Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +518 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,518 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 680,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.14705882352941177,
|
| 13 |
+
"grad_norm": 5.515628814697266,
|
| 14 |
+
"learning_rate": 2.9411764705882354e-05,
|
| 15 |
+
"loss": 1.1436,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.29411764705882354,
|
| 20 |
+
"grad_norm": 2.5779266357421875,
|
| 21 |
+
"learning_rate": 5.882352941176471e-05,
|
| 22 |
+
"loss": 0.4215,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.4411764705882353,
|
| 27 |
+
"grad_norm": 1.5243560075759888,
|
| 28 |
+
"learning_rate": 8.823529411764706e-05,
|
| 29 |
+
"loss": 0.2531,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.5882352941176471,
|
| 34 |
+
"grad_norm": 1.0052002668380737,
|
| 35 |
+
"learning_rate": 9.997871633546257e-05,
|
| 36 |
+
"loss": 0.1957,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.7352941176470589,
|
| 41 |
+
"grad_norm": 1.7932289838790894,
|
| 42 |
+
"learning_rate": 9.98487151097676e-05,
|
| 43 |
+
"loss": 0.1667,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.8823529411764706,
|
| 48 |
+
"grad_norm": 0.6952548623085022,
|
| 49 |
+
"learning_rate": 9.960084393841355e-05,
|
| 50 |
+
"loss": 0.1425,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 1.0294117647058822,
|
| 55 |
+
"grad_norm": 0.944196343421936,
|
| 56 |
+
"learning_rate": 9.923568892600578e-05,
|
| 57 |
+
"loss": 0.1222,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 1.1764705882352942,
|
| 62 |
+
"grad_norm": 0.6027107238769531,
|
| 63 |
+
"learning_rate": 9.875411350104744e-05,
|
| 64 |
+
"loss": 0.1081,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 1.3235294117647058,
|
| 69 |
+
"grad_norm": 0.7133479714393616,
|
| 70 |
+
"learning_rate": 9.815725637431662e-05,
|
| 71 |
+
"loss": 0.1004,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.4705882352941178,
|
| 76 |
+
"grad_norm": 0.6044264435768127,
|
| 77 |
+
"learning_rate": 9.744652884632406e-05,
|
| 78 |
+
"loss": 0.0935,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.6176470588235294,
|
| 83 |
+
"grad_norm": 0.5952054858207703,
|
| 84 |
+
"learning_rate": 9.662361147021779e-05,
|
| 85 |
+
"loss": 0.0938,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.7647058823529411,
|
| 90 |
+
"grad_norm": 0.7661593556404114,
|
| 91 |
+
"learning_rate": 9.569045007802559e-05,
|
| 92 |
+
"loss": 0.0876,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.9117647058823528,
|
| 97 |
+
"grad_norm": 0.5203129649162292,
|
| 98 |
+
"learning_rate": 9.464925117963133e-05,
|
| 99 |
+
"loss": 0.0799,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 2.0588235294117645,
|
| 104 |
+
"grad_norm": 1.392182469367981,
|
| 105 |
+
"learning_rate": 9.35024767453647e-05,
|
| 106 |
+
"loss": 0.0848,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 2.2058823529411766,
|
| 111 |
+
"grad_norm": 0.44986793398857117,
|
| 112 |
+
"learning_rate": 9.225283838454111e-05,
|
| 113 |
+
"loss": 0.0789,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 2.3529411764705883,
|
| 118 |
+
"grad_norm": 0.4528612196445465,
|
| 119 |
+
"learning_rate": 9.090329093371666e-05,
|
| 120 |
+
"loss": 0.0776,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.5,
|
| 125 |
+
"grad_norm": 0.5809573531150818,
|
| 126 |
+
"learning_rate": 8.945702546981969e-05,
|
| 127 |
+
"loss": 0.0715,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.6470588235294117,
|
| 132 |
+
"grad_norm": 0.5827745795249939,
|
| 133 |
+
"learning_rate": 8.791746176467907e-05,
|
| 134 |
+
"loss": 0.0716,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.7941176470588234,
|
| 139 |
+
"grad_norm": 0.5849335789680481,
|
| 140 |
+
"learning_rate": 8.628824019879137e-05,
|
| 141 |
+
"loss": 0.0653,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.9411764705882355,
|
| 146 |
+
"grad_norm": 0.661180317401886,
|
| 147 |
+
"learning_rate": 8.457321315344694e-05,
|
| 148 |
+
"loss": 0.0669,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 3.088235294117647,
|
| 153 |
+
"grad_norm": 0.5251627564430237,
|
| 154 |
+
"learning_rate": 8.277643590156894e-05,
|
| 155 |
+
"loss": 0.069,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 3.235294117647059,
|
| 160 |
+
"grad_norm": 0.471332311630249,
|
| 161 |
+
"learning_rate": 8.090215701880419e-05,
|
| 162 |
+
"loss": 0.0609,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 3.3823529411764706,
|
| 167 |
+
"grad_norm": 0.42382729053497314,
|
| 168 |
+
"learning_rate": 7.89548083375394e-05,
|
| 169 |
+
"loss": 0.0622,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.5294117647058822,
|
| 174 |
+
"grad_norm": 0.899319589138031,
|
| 175 |
+
"learning_rate": 7.693899446759727e-05,
|
| 176 |
+
"loss": 0.0648,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 3.6764705882352944,
|
| 181 |
+
"grad_norm": 0.6775935292243958,
|
| 182 |
+
"learning_rate": 7.485948190839077e-05,
|
| 183 |
+
"loss": 0.0598,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 3.8235294117647056,
|
| 188 |
+
"grad_norm": 0.43540897965431213,
|
| 189 |
+
"learning_rate": 7.272118777828108e-05,
|
| 190 |
+
"loss": 0.0573,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 3.9705882352941178,
|
| 195 |
+
"grad_norm": 0.27909591794013977,
|
| 196 |
+
"learning_rate": 7.052916818778918e-05,
|
| 197 |
+
"loss": 0.0492,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 4.117647058823529,
|
| 202 |
+
"grad_norm": 0.42636606097221375,
|
| 203 |
+
"learning_rate": 6.828860628415253e-05,
|
| 204 |
+
"loss": 0.0557,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 4.264705882352941,
|
| 209 |
+
"grad_norm": 0.4702949821949005,
|
| 210 |
+
"learning_rate": 6.60047999954972e-05,
|
| 211 |
+
"loss": 0.0536,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 4.411764705882353,
|
| 216 |
+
"grad_norm": 0.5331495404243469,
|
| 217 |
+
"learning_rate": 6.368314950360415e-05,
|
| 218 |
+
"loss": 0.0525,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 4.5588235294117645,
|
| 223 |
+
"grad_norm": 0.301176518201828,
|
| 224 |
+
"learning_rate": 6.132914447489137e-05,
|
| 225 |
+
"loss": 0.0568,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 4.705882352941177,
|
| 230 |
+
"grad_norm": 0.2303120642900467,
|
| 231 |
+
"learning_rate": 5.8948351079804875e-05,
|
| 232 |
+
"loss": 0.0432,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 4.852941176470588,
|
| 237 |
+
"grad_norm": 0.37262749671936035,
|
| 238 |
+
"learning_rate": 5.654639883131178e-05,
|
| 239 |
+
"loss": 0.0491,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 5.0,
|
| 244 |
+
"grad_norm": 0.6805188059806824,
|
| 245 |
+
"learning_rate": 5.4128967273616625e-05,
|
| 246 |
+
"loss": 0.0513,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 5.147058823529412,
|
| 251 |
+
"grad_norm": 0.4015423655509949,
|
| 252 |
+
"learning_rate": 5.170177255257618e-05,
|
| 253 |
+
"loss": 0.0456,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 5.294117647058823,
|
| 258 |
+
"grad_norm": 0.3756394386291504,
|
| 259 |
+
"learning_rate": 4.9270553899567686e-05,
|
| 260 |
+
"loss": 0.0535,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 5.4411764705882355,
|
| 265 |
+
"grad_norm": 0.3560592532157898,
|
| 266 |
+
"learning_rate": 4.6841060060770154e-05,
|
| 267 |
+
"loss": 0.0463,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 5.588235294117647,
|
| 272 |
+
"grad_norm": 0.4422471523284912,
|
| 273 |
+
"learning_rate": 4.441903570394739e-05,
|
| 274 |
+
"loss": 0.0417,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 5.735294117647059,
|
| 279 |
+
"grad_norm": 0.577297568321228,
|
| 280 |
+
"learning_rate": 4.201020783487464e-05,
|
| 281 |
+
"loss": 0.0402,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 5.882352941176471,
|
| 286 |
+
"grad_norm": 0.7647914290428162,
|
| 287 |
+
"learning_rate": 3.962027225552807e-05,
|
| 288 |
+
"loss": 0.0402,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 6.029411764705882,
|
| 293 |
+
"grad_norm": 0.8858449459075928,
|
| 294 |
+
"learning_rate": 3.7254880096057073e-05,
|
| 295 |
+
"loss": 0.0479,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 6.176470588235294,
|
| 300 |
+
"grad_norm": 0.49515625834465027,
|
| 301 |
+
"learning_rate": 3.491962445238569e-05,
|
| 302 |
+
"loss": 0.0434,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 6.323529411764706,
|
| 307 |
+
"grad_norm": 0.548555850982666,
|
| 308 |
+
"learning_rate": 3.262002716103897e-05,
|
| 309 |
+
"loss": 0.0442,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 6.470588235294118,
|
| 314 |
+
"grad_norm": 0.35759156942367554,
|
| 315 |
+
"learning_rate": 3.0361525742465973e-05,
|
| 316 |
+
"loss": 0.0411,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 6.617647058823529,
|
| 321 |
+
"grad_norm": 0.31173354387283325,
|
| 322 |
+
"learning_rate": 2.8149460543732664e-05,
|
| 323 |
+
"loss": 0.0376,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 6.764705882352941,
|
| 328 |
+
"grad_norm": 0.24947527050971985,
|
| 329 |
+
"learning_rate": 2.598906211098643e-05,
|
| 330 |
+
"loss": 0.0391,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 6.911764705882353,
|
| 335 |
+
"grad_norm": 0.20982353389263153,
|
| 336 |
+
"learning_rate": 2.388543882155067e-05,
|
| 337 |
+
"loss": 0.0365,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 7.0588235294117645,
|
| 342 |
+
"grad_norm": 0.3836628496646881,
|
| 343 |
+
"learning_rate": 2.184356480489432e-05,
|
| 344 |
+
"loss": 0.0365,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 7.205882352941177,
|
| 349 |
+
"grad_norm": 0.23856157064437866,
|
| 350 |
+
"learning_rate": 1.9868268181037185e-05,
|
| 351 |
+
"loss": 0.0333,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 7.352941176470588,
|
| 356 |
+
"grad_norm": 0.6093345880508423,
|
| 357 |
+
"learning_rate": 1.796421964420285e-05,
|
| 358 |
+
"loss": 0.0389,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 7.5,
|
| 363 |
+
"grad_norm": 0.2536391019821167,
|
| 364 |
+
"learning_rate": 1.6135921418712956e-05,
|
| 365 |
+
"loss": 0.0355,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 7.647058823529412,
|
| 370 |
+
"grad_norm": 0.22027313709259033,
|
| 371 |
+
"learning_rate": 1.4387696613237612e-05,
|
| 372 |
+
"loss": 0.0331,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 7.794117647058823,
|
| 377 |
+
"grad_norm": 0.367398738861084,
|
| 378 |
+
"learning_rate": 1.2723678998574512e-05,
|
| 379 |
+
"loss": 0.0352,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 7.9411764705882355,
|
| 384 |
+
"grad_norm": 0.24603775143623352,
|
| 385 |
+
"learning_rate": 1.114780323312724e-05,
|
| 386 |
+
"loss": 0.0399,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 8.088235294117647,
|
| 391 |
+
"grad_norm": 0.22743625938892365,
|
| 392 |
+
"learning_rate": 9.663795559195733e-06,
|
| 393 |
+
"loss": 0.0315,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 8.235294117647058,
|
| 398 |
+
"grad_norm": 0.3211243152618408,
|
| 399 |
+
"learning_rate": 8.275164992077556e-06,
|
| 400 |
+
"loss": 0.033,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 8.382352941176471,
|
| 405 |
+
"grad_norm": 0.3177715241909027,
|
| 406 |
+
"learning_rate": 6.985195022814067e-06,
|
| 407 |
+
"loss": 0.0366,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 8.529411764705882,
|
| 412 |
+
"grad_norm": 0.4995149075984955,
|
| 413 |
+
"learning_rate": 5.796935854200763e-06,
|
| 414 |
+
"loss": 0.0353,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 8.676470588235293,
|
| 419 |
+
"grad_norm": 0.26444506645202637,
|
| 420 |
+
"learning_rate": 4.713197188420026e-06,
|
| 421 |
+
"loss": 0.0311,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 8.823529411764707,
|
| 426 |
+
"grad_norm": 0.2560294568538666,
|
| 427 |
+
"learning_rate": 3.7365415833504725e-06,
|
| 428 |
+
"loss": 0.0355,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 8.970588235294118,
|
| 433 |
+
"grad_norm": 0.20665033161640167,
|
| 434 |
+
"learning_rate": 2.869278393262226e-06,
|
| 435 |
+
"loss": 0.0367,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 9.117647058823529,
|
| 440 |
+
"grad_norm": 0.4208621382713318,
|
| 441 |
+
"learning_rate": 2.113458308225458e-06,
|
| 442 |
+
"loss": 0.0346,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 9.264705882352942,
|
| 447 |
+
"grad_norm": 0.3151559829711914,
|
| 448 |
+
"learning_rate": 1.4708685051444515e-06,
|
| 449 |
+
"loss": 0.0332,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 9.411764705882353,
|
| 454 |
+
"grad_norm": 0.5470899343490601,
|
| 455 |
+
"learning_rate": 9.430284218824026e-07,
|
| 456 |
+
"loss": 0.0319,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 9.558823529411764,
|
| 461 |
+
"grad_norm": 1.0164936780929565,
|
| 462 |
+
"learning_rate": 5.311861644696048e-07,
|
| 463 |
+
"loss": 0.0378,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 9.705882352941176,
|
| 468 |
+
"grad_norm": 0.18899625539779663,
|
| 469 |
+
"learning_rate": 2.363155558901542e-07,
|
| 470 |
+
"loss": 0.0349,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 9.852941176470589,
|
| 475 |
+
"grad_norm": 0.18862251937389374,
|
| 476 |
+
"learning_rate": 5.911383342556143e-08,
|
| 477 |
+
"loss": 0.036,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 10.0,
|
| 482 |
+
"grad_norm": 0.724671483039856,
|
| 483 |
+
"learning_rate": 0.0,
|
| 484 |
+
"loss": 0.0299,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 10.0,
|
| 489 |
+
"step": 680,
|
| 490 |
+
"total_flos": 9.581997557091456e+16,
|
| 491 |
+
"train_loss": 0.08232775286716573,
|
| 492 |
+
"train_runtime": 870.4339,
|
| 493 |
+
"train_samples_per_second": 49.378,
|
| 494 |
+
"train_steps_per_second": 0.781
|
| 495 |
+
}
|
| 496 |
+
],
|
| 497 |
+
"logging_steps": 10,
|
| 498 |
+
"max_steps": 680,
|
| 499 |
+
"num_input_tokens_seen": 0,
|
| 500 |
+
"num_train_epochs": 10,
|
| 501 |
+
"save_steps": 10000,
|
| 502 |
+
"stateful_callbacks": {
|
| 503 |
+
"TrainerControl": {
|
| 504 |
+
"args": {
|
| 505 |
+
"should_epoch_stop": false,
|
| 506 |
+
"should_evaluate": false,
|
| 507 |
+
"should_log": false,
|
| 508 |
+
"should_save": true,
|
| 509 |
+
"should_training_stop": true
|
| 510 |
+
},
|
| 511 |
+
"attributes": {}
|
| 512 |
+
}
|
| 513 |
+
},
|
| 514 |
+
"total_flos": 9.581997557091456e+16,
|
| 515 |
+
"train_batch_size": 64,
|
| 516 |
+
"trial_name": null,
|
| 517 |
+
"trial_params": null
|
| 518 |
+
}
|