update slot resume pointers for step 1321
Browse files
slots/7/checkpoint-760/config.json
DELETED
|
@@ -1,24 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"architectures": [
|
| 3 |
-
"TwinyForCausalLM"
|
| 4 |
-
],
|
| 5 |
-
"attention_dropout": 0.0,
|
| 6 |
-
"dtype": "float32",
|
| 7 |
-
"hidden_dropout": 0.0,
|
| 8 |
-
"hidden_size": 768,
|
| 9 |
-
"initializer_range": 0.02,
|
| 10 |
-
"intermediate_size": 3072,
|
| 11 |
-
"max_position_embeddings": 128,
|
| 12 |
-
"model_type": "twiny",
|
| 13 |
-
"neftune_alpha": 0.0,
|
| 14 |
-
"num_attention_heads": 12,
|
| 15 |
-
"num_hidden_layers": 3,
|
| 16 |
-
"num_key_value_heads": 3,
|
| 17 |
-
"qk_norm": true,
|
| 18 |
-
"rezero_init": 1.0,
|
| 19 |
-
"rms_norm_eps": 1e-06,
|
| 20 |
-
"rope_theta": 10000.0,
|
| 21 |
-
"transformers_version": "5.0.0",
|
| 22 |
-
"use_cache": false,
|
| 23 |
-
"vocab_size": 32000
|
| 24 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
slots/7/checkpoint-760/trainer_state.json
DELETED
|
@@ -1,307 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": null,
|
| 3 |
-
"best_metric": null,
|
| 4 |
-
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.10160682503739231,
|
| 6 |
-
"eval_steps": 500,
|
| 7 |
-
"global_step": 760,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.0001336931908386741,
|
| 14 |
-
"grad_norm": Infinity,
|
| 15 |
-
"learning_rate": 5e-05,
|
| 16 |
-
"loss": 150.77503967285156,
|
| 17 |
-
"step": 1
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.002673863816773482,
|
| 21 |
-
"grad_norm": 65.5747299194336,
|
| 22 |
-
"learning_rate": 4.999995392022967e-05,
|
| 23 |
-
"loss": 65.57097424958882,
|
| 24 |
-
"step": 20
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.005347727633546964,
|
| 28 |
-
"grad_norm": 29.183452606201172,
|
| 29 |
-
"learning_rate": 4.999976672145381e-05,
|
| 30 |
-
"loss": 12.781006622314454,
|
| 31 |
-
"step": 40
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.008021591450320446,
|
| 35 |
-
"grad_norm": 22.14234161376953,
|
| 36 |
-
"learning_rate": 4.999943552476422e-05,
|
| 37 |
-
"loss": 7.058085632324219,
|
| 38 |
-
"step": 60
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.010695455267093928,
|
| 42 |
-
"grad_norm": 25.116792678833008,
|
| 43 |
-
"learning_rate": 4.999896033206858e-05,
|
| 44 |
-
"loss": 5.17547492980957,
|
| 45 |
-
"step": 80
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.01336931908386741,
|
| 49 |
-
"grad_norm": 17.95206642150879,
|
| 50 |
-
"learning_rate": 4.999834114610398e-05,
|
| 51 |
-
"loss": 3.6641620635986327,
|
| 52 |
-
"step": 100
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.01604318290064089,
|
| 56 |
-
"grad_norm": 13.571253776550293,
|
| 57 |
-
"learning_rate": 4.999757797043691e-05,
|
| 58 |
-
"loss": 2.9684438705444336,
|
| 59 |
-
"step": 120
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.018717046717414372,
|
| 63 |
-
"grad_norm": 11.740418434143066,
|
| 64 |
-
"learning_rate": 4.999667080946324e-05,
|
| 65 |
-
"loss": 2.658764457702637,
|
| 66 |
-
"step": 140
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.021390910534187856,
|
| 70 |
-
"grad_norm": 8.911559104919434,
|
| 71 |
-
"learning_rate": 4.999561966840821e-05,
|
| 72 |
-
"loss": 2.0916315078735352,
|
| 73 |
-
"step": 160
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 0.024064774350961337,
|
| 77 |
-
"grad_norm": 7.613410472869873,
|
| 78 |
-
"learning_rate": 4.9994424553326335e-05,
|
| 79 |
-
"loss": 1.730122184753418,
|
| 80 |
-
"step": 180
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 0.02673863816773482,
|
| 84 |
-
"grad_norm": 10.990550994873047,
|
| 85 |
-
"learning_rate": 4.999308547110146e-05,
|
| 86 |
-
"loss": 1.689904022216797,
|
| 87 |
-
"step": 200
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.029412501984508302,
|
| 91 |
-
"grad_norm": 13.414973258972168,
|
| 92 |
-
"learning_rate": 4.999160242944665e-05,
|
| 93 |
-
"loss": 1.331161880493164,
|
| 94 |
-
"step": 220
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 0.03208636580128178,
|
| 98 |
-
"grad_norm": 9.434646606445312,
|
| 99 |
-
"learning_rate": 4.998997543690418e-05,
|
| 100 |
-
"loss": 1.285347843170166,
|
| 101 |
-
"step": 240
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 0.03476022961805526,
|
| 105 |
-
"grad_norm": 15.048566818237305,
|
| 106 |
-
"learning_rate": 4.998820450284549e-05,
|
| 107 |
-
"loss": 1.1024582862854004,
|
| 108 |
-
"step": 260
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 0.037434093434828744,
|
| 112 |
-
"grad_norm": 7.001116752624512,
|
| 113 |
-
"learning_rate": 4.99862896374711e-05,
|
| 114 |
-
"loss": 1.106878662109375,
|
| 115 |
-
"step": 280
|
| 116 |
-
},
|
| 117 |
-
{
|
| 118 |
-
"epoch": 0.04010795725160223,
|
| 119 |
-
"grad_norm": 10.976016998291016,
|
| 120 |
-
"learning_rate": 4.998423085181056e-05,
|
| 121 |
-
"loss": 1.0516901969909669,
|
| 122 |
-
"step": 300
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.04278182106837571,
|
| 126 |
-
"grad_norm": 10.106513977050781,
|
| 127 |
-
"learning_rate": 4.998202815772245e-05,
|
| 128 |
-
"loss": 0.8515177726745605,
|
| 129 |
-
"step": 320
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.04545568488514919,
|
| 133 |
-
"grad_norm": 3.9024946689605713,
|
| 134 |
-
"learning_rate": 4.9979681567894195e-05,
|
| 135 |
-
"loss": 0.851468849182129,
|
| 136 |
-
"step": 340
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.048129548701922674,
|
| 140 |
-
"grad_norm": 7.649858474731445,
|
| 141 |
-
"learning_rate": 4.997719109584209e-05,
|
| 142 |
-
"loss": 0.8077478408813477,
|
| 143 |
-
"step": 360
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.050803412518696155,
|
| 147 |
-
"grad_norm": 11.689629554748535,
|
| 148 |
-
"learning_rate": 4.997455675591119e-05,
|
| 149 |
-
"loss": 0.6951910495758057,
|
| 150 |
-
"step": 380
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.05347727633546964,
|
| 154 |
-
"grad_norm": 12.950127601623535,
|
| 155 |
-
"learning_rate": 4.9971778563275204e-05,
|
| 156 |
-
"loss": 0.6042420387268066,
|
| 157 |
-
"step": 400
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.05615114015224312,
|
| 161 |
-
"grad_norm": 6.282811641693115,
|
| 162 |
-
"learning_rate": 4.9968856533936436e-05,
|
| 163 |
-
"loss": 0.6595364093780518,
|
| 164 |
-
"step": 420
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.058825003969016604,
|
| 168 |
-
"grad_norm": 3.553151845932007,
|
| 169 |
-
"learning_rate": 4.99657906847257e-05,
|
| 170 |
-
"loss": 0.71713285446167,
|
| 171 |
-
"step": 440
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.061498867785790085,
|
| 175 |
-
"grad_norm": 5.715928554534912,
|
| 176 |
-
"learning_rate": 4.996258103330218e-05,
|
| 177 |
-
"loss": 0.5080134391784668,
|
| 178 |
-
"step": 460
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.06417273160256357,
|
| 182 |
-
"grad_norm": 8.924579620361328,
|
| 183 |
-
"learning_rate": 4.995922759815339e-05,
|
| 184 |
-
"loss": 0.4877929210662842,
|
| 185 |
-
"step": 480
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.06684659541933705,
|
| 189 |
-
"grad_norm": 4.2473649978637695,
|
| 190 |
-
"learning_rate": 4.995573039859501e-05,
|
| 191 |
-
"loss": 0.5352929592132568,
|
| 192 |
-
"step": 500
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.06952045923611053,
|
| 196 |
-
"grad_norm": 5.880599021911621,
|
| 197 |
-
"learning_rate": 4.995208945477081e-05,
|
| 198 |
-
"loss": 0.520254373550415,
|
| 199 |
-
"step": 520
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.07219432305288401,
|
| 203 |
-
"grad_norm": 8.42399787902832,
|
| 204 |
-
"learning_rate": 4.994830478765251e-05,
|
| 205 |
-
"loss": 0.5289288520812988,
|
| 206 |
-
"step": 540
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.07486818686965749,
|
| 210 |
-
"grad_norm": 5.284440040588379,
|
| 211 |
-
"learning_rate": 4.9944376419039684e-05,
|
| 212 |
-
"loss": 0.4858880043029785,
|
| 213 |
-
"step": 560
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.07754205068643098,
|
| 217 |
-
"grad_norm": 4.662903308868408,
|
| 218 |
-
"learning_rate": 4.994030437155961e-05,
|
| 219 |
-
"loss": 0.4253716468811035,
|
| 220 |
-
"step": 580
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.08021591450320446,
|
| 224 |
-
"grad_norm": 2.9051289558410645,
|
| 225 |
-
"learning_rate": 4.993608866866718e-05,
|
| 226 |
-
"loss": 0.32270302772521975,
|
| 227 |
-
"step": 600
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.08288977831997794,
|
| 231 |
-
"grad_norm": 2.804189920425415,
|
| 232 |
-
"learning_rate": 4.993172933464471e-05,
|
| 233 |
-
"loss": 0.329010009765625,
|
| 234 |
-
"step": 620
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 0.08556364213675142,
|
| 238 |
-
"grad_norm": 1.455569863319397,
|
| 239 |
-
"learning_rate": 4.9927226394601815e-05,
|
| 240 |
-
"loss": 0.35200178623199463,
|
| 241 |
-
"step": 640
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 0.0882375059535249,
|
| 245 |
-
"grad_norm": 11.18322467803955,
|
| 246 |
-
"learning_rate": 4.992257987447532e-05,
|
| 247 |
-
"loss": 0.30103607177734376,
|
| 248 |
-
"step": 660
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 0.09091136977029839,
|
| 252 |
-
"grad_norm": 6.4625773429870605,
|
| 253 |
-
"learning_rate": 4.991778980102904e-05,
|
| 254 |
-
"loss": 0.3078582525253296,
|
| 255 |
-
"step": 680
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.09358523358707187,
|
| 259 |
-
"grad_norm": 11.98585033416748,
|
| 260 |
-
"learning_rate": 4.9912856201853644e-05,
|
| 261 |
-
"loss": 0.30774364471435545,
|
| 262 |
-
"step": 700
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 0.09625909740384535,
|
| 266 |
-
"grad_norm": 10.583540916442871,
|
| 267 |
-
"learning_rate": 4.990777910536653e-05,
|
| 268 |
-
"loss": 0.30326850414276124,
|
| 269 |
-
"step": 720
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.09893296122061883,
|
| 273 |
-
"grad_norm": 6.04772424697876,
|
| 274 |
-
"learning_rate": 4.990255854081161e-05,
|
| 275 |
-
"loss": 0.37041378021240234,
|
| 276 |
-
"step": 740
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.10160682503739231,
|
| 280 |
-
"grad_norm": 7.155930519104004,
|
| 281 |
-
"learning_rate": 4.989719453825918e-05,
|
| 282 |
-
"loss": 0.33830211162567136,
|
| 283 |
-
"step": 760
|
| 284 |
-
}
|
| 285 |
-
],
|
| 286 |
-
"logging_steps": 20,
|
| 287 |
-
"max_steps": 26180,
|
| 288 |
-
"num_input_tokens_seen": 0,
|
| 289 |
-
"num_train_epochs": 4,
|
| 290 |
-
"save_steps": 1000000000,
|
| 291 |
-
"stateful_callbacks": {
|
| 292 |
-
"TrainerControl": {
|
| 293 |
-
"args": {
|
| 294 |
-
"should_epoch_stop": false,
|
| 295 |
-
"should_evaluate": false,
|
| 296 |
-
"should_log": false,
|
| 297 |
-
"should_save": true,
|
| 298 |
-
"should_training_stop": false
|
| 299 |
-
},
|
| 300 |
-
"attributes": {}
|
| 301 |
-
}
|
| 302 |
-
},
|
| 303 |
-
"total_flos": 2914835592314880.0,
|
| 304 |
-
"train_batch_size": 1,
|
| 305 |
-
"trial_name": null,
|
| 306 |
-
"trial_params": null
|
| 307 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
slots/7/latest.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"worker_id": "slot:7", "checkpoint": "slots/7/checkpoint-
|
|
|
|
| 1 |
+
{"worker_id": "slot:7", "checkpoint": "slots/7/checkpoint-1321", "step": 1321, "updated_at": 1776682367}
|