_name": null,
|
| 1057 |
+
"trial_params": null
|
| 1058 |
+
}
|