Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +735 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,735 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 99,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.030303030303030304,
|
| 13 |
+
"grad_norm": 1.5466958284378052,
|
| 14 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 15 |
+
"loss": 1.8566,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06060606060606061,
|
| 20 |
+
"grad_norm": 1.5582852363586426,
|
| 21 |
+
"learning_rate": 4.000000000000001e-06,
|
| 22 |
+
"loss": 1.8976,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09090909090909091,
|
| 27 |
+
"grad_norm": 1.524529218673706,
|
| 28 |
+
"learning_rate": 6e-06,
|
| 29 |
+
"loss": 1.7259,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12121212121212122,
|
| 34 |
+
"grad_norm": 1.6974282264709473,
|
| 35 |
+
"learning_rate": 8.000000000000001e-06,
|
| 36 |
+
"loss": 1.814,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.15151515151515152,
|
| 41 |
+
"grad_norm": 1.4745838642120361,
|
| 42 |
+
"learning_rate": 1e-05,
|
| 43 |
+
"loss": 1.5887,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.18181818181818182,
|
| 48 |
+
"grad_norm": 1.522696852684021,
|
| 49 |
+
"learning_rate": 1.2e-05,
|
| 50 |
+
"loss": 1.692,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.21212121212121213,
|
| 55 |
+
"grad_norm": 1.4861280918121338,
|
| 56 |
+
"learning_rate": 1.4e-05,
|
| 57 |
+
"loss": 1.6883,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.24242424242424243,
|
| 62 |
+
"grad_norm": 1.441834807395935,
|
| 63 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 64 |
+
"loss": 1.6722,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2727272727272727,
|
| 69 |
+
"grad_norm": 1.633545160293579,
|
| 70 |
+
"learning_rate": 1.8e-05,
|
| 71 |
+
"loss": 1.6187,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.30303030303030304,
|
| 76 |
+
"grad_norm": 1.495406150817871,
|
| 77 |
+
"learning_rate": 2e-05,
|
| 78 |
+
"loss": 1.5615,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3333333333333333,
|
| 83 |
+
"grad_norm": 1.5129448175430298,
|
| 84 |
+
"learning_rate": 1.9993770622619784e-05,
|
| 85 |
+
"loss": 1.6768,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.36363636363636365,
|
| 90 |
+
"grad_norm": 1.461745023727417,
|
| 91 |
+
"learning_rate": 1.9975090251507637e-05,
|
| 92 |
+
"loss": 1.7377,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3939393939393939,
|
| 97 |
+
"grad_norm": 1.2962136268615723,
|
| 98 |
+
"learning_rate": 1.9943982160079823e-05,
|
| 99 |
+
"loss": 1.7039,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.42424242424242425,
|
| 104 |
+
"grad_norm": 1.5274053812026978,
|
| 105 |
+
"learning_rate": 1.9900485105144544e-05,
|
| 106 |
+
"loss": 1.8094,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.45454545454545453,
|
| 111 |
+
"grad_norm": 1.8526736497879028,
|
| 112 |
+
"learning_rate": 1.9844653278615836e-05,
|
| 113 |
+
"loss": 1.7952,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.48484848484848486,
|
| 118 |
+
"grad_norm": 1.5075675249099731,
|
| 119 |
+
"learning_rate": 1.9776556239997146e-05,
|
| 120 |
+
"loss": 1.7614,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5151515151515151,
|
| 125 |
+
"grad_norm": 1.3172680139541626,
|
| 126 |
+
"learning_rate": 1.9696278829718882e-05,
|
| 127 |
+
"loss": 1.8214,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5454545454545454,
|
| 132 |
+
"grad_norm": 1.2293428182601929,
|
| 133 |
+
"learning_rate": 1.9603921063437795e-05,
|
| 134 |
+
"loss": 1.688,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5757575757575758,
|
| 139 |
+
"grad_norm": 1.6465842723846436,
|
| 140 |
+
"learning_rate": 1.949959800742991e-05,
|
| 141 |
+
"loss": 1.7641,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6060606060606061,
|
| 146 |
+
"grad_norm": 1.2794415950775146,
|
| 147 |
+
"learning_rate": 1.9383439635232296e-05,
|
| 148 |
+
"loss": 1.7237,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6363636363636364,
|
| 153 |
+
"grad_norm": 1.5035333633422852,
|
| 154 |
+
"learning_rate": 1.9255590665712214e-05,
|
| 155 |
+
"loss": 1.7502,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6666666666666666,
|
| 160 |
+
"grad_norm": 1.6453092098236084,
|
| 161 |
+
"learning_rate": 1.911621038276542e-05,
|
| 162 |
+
"loss": 1.4906,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.696969696969697,
|
| 167 |
+
"grad_norm": 1.51984703540802,
|
| 168 |
+
"learning_rate": 1.8965472436868288e-05,
|
| 169 |
+
"loss": 1.5359,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7272727272727273,
|
| 174 |
+
"grad_norm": 1.3503923416137695,
|
| 175 |
+
"learning_rate": 1.8803564628730916e-05,
|
| 176 |
+
"loss": 1.5969,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7575757575757576,
|
| 181 |
+
"grad_norm": 1.3228968381881714,
|
| 182 |
+
"learning_rate": 1.8630688675320844e-05,
|
| 183 |
+
"loss": 1.7511,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.7878787878787878,
|
| 188 |
+
"grad_norm": 1.2232959270477295,
|
| 189 |
+
"learning_rate": 1.8447059958548822e-05,
|
| 190 |
+
"loss": 1.6303,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8181818181818182,
|
| 195 |
+
"grad_norm": 1.2176752090454102,
|
| 196 |
+
"learning_rate": 1.8252907256929777e-05,
|
| 197 |
+
"loss": 1.6744,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.8484848484848485,
|
| 202 |
+
"grad_norm": 1.6348463296890259,
|
| 203 |
+
"learning_rate": 1.804847246055326e-05,
|
| 204 |
+
"loss": 1.7386,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.8787878787878788,
|
| 209 |
+
"grad_norm": 1.3479255437850952,
|
| 210 |
+
"learning_rate": 1.7834010269718526e-05,
|
| 211 |
+
"loss": 1.8378,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9090909090909091,
|
| 216 |
+
"grad_norm": 1.4100639820098877,
|
| 217 |
+
"learning_rate": 1.7609787877609678e-05,
|
| 218 |
+
"loss": 1.6901,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.9393939393939394,
|
| 223 |
+
"grad_norm": 1.289868950843811,
|
| 224 |
+
"learning_rate": 1.7376084637406222e-05,
|
| 225 |
+
"loss": 1.7371,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9696969696969697,
|
| 230 |
+
"grad_norm": 1.2278428077697754,
|
| 231 |
+
"learning_rate": 1.7133191714243805e-05,
|
| 232 |
+
"loss": 1.5087,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.0,
|
| 237 |
+
"grad_norm": 1.5690200328826904,
|
| 238 |
+
"learning_rate": 1.6881411722458688e-05,
|
| 239 |
+
"loss": 1.4644,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.0303030303030303,
|
| 244 |
+
"grad_norm": 1.7996729612350464,
|
| 245 |
+
"learning_rate": 1.6621058348568008e-05,
|
| 246 |
+
"loss": 1.2771,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.0606060606060606,
|
| 251 |
+
"grad_norm": 1.6773903369903564,
|
| 252 |
+
"learning_rate": 1.6352455960455385e-05,
|
| 253 |
+
"loss": 1.324,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.0909090909090908,
|
| 258 |
+
"grad_norm": 1.6019961833953857,
|
| 259 |
+
"learning_rate": 1.607593920324899e-05,
|
| 260 |
+
"loss": 1.2817,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.121212121212121,
|
| 265 |
+
"grad_norm": 1.3165000677108765,
|
| 266 |
+
"learning_rate": 1.5791852582395334e-05,
|
| 267 |
+
"loss": 1.1927,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.1515151515151516,
|
| 272 |
+
"grad_norm": 1.455589771270752,
|
| 273 |
+
"learning_rate": 1.5500550034448415e-05,
|
| 274 |
+
"loss": 1.2133,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.1818181818181819,
|
| 279 |
+
"grad_norm": 1.8257672786712646,
|
| 280 |
+
"learning_rate": 1.5202394486108823e-05,
|
| 281 |
+
"loss": 1.1531,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2121212121212122,
|
| 286 |
+
"grad_norm": 1.873024821281433,
|
| 287 |
+
"learning_rate": 1.4897757402062285e-05,
|
| 288 |
+
"loss": 1.2731,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.2424242424242424,
|
| 293 |
+
"grad_norm": 1.7878319025039673,
|
| 294 |
+
"learning_rate": 1.4587018322180906e-05,
|
| 295 |
+
"loss": 1.2247,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.2727272727272727,
|
| 300 |
+
"grad_norm": 1.2891494035720825,
|
| 301 |
+
"learning_rate": 1.4270564388663761e-05,
|
| 302 |
+
"loss": 1.0763,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.303030303030303,
|
| 307 |
+
"grad_norm": 1.4569677114486694,
|
| 308 |
+
"learning_rate": 1.3948789863705914e-05,
|
| 309 |
+
"loss": 1.1027,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.3333333333333333,
|
| 314 |
+
"grad_norm": 1.5054666996002197,
|
| 315 |
+
"learning_rate": 1.3622095638296827e-05,
|
| 316 |
+
"loss": 1.1478,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.3636363636363638,
|
| 321 |
+
"grad_norm": 1.4277585744857788,
|
| 322 |
+
"learning_rate": 1.32908887327601e-05,
|
| 323 |
+
"loss": 1.1792,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.393939393939394,
|
| 328 |
+
"grad_norm": 1.3641875982284546,
|
| 329 |
+
"learning_rate": 1.2955581789656844e-05,
|
| 330 |
+
"loss": 1.0412,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.4242424242424243,
|
| 335 |
+
"grad_norm": 1.706855297088623,
|
| 336 |
+
"learning_rate": 1.2616592559684408e-05,
|
| 337 |
+
"loss": 1.1198,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.4545454545454546,
|
| 342 |
+
"grad_norm": 1.8636873960494995,
|
| 343 |
+
"learning_rate": 1.2274343381211067e-05,
|
| 344 |
+
"loss": 1.0971,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.4848484848484849,
|
| 349 |
+
"grad_norm": 1.6042084693908691,
|
| 350 |
+
"learning_rate": 1.192926065409497e-05,
|
| 351 |
+
"loss": 1.0331,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.5151515151515151,
|
| 356 |
+
"grad_norm": 1.6517517566680908,
|
| 357 |
+
"learning_rate": 1.1581774308443042e-05,
|
| 358 |
+
"loss": 1.1258,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.5454545454545454,
|
| 363 |
+
"grad_norm": 1.7735017538070679,
|
| 364 |
+
"learning_rate": 1.1232317268971586e-05,
|
| 365 |
+
"loss": 1.211,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.5757575757575757,
|
| 370 |
+
"grad_norm": 1.4876844882965088,
|
| 371 |
+
"learning_rate": 1.088132491563602e-05,
|
| 372 |
+
"loss": 1.0778,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.606060606060606,
|
| 377 |
+
"grad_norm": 1.4725524187088013,
|
| 378 |
+
"learning_rate": 1.0529234541201631e-05,
|
| 379 |
+
"loss": 1.1482,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6363636363636362,
|
| 384 |
+
"grad_norm": 1.6460925340652466,
|
| 385 |
+
"learning_rate": 1.0176484806431288e-05,
|
| 386 |
+
"loss": 1.0857,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.6666666666666665,
|
| 391 |
+
"grad_norm": 1.6207877397537231,
|
| 392 |
+
"learning_rate": 9.823515193568715e-06,
|
| 393 |
+
"loss": 1.1458,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.696969696969697,
|
| 398 |
+
"grad_norm": 1.5556446313858032,
|
| 399 |
+
"learning_rate": 9.470765458798369e-06,
|
| 400 |
+
"loss": 0.9578,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.7272727272727273,
|
| 405 |
+
"grad_norm": 1.459346890449524,
|
| 406 |
+
"learning_rate": 9.118675084363986e-06,
|
| 407 |
+
"loss": 1.1004,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.7575757575757576,
|
| 412 |
+
"grad_norm": 1.6329230070114136,
|
| 413 |
+
"learning_rate": 8.767682731028415e-06,
|
| 414 |
+
"loss": 1.0838,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.7878787878787878,
|
| 419 |
+
"grad_norm": 1.4568471908569336,
|
| 420 |
+
"learning_rate": 8.418225691556962e-06,
|
| 421 |
+
"loss": 0.9716,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.8181818181818183,
|
| 426 |
+
"grad_norm": 1.6114221811294556,
|
| 427 |
+
"learning_rate": 8.070739345905032e-06,
|
| 428 |
+
"loss": 1.2118,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.8484848484848486,
|
| 433 |
+
"grad_norm": 1.4424853324890137,
|
| 434 |
+
"learning_rate": 7.725656618788938e-06,
|
| 435 |
+
"loss": 1.0933,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.878787878787879,
|
| 440 |
+
"grad_norm": 1.3055541515350342,
|
| 441 |
+
"learning_rate": 7.383407440315595e-06,
|
| 442 |
+
"loss": 1.054,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.9090909090909092,
|
| 447 |
+
"grad_norm": 1.608759880065918,
|
| 448 |
+
"learning_rate": 7.044418210343161e-06,
|
| 449 |
+
"loss": 1.2112,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.9393939393939394,
|
| 454 |
+
"grad_norm": 1.381705403327942,
|
| 455 |
+
"learning_rate": 6.7091112672399e-06,
|
| 456 |
+
"loss": 1.06,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.9696969696969697,
|
| 461 |
+
"grad_norm": 1.5174615383148193,
|
| 462 |
+
"learning_rate": 6.3779043617031775e-06,
|
| 463 |
+
"loss": 1.1465,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.0,
|
| 468 |
+
"grad_norm": 1.7489651441574097,
|
| 469 |
+
"learning_rate": 6.051210136294089e-06,
|
| 470 |
+
"loss": 0.8968,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.0303030303030303,
|
| 475 |
+
"grad_norm": 1.7633519172668457,
|
| 476 |
+
"learning_rate": 5.729435611336239e-06,
|
| 477 |
+
"loss": 0.8801,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.0606060606060606,
|
| 482 |
+
"grad_norm": 2.278275728225708,
|
| 483 |
+
"learning_rate": 5.412981677819094e-06,
|
| 484 |
+
"loss": 0.8784,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.090909090909091,
|
| 489 |
+
"grad_norm": 1.8437879085540771,
|
| 490 |
+
"learning_rate": 5.1022425979377174e-06,
|
| 491 |
+
"loss": 0.7264,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.121212121212121,
|
| 496 |
+
"grad_norm": 1.5237785577774048,
|
| 497 |
+
"learning_rate": 4.797605513891179e-06,
|
| 498 |
+
"loss": 0.758,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.1515151515151514,
|
| 503 |
+
"grad_norm": 1.7430484294891357,
|
| 504 |
+
"learning_rate": 4.4994499655515865e-06,
|
| 505 |
+
"loss": 0.8087,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.1818181818181817,
|
| 510 |
+
"grad_norm": 1.7282936573028564,
|
| 511 |
+
"learning_rate": 4.208147417604665e-06,
|
| 512 |
+
"loss": 0.7311,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.212121212121212,
|
| 517 |
+
"grad_norm": 1.907755970954895,
|
| 518 |
+
"learning_rate": 3.924060796751012e-06,
|
| 519 |
+
"loss": 0.7755,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.242424242424242,
|
| 524 |
+
"grad_norm": 2.1145212650299072,
|
| 525 |
+
"learning_rate": 3.647544039544615e-06,
|
| 526 |
+
"loss": 0.7668,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.2727272727272725,
|
| 531 |
+
"grad_norm": 2.2172389030456543,
|
| 532 |
+
"learning_rate": 3.378941651431996e-06,
|
| 533 |
+
"loss": 0.7701,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.303030303030303,
|
| 538 |
+
"grad_norm": 2.0913803577423096,
|
| 539 |
+
"learning_rate": 3.1185882775413123e-06,
|
| 540 |
+
"loss": 0.889,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.3333333333333335,
|
| 545 |
+
"grad_norm": 2.1478118896484375,
|
| 546 |
+
"learning_rate": 2.8668082857562006e-06,
|
| 547 |
+
"loss": 0.7411,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.3636363636363638,
|
| 552 |
+
"grad_norm": 2.0034427642822266,
|
| 553 |
+
"learning_rate": 2.6239153625937786e-06,
|
| 554 |
+
"loss": 0.7564,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.393939393939394,
|
| 559 |
+
"grad_norm": 1.6710702180862427,
|
| 560 |
+
"learning_rate": 2.390212122390323e-06,
|
| 561 |
+
"loss": 0.7695,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.4242424242424243,
|
| 566 |
+
"grad_norm": 1.7885438203811646,
|
| 567 |
+
"learning_rate": 2.165989730281475e-06,
|
| 568 |
+
"loss": 0.8017,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.4545454545454546,
|
| 573 |
+
"grad_norm": 1.3610081672668457,
|
| 574 |
+
"learning_rate": 1.9515275394467446e-06,
|
| 575 |
+
"loss": 0.7561,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.484848484848485,
|
| 580 |
+
"grad_norm": 1.645119547843933,
|
| 581 |
+
"learning_rate": 1.7470927430702277e-06,
|
| 582 |
+
"loss": 0.6155,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.515151515151515,
|
| 587 |
+
"grad_norm": 1.7851331233978271,
|
| 588 |
+
"learning_rate": 1.5529400414511809e-06,
|
| 589 |
+
"loss": 0.6894,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.5454545454545454,
|
| 594 |
+
"grad_norm": 1.4505351781845093,
|
| 595 |
+
"learning_rate": 1.369311324679159e-06,
|
| 596 |
+
"loss": 0.8904,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.5757575757575757,
|
| 601 |
+
"grad_norm": 1.663177728652954,
|
| 602 |
+
"learning_rate": 1.196435371269089e-06,
|
| 603 |
+
"loss": 0.8289,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.606060606060606,
|
| 608 |
+
"grad_norm": 1.7335422039031982,
|
| 609 |
+
"learning_rate": 1.0345275631317165e-06,
|
| 610 |
+
"loss": 0.7425,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.6363636363636362,
|
| 615 |
+
"grad_norm": 1.3842012882232666,
|
| 616 |
+
"learning_rate": 8.837896172345827e-07,
|
| 617 |
+
"loss": 0.7578,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.6666666666666665,
|
| 622 |
+
"grad_norm": 1.770256757736206,
|
| 623 |
+
"learning_rate": 7.4440933428779e-07,
|
| 624 |
+
"loss": 0.6974,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.6969696969696972,
|
| 629 |
+
"grad_norm": 1.5890138149261475,
|
| 630 |
+
"learning_rate": 6.165603647677054e-07,
|
| 631 |
+
"loss": 0.7797,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.7272727272727275,
|
| 636 |
+
"grad_norm": 1.6013965606689453,
|
| 637 |
+
"learning_rate": 5.004019925700921e-07,
|
| 638 |
+
"loss": 0.7053,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.757575757575758,
|
| 643 |
+
"grad_norm": 1.4727004766464233,
|
| 644 |
+
"learning_rate": 3.960789365622075e-07,
|
| 645 |
+
"loss": 0.7264,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.787878787878788,
|
| 650 |
+
"grad_norm": 1.6595274209976196,
|
| 651 |
+
"learning_rate": 3.0372117028111825e-07,
|
| 652 |
+
"loss": 0.7089,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.8181818181818183,
|
| 657 |
+
"grad_norm": 1.8651310205459595,
|
| 658 |
+
"learning_rate": 2.2344376000285606e-07,
|
| 659 |
+
"loss": 0.6721,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.8484848484848486,
|
| 664 |
+
"grad_norm": 1.7887459993362427,
|
| 665 |
+
"learning_rate": 1.553467213841664e-07,
|
| 666 |
+
"loss": 0.788,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.878787878787879,
|
| 671 |
+
"grad_norm": 1.454445719718933,
|
| 672 |
+
"learning_rate": 9.951489485545696e-08,
|
| 673 |
+
"loss": 0.6617,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.909090909090909,
|
| 678 |
+
"grad_norm": 1.662096619606018,
|
| 679 |
+
"learning_rate": 5.6017839920180506e-08,
|
| 680 |
+
"loss": 0.7603,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.9393939393939394,
|
| 685 |
+
"grad_norm": 1.6245455741882324,
|
| 686 |
+
"learning_rate": 2.4909748492362162e-08,
|
| 687 |
+
"loss": 0.7336,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.9696969696969697,
|
| 692 |
+
"grad_norm": 1.5725836753845215,
|
| 693 |
+
"learning_rate": 6.229377380218005e-09,
|
| 694 |
+
"loss": 0.7566,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.0,
|
| 699 |
+
"grad_norm": 1.6165506839752197,
|
| 700 |
+
"learning_rate": 0.0,
|
| 701 |
+
"loss": 0.6936,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.0,
|
| 706 |
+
"step": 99,
|
| 707 |
+
"total_flos": 2.3659576070189875e+17,
|
| 708 |
+
"train_loss": 1.195338128191052,
|
| 709 |
+
"train_runtime": 1048.1424,
|
| 710 |
+
"train_samples_per_second": 2.945,
|
| 711 |
+
"train_steps_per_second": 0.094
|
| 712 |
+
}
|
| 713 |
+
],
|
| 714 |
+
"logging_steps": 1,
|
| 715 |
+
"max_steps": 99,
|
| 716 |
+
"num_input_tokens_seen": 0,
|
| 717 |
+
"num_train_epochs": 3,
|
| 718 |
+
"save_steps": 0,
|
| 719 |
+
"stateful_callbacks": {
|
| 720 |
+
"TrainerControl": {
|
| 721 |
+
"args": {
|
| 722 |
+
"should_epoch_stop": false,
|
| 723 |
+
"should_evaluate": false,
|
| 724 |
+
"should_log": false,
|
| 725 |
+
"should_save": true,
|
| 726 |
+
"should_training_stop": true
|
| 727 |
+
},
|
| 728 |
+
"attributes": {}
|
| 729 |
+
}
|
| 730 |
+
},
|
| 731 |
+
"total_flos": 2.3659576070189875e+17,
|
| 732 |
+
"train_batch_size": 8,
|
| 733 |
+
"trial_name": null,
|
| 734 |
+
"trial_params": null
|
| 735 |
+
}
|