Training in progress, step 600, checkpoint
Browse files- last-checkpoint/trainer_state.json +1053 -3
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 957,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1058,6 +1058,1056 @@
|
|
| 1058 |
"learning_rate": 0.005,
|
| 1059 |
"loss": 3.4181106090545654,
|
| 1060 |
"step": 300
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1061 |
}
|
| 1062 |
],
|
| 1063 |
"logging_steps": 2,
|
|
@@ -1077,7 +2127,7 @@
|
|
| 1077 |
"attributes": {}
|
| 1078 |
}
|
| 1079 |
},
|
| 1080 |
-
"total_flos":
|
| 1081 |
"train_batch_size": 2,
|
| 1082 |
"trial_name": null,
|
| 1083 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.05267356041033252,
|
| 6 |
"eval_steps": 957,
|
| 7 |
+
"global_step": 600,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1058 |
"learning_rate": 0.005,
|
| 1059 |
"loss": 3.4181106090545654,
|
| 1060 |
"step": 300
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.02651235873986737,
|
| 1064 |
+
"grad_norm": 0.1962890625,
|
| 1065 |
+
"learning_rate": 0.005,
|
| 1066 |
+
"loss": 3.4428491592407227,
|
| 1067 |
+
"step": 302
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.026687937274568477,
|
| 1071 |
+
"grad_norm": 0.1865234375,
|
| 1072 |
+
"learning_rate": 0.005,
|
| 1073 |
+
"loss": 3.4298229217529297,
|
| 1074 |
+
"step": 304
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.026863515809269586,
|
| 1078 |
+
"grad_norm": 0.1767578125,
|
| 1079 |
+
"learning_rate": 0.005,
|
| 1080 |
+
"loss": 3.41455078125,
|
| 1081 |
+
"step": 306
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.027039094343970695,
|
| 1085 |
+
"grad_norm": 0.1669921875,
|
| 1086 |
+
"learning_rate": 0.005,
|
| 1087 |
+
"loss": 3.3820443153381348,
|
| 1088 |
+
"step": 308
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.027214672878671804,
|
| 1092 |
+
"grad_norm": 0.2119140625,
|
| 1093 |
+
"learning_rate": 0.005,
|
| 1094 |
+
"loss": 3.3708128929138184,
|
| 1095 |
+
"step": 310
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.027390251413372913,
|
| 1099 |
+
"grad_norm": 0.20703125,
|
| 1100 |
+
"learning_rate": 0.005,
|
| 1101 |
+
"loss": 3.3826849460601807,
|
| 1102 |
+
"step": 312
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.02756582994807402,
|
| 1106 |
+
"grad_norm": 0.212890625,
|
| 1107 |
+
"learning_rate": 0.005,
|
| 1108 |
+
"loss": 3.357750177383423,
|
| 1109 |
+
"step": 314
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.027741408482775128,
|
| 1113 |
+
"grad_norm": 0.1953125,
|
| 1114 |
+
"learning_rate": 0.005,
|
| 1115 |
+
"loss": 3.3657925128936768,
|
| 1116 |
+
"step": 316
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.027916987017476236,
|
| 1120 |
+
"grad_norm": 0.177734375,
|
| 1121 |
+
"learning_rate": 0.005,
|
| 1122 |
+
"loss": 3.367344379425049,
|
| 1123 |
+
"step": 318
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.028092565552177345,
|
| 1127 |
+
"grad_norm": 0.19140625,
|
| 1128 |
+
"learning_rate": 0.005,
|
| 1129 |
+
"loss": 3.3542888164520264,
|
| 1130 |
+
"step": 320
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.028268144086878454,
|
| 1134 |
+
"grad_norm": 0.1904296875,
|
| 1135 |
+
"learning_rate": 0.005,
|
| 1136 |
+
"loss": 3.349008798599243,
|
| 1137 |
+
"step": 322
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.028443722621579563,
|
| 1141 |
+
"grad_norm": 0.2099609375,
|
| 1142 |
+
"learning_rate": 0.005,
|
| 1143 |
+
"loss": 3.319000720977783,
|
| 1144 |
+
"step": 324
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.028619301156280672,
|
| 1148 |
+
"grad_norm": 0.2255859375,
|
| 1149 |
+
"learning_rate": 0.005,
|
| 1150 |
+
"loss": 3.3123488426208496,
|
| 1151 |
+
"step": 326
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.028794879690981778,
|
| 1155 |
+
"grad_norm": 0.275390625,
|
| 1156 |
+
"learning_rate": 0.005,
|
| 1157 |
+
"loss": 3.311730146408081,
|
| 1158 |
+
"step": 328
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.028970458225682887,
|
| 1162 |
+
"grad_norm": 0.21484375,
|
| 1163 |
+
"learning_rate": 0.005,
|
| 1164 |
+
"loss": 3.2922301292419434,
|
| 1165 |
+
"step": 330
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.029146036760383996,
|
| 1169 |
+
"grad_norm": 0.2099609375,
|
| 1170 |
+
"learning_rate": 0.005,
|
| 1171 |
+
"loss": 3.3116979598999023,
|
| 1172 |
+
"step": 332
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.029321615295085104,
|
| 1176 |
+
"grad_norm": 0.17578125,
|
| 1177 |
+
"learning_rate": 0.005,
|
| 1178 |
+
"loss": 3.314528465270996,
|
| 1179 |
+
"step": 334
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.029497193829786213,
|
| 1183 |
+
"grad_norm": 0.1533203125,
|
| 1184 |
+
"learning_rate": 0.005,
|
| 1185 |
+
"loss": 3.3112518787384033,
|
| 1186 |
+
"step": 336
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.029672772364487322,
|
| 1190 |
+
"grad_norm": 0.1484375,
|
| 1191 |
+
"learning_rate": 0.005,
|
| 1192 |
+
"loss": 3.283816337585449,
|
| 1193 |
+
"step": 338
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.029848350899188428,
|
| 1197 |
+
"grad_norm": 0.1337890625,
|
| 1198 |
+
"learning_rate": 0.005,
|
| 1199 |
+
"loss": 3.3049468994140625,
|
| 1200 |
+
"step": 340
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.030023929433889537,
|
| 1204 |
+
"grad_norm": 0.1328125,
|
| 1205 |
+
"learning_rate": 0.005,
|
| 1206 |
+
"loss": 3.2841060161590576,
|
| 1207 |
+
"step": 342
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.030199507968590646,
|
| 1211 |
+
"grad_norm": 0.1533203125,
|
| 1212 |
+
"learning_rate": 0.005,
|
| 1213 |
+
"loss": 3.2544920444488525,
|
| 1214 |
+
"step": 344
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.030375086503291755,
|
| 1218 |
+
"grad_norm": 0.15625,
|
| 1219 |
+
"learning_rate": 0.005,
|
| 1220 |
+
"loss": 3.2736458778381348,
|
| 1221 |
+
"step": 346
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.030550665037992863,
|
| 1225 |
+
"grad_norm": 0.142578125,
|
| 1226 |
+
"learning_rate": 0.005,
|
| 1227 |
+
"loss": 3.2455122470855713,
|
| 1228 |
+
"step": 348
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.030726243572693972,
|
| 1232 |
+
"grad_norm": 0.15234375,
|
| 1233 |
+
"learning_rate": 0.005,
|
| 1234 |
+
"loss": 3.2548434734344482,
|
| 1235 |
+
"step": 350
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.03090182210739508,
|
| 1239 |
+
"grad_norm": 0.146484375,
|
| 1240 |
+
"learning_rate": 0.005,
|
| 1241 |
+
"loss": 3.2415554523468018,
|
| 1242 |
+
"step": 352
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.031077400642096187,
|
| 1246 |
+
"grad_norm": 0.1513671875,
|
| 1247 |
+
"learning_rate": 0.005,
|
| 1248 |
+
"loss": 3.25886607170105,
|
| 1249 |
+
"step": 354
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.0312529791767973,
|
| 1253 |
+
"grad_norm": 0.1494140625,
|
| 1254 |
+
"learning_rate": 0.005,
|
| 1255 |
+
"loss": 3.2604076862335205,
|
| 1256 |
+
"step": 356
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.031428557711498405,
|
| 1260 |
+
"grad_norm": 0.1474609375,
|
| 1261 |
+
"learning_rate": 0.005,
|
| 1262 |
+
"loss": 3.248387336730957,
|
| 1263 |
+
"step": 358
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.03160413624619951,
|
| 1267 |
+
"grad_norm": 0.173828125,
|
| 1268 |
+
"learning_rate": 0.005,
|
| 1269 |
+
"loss": 3.2637195587158203,
|
| 1270 |
+
"step": 360
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.03177971478090062,
|
| 1274 |
+
"grad_norm": 0.1513671875,
|
| 1275 |
+
"learning_rate": 0.005,
|
| 1276 |
+
"loss": 3.2353899478912354,
|
| 1277 |
+
"step": 362
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.03195529331560173,
|
| 1281 |
+
"grad_norm": 0.142578125,
|
| 1282 |
+
"learning_rate": 0.005,
|
| 1283 |
+
"loss": 3.2072153091430664,
|
| 1284 |
+
"step": 364
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.03213087185030284,
|
| 1288 |
+
"grad_norm": 0.1611328125,
|
| 1289 |
+
"learning_rate": 0.005,
|
| 1290 |
+
"loss": 3.249500274658203,
|
| 1291 |
+
"step": 366
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.032306450385003946,
|
| 1295 |
+
"grad_norm": 0.177734375,
|
| 1296 |
+
"learning_rate": 0.005,
|
| 1297 |
+
"loss": 3.2054481506347656,
|
| 1298 |
+
"step": 368
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.03248202891970506,
|
| 1302 |
+
"grad_norm": 0.1767578125,
|
| 1303 |
+
"learning_rate": 0.005,
|
| 1304 |
+
"loss": 3.1932778358459473,
|
| 1305 |
+
"step": 370
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.032657607454406164,
|
| 1309 |
+
"grad_norm": 0.16015625,
|
| 1310 |
+
"learning_rate": 0.005,
|
| 1311 |
+
"loss": 3.197559118270874,
|
| 1312 |
+
"step": 372
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.03283318598910727,
|
| 1316 |
+
"grad_norm": 0.154296875,
|
| 1317 |
+
"learning_rate": 0.005,
|
| 1318 |
+
"loss": 3.158921957015991,
|
| 1319 |
+
"step": 374
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 0.03300876452380838,
|
| 1323 |
+
"grad_norm": 0.125,
|
| 1324 |
+
"learning_rate": 0.005,
|
| 1325 |
+
"loss": 3.163810968399048,
|
| 1326 |
+
"step": 376
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.03318434305850949,
|
| 1330 |
+
"grad_norm": 0.1376953125,
|
| 1331 |
+
"learning_rate": 0.005,
|
| 1332 |
+
"loss": 3.1478772163391113,
|
| 1333 |
+
"step": 378
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 0.0333599215932106,
|
| 1337 |
+
"grad_norm": 0.1328125,
|
| 1338 |
+
"learning_rate": 0.005,
|
| 1339 |
+
"loss": 3.1855390071868896,
|
| 1340 |
+
"step": 380
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.033535500127911705,
|
| 1344 |
+
"grad_norm": 0.142578125,
|
| 1345 |
+
"learning_rate": 0.005,
|
| 1346 |
+
"loss": 3.1721317768096924,
|
| 1347 |
+
"step": 382
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.03371107866261282,
|
| 1351 |
+
"grad_norm": 0.126953125,
|
| 1352 |
+
"learning_rate": 0.005,
|
| 1353 |
+
"loss": 3.1935954093933105,
|
| 1354 |
+
"step": 384
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.03388665719731392,
|
| 1358 |
+
"grad_norm": 0.2236328125,
|
| 1359 |
+
"learning_rate": 0.005,
|
| 1360 |
+
"loss": 3.162994861602783,
|
| 1361 |
+
"step": 386
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 0.03406223573201503,
|
| 1365 |
+
"grad_norm": 0.181640625,
|
| 1366 |
+
"learning_rate": 0.005,
|
| 1367 |
+
"loss": 3.1582443714141846,
|
| 1368 |
+
"step": 388
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.03423781426671614,
|
| 1372 |
+
"grad_norm": 0.1455078125,
|
| 1373 |
+
"learning_rate": 0.005,
|
| 1374 |
+
"loss": 3.1388156414031982,
|
| 1375 |
+
"step": 390
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.034413392801417246,
|
| 1379 |
+
"grad_norm": 0.14453125,
|
| 1380 |
+
"learning_rate": 0.005,
|
| 1381 |
+
"loss": 3.183506727218628,
|
| 1382 |
+
"step": 392
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.03458897133611836,
|
| 1386 |
+
"grad_norm": 0.154296875,
|
| 1387 |
+
"learning_rate": 0.005,
|
| 1388 |
+
"loss": 3.160357713699341,
|
| 1389 |
+
"step": 394
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.034764549870819464,
|
| 1393 |
+
"grad_norm": 0.1591796875,
|
| 1394 |
+
"learning_rate": 0.005,
|
| 1395 |
+
"loss": 3.164640188217163,
|
| 1396 |
+
"step": 396
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.034940128405520576,
|
| 1400 |
+
"grad_norm": 0.169921875,
|
| 1401 |
+
"learning_rate": 0.005,
|
| 1402 |
+
"loss": 3.127124786376953,
|
| 1403 |
+
"step": 398
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 0.03511570694022168,
|
| 1407 |
+
"grad_norm": 0.162109375,
|
| 1408 |
+
"learning_rate": 0.005,
|
| 1409 |
+
"loss": 3.1583406925201416,
|
| 1410 |
+
"step": 400
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.03529128547492279,
|
| 1414 |
+
"grad_norm": 0.140625,
|
| 1415 |
+
"learning_rate": 0.005,
|
| 1416 |
+
"loss": 3.1125779151916504,
|
| 1417 |
+
"step": 402
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.0354668640096239,
|
| 1421 |
+
"grad_norm": 0.1484375,
|
| 1422 |
+
"learning_rate": 0.005,
|
| 1423 |
+
"loss": 3.111574411392212,
|
| 1424 |
+
"step": 404
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.035642442544325005,
|
| 1428 |
+
"grad_norm": 0.2255859375,
|
| 1429 |
+
"learning_rate": 0.005,
|
| 1430 |
+
"loss": 3.1208906173706055,
|
| 1431 |
+
"step": 406
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.03581802107902612,
|
| 1435 |
+
"grad_norm": 0.12451171875,
|
| 1436 |
+
"learning_rate": 0.005,
|
| 1437 |
+
"loss": 3.1485278606414795,
|
| 1438 |
+
"step": 408
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 0.03599359961372722,
|
| 1442 |
+
"grad_norm": 0.11328125,
|
| 1443 |
+
"learning_rate": 0.005,
|
| 1444 |
+
"loss": 3.1263508796691895,
|
| 1445 |
+
"step": 410
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.036169178148428335,
|
| 1449 |
+
"grad_norm": 0.1396484375,
|
| 1450 |
+
"learning_rate": 0.005,
|
| 1451 |
+
"loss": 3.0847690105438232,
|
| 1452 |
+
"step": 412
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.03634475668312944,
|
| 1456 |
+
"grad_norm": 0.1435546875,
|
| 1457 |
+
"learning_rate": 0.005,
|
| 1458 |
+
"loss": 3.104984760284424,
|
| 1459 |
+
"step": 414
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.036520335217830546,
|
| 1463 |
+
"grad_norm": 0.134765625,
|
| 1464 |
+
"learning_rate": 0.005,
|
| 1465 |
+
"loss": 3.102842092514038,
|
| 1466 |
+
"step": 416
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.03669591375253166,
|
| 1470 |
+
"grad_norm": 0.126953125,
|
| 1471 |
+
"learning_rate": 0.005,
|
| 1472 |
+
"loss": 3.097351551055908,
|
| 1473 |
+
"step": 418
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.036871492287232764,
|
| 1477 |
+
"grad_norm": 0.1328125,
|
| 1478 |
+
"learning_rate": 0.005,
|
| 1479 |
+
"loss": 3.116666555404663,
|
| 1480 |
+
"step": 420
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.037047070821933877,
|
| 1484 |
+
"grad_norm": 0.1640625,
|
| 1485 |
+
"learning_rate": 0.005,
|
| 1486 |
+
"loss": 3.0793497562408447,
|
| 1487 |
+
"step": 422
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.03722264935663498,
|
| 1491 |
+
"grad_norm": 0.14453125,
|
| 1492 |
+
"learning_rate": 0.005,
|
| 1493 |
+
"loss": 3.136044502258301,
|
| 1494 |
+
"step": 424
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.037398227891336094,
|
| 1498 |
+
"grad_norm": 0.126953125,
|
| 1499 |
+
"learning_rate": 0.005,
|
| 1500 |
+
"loss": 3.1369822025299072,
|
| 1501 |
+
"step": 426
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 0.0375738064260372,
|
| 1505 |
+
"grad_norm": 0.150390625,
|
| 1506 |
+
"learning_rate": 0.005,
|
| 1507 |
+
"loss": 3.0935890674591064,
|
| 1508 |
+
"step": 428
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.037749384960738305,
|
| 1512 |
+
"grad_norm": 0.142578125,
|
| 1513 |
+
"learning_rate": 0.005,
|
| 1514 |
+
"loss": 3.0664894580841064,
|
| 1515 |
+
"step": 430
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.03792496349543942,
|
| 1519 |
+
"grad_norm": 0.1572265625,
|
| 1520 |
+
"learning_rate": 0.005,
|
| 1521 |
+
"loss": 3.0921502113342285,
|
| 1522 |
+
"step": 432
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.03810054203014052,
|
| 1526 |
+
"grad_norm": 0.1865234375,
|
| 1527 |
+
"learning_rate": 0.005,
|
| 1528 |
+
"loss": 3.082390546798706,
|
| 1529 |
+
"step": 434
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.038276120564841636,
|
| 1533 |
+
"grad_norm": 0.18359375,
|
| 1534 |
+
"learning_rate": 0.005,
|
| 1535 |
+
"loss": 3.10709285736084,
|
| 1536 |
+
"step": 436
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.03845169909954274,
|
| 1540 |
+
"grad_norm": 0.16796875,
|
| 1541 |
+
"learning_rate": 0.005,
|
| 1542 |
+
"loss": 3.0833730697631836,
|
| 1543 |
+
"step": 438
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 0.038627277634243846,
|
| 1547 |
+
"grad_norm": 0.1845703125,
|
| 1548 |
+
"learning_rate": 0.005,
|
| 1549 |
+
"loss": 3.1024534702301025,
|
| 1550 |
+
"step": 440
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.03880285616894496,
|
| 1554 |
+
"grad_norm": 0.15625,
|
| 1555 |
+
"learning_rate": 0.005,
|
| 1556 |
+
"loss": 3.0464351177215576,
|
| 1557 |
+
"step": 442
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.038978434703646064,
|
| 1561 |
+
"grad_norm": 0.14453125,
|
| 1562 |
+
"learning_rate": 0.005,
|
| 1563 |
+
"loss": 3.072038412094116,
|
| 1564 |
+
"step": 444
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 0.03915401323834718,
|
| 1568 |
+
"grad_norm": 0.1318359375,
|
| 1569 |
+
"learning_rate": 0.005,
|
| 1570 |
+
"loss": 3.0556483268737793,
|
| 1571 |
+
"step": 446
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.03932959177304828,
|
| 1575 |
+
"grad_norm": 0.1396484375,
|
| 1576 |
+
"learning_rate": 0.005,
|
| 1577 |
+
"loss": 3.0786564350128174,
|
| 1578 |
+
"step": 448
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.039505170307749395,
|
| 1582 |
+
"grad_norm": 0.1328125,
|
| 1583 |
+
"learning_rate": 0.005,
|
| 1584 |
+
"loss": 3.039013147354126,
|
| 1585 |
+
"step": 450
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.0396807488424505,
|
| 1589 |
+
"grad_norm": 0.12255859375,
|
| 1590 |
+
"learning_rate": 0.005,
|
| 1591 |
+
"loss": 3.100440740585327,
|
| 1592 |
+
"step": 452
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.039856327377151606,
|
| 1596 |
+
"grad_norm": 0.11767578125,
|
| 1597 |
+
"learning_rate": 0.005,
|
| 1598 |
+
"loss": 3.032219171524048,
|
| 1599 |
+
"step": 454
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.04003190591185272,
|
| 1603 |
+
"grad_norm": 0.14453125,
|
| 1604 |
+
"learning_rate": 0.005,
|
| 1605 |
+
"loss": 3.0131278038024902,
|
| 1606 |
+
"step": 456
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 0.04020748444655382,
|
| 1610 |
+
"grad_norm": 0.208984375,
|
| 1611 |
+
"learning_rate": 0.005,
|
| 1612 |
+
"loss": 3.0364959239959717,
|
| 1613 |
+
"step": 458
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 0.040383062981254936,
|
| 1617 |
+
"grad_norm": 0.17578125,
|
| 1618 |
+
"learning_rate": 0.005,
|
| 1619 |
+
"loss": 3.0660479068756104,
|
| 1620 |
+
"step": 460
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 0.04055864151595604,
|
| 1624 |
+
"grad_norm": 0.125,
|
| 1625 |
+
"learning_rate": 0.005,
|
| 1626 |
+
"loss": 3.022470474243164,
|
| 1627 |
+
"step": 462
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 0.040734220050657154,
|
| 1631 |
+
"grad_norm": 0.126953125,
|
| 1632 |
+
"learning_rate": 0.005,
|
| 1633 |
+
"loss": 3.024975538253784,
|
| 1634 |
+
"step": 464
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.04090979858535826,
|
| 1638 |
+
"grad_norm": 0.123046875,
|
| 1639 |
+
"learning_rate": 0.005,
|
| 1640 |
+
"loss": 3.0443012714385986,
|
| 1641 |
+
"step": 466
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.041085377120059365,
|
| 1645 |
+
"grad_norm": 0.14453125,
|
| 1646 |
+
"learning_rate": 0.005,
|
| 1647 |
+
"loss": 3.0197906494140625,
|
| 1648 |
+
"step": 468
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.04126095565476048,
|
| 1652 |
+
"grad_norm": 0.1259765625,
|
| 1653 |
+
"learning_rate": 0.005,
|
| 1654 |
+
"loss": 3.0200116634368896,
|
| 1655 |
+
"step": 470
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.04143653418946158,
|
| 1659 |
+
"grad_norm": 0.1533203125,
|
| 1660 |
+
"learning_rate": 0.005,
|
| 1661 |
+
"loss": 3.0184497833251953,
|
| 1662 |
+
"step": 472
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.041612112724162695,
|
| 1666 |
+
"grad_norm": 0.1552734375,
|
| 1667 |
+
"learning_rate": 0.005,
|
| 1668 |
+
"loss": 3.014349937438965,
|
| 1669 |
+
"step": 474
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.0417876912588638,
|
| 1673 |
+
"grad_norm": 0.1455078125,
|
| 1674 |
+
"learning_rate": 0.005,
|
| 1675 |
+
"loss": 2.9996232986450195,
|
| 1676 |
+
"step": 476
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.04196326979356491,
|
| 1680 |
+
"grad_norm": 0.1318359375,
|
| 1681 |
+
"learning_rate": 0.005,
|
| 1682 |
+
"loss": 2.9938416481018066,
|
| 1683 |
+
"step": 478
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.04213884832826602,
|
| 1687 |
+
"grad_norm": 0.1142578125,
|
| 1688 |
+
"learning_rate": 0.005,
|
| 1689 |
+
"loss": 3.0089657306671143,
|
| 1690 |
+
"step": 480
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.042314426862967124,
|
| 1694 |
+
"grad_norm": 0.1220703125,
|
| 1695 |
+
"learning_rate": 0.005,
|
| 1696 |
+
"loss": 3.0019686222076416,
|
| 1697 |
+
"step": 482
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.042490005397668236,
|
| 1701 |
+
"grad_norm": 0.12451171875,
|
| 1702 |
+
"learning_rate": 0.005,
|
| 1703 |
+
"loss": 3.0172030925750732,
|
| 1704 |
+
"step": 484
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.04266558393236934,
|
| 1708 |
+
"grad_norm": 0.1357421875,
|
| 1709 |
+
"learning_rate": 0.005,
|
| 1710 |
+
"loss": 3.0507822036743164,
|
| 1711 |
+
"step": 486
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.042841162467070454,
|
| 1715 |
+
"grad_norm": 0.1376953125,
|
| 1716 |
+
"learning_rate": 0.005,
|
| 1717 |
+
"loss": 3.0241496562957764,
|
| 1718 |
+
"step": 488
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.04301674100177156,
|
| 1722 |
+
"grad_norm": 0.1259765625,
|
| 1723 |
+
"learning_rate": 0.005,
|
| 1724 |
+
"loss": 3.0176684856414795,
|
| 1725 |
+
"step": 490
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.04319231953647267,
|
| 1729 |
+
"grad_norm": 0.1279296875,
|
| 1730 |
+
"learning_rate": 0.005,
|
| 1731 |
+
"loss": 2.978384256362915,
|
| 1732 |
+
"step": 492
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.04336789807117378,
|
| 1736 |
+
"grad_norm": 0.130859375,
|
| 1737 |
+
"learning_rate": 0.005,
|
| 1738 |
+
"loss": 3.0217716693878174,
|
| 1739 |
+
"step": 494
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.04354347660587488,
|
| 1743 |
+
"grad_norm": 0.13671875,
|
| 1744 |
+
"learning_rate": 0.005,
|
| 1745 |
+
"loss": 3.0396456718444824,
|
| 1746 |
+
"step": 496
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.043719055140575995,
|
| 1750 |
+
"grad_norm": 0.1337890625,
|
| 1751 |
+
"learning_rate": 0.005,
|
| 1752 |
+
"loss": 2.992241621017456,
|
| 1753 |
+
"step": 498
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.0438946336752771,
|
| 1757 |
+
"grad_norm": 0.1416015625,
|
| 1758 |
+
"learning_rate": 0.005,
|
| 1759 |
+
"loss": 3.012587070465088,
|
| 1760 |
+
"step": 500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.04407021220997821,
|
| 1764 |
+
"grad_norm": 0.15234375,
|
| 1765 |
+
"learning_rate": 0.005,
|
| 1766 |
+
"loss": 2.994797945022583,
|
| 1767 |
+
"step": 502
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.04424579074467932,
|
| 1771 |
+
"grad_norm": 0.15625,
|
| 1772 |
+
"learning_rate": 0.005,
|
| 1773 |
+
"loss": 2.9956095218658447,
|
| 1774 |
+
"step": 504
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.044421369279380424,
|
| 1778 |
+
"grad_norm": 0.12353515625,
|
| 1779 |
+
"learning_rate": 0.005,
|
| 1780 |
+
"loss": 2.9908454418182373,
|
| 1781 |
+
"step": 506
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.044596947814081536,
|
| 1785 |
+
"grad_norm": 0.1435546875,
|
| 1786 |
+
"learning_rate": 0.005,
|
| 1787 |
+
"loss": 2.9612441062927246,
|
| 1788 |
+
"step": 508
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.04477252634878264,
|
| 1792 |
+
"grad_norm": 0.154296875,
|
| 1793 |
+
"learning_rate": 0.005,
|
| 1794 |
+
"loss": 2.9770443439483643,
|
| 1795 |
+
"step": 510
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.044948104883483754,
|
| 1799 |
+
"grad_norm": 0.1669921875,
|
| 1800 |
+
"learning_rate": 0.005,
|
| 1801 |
+
"loss": 3.000307083129883,
|
| 1802 |
+
"step": 512
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 0.04512368341818486,
|
| 1806 |
+
"grad_norm": 0.1337890625,
|
| 1807 |
+
"learning_rate": 0.005,
|
| 1808 |
+
"loss": 3.000415086746216,
|
| 1809 |
+
"step": 514
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 0.04529926195288597,
|
| 1813 |
+
"grad_norm": 0.1416015625,
|
| 1814 |
+
"learning_rate": 0.005,
|
| 1815 |
+
"loss": 2.9415054321289062,
|
| 1816 |
+
"step": 516
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 0.04547484048758708,
|
| 1820 |
+
"grad_norm": 0.1298828125,
|
| 1821 |
+
"learning_rate": 0.005,
|
| 1822 |
+
"loss": 2.9683773517608643,
|
| 1823 |
+
"step": 518
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 0.04565041902228818,
|
| 1827 |
+
"grad_norm": 0.12890625,
|
| 1828 |
+
"learning_rate": 0.005,
|
| 1829 |
+
"loss": 2.9754841327667236,
|
| 1830 |
+
"step": 520
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 0.045825997556989295,
|
| 1834 |
+
"grad_norm": 0.1298828125,
|
| 1835 |
+
"learning_rate": 0.005,
|
| 1836 |
+
"loss": 2.9868385791778564,
|
| 1837 |
+
"step": 522
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 0.0460015760916904,
|
| 1841 |
+
"grad_norm": 0.126953125,
|
| 1842 |
+
"learning_rate": 0.005,
|
| 1843 |
+
"loss": 2.96690034866333,
|
| 1844 |
+
"step": 524
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 0.04617715462639151,
|
| 1848 |
+
"grad_norm": 0.1376953125,
|
| 1849 |
+
"learning_rate": 0.005,
|
| 1850 |
+
"loss": 2.970330238342285,
|
| 1851 |
+
"step": 526
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 0.04635273316109262,
|
| 1855 |
+
"grad_norm": 0.1845703125,
|
| 1856 |
+
"learning_rate": 0.005,
|
| 1857 |
+
"loss": 2.960998058319092,
|
| 1858 |
+
"step": 528
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 0.04652831169579373,
|
| 1862 |
+
"grad_norm": 0.1943359375,
|
| 1863 |
+
"learning_rate": 0.005,
|
| 1864 |
+
"loss": 2.980581045150757,
|
| 1865 |
+
"step": 530
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 0.046703890230494836,
|
| 1869 |
+
"grad_norm": 0.146484375,
|
| 1870 |
+
"learning_rate": 0.005,
|
| 1871 |
+
"loss": 2.950021743774414,
|
| 1872 |
+
"step": 532
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 0.04687946876519594,
|
| 1876 |
+
"grad_norm": 0.142578125,
|
| 1877 |
+
"learning_rate": 0.005,
|
| 1878 |
+
"loss": 2.9638140201568604,
|
| 1879 |
+
"step": 534
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 0.047055047299897054,
|
| 1883 |
+
"grad_norm": 0.12158203125,
|
| 1884 |
+
"learning_rate": 0.005,
|
| 1885 |
+
"loss": 2.9714972972869873,
|
| 1886 |
+
"step": 536
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 0.04723062583459816,
|
| 1890 |
+
"grad_norm": 0.10205078125,
|
| 1891 |
+
"learning_rate": 0.005,
|
| 1892 |
+
"loss": 2.9336445331573486,
|
| 1893 |
+
"step": 538
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 0.04740620436929927,
|
| 1897 |
+
"grad_norm": 0.09814453125,
|
| 1898 |
+
"learning_rate": 0.005,
|
| 1899 |
+
"loss": 2.9828338623046875,
|
| 1900 |
+
"step": 540
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 0.04758178290400038,
|
| 1904 |
+
"grad_norm": 0.1142578125,
|
| 1905 |
+
"learning_rate": 0.005,
|
| 1906 |
+
"loss": 2.9587650299072266,
|
| 1907 |
+
"step": 542
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 0.04775736143870149,
|
| 1911 |
+
"grad_norm": 0.111328125,
|
| 1912 |
+
"learning_rate": 0.005,
|
| 1913 |
+
"loss": 2.977285385131836,
|
| 1914 |
+
"step": 544
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 0.047932939973402595,
|
| 1918 |
+
"grad_norm": 0.119140625,
|
| 1919 |
+
"learning_rate": 0.005,
|
| 1920 |
+
"loss": 2.9418492317199707,
|
| 1921 |
+
"step": 546
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 0.0481085185081037,
|
| 1925 |
+
"grad_norm": 0.14453125,
|
| 1926 |
+
"learning_rate": 0.005,
|
| 1927 |
+
"loss": 2.9256904125213623,
|
| 1928 |
+
"step": 548
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 0.04828409704280481,
|
| 1932 |
+
"grad_norm": 0.1376953125,
|
| 1933 |
+
"learning_rate": 0.005,
|
| 1934 |
+
"loss": 2.945709705352783,
|
| 1935 |
+
"step": 550
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 0.04845967557750592,
|
| 1939 |
+
"grad_norm": 0.12890625,
|
| 1940 |
+
"learning_rate": 0.005,
|
| 1941 |
+
"loss": 2.9624128341674805,
|
| 1942 |
+
"step": 552
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 0.04863525411220703,
|
| 1946 |
+
"grad_norm": 0.11962890625,
|
| 1947 |
+
"learning_rate": 0.005,
|
| 1948 |
+
"loss": 2.9454598426818848,
|
| 1949 |
+
"step": 554
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 0.04881083264690814,
|
| 1953 |
+
"grad_norm": 0.1259765625,
|
| 1954 |
+
"learning_rate": 0.005,
|
| 1955 |
+
"loss": 2.925313949584961,
|
| 1956 |
+
"step": 556
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 0.04898641118160925,
|
| 1960 |
+
"grad_norm": 0.11083984375,
|
| 1961 |
+
"learning_rate": 0.005,
|
| 1962 |
+
"loss": 2.933090925216675,
|
| 1963 |
+
"step": 558
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 0.049161989716310354,
|
| 1967 |
+
"grad_norm": 0.1005859375,
|
| 1968 |
+
"learning_rate": 0.005,
|
| 1969 |
+
"loss": 2.94028639793396,
|
| 1970 |
+
"step": 560
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 0.04933756825101146,
|
| 1974 |
+
"grad_norm": 0.11767578125,
|
| 1975 |
+
"learning_rate": 0.005,
|
| 1976 |
+
"loss": 2.9429543018341064,
|
| 1977 |
+
"step": 562
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 0.04951314678571257,
|
| 1981 |
+
"grad_norm": 0.10791015625,
|
| 1982 |
+
"learning_rate": 0.005,
|
| 1983 |
+
"loss": 2.919218063354492,
|
| 1984 |
+
"step": 564
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 0.04968872532041368,
|
| 1988 |
+
"grad_norm": 0.11669921875,
|
| 1989 |
+
"learning_rate": 0.005,
|
| 1990 |
+
"loss": 2.9045145511627197,
|
| 1991 |
+
"step": 566
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 0.04986430385511479,
|
| 1995 |
+
"grad_norm": 0.1279296875,
|
| 1996 |
+
"learning_rate": 0.005,
|
| 1997 |
+
"loss": 2.941321849822998,
|
| 1998 |
+
"step": 568
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 0.050039882389815896,
|
| 2002 |
+
"grad_norm": 0.125,
|
| 2003 |
+
"learning_rate": 0.005,
|
| 2004 |
+
"loss": 2.9178407192230225,
|
| 2005 |
+
"step": 570
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 0.05021546092451701,
|
| 2009 |
+
"grad_norm": 0.154296875,
|
| 2010 |
+
"learning_rate": 0.005,
|
| 2011 |
+
"loss": 2.8982131481170654,
|
| 2012 |
+
"step": 572
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 0.050391039459218114,
|
| 2016 |
+
"grad_norm": 0.115234375,
|
| 2017 |
+
"learning_rate": 0.005,
|
| 2018 |
+
"loss": 2.926523447036743,
|
| 2019 |
+
"step": 574
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.05056661799391922,
|
| 2023 |
+
"grad_norm": 0.11083984375,
|
| 2024 |
+
"learning_rate": 0.005,
|
| 2025 |
+
"loss": 2.936190366744995,
|
| 2026 |
+
"step": 576
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 0.05074219652862033,
|
| 2030 |
+
"grad_norm": 0.1279296875,
|
| 2031 |
+
"learning_rate": 0.005,
|
| 2032 |
+
"loss": 2.920558214187622,
|
| 2033 |
+
"step": 578
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 0.05091777506332144,
|
| 2037 |
+
"grad_norm": 0.1318359375,
|
| 2038 |
+
"learning_rate": 0.005,
|
| 2039 |
+
"loss": 2.8939120769500732,
|
| 2040 |
+
"step": 580
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 0.05109335359802255,
|
| 2044 |
+
"grad_norm": 0.154296875,
|
| 2045 |
+
"learning_rate": 0.005,
|
| 2046 |
+
"loss": 2.9341375827789307,
|
| 2047 |
+
"step": 582
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 0.051268932132723655,
|
| 2051 |
+
"grad_norm": 0.1416015625,
|
| 2052 |
+
"learning_rate": 0.005,
|
| 2053 |
+
"loss": 2.9142417907714844,
|
| 2054 |
+
"step": 584
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 0.05144451066742476,
|
| 2058 |
+
"grad_norm": 0.11279296875,
|
| 2059 |
+
"learning_rate": 0.005,
|
| 2060 |
+
"loss": 2.923351526260376,
|
| 2061 |
+
"step": 586
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 0.05162008920212587,
|
| 2065 |
+
"grad_norm": 0.11474609375,
|
| 2066 |
+
"learning_rate": 0.005,
|
| 2067 |
+
"loss": 2.923654794692993,
|
| 2068 |
+
"step": 588
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 0.05179566773682698,
|
| 2072 |
+
"grad_norm": 0.12451171875,
|
| 2073 |
+
"learning_rate": 0.005,
|
| 2074 |
+
"loss": 2.9181389808654785,
|
| 2075 |
+
"step": 590
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 0.05197124627152809,
|
| 2079 |
+
"grad_norm": 0.1044921875,
|
| 2080 |
+
"learning_rate": 0.005,
|
| 2081 |
+
"loss": 2.8995134830474854,
|
| 2082 |
+
"step": 592
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 0.052146824806229196,
|
| 2086 |
+
"grad_norm": 0.1103515625,
|
| 2087 |
+
"learning_rate": 0.005,
|
| 2088 |
+
"loss": 2.914994716644287,
|
| 2089 |
+
"step": 594
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 0.05232240334093031,
|
| 2093 |
+
"grad_norm": 0.1220703125,
|
| 2094 |
+
"learning_rate": 0.005,
|
| 2095 |
+
"loss": 2.9268240928649902,
|
| 2096 |
+
"step": 596
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.052497981875631414,
|
| 2100 |
+
"grad_norm": 0.11083984375,
|
| 2101 |
+
"learning_rate": 0.005,
|
| 2102 |
+
"loss": 2.901998281478882,
|
| 2103 |
+
"step": 598
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 0.05267356041033252,
|
| 2107 |
+
"grad_norm": 0.1396484375,
|
| 2108 |
+
"learning_rate": 0.005,
|
| 2109 |
+
"loss": 2.892770528793335,
|
| 2110 |
+
"step": 600
|
| 2111 |
}
|
| 2112 |
],
|
| 2113 |
"logging_steps": 2,
|
|
|
|
| 2127 |
"attributes": {}
|
| 2128 |
}
|
| 2129 |
},
|
| 2130 |
+
"total_flos": 1.0142157599816058e+18,
|
| 2131 |
"train_batch_size": 2,
|
| 2132 |
"trial_name": null,
|
| 2133 |
"trial_params": null
|