Upload trainer_state.json
Browse files- trainer_state.json +682 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,682 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.224483333333333,
|
| 5 |
+
"global_step": 60000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.04,
|
| 12 |
+
"learning_rate": 0.002197932219132781,
|
| 13 |
+
"loss": 3.6508,
|
| 14 |
+
"step": 1000
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.09,
|
| 18 |
+
"learning_rate": 0.004419300239533186,
|
| 19 |
+
"loss": 2.1932,
|
| 20 |
+
"step": 2000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.13,
|
| 24 |
+
"learning_rate": 0.006713071372359991,
|
| 25 |
+
"loss": 1.9369,
|
| 26 |
+
"step": 3000
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.17,
|
| 30 |
+
"learning_rate": 0.009095062501728535,
|
| 31 |
+
"loss": 1.6844,
|
| 32 |
+
"step": 4000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.22,
|
| 36 |
+
"learning_rate": 0.011655605398118496,
|
| 37 |
+
"loss": 1.557,
|
| 38 |
+
"step": 5000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.26,
|
| 42 |
+
"learning_rate": 0.014309737831354141,
|
| 43 |
+
"loss": 1.4849,
|
| 44 |
+
"step": 6000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.3,
|
| 48 |
+
"learning_rate": 0.016971396282315254,
|
| 49 |
+
"loss": 1.6145,
|
| 50 |
+
"step": 7000
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.35,
|
| 54 |
+
"learning_rate": 0.01942763663828373,
|
| 55 |
+
"loss": 1.6191,
|
| 56 |
+
"step": 8000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.39,
|
| 60 |
+
"learning_rate": 0.022188229486346245,
|
| 61 |
+
"loss": 1.5503,
|
| 62 |
+
"step": 9000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.43,
|
| 66 |
+
"learning_rate": 0.025435281917452812,
|
| 67 |
+
"loss": 1.5535,
|
| 68 |
+
"step": 10000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.43,
|
| 72 |
+
"eval_bleu": 11.554505529017694,
|
| 73 |
+
"eval_loss": 1.8505867719650269,
|
| 74 |
+
"eval_runtime": 81.2961,
|
| 75 |
+
"eval_samples_per_second": 230.257,
|
| 76 |
+
"eval_steps_per_second": 0.91,
|
| 77 |
+
"step": 10000
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.48,
|
| 81 |
+
"learning_rate": 0.024702614173293114,
|
| 82 |
+
"loss": 1.4991,
|
| 83 |
+
"step": 11000
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.52,
|
| 87 |
+
"learning_rate": 0.024255122989416122,
|
| 88 |
+
"loss": 1.4803,
|
| 89 |
+
"step": 12000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.57,
|
| 93 |
+
"learning_rate": 0.02393551729619503,
|
| 94 |
+
"loss": 1.4415,
|
| 95 |
+
"step": 13000
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 0.61,
|
| 99 |
+
"learning_rate": 0.02371666394174099,
|
| 100 |
+
"loss": 1.4454,
|
| 101 |
+
"step": 14000
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.65,
|
| 105 |
+
"learning_rate": 0.02343085967004299,
|
| 106 |
+
"loss": 1.4106,
|
| 107 |
+
"step": 15000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.7,
|
| 111 |
+
"learning_rate": 0.02309497445821762,
|
| 112 |
+
"loss": 1.3996,
|
| 113 |
+
"step": 16000
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.74,
|
| 117 |
+
"learning_rate": 0.02294100448489189,
|
| 118 |
+
"loss": 1.3815,
|
| 119 |
+
"step": 17000
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.78,
|
| 123 |
+
"learning_rate": 0.022620119154453278,
|
| 124 |
+
"loss": 1.3915,
|
| 125 |
+
"step": 18000
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.83,
|
| 129 |
+
"learning_rate": 0.022044328972697258,
|
| 130 |
+
"loss": 1.5557,
|
| 131 |
+
"step": 19000
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.87,
|
| 135 |
+
"learning_rate": 0.02174384333193302,
|
| 136 |
+
"loss": 1.2879,
|
| 137 |
+
"step": 20000
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.87,
|
| 141 |
+
"eval_bleu": 15.736323415434928,
|
| 142 |
+
"eval_loss": 1.4790531396865845,
|
| 143 |
+
"eval_runtime": 81.7862,
|
| 144 |
+
"eval_samples_per_second": 228.877,
|
| 145 |
+
"eval_steps_per_second": 0.905,
|
| 146 |
+
"step": 20000
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.91,
|
| 150 |
+
"learning_rate": 0.02188653126358986,
|
| 151 |
+
"loss": 1.0018,
|
| 152 |
+
"step": 21000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.48,
|
| 156 |
+
"learning_rate": 0.02189297415316105,
|
| 157 |
+
"loss": 0.9382,
|
| 158 |
+
"step": 22000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.5,
|
| 162 |
+
"learning_rate": 0.02184857614338398,
|
| 163 |
+
"loss": 0.9229,
|
| 164 |
+
"step": 23000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.02,
|
| 168 |
+
"learning_rate": 0.021777568385004997,
|
| 169 |
+
"loss": 1.2887,
|
| 170 |
+
"step": 24000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.04,
|
| 174 |
+
"learning_rate": 0.021719269454479218,
|
| 175 |
+
"loss": 1.3278,
|
| 176 |
+
"step": 24500
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.06,
|
| 180 |
+
"learning_rate": 0.021626409143209457,
|
| 181 |
+
"loss": 1.3052,
|
| 182 |
+
"step": 25000
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.06,
|
| 186 |
+
"eval_bleu": 15.689633530325342,
|
| 187 |
+
"eval_loss": 1.344283103942871,
|
| 188 |
+
"eval_runtime": 81.5254,
|
| 189 |
+
"eval_samples_per_second": 229.61,
|
| 190 |
+
"eval_steps_per_second": 0.908,
|
| 191 |
+
"step": 25000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.07,
|
| 195 |
+
"learning_rate": 0.02152758091688156,
|
| 196 |
+
"loss": 1.2788,
|
| 197 |
+
"step": 25500
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 1.09,
|
| 201 |
+
"learning_rate": 0.021510712802410126,
|
| 202 |
+
"loss": 1.369,
|
| 203 |
+
"step": 26000
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.11,
|
| 207 |
+
"learning_rate": 0.021486839279532433,
|
| 208 |
+
"loss": 1.3663,
|
| 209 |
+
"step": 26500
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 1.12,
|
| 213 |
+
"learning_rate": 0.021426061168313026,
|
| 214 |
+
"loss": 1.2913,
|
| 215 |
+
"step": 27000
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 1.14,
|
| 219 |
+
"learning_rate": 0.021364226937294006,
|
| 220 |
+
"loss": 1.2785,
|
| 221 |
+
"step": 27500
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.16,
|
| 225 |
+
"learning_rate": 0.021301250904798508,
|
| 226 |
+
"loss": 1.2471,
|
| 227 |
+
"step": 28000
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.17,
|
| 231 |
+
"learning_rate": 0.021244272589683533,
|
| 232 |
+
"loss": 1.2191,
|
| 233 |
+
"step": 28500
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.19,
|
| 237 |
+
"learning_rate": 0.021167641505599022,
|
| 238 |
+
"loss": 1.2128,
|
| 239 |
+
"step": 29000
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.21,
|
| 243 |
+
"learning_rate": 0.021093450486660004,
|
| 244 |
+
"loss": 1.1959,
|
| 245 |
+
"step": 29500
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.22,
|
| 249 |
+
"learning_rate": 0.02111968770623207,
|
| 250 |
+
"loss": 1.3139,
|
| 251 |
+
"step": 30000
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 1.22,
|
| 255 |
+
"eval_bleu": 16.118544297247887,
|
| 256 |
+
"eval_loss": 1.2931314706802368,
|
| 257 |
+
"eval_runtime": 81.4634,
|
| 258 |
+
"eval_samples_per_second": 229.784,
|
| 259 |
+
"eval_steps_per_second": 0.908,
|
| 260 |
+
"step": 30000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.18,
|
| 264 |
+
"learning_rate": 0.021165331825613976,
|
| 265 |
+
"loss": 1.3204,
|
| 266 |
+
"step": 30500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.19,
|
| 270 |
+
"learning_rate": 0.021186990663409233,
|
| 271 |
+
"loss": 1.2925,
|
| 272 |
+
"step": 31000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.21,
|
| 276 |
+
"learning_rate": 0.021194208413362503,
|
| 277 |
+
"loss": 1.3002,
|
| 278 |
+
"step": 31500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.22,
|
| 282 |
+
"learning_rate": 0.021185804158449173,
|
| 283 |
+
"loss": 1.2473,
|
| 284 |
+
"step": 32000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.23,
|
| 288 |
+
"learning_rate": 0.021163810044527054,
|
| 289 |
+
"loss": 1.2455,
|
| 290 |
+
"step": 32500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.24,
|
| 294 |
+
"learning_rate": 0.021145079284906387,
|
| 295 |
+
"loss": 1.2411,
|
| 296 |
+
"step": 33000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.26,
|
| 300 |
+
"learning_rate": 0.021122504025697708,
|
| 301 |
+
"loss": 1.2214,
|
| 302 |
+
"step": 33500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 1.27,
|
| 306 |
+
"learning_rate": 0.021091610193252563,
|
| 307 |
+
"loss": 1.2239,
|
| 308 |
+
"step": 34000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.28,
|
| 312 |
+
"learning_rate": 0.021048951894044876,
|
| 313 |
+
"loss": 1.228,
|
| 314 |
+
"step": 34500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 1.29,
|
| 318 |
+
"learning_rate": 0.02102178893983364,
|
| 319 |
+
"loss": 1.2119,
|
| 320 |
+
"step": 35000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.29,
|
| 324 |
+
"eval_bleu": 15.491337721835558,
|
| 325 |
+
"eval_loss": 1.3143055438995361,
|
| 326 |
+
"eval_runtime": 81.4828,
|
| 327 |
+
"eval_samples_per_second": 229.73,
|
| 328 |
+
"eval_steps_per_second": 0.908,
|
| 329 |
+
"step": 35000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.31,
|
| 333 |
+
"learning_rate": 0.020987574011087418,
|
| 334 |
+
"loss": 1.1862,
|
| 335 |
+
"step": 35500
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 1.32,
|
| 339 |
+
"learning_rate": 0.02096775360405445,
|
| 340 |
+
"loss": 1.2127,
|
| 341 |
+
"step": 36000
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 1.33,
|
| 345 |
+
"learning_rate": 0.02094915322959423,
|
| 346 |
+
"loss": 1.228,
|
| 347 |
+
"step": 36500
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.34,
|
| 351 |
+
"learning_rate": 0.02091248333454132,
|
| 352 |
+
"loss": 1.2456,
|
| 353 |
+
"step": 37000
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.36,
|
| 357 |
+
"learning_rate": 0.020875241607427597,
|
| 358 |
+
"loss": 1.2494,
|
| 359 |
+
"step": 37500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.37,
|
| 363 |
+
"learning_rate": 0.020842362195253372,
|
| 364 |
+
"loss": 1.1684,
|
| 365 |
+
"step": 38000
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 1.38,
|
| 369 |
+
"learning_rate": 0.02081664651632309,
|
| 370 |
+
"loss": 1.2389,
|
| 371 |
+
"step": 38500
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.39,
|
| 375 |
+
"learning_rate": 0.020784636959433556,
|
| 376 |
+
"loss": 1.1898,
|
| 377 |
+
"step": 39000
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 1.41,
|
| 381 |
+
"learning_rate": 0.020755469799041748,
|
| 382 |
+
"loss": 1.1915,
|
| 383 |
+
"step": 39500
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.42,
|
| 387 |
+
"learning_rate": 0.02073347009718418,
|
| 388 |
+
"loss": 1.1929,
|
| 389 |
+
"step": 40000
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.42,
|
| 393 |
+
"eval_bleu": 15.669781616522553,
|
| 394 |
+
"eval_loss": 1.3155781030654907,
|
| 395 |
+
"eval_runtime": 81.5967,
|
| 396 |
+
"eval_samples_per_second": 229.409,
|
| 397 |
+
"eval_steps_per_second": 0.907,
|
| 398 |
+
"step": 40000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.29,
|
| 402 |
+
"learning_rate": 0.020710671320557594,
|
| 403 |
+
"loss": 1.2275,
|
| 404 |
+
"step": 40500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.3,
|
| 408 |
+
"learning_rate": 0.020680107176303864,
|
| 409 |
+
"loss": 1.1966,
|
| 410 |
+
"step": 41000
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.3,
|
| 414 |
+
"learning_rate": 0.020645136013627052,
|
| 415 |
+
"loss": 1.2111,
|
| 416 |
+
"step": 41500
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.31,
|
| 420 |
+
"learning_rate": 0.020648332312703133,
|
| 421 |
+
"loss": 1.3908,
|
| 422 |
+
"step": 42000
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.32,
|
| 426 |
+
"learning_rate": 0.02063341997563839,
|
| 427 |
+
"loss": 1.3234,
|
| 428 |
+
"step": 42500
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.33,
|
| 432 |
+
"learning_rate": 0.02064467966556549,
|
| 433 |
+
"loss": 0.9668,
|
| 434 |
+
"step": 43000
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.34,
|
| 438 |
+
"learning_rate": 0.020663069561123848,
|
| 439 |
+
"loss": 0.8413,
|
| 440 |
+
"step": 43500
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.35,
|
| 444 |
+
"learning_rate": 0.0206832867115736,
|
| 445 |
+
"loss": 0.8273,
|
| 446 |
+
"step": 44000
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.35,
|
| 450 |
+
"learning_rate": 0.020704049617052078,
|
| 451 |
+
"loss": 0.7956,
|
| 452 |
+
"step": 44500
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.36,
|
| 456 |
+
"learning_rate": 0.020729372277855873,
|
| 457 |
+
"loss": 0.7875,
|
| 458 |
+
"step": 45000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.36,
|
| 462 |
+
"eval_bleu": 17.782492667022247,
|
| 463 |
+
"eval_loss": 1.2672479152679443,
|
| 464 |
+
"eval_runtime": 81.5877,
|
| 465 |
+
"eval_samples_per_second": 229.434,
|
| 466 |
+
"eval_steps_per_second": 0.907,
|
| 467 |
+
"step": 45000
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 1.37,
|
| 471 |
+
"learning_rate": 0.020742451772093773,
|
| 472 |
+
"loss": 0.7773,
|
| 473 |
+
"step": 45500
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.38,
|
| 477 |
+
"learning_rate": 0.020756695419549942,
|
| 478 |
+
"loss": 0.7604,
|
| 479 |
+
"step": 46000
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.39,
|
| 483 |
+
"learning_rate": 0.020774947479367256,
|
| 484 |
+
"loss": 0.8933,
|
| 485 |
+
"step": 46500
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.01,
|
| 489 |
+
"learning_rate": 0.02075868472456932,
|
| 490 |
+
"loss": 1.2112,
|
| 491 |
+
"step": 47000
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 2.02,
|
| 495 |
+
"learning_rate": 0.020726632326841354,
|
| 496 |
+
"loss": 1.1824,
|
| 497 |
+
"step": 47500
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 2.02,
|
| 501 |
+
"learning_rate": 0.020695462822914124,
|
| 502 |
+
"loss": 1.1919,
|
| 503 |
+
"step": 48000
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.03,
|
| 507 |
+
"learning_rate": 0.020659752190113068,
|
| 508 |
+
"loss": 1.1771,
|
| 509 |
+
"step": 48500
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 2.04,
|
| 513 |
+
"learning_rate": 0.02061435766518116,
|
| 514 |
+
"loss": 1.1045,
|
| 515 |
+
"step": 49000
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.05,
|
| 519 |
+
"learning_rate": 0.02060552127659321,
|
| 520 |
+
"loss": 1.2627,
|
| 521 |
+
"step": 49500
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 2.06,
|
| 525 |
+
"learning_rate": 0.020578160881996155,
|
| 526 |
+
"loss": 1.1741,
|
| 527 |
+
"step": 50000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.06,
|
| 531 |
+
"eval_bleu": 17.817437675612407,
|
| 532 |
+
"eval_loss": 1.2213643789291382,
|
| 533 |
+
"eval_runtime": 81.6429,
|
| 534 |
+
"eval_samples_per_second": 229.279,
|
| 535 |
+
"eval_steps_per_second": 0.906,
|
| 536 |
+
"step": 50000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.07,
|
| 540 |
+
"learning_rate": 0.02054954506456852,
|
| 541 |
+
"loss": 1.1366,
|
| 542 |
+
"step": 50500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 2.07,
|
| 546 |
+
"learning_rate": 0.020517559722065926,
|
| 547 |
+
"loss": 1.1321,
|
| 548 |
+
"step": 51000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.08,
|
| 552 |
+
"learning_rate": 0.020486876368522644,
|
| 553 |
+
"loss": 1.1128,
|
| 554 |
+
"step": 51500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 2.09,
|
| 558 |
+
"learning_rate": 0.020454248413443565,
|
| 559 |
+
"loss": 1.0992,
|
| 560 |
+
"step": 52000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 2.1,
|
| 564 |
+
"learning_rate": 0.020409852266311646,
|
| 565 |
+
"loss": 1.0874,
|
| 566 |
+
"step": 52500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.11,
|
| 570 |
+
"learning_rate": 0.020372800529003143,
|
| 571 |
+
"loss": 1.0964,
|
| 572 |
+
"step": 53000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 2.12,
|
| 576 |
+
"learning_rate": 0.020410917699337006,
|
| 577 |
+
"loss": 1.2197,
|
| 578 |
+
"step": 53500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.12,
|
| 582 |
+
"learning_rate": 0.020421497523784637,
|
| 583 |
+
"loss": 1.1628,
|
| 584 |
+
"step": 54000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 2.13,
|
| 588 |
+
"learning_rate": 0.02043827436864376,
|
| 589 |
+
"loss": 1.1952,
|
| 590 |
+
"step": 54500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.14,
|
| 594 |
+
"learning_rate": 0.020448317751288414,
|
| 595 |
+
"loss": 1.1603,
|
| 596 |
+
"step": 55000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 2.14,
|
| 600 |
+
"eval_bleu": 17.473470393567123,
|
| 601 |
+
"eval_loss": 1.1755515336990356,
|
| 602 |
+
"eval_runtime": 81.7968,
|
| 603 |
+
"eval_samples_per_second": 228.848,
|
| 604 |
+
"eval_steps_per_second": 0.905,
|
| 605 |
+
"step": 55000
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 2.15,
|
| 609 |
+
"learning_rate": 0.020445365458726883,
|
| 610 |
+
"loss": 1.1201,
|
| 611 |
+
"step": 55500
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.16,
|
| 615 |
+
"learning_rate": 0.020447757095098495,
|
| 616 |
+
"loss": 1.1362,
|
| 617 |
+
"step": 56000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 2.17,
|
| 621 |
+
"learning_rate": 0.020444199442863464,
|
| 622 |
+
"loss": 1.151,
|
| 623 |
+
"step": 56500
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 2.17,
|
| 627 |
+
"learning_rate": 0.020440472289919853,
|
| 628 |
+
"loss": 1.125,
|
| 629 |
+
"step": 57000
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 2.18,
|
| 633 |
+
"learning_rate": 0.020431550219655037,
|
| 634 |
+
"loss": 1.1041,
|
| 635 |
+
"step": 57500
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 2.19,
|
| 639 |
+
"learning_rate": 0.0204232819378376,
|
| 640 |
+
"loss": 1.1338,
|
| 641 |
+
"step": 58000
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 2.2,
|
| 645 |
+
"learning_rate": 0.02041398547589779,
|
| 646 |
+
"loss": 1.1279,
|
| 647 |
+
"step": 58500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 2.21,
|
| 651 |
+
"learning_rate": 0.02040201798081398,
|
| 652 |
+
"loss": 1.0957,
|
| 653 |
+
"step": 59000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.22,
|
| 657 |
+
"learning_rate": 0.020392674952745438,
|
| 658 |
+
"loss": 1.1315,
|
| 659 |
+
"step": 59500
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 2.22,
|
| 663 |
+
"learning_rate": 0.020384889096021652,
|
| 664 |
+
"loss": 1.146,
|
| 665 |
+
"step": 60000
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 2.22,
|
| 669 |
+
"eval_bleu": 17.10106003319275,
|
| 670 |
+
"eval_loss": 1.2006129026412964,
|
| 671 |
+
"eval_runtime": 81.8911,
|
| 672 |
+
"eval_samples_per_second": 228.584,
|
| 673 |
+
"eval_steps_per_second": 0.904,
|
| 674 |
+
"step": 60000
|
| 675 |
+
}
|
| 676 |
+
],
|
| 677 |
+
"max_steps": 60000,
|
| 678 |
+
"num_train_epochs": 9223372036854775807,
|
| 679 |
+
"total_flos": 4.6881481692662784e+17,
|
| 680 |
+
"trial_name": null,
|
| 681 |
+
"trial_params": null
|
| 682 |
+
}
|