Initial commit
Browse files- trainer_state.json +708 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,708 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 46266,
|
| 3 |
+
"best_metric": 0.838254282517001,
|
| 4 |
+
"best_model_checkpoint": "dm_training/finetuned_checkpoints/SIM-RAG-full/checkpoint-46266",
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 46266,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.02161414429602732,
|
| 14 |
+
"grad_norm": 0.7032617926597595,
|
| 15 |
+
"learning_rate": 2.994e-05,
|
| 16 |
+
"loss": 0.3292,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.04322828859205464,
|
| 21 |
+
"grad_norm": 8.765708923339844,
|
| 22 |
+
"learning_rate": 2.9782725438685614e-05,
|
| 23 |
+
"loss": 0.3537,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.06484243288808196,
|
| 28 |
+
"grad_norm": 3.9439210891723633,
|
| 29 |
+
"learning_rate": 2.956501545740867e-05,
|
| 30 |
+
"loss": 0.3206,
|
| 31 |
+
"step": 1500
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.08645657718410928,
|
| 35 |
+
"grad_norm": 12.651052474975586,
|
| 36 |
+
"learning_rate": 2.934730547613173e-05,
|
| 37 |
+
"loss": 0.3505,
|
| 38 |
+
"step": 2000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.1080707214801366,
|
| 42 |
+
"grad_norm": 0.14961493015289307,
|
| 43 |
+
"learning_rate": 2.9129595494854787e-05,
|
| 44 |
+
"loss": 0.335,
|
| 45 |
+
"step": 2500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.12968486577616392,
|
| 49 |
+
"grad_norm": 2.6468780040740967,
|
| 50 |
+
"learning_rate": 2.8911885513577847e-05,
|
| 51 |
+
"loss": 0.3601,
|
| 52 |
+
"step": 3000
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.15129901007219124,
|
| 56 |
+
"grad_norm": 27.052431106567383,
|
| 57 |
+
"learning_rate": 2.8694175532300907e-05,
|
| 58 |
+
"loss": 0.3292,
|
| 59 |
+
"step": 3500
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.17291315436821855,
|
| 63 |
+
"grad_norm": 0.17826782166957855,
|
| 64 |
+
"learning_rate": 2.8476465551023964e-05,
|
| 65 |
+
"loss": 0.3218,
|
| 66 |
+
"step": 4000
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.1945272986642459,
|
| 70 |
+
"grad_norm": 0.4498291015625,
|
| 71 |
+
"learning_rate": 2.825875556974702e-05,
|
| 72 |
+
"loss": 0.2925,
|
| 73 |
+
"step": 4500
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.2161414429602732,
|
| 77 |
+
"grad_norm": 0.27510523796081543,
|
| 78 |
+
"learning_rate": 2.804104558847008e-05,
|
| 79 |
+
"loss": 0.3428,
|
| 80 |
+
"step": 5000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.23775558725630053,
|
| 84 |
+
"grad_norm": 38.518646240234375,
|
| 85 |
+
"learning_rate": 2.782333560719314e-05,
|
| 86 |
+
"loss": 0.2883,
|
| 87 |
+
"step": 5500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.25936973155232784,
|
| 91 |
+
"grad_norm": 0.18676885962486267,
|
| 92 |
+
"learning_rate": 2.7605625625916196e-05,
|
| 93 |
+
"loss": 0.3015,
|
| 94 |
+
"step": 6000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.28098387584835516,
|
| 98 |
+
"grad_norm": 0.5126625895500183,
|
| 99 |
+
"learning_rate": 2.7387915644639256e-05,
|
| 100 |
+
"loss": 0.2988,
|
| 101 |
+
"step": 6500
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.3025980201443825,
|
| 105 |
+
"grad_norm": 36.30085754394531,
|
| 106 |
+
"learning_rate": 2.7170205663362313e-05,
|
| 107 |
+
"loss": 0.2914,
|
| 108 |
+
"step": 7000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.3242121644404098,
|
| 112 |
+
"grad_norm": 5.256961345672607,
|
| 113 |
+
"learning_rate": 2.695249568208537e-05,
|
| 114 |
+
"loss": 0.3128,
|
| 115 |
+
"step": 7500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.3458263087364371,
|
| 119 |
+
"grad_norm": 0.18740588426589966,
|
| 120 |
+
"learning_rate": 2.6734785700808432e-05,
|
| 121 |
+
"loss": 0.2849,
|
| 122 |
+
"step": 8000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.3674404530324644,
|
| 126 |
+
"grad_norm": 33.268550872802734,
|
| 127 |
+
"learning_rate": 2.651707571953149e-05,
|
| 128 |
+
"loss": 0.3023,
|
| 129 |
+
"step": 8500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.3890545973284918,
|
| 133 |
+
"grad_norm": 0.0934978798031807,
|
| 134 |
+
"learning_rate": 2.629936573825455e-05,
|
| 135 |
+
"loss": 0.2882,
|
| 136 |
+
"step": 9000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.4106687416245191,
|
| 140 |
+
"grad_norm": 17.90414047241211,
|
| 141 |
+
"learning_rate": 2.6081655756977605e-05,
|
| 142 |
+
"loss": 0.2699,
|
| 143 |
+
"step": 9500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.4322828859205464,
|
| 147 |
+
"grad_norm": 0.9339249134063721,
|
| 148 |
+
"learning_rate": 2.5863945775700662e-05,
|
| 149 |
+
"loss": 0.3039,
|
| 150 |
+
"step": 10000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.45389703021657374,
|
| 154 |
+
"grad_norm": 0.04919858276844025,
|
| 155 |
+
"learning_rate": 2.5646235794423725e-05,
|
| 156 |
+
"loss": 0.2661,
|
| 157 |
+
"step": 10500
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.47551117451260105,
|
| 161 |
+
"grad_norm": 0.06951851397752762,
|
| 162 |
+
"learning_rate": 2.542852581314678e-05,
|
| 163 |
+
"loss": 0.2478,
|
| 164 |
+
"step": 11000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.49712531880862837,
|
| 168 |
+
"grad_norm": 0.058041177690029144,
|
| 169 |
+
"learning_rate": 2.5210815831869838e-05,
|
| 170 |
+
"loss": 0.2489,
|
| 171 |
+
"step": 11500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.5187394631046557,
|
| 175 |
+
"grad_norm": 22.77214813232422,
|
| 176 |
+
"learning_rate": 2.4993105850592898e-05,
|
| 177 |
+
"loss": 0.2744,
|
| 178 |
+
"step": 12000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.5403536074006831,
|
| 182 |
+
"grad_norm": 26.296802520751953,
|
| 183 |
+
"learning_rate": 2.4775395869315954e-05,
|
| 184 |
+
"loss": 0.2627,
|
| 185 |
+
"step": 12500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.5619677516967103,
|
| 189 |
+
"grad_norm": 18.95213508605957,
|
| 190 |
+
"learning_rate": 2.4557685888039014e-05,
|
| 191 |
+
"loss": 0.2405,
|
| 192 |
+
"step": 13000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.5835818959927377,
|
| 196 |
+
"grad_norm": 10.959939002990723,
|
| 197 |
+
"learning_rate": 2.4339975906762074e-05,
|
| 198 |
+
"loss": 0.2472,
|
| 199 |
+
"step": 13500
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.605196040288765,
|
| 203 |
+
"grad_norm": 0.1388530284166336,
|
| 204 |
+
"learning_rate": 2.412226592548513e-05,
|
| 205 |
+
"loss": 0.2532,
|
| 206 |
+
"step": 14000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.6268101845847923,
|
| 210 |
+
"grad_norm": 0.28232917189598083,
|
| 211 |
+
"learning_rate": 2.3904555944208187e-05,
|
| 212 |
+
"loss": 0.2727,
|
| 213 |
+
"step": 14500
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.6484243288808196,
|
| 217 |
+
"grad_norm": 0.17236585915088654,
|
| 218 |
+
"learning_rate": 2.3686845962931247e-05,
|
| 219 |
+
"loss": 0.2593,
|
| 220 |
+
"step": 15000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.670038473176847,
|
| 224 |
+
"grad_norm": 0.013697458431124687,
|
| 225 |
+
"learning_rate": 2.3469135981654307e-05,
|
| 226 |
+
"loss": 0.2163,
|
| 227 |
+
"step": 15500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.6916526174728742,
|
| 231 |
+
"grad_norm": 0.04355601221323013,
|
| 232 |
+
"learning_rate": 2.3251426000377363e-05,
|
| 233 |
+
"loss": 0.2418,
|
| 234 |
+
"step": 16000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.7132667617689016,
|
| 238 |
+
"grad_norm": 0.15288621187210083,
|
| 239 |
+
"learning_rate": 2.3033716019100423e-05,
|
| 240 |
+
"loss": 0.2458,
|
| 241 |
+
"step": 16500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.7348809060649288,
|
| 245 |
+
"grad_norm": 0.11191035062074661,
|
| 246 |
+
"learning_rate": 2.281600603782348e-05,
|
| 247 |
+
"loss": 0.2299,
|
| 248 |
+
"step": 17000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.7564950503609562,
|
| 252 |
+
"grad_norm": 0.06535373628139496,
|
| 253 |
+
"learning_rate": 2.259829605654654e-05,
|
| 254 |
+
"loss": 0.2299,
|
| 255 |
+
"step": 17500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.7781091946569836,
|
| 259 |
+
"grad_norm": 0.10419075191020966,
|
| 260 |
+
"learning_rate": 2.23805860752696e-05,
|
| 261 |
+
"loss": 0.2606,
|
| 262 |
+
"step": 18000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.7997233389530108,
|
| 266 |
+
"grad_norm": 25.499109268188477,
|
| 267 |
+
"learning_rate": 2.2162876093992656e-05,
|
| 268 |
+
"loss": 0.2452,
|
| 269 |
+
"step": 18500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.8213374832490382,
|
| 273 |
+
"grad_norm": 7.74832010269165,
|
| 274 |
+
"learning_rate": 2.1945166112715716e-05,
|
| 275 |
+
"loss": 0.2125,
|
| 276 |
+
"step": 19000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.8429516275450655,
|
| 280 |
+
"grad_norm": 0.14268887042999268,
|
| 281 |
+
"learning_rate": 2.1727456131438773e-05,
|
| 282 |
+
"loss": 0.2155,
|
| 283 |
+
"step": 19500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.8645657718410928,
|
| 287 |
+
"grad_norm": 0.7247521877288818,
|
| 288 |
+
"learning_rate": 2.1509746150161832e-05,
|
| 289 |
+
"loss": 0.2234,
|
| 290 |
+
"step": 20000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.8861799161371201,
|
| 294 |
+
"grad_norm": 14.978250503540039,
|
| 295 |
+
"learning_rate": 2.1292036168884892e-05,
|
| 296 |
+
"loss": 0.2129,
|
| 297 |
+
"step": 20500
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.9077940604331475,
|
| 301 |
+
"grad_norm": 0.235914945602417,
|
| 302 |
+
"learning_rate": 2.107432618760795e-05,
|
| 303 |
+
"loss": 0.2189,
|
| 304 |
+
"step": 21000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.9294082047291747,
|
| 308 |
+
"grad_norm": 20.6143798828125,
|
| 309 |
+
"learning_rate": 2.0856616206331005e-05,
|
| 310 |
+
"loss": 0.2051,
|
| 311 |
+
"step": 21500
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.9510223490252021,
|
| 315 |
+
"grad_norm": 67.3742904663086,
|
| 316 |
+
"learning_rate": 2.0638906225054065e-05,
|
| 317 |
+
"loss": 0.2045,
|
| 318 |
+
"step": 22000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.9726364933212294,
|
| 322 |
+
"grad_norm": 0.51312255859375,
|
| 323 |
+
"learning_rate": 2.0421196243777125e-05,
|
| 324 |
+
"loss": 0.2013,
|
| 325 |
+
"step": 22500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.9942506376172567,
|
| 329 |
+
"grad_norm": 0.40760791301727295,
|
| 330 |
+
"learning_rate": 2.020348626250018e-05,
|
| 331 |
+
"loss": 0.2329,
|
| 332 |
+
"step": 23000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.0,
|
| 336 |
+
"eval_accuracy": 0.8315399845054662,
|
| 337 |
+
"eval_f1_0": 0.8379296066252587,
|
| 338 |
+
"eval_f1_1": 0.8246258625324849,
|
| 339 |
+
"eval_loss": 0.27029550075531006,
|
| 340 |
+
"eval_precision_0": 0.8346807457515262,
|
| 341 |
+
"eval_precision_1": 0.828113750899928,
|
| 342 |
+
"eval_recall_0": 0.8412038576654473,
|
| 343 |
+
"eval_recall_1": 0.8211672318400857,
|
| 344 |
+
"eval_runtime": 6851.1436,
|
| 345 |
+
"eval_samples_per_second": 1.696,
|
| 346 |
+
"eval_steps_per_second": 0.848,
|
| 347 |
+
"step": 23133
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.015864781913284,
|
| 351 |
+
"grad_norm": 0.1842740774154663,
|
| 352 |
+
"learning_rate": 1.998577628122324e-05,
|
| 353 |
+
"loss": 0.1819,
|
| 354 |
+
"step": 23500
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.0374789262093114,
|
| 358 |
+
"grad_norm": 18.2067928314209,
|
| 359 |
+
"learning_rate": 1.9768066299946298e-05,
|
| 360 |
+
"loss": 0.1862,
|
| 361 |
+
"step": 24000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.0590930705053387,
|
| 365 |
+
"grad_norm": 56.502193450927734,
|
| 366 |
+
"learning_rate": 1.9550356318669354e-05,
|
| 367 |
+
"loss": 0.1461,
|
| 368 |
+
"step": 24500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.0807072148013661,
|
| 372 |
+
"grad_norm": 0.013557116501033306,
|
| 373 |
+
"learning_rate": 1.9332646337392418e-05,
|
| 374 |
+
"loss": 0.1521,
|
| 375 |
+
"step": 25000
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.1023213590973933,
|
| 379 |
+
"grad_norm": 0.05783897638320923,
|
| 380 |
+
"learning_rate": 1.9114936356115474e-05,
|
| 381 |
+
"loss": 0.1428,
|
| 382 |
+
"step": 25500
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.1239355033934206,
|
| 386 |
+
"grad_norm": 0.026215313002467155,
|
| 387 |
+
"learning_rate": 1.889722637483853e-05,
|
| 388 |
+
"loss": 0.1491,
|
| 389 |
+
"step": 26000
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.145549647689448,
|
| 393 |
+
"grad_norm": 0.0602092407643795,
|
| 394 |
+
"learning_rate": 1.867951639356159e-05,
|
| 395 |
+
"loss": 0.1442,
|
| 396 |
+
"step": 26500
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.1671637919854754,
|
| 400 |
+
"grad_norm": 0.03851708024740219,
|
| 401 |
+
"learning_rate": 1.8461806412284647e-05,
|
| 402 |
+
"loss": 0.1511,
|
| 403 |
+
"step": 27000
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.1887779362815025,
|
| 407 |
+
"grad_norm": 0.029633022844791412,
|
| 408 |
+
"learning_rate": 1.8244096431007707e-05,
|
| 409 |
+
"loss": 0.1438,
|
| 410 |
+
"step": 27500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.21039208057753,
|
| 414 |
+
"grad_norm": 4.901973247528076,
|
| 415 |
+
"learning_rate": 1.8026386449730767e-05,
|
| 416 |
+
"loss": 0.1541,
|
| 417 |
+
"step": 28000
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.2320062248735573,
|
| 421 |
+
"grad_norm": 0.05654510483145714,
|
| 422 |
+
"learning_rate": 1.7808676468453823e-05,
|
| 423 |
+
"loss": 0.1537,
|
| 424 |
+
"step": 28500
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.2536203691695846,
|
| 428 |
+
"grad_norm": 0.0678759291768074,
|
| 429 |
+
"learning_rate": 1.7590966487176883e-05,
|
| 430 |
+
"loss": 0.1362,
|
| 431 |
+
"step": 29000
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.275234513465612,
|
| 435 |
+
"grad_norm": 14.40847110748291,
|
| 436 |
+
"learning_rate": 1.737325650589994e-05,
|
| 437 |
+
"loss": 0.1611,
|
| 438 |
+
"step": 29500
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.2968486577616392,
|
| 442 |
+
"grad_norm": 0.13242945075035095,
|
| 443 |
+
"learning_rate": 1.7155546524623e-05,
|
| 444 |
+
"loss": 0.1743,
|
| 445 |
+
"step": 30000
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.3184628020576665,
|
| 449 |
+
"grad_norm": 0.05444726720452309,
|
| 450 |
+
"learning_rate": 1.693783654334606e-05,
|
| 451 |
+
"loss": 0.1427,
|
| 452 |
+
"step": 30500
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.340076946353694,
|
| 456 |
+
"grad_norm": 28.47621726989746,
|
| 457 |
+
"learning_rate": 1.6720126562069116e-05,
|
| 458 |
+
"loss": 0.1485,
|
| 459 |
+
"step": 31000
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.361691090649721,
|
| 463 |
+
"grad_norm": 0.04358465224504471,
|
| 464 |
+
"learning_rate": 1.6502416580792172e-05,
|
| 465 |
+
"loss": 0.1422,
|
| 466 |
+
"step": 31500
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.3833052349457484,
|
| 470 |
+
"grad_norm": 0.024943144991993904,
|
| 471 |
+
"learning_rate": 1.6284706599515236e-05,
|
| 472 |
+
"loss": 0.1386,
|
| 473 |
+
"step": 32000
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.4049193792417758,
|
| 477 |
+
"grad_norm": 0.1587284654378891,
|
| 478 |
+
"learning_rate": 1.6066996618238292e-05,
|
| 479 |
+
"loss": 0.1252,
|
| 480 |
+
"step": 32500
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.4265335235378032,
|
| 484 |
+
"grad_norm": 0.054883528500795364,
|
| 485 |
+
"learning_rate": 1.584928663696135e-05,
|
| 486 |
+
"loss": 0.1429,
|
| 487 |
+
"step": 33000
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.4481476678338305,
|
| 491 |
+
"grad_norm": 0.21283945441246033,
|
| 492 |
+
"learning_rate": 1.563157665568441e-05,
|
| 493 |
+
"loss": 0.1465,
|
| 494 |
+
"step": 33500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.4697618121298577,
|
| 498 |
+
"grad_norm": 63.26069259643555,
|
| 499 |
+
"learning_rate": 1.5413866674407465e-05,
|
| 500 |
+
"loss": 0.1157,
|
| 501 |
+
"step": 34000
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.491375956425885,
|
| 505 |
+
"grad_norm": 20.822044372558594,
|
| 506 |
+
"learning_rate": 1.5196156693130527e-05,
|
| 507 |
+
"loss": 0.1505,
|
| 508 |
+
"step": 34500
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.5129901007219124,
|
| 512 |
+
"grad_norm": 0.2792131006717682,
|
| 513 |
+
"learning_rate": 1.4978446711853583e-05,
|
| 514 |
+
"loss": 0.1205,
|
| 515 |
+
"step": 35000
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.5346042450179398,
|
| 519 |
+
"grad_norm": 0.04913631081581116,
|
| 520 |
+
"learning_rate": 1.4760736730576641e-05,
|
| 521 |
+
"loss": 0.124,
|
| 522 |
+
"step": 35500
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.5562183893139672,
|
| 526 |
+
"grad_norm": 0.007014845497906208,
|
| 527 |
+
"learning_rate": 1.45430267492997e-05,
|
| 528 |
+
"loss": 0.1404,
|
| 529 |
+
"step": 36000
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.5778325336099943,
|
| 533 |
+
"grad_norm": 0.03032066859304905,
|
| 534 |
+
"learning_rate": 1.432531676802276e-05,
|
| 535 |
+
"loss": 0.1345,
|
| 536 |
+
"step": 36500
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.5994466779060217,
|
| 540 |
+
"grad_norm": 0.02347446419298649,
|
| 541 |
+
"learning_rate": 1.4107606786745816e-05,
|
| 542 |
+
"loss": 0.141,
|
| 543 |
+
"step": 37000
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.621060822202049,
|
| 547 |
+
"grad_norm": 0.03960123285651207,
|
| 548 |
+
"learning_rate": 1.3889896805468876e-05,
|
| 549 |
+
"loss": 0.1399,
|
| 550 |
+
"step": 37500
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.6426749664980762,
|
| 554 |
+
"grad_norm": 0.0023522686678916216,
|
| 555 |
+
"learning_rate": 1.3672186824191934e-05,
|
| 556 |
+
"loss": 0.1301,
|
| 557 |
+
"step": 38000
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.6642891107941038,
|
| 561 |
+
"grad_norm": 0.10293618589639664,
|
| 562 |
+
"learning_rate": 1.345447684291499e-05,
|
| 563 |
+
"loss": 0.1504,
|
| 564 |
+
"step": 38500
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.685903255090131,
|
| 568 |
+
"grad_norm": 0.6177674531936646,
|
| 569 |
+
"learning_rate": 1.323676686163805e-05,
|
| 570 |
+
"loss": 0.1244,
|
| 571 |
+
"step": 39000
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.7075173993861583,
|
| 575 |
+
"grad_norm": 0.2255789041519165,
|
| 576 |
+
"learning_rate": 1.3019056880361109e-05,
|
| 577 |
+
"loss": 0.1229,
|
| 578 |
+
"step": 39500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.7291315436821857,
|
| 582 |
+
"grad_norm": 0.01264307089149952,
|
| 583 |
+
"learning_rate": 1.2801346899084167e-05,
|
| 584 |
+
"loss": 0.1199,
|
| 585 |
+
"step": 40000
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.7507456879782128,
|
| 589 |
+
"grad_norm": 21.45499610900879,
|
| 590 |
+
"learning_rate": 1.2583636917807225e-05,
|
| 591 |
+
"loss": 0.1234,
|
| 592 |
+
"step": 40500
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.7723598322742402,
|
| 596 |
+
"grad_norm": 0.02090781182050705,
|
| 597 |
+
"learning_rate": 1.2365926936530283e-05,
|
| 598 |
+
"loss": 0.121,
|
| 599 |
+
"step": 41000
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.7939739765702676,
|
| 603 |
+
"grad_norm": 0.054082971066236496,
|
| 604 |
+
"learning_rate": 1.2148216955253343e-05,
|
| 605 |
+
"loss": 0.1158,
|
| 606 |
+
"step": 41500
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.815588120866295,
|
| 610 |
+
"grad_norm": 0.04447195306420326,
|
| 611 |
+
"learning_rate": 1.19305069739764e-05,
|
| 612 |
+
"loss": 0.1183,
|
| 613 |
+
"step": 42000
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.8372022651623223,
|
| 617 |
+
"grad_norm": 38.29100799560547,
|
| 618 |
+
"learning_rate": 1.171279699269946e-05,
|
| 619 |
+
"loss": 0.1302,
|
| 620 |
+
"step": 42500
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.8588164094583495,
|
| 624 |
+
"grad_norm": 0.014447253197431564,
|
| 625 |
+
"learning_rate": 1.1495087011422518e-05,
|
| 626 |
+
"loss": 0.1011,
|
| 627 |
+
"step": 43000
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.8804305537543768,
|
| 631 |
+
"grad_norm": 0.0001848287502070889,
|
| 632 |
+
"learning_rate": 1.1277377030145576e-05,
|
| 633 |
+
"loss": 0.1161,
|
| 634 |
+
"step": 43500
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.9020446980504042,
|
| 638 |
+
"grad_norm": 0.003150364151224494,
|
| 639 |
+
"learning_rate": 1.1059667048868634e-05,
|
| 640 |
+
"loss": 0.1332,
|
| 641 |
+
"step": 44000
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.9236588423464314,
|
| 645 |
+
"grad_norm": 0.2078738808631897,
|
| 646 |
+
"learning_rate": 1.0841957067591692e-05,
|
| 647 |
+
"loss": 0.1409,
|
| 648 |
+
"step": 44500
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.945272986642459,
|
| 652 |
+
"grad_norm": 0.059768468141555786,
|
| 653 |
+
"learning_rate": 1.062424708631475e-05,
|
| 654 |
+
"loss": 0.1326,
|
| 655 |
+
"step": 45000
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.966887130938486,
|
| 659 |
+
"grad_norm": 0.038765549659729004,
|
| 660 |
+
"learning_rate": 1.0406537105037809e-05,
|
| 661 |
+
"loss": 0.0918,
|
| 662 |
+
"step": 45500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.9885012752345135,
|
| 666 |
+
"grad_norm": 18.927526473999023,
|
| 667 |
+
"learning_rate": 1.0188827123760869e-05,
|
| 668 |
+
"loss": 0.1091,
|
| 669 |
+
"step": 46000
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 2.0,
|
| 673 |
+
"eval_accuracy": 0.838254282517001,
|
| 674 |
+
"eval_f1_0": 0.8413944458512703,
|
| 675 |
+
"eval_f1_1": 0.8349872661807324,
|
| 676 |
+
"eval_loss": 0.3542537987232208,
|
| 677 |
+
"eval_precision_0": 0.8544488256471798,
|
| 678 |
+
"eval_precision_1": 0.8219225449515906,
|
| 679 |
+
"eval_recall_0": 0.828732956434985,
|
| 680 |
+
"eval_recall_1": 0.8484740317686953,
|
| 681 |
+
"eval_runtime": 6835.7043,
|
| 682 |
+
"eval_samples_per_second": 1.699,
|
| 683 |
+
"eval_steps_per_second": 0.85,
|
| 684 |
+
"step": 46266
|
| 685 |
+
}
|
| 686 |
+
],
|
| 687 |
+
"logging_steps": 500,
|
| 688 |
+
"max_steps": 69399,
|
| 689 |
+
"num_input_tokens_seen": 0,
|
| 690 |
+
"num_train_epochs": 3,
|
| 691 |
+
"save_steps": 500,
|
| 692 |
+
"stateful_callbacks": {
|
| 693 |
+
"TrainerControl": {
|
| 694 |
+
"args": {
|
| 695 |
+
"should_epoch_stop": false,
|
| 696 |
+
"should_evaluate": false,
|
| 697 |
+
"should_log": false,
|
| 698 |
+
"should_save": true,
|
| 699 |
+
"should_training_stop": false
|
| 700 |
+
},
|
| 701 |
+
"attributes": {}
|
| 702 |
+
}
|
| 703 |
+
},
|
| 704 |
+
"total_flos": 1.5826915766462054e+18,
|
| 705 |
+
"train_batch_size": 2,
|
| 706 |
+
"trial_name": null,
|
| 707 |
+
"trial_params": null
|
| 708 |
+
}
|