Training in progress, epoch 5, checkpoint
Browse files
checkpoint-500/config.json
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "microsoft/deberta-large",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"DebertaForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.1,
|
| 9 |
+
"hidden_size": 1024,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 4096,
|
| 12 |
+
"layer_norm_eps": 1e-07,
|
| 13 |
+
"legacy": true,
|
| 14 |
+
"max_position_embeddings": 512,
|
| 15 |
+
"max_relative_positions": -1,
|
| 16 |
+
"model_type": "deberta",
|
| 17 |
+
"num_attention_heads": 16,
|
| 18 |
+
"num_hidden_layers": 24,
|
| 19 |
+
"pad_token_id": 0,
|
| 20 |
+
"pooler_dropout": 0,
|
| 21 |
+
"pooler_hidden_act": "gelu",
|
| 22 |
+
"pooler_hidden_size": 1024,
|
| 23 |
+
"pos_att_type": [
|
| 24 |
+
"c2p",
|
| 25 |
+
"p2c"
|
| 26 |
+
],
|
| 27 |
+
"position_biased_input": false,
|
| 28 |
+
"relative_attention": true,
|
| 29 |
+
"torch_dtype": "float32",
|
| 30 |
+
"transformers_version": "4.48.3",
|
| 31 |
+
"type_vocab_size": 0,
|
| 32 |
+
"vocab_size": 50265
|
| 33 |
+
}
|
checkpoint-500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:25ec88b0210395a17c4fb3f83603411da49b1110ac6aec88a7cb415ddcaa3eee
|
| 3 |
+
size 1624907048
|
checkpoint-500/trainer_state.json
ADDED
|
@@ -0,0 +1,263 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 50,
|
| 6 |
+
"global_step": 500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.5,
|
| 13 |
+
"grad_norm": 9.394545555114746,
|
| 14 |
+
"learning_rate": 2.75e-05,
|
| 15 |
+
"loss": 0.52,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.5,
|
| 20 |
+
"eval_pet_AUC": 0.9145265189137036,
|
| 21 |
+
"eval_pet_accuracy": 0.845,
|
| 22 |
+
"eval_pet_f1": 0.7703703703703704,
|
| 23 |
+
"eval_pet_loss": 0.3643406927585602,
|
| 24 |
+
"eval_pet_macro_f1": 0.8266946191474493,
|
| 25 |
+
"eval_pet_micro_f1": 0.845,
|
| 26 |
+
"eval_pet_precision": 0.7482014388489209,
|
| 27 |
+
"eval_pet_recall": 0.7938931297709924,
|
| 28 |
+
"eval_pet_runtime": 0.9635,
|
| 29 |
+
"eval_pet_samples_per_second": 415.172,
|
| 30 |
+
"eval_pet_specificity": 0.8698884758364313,
|
| 31 |
+
"eval_pet_steps_per_second": 1.038,
|
| 32 |
+
"step": 50
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 1.0,
|
| 36 |
+
"grad_norm": 6.517026901245117,
|
| 37 |
+
"learning_rate": 2.5e-05,
|
| 38 |
+
"loss": 0.4033,
|
| 39 |
+
"step": 100
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.0,
|
| 43 |
+
"eval_pet_AUC": 0.9307017792786401,
|
| 44 |
+
"eval_pet_accuracy": 0.85,
|
| 45 |
+
"eval_pet_f1": 0.7744360902255639,
|
| 46 |
+
"eval_pet_loss": 0.32512709498405457,
|
| 47 |
+
"eval_pet_macro_f1": 0.831038269831883,
|
| 48 |
+
"eval_pet_micro_f1": 0.85,
|
| 49 |
+
"eval_pet_precision": 0.762962962962963,
|
| 50 |
+
"eval_pet_recall": 0.7862595419847328,
|
| 51 |
+
"eval_pet_runtime": 0.9568,
|
| 52 |
+
"eval_pet_samples_per_second": 418.039,
|
| 53 |
+
"eval_pet_specificity": 0.8810408921933085,
|
| 54 |
+
"eval_pet_steps_per_second": 1.045,
|
| 55 |
+
"step": 100
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 1.5,
|
| 59 |
+
"grad_norm": 4.107645034790039,
|
| 60 |
+
"learning_rate": 2.25e-05,
|
| 61 |
+
"loss": 0.2658,
|
| 62 |
+
"step": 150
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.5,
|
| 66 |
+
"eval_pet_AUC": 0.945259513607083,
|
| 67 |
+
"eval_pet_accuracy": 0.855,
|
| 68 |
+
"eval_pet_f1": 0.7943262411347518,
|
| 69 |
+
"eval_pet_loss": 0.375436931848526,
|
| 70 |
+
"eval_pet_macro_f1": 0.8411785645828199,
|
| 71 |
+
"eval_pet_micro_f1": 0.855,
|
| 72 |
+
"eval_pet_precision": 0.7417218543046358,
|
| 73 |
+
"eval_pet_recall": 0.8549618320610687,
|
| 74 |
+
"eval_pet_runtime": 0.9663,
|
| 75 |
+
"eval_pet_samples_per_second": 413.932,
|
| 76 |
+
"eval_pet_specificity": 0.8550185873605948,
|
| 77 |
+
"eval_pet_steps_per_second": 1.035,
|
| 78 |
+
"step": 150
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 2.0,
|
| 82 |
+
"grad_norm": 9.23839282989502,
|
| 83 |
+
"learning_rate": 1.9999999999999998e-05,
|
| 84 |
+
"loss": 0.2444,
|
| 85 |
+
"step": 200
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 2.0,
|
| 89 |
+
"eval_pet_AUC": 0.9240614092340873,
|
| 90 |
+
"eval_pet_accuracy": 0.845,
|
| 91 |
+
"eval_pet_f1": 0.7785714285714286,
|
| 92 |
+
"eval_pet_loss": 0.38033241033554077,
|
| 93 |
+
"eval_pet_macro_f1": 0.8296703296703296,
|
| 94 |
+
"eval_pet_micro_f1": 0.845,
|
| 95 |
+
"eval_pet_precision": 0.7315436241610739,
|
| 96 |
+
"eval_pet_recall": 0.8320610687022901,
|
| 97 |
+
"eval_pet_runtime": 0.9687,
|
| 98 |
+
"eval_pet_samples_per_second": 412.922,
|
| 99 |
+
"eval_pet_specificity": 0.8513011152416357,
|
| 100 |
+
"eval_pet_steps_per_second": 1.032,
|
| 101 |
+
"step": 200
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 2.5,
|
| 105 |
+
"grad_norm": 15.208889961242676,
|
| 106 |
+
"learning_rate": 1.7500000000000002e-05,
|
| 107 |
+
"loss": 0.1705,
|
| 108 |
+
"step": 250
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.5,
|
| 112 |
+
"eval_pet_AUC": 0.9323760606146599,
|
| 113 |
+
"eval_pet_accuracy": 0.8625,
|
| 114 |
+
"eval_pet_f1": 0.7639484978540773,
|
| 115 |
+
"eval_pet_loss": 0.6355132460594177,
|
| 116 |
+
"eval_pet_macro_f1": 0.8334733670928234,
|
| 117 |
+
"eval_pet_micro_f1": 0.8625,
|
| 118 |
+
"eval_pet_precision": 0.8725490196078431,
|
| 119 |
+
"eval_pet_recall": 0.6793893129770993,
|
| 120 |
+
"eval_pet_runtime": 0.9639,
|
| 121 |
+
"eval_pet_samples_per_second": 415.002,
|
| 122 |
+
"eval_pet_specificity": 0.9516728624535316,
|
| 123 |
+
"eval_pet_steps_per_second": 1.038,
|
| 124 |
+
"step": 250
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 3.0,
|
| 128 |
+
"grad_norm": 9.768027305603027,
|
| 129 |
+
"learning_rate": 1.5e-05,
|
| 130 |
+
"loss": 0.1972,
|
| 131 |
+
"step": 300
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 3.0,
|
| 135 |
+
"eval_pet_AUC": 0.9315531087715316,
|
| 136 |
+
"eval_pet_accuracy": 0.8525,
|
| 137 |
+
"eval_pet_f1": 0.7773584905660378,
|
| 138 |
+
"eval_pet_loss": 0.43593117594718933,
|
| 139 |
+
"eval_pet_macro_f1": 0.833539058367131,
|
| 140 |
+
"eval_pet_micro_f1": 0.8525,
|
| 141 |
+
"eval_pet_precision": 0.7686567164179104,
|
| 142 |
+
"eval_pet_recall": 0.7862595419847328,
|
| 143 |
+
"eval_pet_runtime": 0.9547,
|
| 144 |
+
"eval_pet_samples_per_second": 418.986,
|
| 145 |
+
"eval_pet_specificity": 0.8847583643122676,
|
| 146 |
+
"eval_pet_steps_per_second": 1.047,
|
| 147 |
+
"step": 300
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 3.5,
|
| 151 |
+
"grad_norm": 0.043738868087530136,
|
| 152 |
+
"learning_rate": 1.25e-05,
|
| 153 |
+
"loss": 0.0835,
|
| 154 |
+
"step": 350
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 3.5,
|
| 158 |
+
"eval_pet_AUC": 0.9468486619938137,
|
| 159 |
+
"eval_pet_accuracy": 0.8775,
|
| 160 |
+
"eval_pet_f1": 0.7949790794979079,
|
| 161 |
+
"eval_pet_loss": 0.6166596412658691,
|
| 162 |
+
"eval_pet_macro_f1": 0.8538175254887044,
|
| 163 |
+
"eval_pet_micro_f1": 0.8775,
|
| 164 |
+
"eval_pet_precision": 0.8796296296296297,
|
| 165 |
+
"eval_pet_recall": 0.7251908396946565,
|
| 166 |
+
"eval_pet_runtime": 0.9741,
|
| 167 |
+
"eval_pet_samples_per_second": 410.652,
|
| 168 |
+
"eval_pet_specificity": 0.9516728624535316,
|
| 169 |
+
"eval_pet_steps_per_second": 1.027,
|
| 170 |
+
"step": 350
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.0,
|
| 174 |
+
"grad_norm": 0.04784310236573219,
|
| 175 |
+
"learning_rate": 9.999999999999999e-06,
|
| 176 |
+
"loss": 0.073,
|
| 177 |
+
"step": 400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 4.0,
|
| 181 |
+
"eval_pet_AUC": 0.94308862340021,
|
| 182 |
+
"eval_pet_accuracy": 0.8725,
|
| 183 |
+
"eval_pet_f1": 0.8104089219330854,
|
| 184 |
+
"eval_pet_loss": 0.5896663069725037,
|
| 185 |
+
"eval_pet_macro_f1": 0.8571818620964862,
|
| 186 |
+
"eval_pet_micro_f1": 0.8725,
|
| 187 |
+
"eval_pet_precision": 0.7898550724637681,
|
| 188 |
+
"eval_pet_recall": 0.8320610687022901,
|
| 189 |
+
"eval_pet_runtime": 0.9603,
|
| 190 |
+
"eval_pet_samples_per_second": 416.535,
|
| 191 |
+
"eval_pet_specificity": 0.8921933085501859,
|
| 192 |
+
"eval_pet_steps_per_second": 1.041,
|
| 193 |
+
"step": 400
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 4.5,
|
| 197 |
+
"grad_norm": 0.009993120096623898,
|
| 198 |
+
"learning_rate": 7.5e-06,
|
| 199 |
+
"loss": 0.0352,
|
| 200 |
+
"step": 450
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 4.5,
|
| 204 |
+
"eval_pet_AUC": 0.9398393825023411,
|
| 205 |
+
"eval_pet_accuracy": 0.8775,
|
| 206 |
+
"eval_pet_f1": 0.7983539094650206,
|
| 207 |
+
"eval_pet_loss": 0.822050154209137,
|
| 208 |
+
"eval_pet_macro_f1": 0.8551913173896019,
|
| 209 |
+
"eval_pet_micro_f1": 0.8775,
|
| 210 |
+
"eval_pet_precision": 0.8660714285714286,
|
| 211 |
+
"eval_pet_recall": 0.7404580152671756,
|
| 212 |
+
"eval_pet_runtime": 0.9738,
|
| 213 |
+
"eval_pet_samples_per_second": 410.779,
|
| 214 |
+
"eval_pet_specificity": 0.9442379182156134,
|
| 215 |
+
"eval_pet_steps_per_second": 1.027,
|
| 216 |
+
"step": 450
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 5.0,
|
| 220 |
+
"grad_norm": 4.358645915985107,
|
| 221 |
+
"learning_rate": 4.9999999999999996e-06,
|
| 222 |
+
"loss": 0.0281,
|
| 223 |
+
"step": 500
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 5.0,
|
| 227 |
+
"eval_pet_AUC": 0.9366327080791168,
|
| 228 |
+
"eval_pet_accuracy": 0.87,
|
| 229 |
+
"eval_pet_f1": 0.7903225806451613,
|
| 230 |
+
"eval_pet_loss": 0.8214367628097534,
|
| 231 |
+
"eval_pet_macro_f1": 0.8480598410472183,
|
| 232 |
+
"eval_pet_micro_f1": 0.87,
|
| 233 |
+
"eval_pet_precision": 0.8376068376068376,
|
| 234 |
+
"eval_pet_recall": 0.7480916030534351,
|
| 235 |
+
"eval_pet_runtime": 0.9675,
|
| 236 |
+
"eval_pet_samples_per_second": 413.416,
|
| 237 |
+
"eval_pet_specificity": 0.929368029739777,
|
| 238 |
+
"eval_pet_steps_per_second": 1.034,
|
| 239 |
+
"step": 500
|
| 240 |
+
}
|
| 241 |
+
],
|
| 242 |
+
"logging_steps": 50,
|
| 243 |
+
"max_steps": 600,
|
| 244 |
+
"num_input_tokens_seen": 0,
|
| 245 |
+
"num_train_epochs": 6,
|
| 246 |
+
"save_steps": 500,
|
| 247 |
+
"stateful_callbacks": {
|
| 248 |
+
"TrainerControl": {
|
| 249 |
+
"args": {
|
| 250 |
+
"should_epoch_stop": false,
|
| 251 |
+
"should_evaluate": false,
|
| 252 |
+
"should_log": false,
|
| 253 |
+
"should_save": true,
|
| 254 |
+
"should_training_stop": false
|
| 255 |
+
},
|
| 256 |
+
"attributes": {}
|
| 257 |
+
}
|
| 258 |
+
},
|
| 259 |
+
"total_flos": 679093866240000.0,
|
| 260 |
+
"train_batch_size": 16,
|
| 261 |
+
"trial_name": null,
|
| 262 |
+
"trial_params": null
|
| 263 |
+
}
|
checkpoint-500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:782038bc3d4de4077e2809a63f4570c5ba00c054cf84c7f0402dfaeb26a11ccd
|
| 3 |
+
size 5432
|