Training in progress, epoch 0
Browse files- adapter_config.json +4 -2
- adapter_model.safetensors +2 -2
- trainer_state.json +639 -0
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -20,10 +20,12 @@
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
-
"k_proj",
|
| 24 |
"v_proj",
|
|
|
|
| 25 |
"q_proj",
|
| 26 |
-
"
|
|
|
|
|
|
|
| 27 |
],
|
| 28 |
"task_type": "CAUSAL_LM",
|
| 29 |
"use_dora": false,
|
|
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
|
|
|
| 23 |
"v_proj",
|
| 24 |
+
"fc1",
|
| 25 |
"q_proj",
|
| 26 |
+
"k_proj",
|
| 27 |
+
"fc2",
|
| 28 |
+
"dense"
|
| 29 |
],
|
| 30 |
"task_type": "CAUSAL_LM",
|
| 31 |
"use_dora": false,
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4b1b4c2bf412d25e81a1e930e6646e340fdb9da06243a818484153349e9339cb
|
| 3 |
+
size 94422368
|
trainer_state.json
ADDED
|
@@ -0,0 +1,639 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.914529914529915,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 870,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.11396011396011396,
|
| 13 |
+
"grad_norm": 0.158817857503891,
|
| 14 |
+
"learning_rate": 0.0001999348095389677,
|
| 15 |
+
"loss": 0.9924,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.22792022792022792,
|
| 20 |
+
"grad_norm": 0.21280939877033234,
|
| 21 |
+
"learning_rate": 0.000199739323151795,
|
| 22 |
+
"loss": 0.819,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.3418803418803419,
|
| 27 |
+
"grad_norm": 0.22974510490894318,
|
| 28 |
+
"learning_rate": 0.00019941379571543596,
|
| 29 |
+
"loss": 0.767,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.45584045584045585,
|
| 34 |
+
"grad_norm": 0.20720455050468445,
|
| 35 |
+
"learning_rate": 0.00019895865165556377,
|
| 36 |
+
"loss": 0.6948,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.5698005698005698,
|
| 41 |
+
"grad_norm": 0.1902514398097992,
|
| 42 |
+
"learning_rate": 0.00019837448439320027,
|
| 43 |
+
"loss": 0.6509,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.6837606837606838,
|
| 48 |
+
"grad_norm": 0.18335820734500885,
|
| 49 |
+
"learning_rate": 0.00019766205557100868,
|
| 50 |
+
"loss": 0.6344,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.7977207977207977,
|
| 55 |
+
"grad_norm": 0.17900069057941437,
|
| 56 |
+
"learning_rate": 0.00019682229406025635,
|
| 57 |
+
"loss": 0.6447,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.9116809116809117,
|
| 62 |
+
"grad_norm": 0.16915330290794373,
|
| 63 |
+
"learning_rate": 0.00019585629474974415,
|
| 64 |
+
"loss": 0.6335,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 1.0256410256410255,
|
| 69 |
+
"grad_norm": 0.16036000847816467,
|
| 70 |
+
"learning_rate": 0.00019476531711828027,
|
| 71 |
+
"loss": 0.634,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.1396011396011396,
|
| 76 |
+
"grad_norm": 0.16852639615535736,
|
| 77 |
+
"learning_rate": 0.0001935507835925601,
|
| 78 |
+
"loss": 0.6058,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.2535612535612537,
|
| 83 |
+
"grad_norm": 0.15856905281543732,
|
| 84 |
+
"learning_rate": 0.00019221427769259333,
|
| 85 |
+
"loss": 0.5902,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.3675213675213675,
|
| 90 |
+
"grad_norm": 0.16909192502498627,
|
| 91 |
+
"learning_rate": 0.00019075754196709572,
|
| 92 |
+
"loss": 0.6051,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.4814814814814814,
|
| 97 |
+
"grad_norm": 0.1899166703224182,
|
| 98 |
+
"learning_rate": 0.00018918247572153823,
|
| 99 |
+
"loss": 0.6098,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.5954415954415955,
|
| 104 |
+
"grad_norm": 0.17596793174743652,
|
| 105 |
+
"learning_rate": 0.00018749113254181498,
|
| 106 |
+
"loss": 0.597,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.7094017094017095,
|
| 111 |
+
"grad_norm": 0.16560517251491547,
|
| 112 |
+
"learning_rate": 0.00018568571761675893,
|
| 113 |
+
"loss": 0.5899,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.8233618233618234,
|
| 118 |
+
"grad_norm": 0.16513986885547638,
|
| 119 |
+
"learning_rate": 0.00018376858486299647,
|
| 120 |
+
"loss": 0.5989,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.9373219373219372,
|
| 125 |
+
"grad_norm": 0.20360782742500305,
|
| 126 |
+
"learning_rate": 0.00018174223385588917,
|
| 127 |
+
"loss": 0.5982,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.051282051282051,
|
| 132 |
+
"grad_norm": 0.16155321896076202,
|
| 133 |
+
"learning_rate": 0.00017960930657056438,
|
| 134 |
+
"loss": 0.593,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.1652421652421654,
|
| 139 |
+
"grad_norm": 0.1811763048171997,
|
| 140 |
+
"learning_rate": 0.00017737258393728364,
|
| 141 |
+
"loss": 0.6077,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.2792022792022792,
|
| 146 |
+
"grad_norm": 0.16952063143253326,
|
| 147 |
+
"learning_rate": 0.00017503498221564025,
|
| 148 |
+
"loss": 0.5749,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 2.393162393162393,
|
| 153 |
+
"grad_norm": 0.17240603268146515,
|
| 154 |
+
"learning_rate": 0.0001725995491923131,
|
| 155 |
+
"loss": 0.5592,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 2.5071225071225074,
|
| 160 |
+
"grad_norm": 0.1657334417104721,
|
| 161 |
+
"learning_rate": 0.00017006946020733425,
|
| 162 |
+
"loss": 0.5779,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 2.6210826210826212,
|
| 167 |
+
"grad_norm": 0.16417497396469116,
|
| 168 |
+
"learning_rate": 0.0001674480140140514,
|
| 169 |
+
"loss": 0.5675,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 2.735042735042735,
|
| 174 |
+
"grad_norm": 0.174308180809021,
|
| 175 |
+
"learning_rate": 0.00016473862847818277,
|
| 176 |
+
"loss": 0.5977,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 2.849002849002849,
|
| 181 |
+
"grad_norm": 0.17116901278495789,
|
| 182 |
+
"learning_rate": 0.0001619448361215723,
|
| 183 |
+
"loss": 0.5582,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 2.962962962962963,
|
| 188 |
+
"grad_norm": 0.16816489398479462,
|
| 189 |
+
"learning_rate": 0.0001590702795164551,
|
| 190 |
+
"loss": 0.5813,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 3.076923076923077,
|
| 195 |
+
"grad_norm": 0.17530137300491333,
|
| 196 |
+
"learning_rate": 0.00015611870653623825,
|
| 197 |
+
"loss": 0.559,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 3.190883190883191,
|
| 202 |
+
"grad_norm": 0.1744232326745987,
|
| 203 |
+
"learning_rate": 0.0001530939654689887,
|
| 204 |
+
"loss": 0.5668,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 3.304843304843305,
|
| 209 |
+
"grad_norm": 0.1809006929397583,
|
| 210 |
+
"learning_rate": 0.00015000000000000001,
|
| 211 |
+
"loss": 0.5754,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.4188034188034186,
|
| 216 |
+
"grad_norm": 0.16484159231185913,
|
| 217 |
+
"learning_rate": 0.00014684084406997903,
|
| 218 |
+
"loss": 0.5731,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.532763532763533,
|
| 223 |
+
"grad_norm": 0.19075918197631836,
|
| 224 |
+
"learning_rate": 0.00014362061661555675,
|
| 225 |
+
"loss": 0.5496,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 3.646723646723647,
|
| 230 |
+
"grad_norm": 0.18451079726219177,
|
| 231 |
+
"learning_rate": 0.00014034351619898088,
|
| 232 |
+
"loss": 0.5463,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 3.7606837606837606,
|
| 237 |
+
"grad_norm": 0.18566997349262238,
|
| 238 |
+
"learning_rate": 0.00013701381553399145,
|
| 239 |
+
"loss": 0.5768,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 3.8746438746438745,
|
| 244 |
+
"grad_norm": 0.1669853925704956,
|
| 245 |
+
"learning_rate": 0.0001336358559150175,
|
| 246 |
+
"loss": 0.5606,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 3.9886039886039883,
|
| 251 |
+
"grad_norm": 0.17847082018852234,
|
| 252 |
+
"learning_rate": 0.00013021404155695725,
|
| 253 |
+
"loss": 0.5756,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 4.102564102564102,
|
| 258 |
+
"grad_norm": 0.16660483181476593,
|
| 259 |
+
"learning_rate": 0.00012675283385292212,
|
| 260 |
+
"loss": 0.5585,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 4.216524216524217,
|
| 265 |
+
"grad_norm": 0.17163340747356415,
|
| 266 |
+
"learning_rate": 0.00012325674555743106,
|
| 267 |
+
"loss": 0.5434,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 4.330484330484331,
|
| 272 |
+
"grad_norm": 0.16264410316944122,
|
| 273 |
+
"learning_rate": 0.00011973033490264001,
|
| 274 |
+
"loss": 0.5449,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 4.444444444444445,
|
| 279 |
+
"grad_norm": 0.17614829540252686,
|
| 280 |
+
"learning_rate": 0.0001161781996552765,
|
| 281 |
+
"loss": 0.5574,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 4.5584045584045585,
|
| 286 |
+
"grad_norm": 0.19437584280967712,
|
| 287 |
+
"learning_rate": 0.00011260497112202895,
|
| 288 |
+
"loss": 0.5448,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 4.672364672364672,
|
| 293 |
+
"grad_norm": 0.19045701622962952,
|
| 294 |
+
"learning_rate": 0.00010901530811120655,
|
| 295 |
+
"loss": 0.5474,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 4.786324786324786,
|
| 300 |
+
"grad_norm": 0.21330882608890533,
|
| 301 |
+
"learning_rate": 0.00010541389085854176,
|
| 302 |
+
"loss": 0.5552,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 4.9002849002849,
|
| 307 |
+
"grad_norm": 0.17429402470588684,
|
| 308 |
+
"learning_rate": 0.00010180541492505604,
|
| 309 |
+
"loss": 0.5495,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.014245014245014,
|
| 314 |
+
"grad_norm": 0.17785826325416565,
|
| 315 |
+
"learning_rate": 9.819458507494394e-05,
|
| 316 |
+
"loss": 0.5583,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.128205128205128,
|
| 321 |
+
"grad_norm": 0.19076977670192719,
|
| 322 |
+
"learning_rate": 9.458610914145826e-05,
|
| 323 |
+
"loss": 0.5291,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.2421652421652425,
|
| 328 |
+
"grad_norm": 0.19988471269607544,
|
| 329 |
+
"learning_rate": 9.098469188879349e-05,
|
| 330 |
+
"loss": 0.5311,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.356125356125356,
|
| 335 |
+
"grad_norm": 0.19638335704803467,
|
| 336 |
+
"learning_rate": 8.739502887797107e-05,
|
| 337 |
+
"loss": 0.5684,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.47008547008547,
|
| 342 |
+
"grad_norm": 0.2043437659740448,
|
| 343 |
+
"learning_rate": 8.382180034472353e-05,
|
| 344 |
+
"loss": 0.5371,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.584045584045584,
|
| 349 |
+
"grad_norm": 0.2045976221561432,
|
| 350 |
+
"learning_rate": 8.026966509736001e-05,
|
| 351 |
+
"loss": 0.5307,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.698005698005698,
|
| 356 |
+
"grad_norm": 0.21237310767173767,
|
| 357 |
+
"learning_rate": 7.674325444256899e-05,
|
| 358 |
+
"loss": 0.5483,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.811965811965812,
|
| 363 |
+
"grad_norm": 0.22306476533412933,
|
| 364 |
+
"learning_rate": 7.324716614707793e-05,
|
| 365 |
+
"loss": 0.5572,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 5.925925925925926,
|
| 370 |
+
"grad_norm": 0.20065273344516754,
|
| 371 |
+
"learning_rate": 6.978595844304271e-05,
|
| 372 |
+
"loss": 0.5363,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 6.0398860398860394,
|
| 377 |
+
"grad_norm": 0.21213628351688385,
|
| 378 |
+
"learning_rate": 6.636414408498249e-05,
|
| 379 |
+
"loss": 0.521,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.153846153846154,
|
| 384 |
+
"grad_norm": 0.1936779022216797,
|
| 385 |
+
"learning_rate": 6.298618446600856e-05,
|
| 386 |
+
"loss": 0.5283,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 6.267806267806268,
|
| 391 |
+
"grad_norm": 0.19564631581306458,
|
| 392 |
+
"learning_rate": 5.965648380101916e-05,
|
| 393 |
+
"loss": 0.5301,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 6.381766381766382,
|
| 398 |
+
"grad_norm": 0.20069913566112518,
|
| 399 |
+
"learning_rate": 5.6379383384443255e-05,
|
| 400 |
+
"loss": 0.5204,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 6.495726495726496,
|
| 405 |
+
"grad_norm": 0.21325626969337463,
|
| 406 |
+
"learning_rate": 5.3159155930021e-05,
|
| 407 |
+
"loss": 0.5419,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 6.60968660968661,
|
| 412 |
+
"grad_norm": 0.21303197741508484,
|
| 413 |
+
"learning_rate": 5.000000000000002e-05,
|
| 414 |
+
"loss": 0.543,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 6.7236467236467234,
|
| 419 |
+
"grad_norm": 0.21136346459388733,
|
| 420 |
+
"learning_rate": 4.6906034531011346e-05,
|
| 421 |
+
"loss": 0.5217,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 6.837606837606837,
|
| 426 |
+
"grad_norm": 0.21392931044101715,
|
| 427 |
+
"learning_rate": 4.388129346376178e-05,
|
| 428 |
+
"loss": 0.5288,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 6.951566951566951,
|
| 433 |
+
"grad_norm": 0.22880437970161438,
|
| 434 |
+
"learning_rate": 4.092972048354491e-05,
|
| 435 |
+
"loss": 0.5273,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 7.065527065527066,
|
| 440 |
+
"grad_norm": 0.21491903066635132,
|
| 441 |
+
"learning_rate": 3.80551638784277e-05,
|
| 442 |
+
"loss": 0.5332,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 7.17948717948718,
|
| 447 |
+
"grad_norm": 0.26633119583129883,
|
| 448 |
+
"learning_rate": 3.5261371521817244e-05,
|
| 449 |
+
"loss": 0.5239,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 7.293447293447294,
|
| 454 |
+
"grad_norm": 0.23685385286808014,
|
| 455 |
+
"learning_rate": 3.2551985985948616e-05,
|
| 456 |
+
"loss": 0.5309,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 7.407407407407407,
|
| 461 |
+
"grad_norm": 0.22292840480804443,
|
| 462 |
+
"learning_rate": 2.993053979266577e-05,
|
| 463 |
+
"loss": 0.5372,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 7.521367521367521,
|
| 468 |
+
"grad_norm": 0.2220107614994049,
|
| 469 |
+
"learning_rate": 2.7400450807686938e-05,
|
| 470 |
+
"loss": 0.5083,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 7.635327635327635,
|
| 475 |
+
"grad_norm": 0.2191537618637085,
|
| 476 |
+
"learning_rate": 2.496501778435977e-05,
|
| 477 |
+
"loss": 0.5164,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 7.749287749287749,
|
| 482 |
+
"grad_norm": 0.22593119740486145,
|
| 483 |
+
"learning_rate": 2.2627416062716366e-05,
|
| 484 |
+
"loss": 0.5152,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 7.863247863247864,
|
| 489 |
+
"grad_norm": 0.23532789945602417,
|
| 490 |
+
"learning_rate": 2.0390693429435627e-05,
|
| 491 |
+
"loss": 0.5269,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 7.977207977207978,
|
| 496 |
+
"grad_norm": 0.25111591815948486,
|
| 497 |
+
"learning_rate": 1.825776614411082e-05,
|
| 498 |
+
"loss": 0.5335,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 8.091168091168091,
|
| 503 |
+
"grad_norm": 0.21956747770309448,
|
| 504 |
+
"learning_rate": 1.6231415137003537e-05,
|
| 505 |
+
"loss": 0.5144,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 8.205128205128204,
|
| 510 |
+
"grad_norm": 0.23355403542518616,
|
| 511 |
+
"learning_rate": 1.4314282383241096e-05,
|
| 512 |
+
"loss": 0.5294,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 8.31908831908832,
|
| 517 |
+
"grad_norm": 0.23712006211280823,
|
| 518 |
+
"learning_rate": 1.2508867458185037e-05,
|
| 519 |
+
"loss": 0.5229,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 8.433048433048434,
|
| 524 |
+
"grad_norm": 0.22506175935268402,
|
| 525 |
+
"learning_rate": 1.0817524278461776e-05,
|
| 526 |
+
"loss": 0.5212,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 8.547008547008547,
|
| 531 |
+
"grad_norm": 0.21853385865688324,
|
| 532 |
+
"learning_rate": 9.242458032904311e-06,
|
| 533 |
+
"loss": 0.5193,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 8.660968660968662,
|
| 538 |
+
"grad_norm": 0.23257511854171753,
|
| 539 |
+
"learning_rate": 7.785722307406684e-06,
|
| 540 |
+
"loss": 0.5039,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 8.774928774928775,
|
| 545 |
+
"grad_norm": 0.21563945710659027,
|
| 546 |
+
"learning_rate": 6.4492164074399065e-06,
|
| 547 |
+
"loss": 0.5232,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 8.88888888888889,
|
| 552 |
+
"grad_norm": 0.22108329832553864,
|
| 553 |
+
"learning_rate": 5.2346828817197655e-06,
|
| 554 |
+
"loss": 0.5309,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 9.002849002849002,
|
| 559 |
+
"grad_norm": 0.22330021858215332,
|
| 560 |
+
"learning_rate": 4.143705250255869e-06,
|
| 561 |
+
"loss": 0.5287,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 9.116809116809117,
|
| 566 |
+
"grad_norm": 0.22394247353076935,
|
| 567 |
+
"learning_rate": 3.1777059397436692e-06,
|
| 568 |
+
"loss": 0.5007,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 9.23076923076923,
|
| 573 |
+
"grad_norm": 0.2144930511713028,
|
| 574 |
+
"learning_rate": 2.3379444289913342e-06,
|
| 575 |
+
"loss": 0.5277,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 9.344729344729345,
|
| 580 |
+
"grad_norm": 0.2214236557483673,
|
| 581 |
+
"learning_rate": 1.6255156067997323e-06,
|
| 582 |
+
"loss": 0.5173,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 9.45868945868946,
|
| 587 |
+
"grad_norm": 0.2192196100950241,
|
| 588 |
+
"learning_rate": 1.0413483444362771e-06,
|
| 589 |
+
"loss": 0.5123,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 9.572649572649572,
|
| 594 |
+
"grad_norm": 0.22837017476558685,
|
| 595 |
+
"learning_rate": 5.862042845640403e-07,
|
| 596 |
+
"loss": 0.5279,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 9.686609686609687,
|
| 601 |
+
"grad_norm": 0.21172335743904114,
|
| 602 |
+
"learning_rate": 2.606768482050215e-07,
|
| 603 |
+
"loss": 0.5263,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 9.8005698005698,
|
| 608 |
+
"grad_norm": 0.23530949652194977,
|
| 609 |
+
"learning_rate": 6.519046103230508e-08,
|
| 610 |
+
"loss": 0.5202,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 9.914529914529915,
|
| 615 |
+
"grad_norm": 0.23058444261550903,
|
| 616 |
+
"learning_rate": 0.0,
|
| 617 |
+
"loss": 0.5243,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 9.914529914529915,
|
| 622 |
+
"step": 870,
|
| 623 |
+
"total_flos": 5.67984355540992e+16,
|
| 624 |
+
"train_loss": 0.5655439464525245,
|
| 625 |
+
"train_runtime": 2716.1071,
|
| 626 |
+
"train_samples_per_second": 1.292,
|
| 627 |
+
"train_steps_per_second": 0.32
|
| 628 |
+
}
|
| 629 |
+
],
|
| 630 |
+
"logging_steps": 10,
|
| 631 |
+
"max_steps": 870,
|
| 632 |
+
"num_input_tokens_seen": 0,
|
| 633 |
+
"num_train_epochs": 10,
|
| 634 |
+
"save_steps": 500,
|
| 635 |
+
"total_flos": 5.67984355540992e+16,
|
| 636 |
+
"train_batch_size": 1,
|
| 637 |
+
"trial_name": null,
|
| 638 |
+
"trial_params": null
|
| 639 |
+
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:397e037153240ddceac73ed48aa51e61256700ed0043d6463824f5d88fa3beb8
|
| 3 |
size 4984
|