Upload folder using huggingface_hub
Browse files
checkpoints/checkpoint-408/adapter.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aaf5abcb61d1492002aac91a16ea3bdecba80ccfa546de2e84418b8fa0168721
|
| 3 |
+
size 6439640
|
checkpoints/checkpoint-408/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8daf7c41f9dfd692fa1bc1d86914b7c497e2c39d1891b6c0e352de5ab5ff173
|
| 3 |
+
size 3304962
|
checkpoints/checkpoint-408/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5ed0adb262ae5a32b0426ae5f447cd38d4498b7741803b88609e5036322fb8d3
|
| 3 |
+
size 14244
|
checkpoints/checkpoint-408/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e8c0d01d51244373dd2dd86b408e09445988f26a5dd6fa08072fd5289a38c0ab
|
| 3 |
+
size 1064
|
checkpoints/checkpoint-408/trainer_state.json
ADDED
|
@@ -0,0 +1,360 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 8.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 408,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.3137254901960784,
|
| 13 |
+
"grad_norm": 2.9848222732543945,
|
| 14 |
+
"learning_rate": 1.4634146341463413e-07,
|
| 15 |
+
"loss": 9.4214,
|
| 16 |
+
"step": 16
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.6274509803921569,
|
| 20 |
+
"grad_norm": 3.9102113246917725,
|
| 21 |
+
"learning_rate": 3.4146341463414634e-07,
|
| 22 |
+
"loss": 9.4215,
|
| 23 |
+
"step": 32
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.9411764705882353,
|
| 27 |
+
"grad_norm": 3.406308174133301,
|
| 28 |
+
"learning_rate": 5.365853658536586e-07,
|
| 29 |
+
"loss": 9.421,
|
| 30 |
+
"step": 48
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 1.0,
|
| 34 |
+
"eval_bleu": 0.27230328735060844,
|
| 35 |
+
"eval_cap_loss": 2.1642436525400948,
|
| 36 |
+
"eval_con_loss": 2.0591190656026206,
|
| 37 |
+
"eval_loss": 5.200359978512222,
|
| 38 |
+
"step": 51
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.0,
|
| 42 |
+
"eval_bleu": 0.27230328735060844,
|
| 43 |
+
"eval_cap_loss": 2.1642436525400948,
|
| 44 |
+
"eval_con_loss": 2.0591190656026206,
|
| 45 |
+
"eval_loss": 5.200359978512222,
|
| 46 |
+
"eval_runtime": 30.9761,
|
| 47 |
+
"eval_samples_per_second": 26.117,
|
| 48 |
+
"eval_steps_per_second": 3.293,
|
| 49 |
+
"step": 51
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 1.2549019607843137,
|
| 53 |
+
"grad_norm": 2.83813738822937,
|
| 54 |
+
"learning_rate": 7.317073170731707e-07,
|
| 55 |
+
"loss": 9.2807,
|
| 56 |
+
"step": 64
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.5686274509803921,
|
| 60 |
+
"grad_norm": 2.936375379562378,
|
| 61 |
+
"learning_rate": 9.146341463414634e-07,
|
| 62 |
+
"loss": 9.406,
|
| 63 |
+
"step": 80
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.8823529411764706,
|
| 67 |
+
"grad_norm": 3.668606996536255,
|
| 68 |
+
"learning_rate": 9.981206082534287e-07,
|
| 69 |
+
"loss": 9.3986,
|
| 70 |
+
"step": 96
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 2.0,
|
| 74 |
+
"eval_bleu": 0.27320616813251447,
|
| 75 |
+
"eval_cap_loss": 2.1230438223072126,
|
| 76 |
+
"eval_con_loss": 2.059116809975867,
|
| 77 |
+
"eval_loss": 5.179755524677389,
|
| 78 |
+
"step": 102
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 2.0,
|
| 82 |
+
"eval_bleu": 0.27320616813251447,
|
| 83 |
+
"eval_cap_loss": 2.1230438223072126,
|
| 84 |
+
"eval_con_loss": 2.059116809975867,
|
| 85 |
+
"eval_loss": 5.179755524677389,
|
| 86 |
+
"eval_runtime": 31.0,
|
| 87 |
+
"eval_samples_per_second": 26.097,
|
| 88 |
+
"eval_steps_per_second": 3.29,
|
| 89 |
+
"step": 102
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 2.196078431372549,
|
| 93 |
+
"grad_norm": 2.896123170852661,
|
| 94 |
+
"learning_rate": 9.855594766321122e-07,
|
| 95 |
+
"loss": 9.2709,
|
| 96 |
+
"step": 112
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 2.5098039215686274,
|
| 100 |
+
"grad_norm": 2.431732177734375,
|
| 101 |
+
"learning_rate": 9.614774462458572e-07,
|
| 102 |
+
"loss": 9.3793,
|
| 103 |
+
"step": 128
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 2.8235294117647056,
|
| 107 |
+
"grad_norm": 2.8162143230438232,
|
| 108 |
+
"learning_rate": 9.264459128710549e-07,
|
| 109 |
+
"loss": 9.382,
|
| 110 |
+
"step": 144
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 3.0,
|
| 114 |
+
"eval_bleu": 0.27320213706102187,
|
| 115 |
+
"eval_cap_loss": 2.082164978279787,
|
| 116 |
+
"eval_con_loss": 2.0591136591107237,
|
| 117 |
+
"eval_loss": 5.159309838332382,
|
| 118 |
+
"step": 153
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 3.0,
|
| 122 |
+
"eval_bleu": 0.27320213706102187,
|
| 123 |
+
"eval_cap_loss": 2.082164978279787,
|
| 124 |
+
"eval_con_loss": 2.0591136591107237,
|
| 125 |
+
"eval_loss": 5.159309838332382,
|
| 126 |
+
"eval_runtime": 31.2438,
|
| 127 |
+
"eval_samples_per_second": 25.893,
|
| 128 |
+
"eval_steps_per_second": 3.265,
|
| 129 |
+
"step": 153
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 3.1372549019607843,
|
| 133 |
+
"grad_norm": 2.2468996047973633,
|
| 134 |
+
"learning_rate": 8.812960717968818e-07,
|
| 135 |
+
"loss": 9.2376,
|
| 136 |
+
"step": 160
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 3.450980392156863,
|
| 140 |
+
"grad_norm": 2.2097907066345215,
|
| 141 |
+
"learning_rate": 8.27099196004923e-07,
|
| 142 |
+
"loss": 9.3674,
|
| 143 |
+
"step": 176
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 3.764705882352941,
|
| 147 |
+
"grad_norm": 2.2787301540374756,
|
| 148 |
+
"learning_rate": 7.651412180110175e-07,
|
| 149 |
+
"loss": 9.3575,
|
| 150 |
+
"step": 192
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 4.0,
|
| 154 |
+
"eval_bleu": 0.27339830557499767,
|
| 155 |
+
"eval_cap_loss": 2.052475500340555,
|
| 156 |
+
"eval_con_loss": 2.059111627877927,
|
| 157 |
+
"eval_loss": 5.144460982551761,
|
| 158 |
+
"step": 204
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 4.0,
|
| 162 |
+
"eval_bleu": 0.27339830557499767,
|
| 163 |
+
"eval_cap_loss": 2.052475500340555,
|
| 164 |
+
"eval_con_loss": 2.059111627877927,
|
| 165 |
+
"eval_loss": 5.144460982551761,
|
| 166 |
+
"eval_runtime": 31.4567,
|
| 167 |
+
"eval_samples_per_second": 25.718,
|
| 168 |
+
"eval_steps_per_second": 3.243,
|
| 169 |
+
"step": 204
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 4.078431372549019,
|
| 173 |
+
"grad_norm": 2.41097092628479,
|
| 174 |
+
"learning_rate": 6.968922184674867e-07,
|
| 175 |
+
"loss": 9.2198,
|
| 176 |
+
"step": 208
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.392156862745098,
|
| 180 |
+
"grad_norm": 2.1482274532318115,
|
| 181 |
+
"learning_rate": 6.239715454715053e-07,
|
| 182 |
+
"loss": 9.3524,
|
| 183 |
+
"step": 224
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 4.705882352941177,
|
| 187 |
+
"grad_norm": 4.837054252624512,
|
| 188 |
+
"learning_rate": 5.481093921958749e-07,
|
| 189 |
+
"loss": 9.3413,
|
| 190 |
+
"step": 240
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 5.0,
|
| 194 |
+
"eval_bleu": 0.27362478077654184,
|
| 195 |
+
"eval_cap_loss": 2.0312489189353644,
|
| 196 |
+
"eval_con_loss": 2.059109935573503,
|
| 197 |
+
"eval_loss": 5.133844311330833,
|
| 198 |
+
"step": 255
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 5.0,
|
| 202 |
+
"eval_bleu": 0.27362478077654184,
|
| 203 |
+
"eval_cap_loss": 2.0312489189353644,
|
| 204 |
+
"eval_con_loss": 2.059109935573503,
|
| 205 |
+
"eval_loss": 5.133844311330833,
|
| 206 |
+
"eval_runtime": 30.935,
|
| 207 |
+
"eval_samples_per_second": 26.152,
|
| 208 |
+
"eval_steps_per_second": 3.297,
|
| 209 |
+
"step": 255
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 5.019607843137255,
|
| 213 |
+
"grad_norm": 2.2715868949890137,
|
| 214 |
+
"learning_rate": 4.7110574449205214e-07,
|
| 215 |
+
"loss": 9.2202,
|
| 216 |
+
"step": 256
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 5.333333333333333,
|
| 220 |
+
"grad_norm": 2.1526544094085693,
|
| 221 |
+
"learning_rate": 3.9478767251811595e-07,
|
| 222 |
+
"loss": 9.3375,
|
| 223 |
+
"step": 272
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 5.647058823529412,
|
| 227 |
+
"grad_norm": 1.8560357093811035,
|
| 228 |
+
"learning_rate": 3.209659797357669e-07,
|
| 229 |
+
"loss": 9.338,
|
| 230 |
+
"step": 288
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 5.96078431372549,
|
| 234 |
+
"grad_norm": 1.7859069108963013,
|
| 235 |
+
"learning_rate": 2.5139223786820744e-07,
|
| 236 |
+
"loss": 9.3379,
|
| 237 |
+
"step": 304
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 6.0,
|
| 241 |
+
"eval_bleu": 0.27264795125070107,
|
| 242 |
+
"eval_cap_loss": 2.0194136815912582,
|
| 243 |
+
"eval_con_loss": 2.0591090099484313,
|
| 244 |
+
"eval_loss": 5.127924858355055,
|
| 245 |
+
"step": 306
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 6.0,
|
| 249 |
+
"eval_bleu": 0.27264795125070107,
|
| 250 |
+
"eval_cap_loss": 2.0194136815912582,
|
| 251 |
+
"eval_con_loss": 2.0591090099484313,
|
| 252 |
+
"eval_loss": 5.127924858355055,
|
| 253 |
+
"eval_runtime": 31.0894,
|
| 254 |
+
"eval_samples_per_second": 26.022,
|
| 255 |
+
"eval_steps_per_second": 3.281,
|
| 256 |
+
"step": 306
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 6.2745098039215685,
|
| 260 |
+
"grad_norm": 2.1876094341278076,
|
| 261 |
+
"learning_rate": 1.877172272530264e-07,
|
| 262 |
+
"loss": 9.2022,
|
| 263 |
+
"step": 320
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 6.588235294117647,
|
| 267 |
+
"grad_norm": 2.331590414047241,
|
| 268 |
+
"learning_rate": 1.3145176867832165e-07,
|
| 269 |
+
"loss": 9.3296,
|
| 270 |
+
"step": 336
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 6.901960784313726,
|
| 274 |
+
"grad_norm": 2.335200309753418,
|
| 275 |
+
"learning_rate": 8.393087604743283e-08,
|
| 276 |
+
"loss": 9.337,
|
| 277 |
+
"step": 352
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 7.0,
|
| 281 |
+
"eval_bleu": 0.272648714102529,
|
| 282 |
+
"eval_cap_loss": 2.0149140498217415,
|
| 283 |
+
"eval_con_loss": 2.059108528436399,
|
| 284 |
+
"eval_loss": 5.125674068343406,
|
| 285 |
+
"step": 357
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 7.0,
|
| 289 |
+
"eval_bleu": 0.272648714102529,
|
| 290 |
+
"eval_cap_loss": 2.0149140498217415,
|
| 291 |
+
"eval_con_loss": 2.059108528436399,
|
| 292 |
+
"eval_loss": 5.125674068343406,
|
| 293 |
+
"eval_runtime": 31.305,
|
| 294 |
+
"eval_samples_per_second": 25.843,
|
| 295 |
+
"eval_steps_per_second": 3.258,
|
| 296 |
+
"step": 357
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.215686274509804,
|
| 300 |
+
"grad_norm": 2.2809042930603027,
|
| 301 |
+
"learning_rate": 4.6282080424148886e-08,
|
| 302 |
+
"loss": 9.2029,
|
| 303 |
+
"step": 368
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 7.529411764705882,
|
| 307 |
+
"grad_norm": 1.9819755554199219,
|
| 308 |
+
"learning_rate": 1.9398677035671218e-08,
|
| 309 |
+
"loss": 9.3332,
|
| 310 |
+
"step": 384
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 7.8431372549019605,
|
| 314 |
+
"grad_norm": 2.1756012439727783,
|
| 315 |
+
"learning_rate": 3.9185300032889e-09,
|
| 316 |
+
"loss": 9.3308,
|
| 317 |
+
"step": 400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 8.0,
|
| 321 |
+
"eval_bleu": 0.2724553369227519,
|
| 322 |
+
"eval_cap_loss": 2.01423061244628,
|
| 323 |
+
"eval_con_loss": 2.0591084513009763,
|
| 324 |
+
"eval_loss": 5.125332222265356,
|
| 325 |
+
"step": 408
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 8.0,
|
| 329 |
+
"eval_bleu": 0.2724553369227519,
|
| 330 |
+
"eval_cap_loss": 2.01423061244628,
|
| 331 |
+
"eval_con_loss": 2.0591084513009763,
|
| 332 |
+
"eval_loss": 5.125332222265356,
|
| 333 |
+
"eval_runtime": 31.0026,
|
| 334 |
+
"eval_samples_per_second": 26.095,
|
| 335 |
+
"eval_steps_per_second": 3.29,
|
| 336 |
+
"step": 408
|
| 337 |
+
}
|
| 338 |
+
],
|
| 339 |
+
"logging_steps": 16,
|
| 340 |
+
"max_steps": 408,
|
| 341 |
+
"num_input_tokens_seen": 0,
|
| 342 |
+
"num_train_epochs": 8,
|
| 343 |
+
"save_steps": 500,
|
| 344 |
+
"stateful_callbacks": {
|
| 345 |
+
"TrainerControl": {
|
| 346 |
+
"args": {
|
| 347 |
+
"should_epoch_stop": false,
|
| 348 |
+
"should_evaluate": false,
|
| 349 |
+
"should_log": false,
|
| 350 |
+
"should_save": true,
|
| 351 |
+
"should_training_stop": true
|
| 352 |
+
},
|
| 353 |
+
"attributes": {}
|
| 354 |
+
}
|
| 355 |
+
},
|
| 356 |
+
"total_flos": 0.0,
|
| 357 |
+
"train_batch_size": 64,
|
| 358 |
+
"trial_name": null,
|
| 359 |
+
"trial_params": null
|
| 360 |
+
}
|