Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +633 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,633 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.001000500250125,
|
| 5 |
+
"eval_steps": 25,
|
| 6 |
+
"global_step": 1000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.05002501250625312,
|
| 13 |
+
"grad_norm": 0.517996072769165,
|
| 14 |
+
"learning_rate": 0.0001951951951951952,
|
| 15 |
+
"loss": 1.677,
|
| 16 |
+
"step": 25
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05002501250625312,
|
| 20 |
+
"eval_loss": 1.3813503980636597,
|
| 21 |
+
"eval_runtime": 148.0614,
|
| 22 |
+
"eval_samples_per_second": 3.37,
|
| 23 |
+
"eval_steps_per_second": 0.425,
|
| 24 |
+
"step": 25
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.10005002501250625,
|
| 28 |
+
"grad_norm": 0.5020231604576111,
|
| 29 |
+
"learning_rate": 0.0001901901901901902,
|
| 30 |
+
"loss": 1.2016,
|
| 31 |
+
"step": 50
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.10005002501250625,
|
| 35 |
+
"eval_loss": 1.347744107246399,
|
| 36 |
+
"eval_runtime": 151.7258,
|
| 37 |
+
"eval_samples_per_second": 3.289,
|
| 38 |
+
"eval_steps_per_second": 0.415,
|
| 39 |
+
"step": 50
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.1500750375187594,
|
| 43 |
+
"grad_norm": 0.3798060119152069,
|
| 44 |
+
"learning_rate": 0.0001851851851851852,
|
| 45 |
+
"loss": 1.4491,
|
| 46 |
+
"step": 75
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.1500750375187594,
|
| 50 |
+
"eval_loss": 1.3210723400115967,
|
| 51 |
+
"eval_runtime": 150.0032,
|
| 52 |
+
"eval_samples_per_second": 3.327,
|
| 53 |
+
"eval_steps_per_second": 0.42,
|
| 54 |
+
"step": 75
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.2001000500250125,
|
| 58 |
+
"grad_norm": 0.3365944027900696,
|
| 59 |
+
"learning_rate": 0.00018018018018018018,
|
| 60 |
+
"loss": 1.2076,
|
| 61 |
+
"step": 100
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.2001000500250125,
|
| 65 |
+
"eval_loss": 1.3334178924560547,
|
| 66 |
+
"eval_runtime": 151.2551,
|
| 67 |
+
"eval_samples_per_second": 3.299,
|
| 68 |
+
"eval_steps_per_second": 0.417,
|
| 69 |
+
"step": 100
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.25012506253126565,
|
| 73 |
+
"grad_norm": 0.22820694744586945,
|
| 74 |
+
"learning_rate": 0.0001751751751751752,
|
| 75 |
+
"loss": 1.4415,
|
| 76 |
+
"step": 125
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.25012506253126565,
|
| 80 |
+
"eval_loss": 1.309592366218567,
|
| 81 |
+
"eval_runtime": 149.299,
|
| 82 |
+
"eval_samples_per_second": 3.342,
|
| 83 |
+
"eval_steps_per_second": 0.422,
|
| 84 |
+
"step": 125
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.3001500750375188,
|
| 88 |
+
"grad_norm": 0.3848935663700104,
|
| 89 |
+
"learning_rate": 0.0001701701701701702,
|
| 90 |
+
"loss": 1.139,
|
| 91 |
+
"step": 150
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.3001500750375188,
|
| 95 |
+
"eval_loss": 1.3208202123641968,
|
| 96 |
+
"eval_runtime": 149.5811,
|
| 97 |
+
"eval_samples_per_second": 3.336,
|
| 98 |
+
"eval_steps_per_second": 0.421,
|
| 99 |
+
"step": 150
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.3501750875437719,
|
| 103 |
+
"grad_norm": 0.2774136960506439,
|
| 104 |
+
"learning_rate": 0.00016516516516516518,
|
| 105 |
+
"loss": 1.4055,
|
| 106 |
+
"step": 175
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.3501750875437719,
|
| 110 |
+
"eval_loss": 1.3086917400360107,
|
| 111 |
+
"eval_runtime": 150.1042,
|
| 112 |
+
"eval_samples_per_second": 3.324,
|
| 113 |
+
"eval_steps_per_second": 0.42,
|
| 114 |
+
"step": 175
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.400200100050025,
|
| 118 |
+
"grad_norm": 0.32166117429733276,
|
| 119 |
+
"learning_rate": 0.00016016016016016018,
|
| 120 |
+
"loss": 1.1459,
|
| 121 |
+
"step": 200
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.400200100050025,
|
| 125 |
+
"eval_loss": 1.306862473487854,
|
| 126 |
+
"eval_runtime": 150.7168,
|
| 127 |
+
"eval_samples_per_second": 3.311,
|
| 128 |
+
"eval_steps_per_second": 0.418,
|
| 129 |
+
"step": 200
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.4502251125562781,
|
| 133 |
+
"grad_norm": 0.23773141205310822,
|
| 134 |
+
"learning_rate": 0.00015515515515515516,
|
| 135 |
+
"loss": 1.4444,
|
| 136 |
+
"step": 225
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.4502251125562781,
|
| 140 |
+
"eval_loss": 1.3020325899124146,
|
| 141 |
+
"eval_runtime": 148.5364,
|
| 142 |
+
"eval_samples_per_second": 3.359,
|
| 143 |
+
"eval_steps_per_second": 0.424,
|
| 144 |
+
"step": 225
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.5002501250625313,
|
| 148 |
+
"grad_norm": 0.37095341086387634,
|
| 149 |
+
"learning_rate": 0.00015015015015015014,
|
| 150 |
+
"loss": 1.2264,
|
| 151 |
+
"step": 250
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.5002501250625313,
|
| 155 |
+
"eval_loss": 1.3001904487609863,
|
| 156 |
+
"eval_runtime": 152.658,
|
| 157 |
+
"eval_samples_per_second": 3.269,
|
| 158 |
+
"eval_steps_per_second": 0.413,
|
| 159 |
+
"step": 250
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.5502751375687844,
|
| 163 |
+
"grad_norm": 0.2519828677177429,
|
| 164 |
+
"learning_rate": 0.00014514514514514515,
|
| 165 |
+
"loss": 1.4605,
|
| 166 |
+
"step": 275
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.5502751375687844,
|
| 170 |
+
"eval_loss": 1.299567699432373,
|
| 171 |
+
"eval_runtime": 148.4653,
|
| 172 |
+
"eval_samples_per_second": 3.361,
|
| 173 |
+
"eval_steps_per_second": 0.424,
|
| 174 |
+
"step": 275
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.6003001500750376,
|
| 178 |
+
"grad_norm": 0.3685779273509979,
|
| 179 |
+
"learning_rate": 0.00014014014014014013,
|
| 180 |
+
"loss": 1.1655,
|
| 181 |
+
"step": 300
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 0.6003001500750376,
|
| 185 |
+
"eval_loss": 1.2988265752792358,
|
| 186 |
+
"eval_runtime": 151.1788,
|
| 187 |
+
"eval_samples_per_second": 3.301,
|
| 188 |
+
"eval_steps_per_second": 0.417,
|
| 189 |
+
"step": 300
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.6503251625812907,
|
| 193 |
+
"grad_norm": 0.26966241002082825,
|
| 194 |
+
"learning_rate": 0.00013513513513513514,
|
| 195 |
+
"loss": 1.4313,
|
| 196 |
+
"step": 325
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.6503251625812907,
|
| 200 |
+
"eval_loss": 1.298296332359314,
|
| 201 |
+
"eval_runtime": 152.0718,
|
| 202 |
+
"eval_samples_per_second": 3.281,
|
| 203 |
+
"eval_steps_per_second": 0.414,
|
| 204 |
+
"step": 325
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.7003501750875438,
|
| 208 |
+
"grad_norm": 0.35637611150741577,
|
| 209 |
+
"learning_rate": 0.00013013013013013014,
|
| 210 |
+
"loss": 1.2002,
|
| 211 |
+
"step": 350
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.7003501750875438,
|
| 215 |
+
"eval_loss": 1.2959158420562744,
|
| 216 |
+
"eval_runtime": 151.1585,
|
| 217 |
+
"eval_samples_per_second": 3.301,
|
| 218 |
+
"eval_steps_per_second": 0.417,
|
| 219 |
+
"step": 350
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.7503751875937968,
|
| 223 |
+
"grad_norm": 0.22513383626937866,
|
| 224 |
+
"learning_rate": 0.00012512512512512512,
|
| 225 |
+
"loss": 1.3994,
|
| 226 |
+
"step": 375
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.7503751875937968,
|
| 230 |
+
"eval_loss": 1.2951635122299194,
|
| 231 |
+
"eval_runtime": 148.5372,
|
| 232 |
+
"eval_samples_per_second": 3.359,
|
| 233 |
+
"eval_steps_per_second": 0.424,
|
| 234 |
+
"step": 375
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.80040020010005,
|
| 238 |
+
"grad_norm": 0.35314086079597473,
|
| 239 |
+
"learning_rate": 0.00012012012012012013,
|
| 240 |
+
"loss": 1.1836,
|
| 241 |
+
"step": 400
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.80040020010005,
|
| 245 |
+
"eval_loss": 1.294690728187561,
|
| 246 |
+
"eval_runtime": 149.3769,
|
| 247 |
+
"eval_samples_per_second": 3.341,
|
| 248 |
+
"eval_steps_per_second": 0.422,
|
| 249 |
+
"step": 400
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.8504252126063031,
|
| 253 |
+
"grad_norm": 0.240916907787323,
|
| 254 |
+
"learning_rate": 0.00011511511511511512,
|
| 255 |
+
"loss": 1.4378,
|
| 256 |
+
"step": 425
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.8504252126063031,
|
| 260 |
+
"eval_loss": 1.2916043996810913,
|
| 261 |
+
"eval_runtime": 152.0772,
|
| 262 |
+
"eval_samples_per_second": 3.281,
|
| 263 |
+
"eval_steps_per_second": 0.414,
|
| 264 |
+
"step": 425
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.9004502251125562,
|
| 268 |
+
"grad_norm": 0.31087398529052734,
|
| 269 |
+
"learning_rate": 0.00011011011011011012,
|
| 270 |
+
"loss": 1.1989,
|
| 271 |
+
"step": 450
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.9004502251125562,
|
| 275 |
+
"eval_loss": 1.2893831729888916,
|
| 276 |
+
"eval_runtime": 150.4895,
|
| 277 |
+
"eval_samples_per_second": 3.316,
|
| 278 |
+
"eval_steps_per_second": 0.419,
|
| 279 |
+
"step": 450
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.9504752376188094,
|
| 283 |
+
"grad_norm": 0.2413586527109146,
|
| 284 |
+
"learning_rate": 0.00010510510510510511,
|
| 285 |
+
"loss": 1.4508,
|
| 286 |
+
"step": 475
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 0.9504752376188094,
|
| 290 |
+
"eval_loss": 1.2888984680175781,
|
| 291 |
+
"eval_runtime": 151.1108,
|
| 292 |
+
"eval_samples_per_second": 3.302,
|
| 293 |
+
"eval_steps_per_second": 0.417,
|
| 294 |
+
"step": 475
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.0005002501250626,
|
| 298 |
+
"grad_norm": 0.40069064497947693,
|
| 299 |
+
"learning_rate": 0.00010010010010010012,
|
| 300 |
+
"loss": 1.2076,
|
| 301 |
+
"step": 500
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 1.0005002501250626,
|
| 305 |
+
"eval_loss": 1.2911962270736694,
|
| 306 |
+
"eval_runtime": 148.6843,
|
| 307 |
+
"eval_samples_per_second": 3.356,
|
| 308 |
+
"eval_steps_per_second": 0.424,
|
| 309 |
+
"step": 500
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.0505252626313157,
|
| 313 |
+
"grad_norm": 0.22050493955612183,
|
| 314 |
+
"learning_rate": 9.50950950950951e-05,
|
| 315 |
+
"loss": 1.3994,
|
| 316 |
+
"step": 525
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.0505252626313157,
|
| 320 |
+
"eval_loss": 1.2921332120895386,
|
| 321 |
+
"eval_runtime": 149.3015,
|
| 322 |
+
"eval_samples_per_second": 3.342,
|
| 323 |
+
"eval_steps_per_second": 0.422,
|
| 324 |
+
"step": 525
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.1005502751375689,
|
| 328 |
+
"grad_norm": 0.3588818907737732,
|
| 329 |
+
"learning_rate": 9.009009009009009e-05,
|
| 330 |
+
"loss": 1.177,
|
| 331 |
+
"step": 550
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.1005502751375689,
|
| 335 |
+
"eval_loss": 1.2903811931610107,
|
| 336 |
+
"eval_runtime": 149.8093,
|
| 337 |
+
"eval_samples_per_second": 3.331,
|
| 338 |
+
"eval_steps_per_second": 0.421,
|
| 339 |
+
"step": 550
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.150575287643822,
|
| 343 |
+
"grad_norm": 0.2672303020954132,
|
| 344 |
+
"learning_rate": 8.50850850850851e-05,
|
| 345 |
+
"loss": 1.4015,
|
| 346 |
+
"step": 575
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.150575287643822,
|
| 350 |
+
"eval_loss": 1.2898900508880615,
|
| 351 |
+
"eval_runtime": 149.8311,
|
| 352 |
+
"eval_samples_per_second": 3.33,
|
| 353 |
+
"eval_steps_per_second": 0.42,
|
| 354 |
+
"step": 575
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.2006003001500751,
|
| 358 |
+
"grad_norm": 0.31220486760139465,
|
| 359 |
+
"learning_rate": 8.008008008008009e-05,
|
| 360 |
+
"loss": 1.192,
|
| 361 |
+
"step": 600
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.2006003001500751,
|
| 365 |
+
"eval_loss": 1.288824439048767,
|
| 366 |
+
"eval_runtime": 151.038,
|
| 367 |
+
"eval_samples_per_second": 3.304,
|
| 368 |
+
"eval_steps_per_second": 0.417,
|
| 369 |
+
"step": 600
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.2506253126563283,
|
| 373 |
+
"grad_norm": 0.2526504695415497,
|
| 374 |
+
"learning_rate": 7.507507507507507e-05,
|
| 375 |
+
"loss": 1.3829,
|
| 376 |
+
"step": 625
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 1.2506253126563283,
|
| 380 |
+
"eval_loss": 1.2878332138061523,
|
| 381 |
+
"eval_runtime": 151.5015,
|
| 382 |
+
"eval_samples_per_second": 3.294,
|
| 383 |
+
"eval_steps_per_second": 0.416,
|
| 384 |
+
"step": 625
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.3006503251625814,
|
| 388 |
+
"grad_norm": 0.28051283955574036,
|
| 389 |
+
"learning_rate": 7.007007007007007e-05,
|
| 390 |
+
"loss": 1.1514,
|
| 391 |
+
"step": 650
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 1.3006503251625814,
|
| 395 |
+
"eval_loss": 1.2859280109405518,
|
| 396 |
+
"eval_runtime": 150.4738,
|
| 397 |
+
"eval_samples_per_second": 3.316,
|
| 398 |
+
"eval_steps_per_second": 0.419,
|
| 399 |
+
"step": 650
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.3506753376688345,
|
| 403 |
+
"grad_norm": 0.26419979333877563,
|
| 404 |
+
"learning_rate": 6.506506506506507e-05,
|
| 405 |
+
"loss": 1.4028,
|
| 406 |
+
"step": 675
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.3506753376688345,
|
| 410 |
+
"eval_loss": 1.2848296165466309,
|
| 411 |
+
"eval_runtime": 149.0963,
|
| 412 |
+
"eval_samples_per_second": 3.347,
|
| 413 |
+
"eval_steps_per_second": 0.423,
|
| 414 |
+
"step": 675
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 1.4007003501750876,
|
| 418 |
+
"grad_norm": 0.3227976858615875,
|
| 419 |
+
"learning_rate": 6.0060060060060066e-05,
|
| 420 |
+
"loss": 1.1778,
|
| 421 |
+
"step": 700
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 1.4007003501750876,
|
| 425 |
+
"eval_loss": 1.285400152206421,
|
| 426 |
+
"eval_runtime": 149.1519,
|
| 427 |
+
"eval_samples_per_second": 3.346,
|
| 428 |
+
"eval_steps_per_second": 0.422,
|
| 429 |
+
"step": 700
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.4507253626813408,
|
| 433 |
+
"grad_norm": 0.24903441965579987,
|
| 434 |
+
"learning_rate": 5.505505505505506e-05,
|
| 435 |
+
"loss": 1.4058,
|
| 436 |
+
"step": 725
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.4507253626813408,
|
| 440 |
+
"eval_loss": 1.2824435234069824,
|
| 441 |
+
"eval_runtime": 149.5232,
|
| 442 |
+
"eval_samples_per_second": 3.337,
|
| 443 |
+
"eval_steps_per_second": 0.421,
|
| 444 |
+
"step": 725
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.500750375187594,
|
| 448 |
+
"grad_norm": 0.31187903881073,
|
| 449 |
+
"learning_rate": 5.005005005005006e-05,
|
| 450 |
+
"loss": 1.1698,
|
| 451 |
+
"step": 750
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.500750375187594,
|
| 455 |
+
"eval_loss": 1.2831988334655762,
|
| 456 |
+
"eval_runtime": 150.4227,
|
| 457 |
+
"eval_samples_per_second": 3.317,
|
| 458 |
+
"eval_steps_per_second": 0.419,
|
| 459 |
+
"step": 750
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.550775387693847,
|
| 463 |
+
"grad_norm": 0.2889004051685333,
|
| 464 |
+
"learning_rate": 4.5045045045045046e-05,
|
| 465 |
+
"loss": 1.3516,
|
| 466 |
+
"step": 775
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.550775387693847,
|
| 470 |
+
"eval_loss": 1.2823545932769775,
|
| 471 |
+
"eval_runtime": 149.8614,
|
| 472 |
+
"eval_samples_per_second": 3.33,
|
| 473 |
+
"eval_steps_per_second": 0.42,
|
| 474 |
+
"step": 775
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 1.6008004002001002,
|
| 478 |
+
"grad_norm": 0.37189939618110657,
|
| 479 |
+
"learning_rate": 4.0040040040040046e-05,
|
| 480 |
+
"loss": 1.1264,
|
| 481 |
+
"step": 800
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 1.6008004002001002,
|
| 485 |
+
"eval_loss": 1.2828818559646606,
|
| 486 |
+
"eval_runtime": 150.672,
|
| 487 |
+
"eval_samples_per_second": 3.312,
|
| 488 |
+
"eval_steps_per_second": 0.418,
|
| 489 |
+
"step": 800
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.6508254127063533,
|
| 493 |
+
"grad_norm": 0.25290611386299133,
|
| 494 |
+
"learning_rate": 3.503503503503503e-05,
|
| 495 |
+
"loss": 1.4113,
|
| 496 |
+
"step": 825
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 1.6508254127063533,
|
| 500 |
+
"eval_loss": 1.2822470664978027,
|
| 501 |
+
"eval_runtime": 149.3988,
|
| 502 |
+
"eval_samples_per_second": 3.34,
|
| 503 |
+
"eval_steps_per_second": 0.422,
|
| 504 |
+
"step": 825
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 1.7008504252126064,
|
| 508 |
+
"grad_norm": 0.3559873104095459,
|
| 509 |
+
"learning_rate": 3.0030030030030033e-05,
|
| 510 |
+
"loss": 1.1248,
|
| 511 |
+
"step": 850
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 1.7008504252126064,
|
| 515 |
+
"eval_loss": 1.2828270196914673,
|
| 516 |
+
"eval_runtime": 149.9897,
|
| 517 |
+
"eval_samples_per_second": 3.327,
|
| 518 |
+
"eval_steps_per_second": 0.42,
|
| 519 |
+
"step": 850
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.7508754377188596,
|
| 523 |
+
"grad_norm": 0.3052867352962494,
|
| 524 |
+
"learning_rate": 2.502502502502503e-05,
|
| 525 |
+
"loss": 1.336,
|
| 526 |
+
"step": 875
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 1.7508754377188596,
|
| 530 |
+
"eval_loss": 1.282852053642273,
|
| 531 |
+
"eval_runtime": 151.397,
|
| 532 |
+
"eval_samples_per_second": 3.296,
|
| 533 |
+
"eval_steps_per_second": 0.416,
|
| 534 |
+
"step": 875
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.8009004502251127,
|
| 538 |
+
"grad_norm": 0.33662667870521545,
|
| 539 |
+
"learning_rate": 2.0020020020020023e-05,
|
| 540 |
+
"loss": 1.0725,
|
| 541 |
+
"step": 900
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.8009004502251127,
|
| 545 |
+
"eval_loss": 1.2822794914245605,
|
| 546 |
+
"eval_runtime": 150.7632,
|
| 547 |
+
"eval_samples_per_second": 3.31,
|
| 548 |
+
"eval_steps_per_second": 0.418,
|
| 549 |
+
"step": 900
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.8509254627313658,
|
| 553 |
+
"grad_norm": 0.29956212639808655,
|
| 554 |
+
"learning_rate": 1.5015015015015016e-05,
|
| 555 |
+
"loss": 1.3989,
|
| 556 |
+
"step": 925
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.8509254627313658,
|
| 560 |
+
"eval_loss": 1.2824186086654663,
|
| 561 |
+
"eval_runtime": 150.6938,
|
| 562 |
+
"eval_samples_per_second": 3.311,
|
| 563 |
+
"eval_steps_per_second": 0.418,
|
| 564 |
+
"step": 925
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.900950475237619,
|
| 568 |
+
"grad_norm": 0.3255136013031006,
|
| 569 |
+
"learning_rate": 1.0010010010010011e-05,
|
| 570 |
+
"loss": 1.112,
|
| 571 |
+
"step": 950
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.900950475237619,
|
| 575 |
+
"eval_loss": 1.28144371509552,
|
| 576 |
+
"eval_runtime": 149.8969,
|
| 577 |
+
"eval_samples_per_second": 3.329,
|
| 578 |
+
"eval_steps_per_second": 0.42,
|
| 579 |
+
"step": 950
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 1.950975487743872,
|
| 583 |
+
"grad_norm": 0.2689700424671173,
|
| 584 |
+
"learning_rate": 5.005005005005006e-06,
|
| 585 |
+
"loss": 1.3972,
|
| 586 |
+
"step": 975
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.950975487743872,
|
| 590 |
+
"eval_loss": 1.280760645866394,
|
| 591 |
+
"eval_runtime": 149.8977,
|
| 592 |
+
"eval_samples_per_second": 3.329,
|
| 593 |
+
"eval_steps_per_second": 0.42,
|
| 594 |
+
"step": 975
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.001000500250125,
|
| 598 |
+
"grad_norm": 0.3633726239204407,
|
| 599 |
+
"learning_rate": 0.0,
|
| 600 |
+
"loss": 1.1746,
|
| 601 |
+
"step": 1000
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.001000500250125,
|
| 605 |
+
"eval_loss": 1.2818013429641724,
|
| 606 |
+
"eval_runtime": 149.8121,
|
| 607 |
+
"eval_samples_per_second": 3.331,
|
| 608 |
+
"eval_steps_per_second": 0.421,
|
| 609 |
+
"step": 1000
|
| 610 |
+
}
|
| 611 |
+
],
|
| 612 |
+
"logging_steps": 25,
|
| 613 |
+
"max_steps": 1000,
|
| 614 |
+
"num_input_tokens_seen": 0,
|
| 615 |
+
"num_train_epochs": 3,
|
| 616 |
+
"save_steps": 25,
|
| 617 |
+
"stateful_callbacks": {
|
| 618 |
+
"TrainerControl": {
|
| 619 |
+
"args": {
|
| 620 |
+
"should_epoch_stop": false,
|
| 621 |
+
"should_evaluate": false,
|
| 622 |
+
"should_log": false,
|
| 623 |
+
"should_save": true,
|
| 624 |
+
"should_training_stop": true
|
| 625 |
+
},
|
| 626 |
+
"attributes": {}
|
| 627 |
+
}
|
| 628 |
+
},
|
| 629 |
+
"total_flos": 1.844485620424704e+16,
|
| 630 |
+
"train_batch_size": 1,
|
| 631 |
+
"trial_name": null,
|
| 632 |
+
"trial_params": null
|
| 633 |
+
}
|