Training in progress, step 900, 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,
|
|
@@ -2108,6 +2108,1056 @@
|
|
| 2108 |
"learning_rate": 0.005,
|
| 2109 |
"loss": 2.892770528793335,
|
| 2110 |
"step": 600
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2111 |
}
|
| 2112 |
],
|
| 2113 |
"logging_steps": 2,
|
|
@@ -2127,7 +3177,7 @@
|
|
| 2127 |
"attributes": {}
|
| 2128 |
}
|
| 2129 |
},
|
| 2130 |
-
"total_flos": 1.
|
| 2131 |
"train_batch_size": 2,
|
| 2132 |
"trial_name": null,
|
| 2133 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.07901034061549879,
|
| 6 |
"eval_steps": 957,
|
| 7 |
+
"global_step": 900,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 2108 |
"learning_rate": 0.005,
|
| 2109 |
"loss": 2.892770528793335,
|
| 2110 |
"step": 600
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 0.05284913894503363,
|
| 2114 |
+
"grad_norm": 0.1474609375,
|
| 2115 |
+
"learning_rate": 0.005,
|
| 2116 |
+
"loss": 2.9107680320739746,
|
| 2117 |
+
"step": 602
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 0.05302471747973474,
|
| 2121 |
+
"grad_norm": 0.1728515625,
|
| 2122 |
+
"learning_rate": 0.005,
|
| 2123 |
+
"loss": 2.940669298171997,
|
| 2124 |
+
"step": 604
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 0.05320029601443585,
|
| 2128 |
+
"grad_norm": 0.1337890625,
|
| 2129 |
+
"learning_rate": 0.005,
|
| 2130 |
+
"loss": 2.8937621116638184,
|
| 2131 |
+
"step": 606
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 0.053375874549136955,
|
| 2135 |
+
"grad_norm": 0.1171875,
|
| 2136 |
+
"learning_rate": 0.005,
|
| 2137 |
+
"loss": 2.9157919883728027,
|
| 2138 |
+
"step": 608
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 0.05355145308383807,
|
| 2142 |
+
"grad_norm": 0.1376953125,
|
| 2143 |
+
"learning_rate": 0.005,
|
| 2144 |
+
"loss": 2.9084694385528564,
|
| 2145 |
+
"step": 610
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 0.05372703161853917,
|
| 2149 |
+
"grad_norm": 0.1259765625,
|
| 2150 |
+
"learning_rate": 0.005,
|
| 2151 |
+
"loss": 2.89589786529541,
|
| 2152 |
+
"step": 612
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 0.05390261015324028,
|
| 2156 |
+
"grad_norm": 0.10546875,
|
| 2157 |
+
"learning_rate": 0.005,
|
| 2158 |
+
"loss": 2.866093397140503,
|
| 2159 |
+
"step": 614
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 0.05407818868794139,
|
| 2163 |
+
"grad_norm": 0.1123046875,
|
| 2164 |
+
"learning_rate": 0.005,
|
| 2165 |
+
"loss": 2.8685555458068848,
|
| 2166 |
+
"step": 616
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 0.054253767222642496,
|
| 2170 |
+
"grad_norm": 0.1337890625,
|
| 2171 |
+
"learning_rate": 0.005,
|
| 2172 |
+
"loss": 2.934528112411499,
|
| 2173 |
+
"step": 618
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 0.05442934575734361,
|
| 2177 |
+
"grad_norm": 0.13671875,
|
| 2178 |
+
"learning_rate": 0.005,
|
| 2179 |
+
"loss": 2.915015459060669,
|
| 2180 |
+
"step": 620
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 0.054604924292044714,
|
| 2184 |
+
"grad_norm": 0.130859375,
|
| 2185 |
+
"learning_rate": 0.005,
|
| 2186 |
+
"loss": 2.930428981781006,
|
| 2187 |
+
"step": 622
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 0.054780502826745826,
|
| 2191 |
+
"grad_norm": 0.1455078125,
|
| 2192 |
+
"learning_rate": 0.005,
|
| 2193 |
+
"loss": 2.885199546813965,
|
| 2194 |
+
"step": 624
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 0.05495608136144693,
|
| 2198 |
+
"grad_norm": 0.14453125,
|
| 2199 |
+
"learning_rate": 0.005,
|
| 2200 |
+
"loss": 2.9132332801818848,
|
| 2201 |
+
"step": 626
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.05513165989614804,
|
| 2205 |
+
"grad_norm": 0.1220703125,
|
| 2206 |
+
"learning_rate": 0.005,
|
| 2207 |
+
"loss": 2.8978776931762695,
|
| 2208 |
+
"step": 628
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.05530723843084915,
|
| 2212 |
+
"grad_norm": 0.140625,
|
| 2213 |
+
"learning_rate": 0.005,
|
| 2214 |
+
"loss": 2.8665482997894287,
|
| 2215 |
+
"step": 630
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.055482816965550255,
|
| 2219 |
+
"grad_norm": 0.1279296875,
|
| 2220 |
+
"learning_rate": 0.005,
|
| 2221 |
+
"loss": 2.8479433059692383,
|
| 2222 |
+
"step": 632
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.05565839550025137,
|
| 2226 |
+
"grad_norm": 0.1591796875,
|
| 2227 |
+
"learning_rate": 0.005,
|
| 2228 |
+
"loss": 2.902829647064209,
|
| 2229 |
+
"step": 634
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.05583397403495247,
|
| 2233 |
+
"grad_norm": 0.1376953125,
|
| 2234 |
+
"learning_rate": 0.005,
|
| 2235 |
+
"loss": 2.8622169494628906,
|
| 2236 |
+
"step": 636
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 0.056009552569653585,
|
| 2240 |
+
"grad_norm": 0.126953125,
|
| 2241 |
+
"learning_rate": 0.005,
|
| 2242 |
+
"loss": 2.8923892974853516,
|
| 2243 |
+
"step": 638
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.05618513110435469,
|
| 2247 |
+
"grad_norm": 0.12451171875,
|
| 2248 |
+
"learning_rate": 0.005,
|
| 2249 |
+
"loss": 2.8786332607269287,
|
| 2250 |
+
"step": 640
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 0.056360709639055796,
|
| 2254 |
+
"grad_norm": 0.1064453125,
|
| 2255 |
+
"learning_rate": 0.005,
|
| 2256 |
+
"loss": 2.8552439212799072,
|
| 2257 |
+
"step": 642
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.05653628817375691,
|
| 2261 |
+
"grad_norm": 0.10595703125,
|
| 2262 |
+
"learning_rate": 0.005,
|
| 2263 |
+
"loss": 2.848996877670288,
|
| 2264 |
+
"step": 644
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 0.056711866708458014,
|
| 2268 |
+
"grad_norm": 0.11767578125,
|
| 2269 |
+
"learning_rate": 0.005,
|
| 2270 |
+
"loss": 2.9026734828948975,
|
| 2271 |
+
"step": 646
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.05688744524315913,
|
| 2275 |
+
"grad_norm": 0.12890625,
|
| 2276 |
+
"learning_rate": 0.005,
|
| 2277 |
+
"loss": 2.907194137573242,
|
| 2278 |
+
"step": 648
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 0.05706302377786023,
|
| 2282 |
+
"grad_norm": 0.130859375,
|
| 2283 |
+
"learning_rate": 0.005,
|
| 2284 |
+
"loss": 2.885383129119873,
|
| 2285 |
+
"step": 650
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 0.057238602312561344,
|
| 2289 |
+
"grad_norm": 0.12353515625,
|
| 2290 |
+
"learning_rate": 0.005,
|
| 2291 |
+
"loss": 2.8656253814697266,
|
| 2292 |
+
"step": 652
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 0.05741418084726245,
|
| 2296 |
+
"grad_norm": 0.1259765625,
|
| 2297 |
+
"learning_rate": 0.005,
|
| 2298 |
+
"loss": 2.873291254043579,
|
| 2299 |
+
"step": 654
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.057589759381963555,
|
| 2303 |
+
"grad_norm": 0.10546875,
|
| 2304 |
+
"learning_rate": 0.005,
|
| 2305 |
+
"loss": 2.8469748497009277,
|
| 2306 |
+
"step": 656
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.05776533791666467,
|
| 2310 |
+
"grad_norm": 0.1162109375,
|
| 2311 |
+
"learning_rate": 0.005,
|
| 2312 |
+
"loss": 2.8983521461486816,
|
| 2313 |
+
"step": 658
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.05794091645136577,
|
| 2317 |
+
"grad_norm": 0.099609375,
|
| 2318 |
+
"learning_rate": 0.005,
|
| 2319 |
+
"loss": 2.8409366607666016,
|
| 2320 |
+
"step": 660
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 0.058116494986066886,
|
| 2324 |
+
"grad_norm": 0.115234375,
|
| 2325 |
+
"learning_rate": 0.005,
|
| 2326 |
+
"loss": 2.9008703231811523,
|
| 2327 |
+
"step": 662
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 0.05829207352076799,
|
| 2331 |
+
"grad_norm": 0.1328125,
|
| 2332 |
+
"learning_rate": 0.005,
|
| 2333 |
+
"loss": 2.853753089904785,
|
| 2334 |
+
"step": 664
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 0.058467652055469096,
|
| 2338 |
+
"grad_norm": 0.11376953125,
|
| 2339 |
+
"learning_rate": 0.005,
|
| 2340 |
+
"loss": 2.8707268238067627,
|
| 2341 |
+
"step": 666
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.05864323059017021,
|
| 2345 |
+
"grad_norm": 0.099609375,
|
| 2346 |
+
"learning_rate": 0.005,
|
| 2347 |
+
"loss": 2.889103889465332,
|
| 2348 |
+
"step": 668
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 0.058818809124871314,
|
| 2352 |
+
"grad_norm": 0.09716796875,
|
| 2353 |
+
"learning_rate": 0.005,
|
| 2354 |
+
"loss": 2.8374733924865723,
|
| 2355 |
+
"step": 670
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 0.05899438765957243,
|
| 2359 |
+
"grad_norm": 0.109375,
|
| 2360 |
+
"learning_rate": 0.005,
|
| 2361 |
+
"loss": 2.8331263065338135,
|
| 2362 |
+
"step": 672
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 0.05916996619427353,
|
| 2366 |
+
"grad_norm": 0.11279296875,
|
| 2367 |
+
"learning_rate": 0.005,
|
| 2368 |
+
"loss": 2.880403757095337,
|
| 2369 |
+
"step": 674
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 0.059345544728974645,
|
| 2373 |
+
"grad_norm": 0.1025390625,
|
| 2374 |
+
"learning_rate": 0.005,
|
| 2375 |
+
"loss": 2.8748178482055664,
|
| 2376 |
+
"step": 676
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 0.05952112326367575,
|
| 2380 |
+
"grad_norm": 0.09619140625,
|
| 2381 |
+
"learning_rate": 0.005,
|
| 2382 |
+
"loss": 2.8325412273406982,
|
| 2383 |
+
"step": 678
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 0.059696701798376856,
|
| 2387 |
+
"grad_norm": 0.0966796875,
|
| 2388 |
+
"learning_rate": 0.005,
|
| 2389 |
+
"loss": 2.826225519180298,
|
| 2390 |
+
"step": 680
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 0.05987228033307797,
|
| 2394 |
+
"grad_norm": 0.11767578125,
|
| 2395 |
+
"learning_rate": 0.005,
|
| 2396 |
+
"loss": 2.858710527420044,
|
| 2397 |
+
"step": 682
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 0.06004785886777907,
|
| 2401 |
+
"grad_norm": 0.1083984375,
|
| 2402 |
+
"learning_rate": 0.005,
|
| 2403 |
+
"loss": 2.8498644828796387,
|
| 2404 |
+
"step": 684
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 0.060223437402480186,
|
| 2408 |
+
"grad_norm": 0.099609375,
|
| 2409 |
+
"learning_rate": 0.005,
|
| 2410 |
+
"loss": 2.8439319133758545,
|
| 2411 |
+
"step": 686
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 0.06039901593718129,
|
| 2415 |
+
"grad_norm": 0.0986328125,
|
| 2416 |
+
"learning_rate": 0.005,
|
| 2417 |
+
"loss": 2.816505193710327,
|
| 2418 |
+
"step": 688
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 0.060574594471882404,
|
| 2422 |
+
"grad_norm": 0.11181640625,
|
| 2423 |
+
"learning_rate": 0.005,
|
| 2424 |
+
"loss": 2.8512182235717773,
|
| 2425 |
+
"step": 690
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 0.06075017300658351,
|
| 2429 |
+
"grad_norm": 0.09619140625,
|
| 2430 |
+
"learning_rate": 0.005,
|
| 2431 |
+
"loss": 2.8385231494903564,
|
| 2432 |
+
"step": 692
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 0.060925751541284615,
|
| 2436 |
+
"grad_norm": 0.09619140625,
|
| 2437 |
+
"learning_rate": 0.005,
|
| 2438 |
+
"loss": 2.8305373191833496,
|
| 2439 |
+
"step": 694
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 0.06110133007598573,
|
| 2443 |
+
"grad_norm": 0.1123046875,
|
| 2444 |
+
"learning_rate": 0.005,
|
| 2445 |
+
"loss": 2.817195415496826,
|
| 2446 |
+
"step": 696
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 0.06127690861068683,
|
| 2450 |
+
"grad_norm": 0.10888671875,
|
| 2451 |
+
"learning_rate": 0.005,
|
| 2452 |
+
"loss": 2.869903087615967,
|
| 2453 |
+
"step": 698
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 0.061452487145387945,
|
| 2457 |
+
"grad_norm": 0.14453125,
|
| 2458 |
+
"learning_rate": 0.005,
|
| 2459 |
+
"loss": 2.856109619140625,
|
| 2460 |
+
"step": 700
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 0.06162806568008905,
|
| 2464 |
+
"grad_norm": 0.1025390625,
|
| 2465 |
+
"learning_rate": 0.005,
|
| 2466 |
+
"loss": 2.8376383781433105,
|
| 2467 |
+
"step": 702
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 0.06180364421479016,
|
| 2471 |
+
"grad_norm": 0.10595703125,
|
| 2472 |
+
"learning_rate": 0.005,
|
| 2473 |
+
"loss": 2.7832024097442627,
|
| 2474 |
+
"step": 704
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 0.06197922274949127,
|
| 2478 |
+
"grad_norm": 0.1220703125,
|
| 2479 |
+
"learning_rate": 0.005,
|
| 2480 |
+
"loss": 2.8388211727142334,
|
| 2481 |
+
"step": 706
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 0.062154801284192374,
|
| 2485 |
+
"grad_norm": 0.107421875,
|
| 2486 |
+
"learning_rate": 0.005,
|
| 2487 |
+
"loss": 2.8409483432769775,
|
| 2488 |
+
"step": 708
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 0.062330379818893486,
|
| 2492 |
+
"grad_norm": 0.09912109375,
|
| 2493 |
+
"learning_rate": 0.005,
|
| 2494 |
+
"loss": 2.8188676834106445,
|
| 2495 |
+
"step": 710
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 0.0625059583535946,
|
| 2499 |
+
"grad_norm": 0.103515625,
|
| 2500 |
+
"learning_rate": 0.005,
|
| 2501 |
+
"loss": 2.8468964099884033,
|
| 2502 |
+
"step": 712
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 0.0626815368882957,
|
| 2506 |
+
"grad_norm": 0.10302734375,
|
| 2507 |
+
"learning_rate": 0.005,
|
| 2508 |
+
"loss": 2.798382520675659,
|
| 2509 |
+
"step": 714
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 0.06285711542299681,
|
| 2513 |
+
"grad_norm": 0.130859375,
|
| 2514 |
+
"learning_rate": 0.005,
|
| 2515 |
+
"loss": 2.8014094829559326,
|
| 2516 |
+
"step": 716
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 0.06303269395769792,
|
| 2520 |
+
"grad_norm": 0.126953125,
|
| 2521 |
+
"learning_rate": 0.005,
|
| 2522 |
+
"loss": 2.8606605529785156,
|
| 2523 |
+
"step": 718
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 0.06320827249239902,
|
| 2527 |
+
"grad_norm": 0.1201171875,
|
| 2528 |
+
"learning_rate": 0.005,
|
| 2529 |
+
"loss": 2.8185365200042725,
|
| 2530 |
+
"step": 720
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 0.06338385102710013,
|
| 2534 |
+
"grad_norm": 0.12109375,
|
| 2535 |
+
"learning_rate": 0.005,
|
| 2536 |
+
"loss": 2.79809832572937,
|
| 2537 |
+
"step": 722
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 0.06355942956180125,
|
| 2541 |
+
"grad_norm": 0.12890625,
|
| 2542 |
+
"learning_rate": 0.005,
|
| 2543 |
+
"loss": 2.829028606414795,
|
| 2544 |
+
"step": 724
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 0.06373500809650236,
|
| 2548 |
+
"grad_norm": 0.09716796875,
|
| 2549 |
+
"learning_rate": 0.005,
|
| 2550 |
+
"loss": 2.775477886199951,
|
| 2551 |
+
"step": 726
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 0.06391058663120346,
|
| 2555 |
+
"grad_norm": 0.09521484375,
|
| 2556 |
+
"learning_rate": 0.005,
|
| 2557 |
+
"loss": 2.806027412414551,
|
| 2558 |
+
"step": 728
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 0.06408616516590457,
|
| 2562 |
+
"grad_norm": 0.10498046875,
|
| 2563 |
+
"learning_rate": 0.005,
|
| 2564 |
+
"loss": 2.8189749717712402,
|
| 2565 |
+
"step": 730
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 0.06426174370060568,
|
| 2569 |
+
"grad_norm": 0.103515625,
|
| 2570 |
+
"learning_rate": 0.005,
|
| 2571 |
+
"loss": 2.809892416000366,
|
| 2572 |
+
"step": 732
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 0.06443732223530678,
|
| 2576 |
+
"grad_norm": 0.10009765625,
|
| 2577 |
+
"learning_rate": 0.005,
|
| 2578 |
+
"loss": 2.8056528568267822,
|
| 2579 |
+
"step": 734
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 0.06461290077000789,
|
| 2583 |
+
"grad_norm": 0.11328125,
|
| 2584 |
+
"learning_rate": 0.005,
|
| 2585 |
+
"loss": 2.851624011993408,
|
| 2586 |
+
"step": 736
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 0.064788479304709,
|
| 2590 |
+
"grad_norm": 0.09375,
|
| 2591 |
+
"learning_rate": 0.005,
|
| 2592 |
+
"loss": 2.839448928833008,
|
| 2593 |
+
"step": 738
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 0.06496405783941012,
|
| 2597 |
+
"grad_norm": 0.09326171875,
|
| 2598 |
+
"learning_rate": 0.005,
|
| 2599 |
+
"loss": 2.832307815551758,
|
| 2600 |
+
"step": 740
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 0.06513963637411121,
|
| 2604 |
+
"grad_norm": 0.10498046875,
|
| 2605 |
+
"learning_rate": 0.005,
|
| 2606 |
+
"loss": 2.816222667694092,
|
| 2607 |
+
"step": 742
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 0.06531521490881233,
|
| 2611 |
+
"grad_norm": 0.11328125,
|
| 2612 |
+
"learning_rate": 0.005,
|
| 2613 |
+
"loss": 2.814714193344116,
|
| 2614 |
+
"step": 744
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 0.06549079344351344,
|
| 2618 |
+
"grad_norm": 0.10302734375,
|
| 2619 |
+
"learning_rate": 0.005,
|
| 2620 |
+
"loss": 2.8213393688201904,
|
| 2621 |
+
"step": 746
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 0.06566637197821454,
|
| 2625 |
+
"grad_norm": 0.12451171875,
|
| 2626 |
+
"learning_rate": 0.005,
|
| 2627 |
+
"loss": 2.838747024536133,
|
| 2628 |
+
"step": 748
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 0.06584195051291565,
|
| 2632 |
+
"grad_norm": 0.11669921875,
|
| 2633 |
+
"learning_rate": 0.005,
|
| 2634 |
+
"loss": 2.8111016750335693,
|
| 2635 |
+
"step": 750
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 0.06601752904761676,
|
| 2639 |
+
"grad_norm": 0.1357421875,
|
| 2640 |
+
"learning_rate": 0.005,
|
| 2641 |
+
"loss": 2.8226990699768066,
|
| 2642 |
+
"step": 752
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 0.06619310758231788,
|
| 2646 |
+
"grad_norm": 0.12060546875,
|
| 2647 |
+
"learning_rate": 0.005,
|
| 2648 |
+
"loss": 2.8317198753356934,
|
| 2649 |
+
"step": 754
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 0.06636868611701897,
|
| 2653 |
+
"grad_norm": 0.1015625,
|
| 2654 |
+
"learning_rate": 0.005,
|
| 2655 |
+
"loss": 2.795905113220215,
|
| 2656 |
+
"step": 756
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 0.06654426465172009,
|
| 2660 |
+
"grad_norm": 0.10205078125,
|
| 2661 |
+
"learning_rate": 0.005,
|
| 2662 |
+
"loss": 2.8132882118225098,
|
| 2663 |
+
"step": 758
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 0.0667198431864212,
|
| 2667 |
+
"grad_norm": 0.10205078125,
|
| 2668 |
+
"learning_rate": 0.005,
|
| 2669 |
+
"loss": 2.8165059089660645,
|
| 2670 |
+
"step": 760
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 0.0668954217211223,
|
| 2674 |
+
"grad_norm": 0.09228515625,
|
| 2675 |
+
"learning_rate": 0.005,
|
| 2676 |
+
"loss": 2.8166255950927734,
|
| 2677 |
+
"step": 762
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 0.06707100025582341,
|
| 2681 |
+
"grad_norm": 0.091796875,
|
| 2682 |
+
"learning_rate": 0.005,
|
| 2683 |
+
"loss": 2.822350025177002,
|
| 2684 |
+
"step": 764
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 0.06724657879052452,
|
| 2688 |
+
"grad_norm": 0.095703125,
|
| 2689 |
+
"learning_rate": 0.005,
|
| 2690 |
+
"loss": 2.8084328174591064,
|
| 2691 |
+
"step": 766
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 0.06742215732522563,
|
| 2695 |
+
"grad_norm": 0.11279296875,
|
| 2696 |
+
"learning_rate": 0.005,
|
| 2697 |
+
"loss": 2.803453207015991,
|
| 2698 |
+
"step": 768
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 0.06759773585992673,
|
| 2702 |
+
"grad_norm": 0.11376953125,
|
| 2703 |
+
"learning_rate": 0.005,
|
| 2704 |
+
"loss": 2.7944207191467285,
|
| 2705 |
+
"step": 770
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 0.06777331439462785,
|
| 2709 |
+
"grad_norm": 0.1201171875,
|
| 2710 |
+
"learning_rate": 0.005,
|
| 2711 |
+
"loss": 2.778062582015991,
|
| 2712 |
+
"step": 772
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 0.06794889292932896,
|
| 2716 |
+
"grad_norm": 0.1025390625,
|
| 2717 |
+
"learning_rate": 0.005,
|
| 2718 |
+
"loss": 2.7989413738250732,
|
| 2719 |
+
"step": 774
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 0.06812447146403006,
|
| 2723 |
+
"grad_norm": 0.12353515625,
|
| 2724 |
+
"learning_rate": 0.005,
|
| 2725 |
+
"loss": 2.784641742706299,
|
| 2726 |
+
"step": 776
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 0.06830004999873117,
|
| 2730 |
+
"grad_norm": 0.1103515625,
|
| 2731 |
+
"learning_rate": 0.005,
|
| 2732 |
+
"loss": 2.8191514015197754,
|
| 2733 |
+
"step": 778
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 0.06847562853343228,
|
| 2737 |
+
"grad_norm": 0.10791015625,
|
| 2738 |
+
"learning_rate": 0.005,
|
| 2739 |
+
"loss": 2.793182373046875,
|
| 2740 |
+
"step": 780
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 0.0686512070681334,
|
| 2744 |
+
"grad_norm": 0.1201171875,
|
| 2745 |
+
"learning_rate": 0.005,
|
| 2746 |
+
"loss": 2.7880682945251465,
|
| 2747 |
+
"step": 782
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 0.06882678560283449,
|
| 2751 |
+
"grad_norm": 0.1396484375,
|
| 2752 |
+
"learning_rate": 0.005,
|
| 2753 |
+
"loss": 2.8155131340026855,
|
| 2754 |
+
"step": 784
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.0690023641375356,
|
| 2758 |
+
"grad_norm": 0.1083984375,
|
| 2759 |
+
"learning_rate": 0.005,
|
| 2760 |
+
"loss": 2.8056650161743164,
|
| 2761 |
+
"step": 786
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.06917794267223672,
|
| 2765 |
+
"grad_norm": 0.1103515625,
|
| 2766 |
+
"learning_rate": 0.005,
|
| 2767 |
+
"loss": 2.794665575027466,
|
| 2768 |
+
"step": 788
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 0.06935352120693782,
|
| 2772 |
+
"grad_norm": 0.11474609375,
|
| 2773 |
+
"learning_rate": 0.005,
|
| 2774 |
+
"loss": 2.8043224811553955,
|
| 2775 |
+
"step": 790
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 0.06952909974163893,
|
| 2779 |
+
"grad_norm": 0.09716796875,
|
| 2780 |
+
"learning_rate": 0.005,
|
| 2781 |
+
"loss": 2.7930874824523926,
|
| 2782 |
+
"step": 792
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 0.06970467827634004,
|
| 2786 |
+
"grad_norm": 0.095703125,
|
| 2787 |
+
"learning_rate": 0.005,
|
| 2788 |
+
"loss": 2.7990355491638184,
|
| 2789 |
+
"step": 794
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 0.06988025681104115,
|
| 2793 |
+
"grad_norm": 0.0947265625,
|
| 2794 |
+
"learning_rate": 0.005,
|
| 2795 |
+
"loss": 2.7672834396362305,
|
| 2796 |
+
"step": 796
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.07005583534574225,
|
| 2800 |
+
"grad_norm": 0.09716796875,
|
| 2801 |
+
"learning_rate": 0.005,
|
| 2802 |
+
"loss": 2.8196029663085938,
|
| 2803 |
+
"step": 798
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 0.07023141388044336,
|
| 2807 |
+
"grad_norm": 0.08837890625,
|
| 2808 |
+
"learning_rate": 0.005,
|
| 2809 |
+
"loss": 2.78668475151062,
|
| 2810 |
+
"step": 800
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 0.07040699241514448,
|
| 2814 |
+
"grad_norm": 0.099609375,
|
| 2815 |
+
"learning_rate": 0.005,
|
| 2816 |
+
"loss": 2.755502462387085,
|
| 2817 |
+
"step": 802
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 0.07058257094984557,
|
| 2821 |
+
"grad_norm": 0.1064453125,
|
| 2822 |
+
"learning_rate": 0.005,
|
| 2823 |
+
"loss": 2.756058692932129,
|
| 2824 |
+
"step": 804
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 0.07075814948454669,
|
| 2828 |
+
"grad_norm": 0.0927734375,
|
| 2829 |
+
"learning_rate": 0.005,
|
| 2830 |
+
"loss": 2.7573466300964355,
|
| 2831 |
+
"step": 806
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 0.0709337280192478,
|
| 2835 |
+
"grad_norm": 0.09619140625,
|
| 2836 |
+
"learning_rate": 0.005,
|
| 2837 |
+
"loss": 2.804060220718384,
|
| 2838 |
+
"step": 808
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 0.07110930655394891,
|
| 2842 |
+
"grad_norm": 0.095703125,
|
| 2843 |
+
"learning_rate": 0.005,
|
| 2844 |
+
"loss": 2.8130621910095215,
|
| 2845 |
+
"step": 810
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 0.07128488508865001,
|
| 2849 |
+
"grad_norm": 0.0869140625,
|
| 2850 |
+
"learning_rate": 0.005,
|
| 2851 |
+
"loss": 2.744436264038086,
|
| 2852 |
+
"step": 812
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 0.07146046362335112,
|
| 2856 |
+
"grad_norm": 0.09619140625,
|
| 2857 |
+
"learning_rate": 0.005,
|
| 2858 |
+
"loss": 2.813497304916382,
|
| 2859 |
+
"step": 814
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 0.07163604215805223,
|
| 2863 |
+
"grad_norm": 0.11669921875,
|
| 2864 |
+
"learning_rate": 0.005,
|
| 2865 |
+
"loss": 2.8008008003234863,
|
| 2866 |
+
"step": 816
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 0.07181162069275333,
|
| 2870 |
+
"grad_norm": 0.091796875,
|
| 2871 |
+
"learning_rate": 0.005,
|
| 2872 |
+
"loss": 2.7890126705169678,
|
| 2873 |
+
"step": 818
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 0.07198719922745445,
|
| 2877 |
+
"grad_norm": 0.119140625,
|
| 2878 |
+
"learning_rate": 0.005,
|
| 2879 |
+
"loss": 2.784027576446533,
|
| 2880 |
+
"step": 820
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 0.07216277776215556,
|
| 2884 |
+
"grad_norm": 0.0869140625,
|
| 2885 |
+
"learning_rate": 0.005,
|
| 2886 |
+
"loss": 2.790325880050659,
|
| 2887 |
+
"step": 822
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 0.07233835629685667,
|
| 2891 |
+
"grad_norm": 0.0908203125,
|
| 2892 |
+
"learning_rate": 0.005,
|
| 2893 |
+
"loss": 2.780329704284668,
|
| 2894 |
+
"step": 824
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 0.07251393483155777,
|
| 2898 |
+
"grad_norm": 0.11865234375,
|
| 2899 |
+
"learning_rate": 0.005,
|
| 2900 |
+
"loss": 2.7846665382385254,
|
| 2901 |
+
"step": 826
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 0.07268951336625888,
|
| 2905 |
+
"grad_norm": 0.1171875,
|
| 2906 |
+
"learning_rate": 0.005,
|
| 2907 |
+
"loss": 2.786909818649292,
|
| 2908 |
+
"step": 828
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 0.07286509190096,
|
| 2912 |
+
"grad_norm": 0.11083984375,
|
| 2913 |
+
"learning_rate": 0.005,
|
| 2914 |
+
"loss": 2.808149814605713,
|
| 2915 |
+
"step": 830
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 0.07304067043566109,
|
| 2919 |
+
"grad_norm": 0.1025390625,
|
| 2920 |
+
"learning_rate": 0.005,
|
| 2921 |
+
"loss": 2.758054733276367,
|
| 2922 |
+
"step": 832
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 0.0732162489703622,
|
| 2926 |
+
"grad_norm": 0.09716796875,
|
| 2927 |
+
"learning_rate": 0.005,
|
| 2928 |
+
"loss": 2.8043644428253174,
|
| 2929 |
+
"step": 834
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 0.07339182750506332,
|
| 2933 |
+
"grad_norm": 0.0947265625,
|
| 2934 |
+
"learning_rate": 0.005,
|
| 2935 |
+
"loss": 2.7352335453033447,
|
| 2936 |
+
"step": 836
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 0.07356740603976443,
|
| 2940 |
+
"grad_norm": 0.11328125,
|
| 2941 |
+
"learning_rate": 0.005,
|
| 2942 |
+
"loss": 2.762988567352295,
|
| 2943 |
+
"step": 838
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 0.07374298457446553,
|
| 2947 |
+
"grad_norm": 0.10595703125,
|
| 2948 |
+
"learning_rate": 0.005,
|
| 2949 |
+
"loss": 2.7505500316619873,
|
| 2950 |
+
"step": 840
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 0.07391856310916664,
|
| 2954 |
+
"grad_norm": 0.08837890625,
|
| 2955 |
+
"learning_rate": 0.005,
|
| 2956 |
+
"loss": 2.756007194519043,
|
| 2957 |
+
"step": 842
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 0.07409414164386775,
|
| 2961 |
+
"grad_norm": 0.08203125,
|
| 2962 |
+
"learning_rate": 0.005,
|
| 2963 |
+
"loss": 2.740776538848877,
|
| 2964 |
+
"step": 844
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 0.07426972017856885,
|
| 2968 |
+
"grad_norm": 0.103515625,
|
| 2969 |
+
"learning_rate": 0.005,
|
| 2970 |
+
"loss": 2.766468048095703,
|
| 2971 |
+
"step": 846
|
| 2972 |
+
},
|
| 2973 |
+
{
|
| 2974 |
+
"epoch": 0.07444529871326996,
|
| 2975 |
+
"grad_norm": 0.10009765625,
|
| 2976 |
+
"learning_rate": 0.005,
|
| 2977 |
+
"loss": 2.768131971359253,
|
| 2978 |
+
"step": 848
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"epoch": 0.07462087724797108,
|
| 2982 |
+
"grad_norm": 0.09326171875,
|
| 2983 |
+
"learning_rate": 0.005,
|
| 2984 |
+
"loss": 2.7882707118988037,
|
| 2985 |
+
"step": 850
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 0.07479645578267219,
|
| 2989 |
+
"grad_norm": 0.08935546875,
|
| 2990 |
+
"learning_rate": 0.005,
|
| 2991 |
+
"loss": 2.759927988052368,
|
| 2992 |
+
"step": 852
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 0.07497203431737329,
|
| 2996 |
+
"grad_norm": 0.09814453125,
|
| 2997 |
+
"learning_rate": 0.005,
|
| 2998 |
+
"loss": 2.772904634475708,
|
| 2999 |
+
"step": 854
|
| 3000 |
+
},
|
| 3001 |
+
{
|
| 3002 |
+
"epoch": 0.0751476128520744,
|
| 3003 |
+
"grad_norm": 0.11279296875,
|
| 3004 |
+
"learning_rate": 0.005,
|
| 3005 |
+
"loss": 2.7328786849975586,
|
| 3006 |
+
"step": 856
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 0.07532319138677551,
|
| 3010 |
+
"grad_norm": 0.1025390625,
|
| 3011 |
+
"learning_rate": 0.005,
|
| 3012 |
+
"loss": 2.789806842803955,
|
| 3013 |
+
"step": 858
|
| 3014 |
+
},
|
| 3015 |
+
{
|
| 3016 |
+
"epoch": 0.07549876992147661,
|
| 3017 |
+
"grad_norm": 0.091796875,
|
| 3018 |
+
"learning_rate": 0.005,
|
| 3019 |
+
"loss": 2.771458148956299,
|
| 3020 |
+
"step": 860
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 0.07567434845617772,
|
| 3024 |
+
"grad_norm": 0.11962890625,
|
| 3025 |
+
"learning_rate": 0.005,
|
| 3026 |
+
"loss": 2.7628495693206787,
|
| 3027 |
+
"step": 862
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 0.07584992699087884,
|
| 3031 |
+
"grad_norm": 0.0966796875,
|
| 3032 |
+
"learning_rate": 0.005,
|
| 3033 |
+
"loss": 2.770573377609253,
|
| 3034 |
+
"step": 864
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 0.07602550552557993,
|
| 3038 |
+
"grad_norm": 0.09228515625,
|
| 3039 |
+
"learning_rate": 0.005,
|
| 3040 |
+
"loss": 2.793875217437744,
|
| 3041 |
+
"step": 866
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 0.07620108406028105,
|
| 3045 |
+
"grad_norm": 0.091796875,
|
| 3046 |
+
"learning_rate": 0.005,
|
| 3047 |
+
"loss": 2.7835752964019775,
|
| 3048 |
+
"step": 868
|
| 3049 |
+
},
|
| 3050 |
+
{
|
| 3051 |
+
"epoch": 0.07637666259498216,
|
| 3052 |
+
"grad_norm": 0.08544921875,
|
| 3053 |
+
"learning_rate": 0.005,
|
| 3054 |
+
"loss": 2.775233745574951,
|
| 3055 |
+
"step": 870
|
| 3056 |
+
},
|
| 3057 |
+
{
|
| 3058 |
+
"epoch": 0.07655224112968327,
|
| 3059 |
+
"grad_norm": 0.08544921875,
|
| 3060 |
+
"learning_rate": 0.005,
|
| 3061 |
+
"loss": 2.7418532371520996,
|
| 3062 |
+
"step": 872
|
| 3063 |
+
},
|
| 3064 |
+
{
|
| 3065 |
+
"epoch": 0.07672781966438437,
|
| 3066 |
+
"grad_norm": 0.08740234375,
|
| 3067 |
+
"learning_rate": 0.005,
|
| 3068 |
+
"loss": 2.754110097885132,
|
| 3069 |
+
"step": 874
|
| 3070 |
+
},
|
| 3071 |
+
{
|
| 3072 |
+
"epoch": 0.07690339819908548,
|
| 3073 |
+
"grad_norm": 0.0810546875,
|
| 3074 |
+
"learning_rate": 0.005,
|
| 3075 |
+
"loss": 2.773123264312744,
|
| 3076 |
+
"step": 876
|
| 3077 |
+
},
|
| 3078 |
+
{
|
| 3079 |
+
"epoch": 0.0770789767337866,
|
| 3080 |
+
"grad_norm": 0.083984375,
|
| 3081 |
+
"learning_rate": 0.005,
|
| 3082 |
+
"loss": 2.750389337539673,
|
| 3083 |
+
"step": 878
|
| 3084 |
+
},
|
| 3085 |
+
{
|
| 3086 |
+
"epoch": 0.07725455526848769,
|
| 3087 |
+
"grad_norm": 0.0830078125,
|
| 3088 |
+
"learning_rate": 0.005,
|
| 3089 |
+
"loss": 2.733523368835449,
|
| 3090 |
+
"step": 880
|
| 3091 |
+
},
|
| 3092 |
+
{
|
| 3093 |
+
"epoch": 0.0774301338031888,
|
| 3094 |
+
"grad_norm": 0.08984375,
|
| 3095 |
+
"learning_rate": 0.005,
|
| 3096 |
+
"loss": 2.7816543579101562,
|
| 3097 |
+
"step": 882
|
| 3098 |
+
},
|
| 3099 |
+
{
|
| 3100 |
+
"epoch": 0.07760571233788992,
|
| 3101 |
+
"grad_norm": 0.0947265625,
|
| 3102 |
+
"learning_rate": 0.005,
|
| 3103 |
+
"loss": 2.7417476177215576,
|
| 3104 |
+
"step": 884
|
| 3105 |
+
},
|
| 3106 |
+
{
|
| 3107 |
+
"epoch": 0.07778129087259103,
|
| 3108 |
+
"grad_norm": 0.10302734375,
|
| 3109 |
+
"learning_rate": 0.005,
|
| 3110 |
+
"loss": 2.7590861320495605,
|
| 3111 |
+
"step": 886
|
| 3112 |
+
},
|
| 3113 |
+
{
|
| 3114 |
+
"epoch": 0.07795686940729213,
|
| 3115 |
+
"grad_norm": 0.099609375,
|
| 3116 |
+
"learning_rate": 0.005,
|
| 3117 |
+
"loss": 2.74955677986145,
|
| 3118 |
+
"step": 888
|
| 3119 |
+
},
|
| 3120 |
+
{
|
| 3121 |
+
"epoch": 0.07813244794199324,
|
| 3122 |
+
"grad_norm": 0.09375,
|
| 3123 |
+
"learning_rate": 0.005,
|
| 3124 |
+
"loss": 2.7264111042022705,
|
| 3125 |
+
"step": 890
|
| 3126 |
+
},
|
| 3127 |
+
{
|
| 3128 |
+
"epoch": 0.07830802647669435,
|
| 3129 |
+
"grad_norm": 0.09521484375,
|
| 3130 |
+
"learning_rate": 0.005,
|
| 3131 |
+
"loss": 2.7831571102142334,
|
| 3132 |
+
"step": 892
|
| 3133 |
+
},
|
| 3134 |
+
{
|
| 3135 |
+
"epoch": 0.07848360501139545,
|
| 3136 |
+
"grad_norm": 0.09716796875,
|
| 3137 |
+
"learning_rate": 0.005,
|
| 3138 |
+
"loss": 2.777005672454834,
|
| 3139 |
+
"step": 894
|
| 3140 |
+
},
|
| 3141 |
+
{
|
| 3142 |
+
"epoch": 0.07865918354609656,
|
| 3143 |
+
"grad_norm": 0.09716796875,
|
| 3144 |
+
"learning_rate": 0.005,
|
| 3145 |
+
"loss": 2.759586811065674,
|
| 3146 |
+
"step": 896
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 0.07883476208079768,
|
| 3150 |
+
"grad_norm": 0.083984375,
|
| 3151 |
+
"learning_rate": 0.005,
|
| 3152 |
+
"loss": 2.750412940979004,
|
| 3153 |
+
"step": 898
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 0.07901034061549879,
|
| 3157 |
+
"grad_norm": 0.08447265625,
|
| 3158 |
+
"learning_rate": 0.005,
|
| 3159 |
+
"loss": 2.7232556343078613,
|
| 3160 |
+
"step": 900
|
| 3161 |
}
|
| 3162 |
],
|
| 3163 |
"logging_steps": 2,
|
|
|
|
| 3177 |
"attributes": {}
|
| 3178 |
}
|
| 3179 |
},
|
| 3180 |
+
"total_flos": 1.5213476081262559e+18,
|
| 3181 |
"train_batch_size": 2,
|
| 3182 |
"trial_name": null,
|
| 3183 |
"trial_params": null
|