Training in progress, step 400, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 723674912
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8c3b20dc5667964ffe2baf981e662e4a77f20dd3ac7eecffcec9fa4d6eb9d179
|
| 3 |
size 723674912
|
last-checkpoint/optimizer.pt
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 sha256:75f80a3817484e177dce1b009d021deef68bbea27b4c0296f59745f17eb4d67c
|
| 3 |
+
size 735625626
|
last-checkpoint/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9feae33b2fec0a6229240e7adaee6ecc8f5cfdf1a8bd0e827b1d8a241424e3c0
|
| 3 |
+
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a673aaf85c0fe6b6c29cb8f3e7dbd829eef637110e4ad9a775f3fcf001c92591
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,71 +1,323 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
-
"eval_steps":
|
| 6 |
-
"global_step":
|
| 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.
|
| 13 |
-
"eval_loss": 3.
|
| 14 |
-
"eval_runtime":
|
| 15 |
-
"eval_samples_per_second":
|
| 16 |
-
"eval_steps_per_second":
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
-
"epoch": 0.
|
| 21 |
-
"grad_norm":
|
| 22 |
"learning_rate": 1.6000000000000003e-05,
|
| 23 |
-
"loss": 2.
|
| 24 |
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"epoch": 0.
|
| 28 |
-
"grad_norm":
|
| 29 |
"learning_rate": 3.2000000000000005e-05,
|
| 30 |
-
"loss": 2.
|
| 31 |
-
"step": 20
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.023668639053254437,
|
| 35 |
-
"eval_loss": 3.0602526664733887,
|
| 36 |
-
"eval_runtime": 5.3059,
|
| 37 |
-
"eval_samples_per_second": 283.082,
|
| 38 |
-
"eval_steps_per_second": 17.716,
|
| 39 |
"step": 20
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"grad_norm":
|
| 44 |
"learning_rate": 4.8e-05,
|
| 45 |
-
"loss":
|
| 46 |
"step": 30
|
| 47 |
},
|
| 48 |
{
|
| 49 |
-
"epoch": 0.
|
| 50 |
-
"grad_norm":
|
| 51 |
"learning_rate": 6.400000000000001e-05,
|
| 52 |
-
"loss": 2.
|
| 53 |
"step": 40
|
| 54 |
},
|
| 55 |
{
|
| 56 |
-
"epoch": 0.
|
| 57 |
-
"
|
| 58 |
-
"
|
| 59 |
-
"
|
| 60 |
-
"
|
| 61 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 62 |
}
|
| 63 |
],
|
| 64 |
"logging_steps": 10,
|
| 65 |
"max_steps": 2500,
|
| 66 |
"num_input_tokens_seen": 0,
|
| 67 |
-
"num_train_epochs":
|
| 68 |
-
"save_steps":
|
| 69 |
"stateful_callbacks": {
|
| 70 |
"TrainerControl": {
|
| 71 |
"args": {
|
|
@@ -78,8 +330,8 @@
|
|
| 78 |
"attributes": {}
|
| 79 |
}
|
| 80 |
},
|
| 81 |
-
"total_flos":
|
| 82 |
-
"train_batch_size":
|
| 83 |
"trial_name": null,
|
| 84 |
"trial_params": null
|
| 85 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.11837821840781296,
|
| 5 |
+
"eval_steps": 200,
|
| 6 |
+
"global_step": 400,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.0002959455460195324,
|
| 13 |
+
"eval_loss": 3.0371296405792236,
|
| 14 |
+
"eval_runtime": 40.2878,
|
| 15 |
+
"eval_samples_per_second": 37.282,
|
| 16 |
+
"eval_steps_per_second": 9.333,
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.002959455460195324,
|
| 21 |
+
"grad_norm": 2.84375,
|
| 22 |
"learning_rate": 1.6000000000000003e-05,
|
| 23 |
+
"loss": 2.5732,
|
| 24 |
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.005918910920390648,
|
| 28 |
+
"grad_norm": 4.40625,
|
| 29 |
"learning_rate": 3.2000000000000005e-05,
|
| 30 |
+
"loss": 2.6895,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 31 |
"step": 20
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.008878366380585973,
|
| 35 |
+
"grad_norm": 5.375,
|
| 36 |
"learning_rate": 4.8e-05,
|
| 37 |
+
"loss": 2.5336,
|
| 38 |
"step": 30
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.011837821840781295,
|
| 42 |
+
"grad_norm": 7.90625,
|
| 43 |
"learning_rate": 6.400000000000001e-05,
|
| 44 |
+
"loss": 2.9661,
|
| 45 |
"step": 40
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.01479727730097662,
|
| 49 |
+
"grad_norm": 25.375,
|
| 50 |
+
"learning_rate": 8e-05,
|
| 51 |
+
"loss": 3.4081,
|
| 52 |
+
"step": 50
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.017756732761171946,
|
| 56 |
+
"grad_norm": 3.171875,
|
| 57 |
+
"learning_rate": 9.6e-05,
|
| 58 |
+
"loss": 2.3133,
|
| 59 |
+
"step": 60
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.020716188221367268,
|
| 63 |
+
"grad_norm": 2.953125,
|
| 64 |
+
"learning_rate": 0.00011200000000000001,
|
| 65 |
+
"loss": 2.3044,
|
| 66 |
+
"step": 70
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.02367564368156259,
|
| 70 |
+
"grad_norm": 4.34375,
|
| 71 |
+
"learning_rate": 0.00012800000000000002,
|
| 72 |
+
"loss": 2.411,
|
| 73 |
+
"step": 80
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.026635099141757917,
|
| 77 |
+
"grad_norm": 5.5,
|
| 78 |
+
"learning_rate": 0.000144,
|
| 79 |
+
"loss": 2.326,
|
| 80 |
+
"step": 90
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.02959455460195324,
|
| 84 |
+
"grad_norm": 22.375,
|
| 85 |
+
"learning_rate": 0.00016,
|
| 86 |
+
"loss": 3.3354,
|
| 87 |
+
"step": 100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.032554010062148565,
|
| 91 |
+
"grad_norm": 3.875,
|
| 92 |
+
"learning_rate": 0.00017600000000000002,
|
| 93 |
+
"loss": 2.4824,
|
| 94 |
+
"step": 110
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.03551346552234389,
|
| 98 |
+
"grad_norm": 2.90625,
|
| 99 |
+
"learning_rate": 0.000192,
|
| 100 |
+
"loss": 2.2624,
|
| 101 |
+
"step": 120
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.03847292098253921,
|
| 105 |
+
"grad_norm": 2.890625,
|
| 106 |
+
"learning_rate": 0.0001999978128380225,
|
| 107 |
+
"loss": 2.3521,
|
| 108 |
+
"step": 130
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.041432376442734536,
|
| 112 |
+
"grad_norm": 7.09375,
|
| 113 |
+
"learning_rate": 0.0001999803161162393,
|
| 114 |
+
"loss": 2.3626,
|
| 115 |
+
"step": 140
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.04439183190292986,
|
| 119 |
+
"grad_norm": 31.625,
|
| 120 |
+
"learning_rate": 0.00019994532573409262,
|
| 121 |
+
"loss": 2.2468,
|
| 122 |
+
"step": 150
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.04735128736312518,
|
| 126 |
+
"grad_norm": 5.6875,
|
| 127 |
+
"learning_rate": 0.00019989284781388617,
|
| 128 |
+
"loss": 2.3798,
|
| 129 |
+
"step": 160
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.05031074282332051,
|
| 133 |
+
"grad_norm": 2.578125,
|
| 134 |
+
"learning_rate": 0.00019982289153773646,
|
| 135 |
+
"loss": 2.3013,
|
| 136 |
+
"step": 170
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.053270198283515834,
|
| 140 |
+
"grad_norm": 4.375,
|
| 141 |
+
"learning_rate": 0.00019973546914596623,
|
| 142 |
+
"loss": 2.4314,
|
| 143 |
+
"step": 180
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.05622965374371116,
|
| 147 |
+
"grad_norm": 6.90625,
|
| 148 |
+
"learning_rate": 0.00019963059593496268,
|
| 149 |
+
"loss": 2.225,
|
| 150 |
+
"step": 190
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.05918910920390648,
|
| 154 |
+
"grad_norm": 25.75,
|
| 155 |
+
"learning_rate": 0.00019950829025450114,
|
| 156 |
+
"loss": 2.5078,
|
| 157 |
+
"step": 200
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.05918910920390648,
|
| 161 |
+
"eval_loss": 2.297968864440918,
|
| 162 |
+
"eval_runtime": 38.145,
|
| 163 |
+
"eval_samples_per_second": 39.376,
|
| 164 |
+
"eval_steps_per_second": 9.857,
|
| 165 |
+
"step": 200
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.062148564664101805,
|
| 169 |
+
"grad_norm": 2.5625,
|
| 170 |
+
"learning_rate": 0.0001993685735045343,
|
| 171 |
+
"loss": 2.4115,
|
| 172 |
+
"step": 210
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.06510802012429713,
|
| 176 |
+
"grad_norm": 2.890625,
|
| 177 |
+
"learning_rate": 0.0001992114701314478,
|
| 178 |
+
"loss": 2.3576,
|
| 179 |
+
"step": 220
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.06806747558449246,
|
| 183 |
+
"grad_norm": 3.03125,
|
| 184 |
+
"learning_rate": 0.000199037007623783,
|
| 185 |
+
"loss": 2.2509,
|
| 186 |
+
"step": 230
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.07102693104468778,
|
| 190 |
+
"grad_norm": 6.65625,
|
| 191 |
+
"learning_rate": 0.00019884521650742715,
|
| 192 |
+
"loss": 2.2513,
|
| 193 |
+
"step": 240
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.0739863865048831,
|
| 197 |
+
"grad_norm": 39.25,
|
| 198 |
+
"learning_rate": 0.00019863613034027224,
|
| 199 |
+
"loss": 2.7071,
|
| 200 |
+
"step": 250
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.07694584196507842,
|
| 204 |
+
"grad_norm": 5.125,
|
| 205 |
+
"learning_rate": 0.0001984097857063434,
|
| 206 |
+
"loss": 2.1904,
|
| 207 |
+
"step": 260
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.07990529742527375,
|
| 211 |
+
"grad_norm": 2.734375,
|
| 212 |
+
"learning_rate": 0.0001981662222093976,
|
| 213 |
+
"loss": 2.4093,
|
| 214 |
+
"step": 270
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.08286475288546907,
|
| 218 |
+
"grad_norm": 3.140625,
|
| 219 |
+
"learning_rate": 0.00019790548246599447,
|
| 220 |
+
"loss": 2.2674,
|
| 221 |
+
"step": 280
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.0858242083456644,
|
| 225 |
+
"grad_norm": 5.34375,
|
| 226 |
+
"learning_rate": 0.00019762761209803927,
|
| 227 |
+
"loss": 2.052,
|
| 228 |
+
"step": 290
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.08878366380585972,
|
| 232 |
+
"grad_norm": 15.5,
|
| 233 |
+
"learning_rate": 0.0001973326597248006,
|
| 234 |
+
"loss": 2.2147,
|
| 235 |
+
"step": 300
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.09174311926605505,
|
| 239 |
+
"grad_norm": 3.234375,
|
| 240 |
+
"learning_rate": 0.00019702067695440332,
|
| 241 |
+
"loss": 2.4276,
|
| 242 |
+
"step": 310
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.09470257472625036,
|
| 246 |
+
"grad_norm": 2.875,
|
| 247 |
+
"learning_rate": 0.00019669171837479873,
|
| 248 |
+
"loss": 2.0212,
|
| 249 |
+
"step": 320
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.09766203018644569,
|
| 253 |
+
"grad_norm": 3.0,
|
| 254 |
+
"learning_rate": 0.00019634584154421317,
|
| 255 |
+
"loss": 2.2913,
|
| 256 |
+
"step": 330
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.10062148564664102,
|
| 260 |
+
"grad_norm": 5.53125,
|
| 261 |
+
"learning_rate": 0.00019598310698107702,
|
| 262 |
+
"loss": 2.1126,
|
| 263 |
+
"step": 340
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.10358094110683634,
|
| 267 |
+
"grad_norm": 23.375,
|
| 268 |
+
"learning_rate": 0.00019560357815343577,
|
| 269 |
+
"loss": 1.9027,
|
| 270 |
+
"step": 350
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.10654039656703167,
|
| 274 |
+
"grad_norm": 2.515625,
|
| 275 |
+
"learning_rate": 0.00019520732146784491,
|
| 276 |
+
"loss": 2.2814,
|
| 277 |
+
"step": 360
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.109499852027227,
|
| 281 |
+
"grad_norm": 3.09375,
|
| 282 |
+
"learning_rate": 0.0001947944062577507,
|
| 283 |
+
"loss": 2.0483,
|
| 284 |
+
"step": 370
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.11245930748742232,
|
| 288 |
+
"grad_norm": 4.125,
|
| 289 |
+
"learning_rate": 0.00019436490477135878,
|
| 290 |
+
"loss": 2.2125,
|
| 291 |
+
"step": 380
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.11541876294761765,
|
| 295 |
+
"grad_norm": 5.53125,
|
| 296 |
+
"learning_rate": 0.00019391889215899299,
|
| 297 |
+
"loss": 1.9433,
|
| 298 |
+
"step": 390
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 0.11837821840781296,
|
| 302 |
+
"grad_norm": 25.375,
|
| 303 |
+
"learning_rate": 0.0001934564464599461,
|
| 304 |
+
"loss": 1.858,
|
| 305 |
+
"step": 400
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.11837821840781296,
|
| 309 |
+
"eval_loss": 2.2061269283294678,
|
| 310 |
+
"eval_runtime": 37.9861,
|
| 311 |
+
"eval_samples_per_second": 39.541,
|
| 312 |
+
"eval_steps_per_second": 9.898,
|
| 313 |
+
"step": 400
|
| 314 |
}
|
| 315 |
],
|
| 316 |
"logging_steps": 10,
|
| 317 |
"max_steps": 2500,
|
| 318 |
"num_input_tokens_seen": 0,
|
| 319 |
+
"num_train_epochs": 1,
|
| 320 |
+
"save_steps": 400,
|
| 321 |
"stateful_callbacks": {
|
| 322 |
"TrainerControl": {
|
| 323 |
"args": {
|
|
|
|
| 330 |
"attributes": {}
|
| 331 |
}
|
| 332 |
},
|
| 333 |
+
"total_flos": 6216909638860800.0,
|
| 334 |
+
"train_batch_size": 4,
|
| 335 |
"trial_name": null,
|
| 336 |
"trial_params": null
|
| 337 |
}
|
last-checkpoint/training_args.bin
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 sha256:8728015aed6a4d50cfee4eda0cc3ff69e5a4901af146b3fd10ebc2972270b3a8
|
| 3 |
+
size 6840
|