Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +730 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,730 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 476,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.004201680672268907,
|
| 13 |
+
"grad_norm": 6.843970053848994,
|
| 14 |
+
"learning_rate": 4.1666666666666667e-07,
|
| 15 |
+
"loss": 1.4834,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02100840336134454,
|
| 20 |
+
"grad_norm": 5.888026413386213,
|
| 21 |
+
"learning_rate": 2.0833333333333334e-06,
|
| 22 |
+
"loss": 1.4466,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04201680672268908,
|
| 27 |
+
"grad_norm": 1.672754748671725,
|
| 28 |
+
"learning_rate": 4.166666666666667e-06,
|
| 29 |
+
"loss": 1.29,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06302521008403361,
|
| 34 |
+
"grad_norm": 1.2549304333706048,
|
| 35 |
+
"learning_rate": 6.25e-06,
|
| 36 |
+
"loss": 1.265,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08403361344537816,
|
| 41 |
+
"grad_norm": 1.0804048514467623,
|
| 42 |
+
"learning_rate": 8.333333333333334e-06,
|
| 43 |
+
"loss": 1.218,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.10504201680672269,
|
| 48 |
+
"grad_norm": 1.0012856727023651,
|
| 49 |
+
"learning_rate": 1.0416666666666668e-05,
|
| 50 |
+
"loss": 1.1874,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12605042016806722,
|
| 55 |
+
"grad_norm": 1.0428770541540702,
|
| 56 |
+
"learning_rate": 1.25e-05,
|
| 57 |
+
"loss": 1.1738,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14705882352941177,
|
| 62 |
+
"grad_norm": 0.8971861131120643,
|
| 63 |
+
"learning_rate": 1.4583333333333333e-05,
|
| 64 |
+
"loss": 1.1823,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.16806722689075632,
|
| 69 |
+
"grad_norm": 0.8761555093987056,
|
| 70 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 71 |
+
"loss": 1.1509,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.18907563025210083,
|
| 76 |
+
"grad_norm": 0.9022691166860466,
|
| 77 |
+
"learning_rate": 1.8750000000000002e-05,
|
| 78 |
+
"loss": 1.1738,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.21008403361344538,
|
| 83 |
+
"grad_norm": 0.8595698667749491,
|
| 84 |
+
"learning_rate": 1.9998922457512608e-05,
|
| 85 |
+
"loss": 1.1827,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.23109243697478993,
|
| 90 |
+
"grad_norm": 0.8131261419160233,
|
| 91 |
+
"learning_rate": 1.9986802771267902e-05,
|
| 92 |
+
"loss": 1.1595,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.25210084033613445,
|
| 97 |
+
"grad_norm": 0.9352189654627542,
|
| 98 |
+
"learning_rate": 1.996123284790336e-05,
|
| 99 |
+
"loss": 1.1751,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.27310924369747897,
|
| 104 |
+
"grad_norm": 0.8183133797776339,
|
| 105 |
+
"learning_rate": 1.9922247125020307e-05,
|
| 106 |
+
"loss": 1.1368,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.29411764705882354,
|
| 111 |
+
"grad_norm": 0.7919702228536482,
|
| 112 |
+
"learning_rate": 1.9869898108633834e-05,
|
| 113 |
+
"loss": 1.1576,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.31512605042016806,
|
| 118 |
+
"grad_norm": 0.809220998201082,
|
| 119 |
+
"learning_rate": 1.9804256302457653e-05,
|
| 120 |
+
"loss": 1.1214,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.33613445378151263,
|
| 125 |
+
"grad_norm": 0.7934942224774806,
|
| 126 |
+
"learning_rate": 1.972541011294959e-05,
|
| 127 |
+
"loss": 1.1665,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.35714285714285715,
|
| 132 |
+
"grad_norm": 0.8016460068505575,
|
| 133 |
+
"learning_rate": 1.963346573024568e-05,
|
| 134 |
+
"loss": 1.14,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.37815126050420167,
|
| 139 |
+
"grad_norm": 0.7959323509300636,
|
| 140 |
+
"learning_rate": 1.952854698514318e-05,
|
| 141 |
+
"loss": 1.1427,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.39915966386554624,
|
| 146 |
+
"grad_norm": 0.834564941958375,
|
| 147 |
+
"learning_rate": 1.9410795182325113e-05,
|
| 148 |
+
"loss": 1.1582,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.42016806722689076,
|
| 153 |
+
"grad_norm": 0.8021887426628376,
|
| 154 |
+
"learning_rate": 1.9280368910050943e-05,
|
| 155 |
+
"loss": 1.1425,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.4411764705882353,
|
| 160 |
+
"grad_norm": 0.8535114934490843,
|
| 161 |
+
"learning_rate": 1.9137443826569758e-05,
|
| 162 |
+
"loss": 1.1507,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.46218487394957986,
|
| 167 |
+
"grad_norm": 0.8097650803240342,
|
| 168 |
+
"learning_rate": 1.898221242354353e-05,
|
| 169 |
+
"loss": 1.1376,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4831932773109244,
|
| 174 |
+
"grad_norm": 0.854806698328117,
|
| 175 |
+
"learning_rate": 1.881488376679912e-05,
|
| 176 |
+
"loss": 1.1538,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.5042016806722689,
|
| 181 |
+
"grad_norm": 0.7411750487736666,
|
| 182 |
+
"learning_rate": 1.8635683214758213e-05,
|
| 183 |
+
"loss": 1.1548,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.5252100840336135,
|
| 188 |
+
"grad_norm": 0.7470899750611266,
|
| 189 |
+
"learning_rate": 1.8444852114924325e-05,
|
| 190 |
+
"loss": 1.1285,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.5462184873949579,
|
| 195 |
+
"grad_norm": 0.7855863245637623,
|
| 196 |
+
"learning_rate": 1.8242647478835717e-05,
|
| 197 |
+
"loss": 1.1428,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5672268907563025,
|
| 202 |
+
"grad_norm": 0.74562538616598,
|
| 203 |
+
"learning_rate": 1.8029341635921985e-05,
|
| 204 |
+
"loss": 1.1303,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5882352941176471,
|
| 209 |
+
"grad_norm": 0.751956755942676,
|
| 210 |
+
"learning_rate": 1.780522186673046e-05,
|
| 211 |
+
"loss": 1.157,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.6092436974789915,
|
| 216 |
+
"grad_norm": 0.7490360146750048,
|
| 217 |
+
"learning_rate": 1.7570590016016455e-05,
|
| 218 |
+
"loss": 1.1245,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.6302521008403361,
|
| 223 |
+
"grad_norm": 0.7962243135552697,
|
| 224 |
+
"learning_rate": 1.7325762086218415e-05,
|
| 225 |
+
"loss": 1.1577,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.6512605042016807,
|
| 230 |
+
"grad_norm": 0.7647261906536346,
|
| 231 |
+
"learning_rate": 1.7071067811865477e-05,
|
| 232 |
+
"loss": 1.1378,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.6722689075630253,
|
| 237 |
+
"grad_norm": 0.7569274861307836,
|
| 238 |
+
"learning_rate": 1.680685021549063e-05,
|
| 239 |
+
"loss": 1.1249,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6932773109243697,
|
| 244 |
+
"grad_norm": 0.7606095560211281,
|
| 245 |
+
"learning_rate": 1.6533465145647598e-05,
|
| 246 |
+
"loss": 1.1405,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.7142857142857143,
|
| 251 |
+
"grad_norm": 0.7797117802105004,
|
| 252 |
+
"learning_rate": 1.6251280797653606e-05,
|
| 253 |
+
"loss": 1.1432,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.7352941176470589,
|
| 258 |
+
"grad_norm": 0.7411898433127342,
|
| 259 |
+
"learning_rate": 1.5960677217703512e-05,
|
| 260 |
+
"loss": 1.1341,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.7563025210084033,
|
| 265 |
+
"grad_norm": 0.7937630835311652,
|
| 266 |
+
"learning_rate": 1.566204579102317e-05,
|
| 267 |
+
"loss": 1.1771,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.7773109243697479,
|
| 272 |
+
"grad_norm": 0.7831403718494921,
|
| 273 |
+
"learning_rate": 1.5355788714751378e-05,
|
| 274 |
+
"loss": 1.1346,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.7983193277310925,
|
| 279 |
+
"grad_norm": 0.7670962454375009,
|
| 280 |
+
"learning_rate": 1.5042318456260305e-05,
|
| 281 |
+
"loss": 1.1411,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.819327731092437,
|
| 286 |
+
"grad_norm": 0.8187182133488576,
|
| 287 |
+
"learning_rate": 1.4722057197643986e-05,
|
| 288 |
+
"loss": 1.1287,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.8403361344537815,
|
| 293 |
+
"grad_norm": 0.7707658033677821,
|
| 294 |
+
"learning_rate": 1.4395436267123017e-05,
|
| 295 |
+
"loss": 1.1686,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.8613445378151261,
|
| 300 |
+
"grad_norm": 0.7929027157964723,
|
| 301 |
+
"learning_rate": 1.40628955581312e-05,
|
| 302 |
+
"loss": 1.1095,
|
| 303 |
+
"step": 205
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.8823529411764706,
|
| 307 |
+
"grad_norm": 0.7671917924588447,
|
| 308 |
+
"learning_rate": 1.3724882936866596e-05,
|
| 309 |
+
"loss": 1.1532,
|
| 310 |
+
"step": 210
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.9033613445378151,
|
| 314 |
+
"grad_norm": 0.743350743534685,
|
| 315 |
+
"learning_rate": 1.3381853639104817e-05,
|
| 316 |
+
"loss": 1.1177,
|
| 317 |
+
"step": 215
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.9243697478991597,
|
| 321 |
+
"grad_norm": 0.7994542205340992,
|
| 322 |
+
"learning_rate": 1.3034269657086993e-05,
|
| 323 |
+
"loss": 1.0984,
|
| 324 |
+
"step": 220
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.9453781512605042,
|
| 328 |
+
"grad_norm": 0.8028923468414233,
|
| 329 |
+
"learning_rate": 1.2682599117308083e-05,
|
| 330 |
+
"loss": 1.1225,
|
| 331 |
+
"step": 225
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.9663865546218487,
|
| 335 |
+
"grad_norm": 0.7701315329198188,
|
| 336 |
+
"learning_rate": 1.2327315650043605e-05,
|
| 337 |
+
"loss": 1.1225,
|
| 338 |
+
"step": 230
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.9873949579831933,
|
| 342 |
+
"grad_norm": 0.749625186589024,
|
| 343 |
+
"learning_rate": 1.1968897751463841e-05,
|
| 344 |
+
"loss": 1.1205,
|
| 345 |
+
"step": 235
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.0,
|
| 349 |
+
"eval_loss": 1.122010350227356,
|
| 350 |
+
"eval_runtime": 4.1604,
|
| 351 |
+
"eval_samples_per_second": 35.574,
|
| 352 |
+
"eval_steps_per_second": 0.721,
|
| 353 |
+
"step": 238
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.0084033613445378,
|
| 357 |
+
"grad_norm": 0.8685596492224139,
|
| 358 |
+
"learning_rate": 1.1607828139194683e-05,
|
| 359 |
+
"loss": 1.0392,
|
| 360 |
+
"step": 240
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.0294117647058822,
|
| 364 |
+
"grad_norm": 0.8195681790245358,
|
| 365 |
+
"learning_rate": 1.1244593102192961e-05,
|
| 366 |
+
"loss": 0.8804,
|
| 367 |
+
"step": 245
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 1.050420168067227,
|
| 371 |
+
"grad_norm": 0.7862106935161252,
|
| 372 |
+
"learning_rate": 1.0879681845811964e-05,
|
| 373 |
+
"loss": 0.8759,
|
| 374 |
+
"step": 250
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.0714285714285714,
|
| 378 |
+
"grad_norm": 0.7695442905422751,
|
| 379 |
+
"learning_rate": 1.0513585832939104e-05,
|
| 380 |
+
"loss": 0.8921,
|
| 381 |
+
"step": 255
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.092436974789916,
|
| 385 |
+
"grad_norm": 0.7197984706671763,
|
| 386 |
+
"learning_rate": 1.0146798122093167e-05,
|
| 387 |
+
"loss": 0.8465,
|
| 388 |
+
"step": 260
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.1134453781512605,
|
| 392 |
+
"grad_norm": 0.7436730962206078,
|
| 393 |
+
"learning_rate": 9.779812703372538e-06,
|
| 394 |
+
"loss": 0.8983,
|
| 395 |
+
"step": 265
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.134453781512605,
|
| 399 |
+
"grad_norm": 0.7419938040761785,
|
| 400 |
+
"learning_rate": 9.41312383314878e-06,
|
| 401 |
+
"loss": 0.86,
|
| 402 |
+
"step": 270
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.1554621848739495,
|
| 406 |
+
"grad_norm": 0.7727780995520752,
|
| 407 |
+
"learning_rate": 9.047225368401622e-06,
|
| 408 |
+
"loss": 0.8751,
|
| 409 |
+
"step": 275
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.1764705882352942,
|
| 413 |
+
"grad_norm": 0.7184055924425042,
|
| 414 |
+
"learning_rate": 8.682610101591813e-06,
|
| 415 |
+
"loss": 0.8633,
|
| 416 |
+
"step": 280
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.1974789915966386,
|
| 420 |
+
"grad_norm": 0.7424505911824795,
|
| 421 |
+
"learning_rate": 8.319769096967681e-06,
|
| 422 |
+
"loss": 0.881,
|
| 423 |
+
"step": 285
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.2184873949579833,
|
| 427 |
+
"grad_norm": 0.7831013281249656,
|
| 428 |
+
"learning_rate": 7.95919102919926e-06,
|
| 429 |
+
"loss": 0.8758,
|
| 430 |
+
"step": 290
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.2394957983193278,
|
| 434 |
+
"grad_norm": 0.7218415137209162,
|
| 435 |
+
"learning_rate": 7.601361525230713e-06,
|
| 436 |
+
"loss": 0.848,
|
| 437 |
+
"step": 295
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.2605042016806722,
|
| 441 |
+
"grad_norm": 0.727646718952851,
|
| 442 |
+
"learning_rate": 7.246762510237404e-06,
|
| 443 |
+
"loss": 0.8928,
|
| 444 |
+
"step": 300
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.2815126050420167,
|
| 448 |
+
"grad_norm": 0.7318861363154572,
|
| 449 |
+
"learning_rate": 6.89587155856853e-06,
|
| 450 |
+
"loss": 0.8861,
|
| 451 |
+
"step": 305
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.3025210084033614,
|
| 455 |
+
"grad_norm": 0.7850969062028121,
|
| 456 |
+
"learning_rate": 6.549161250549474e-06,
|
| 457 |
+
"loss": 0.8785,
|
| 458 |
+
"step": 310
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.3235294117647058,
|
| 462 |
+
"grad_norm": 0.7424275388998128,
|
| 463 |
+
"learning_rate": 6.207098536010083e-06,
|
| 464 |
+
"loss": 0.8732,
|
| 465 |
+
"step": 315
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.3445378151260505,
|
| 469 |
+
"grad_norm": 0.7785063907714577,
|
| 470 |
+
"learning_rate": 5.8701441053961185e-06,
|
| 471 |
+
"loss": 0.8923,
|
| 472 |
+
"step": 320
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.365546218487395,
|
| 476 |
+
"grad_norm": 0.7101355435294052,
|
| 477 |
+
"learning_rate": 5.538751769310842e-06,
|
| 478 |
+
"loss": 0.8756,
|
| 479 |
+
"step": 325
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.3865546218487395,
|
| 483 |
+
"grad_norm": 0.7360750873465238,
|
| 484 |
+
"learning_rate": 5.213367847322408e-06,
|
| 485 |
+
"loss": 0.8657,
|
| 486 |
+
"step": 330
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.407563025210084,
|
| 490 |
+
"grad_norm": 0.7493101168945231,
|
| 491 |
+
"learning_rate": 4.894430566860144e-06,
|
| 492 |
+
"loss": 0.8565,
|
| 493 |
+
"step": 335
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.4285714285714286,
|
| 497 |
+
"grad_norm": 0.7246837540339062,
|
| 498 |
+
"learning_rate": 4.58236947300939e-06,
|
| 499 |
+
"loss": 0.867,
|
| 500 |
+
"step": 340
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.449579831932773,
|
| 504 |
+
"grad_norm": 0.7052644676750524,
|
| 505 |
+
"learning_rate": 4.277604849999666e-06,
|
| 506 |
+
"loss": 0.8594,
|
| 507 |
+
"step": 345
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.4705882352941178,
|
| 511 |
+
"grad_norm": 0.7349952399246719,
|
| 512 |
+
"learning_rate": 3.980547155165429e-06,
|
| 513 |
+
"loss": 0.8533,
|
| 514 |
+
"step": 350
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.4915966386554622,
|
| 518 |
+
"grad_norm": 0.7284230275061265,
|
| 519 |
+
"learning_rate": 3.691596466141666e-06,
|
| 520 |
+
"loss": 0.9056,
|
| 521 |
+
"step": 355
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.5126050420168067,
|
| 525 |
+
"grad_norm": 0.7193222649134673,
|
| 526 |
+
"learning_rate": 3.4111419420388904e-06,
|
| 527 |
+
"loss": 0.8662,
|
| 528 |
+
"step": 360
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.5336134453781511,
|
| 532 |
+
"grad_norm": 0.731566479431413,
|
| 533 |
+
"learning_rate": 3.139561299323206e-06,
|
| 534 |
+
"loss": 0.8587,
|
| 535 |
+
"step": 365
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.5546218487394958,
|
| 539 |
+
"grad_norm": 0.7391733181416821,
|
| 540 |
+
"learning_rate": 2.877220303107373e-06,
|
| 541 |
+
"loss": 0.8602,
|
| 542 |
+
"step": 370
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.5756302521008403,
|
| 546 |
+
"grad_norm": 0.7390429200298151,
|
| 547 |
+
"learning_rate": 2.624472274537925e-06,
|
| 548 |
+
"loss": 0.8701,
|
| 549 |
+
"step": 375
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.596638655462185,
|
| 553 |
+
"grad_norm": 0.741782430807988,
|
| 554 |
+
"learning_rate": 2.381657614941858e-06,
|
| 555 |
+
"loss": 0.8588,
|
| 556 |
+
"step": 380
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.6176470588235294,
|
| 560 |
+
"grad_norm": 0.7082521628110617,
|
| 561 |
+
"learning_rate": 2.149103347373753e-06,
|
| 562 |
+
"loss": 0.8738,
|
| 563 |
+
"step": 385
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.638655462184874,
|
| 567 |
+
"grad_norm": 0.7382222604301619,
|
| 568 |
+
"learning_rate": 1.927122676180756e-06,
|
| 569 |
+
"loss": 0.8501,
|
| 570 |
+
"step": 390
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.6596638655462184,
|
| 574 |
+
"grad_norm": 0.7170057238227614,
|
| 575 |
+
"learning_rate": 1.7160145651786098e-06,
|
| 576 |
+
"loss": 0.8628,
|
| 577 |
+
"step": 395
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.680672268907563,
|
| 581 |
+
"grad_norm": 0.7735455354416457,
|
| 582 |
+
"learning_rate": 1.516063335006851e-06,
|
| 583 |
+
"loss": 0.8799,
|
| 584 |
+
"step": 400
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.7016806722689075,
|
| 588 |
+
"grad_norm": 0.7544008999017865,
|
| 589 |
+
"learning_rate": 1.3275382802054704e-06,
|
| 590 |
+
"loss": 0.8584,
|
| 591 |
+
"step": 405
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.7226890756302522,
|
| 595 |
+
"grad_norm": 0.7234290509257019,
|
| 596 |
+
"learning_rate": 1.1506933065287062e-06,
|
| 597 |
+
"loss": 0.8878,
|
| 598 |
+
"step": 410
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.7436974789915967,
|
| 602 |
+
"grad_norm": 0.7179219619602794,
|
| 603 |
+
"learning_rate": 9.85766588984508e-07,
|
| 604 |
+
"loss": 0.8701,
|
| 605 |
+
"step": 415
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.7647058823529411,
|
| 609 |
+
"grad_norm": 0.710095564820816,
|
| 610 |
+
"learning_rate": 8.329802510601559e-07,
|
| 611 |
+
"loss": 0.8732,
|
| 612 |
+
"step": 420
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.7857142857142856,
|
| 616 |
+
"grad_norm": 0.7465304276755028,
|
| 617 |
+
"learning_rate": 6.925400655661118e-07,
|
| 618 |
+
"loss": 0.897,
|
| 619 |
+
"step": 425
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.8067226890756303,
|
| 623 |
+
"grad_norm": 0.7039905037617595,
|
| 624 |
+
"learning_rate": 5.646351775009617e-07,
|
| 625 |
+
"loss": 0.8484,
|
| 626 |
+
"step": 430
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.8277310924369747,
|
| 630 |
+
"grad_norm": 0.7170317726401236,
|
| 631 |
+
"learning_rate": 4.494378493107232e-07,
|
| 632 |
+
"loss": 0.8518,
|
| 633 |
+
"step": 435
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.8487394957983194,
|
| 637 |
+
"grad_norm": 0.7399232574218224,
|
| 638 |
+
"learning_rate": 3.471032288855869e-07,
|
| 639 |
+
"loss": 0.8629,
|
| 640 |
+
"step": 440
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.8697478991596639,
|
| 644 |
+
"grad_norm": 0.7309518709786924,
|
| 645 |
+
"learning_rate": 2.577691406065708e-07,
|
| 646 |
+
"loss": 0.8723,
|
| 647 |
+
"step": 445
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.8907563025210083,
|
| 651 |
+
"grad_norm": 0.7524923842002688,
|
| 652 |
+
"learning_rate": 1.8155589972348453e-07,
|
| 653 |
+
"loss": 0.8655,
|
| 654 |
+
"step": 450
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.9117647058823528,
|
| 658 |
+
"grad_norm": 0.7470763175141618,
|
| 659 |
+
"learning_rate": 1.1856615031422214e-07,
|
| 660 |
+
"loss": 0.8656,
|
| 661 |
+
"step": 455
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.9327731092436975,
|
| 665 |
+
"grad_norm": 0.7413561376156461,
|
| 666 |
+
"learning_rate": 6.888472704359661e-08,
|
| 667 |
+
"loss": 0.8555,
|
| 668 |
+
"step": 460
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.9537815126050422,
|
| 672 |
+
"grad_norm": 0.725024144075873,
|
| 673 |
+
"learning_rate": 3.2578540907926e-08,
|
| 674 |
+
"loss": 0.8683,
|
| 675 |
+
"step": 465
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.9747899159663866,
|
| 679 |
+
"grad_norm": 0.7137402433851012,
|
| 680 |
+
"learning_rate": 9.696489119221942e-09,
|
| 681 |
+
"loss": 0.8736,
|
| 682 |
+
"step": 470
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.995798319327731,
|
| 686 |
+
"grad_norm": 0.7264722107888033,
|
| 687 |
+
"learning_rate": 2.6938925037689467e-10,
|
| 688 |
+
"loss": 0.8776,
|
| 689 |
+
"step": 475
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 2.0,
|
| 693 |
+
"eval_loss": 1.1365480422973633,
|
| 694 |
+
"eval_runtime": 5.9679,
|
| 695 |
+
"eval_samples_per_second": 24.799,
|
| 696 |
+
"eval_steps_per_second": 0.503,
|
| 697 |
+
"step": 476
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 2.0,
|
| 701 |
+
"step": 476,
|
| 702 |
+
"total_flos": 124772558045184.0,
|
| 703 |
+
"train_loss": 1.015126781929441,
|
| 704 |
+
"train_runtime": 3305.096,
|
| 705 |
+
"train_samples_per_second": 9.214,
|
| 706 |
+
"train_steps_per_second": 0.144
|
| 707 |
+
}
|
| 708 |
+
],
|
| 709 |
+
"logging_steps": 5,
|
| 710 |
+
"max_steps": 476,
|
| 711 |
+
"num_input_tokens_seen": 0,
|
| 712 |
+
"num_train_epochs": 2,
|
| 713 |
+
"save_steps": 500,
|
| 714 |
+
"stateful_callbacks": {
|
| 715 |
+
"TrainerControl": {
|
| 716 |
+
"args": {
|
| 717 |
+
"should_epoch_stop": false,
|
| 718 |
+
"should_evaluate": false,
|
| 719 |
+
"should_log": false,
|
| 720 |
+
"should_save": true,
|
| 721 |
+
"should_training_stop": true
|
| 722 |
+
},
|
| 723 |
+
"attributes": {}
|
| 724 |
+
}
|
| 725 |
+
},
|
| 726 |
+
"total_flos": 124772558045184.0,
|
| 727 |
+
"train_batch_size": 8,
|
| 728 |
+
"trial_name": null,
|
| 729 |
+
"trial_params": null
|
| 730 |
+
}
|