Spaces:
Sleeping
Sleeping
add all model
Browse files- ModelBert/rng_state.pth +3 -0
- ModelBert/scheduler.pt +3 -0
- ModelBert/trainer_state.json +719 -0
- ModelBert/training_args.bin +3 -0
- ModelBert/vocab.txt +0 -0
ModelBert/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4f91ceb729f390fd5d88dd44c4f32d4927be0346d04d356227f656f2dad30bb
|
| 3 |
+
size 14244
|
ModelBert/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dba4500f662814fce6941b872891c1d53b56fcf56d6faf8c90938b9b2c92f9ee
|
| 3 |
+
size 1064
|
ModelBert/trainer_state.json
ADDED
|
@@ -0,0 +1,719 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.965426072808624,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 49000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1016880211511084,
|
| 13 |
+
"grad_norm": 8.324226379394531,
|
| 14 |
+
"learning_rate": 4.830519964748153e-05,
|
| 15 |
+
"loss": 2.3588,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.2033760423022168,
|
| 20 |
+
"grad_norm": 8.197102546691895,
|
| 21 |
+
"learning_rate": 4.661039929496305e-05,
|
| 22 |
+
"loss": 2.1044,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.3050640634533252,
|
| 27 |
+
"grad_norm": 7.696329593658447,
|
| 28 |
+
"learning_rate": 4.4915598942444584e-05,
|
| 29 |
+
"loss": 2.0132,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.4067520846044336,
|
| 34 |
+
"grad_norm": 7.981374263763428,
|
| 35 |
+
"learning_rate": 4.322079858992611e-05,
|
| 36 |
+
"loss": 1.9548,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.508440105755542,
|
| 41 |
+
"grad_norm": 8.823938369750977,
|
| 42 |
+
"learning_rate": 4.152599823740764e-05,
|
| 43 |
+
"loss": 1.899,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.6101281269066504,
|
| 48 |
+
"grad_norm": 8.75110149383545,
|
| 49 |
+
"learning_rate": 3.983119788488916e-05,
|
| 50 |
+
"loss": 1.8601,
|
| 51 |
+
"step": 3000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.7118161480577588,
|
| 55 |
+
"grad_norm": 11.115782737731934,
|
| 56 |
+
"learning_rate": 3.8136397532370685e-05,
|
| 57 |
+
"loss": 1.8438,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.8135041692088671,
|
| 62 |
+
"grad_norm": 8.88609790802002,
|
| 63 |
+
"learning_rate": 3.644159717985222e-05,
|
| 64 |
+
"loss": 1.8221,
|
| 65 |
+
"step": 4000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.9151921903599756,
|
| 69 |
+
"grad_norm": 9.193504333496094,
|
| 70 |
+
"learning_rate": 3.4746796827333746e-05,
|
| 71 |
+
"loss": 1.8052,
|
| 72 |
+
"step": 4500
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.016880211511084,
|
| 76 |
+
"grad_norm": 9.435020446777344,
|
| 77 |
+
"learning_rate": 3.3051996474815266e-05,
|
| 78 |
+
"loss": 1.7687,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.1185682326621924,
|
| 83 |
+
"grad_norm": 8.944482803344727,
|
| 84 |
+
"learning_rate": 3.135719612229679e-05,
|
| 85 |
+
"loss": 1.7001,
|
| 86 |
+
"step": 5500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.2202562538133008,
|
| 90 |
+
"grad_norm": 9.68911075592041,
|
| 91 |
+
"learning_rate": 2.9662395769778324e-05,
|
| 92 |
+
"loss": 1.7166,
|
| 93 |
+
"step": 6000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.3219442749644092,
|
| 97 |
+
"grad_norm": 8.70807933807373,
|
| 98 |
+
"learning_rate": 2.796759541725985e-05,
|
| 99 |
+
"loss": 1.6879,
|
| 100 |
+
"step": 6500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.4236322961155177,
|
| 104 |
+
"grad_norm": 6.220630645751953,
|
| 105 |
+
"learning_rate": 2.6272795064741374e-05,
|
| 106 |
+
"loss": 1.6574,
|
| 107 |
+
"step": 7000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.525320317266626,
|
| 111 |
+
"grad_norm": 10.505532264709473,
|
| 112 |
+
"learning_rate": 2.45779947122229e-05,
|
| 113 |
+
"loss": 1.6609,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.6270083384177343,
|
| 118 |
+
"grad_norm": 6.303832054138184,
|
| 119 |
+
"learning_rate": 2.2883194359704428e-05,
|
| 120 |
+
"loss": 1.6455,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.728696359568843,
|
| 125 |
+
"grad_norm": 8.265469551086426,
|
| 126 |
+
"learning_rate": 2.1188394007185955e-05,
|
| 127 |
+
"loss": 1.6164,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.8303843807199511,
|
| 132 |
+
"grad_norm": 7.970987319946289,
|
| 133 |
+
"learning_rate": 1.949359365466748e-05,
|
| 134 |
+
"loss": 1.6029,
|
| 135 |
+
"step": 9000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.9320724018710596,
|
| 139 |
+
"grad_norm": 9.984902381896973,
|
| 140 |
+
"learning_rate": 1.779879330214901e-05,
|
| 141 |
+
"loss": 1.6018,
|
| 142 |
+
"step": 9500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.033760423022168,
|
| 146 |
+
"grad_norm": 7.69148588180542,
|
| 147 |
+
"learning_rate": 1.6103992949630533e-05,
|
| 148 |
+
"loss": 1.6092,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 2.135448444173276,
|
| 153 |
+
"grad_norm": 6.372833251953125,
|
| 154 |
+
"learning_rate": 1.4409192597112062e-05,
|
| 155 |
+
"loss": 1.5372,
|
| 156 |
+
"step": 10500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 2.237136465324385,
|
| 160 |
+
"grad_norm": 7.225351333618164,
|
| 161 |
+
"learning_rate": 1.2714392244593587e-05,
|
| 162 |
+
"loss": 1.5452,
|
| 163 |
+
"step": 11000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 2.338824486475493,
|
| 167 |
+
"grad_norm": 8.18689250946045,
|
| 168 |
+
"learning_rate": 1.1019591892075114e-05,
|
| 169 |
+
"loss": 1.539,
|
| 170 |
+
"step": 11500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 2.4405125076266017,
|
| 174 |
+
"grad_norm": 7.346330165863037,
|
| 175 |
+
"learning_rate": 9.324791539556641e-06,
|
| 176 |
+
"loss": 1.5227,
|
| 177 |
+
"step": 12000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 2.54220052877771,
|
| 181 |
+
"grad_norm": 9.162985801696777,
|
| 182 |
+
"learning_rate": 7.629991187038167e-06,
|
| 183 |
+
"loss": 1.5207,
|
| 184 |
+
"step": 12500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 2.6438885499288185,
|
| 188 |
+
"grad_norm": 7.583765029907227,
|
| 189 |
+
"learning_rate": 5.935190834519694e-06,
|
| 190 |
+
"loss": 1.5172,
|
| 191 |
+
"step": 13000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 2.7455765710799267,
|
| 195 |
+
"grad_norm": 7.970169544219971,
|
| 196 |
+
"learning_rate": 4.240390482001221e-06,
|
| 197 |
+
"loss": 1.5272,
|
| 198 |
+
"step": 13500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 2.8472645922310353,
|
| 202 |
+
"grad_norm": 9.05431079864502,
|
| 203 |
+
"learning_rate": 2.5455901294827473e-06,
|
| 204 |
+
"loss": 1.5014,
|
| 205 |
+
"step": 14000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 2.9489526133821435,
|
| 209 |
+
"grad_norm": 8.572937965393066,
|
| 210 |
+
"learning_rate": 8.507897769642736e-07,
|
| 211 |
+
"loss": 1.4757,
|
| 212 |
+
"step": 14500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.050640634533252,
|
| 216 |
+
"grad_norm": 8.990861892700195,
|
| 217 |
+
"learning_rate": 3.0933496034167175e-05,
|
| 218 |
+
"loss": 1.5368,
|
| 219 |
+
"step": 15000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.1523286556843604,
|
| 223 |
+
"grad_norm": 8.719560623168945,
|
| 224 |
+
"learning_rate": 3.0297945901972747e-05,
|
| 225 |
+
"loss": 1.5287,
|
| 226 |
+
"step": 15500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 3.2540166768354686,
|
| 230 |
+
"grad_norm": 8.392940521240234,
|
| 231 |
+
"learning_rate": 2.9662395769778324e-05,
|
| 232 |
+
"loss": 1.5461,
|
| 233 |
+
"step": 16000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 3.3557046979865772,
|
| 237 |
+
"grad_norm": 8.443445205688477,
|
| 238 |
+
"learning_rate": 2.9026845637583893e-05,
|
| 239 |
+
"loss": 1.5664,
|
| 240 |
+
"step": 16500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 3.4573927191376854,
|
| 244 |
+
"grad_norm": 7.7863850593566895,
|
| 245 |
+
"learning_rate": 2.8391295505389465e-05,
|
| 246 |
+
"loss": 1.5471,
|
| 247 |
+
"step": 17000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 3.559080740288794,
|
| 251 |
+
"grad_norm": 9.05781364440918,
|
| 252 |
+
"learning_rate": 2.775574537319504e-05,
|
| 253 |
+
"loss": 1.5505,
|
| 254 |
+
"step": 17500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 3.6607687614399023,
|
| 258 |
+
"grad_norm": 6.635441303253174,
|
| 259 |
+
"learning_rate": 2.712019524100061e-05,
|
| 260 |
+
"loss": 1.518,
|
| 261 |
+
"step": 18000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 3.762456782591011,
|
| 265 |
+
"grad_norm": 8.673697471618652,
|
| 266 |
+
"learning_rate": 2.6484645108806183e-05,
|
| 267 |
+
"loss": 1.5018,
|
| 268 |
+
"step": 18500
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 3.864144803742119,
|
| 272 |
+
"grad_norm": 9.876930236816406,
|
| 273 |
+
"learning_rate": 2.5849094976611753e-05,
|
| 274 |
+
"loss": 1.5348,
|
| 275 |
+
"step": 19000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 3.9658328248932273,
|
| 279 |
+
"grad_norm": 8.771023750305176,
|
| 280 |
+
"learning_rate": 2.5213544844417332e-05,
|
| 281 |
+
"loss": 1.4903,
|
| 282 |
+
"step": 19500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 4.067520846044336,
|
| 286 |
+
"grad_norm": 7.6693501472473145,
|
| 287 |
+
"learning_rate": 2.45779947122229e-05,
|
| 288 |
+
"loss": 1.4806,
|
| 289 |
+
"step": 20000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 4.169208867195445,
|
| 293 |
+
"grad_norm": 6.399499416351318,
|
| 294 |
+
"learning_rate": 2.3942444580028474e-05,
|
| 295 |
+
"loss": 1.4807,
|
| 296 |
+
"step": 20500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 4.270896888346552,
|
| 300 |
+
"grad_norm": 8.920206069946289,
|
| 301 |
+
"learning_rate": 2.3306894447834047e-05,
|
| 302 |
+
"loss": 1.4578,
|
| 303 |
+
"step": 21000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 4.372584909497661,
|
| 307 |
+
"grad_norm": 8.65735912322998,
|
| 308 |
+
"learning_rate": 2.267134431563962e-05,
|
| 309 |
+
"loss": 1.468,
|
| 310 |
+
"step": 21500
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 4.47427293064877,
|
| 314 |
+
"grad_norm": 5.02122688293457,
|
| 315 |
+
"learning_rate": 2.2035794183445192e-05,
|
| 316 |
+
"loss": 1.4579,
|
| 317 |
+
"step": 22000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 4.575960951799878,
|
| 321 |
+
"grad_norm": 8.269110679626465,
|
| 322 |
+
"learning_rate": 2.1400244051250764e-05,
|
| 323 |
+
"loss": 1.4484,
|
| 324 |
+
"step": 22500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 4.677648972950986,
|
| 328 |
+
"grad_norm": 7.763547897338867,
|
| 329 |
+
"learning_rate": 2.0764693919056337e-05,
|
| 330 |
+
"loss": 1.4554,
|
| 331 |
+
"step": 23000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 4.779336994102095,
|
| 335 |
+
"grad_norm": 9.097722053527832,
|
| 336 |
+
"learning_rate": 2.012914378686191e-05,
|
| 337 |
+
"loss": 1.4531,
|
| 338 |
+
"step": 23500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 4.881025015253203,
|
| 342 |
+
"grad_norm": 8.873741149902344,
|
| 343 |
+
"learning_rate": 1.949359365466748e-05,
|
| 344 |
+
"loss": 1.4677,
|
| 345 |
+
"step": 24000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 4.982713036404312,
|
| 349 |
+
"grad_norm": 8.350312232971191,
|
| 350 |
+
"learning_rate": 1.8858043522473055e-05,
|
| 351 |
+
"loss": 1.4306,
|
| 352 |
+
"step": 24500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.08440105755542,
|
| 356 |
+
"grad_norm": 5.5741143226623535,
|
| 357 |
+
"learning_rate": 1.8222493390278624e-05,
|
| 358 |
+
"loss": 1.4203,
|
| 359 |
+
"step": 25000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.186089078706528,
|
| 363 |
+
"grad_norm": 9.378081321716309,
|
| 364 |
+
"learning_rate": 1.75869432580842e-05,
|
| 365 |
+
"loss": 1.4165,
|
| 366 |
+
"step": 25500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 5.287777099857637,
|
| 370 |
+
"grad_norm": 8.990020751953125,
|
| 371 |
+
"learning_rate": 1.6951393125889773e-05,
|
| 372 |
+
"loss": 1.3941,
|
| 373 |
+
"step": 26000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 5.389465121008746,
|
| 377 |
+
"grad_norm": 8.84300708770752,
|
| 378 |
+
"learning_rate": 1.6315842993695342e-05,
|
| 379 |
+
"loss": 1.3888,
|
| 380 |
+
"step": 26500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 5.491153142159853,
|
| 384 |
+
"grad_norm": 7.583876132965088,
|
| 385 |
+
"learning_rate": 1.5680292861500918e-05,
|
| 386 |
+
"loss": 1.3947,
|
| 387 |
+
"step": 27000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 5.592841163310962,
|
| 391 |
+
"grad_norm": 7.263258934020996,
|
| 392 |
+
"learning_rate": 1.5044742729306487e-05,
|
| 393 |
+
"loss": 1.3716,
|
| 394 |
+
"step": 27500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 5.694529184462071,
|
| 398 |
+
"grad_norm": 9.45223331451416,
|
| 399 |
+
"learning_rate": 1.4409192597112062e-05,
|
| 400 |
+
"loss": 1.3853,
|
| 401 |
+
"step": 28000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 5.796217205613178,
|
| 405 |
+
"grad_norm": 6.165616989135742,
|
| 406 |
+
"learning_rate": 1.3773642464917633e-05,
|
| 407 |
+
"loss": 1.3963,
|
| 408 |
+
"step": 28500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 5.897905226764287,
|
| 412 |
+
"grad_norm": 7.537690162658691,
|
| 413 |
+
"learning_rate": 1.3138092332723207e-05,
|
| 414 |
+
"loss": 1.3644,
|
| 415 |
+
"step": 29000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 5.999593247915396,
|
| 419 |
+
"grad_norm": 9.683405876159668,
|
| 420 |
+
"learning_rate": 1.2502542200528778e-05,
|
| 421 |
+
"loss": 1.3892,
|
| 422 |
+
"step": 29500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 6.101281269066504,
|
| 426 |
+
"grad_norm": 8.821761131286621,
|
| 427 |
+
"learning_rate": 1.186699206833435e-05,
|
| 428 |
+
"loss": 1.3335,
|
| 429 |
+
"step": 30000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 6.202969290217612,
|
| 433 |
+
"grad_norm": 12.081774711608887,
|
| 434 |
+
"learning_rate": 1.1231441936139923e-05,
|
| 435 |
+
"loss": 1.3526,
|
| 436 |
+
"step": 30500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 6.304657311368721,
|
| 440 |
+
"grad_norm": 6.325799465179443,
|
| 441 |
+
"learning_rate": 1.0595891803945496e-05,
|
| 442 |
+
"loss": 1.3424,
|
| 443 |
+
"step": 31000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 6.406345332519829,
|
| 447 |
+
"grad_norm": 9.028151512145996,
|
| 448 |
+
"learning_rate": 9.960341671751069e-06,
|
| 449 |
+
"loss": 1.3693,
|
| 450 |
+
"step": 31500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 6.508033353670937,
|
| 454 |
+
"grad_norm": 6.508680820465088,
|
| 455 |
+
"learning_rate": 9.324791539556641e-06,
|
| 456 |
+
"loss": 1.3506,
|
| 457 |
+
"step": 32000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 6.609721374822046,
|
| 461 |
+
"grad_norm": 6.9301629066467285,
|
| 462 |
+
"learning_rate": 8.689241407362212e-06,
|
| 463 |
+
"loss": 1.3426,
|
| 464 |
+
"step": 32500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 6.7114093959731544,
|
| 468 |
+
"grad_norm": 8.2422456741333,
|
| 469 |
+
"learning_rate": 8.053691275167785e-06,
|
| 470 |
+
"loss": 1.3337,
|
| 471 |
+
"step": 33000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 6.813097417124263,
|
| 475 |
+
"grad_norm": 7.615689277648926,
|
| 476 |
+
"learning_rate": 7.4181411429733575e-06,
|
| 477 |
+
"loss": 1.3322,
|
| 478 |
+
"step": 33500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 6.914785438275371,
|
| 482 |
+
"grad_norm": 7.60098123550415,
|
| 483 |
+
"learning_rate": 6.782591010778931e-06,
|
| 484 |
+
"loss": 1.3182,
|
| 485 |
+
"step": 34000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 7.0164734594264795,
|
| 489 |
+
"grad_norm": 8.578591346740723,
|
| 490 |
+
"learning_rate": 6.147040878584503e-06,
|
| 491 |
+
"loss": 1.3333,
|
| 492 |
+
"step": 34500
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 7.118161480577588,
|
| 496 |
+
"grad_norm": 8.674110412597656,
|
| 497 |
+
"learning_rate": 5.511490746390076e-06,
|
| 498 |
+
"loss": 1.2914,
|
| 499 |
+
"step": 35000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 7.219849501728697,
|
| 503 |
+
"grad_norm": 7.355325222015381,
|
| 504 |
+
"learning_rate": 4.875940614195648e-06,
|
| 505 |
+
"loss": 1.3241,
|
| 506 |
+
"step": 35500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 7.3215375228798045,
|
| 510 |
+
"grad_norm": 8.176190376281738,
|
| 511 |
+
"learning_rate": 4.240390482001221e-06,
|
| 512 |
+
"loss": 1.3184,
|
| 513 |
+
"step": 36000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 7.423225544030913,
|
| 517 |
+
"grad_norm": 8.22246265411377,
|
| 518 |
+
"learning_rate": 3.604840349806793e-06,
|
| 519 |
+
"loss": 1.3011,
|
| 520 |
+
"step": 36500
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 7.524913565182022,
|
| 524 |
+
"grad_norm": 9.817005157470703,
|
| 525 |
+
"learning_rate": 2.9692902176123655e-06,
|
| 526 |
+
"loss": 1.3044,
|
| 527 |
+
"step": 37000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 7.6266015863331305,
|
| 531 |
+
"grad_norm": 8.255529403686523,
|
| 532 |
+
"learning_rate": 2.3337400854179377e-06,
|
| 533 |
+
"loss": 1.2964,
|
| 534 |
+
"step": 37500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 7.728289607484238,
|
| 538 |
+
"grad_norm": 6.6958160400390625,
|
| 539 |
+
"learning_rate": 1.6981899532235102e-06,
|
| 540 |
+
"loss": 1.3092,
|
| 541 |
+
"step": 38000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 7.829977628635347,
|
| 545 |
+
"grad_norm": 8.960187911987305,
|
| 546 |
+
"learning_rate": 1.062639821029083e-06,
|
| 547 |
+
"loss": 1.2976,
|
| 548 |
+
"step": 38500
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 7.9316656497864555,
|
| 552 |
+
"grad_norm": 7.582452297210693,
|
| 553 |
+
"learning_rate": 4.2708968883465534e-07,
|
| 554 |
+
"loss": 1.2956,
|
| 555 |
+
"step": 39000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 8.033353670937563,
|
| 559 |
+
"grad_norm": 7.20064115524292,
|
| 560 |
+
"learning_rate": 9.833231645312182e-06,
|
| 561 |
+
"loss": 1.3157,
|
| 562 |
+
"step": 39500
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 8.135041692088672,
|
| 566 |
+
"grad_norm": 7.36190128326416,
|
| 567 |
+
"learning_rate": 9.324791539556641e-06,
|
| 568 |
+
"loss": 1.3021,
|
| 569 |
+
"step": 40000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 8.23672971323978,
|
| 573 |
+
"grad_norm": 9.36536979675293,
|
| 574 |
+
"learning_rate": 8.816351433801099e-06,
|
| 575 |
+
"loss": 1.3073,
|
| 576 |
+
"step": 40500
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 8.33841773439089,
|
| 580 |
+
"grad_norm": 7.667935848236084,
|
| 581 |
+
"learning_rate": 8.307911328045556e-06,
|
| 582 |
+
"loss": 1.2902,
|
| 583 |
+
"step": 41000
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 8.440105755541998,
|
| 587 |
+
"grad_norm": 8.367385864257812,
|
| 588 |
+
"learning_rate": 7.799471222290014e-06,
|
| 589 |
+
"loss": 1.3164,
|
| 590 |
+
"step": 41500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 8.541793776693105,
|
| 594 |
+
"grad_norm": 8.879359245300293,
|
| 595 |
+
"learning_rate": 7.291031116534472e-06,
|
| 596 |
+
"loss": 1.2927,
|
| 597 |
+
"step": 42000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 8.643481797844213,
|
| 601 |
+
"grad_norm": 7.432185649871826,
|
| 602 |
+
"learning_rate": 6.782591010778931e-06,
|
| 603 |
+
"loss": 1.3027,
|
| 604 |
+
"step": 42500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 8.745169818995322,
|
| 608 |
+
"grad_norm": 7.75663423538208,
|
| 609 |
+
"learning_rate": 6.274150905023389e-06,
|
| 610 |
+
"loss": 1.3266,
|
| 611 |
+
"step": 43000
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 8.84685784014643,
|
| 615 |
+
"grad_norm": 10.648377418518066,
|
| 616 |
+
"learning_rate": 5.765710799267846e-06,
|
| 617 |
+
"loss": 1.263,
|
| 618 |
+
"step": 43500
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 8.94854586129754,
|
| 622 |
+
"grad_norm": 8.594388961791992,
|
| 623 |
+
"learning_rate": 5.257270693512305e-06,
|
| 624 |
+
"loss": 1.3049,
|
| 625 |
+
"step": 44000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 9.050233882448648,
|
| 629 |
+
"grad_norm": 8.348444938659668,
|
| 630 |
+
"learning_rate": 4.748830587756762e-06,
|
| 631 |
+
"loss": 1.2984,
|
| 632 |
+
"step": 44500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 9.151921903599757,
|
| 636 |
+
"grad_norm": 9.689435958862305,
|
| 637 |
+
"learning_rate": 4.240390482001221e-06,
|
| 638 |
+
"loss": 1.2609,
|
| 639 |
+
"step": 45000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 9.253609924750865,
|
| 643 |
+
"grad_norm": 7.50744104385376,
|
| 644 |
+
"learning_rate": 3.7319503762456785e-06,
|
| 645 |
+
"loss": 1.2705,
|
| 646 |
+
"step": 45500
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 9.355297945901972,
|
| 650 |
+
"grad_norm": 7.884949684143066,
|
| 651 |
+
"learning_rate": 3.2235102704901364e-06,
|
| 652 |
+
"loss": 1.2949,
|
| 653 |
+
"step": 46000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 9.45698596705308,
|
| 657 |
+
"grad_norm": 7.593263626098633,
|
| 658 |
+
"learning_rate": 2.7150701647345947e-06,
|
| 659 |
+
"loss": 1.2635,
|
| 660 |
+
"step": 46500
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 9.55867398820419,
|
| 664 |
+
"grad_norm": 6.893133640289307,
|
| 665 |
+
"learning_rate": 2.2066300589790525e-06,
|
| 666 |
+
"loss": 1.2811,
|
| 667 |
+
"step": 47000
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 9.660362009355298,
|
| 671 |
+
"grad_norm": 9.131689071655273,
|
| 672 |
+
"learning_rate": 1.6981899532235102e-06,
|
| 673 |
+
"loss": 1.2854,
|
| 674 |
+
"step": 47500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 9.762050030506407,
|
| 678 |
+
"grad_norm": 6.920133113861084,
|
| 679 |
+
"learning_rate": 1.1897498474679682e-06,
|
| 680 |
+
"loss": 1.306,
|
| 681 |
+
"step": 48000
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 9.863738051657515,
|
| 685 |
+
"grad_norm": 7.8755950927734375,
|
| 686 |
+
"learning_rate": 6.813097417124263e-07,
|
| 687 |
+
"loss": 1.2769,
|
| 688 |
+
"step": 48500
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 9.965426072808624,
|
| 692 |
+
"grad_norm": 9.283025741577148,
|
| 693 |
+
"learning_rate": 1.728696359568843e-07,
|
| 694 |
+
"loss": 1.2504,
|
| 695 |
+
"step": 49000
|
| 696 |
+
}
|
| 697 |
+
],
|
| 698 |
+
"logging_steps": 500,
|
| 699 |
+
"max_steps": 49170,
|
| 700 |
+
"num_input_tokens_seen": 0,
|
| 701 |
+
"num_train_epochs": 10,
|
| 702 |
+
"save_steps": 500,
|
| 703 |
+
"stateful_callbacks": {
|
| 704 |
+
"TrainerControl": {
|
| 705 |
+
"args": {
|
| 706 |
+
"should_epoch_stop": false,
|
| 707 |
+
"should_evaluate": false,
|
| 708 |
+
"should_log": false,
|
| 709 |
+
"should_save": true,
|
| 710 |
+
"should_training_stop": false
|
| 711 |
+
},
|
| 712 |
+
"attributes": {}
|
| 713 |
+
}
|
| 714 |
+
},
|
| 715 |
+
"total_flos": 2.57905183894272e+16,
|
| 716 |
+
"train_batch_size": 8,
|
| 717 |
+
"trial_name": null,
|
| 718 |
+
"trial_params": null
|
| 719 |
+
}
|
ModelBert/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fd54889ffc58d6429775286ff45dc3e4aae0b2b0e45c442d5f31bbafca661ee6
|
| 3 |
+
size 5112
|
ModelBert/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|