Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +751 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,751 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 492,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0040650406504065045,
|
| 13 |
+
"grad_norm": 16.96013186396094,
|
| 14 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 15 |
+
"loss": 1.5409,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02032520325203252,
|
| 20 |
+
"grad_norm": 16.916170957900782,
|
| 21 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 22 |
+
"loss": 1.5345,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04065040650406504,
|
| 27 |
+
"grad_norm": 4.992436085095587,
|
| 28 |
+
"learning_rate": 4.000000000000001e-06,
|
| 29 |
+
"loss": 1.3026,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06097560975609756,
|
| 34 |
+
"grad_norm": 2.654639415277273,
|
| 35 |
+
"learning_rate": 6e-06,
|
| 36 |
+
"loss": 1.2217,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08130081300813008,
|
| 41 |
+
"grad_norm": 2.2317048430147577,
|
| 42 |
+
"learning_rate": 8.000000000000001e-06,
|
| 43 |
+
"loss": 1.162,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.1016260162601626,
|
| 48 |
+
"grad_norm": 1.3761478430076108,
|
| 49 |
+
"learning_rate": 1e-05,
|
| 50 |
+
"loss": 1.124,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12195121951219512,
|
| 55 |
+
"grad_norm": 1.1540940483296693,
|
| 56 |
+
"learning_rate": 1.2e-05,
|
| 57 |
+
"loss": 1.1148,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14227642276422764,
|
| 62 |
+
"grad_norm": 1.1928473894163938,
|
| 63 |
+
"learning_rate": 1.4e-05,
|
| 64 |
+
"loss": 1.1046,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.16260162601626016,
|
| 69 |
+
"grad_norm": 1.185131002730624,
|
| 70 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 71 |
+
"loss": 1.1034,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.18292682926829268,
|
| 76 |
+
"grad_norm": 1.3370638997381883,
|
| 77 |
+
"learning_rate": 1.8e-05,
|
| 78 |
+
"loss": 1.0685,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.2032520325203252,
|
| 83 |
+
"grad_norm": 1.4250484136195543,
|
| 84 |
+
"learning_rate": 2e-05,
|
| 85 |
+
"loss": 1.0857,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.22357723577235772,
|
| 90 |
+
"grad_norm": 1.3017876025421216,
|
| 91 |
+
"learning_rate": 1.9993685782892184e-05,
|
| 92 |
+
"loss": 1.0214,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.24390243902439024,
|
| 97 |
+
"grad_norm": 1.486201507551871,
|
| 98 |
+
"learning_rate": 1.9974751105436266e-05,
|
| 99 |
+
"loss": 1.0418,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.26422764227642276,
|
| 104 |
+
"grad_norm": 1.2014188024997923,
|
| 105 |
+
"learning_rate": 1.9943219879165113e-05,
|
| 106 |
+
"loss": 1.0753,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.2845528455284553,
|
| 111 |
+
"grad_norm": 1.3819404641565476,
|
| 112 |
+
"learning_rate": 1.98991319230804e-05,
|
| 113 |
+
"loss": 1.0214,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.3048780487804878,
|
| 118 |
+
"grad_norm": 1.1959378462736852,
|
| 119 |
+
"learning_rate": 1.984254291336743e-05,
|
| 120 |
+
"loss": 1.0168,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3252032520325203,
|
| 125 |
+
"grad_norm": 1.183753776787632,
|
| 126 |
+
"learning_rate": 1.9773524313084857e-05,
|
| 127 |
+
"loss": 1.0293,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.34552845528455284,
|
| 132 |
+
"grad_norm": 1.2267547234529148,
|
| 133 |
+
"learning_rate": 1.9692163281918016e-05,
|
| 134 |
+
"loss": 1.0394,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.36585365853658536,
|
| 139 |
+
"grad_norm": 1.1369844018691624,
|
| 140 |
+
"learning_rate": 1.959856256610988e-05,
|
| 141 |
+
"loss": 1.0491,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3861788617886179,
|
| 146 |
+
"grad_norm": 1.1999534529754976,
|
| 147 |
+
"learning_rate": 1.9492840368708668e-05,
|
| 148 |
+
"loss": 1.0136,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.4065040650406504,
|
| 153 |
+
"grad_norm": 1.1735600472638217,
|
| 154 |
+
"learning_rate": 1.937513020029588e-05,
|
| 155 |
+
"loss": 1.0361,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.4268292682926829,
|
| 160 |
+
"grad_norm": 1.250115647846771,
|
| 161 |
+
"learning_rate": 1.9245580710383344e-05,
|
| 162 |
+
"loss": 1.0283,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.44715447154471544,
|
| 167 |
+
"grad_norm": 1.1840318671469985,
|
| 168 |
+
"learning_rate": 1.9104355499692166e-05,
|
| 169 |
+
"loss": 1.0483,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.46747967479674796,
|
| 174 |
+
"grad_norm": 1.1355456855985984,
|
| 175 |
+
"learning_rate": 1.8951632913550625e-05,
|
| 176 |
+
"loss": 1.0076,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4878048780487805,
|
| 181 |
+
"grad_norm": 1.176376978993297,
|
| 182 |
+
"learning_rate": 1.8787605816671956e-05,
|
| 183 |
+
"loss": 1.0451,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.508130081300813,
|
| 188 |
+
"grad_norm": 1.2157595019570657,
|
| 189 |
+
"learning_rate": 1.8612481349596406e-05,
|
| 190 |
+
"loss": 0.9807,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.5284552845528455,
|
| 195 |
+
"grad_norm": 1.617382778432822,
|
| 196 |
+
"learning_rate": 1.8426480667105178e-05,
|
| 197 |
+
"loss": 1.0273,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5487804878048781,
|
| 202 |
+
"grad_norm": 1.072969358036799,
|
| 203 |
+
"learning_rate": 1.8229838658936566e-05,
|
| 204 |
+
"loss": 1.0239,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5691056910569106,
|
| 209 |
+
"grad_norm": 1.1349667239072485,
|
| 210 |
+
"learning_rate": 1.8022803653156983e-05,
|
| 211 |
+
"loss": 1.0191,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5894308943089431,
|
| 216 |
+
"grad_norm": 1.061992535593682,
|
| 217 |
+
"learning_rate": 1.7805637102561516e-05,
|
| 218 |
+
"loss": 1.004,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.6097560975609756,
|
| 223 |
+
"grad_norm": 1.1127139653197202,
|
| 224 |
+
"learning_rate": 1.757861325449997e-05,
|
| 225 |
+
"loss": 0.9924,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.6300813008130082,
|
| 230 |
+
"grad_norm": 1.1022251104365055,
|
| 231 |
+
"learning_rate": 1.73420188045454e-05,
|
| 232 |
+
"loss": 0.9971,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.6504065040650406,
|
| 237 |
+
"grad_norm": 1.273269943630818,
|
| 238 |
+
"learning_rate": 1.7096152534442515e-05,
|
| 239 |
+
"loss": 1.0056,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6707317073170732,
|
| 244 |
+
"grad_norm": 1.0563584964584172,
|
| 245 |
+
"learning_rate": 1.6841324934793096e-05,
|
| 246 |
+
"loss": 0.9856,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6910569105691057,
|
| 251 |
+
"grad_norm": 1.145292649525439,
|
| 252 |
+
"learning_rate": 1.6577857812954994e-05,
|
| 253 |
+
"loss": 1.0083,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.7113821138211383,
|
| 258 |
+
"grad_norm": 1.0928885418057457,
|
| 259 |
+
"learning_rate": 1.6306083886649823e-05,
|
| 260 |
+
"loss": 0.9936,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.7317073170731707,
|
| 265 |
+
"grad_norm": 1.13829615874657,
|
| 266 |
+
"learning_rate": 1.6026346363792565e-05,
|
| 267 |
+
"loss": 0.9874,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.7520325203252033,
|
| 272 |
+
"grad_norm": 1.0862202261960012,
|
| 273 |
+
"learning_rate": 1.573899850907373e-05,
|
| 274 |
+
"loss": 0.9856,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.7723577235772358,
|
| 279 |
+
"grad_norm": 1.0544760585534974,
|
| 280 |
+
"learning_rate": 1.5444403197841345e-05,
|
| 281 |
+
"loss": 1.0079,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7926829268292683,
|
| 286 |
+
"grad_norm": 1.1972116619581246,
|
| 287 |
+
"learning_rate": 1.514293245784623e-05,
|
| 288 |
+
"loss": 0.9818,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.8130081300813008,
|
| 293 |
+
"grad_norm": 1.0551988260235088,
|
| 294 |
+
"learning_rate": 1.4834966999429179e-05,
|
| 295 |
+
"loss": 0.9762,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.8333333333333334,
|
| 300 |
+
"grad_norm": 1.0937895071472443,
|
| 301 |
+
"learning_rate": 1.4520895734743419e-05,
|
| 302 |
+
"loss": 0.9967,
|
| 303 |
+
"step": 205
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.8536585365853658,
|
| 307 |
+
"grad_norm": 2.1677819071037283,
|
| 308 |
+
"learning_rate": 1.4201115286619464e-05,
|
| 309 |
+
"loss": 0.9642,
|
| 310 |
+
"step": 210
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.8739837398373984,
|
| 314 |
+
"grad_norm": 1.0198723964107377,
|
| 315 |
+
"learning_rate": 1.387602948769257e-05,
|
| 316 |
+
"loss": 0.9717,
|
| 317 |
+
"step": 215
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8943089430894309,
|
| 321 |
+
"grad_norm": 1.0362342965679063,
|
| 322 |
+
"learning_rate": 1.3546048870425356e-05,
|
| 323 |
+
"loss": 0.9714,
|
| 324 |
+
"step": 220
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.9146341463414634,
|
| 328 |
+
"grad_norm": 0.997164008308899,
|
| 329 |
+
"learning_rate": 1.3211590148669586e-05,
|
| 330 |
+
"loss": 0.9713,
|
| 331 |
+
"step": 225
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.9349593495934959,
|
| 335 |
+
"grad_norm": 1.1623269423245046,
|
| 336 |
+
"learning_rate": 1.2873075691421808e-05,
|
| 337 |
+
"loss": 0.9454,
|
| 338 |
+
"step": 230
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.9552845528455285,
|
| 342 |
+
"grad_norm": 1.0749975590040124,
|
| 343 |
+
"learning_rate": 1.2530932989437463e-05,
|
| 344 |
+
"loss": 0.9845,
|
| 345 |
+
"step": 235
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.975609756097561,
|
| 349 |
+
"grad_norm": 0.988624506908827,
|
| 350 |
+
"learning_rate": 1.2185594115376991e-05,
|
| 351 |
+
"loss": 0.9547,
|
| 352 |
+
"step": 240
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.9959349593495935,
|
| 356 |
+
"grad_norm": 1.0165003436015188,
|
| 357 |
+
"learning_rate": 1.1837495178165706e-05,
|
| 358 |
+
"loss": 0.9962,
|
| 359 |
+
"step": 245
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.0,
|
| 363 |
+
"eval_loss": 0.968990683555603,
|
| 364 |
+
"eval_runtime": 4.2249,
|
| 365 |
+
"eval_samples_per_second": 37.397,
|
| 366 |
+
"eval_steps_per_second": 0.71,
|
| 367 |
+
"step": 246
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 1.016260162601626,
|
| 371 |
+
"grad_norm": 2.2206925031006675,
|
| 372 |
+
"learning_rate": 1.1487075772256517e-05,
|
| 373 |
+
"loss": 0.7411,
|
| 374 |
+
"step": 250
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.0365853658536586,
|
| 378 |
+
"grad_norm": 1.5057083208667505,
|
| 379 |
+
"learning_rate": 1.1134778422490971e-05,
|
| 380 |
+
"loss": 0.6718,
|
| 381 |
+
"step": 255
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.056910569105691,
|
| 385 |
+
"grad_norm": 1.1997738500067352,
|
| 386 |
+
"learning_rate": 1.0781048025259648e-05,
|
| 387 |
+
"loss": 0.6885,
|
| 388 |
+
"step": 260
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.0772357723577235,
|
| 392 |
+
"grad_norm": 1.129770949868323,
|
| 393 |
+
"learning_rate": 1.0426331286667701e-05,
|
| 394 |
+
"loss": 0.6459,
|
| 395 |
+
"step": 265
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.0975609756097562,
|
| 399 |
+
"grad_norm": 1.0821575635964367,
|
| 400 |
+
"learning_rate": 1.0071076158414977e-05,
|
| 401 |
+
"loss": 0.6578,
|
| 402 |
+
"step": 270
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.1178861788617886,
|
| 406 |
+
"grad_norm": 0.9858678711249164,
|
| 407 |
+
"learning_rate": 9.715731272103172e-06,
|
| 408 |
+
"loss": 0.6665,
|
| 409 |
+
"step": 275
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.1382113821138211,
|
| 413 |
+
"grad_norm": 1.0404542015906177,
|
| 414 |
+
"learning_rate": 9.360745372684346e-06,
|
| 415 |
+
"loss": 0.6682,
|
| 416 |
+
"step": 280
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.1585365853658536,
|
| 420 |
+
"grad_norm": 1.0009568415817596,
|
| 421 |
+
"learning_rate": 9.00656675176633e-06,
|
| 422 |
+
"loss": 0.6557,
|
| 423 |
+
"step": 285
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.1788617886178863,
|
| 427 |
+
"grad_norm": 1.0425691363766654,
|
| 428 |
+
"learning_rate": 8.653642681490608e-06,
|
| 429 |
+
"loss": 0.6632,
|
| 430 |
+
"step": 290
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.1991869918699187,
|
| 434 |
+
"grad_norm": 0.9980360375332701,
|
| 435 |
+
"learning_rate": 8.30241884969764e-06,
|
| 436 |
+
"loss": 0.6557,
|
| 437 |
+
"step": 295
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.2195121951219512,
|
| 441 |
+
"grad_norm": 1.1151520476299475,
|
| 442 |
+
"learning_rate": 7.953338797092902e-06,
|
| 443 |
+
"loss": 0.676,
|
| 444 |
+
"step": 300
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.2398373983739837,
|
| 448 |
+
"grad_norm": 1.1095243103350756,
|
| 449 |
+
"learning_rate": 7.606843357124426e-06,
|
| 450 |
+
"loss": 0.6514,
|
| 451 |
+
"step": 305
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.2601626016260163,
|
| 455 |
+
"grad_norm": 1.0773726823725258,
|
| 456 |
+
"learning_rate": 7.263370099279173e-06,
|
| 457 |
+
"loss": 0.6592,
|
| 458 |
+
"step": 310
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.2804878048780488,
|
| 462 |
+
"grad_norm": 1.0415539344975482,
|
| 463 |
+
"learning_rate": 6.923352776501302e-06,
|
| 464 |
+
"loss": 0.663,
|
| 465 |
+
"step": 315
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.3008130081300813,
|
| 469 |
+
"grad_norm": 0.9921263103830474,
|
| 470 |
+
"learning_rate": 6.587220777430097e-06,
|
| 471 |
+
"loss": 0.6554,
|
| 472 |
+
"step": 320
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.321138211382114,
|
| 476 |
+
"grad_norm": 1.0067612241699029,
|
| 477 |
+
"learning_rate": 6.255398584149366e-06,
|
| 478 |
+
"loss": 0.6592,
|
| 479 |
+
"step": 325
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.3414634146341464,
|
| 483 |
+
"grad_norm": 0.9539630263562321,
|
| 484 |
+
"learning_rate": 5.928305236133016e-06,
|
| 485 |
+
"loss": 0.638,
|
| 486 |
+
"step": 330
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.3617886178861789,
|
| 490 |
+
"grad_norm": 0.9724353060112348,
|
| 491 |
+
"learning_rate": 5.60635380106383e-06,
|
| 492 |
+
"loss": 0.6664,
|
| 493 |
+
"step": 335
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.3821138211382114,
|
| 497 |
+
"grad_norm": 0.9667751887299418,
|
| 498 |
+
"learning_rate": 5.2899508531936526e-06,
|
| 499 |
+
"loss": 0.6582,
|
| 500 |
+
"step": 340
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.4024390243902438,
|
| 504 |
+
"grad_norm": 1.0179726164365328,
|
| 505 |
+
"learning_rate": 4.979495959903759e-06,
|
| 506 |
+
"loss": 0.658,
|
| 507 |
+
"step": 345
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.4227642276422765,
|
| 511 |
+
"grad_norm": 0.9892639534813169,
|
| 512 |
+
"learning_rate": 4.675381177113837e-06,
|
| 513 |
+
"loss": 0.6588,
|
| 514 |
+
"step": 350
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.443089430894309,
|
| 518 |
+
"grad_norm": 0.9663435093514263,
|
| 519 |
+
"learning_rate": 4.377990554176729e-06,
|
| 520 |
+
"loss": 0.6295,
|
| 521 |
+
"step": 355
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.4634146341463414,
|
| 525 |
+
"grad_norm": 0.9475872699295589,
|
| 526 |
+
"learning_rate": 4.087699648884248e-06,
|
| 527 |
+
"loss": 0.6309,
|
| 528 |
+
"step": 360
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.4837398373983741,
|
| 532 |
+
"grad_norm": 1.0174077533995098,
|
| 533 |
+
"learning_rate": 3.804875053196477e-06,
|
| 534 |
+
"loss": 0.6548,
|
| 535 |
+
"step": 365
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.5040650406504064,
|
| 539 |
+
"grad_norm": 0.9473261301601033,
|
| 540 |
+
"learning_rate": 3.529873930293546e-06,
|
| 541 |
+
"loss": 0.6585,
|
| 542 |
+
"step": 370
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.524390243902439,
|
| 546 |
+
"grad_norm": 0.9724565020578093,
|
| 547 |
+
"learning_rate": 3.2630435635344283e-06,
|
| 548 |
+
"loss": 0.6579,
|
| 549 |
+
"step": 375
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.5447154471544715,
|
| 553 |
+
"grad_norm": 0.9377507057850171,
|
| 554 |
+
"learning_rate": 3.004720917892464e-06,
|
| 555 |
+
"loss": 0.6325,
|
| 556 |
+
"step": 380
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.565040650406504,
|
| 560 |
+
"grad_norm": 1.220689589737471,
|
| 561 |
+
"learning_rate": 2.7552322144213405e-06,
|
| 562 |
+
"loss": 0.6656,
|
| 563 |
+
"step": 385
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.5853658536585367,
|
| 567 |
+
"grad_norm": 0.9445474318613094,
|
| 568 |
+
"learning_rate": 2.514892518288988e-06,
|
| 569 |
+
"loss": 0.6484,
|
| 570 |
+
"step": 390
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.6056910569105691,
|
| 574 |
+
"grad_norm": 0.915729013522606,
|
| 575 |
+
"learning_rate": 2.2840053408996154e-06,
|
| 576 |
+
"loss": 0.6611,
|
| 577 |
+
"step": 395
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.6260162601626016,
|
| 581 |
+
"grad_norm": 0.9261545603143331,
|
| 582 |
+
"learning_rate": 2.0628622566063063e-06,
|
| 583 |
+
"loss": 0.6366,
|
| 584 |
+
"step": 400
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.6463414634146343,
|
| 588 |
+
"grad_norm": 0.9879532897121964,
|
| 589 |
+
"learning_rate": 1.8517425344982831e-06,
|
| 590 |
+
"loss": 0.6349,
|
| 591 |
+
"step": 405
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.6666666666666665,
|
| 595 |
+
"grad_norm": 0.9360662238523524,
|
| 596 |
+
"learning_rate": 1.6509127857277784e-06,
|
| 597 |
+
"loss": 0.654,
|
| 598 |
+
"step": 410
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.6869918699186992,
|
| 602 |
+
"grad_norm": 0.9548880407020373,
|
| 603 |
+
"learning_rate": 1.4606266268218783e-06,
|
| 604 |
+
"loss": 0.6625,
|
| 605 |
+
"step": 415
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.7073170731707317,
|
| 609 |
+
"grad_norm": 0.9497166098139703,
|
| 610 |
+
"learning_rate": 1.2811243594045697e-06,
|
| 611 |
+
"loss": 0.6593,
|
| 612 |
+
"step": 420
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.7276422764227641,
|
| 616 |
+
"grad_norm": 0.937994818106022,
|
| 617 |
+
"learning_rate": 1.1126326667334196e-06,
|
| 618 |
+
"loss": 0.657,
|
| 619 |
+
"step": 425
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.7479674796747968,
|
| 623 |
+
"grad_norm": 0.9426215224621286,
|
| 624 |
+
"learning_rate": 9.55364327434105e-07,
|
| 625 |
+
"loss": 0.6572,
|
| 626 |
+
"step": 430
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.7682926829268293,
|
| 630 |
+
"grad_norm": 0.956528393469926,
|
| 631 |
+
"learning_rate": 8.095179467943293e-07,
|
| 632 |
+
"loss": 0.6538,
|
| 633 |
+
"step": 435
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.7886178861788617,
|
| 637 |
+
"grad_norm": 0.9816007976764832,
|
| 638 |
+
"learning_rate": 6.752777059564431e-07,
|
| 639 |
+
"loss": 0.6683,
|
| 640 |
+
"step": 440
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.8089430894308944,
|
| 644 |
+
"grad_norm": 0.9327650063702297,
|
| 645 |
+
"learning_rate": 5.528131293254957e-07,
|
| 646 |
+
"loss": 0.6204,
|
| 647 |
+
"step": 445
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.8292682926829267,
|
| 651 |
+
"grad_norm": 0.9411103321287626,
|
| 652 |
+
"learning_rate": 4.4227887048646335e-07,
|
| 653 |
+
"loss": 0.6595,
|
| 654 |
+
"step": 450
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.8495934959349594,
|
| 658 |
+
"grad_norm": 0.9788741395392418,
|
| 659 |
+
"learning_rate": 3.4381451690097653e-07,
|
| 660 |
+
"loss": 0.6425,
|
| 661 |
+
"step": 455
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.8699186991869918,
|
| 665 |
+
"grad_norm": 0.9336996330937468,
|
| 666 |
+
"learning_rate": 2.5754441363021854e-07,
|
| 667 |
+
"loss": 0.6345,
|
| 668 |
+
"step": 460
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.8902439024390243,
|
| 672 |
+
"grad_norm": 0.9453351615169212,
|
| 673 |
+
"learning_rate": 1.8357750630658367e-07,
|
| 674 |
+
"loss": 0.6489,
|
| 675 |
+
"step": 465
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.910569105691057,
|
| 679 |
+
"grad_norm": 0.9708668647842268,
|
| 680 |
+
"learning_rate": 1.2200720355239893e-07,
|
| 681 |
+
"loss": 0.6509,
|
| 682 |
+
"step": 470
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.9308943089430894,
|
| 686 |
+
"grad_norm": 0.984113893821567,
|
| 687 |
+
"learning_rate": 7.291125901946027e-08,
|
| 688 |
+
"loss": 0.6546,
|
| 689 |
+
"step": 475
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.951219512195122,
|
| 693 |
+
"grad_norm": 0.9685414915217074,
|
| 694 |
+
"learning_rate": 3.635167319834709e-08,
|
| 695 |
+
"loss": 0.6384,
|
| 696 |
+
"step": 480
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.9715447154471546,
|
| 700 |
+
"grad_norm": 0.9169491016884999,
|
| 701 |
+
"learning_rate": 1.2374615121508726e-08,
|
| 702 |
+
"loss": 0.6367,
|
| 703 |
+
"step": 485
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.9918699186991868,
|
| 707 |
+
"grad_norm": 0.9084443808404363,
|
| 708 |
+
"learning_rate": 1.0103640590064524e-09,
|
| 709 |
+
"loss": 0.6291,
|
| 710 |
+
"step": 490
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 2.0,
|
| 714 |
+
"eval_loss": 0.9818947911262512,
|
| 715 |
+
"eval_runtime": 6.2779,
|
| 716 |
+
"eval_samples_per_second": 25.167,
|
| 717 |
+
"eval_steps_per_second": 0.478,
|
| 718 |
+
"step": 492
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 2.0,
|
| 722 |
+
"step": 492,
|
| 723 |
+
"total_flos": 103014790594560.0,
|
| 724 |
+
"train_loss": 0.8469056426509609,
|
| 725 |
+
"train_runtime": 3499.6534,
|
| 726 |
+
"train_samples_per_second": 8.98,
|
| 727 |
+
"train_steps_per_second": 0.141
|
| 728 |
+
}
|
| 729 |
+
],
|
| 730 |
+
"logging_steps": 5,
|
| 731 |
+
"max_steps": 492,
|
| 732 |
+
"num_input_tokens_seen": 0,
|
| 733 |
+
"num_train_epochs": 2,
|
| 734 |
+
"save_steps": 500,
|
| 735 |
+
"stateful_callbacks": {
|
| 736 |
+
"TrainerControl": {
|
| 737 |
+
"args": {
|
| 738 |
+
"should_epoch_stop": false,
|
| 739 |
+
"should_evaluate": false,
|
| 740 |
+
"should_log": false,
|
| 741 |
+
"should_save": true,
|
| 742 |
+
"should_training_stop": true
|
| 743 |
+
},
|
| 744 |
+
"attributes": {}
|
| 745 |
+
}
|
| 746 |
+
},
|
| 747 |
+
"total_flos": 103014790594560.0,
|
| 748 |
+
"train_batch_size": 8,
|
| 749 |
+
"trial_name": null,
|
| 750 |
+
"trial_params": null
|
| 751 |
+
}
|