Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +807 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,807 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 530,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0037735849056603774,
|
| 13 |
+
"grad_norm": 15.33187898922165,
|
| 14 |
+
"learning_rate": 3.773584905660378e-07,
|
| 15 |
+
"loss": 1.3947,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.018867924528301886,
|
| 20 |
+
"grad_norm": 14.587979298665214,
|
| 21 |
+
"learning_rate": 1.8867924528301889e-06,
|
| 22 |
+
"loss": 1.2879,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03773584905660377,
|
| 27 |
+
"grad_norm": 5.156574643577073,
|
| 28 |
+
"learning_rate": 3.7735849056603777e-06,
|
| 29 |
+
"loss": 1.1356,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.05660377358490566,
|
| 34 |
+
"grad_norm": 2.4413584926483494,
|
| 35 |
+
"learning_rate": 5.660377358490566e-06,
|
| 36 |
+
"loss": 1.056,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.07547169811320754,
|
| 41 |
+
"grad_norm": 1.939557281184319,
|
| 42 |
+
"learning_rate": 7.5471698113207555e-06,
|
| 43 |
+
"loss": 1.0035,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09433962264150944,
|
| 48 |
+
"grad_norm": 1.445408046384172,
|
| 49 |
+
"learning_rate": 9.433962264150944e-06,
|
| 50 |
+
"loss": 0.9666,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.11320754716981132,
|
| 55 |
+
"grad_norm": 1.5429900501633669,
|
| 56 |
+
"learning_rate": 1.1320754716981132e-05,
|
| 57 |
+
"loss": 0.9673,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.1320754716981132,
|
| 62 |
+
"grad_norm": 1.2999242817660164,
|
| 63 |
+
"learning_rate": 1.320754716981132e-05,
|
| 64 |
+
"loss": 0.9161,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1509433962264151,
|
| 69 |
+
"grad_norm": 1.1690976259696864,
|
| 70 |
+
"learning_rate": 1.5094339622641511e-05,
|
| 71 |
+
"loss": 0.9029,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.16981132075471697,
|
| 76 |
+
"grad_norm": 1.2404008101715356,
|
| 77 |
+
"learning_rate": 1.69811320754717e-05,
|
| 78 |
+
"loss": 0.8976,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.18867924528301888,
|
| 83 |
+
"grad_norm": 1.2490209366453322,
|
| 84 |
+
"learning_rate": 1.8867924528301888e-05,
|
| 85 |
+
"loss": 0.9202,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.20754716981132076,
|
| 90 |
+
"grad_norm": 1.2687062081137546,
|
| 91 |
+
"learning_rate": 1.9999132465602526e-05,
|
| 92 |
+
"loss": 0.9125,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.22641509433962265,
|
| 97 |
+
"grad_norm": 1.2693770693981177,
|
| 98 |
+
"learning_rate": 1.998937443221316e-05,
|
| 99 |
+
"loss": 0.918,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.24528301886792453,
|
| 104 |
+
"grad_norm": 1.17511999372123,
|
| 105 |
+
"learning_rate": 1.9968784563700586e-05,
|
| 106 |
+
"loss": 0.8731,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.2641509433962264,
|
| 111 |
+
"grad_norm": 1.1148641046803303,
|
| 112 |
+
"learning_rate": 1.9937385186393888e-05,
|
| 113 |
+
"loss": 0.9063,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.2830188679245283,
|
| 118 |
+
"grad_norm": 1.2209470360517312,
|
| 119 |
+
"learning_rate": 1.9895210347758233e-05,
|
| 120 |
+
"loss": 0.8973,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3018867924528302,
|
| 125 |
+
"grad_norm": 1.2489264571555163,
|
| 126 |
+
"learning_rate": 1.984230577947597e-05,
|
| 127 |
+
"loss": 0.9086,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.32075471698113206,
|
| 132 |
+
"grad_norm": 1.7088824250739199,
|
| 133 |
+
"learning_rate": 1.977872884785815e-05,
|
| 134 |
+
"loss": 0.891,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.33962264150943394,
|
| 139 |
+
"grad_norm": 1.1609543047434623,
|
| 140 |
+
"learning_rate": 1.9704548491640195e-05,
|
| 141 |
+
"loss": 0.9064,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3584905660377358,
|
| 146 |
+
"grad_norm": 1.1548579761616171,
|
| 147 |
+
"learning_rate": 1.961984514722914e-05,
|
| 148 |
+
"loss": 0.886,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.37735849056603776,
|
| 153 |
+
"grad_norm": 1.2072363172467877,
|
| 154 |
+
"learning_rate": 1.9524710661483594e-05,
|
| 155 |
+
"loss": 0.914,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.39622641509433965,
|
| 160 |
+
"grad_norm": 1.2486639912473576,
|
| 161 |
+
"learning_rate": 1.94192481921209e-05,
|
| 162 |
+
"loss": 0.8858,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.41509433962264153,
|
| 167 |
+
"grad_norm": 1.1907067530146591,
|
| 168 |
+
"learning_rate": 1.9303572095859545e-05,
|
| 169 |
+
"loss": 0.883,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4339622641509434,
|
| 174 |
+
"grad_norm": 1.205701711118422,
|
| 175 |
+
"learning_rate": 1.91778078044181e-05,
|
| 176 |
+
"loss": 0.903,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4528301886792453,
|
| 181 |
+
"grad_norm": 1.4297409544875106,
|
| 182 |
+
"learning_rate": 1.9042091688505104e-05,
|
| 183 |
+
"loss": 0.8903,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.4716981132075472,
|
| 188 |
+
"grad_norm": 1.3015347726404307,
|
| 189 |
+
"learning_rate": 1.8896570909947477e-05,
|
| 190 |
+
"loss": 0.8975,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.49056603773584906,
|
| 195 |
+
"grad_norm": 1.1056494521961182,
|
| 196 |
+
"learning_rate": 1.874140326211766e-05,
|
| 197 |
+
"loss": 0.8973,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5094339622641509,
|
| 202 |
+
"grad_norm": 1.3117079939596017,
|
| 203 |
+
"learning_rate": 1.8576756998832667e-05,
|
| 204 |
+
"loss": 0.8665,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5283018867924528,
|
| 209 |
+
"grad_norm": 1.0536496719085662,
|
| 210 |
+
"learning_rate": 1.8402810651910444e-05,
|
| 211 |
+
"loss": 0.8638,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5471698113207547,
|
| 216 |
+
"grad_norm": 2.702397102319974,
|
| 217 |
+
"learning_rate": 1.8219752837581466e-05,
|
| 218 |
+
"loss": 0.882,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5660377358490566,
|
| 223 |
+
"grad_norm": 1.0422440813400673,
|
| 224 |
+
"learning_rate": 1.8027782051965408e-05,
|
| 225 |
+
"loss": 0.8657,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5849056603773585,
|
| 230 |
+
"grad_norm": 1.3024995956700887,
|
| 231 |
+
"learning_rate": 1.782710645583473e-05,
|
| 232 |
+
"loss": 0.8599,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.6037735849056604,
|
| 237 |
+
"grad_norm": 1.1482394390531847,
|
| 238 |
+
"learning_rate": 1.761794364889855e-05,
|
| 239 |
+
"loss": 0.8926,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6226415094339622,
|
| 244 |
+
"grad_norm": 1.079599574770899,
|
| 245 |
+
"learning_rate": 1.7400520433851457e-05,
|
| 246 |
+
"loss": 0.8428,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6415094339622641,
|
| 251 |
+
"grad_norm": 1.0901592446951243,
|
| 252 |
+
"learning_rate": 1.717507257044331e-05,
|
| 253 |
+
"loss": 0.8496,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.660377358490566,
|
| 258 |
+
"grad_norm": 1.1077239121822327,
|
| 259 |
+
"learning_rate": 1.694184451983651e-05,
|
| 260 |
+
"loss": 0.9012,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6792452830188679,
|
| 265 |
+
"grad_norm": 1.0853595621923493,
|
| 266 |
+
"learning_rate": 1.6701089179528032e-05,
|
| 267 |
+
"loss": 0.8681,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6981132075471698,
|
| 272 |
+
"grad_norm": 1.0734841287772214,
|
| 273 |
+
"learning_rate": 1.6453067609123656e-05,
|
| 274 |
+
"loss": 0.856,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.7169811320754716,
|
| 279 |
+
"grad_norm": 1.088487035960174,
|
| 280 |
+
"learning_rate": 1.619804874726171e-05,
|
| 281 |
+
"loss": 0.8652,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7358490566037735,
|
| 286 |
+
"grad_norm": 1.0825029747031027,
|
| 287 |
+
"learning_rate": 1.5936309119993333e-05,
|
| 288 |
+
"loss": 0.8565,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7547169811320755,
|
| 293 |
+
"grad_norm": 1.200978191088004,
|
| 294 |
+
"learning_rate": 1.566813254093538e-05,
|
| 295 |
+
"loss": 0.8438,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7735849056603774,
|
| 300 |
+
"grad_norm": 1.0577639466302755,
|
| 301 |
+
"learning_rate": 1.5393809803521213e-05,
|
| 302 |
+
"loss": 0.8681,
|
| 303 |
+
"step": 205
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.7924528301886793,
|
| 307 |
+
"grad_norm": 1.1116902648533118,
|
| 308 |
+
"learning_rate": 1.5113638365682996e-05,
|
| 309 |
+
"loss": 0.8534,
|
| 310 |
+
"step": 210
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.8113207547169812,
|
| 314 |
+
"grad_norm": 1.0957222267960907,
|
| 315 |
+
"learning_rate": 1.482792202730745e-05,
|
| 316 |
+
"loss": 0.8598,
|
| 317 |
+
"step": 215
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8301886792452831,
|
| 321 |
+
"grad_norm": 1.0870428672337984,
|
| 322 |
+
"learning_rate": 1.4536970600814789e-05,
|
| 323 |
+
"loss": 0.8854,
|
| 324 |
+
"step": 220
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.8490566037735849,
|
| 328 |
+
"grad_norm": 1.0340161162672956,
|
| 329 |
+
"learning_rate": 1.424109957521806e-05,
|
| 330 |
+
"loss": 0.865,
|
| 331 |
+
"step": 225
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.8679245283018868,
|
| 335 |
+
"grad_norm": 1.062102724438973,
|
| 336 |
+
"learning_rate": 1.394062977402717e-05,
|
| 337 |
+
"loss": 0.8664,
|
| 338 |
+
"step": 230
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.8867924528301887,
|
| 342 |
+
"grad_norm": 1.0590868294575073,
|
| 343 |
+
"learning_rate": 1.3635887007368467e-05,
|
| 344 |
+
"loss": 0.8419,
|
| 345 |
+
"step": 235
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.9056603773584906,
|
| 349 |
+
"grad_norm": 1.0864493134204916,
|
| 350 |
+
"learning_rate": 1.3327201718697232e-05,
|
| 351 |
+
"loss": 0.8519,
|
| 352 |
+
"step": 240
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.9245283018867925,
|
| 356 |
+
"grad_norm": 1.0168196512793692,
|
| 357 |
+
"learning_rate": 1.3014908626486032e-05,
|
| 358 |
+
"loss": 0.8629,
|
| 359 |
+
"step": 245
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9433962264150944,
|
| 363 |
+
"grad_norm": 1.0404726364027557,
|
| 364 |
+
"learning_rate": 1.2699346361277538e-05,
|
| 365 |
+
"loss": 0.8337,
|
| 366 |
+
"step": 250
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.9622641509433962,
|
| 370 |
+
"grad_norm": 1.018131650974526,
|
| 371 |
+
"learning_rate": 1.2380857098495355e-05,
|
| 372 |
+
"loss": 0.9745,
|
| 373 |
+
"step": 255
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.9811320754716981,
|
| 377 |
+
"grad_norm": 1.0037862694896684,
|
| 378 |
+
"learning_rate": 1.2059786187410984e-05,
|
| 379 |
+
"loss": 0.8712,
|
| 380 |
+
"step": 260
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.0,
|
| 384 |
+
"grad_norm": 0.989425132088932,
|
| 385 |
+
"learning_rate": 1.1736481776669307e-05,
|
| 386 |
+
"loss": 0.8495,
|
| 387 |
+
"step": 265
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.0,
|
| 391 |
+
"eval_loss": 0.8464146852493286,
|
| 392 |
+
"eval_runtime": 4.2653,
|
| 393 |
+
"eval_samples_per_second": 40.091,
|
| 394 |
+
"eval_steps_per_second": 0.703,
|
| 395 |
+
"step": 265
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.0188679245283019,
|
| 399 |
+
"grad_norm": 3.0712777016007875,
|
| 400 |
+
"learning_rate": 1.1411294436778562e-05,
|
| 401 |
+
"loss": 0.5752,
|
| 402 |
+
"step": 270
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.0377358490566038,
|
| 406 |
+
"grad_norm": 1.367908639519453,
|
| 407 |
+
"learning_rate": 1.1084576779974257e-05,
|
| 408 |
+
"loss": 0.5572,
|
| 409 |
+
"step": 275
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.0566037735849056,
|
| 413 |
+
"grad_norm": 1.1027538267727532,
|
| 414 |
+
"learning_rate": 1.0756683077869133e-05,
|
| 415 |
+
"loss": 0.5489,
|
| 416 |
+
"step": 280
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.0754716981132075,
|
| 420 |
+
"grad_norm": 1.1284473887078348,
|
| 421 |
+
"learning_rate": 1.0427968877303809e-05,
|
| 422 |
+
"loss": 0.549,
|
| 423 |
+
"step": 285
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.0943396226415094,
|
| 427 |
+
"grad_norm": 1.0579765238733445,
|
| 428 |
+
"learning_rate": 1.0098790614814658e-05,
|
| 429 |
+
"loss": 0.5584,
|
| 430 |
+
"step": 290
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.1132075471698113,
|
| 434 |
+
"grad_norm": 1.0510922170571144,
|
| 435 |
+
"learning_rate": 9.769505230136962e-06,
|
| 436 |
+
"loss": 0.5326,
|
| 437 |
+
"step": 295
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.1320754716981132,
|
| 441 |
+
"grad_norm": 1.0257584558372113,
|
| 442 |
+
"learning_rate": 9.440469779162407e-06,
|
| 443 |
+
"loss": 0.5457,
|
| 444 |
+
"step": 300
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.150943396226415,
|
| 448 |
+
"grad_norm": 1.0531732710979618,
|
| 449 |
+
"learning_rate": 9.112041046770653e-06,
|
| 450 |
+
"loss": 0.5378,
|
| 451 |
+
"step": 305
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.169811320754717,
|
| 455 |
+
"grad_norm": 1.0708673795759969,
|
| 456 |
+
"learning_rate": 8.784575159954748e-06,
|
| 457 |
+
"loss": 0.5425,
|
| 458 |
+
"step": 310
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.1886792452830188,
|
| 462 |
+
"grad_norm": 1.0132141363080573,
|
| 463 |
+
"learning_rate": 8.458427201659926e-06,
|
| 464 |
+
"loss": 0.5391,
|
| 465 |
+
"step": 315
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.2075471698113207,
|
| 469 |
+
"grad_norm": 1.0680470294860878,
|
| 470 |
+
"learning_rate": 8.133950825754511e-06,
|
| 471 |
+
"loss": 0.5447,
|
| 472 |
+
"step": 320
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.2264150943396226,
|
| 476 |
+
"grad_norm": 0.9666850547234715,
|
| 477 |
+
"learning_rate": 7.81149787355039e-06,
|
| 478 |
+
"loss": 0.5379,
|
| 479 |
+
"step": 325
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.2452830188679245,
|
| 483 |
+
"grad_norm": 0.9710527007406873,
|
| 484 |
+
"learning_rate": 7.491417992288927e-06,
|
| 485 |
+
"loss": 0.5326,
|
| 486 |
+
"step": 330
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.2641509433962264,
|
| 490 |
+
"grad_norm": 1.1620230853698044,
|
| 491 |
+
"learning_rate": 7.174058256006012e-06,
|
| 492 |
+
"loss": 0.5458,
|
| 493 |
+
"step": 335
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.2830188679245282,
|
| 497 |
+
"grad_norm": 1.0264570630272518,
|
| 498 |
+
"learning_rate": 6.859762789187259e-06,
|
| 499 |
+
"loss": 0.5521,
|
| 500 |
+
"step": 340
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.3018867924528301,
|
| 504 |
+
"grad_norm": 1.0303843487410083,
|
| 505 |
+
"learning_rate": 6.548872393621578e-06,
|
| 506 |
+
"loss": 0.5465,
|
| 507 |
+
"step": 345
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.320754716981132,
|
| 511 |
+
"grad_norm": 0.9494639965897911,
|
| 512 |
+
"learning_rate": 6.241724178857621e-06,
|
| 513 |
+
"loss": 0.5531,
|
| 514 |
+
"step": 350
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.3396226415094339,
|
| 518 |
+
"grad_norm": 0.998464274247554,
|
| 519 |
+
"learning_rate": 5.938651196663865e-06,
|
| 520 |
+
"loss": 0.5528,
|
| 521 |
+
"step": 355
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.3584905660377358,
|
| 525 |
+
"grad_norm": 0.9719381109648194,
|
| 526 |
+
"learning_rate": 5.6399820798887266e-06,
|
| 527 |
+
"loss": 0.5351,
|
| 528 |
+
"step": 360
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.3773584905660377,
|
| 532 |
+
"grad_norm": 0.9816121465159701,
|
| 533 |
+
"learning_rate": 5.346040686112189e-06,
|
| 534 |
+
"loss": 0.551,
|
| 535 |
+
"step": 365
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.3962264150943398,
|
| 539 |
+
"grad_norm": 0.9743057773598173,
|
| 540 |
+
"learning_rate": 5.0571457464755226e-06,
|
| 541 |
+
"loss": 0.5345,
|
| 542 |
+
"step": 370
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.4150943396226414,
|
| 546 |
+
"grad_norm": 0.9293281745891194,
|
| 547 |
+
"learning_rate": 4.773610520069706e-06,
|
| 548 |
+
"loss": 0.5455,
|
| 549 |
+
"step": 375
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.4339622641509435,
|
| 553 |
+
"grad_norm": 0.9573637805384148,
|
| 554 |
+
"learning_rate": 4.495742454257418e-06,
|
| 555 |
+
"loss": 0.546,
|
| 556 |
+
"step": 380
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.4528301886792452,
|
| 560 |
+
"grad_norm": 0.9286559221860932,
|
| 561 |
+
"learning_rate": 4.223842851296907e-06,
|
| 562 |
+
"loss": 0.5179,
|
| 563 |
+
"step": 385
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.4716981132075473,
|
| 567 |
+
"grad_norm": 0.9387936040725331,
|
| 568 |
+
"learning_rate": 3.9582065416291926e-06,
|
| 569 |
+
"loss": 0.5561,
|
| 570 |
+
"step": 390
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.490566037735849,
|
| 574 |
+
"grad_norm": 0.9706784805714281,
|
| 575 |
+
"learning_rate": 3.6991215641828903e-06,
|
| 576 |
+
"loss": 0.5471,
|
| 577 |
+
"step": 395
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.509433962264151,
|
| 581 |
+
"grad_norm": 0.998331330776895,
|
| 582 |
+
"learning_rate": 3.4468688540433425e-06,
|
| 583 |
+
"loss": 0.5462,
|
| 584 |
+
"step": 400
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.5283018867924527,
|
| 588 |
+
"grad_norm": 0.9387337728890874,
|
| 589 |
+
"learning_rate": 3.2017219378246734e-06,
|
| 590 |
+
"loss": 0.5658,
|
| 591 |
+
"step": 405
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.5471698113207548,
|
| 595 |
+
"grad_norm": 0.9419101032950977,
|
| 596 |
+
"learning_rate": 2.963946637075107e-06,
|
| 597 |
+
"loss": 0.5416,
|
| 598 |
+
"step": 410
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.5660377358490565,
|
| 602 |
+
"grad_norm": 0.9575101628921765,
|
| 603 |
+
"learning_rate": 2.7338007800372024e-06,
|
| 604 |
+
"loss": 0.5323,
|
| 605 |
+
"step": 415
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.5849056603773586,
|
| 609 |
+
"grad_norm": 1.0034727804533015,
|
| 610 |
+
"learning_rate": 2.5115339220754796e-06,
|
| 611 |
+
"loss": 0.5402,
|
| 612 |
+
"step": 420
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.6037735849056602,
|
| 616 |
+
"grad_norm": 7.5029005576343195,
|
| 617 |
+
"learning_rate": 2.2973870750746253e-06,
|
| 618 |
+
"loss": 0.5655,
|
| 619 |
+
"step": 425
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.6226415094339623,
|
| 623 |
+
"grad_norm": 1.0027923672255161,
|
| 624 |
+
"learning_rate": 2.09159244610172e-06,
|
| 625 |
+
"loss": 0.5542,
|
| 626 |
+
"step": 430
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.641509433962264,
|
| 630 |
+
"grad_norm": 0.9357231473507377,
|
| 631 |
+
"learning_rate": 1.8943731856158299e-06,
|
| 632 |
+
"loss": 0.6383,
|
| 633 |
+
"step": 435
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.6603773584905661,
|
| 637 |
+
"grad_norm": 0.9686189698665412,
|
| 638 |
+
"learning_rate": 1.7059431454979825e-06,
|
| 639 |
+
"loss": 0.5562,
|
| 640 |
+
"step": 440
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.6792452830188678,
|
| 644 |
+
"grad_norm": 0.9045883081564006,
|
| 645 |
+
"learning_rate": 1.5265066471639701e-06,
|
| 646 |
+
"loss": 0.5348,
|
| 647 |
+
"step": 445
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.6981132075471699,
|
| 651 |
+
"grad_norm": 1.0056737797057134,
|
| 652 |
+
"learning_rate": 1.3562582600113295e-06,
|
| 653 |
+
"loss": 0.5352,
|
| 654 |
+
"step": 450
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.7169811320754715,
|
| 658 |
+
"grad_norm": 0.9512535781198771,
|
| 659 |
+
"learning_rate": 1.1953825904408033e-06,
|
| 660 |
+
"loss": 0.5464,
|
| 661 |
+
"step": 455
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.7358490566037736,
|
| 665 |
+
"grad_norm": 0.9930921183835165,
|
| 666 |
+
"learning_rate": 1.0440540816810395e-06,
|
| 667 |
+
"loss": 0.561,
|
| 668 |
+
"step": 460
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.7547169811320755,
|
| 672 |
+
"grad_norm": 0.9410700940787784,
|
| 673 |
+
"learning_rate": 9.024368246335735e-07,
|
| 674 |
+
"loss": 0.5456,
|
| 675 |
+
"step": 465
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.7735849056603774,
|
| 679 |
+
"grad_norm": 0.9323535925403078,
|
| 680 |
+
"learning_rate": 7.706843799431985e-07,
|
| 681 |
+
"loss": 0.5243,
|
| 682 |
+
"step": 470
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.7924528301886793,
|
| 686 |
+
"grad_norm": 0.9689906420605915,
|
| 687 |
+
"learning_rate": 6.489396114866942e-07,
|
| 688 |
+
"loss": 0.5296,
|
| 689 |
+
"step": 475
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.8113207547169812,
|
| 693 |
+
"grad_norm": 1.2229314873705728,
|
| 694 |
+
"learning_rate": 5.373345314604206e-07,
|
| 695 |
+
"loss": 0.53,
|
| 696 |
+
"step": 480
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.830188679245283,
|
| 700 |
+
"grad_norm": 0.9545280104598107,
|
| 701 |
+
"learning_rate": 4.359901572347758e-07,
|
| 702 |
+
"loss": 0.5031,
|
| 703 |
+
"step": 485
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.849056603773585,
|
| 707 |
+
"grad_norm": 0.9636644642636012,
|
| 708 |
+
"learning_rate": 3.450163801307582e-07,
|
| 709 |
+
"loss": 0.5159,
|
| 710 |
+
"step": 490
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.8679245283018868,
|
| 714 |
+
"grad_norm": 0.9314616621215778,
|
| 715 |
+
"learning_rate": 2.6451184626087646e-07,
|
| 716 |
+
"loss": 0.5212,
|
| 717 |
+
"step": 495
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.8867924528301887,
|
| 721 |
+
"grad_norm": 0.9019228681615893,
|
| 722 |
+
"learning_rate": 1.9456384956365149e-07,
|
| 723 |
+
"loss": 0.539,
|
| 724 |
+
"step": 500
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.9056603773584906,
|
| 728 |
+
"grad_norm": 0.8782721096424342,
|
| 729 |
+
"learning_rate": 1.3524823714768375e-07,
|
| 730 |
+
"loss": 0.5462,
|
| 731 |
+
"step": 505
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.9245283018867925,
|
| 735 |
+
"grad_norm": 0.9377029041620999,
|
| 736 |
+
"learning_rate": 8.662932704792793e-08,
|
| 737 |
+
"loss": 0.5352,
|
| 738 |
+
"step": 510
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.9433962264150944,
|
| 742 |
+
"grad_norm": 0.9313929690259365,
|
| 743 |
+
"learning_rate": 4.8759838483358745e-08,
|
| 744 |
+
"loss": 0.5129,
|
| 745 |
+
"step": 515
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.9622641509433962,
|
| 749 |
+
"grad_norm": 0.9729472297721098,
|
| 750 |
+
"learning_rate": 2.1680834691628627e-08,
|
| 751 |
+
"loss": 0.5358,
|
| 752 |
+
"step": 520
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.9811320754716981,
|
| 756 |
+
"grad_norm": 0.9341412228490686,
|
| 757 |
+
"learning_rate": 5.421678402741659e-09,
|
| 758 |
+
"loss": 0.5717,
|
| 759 |
+
"step": 525
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 2.0,
|
| 763 |
+
"grad_norm": 0.945793151587699,
|
| 764 |
+
"learning_rate": 0.0,
|
| 765 |
+
"loss": 0.5006,
|
| 766 |
+
"step": 530
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 2.0,
|
| 770 |
+
"eval_loss": 0.852841854095459,
|
| 771 |
+
"eval_runtime": 6.8692,
|
| 772 |
+
"eval_samples_per_second": 24.894,
|
| 773 |
+
"eval_steps_per_second": 0.437,
|
| 774 |
+
"step": 530
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 2.0,
|
| 778 |
+
"step": 530,
|
| 779 |
+
"total_flos": 110971217510400.0,
|
| 780 |
+
"train_loss": 0.7231714307137256,
|
| 781 |
+
"train_runtime": 3756.5168,
|
| 782 |
+
"train_samples_per_second": 9.007,
|
| 783 |
+
"train_steps_per_second": 0.141
|
| 784 |
+
}
|
| 785 |
+
],
|
| 786 |
+
"logging_steps": 5,
|
| 787 |
+
"max_steps": 530,
|
| 788 |
+
"num_input_tokens_seen": 0,
|
| 789 |
+
"num_train_epochs": 2,
|
| 790 |
+
"save_steps": 500,
|
| 791 |
+
"stateful_callbacks": {
|
| 792 |
+
"TrainerControl": {
|
| 793 |
+
"args": {
|
| 794 |
+
"should_epoch_stop": false,
|
| 795 |
+
"should_evaluate": false,
|
| 796 |
+
"should_log": false,
|
| 797 |
+
"should_save": true,
|
| 798 |
+
"should_training_stop": true
|
| 799 |
+
},
|
| 800 |
+
"attributes": {}
|
| 801 |
+
}
|
| 802 |
+
},
|
| 803 |
+
"total_flos": 110971217510400.0,
|
| 804 |
+
"train_batch_size": 8,
|
| 805 |
+
"trial_name": null,
|
| 806 |
+
"trial_params": null
|
| 807 |
+
}
|