Upload 10 files
Browse files- trainer_state.json +691 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,691 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 47226,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.010587388303053403,
|
| 13 |
+
"grad_norm": 1.734375,
|
| 14 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 15 |
+
"loss": 7.2559,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.021174776606106806,
|
| 20 |
+
"grad_norm": 0.875,
|
| 21 |
+
"learning_rate": 4.000000000000001e-06,
|
| 22 |
+
"loss": 0.5321,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03176216490916021,
|
| 27 |
+
"grad_norm": 0.9296875,
|
| 28 |
+
"learning_rate": 6e-06,
|
| 29 |
+
"loss": 0.4789,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.04234955321221361,
|
| 34 |
+
"grad_norm": 2.5625,
|
| 35 |
+
"learning_rate": 8.000000000000001e-06,
|
| 36 |
+
"loss": 0.4361,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05293694151526701,
|
| 41 |
+
"grad_norm": 0.66796875,
|
| 42 |
+
"learning_rate": 1e-05,
|
| 43 |
+
"loss": 0.4095,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06352432981832042,
|
| 48 |
+
"grad_norm": 0.58984375,
|
| 49 |
+
"learning_rate": 1.2e-05,
|
| 50 |
+
"loss": 0.3739,
|
| 51 |
+
"step": 3000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07411171812137382,
|
| 55 |
+
"grad_norm": 0.76171875,
|
| 56 |
+
"learning_rate": 1.4e-05,
|
| 57 |
+
"loss": 0.3593,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.08469910642442723,
|
| 62 |
+
"grad_norm": 1.0703125,
|
| 63 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 64 |
+
"loss": 0.3277,
|
| 65 |
+
"step": 4000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09528649472748063,
|
| 69 |
+
"grad_norm": 0.83984375,
|
| 70 |
+
"learning_rate": 1.8e-05,
|
| 71 |
+
"loss": 0.3148,
|
| 72 |
+
"step": 4500
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.10587388303053402,
|
| 76 |
+
"grad_norm": 0.84765625,
|
| 77 |
+
"learning_rate": 2e-05,
|
| 78 |
+
"loss": 0.2913,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.11646127133358743,
|
| 83 |
+
"grad_norm": 0.6328125,
|
| 84 |
+
"learning_rate": 1.9763179083976696e-05,
|
| 85 |
+
"loss": 0.2715,
|
| 86 |
+
"step": 5500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.12704865963664083,
|
| 90 |
+
"grad_norm": 0.71484375,
|
| 91 |
+
"learning_rate": 1.9526358167953394e-05,
|
| 92 |
+
"loss": 0.2621,
|
| 93 |
+
"step": 6000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.13763604793969422,
|
| 97 |
+
"grad_norm": 0.80859375,
|
| 98 |
+
"learning_rate": 1.9289537251930092e-05,
|
| 99 |
+
"loss": 0.2484,
|
| 100 |
+
"step": 6500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.14822343624274764,
|
| 104 |
+
"grad_norm": 0.57421875,
|
| 105 |
+
"learning_rate": 1.905271633590679e-05,
|
| 106 |
+
"loss": 0.2359,
|
| 107 |
+
"step": 7000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.15881082454580103,
|
| 111 |
+
"grad_norm": 0.90625,
|
| 112 |
+
"learning_rate": 1.8815895419883485e-05,
|
| 113 |
+
"loss": 0.2284,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.16939821284885445,
|
| 118 |
+
"grad_norm": 0.9921875,
|
| 119 |
+
"learning_rate": 1.8579074503860183e-05,
|
| 120 |
+
"loss": 0.229,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.17998560115190784,
|
| 125 |
+
"grad_norm": 0.875,
|
| 126 |
+
"learning_rate": 1.834225358783688e-05,
|
| 127 |
+
"loss": 0.2303,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.19057298945496126,
|
| 132 |
+
"grad_norm": 1.15625,
|
| 133 |
+
"learning_rate": 1.8105432671813576e-05,
|
| 134 |
+
"loss": 0.2227,
|
| 135 |
+
"step": 9000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.20116037775801465,
|
| 139 |
+
"grad_norm": 0.515625,
|
| 140 |
+
"learning_rate": 1.7868611755790274e-05,
|
| 141 |
+
"loss": 0.216,
|
| 142 |
+
"step": 9500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.21174776606106804,
|
| 146 |
+
"grad_norm": 0.6875,
|
| 147 |
+
"learning_rate": 1.763179083976697e-05,
|
| 148 |
+
"loss": 0.2148,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.22233515436412146,
|
| 153 |
+
"grad_norm": 0.703125,
|
| 154 |
+
"learning_rate": 1.7394969923743667e-05,
|
| 155 |
+
"loss": 0.2106,
|
| 156 |
+
"step": 10500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.23292254266717485,
|
| 160 |
+
"grad_norm": 0.59375,
|
| 161 |
+
"learning_rate": 1.7158149007720365e-05,
|
| 162 |
+
"loss": 0.205,
|
| 163 |
+
"step": 11000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.24350993097022827,
|
| 167 |
+
"grad_norm": 1.0703125,
|
| 168 |
+
"learning_rate": 1.692132809169706e-05,
|
| 169 |
+
"loss": 0.211,
|
| 170 |
+
"step": 11500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.25409731927328166,
|
| 174 |
+
"grad_norm": 1.015625,
|
| 175 |
+
"learning_rate": 1.6684507175673758e-05,
|
| 176 |
+
"loss": 0.2051,
|
| 177 |
+
"step": 12000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.26468470757633505,
|
| 181 |
+
"grad_norm": 1.015625,
|
| 182 |
+
"learning_rate": 1.6447686259650452e-05,
|
| 183 |
+
"loss": 0.1985,
|
| 184 |
+
"step": 12500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.27527209587938845,
|
| 188 |
+
"grad_norm": 0.734375,
|
| 189 |
+
"learning_rate": 1.621086534362715e-05,
|
| 190 |
+
"loss": 0.2035,
|
| 191 |
+
"step": 13000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.2858594841824419,
|
| 195 |
+
"grad_norm": 0.61328125,
|
| 196 |
+
"learning_rate": 1.5974044427603848e-05,
|
| 197 |
+
"loss": 0.2019,
|
| 198 |
+
"step": 13500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.2964468724854953,
|
| 202 |
+
"grad_norm": 0.953125,
|
| 203 |
+
"learning_rate": 1.5737223511580543e-05,
|
| 204 |
+
"loss": 0.1921,
|
| 205 |
+
"step": 14000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.3070342607885487,
|
| 209 |
+
"grad_norm": 0.83984375,
|
| 210 |
+
"learning_rate": 1.550040259555724e-05,
|
| 211 |
+
"loss": 0.1952,
|
| 212 |
+
"step": 14500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.31762164909160207,
|
| 216 |
+
"grad_norm": 1.03125,
|
| 217 |
+
"learning_rate": 1.526358167953394e-05,
|
| 218 |
+
"loss": 0.1988,
|
| 219 |
+
"step": 15000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.3282090373946555,
|
| 223 |
+
"grad_norm": 0.84375,
|
| 224 |
+
"learning_rate": 1.5026760763510635e-05,
|
| 225 |
+
"loss": 0.193,
|
| 226 |
+
"step": 15500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.3387964256977089,
|
| 230 |
+
"grad_norm": 0.60546875,
|
| 231 |
+
"learning_rate": 1.4789939847487332e-05,
|
| 232 |
+
"loss": 0.1946,
|
| 233 |
+
"step": 16000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.3493838140007623,
|
| 237 |
+
"grad_norm": 0.6484375,
|
| 238 |
+
"learning_rate": 1.4553118931464028e-05,
|
| 239 |
+
"loss": 0.1953,
|
| 240 |
+
"step": 16500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.3599712023038157,
|
| 244 |
+
"grad_norm": 1.09375,
|
| 245 |
+
"learning_rate": 1.4316298015440725e-05,
|
| 246 |
+
"loss": 0.1902,
|
| 247 |
+
"step": 17000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.3705585906068691,
|
| 251 |
+
"grad_norm": 0.578125,
|
| 252 |
+
"learning_rate": 1.4079477099417423e-05,
|
| 253 |
+
"loss": 0.1859,
|
| 254 |
+
"step": 17500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.3811459789099225,
|
| 258 |
+
"grad_norm": 1.046875,
|
| 259 |
+
"learning_rate": 1.3842656183394119e-05,
|
| 260 |
+
"loss": 0.1906,
|
| 261 |
+
"step": 18000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.3917333672129759,
|
| 265 |
+
"grad_norm": 0.51171875,
|
| 266 |
+
"learning_rate": 1.3605835267370815e-05,
|
| 267 |
+
"loss": 0.1908,
|
| 268 |
+
"step": 18500
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.4023207555160293,
|
| 272 |
+
"grad_norm": 0.73046875,
|
| 273 |
+
"learning_rate": 1.3369014351347512e-05,
|
| 274 |
+
"loss": 0.1805,
|
| 275 |
+
"step": 19000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.4129081438190827,
|
| 279 |
+
"grad_norm": 0.7734375,
|
| 280 |
+
"learning_rate": 1.313219343532421e-05,
|
| 281 |
+
"loss": 0.1855,
|
| 282 |
+
"step": 19500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.4234955321221361,
|
| 286 |
+
"grad_norm": 0.7265625,
|
| 287 |
+
"learning_rate": 1.2895372519300906e-05,
|
| 288 |
+
"loss": 0.1906,
|
| 289 |
+
"step": 20000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.43408292042518953,
|
| 293 |
+
"grad_norm": 0.87890625,
|
| 294 |
+
"learning_rate": 1.2658551603277602e-05,
|
| 295 |
+
"loss": 0.1801,
|
| 296 |
+
"step": 20500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.4446703087282429,
|
| 300 |
+
"grad_norm": 0.5234375,
|
| 301 |
+
"learning_rate": 1.2421730687254299e-05,
|
| 302 |
+
"loss": 0.1827,
|
| 303 |
+
"step": 21000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.4552576970312963,
|
| 307 |
+
"grad_norm": 0.9921875,
|
| 308 |
+
"learning_rate": 1.2184909771230997e-05,
|
| 309 |
+
"loss": 0.1832,
|
| 310 |
+
"step": 21500
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.4658450853343497,
|
| 314 |
+
"grad_norm": 0.66796875,
|
| 315 |
+
"learning_rate": 1.1948088855207693e-05,
|
| 316 |
+
"loss": 0.1875,
|
| 317 |
+
"step": 22000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.4764324736374031,
|
| 321 |
+
"grad_norm": 1.3203125,
|
| 322 |
+
"learning_rate": 1.171126793918439e-05,
|
| 323 |
+
"loss": 0.1794,
|
| 324 |
+
"step": 22500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.48701986194045654,
|
| 328 |
+
"grad_norm": 0.9765625,
|
| 329 |
+
"learning_rate": 1.1474447023161086e-05,
|
| 330 |
+
"loss": 0.1831,
|
| 331 |
+
"step": 23000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.49760725024350994,
|
| 335 |
+
"grad_norm": 0.59765625,
|
| 336 |
+
"learning_rate": 1.1237626107137782e-05,
|
| 337 |
+
"loss": 0.1841,
|
| 338 |
+
"step": 23500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.5081946385465633,
|
| 342 |
+
"grad_norm": 0.953125,
|
| 343 |
+
"learning_rate": 1.100080519111448e-05,
|
| 344 |
+
"loss": 0.186,
|
| 345 |
+
"step": 24000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.5187820268496167,
|
| 349 |
+
"grad_norm": 0.58984375,
|
| 350 |
+
"learning_rate": 1.0763984275091177e-05,
|
| 351 |
+
"loss": 0.1849,
|
| 352 |
+
"step": 24500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.5293694151526701,
|
| 356 |
+
"grad_norm": 1.328125,
|
| 357 |
+
"learning_rate": 1.0527163359067873e-05,
|
| 358 |
+
"loss": 0.1797,
|
| 359 |
+
"step": 25000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.5399568034557235,
|
| 363 |
+
"grad_norm": 0.734375,
|
| 364 |
+
"learning_rate": 1.029034244304457e-05,
|
| 365 |
+
"loss": 0.1867,
|
| 366 |
+
"step": 25500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5505441917587769,
|
| 370 |
+
"grad_norm": 1.09375,
|
| 371 |
+
"learning_rate": 1.0053521527021268e-05,
|
| 372 |
+
"loss": 0.1792,
|
| 373 |
+
"step": 26000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.5611315800618304,
|
| 377 |
+
"grad_norm": 0.6875,
|
| 378 |
+
"learning_rate": 9.816700610997964e-06,
|
| 379 |
+
"loss": 0.1839,
|
| 380 |
+
"step": 26500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.5717189683648838,
|
| 384 |
+
"grad_norm": 0.83984375,
|
| 385 |
+
"learning_rate": 9.57987969497466e-06,
|
| 386 |
+
"loss": 0.1816,
|
| 387 |
+
"step": 27000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.5823063566679372,
|
| 391 |
+
"grad_norm": 0.92578125,
|
| 392 |
+
"learning_rate": 9.343058778951358e-06,
|
| 393 |
+
"loss": 0.1816,
|
| 394 |
+
"step": 27500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.5928937449709906,
|
| 398 |
+
"grad_norm": 0.58984375,
|
| 399 |
+
"learning_rate": 9.106237862928055e-06,
|
| 400 |
+
"loss": 0.1779,
|
| 401 |
+
"step": 28000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.603481133274044,
|
| 405 |
+
"grad_norm": 0.609375,
|
| 406 |
+
"learning_rate": 8.869416946904751e-06,
|
| 407 |
+
"loss": 0.1821,
|
| 408 |
+
"step": 28500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.6140685215770973,
|
| 412 |
+
"grad_norm": 1.2109375,
|
| 413 |
+
"learning_rate": 8.632596030881449e-06,
|
| 414 |
+
"loss": 0.1824,
|
| 415 |
+
"step": 29000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.6246559098801507,
|
| 419 |
+
"grad_norm": 0.49609375,
|
| 420 |
+
"learning_rate": 8.395775114858144e-06,
|
| 421 |
+
"loss": 0.179,
|
| 422 |
+
"step": 29500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.6352432981832041,
|
| 426 |
+
"grad_norm": 0.5625,
|
| 427 |
+
"learning_rate": 8.158954198834842e-06,
|
| 428 |
+
"loss": 0.1846,
|
| 429 |
+
"step": 30000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.6458306864862575,
|
| 433 |
+
"grad_norm": 0.67578125,
|
| 434 |
+
"learning_rate": 7.922133282811538e-06,
|
| 435 |
+
"loss": 0.1853,
|
| 436 |
+
"step": 30500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.656418074789311,
|
| 440 |
+
"grad_norm": 1.0,
|
| 441 |
+
"learning_rate": 7.685312366788236e-06,
|
| 442 |
+
"loss": 0.1766,
|
| 443 |
+
"step": 31000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.6670054630923644,
|
| 447 |
+
"grad_norm": 0.82421875,
|
| 448 |
+
"learning_rate": 7.448491450764932e-06,
|
| 449 |
+
"loss": 0.1819,
|
| 450 |
+
"step": 31500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.6775928513954178,
|
| 454 |
+
"grad_norm": 0.90234375,
|
| 455 |
+
"learning_rate": 7.211670534741629e-06,
|
| 456 |
+
"loss": 0.1807,
|
| 457 |
+
"step": 32000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.6881802396984712,
|
| 461 |
+
"grad_norm": 0.546875,
|
| 462 |
+
"learning_rate": 6.974849618718325e-06,
|
| 463 |
+
"loss": 0.1824,
|
| 464 |
+
"step": 32500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.6987676280015246,
|
| 468 |
+
"grad_norm": 0.52734375,
|
| 469 |
+
"learning_rate": 6.738028702695023e-06,
|
| 470 |
+
"loss": 0.1829,
|
| 471 |
+
"step": 33000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.709355016304578,
|
| 475 |
+
"grad_norm": 0.91015625,
|
| 476 |
+
"learning_rate": 6.50120778667172e-06,
|
| 477 |
+
"loss": 0.1812,
|
| 478 |
+
"step": 33500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.7199424046076314,
|
| 482 |
+
"grad_norm": 0.5703125,
|
| 483 |
+
"learning_rate": 6.264386870648415e-06,
|
| 484 |
+
"loss": 0.1778,
|
| 485 |
+
"step": 34000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.7305297929106848,
|
| 489 |
+
"grad_norm": 1.09375,
|
| 490 |
+
"learning_rate": 6.027565954625113e-06,
|
| 491 |
+
"loss": 0.1868,
|
| 492 |
+
"step": 34500
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.7411171812137382,
|
| 496 |
+
"grad_norm": 0.91796875,
|
| 497 |
+
"learning_rate": 5.79074503860181e-06,
|
| 498 |
+
"loss": 0.1806,
|
| 499 |
+
"step": 35000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.7517045695167915,
|
| 503 |
+
"grad_norm": 0.69140625,
|
| 504 |
+
"learning_rate": 5.553924122578507e-06,
|
| 505 |
+
"loss": 0.1843,
|
| 506 |
+
"step": 35500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.762291957819845,
|
| 510 |
+
"grad_norm": 1.3046875,
|
| 511 |
+
"learning_rate": 5.317103206555203e-06,
|
| 512 |
+
"loss": 0.1778,
|
| 513 |
+
"step": 36000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.7728793461228984,
|
| 517 |
+
"grad_norm": 0.67578125,
|
| 518 |
+
"learning_rate": 5.0802822905319005e-06,
|
| 519 |
+
"loss": 0.1752,
|
| 520 |
+
"step": 36500
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.7834667344259518,
|
| 524 |
+
"grad_norm": 1.1875,
|
| 525 |
+
"learning_rate": 4.843461374508597e-06,
|
| 526 |
+
"loss": 0.1829,
|
| 527 |
+
"step": 37000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.7940541227290052,
|
| 531 |
+
"grad_norm": 0.9921875,
|
| 532 |
+
"learning_rate": 4.606640458485293e-06,
|
| 533 |
+
"loss": 0.1758,
|
| 534 |
+
"step": 37500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.8046415110320586,
|
| 538 |
+
"grad_norm": 0.859375,
|
| 539 |
+
"learning_rate": 4.3698195424619904e-06,
|
| 540 |
+
"loss": 0.1817,
|
| 541 |
+
"step": 38000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.815228899335112,
|
| 545 |
+
"grad_norm": 1.0234375,
|
| 546 |
+
"learning_rate": 4.132998626438688e-06,
|
| 547 |
+
"loss": 0.1872,
|
| 548 |
+
"step": 38500
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.8258162876381654,
|
| 552 |
+
"grad_norm": 0.6953125,
|
| 553 |
+
"learning_rate": 3.896177710415384e-06,
|
| 554 |
+
"loss": 0.1775,
|
| 555 |
+
"step": 39000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.8364036759412188,
|
| 559 |
+
"grad_norm": 1.5234375,
|
| 560 |
+
"learning_rate": 3.659356794392081e-06,
|
| 561 |
+
"loss": 0.1839,
|
| 562 |
+
"step": 39500
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.8469910642442722,
|
| 566 |
+
"grad_norm": 0.5625,
|
| 567 |
+
"learning_rate": 3.422535878368778e-06,
|
| 568 |
+
"loss": 0.1766,
|
| 569 |
+
"step": 40000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.8575784525473257,
|
| 573 |
+
"grad_norm": 0.875,
|
| 574 |
+
"learning_rate": 3.1857149623454748e-06,
|
| 575 |
+
"loss": 0.1782,
|
| 576 |
+
"step": 40500
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.8681658408503791,
|
| 580 |
+
"grad_norm": 1.359375,
|
| 581 |
+
"learning_rate": 2.9488940463221716e-06,
|
| 582 |
+
"loss": 0.1804,
|
| 583 |
+
"step": 41000
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.8787532291534325,
|
| 587 |
+
"grad_norm": 0.60546875,
|
| 588 |
+
"learning_rate": 2.7120731302988684e-06,
|
| 589 |
+
"loss": 0.178,
|
| 590 |
+
"step": 41500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.8893406174564858,
|
| 594 |
+
"grad_norm": 1.1328125,
|
| 595 |
+
"learning_rate": 2.475252214275565e-06,
|
| 596 |
+
"loss": 0.1836,
|
| 597 |
+
"step": 42000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.8999280057595392,
|
| 601 |
+
"grad_norm": 0.5859375,
|
| 602 |
+
"learning_rate": 2.238431298252262e-06,
|
| 603 |
+
"loss": 0.176,
|
| 604 |
+
"step": 42500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.9105153940625926,
|
| 608 |
+
"grad_norm": 2.203125,
|
| 609 |
+
"learning_rate": 2.0016103822289587e-06,
|
| 610 |
+
"loss": 0.1826,
|
| 611 |
+
"step": 43000
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.921102782365646,
|
| 615 |
+
"grad_norm": 0.466796875,
|
| 616 |
+
"learning_rate": 1.7647894662056555e-06,
|
| 617 |
+
"loss": 0.1837,
|
| 618 |
+
"step": 43500
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.9316901706686994,
|
| 622 |
+
"grad_norm": 1.265625,
|
| 623 |
+
"learning_rate": 1.527968550182352e-06,
|
| 624 |
+
"loss": 0.1778,
|
| 625 |
+
"step": 44000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.9422775589717528,
|
| 629 |
+
"grad_norm": 0.6171875,
|
| 630 |
+
"learning_rate": 1.291147634159049e-06,
|
| 631 |
+
"loss": 0.1817,
|
| 632 |
+
"step": 44500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.9528649472748062,
|
| 636 |
+
"grad_norm": 0.60546875,
|
| 637 |
+
"learning_rate": 1.0543267181357459e-06,
|
| 638 |
+
"loss": 0.1775,
|
| 639 |
+
"step": 45000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.9634523355778597,
|
| 643 |
+
"grad_norm": 0.5703125,
|
| 644 |
+
"learning_rate": 8.175058021124427e-07,
|
| 645 |
+
"loss": 0.1759,
|
| 646 |
+
"step": 45500
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.9740397238809131,
|
| 650 |
+
"grad_norm": 0.70703125,
|
| 651 |
+
"learning_rate": 5.806848860891394e-07,
|
| 652 |
+
"loss": 0.1723,
|
| 653 |
+
"step": 46000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.9846271121839665,
|
| 657 |
+
"grad_norm": 1.7265625,
|
| 658 |
+
"learning_rate": 3.4386397006583623e-07,
|
| 659 |
+
"loss": 0.1822,
|
| 660 |
+
"step": 46500
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.9952145004870199,
|
| 664 |
+
"grad_norm": 1.0234375,
|
| 665 |
+
"learning_rate": 1.0704305404253304e-07,
|
| 666 |
+
"loss": 0.1825,
|
| 667 |
+
"step": 47000
|
| 668 |
+
}
|
| 669 |
+
],
|
| 670 |
+
"logging_steps": 500,
|
| 671 |
+
"max_steps": 47226,
|
| 672 |
+
"num_input_tokens_seen": 0,
|
| 673 |
+
"num_train_epochs": 1,
|
| 674 |
+
"save_steps": 10000,
|
| 675 |
+
"stateful_callbacks": {
|
| 676 |
+
"TrainerControl": {
|
| 677 |
+
"args": {
|
| 678 |
+
"should_epoch_stop": false,
|
| 679 |
+
"should_evaluate": false,
|
| 680 |
+
"should_log": false,
|
| 681 |
+
"should_save": true,
|
| 682 |
+
"should_training_stop": true
|
| 683 |
+
},
|
| 684 |
+
"attributes": {}
|
| 685 |
+
}
|
| 686 |
+
},
|
| 687 |
+
"total_flos": 5.979597801902148e+18,
|
| 688 |
+
"train_batch_size": 10,
|
| 689 |
+
"trial_name": null,
|
| 690 |
+
"trial_params": null
|
| 691 |
+
}
|