Upload trainer_state.json
Browse files- trainer_state.json +676 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,676 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9702315325248071,
|
| 5 |
+
"global_step": 440,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.02,
|
| 12 |
+
"learning_rate": 9.77924944812362e-05,
|
| 13 |
+
"loss": 0.4213,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.02,
|
| 18 |
+
"eval_accuracy": 0.9030836820602417,
|
| 19 |
+
"eval_loss": 0.24132861196994781,
|
| 20 |
+
"eval_runtime": 17.6613,
|
| 21 |
+
"eval_samples_per_second": 25.706,
|
| 22 |
+
"eval_steps_per_second": 6.455,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04,
|
| 27 |
+
"learning_rate": 9.55849889624724e-05,
|
| 28 |
+
"loss": 0.4863,
|
| 29 |
+
"step": 20
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 0.04,
|
| 33 |
+
"eval_accuracy": 0.865638792514801,
|
| 34 |
+
"eval_loss": 0.2710409164428711,
|
| 35 |
+
"eval_runtime": 19.2736,
|
| 36 |
+
"eval_samples_per_second": 23.556,
|
| 37 |
+
"eval_steps_per_second": 5.915,
|
| 38 |
+
"step": 20
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07,
|
| 42 |
+
"learning_rate": 9.337748344370862e-05,
|
| 43 |
+
"loss": 0.3922,
|
| 44 |
+
"step": 30
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.07,
|
| 48 |
+
"eval_accuracy": 0.8722466826438904,
|
| 49 |
+
"eval_loss": 0.3421204090118408,
|
| 50 |
+
"eval_runtime": 16.7045,
|
| 51 |
+
"eval_samples_per_second": 27.178,
|
| 52 |
+
"eval_steps_per_second": 6.825,
|
| 53 |
+
"step": 30
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.09,
|
| 57 |
+
"learning_rate": 9.116997792494482e-05,
|
| 58 |
+
"loss": 0.1911,
|
| 59 |
+
"step": 40
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.09,
|
| 63 |
+
"eval_accuracy": 0.8127753138542175,
|
| 64 |
+
"eval_loss": 0.4282372295856476,
|
| 65 |
+
"eval_runtime": 18.8864,
|
| 66 |
+
"eval_samples_per_second": 24.039,
|
| 67 |
+
"eval_steps_per_second": 6.036,
|
| 68 |
+
"step": 40
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.11,
|
| 72 |
+
"learning_rate": 8.896247240618102e-05,
|
| 73 |
+
"loss": 0.3023,
|
| 74 |
+
"step": 50
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.11,
|
| 78 |
+
"eval_accuracy": 0.8942731022834778,
|
| 79 |
+
"eval_loss": 0.4062748849391937,
|
| 80 |
+
"eval_runtime": 17.7274,
|
| 81 |
+
"eval_samples_per_second": 25.61,
|
| 82 |
+
"eval_steps_per_second": 6.431,
|
| 83 |
+
"step": 50
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.13,
|
| 87 |
+
"learning_rate": 8.675496688741722e-05,
|
| 88 |
+
"loss": 0.3875,
|
| 89 |
+
"step": 60
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.13,
|
| 93 |
+
"eval_accuracy": 0.8722466826438904,
|
| 94 |
+
"eval_loss": 0.341449499130249,
|
| 95 |
+
"eval_runtime": 16.6774,
|
| 96 |
+
"eval_samples_per_second": 27.222,
|
| 97 |
+
"eval_steps_per_second": 6.836,
|
| 98 |
+
"step": 60
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.15,
|
| 102 |
+
"learning_rate": 8.454746136865342e-05,
|
| 103 |
+
"loss": 0.2723,
|
| 104 |
+
"step": 70
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.15,
|
| 108 |
+
"eval_accuracy": 0.8744493126869202,
|
| 109 |
+
"eval_loss": 0.3896920084953308,
|
| 110 |
+
"eval_runtime": 16.9212,
|
| 111 |
+
"eval_samples_per_second": 26.83,
|
| 112 |
+
"eval_steps_per_second": 6.737,
|
| 113 |
+
"step": 70
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.18,
|
| 117 |
+
"learning_rate": 8.233995584988962e-05,
|
| 118 |
+
"loss": 0.1704,
|
| 119 |
+
"step": 80
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.18,
|
| 123 |
+
"eval_accuracy": 0.8700440526008606,
|
| 124 |
+
"eval_loss": 0.3918803632259369,
|
| 125 |
+
"eval_runtime": 18.7956,
|
| 126 |
+
"eval_samples_per_second": 24.155,
|
| 127 |
+
"eval_steps_per_second": 6.065,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.2,
|
| 132 |
+
"learning_rate": 8.013245033112584e-05,
|
| 133 |
+
"loss": 0.3443,
|
| 134 |
+
"step": 90
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.2,
|
| 138 |
+
"eval_accuracy": 0.8942731022834778,
|
| 139 |
+
"eval_loss": 0.3114801049232483,
|
| 140 |
+
"eval_runtime": 16.9667,
|
| 141 |
+
"eval_samples_per_second": 26.758,
|
| 142 |
+
"eval_steps_per_second": 6.719,
|
| 143 |
+
"step": 90
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.22,
|
| 147 |
+
"learning_rate": 7.792494481236204e-05,
|
| 148 |
+
"loss": 0.2096,
|
| 149 |
+
"step": 100
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.22,
|
| 153 |
+
"eval_accuracy": 0.9030836820602417,
|
| 154 |
+
"eval_loss": 0.2851845622062683,
|
| 155 |
+
"eval_runtime": 16.7531,
|
| 156 |
+
"eval_samples_per_second": 27.1,
|
| 157 |
+
"eval_steps_per_second": 6.805,
|
| 158 |
+
"step": 100
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.24,
|
| 162 |
+
"learning_rate": 7.571743929359824e-05,
|
| 163 |
+
"loss": 0.2619,
|
| 164 |
+
"step": 110
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.24,
|
| 168 |
+
"eval_accuracy": 0.8964757919311523,
|
| 169 |
+
"eval_loss": 0.22432558238506317,
|
| 170 |
+
"eval_runtime": 18.4437,
|
| 171 |
+
"eval_samples_per_second": 24.615,
|
| 172 |
+
"eval_steps_per_second": 6.181,
|
| 173 |
+
"step": 110
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.26,
|
| 177 |
+
"learning_rate": 7.350993377483444e-05,
|
| 178 |
+
"loss": 0.2568,
|
| 179 |
+
"step": 120
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.26,
|
| 183 |
+
"eval_accuracy": 0.8832598924636841,
|
| 184 |
+
"eval_loss": 0.3484536409378052,
|
| 185 |
+
"eval_runtime": 23.929,
|
| 186 |
+
"eval_samples_per_second": 18.973,
|
| 187 |
+
"eval_steps_per_second": 4.764,
|
| 188 |
+
"step": 120
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.29,
|
| 192 |
+
"learning_rate": 7.130242825607064e-05,
|
| 193 |
+
"loss": 0.171,
|
| 194 |
+
"step": 130
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.29,
|
| 198 |
+
"eval_accuracy": 0.8854625821113586,
|
| 199 |
+
"eval_loss": 0.36249256134033203,
|
| 200 |
+
"eval_runtime": 16.6508,
|
| 201 |
+
"eval_samples_per_second": 27.266,
|
| 202 |
+
"eval_steps_per_second": 6.847,
|
| 203 |
+
"step": 130
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.31,
|
| 207 |
+
"learning_rate": 6.909492273730685e-05,
|
| 208 |
+
"loss": 0.2208,
|
| 209 |
+
"step": 140
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.31,
|
| 213 |
+
"eval_accuracy": 0.9096916317939758,
|
| 214 |
+
"eval_loss": 0.27433687448501587,
|
| 215 |
+
"eval_runtime": 17.0236,
|
| 216 |
+
"eval_samples_per_second": 26.669,
|
| 217 |
+
"eval_steps_per_second": 6.697,
|
| 218 |
+
"step": 140
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.33,
|
| 222 |
+
"learning_rate": 6.688741721854305e-05,
|
| 223 |
+
"loss": 0.1581,
|
| 224 |
+
"step": 150
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.33,
|
| 228 |
+
"eval_accuracy": 0.9140968918800354,
|
| 229 |
+
"eval_loss": 0.2508123517036438,
|
| 230 |
+
"eval_runtime": 18.1498,
|
| 231 |
+
"eval_samples_per_second": 25.014,
|
| 232 |
+
"eval_steps_per_second": 6.281,
|
| 233 |
+
"step": 150
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.35,
|
| 237 |
+
"learning_rate": 6.467991169977925e-05,
|
| 238 |
+
"loss": 0.2796,
|
| 239 |
+
"step": 160
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.35,
|
| 243 |
+
"eval_accuracy": 0.9118942618370056,
|
| 244 |
+
"eval_loss": 0.24465535581111908,
|
| 245 |
+
"eval_runtime": 16.9303,
|
| 246 |
+
"eval_samples_per_second": 26.816,
|
| 247 |
+
"eval_steps_per_second": 6.733,
|
| 248 |
+
"step": 160
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.37,
|
| 252 |
+
"learning_rate": 6.247240618101545e-05,
|
| 253 |
+
"loss": 0.2155,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.37,
|
| 258 |
+
"eval_accuracy": 0.9096916317939758,
|
| 259 |
+
"eval_loss": 0.250988245010376,
|
| 260 |
+
"eval_runtime": 17.1735,
|
| 261 |
+
"eval_samples_per_second": 26.436,
|
| 262 |
+
"eval_steps_per_second": 6.638,
|
| 263 |
+
"step": 170
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.4,
|
| 267 |
+
"learning_rate": 6.026490066225165e-05,
|
| 268 |
+
"loss": 0.2012,
|
| 269 |
+
"step": 180
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.4,
|
| 273 |
+
"eval_accuracy": 0.9118942618370056,
|
| 274 |
+
"eval_loss": 0.2246789187192917,
|
| 275 |
+
"eval_runtime": 16.6423,
|
| 276 |
+
"eval_samples_per_second": 27.28,
|
| 277 |
+
"eval_steps_per_second": 6.85,
|
| 278 |
+
"step": 180
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.42,
|
| 282 |
+
"learning_rate": 5.805739514348786e-05,
|
| 283 |
+
"loss": 0.251,
|
| 284 |
+
"step": 190
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.42,
|
| 288 |
+
"eval_accuracy": 0.91629958152771,
|
| 289 |
+
"eval_loss": 0.24300266802310944,
|
| 290 |
+
"eval_runtime": 17.9462,
|
| 291 |
+
"eval_samples_per_second": 25.298,
|
| 292 |
+
"eval_steps_per_second": 6.352,
|
| 293 |
+
"step": 190
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 0.44,
|
| 297 |
+
"learning_rate": 5.607064017660044e-05,
|
| 298 |
+
"loss": 0.3025,
|
| 299 |
+
"step": 200
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.44,
|
| 303 |
+
"eval_accuracy": 0.91629958152771,
|
| 304 |
+
"eval_loss": 0.28177693486213684,
|
| 305 |
+
"eval_runtime": 16.5094,
|
| 306 |
+
"eval_samples_per_second": 27.499,
|
| 307 |
+
"eval_steps_per_second": 6.905,
|
| 308 |
+
"step": 200
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.46,
|
| 312 |
+
"learning_rate": 5.386313465783664e-05,
|
| 313 |
+
"loss": 0.2298,
|
| 314 |
+
"step": 210
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.46,
|
| 318 |
+
"eval_accuracy": 0.907489001750946,
|
| 319 |
+
"eval_loss": 0.26173070073127747,
|
| 320 |
+
"eval_runtime": 16.7193,
|
| 321 |
+
"eval_samples_per_second": 27.154,
|
| 322 |
+
"eval_steps_per_second": 6.818,
|
| 323 |
+
"step": 210
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 0.49,
|
| 327 |
+
"learning_rate": 5.165562913907285e-05,
|
| 328 |
+
"loss": 0.1938,
|
| 329 |
+
"step": 220
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.49,
|
| 333 |
+
"eval_accuracy": 0.8964757919311523,
|
| 334 |
+
"eval_loss": 0.3249448239803314,
|
| 335 |
+
"eval_runtime": 17.1352,
|
| 336 |
+
"eval_samples_per_second": 26.495,
|
| 337 |
+
"eval_steps_per_second": 6.653,
|
| 338 |
+
"step": 220
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.51,
|
| 342 |
+
"learning_rate": 4.9448123620309056e-05,
|
| 343 |
+
"loss": 0.2256,
|
| 344 |
+
"step": 230
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.51,
|
| 348 |
+
"eval_accuracy": 0.8766520023345947,
|
| 349 |
+
"eval_loss": 0.3929162621498108,
|
| 350 |
+
"eval_runtime": 27.4825,
|
| 351 |
+
"eval_samples_per_second": 16.52,
|
| 352 |
+
"eval_steps_per_second": 4.148,
|
| 353 |
+
"step": 230
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.53,
|
| 357 |
+
"learning_rate": 4.7240618101545256e-05,
|
| 358 |
+
"loss": 0.1327,
|
| 359 |
+
"step": 240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.53,
|
| 363 |
+
"eval_accuracy": 0.907489001750946,
|
| 364 |
+
"eval_loss": 0.2901674509048462,
|
| 365 |
+
"eval_runtime": 17.1801,
|
| 366 |
+
"eval_samples_per_second": 26.426,
|
| 367 |
+
"eval_steps_per_second": 6.636,
|
| 368 |
+
"step": 240
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.55,
|
| 372 |
+
"learning_rate": 4.5033112582781463e-05,
|
| 373 |
+
"loss": 0.1238,
|
| 374 |
+
"step": 250
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.55,
|
| 378 |
+
"eval_accuracy": 0.9140968918800354,
|
| 379 |
+
"eval_loss": 0.2505219280719757,
|
| 380 |
+
"eval_runtime": 17.0203,
|
| 381 |
+
"eval_samples_per_second": 26.674,
|
| 382 |
+
"eval_steps_per_second": 6.698,
|
| 383 |
+
"step": 250
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.57,
|
| 387 |
+
"learning_rate": 4.2825607064017664e-05,
|
| 388 |
+
"loss": 0.1279,
|
| 389 |
+
"step": 260
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.57,
|
| 393 |
+
"eval_accuracy": 0.9030836820602417,
|
| 394 |
+
"eval_loss": 0.2885590195655823,
|
| 395 |
+
"eval_runtime": 16.8884,
|
| 396 |
+
"eval_samples_per_second": 26.882,
|
| 397 |
+
"eval_steps_per_second": 6.75,
|
| 398 |
+
"step": 260
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.6,
|
| 402 |
+
"learning_rate": 4.0618101545253864e-05,
|
| 403 |
+
"loss": 0.3321,
|
| 404 |
+
"step": 270
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.6,
|
| 408 |
+
"eval_accuracy": 0.9008810520172119,
|
| 409 |
+
"eval_loss": 0.301535964012146,
|
| 410 |
+
"eval_runtime": 17.9274,
|
| 411 |
+
"eval_samples_per_second": 25.324,
|
| 412 |
+
"eval_steps_per_second": 6.359,
|
| 413 |
+
"step": 270
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.62,
|
| 417 |
+
"learning_rate": 3.8631346578366446e-05,
|
| 418 |
+
"loss": 0.2077,
|
| 419 |
+
"step": 280
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.62,
|
| 423 |
+
"eval_accuracy": 0.9096916317939758,
|
| 424 |
+
"eval_loss": 0.28936102986335754,
|
| 425 |
+
"eval_runtime": 18.9025,
|
| 426 |
+
"eval_samples_per_second": 24.018,
|
| 427 |
+
"eval_steps_per_second": 6.031,
|
| 428 |
+
"step": 280
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.64,
|
| 432 |
+
"learning_rate": 3.6423841059602646e-05,
|
| 433 |
+
"loss": 0.2746,
|
| 434 |
+
"step": 290
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.64,
|
| 438 |
+
"eval_accuracy": 0.9096916317939758,
|
| 439 |
+
"eval_loss": 0.26631271839141846,
|
| 440 |
+
"eval_runtime": 17.2835,
|
| 441 |
+
"eval_samples_per_second": 26.268,
|
| 442 |
+
"eval_steps_per_second": 6.596,
|
| 443 |
+
"step": 290
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.66,
|
| 447 |
+
"learning_rate": 3.4216335540838853e-05,
|
| 448 |
+
"loss": 0.1774,
|
| 449 |
+
"step": 300
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.66,
|
| 453 |
+
"eval_accuracy": 0.9140968918800354,
|
| 454 |
+
"eval_loss": 0.2531871497631073,
|
| 455 |
+
"eval_runtime": 18.4185,
|
| 456 |
+
"eval_samples_per_second": 24.649,
|
| 457 |
+
"eval_steps_per_second": 6.189,
|
| 458 |
+
"step": 300
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.68,
|
| 462 |
+
"learning_rate": 3.200883002207506e-05,
|
| 463 |
+
"loss": 0.1757,
|
| 464 |
+
"step": 310
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.68,
|
| 468 |
+
"eval_accuracy": 0.91629958152771,
|
| 469 |
+
"eval_loss": 0.21960243582725525,
|
| 470 |
+
"eval_runtime": 17.2208,
|
| 471 |
+
"eval_samples_per_second": 26.363,
|
| 472 |
+
"eval_steps_per_second": 6.62,
|
| 473 |
+
"step": 310
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.71,
|
| 477 |
+
"learning_rate": 2.980132450331126e-05,
|
| 478 |
+
"loss": 0.2271,
|
| 479 |
+
"step": 320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.71,
|
| 483 |
+
"eval_accuracy": 0.9052863717079163,
|
| 484 |
+
"eval_loss": 0.25458627939224243,
|
| 485 |
+
"eval_runtime": 16.7182,
|
| 486 |
+
"eval_samples_per_second": 27.156,
|
| 487 |
+
"eval_steps_per_second": 6.819,
|
| 488 |
+
"step": 320
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.73,
|
| 492 |
+
"learning_rate": 2.7593818984547465e-05,
|
| 493 |
+
"loss": 0.1546,
|
| 494 |
+
"step": 330
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.73,
|
| 498 |
+
"eval_accuracy": 0.8854625821113586,
|
| 499 |
+
"eval_loss": 0.32153868675231934,
|
| 500 |
+
"eval_runtime": 16.7413,
|
| 501 |
+
"eval_samples_per_second": 27.119,
|
| 502 |
+
"eval_steps_per_second": 6.809,
|
| 503 |
+
"step": 330
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.75,
|
| 507 |
+
"learning_rate": 2.538631346578367e-05,
|
| 508 |
+
"loss": 0.1999,
|
| 509 |
+
"step": 340
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.75,
|
| 513 |
+
"eval_accuracy": 0.9008810520172119,
|
| 514 |
+
"eval_loss": 0.25611090660095215,
|
| 515 |
+
"eval_runtime": 27.5096,
|
| 516 |
+
"eval_samples_per_second": 16.503,
|
| 517 |
+
"eval_steps_per_second": 4.144,
|
| 518 |
+
"step": 340
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.77,
|
| 522 |
+
"learning_rate": 2.317880794701987e-05,
|
| 523 |
+
"loss": 0.2098,
|
| 524 |
+
"step": 350
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.77,
|
| 528 |
+
"eval_accuracy": 0.892070472240448,
|
| 529 |
+
"eval_loss": 0.24251103401184082,
|
| 530 |
+
"eval_runtime": 17.0522,
|
| 531 |
+
"eval_samples_per_second": 26.624,
|
| 532 |
+
"eval_steps_per_second": 6.685,
|
| 533 |
+
"step": 350
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.79,
|
| 537 |
+
"learning_rate": 2.097130242825607e-05,
|
| 538 |
+
"loss": 0.1623,
|
| 539 |
+
"step": 360
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 0.79,
|
| 543 |
+
"eval_accuracy": 0.9052863717079163,
|
| 544 |
+
"eval_loss": 0.23717546463012695,
|
| 545 |
+
"eval_runtime": 16.9552,
|
| 546 |
+
"eval_samples_per_second": 26.776,
|
| 547 |
+
"eval_steps_per_second": 6.724,
|
| 548 |
+
"step": 360
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.82,
|
| 552 |
+
"learning_rate": 1.8763796909492273e-05,
|
| 553 |
+
"loss": 0.2701,
|
| 554 |
+
"step": 370
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.82,
|
| 558 |
+
"eval_accuracy": 0.9229074716567993,
|
| 559 |
+
"eval_loss": 0.24308158457279205,
|
| 560 |
+
"eval_runtime": 17.2488,
|
| 561 |
+
"eval_samples_per_second": 26.321,
|
| 562 |
+
"eval_steps_per_second": 6.609,
|
| 563 |
+
"step": 370
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.84,
|
| 567 |
+
"learning_rate": 1.655629139072848e-05,
|
| 568 |
+
"loss": 0.3512,
|
| 569 |
+
"step": 380
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.84,
|
| 573 |
+
"eval_accuracy": 0.9096916317939758,
|
| 574 |
+
"eval_loss": 0.22931186854839325,
|
| 575 |
+
"eval_runtime": 18.2255,
|
| 576 |
+
"eval_samples_per_second": 24.91,
|
| 577 |
+
"eval_steps_per_second": 6.255,
|
| 578 |
+
"step": 380
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.86,
|
| 582 |
+
"learning_rate": 1.4348785871964682e-05,
|
| 583 |
+
"loss": 0.2043,
|
| 584 |
+
"step": 390
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.86,
|
| 588 |
+
"eval_accuracy": 0.9030836820602417,
|
| 589 |
+
"eval_loss": 0.22873848676681519,
|
| 590 |
+
"eval_runtime": 17.3783,
|
| 591 |
+
"eval_samples_per_second": 26.124,
|
| 592 |
+
"eval_steps_per_second": 6.56,
|
| 593 |
+
"step": 390
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.88,
|
| 597 |
+
"learning_rate": 1.2141280353200883e-05,
|
| 598 |
+
"loss": 0.1794,
|
| 599 |
+
"step": 400
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.88,
|
| 603 |
+
"eval_accuracy": 0.9096916317939758,
|
| 604 |
+
"eval_loss": 0.2241315245628357,
|
| 605 |
+
"eval_runtime": 17.0594,
|
| 606 |
+
"eval_samples_per_second": 26.613,
|
| 607 |
+
"eval_steps_per_second": 6.683,
|
| 608 |
+
"step": 400
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.9,
|
| 612 |
+
"learning_rate": 9.933774834437086e-06,
|
| 613 |
+
"loss": 0.1664,
|
| 614 |
+
"step": 410
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.9,
|
| 618 |
+
"eval_accuracy": 0.9052863717079163,
|
| 619 |
+
"eval_loss": 0.24238458275794983,
|
| 620 |
+
"eval_runtime": 17.031,
|
| 621 |
+
"eval_samples_per_second": 26.657,
|
| 622 |
+
"eval_steps_per_second": 6.694,
|
| 623 |
+
"step": 410
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 0.93,
|
| 627 |
+
"learning_rate": 7.72626931567329e-06,
|
| 628 |
+
"loss": 0.1018,
|
| 629 |
+
"step": 420
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 0.93,
|
| 633 |
+
"eval_accuracy": 0.9052863717079163,
|
| 634 |
+
"eval_loss": 0.23826861381530762,
|
| 635 |
+
"eval_runtime": 18.4005,
|
| 636 |
+
"eval_samples_per_second": 24.673,
|
| 637 |
+
"eval_steps_per_second": 6.195,
|
| 638 |
+
"step": 420
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.95,
|
| 642 |
+
"learning_rate": 5.518763796909492e-06,
|
| 643 |
+
"loss": 0.2466,
|
| 644 |
+
"step": 430
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.95,
|
| 648 |
+
"eval_accuracy": 0.907489001750946,
|
| 649 |
+
"eval_loss": 0.23324483633041382,
|
| 650 |
+
"eval_runtime": 16.9891,
|
| 651 |
+
"eval_samples_per_second": 26.723,
|
| 652 |
+
"eval_steps_per_second": 6.71,
|
| 653 |
+
"step": 430
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.97,
|
| 657 |
+
"learning_rate": 3.3112582781456956e-06,
|
| 658 |
+
"loss": 0.2018,
|
| 659 |
+
"step": 440
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 0.97,
|
| 663 |
+
"eval_accuracy": 0.9140968918800354,
|
| 664 |
+
"eval_loss": 0.2161666452884674,
|
| 665 |
+
"eval_runtime": 17.0773,
|
| 666 |
+
"eval_samples_per_second": 26.585,
|
| 667 |
+
"eval_steps_per_second": 6.676,
|
| 668 |
+
"step": 440
|
| 669 |
+
}
|
| 670 |
+
],
|
| 671 |
+
"max_steps": 453,
|
| 672 |
+
"num_train_epochs": 1,
|
| 673 |
+
"total_flos": 1.0743559886592e+17,
|
| 674 |
+
"trial_name": null,
|
| 675 |
+
"trial_params": null
|
| 676 |
+
}
|