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:e9ee55187d13c358633b1d5edec5396b34a5add595430b30e07c6add3ecff8ef
|
| 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": 11.574865341186523,
|
| 14 |
+
"learning_rate": 1.8333333333333333e-05,
|
| 15 |
+
"loss": 0.6129,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.5,
|
| 20 |
+
"eval_sport_AUC": 0.7282902151975048,
|
| 21 |
+
"eval_sport_accuracy": 0.7025,
|
| 22 |
+
"eval_sport_f1": 0.5734767025089605,
|
| 23 |
+
"eval_sport_loss": 0.6572355628013611,
|
| 24 |
+
"eval_sport_macro_f1": 0.6725348963600465,
|
| 25 |
+
"eval_sport_micro_f1": 0.7025,
|
| 26 |
+
"eval_sport_precision": 0.6557377049180327,
|
| 27 |
+
"eval_sport_recall": 0.5095541401273885,
|
| 28 |
+
"eval_sport_runtime": 1.1513,
|
| 29 |
+
"eval_sport_samples_per_second": 347.442,
|
| 30 |
+
"eval_sport_specificity": 0.8271604938271605,
|
| 31 |
+
"eval_sport_steps_per_second": 0.869,
|
| 32 |
+
"step": 50
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 1.0,
|
| 36 |
+
"grad_norm": 4.967200756072998,
|
| 37 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 38 |
+
"loss": 0.588,
|
| 39 |
+
"step": 100
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.0,
|
| 43 |
+
"eval_sport_AUC": 0.7831511624859111,
|
| 44 |
+
"eval_sport_accuracy": 0.745,
|
| 45 |
+
"eval_sport_f1": 0.5819672131147541,
|
| 46 |
+
"eval_sport_loss": 0.5513046979904175,
|
| 47 |
+
"eval_sport_macro_f1": 0.6992569878523411,
|
| 48 |
+
"eval_sport_micro_f1": 0.745,
|
| 49 |
+
"eval_sport_precision": 0.8160919540229885,
|
| 50 |
+
"eval_sport_recall": 0.45222929936305734,
|
| 51 |
+
"eval_sport_runtime": 1.1692,
|
| 52 |
+
"eval_sport_samples_per_second": 342.11,
|
| 53 |
+
"eval_sport_specificity": 0.934156378600823,
|
| 54 |
+
"eval_sport_steps_per_second": 0.855,
|
| 55 |
+
"step": 100
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 1.5,
|
| 59 |
+
"grad_norm": 4.148317813873291,
|
| 60 |
+
"learning_rate": 1.5000000000000002e-05,
|
| 61 |
+
"loss": 0.4635,
|
| 62 |
+
"step": 150
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.5,
|
| 66 |
+
"eval_sport_AUC": 0.8006867447773324,
|
| 67 |
+
"eval_sport_accuracy": 0.78,
|
| 68 |
+
"eval_sport_f1": 0.6879432624113475,
|
| 69 |
+
"eval_sport_loss": 0.523375391960144,
|
| 70 |
+
"eval_sport_macro_f1": 0.7590295462635888,
|
| 71 |
+
"eval_sport_micro_f1": 0.78,
|
| 72 |
+
"eval_sport_precision": 0.776,
|
| 73 |
+
"eval_sport_recall": 0.6178343949044586,
|
| 74 |
+
"eval_sport_runtime": 1.1639,
|
| 75 |
+
"eval_sport_samples_per_second": 343.677,
|
| 76 |
+
"eval_sport_specificity": 0.8847736625514403,
|
| 77 |
+
"eval_sport_steps_per_second": 0.859,
|
| 78 |
+
"step": 150
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 2.0,
|
| 82 |
+
"grad_norm": 6.097774982452393,
|
| 83 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 84 |
+
"loss": 0.4886,
|
| 85 |
+
"step": 200
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 2.0,
|
| 89 |
+
"eval_sport_AUC": 0.8396896542685643,
|
| 90 |
+
"eval_sport_accuracy": 0.7675,
|
| 91 |
+
"eval_sport_f1": 0.6910299003322259,
|
| 92 |
+
"eval_sport_loss": 0.5737913250923157,
|
| 93 |
+
"eval_sport_macro_f1": 0.752328577420622,
|
| 94 |
+
"eval_sport_micro_f1": 0.7675,
|
| 95 |
+
"eval_sport_precision": 0.7222222222222222,
|
| 96 |
+
"eval_sport_recall": 0.6624203821656051,
|
| 97 |
+
"eval_sport_runtime": 1.1526,
|
| 98 |
+
"eval_sport_samples_per_second": 347.033,
|
| 99 |
+
"eval_sport_specificity": 0.8353909465020576,
|
| 100 |
+
"eval_sport_steps_per_second": 0.868,
|
| 101 |
+
"step": 200
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 2.5,
|
| 105 |
+
"grad_norm": 2.1180505752563477,
|
| 106 |
+
"learning_rate": 1.1666666666666668e-05,
|
| 107 |
+
"loss": 0.4257,
|
| 108 |
+
"step": 250
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 2.5,
|
| 112 |
+
"eval_sport_AUC": 0.825836806374669,
|
| 113 |
+
"eval_sport_accuracy": 0.7725,
|
| 114 |
+
"eval_sport_f1": 0.6915254237288135,
|
| 115 |
+
"eval_sport_loss": 0.6375756859779358,
|
| 116 |
+
"eval_sport_macro_f1": 0.7556637019634167,
|
| 117 |
+
"eval_sport_micro_f1": 0.7725,
|
| 118 |
+
"eval_sport_precision": 0.7391304347826086,
|
| 119 |
+
"eval_sport_recall": 0.6496815286624203,
|
| 120 |
+
"eval_sport_runtime": 1.1419,
|
| 121 |
+
"eval_sport_samples_per_second": 350.281,
|
| 122 |
+
"eval_sport_specificity": 0.8518518518518519,
|
| 123 |
+
"eval_sport_steps_per_second": 0.876,
|
| 124 |
+
"step": 250
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 3.0,
|
| 128 |
+
"grad_norm": 8.99002742767334,
|
| 129 |
+
"learning_rate": 1e-05,
|
| 130 |
+
"loss": 0.4228,
|
| 131 |
+
"step": 300
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 3.0,
|
| 135 |
+
"eval_sport_AUC": 0.837107808445388,
|
| 136 |
+
"eval_sport_accuracy": 0.755,
|
| 137 |
+
"eval_sport_f1": 0.711764705882353,
|
| 138 |
+
"eval_sport_loss": 0.7621915936470032,
|
| 139 |
+
"eval_sport_macro_f1": 0.7493606138107417,
|
| 140 |
+
"eval_sport_micro_f1": 0.755,
|
| 141 |
+
"eval_sport_precision": 0.6612021857923497,
|
| 142 |
+
"eval_sport_recall": 0.7707006369426752,
|
| 143 |
+
"eval_sport_runtime": 1.1477,
|
| 144 |
+
"eval_sport_samples_per_second": 348.521,
|
| 145 |
+
"eval_sport_specificity": 0.7448559670781894,
|
| 146 |
+
"eval_sport_steps_per_second": 0.871,
|
| 147 |
+
"step": 300
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 3.5,
|
| 151 |
+
"grad_norm": 4.307883262634277,
|
| 152 |
+
"learning_rate": 8.333333333333334e-06,
|
| 153 |
+
"loss": 0.3663,
|
| 154 |
+
"step": 350
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 3.5,
|
| 158 |
+
"eval_sport_AUC": 0.8288904615868523,
|
| 159 |
+
"eval_sport_accuracy": 0.755,
|
| 160 |
+
"eval_sport_f1": 0.6754966887417219,
|
| 161 |
+
"eval_sport_loss": 0.6865992546081543,
|
| 162 |
+
"eval_sport_macro_f1": 0.7393547700736722,
|
| 163 |
+
"eval_sport_micro_f1": 0.755,
|
| 164 |
+
"eval_sport_precision": 0.7034482758620689,
|
| 165 |
+
"eval_sport_recall": 0.6496815286624203,
|
| 166 |
+
"eval_sport_runtime": 1.158,
|
| 167 |
+
"eval_sport_samples_per_second": 345.415,
|
| 168 |
+
"eval_sport_specificity": 0.823045267489712,
|
| 169 |
+
"eval_sport_steps_per_second": 0.864,
|
| 170 |
+
"step": 350
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.0,
|
| 174 |
+
"grad_norm": 25.94264030456543,
|
| 175 |
+
"learning_rate": 6.666666666666667e-06,
|
| 176 |
+
"loss": 0.3632,
|
| 177 |
+
"step": 400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 4.0,
|
| 181 |
+
"eval_sport_AUC": 0.8350108778275799,
|
| 182 |
+
"eval_sport_accuracy": 0.765,
|
| 183 |
+
"eval_sport_f1": 0.6907894736842105,
|
| 184 |
+
"eval_sport_loss": 0.7082327008247375,
|
| 185 |
+
"eval_sport_macro_f1": 0.7506366723259762,
|
| 186 |
+
"eval_sport_micro_f1": 0.765,
|
| 187 |
+
"eval_sport_precision": 0.7142857142857143,
|
| 188 |
+
"eval_sport_recall": 0.6687898089171974,
|
| 189 |
+
"eval_sport_runtime": 1.1439,
|
| 190 |
+
"eval_sport_samples_per_second": 349.69,
|
| 191 |
+
"eval_sport_specificity": 0.8271604938271605,
|
| 192 |
+
"eval_sport_steps_per_second": 0.874,
|
| 193 |
+
"step": 400
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 4.5,
|
| 197 |
+
"grad_norm": 3.1994056701660156,
|
| 198 |
+
"learning_rate": 5e-06,
|
| 199 |
+
"loss": 0.3067,
|
| 200 |
+
"step": 450
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 4.5,
|
| 204 |
+
"eval_sport_AUC": 0.8278682079106707,
|
| 205 |
+
"eval_sport_accuracy": 0.7825,
|
| 206 |
+
"eval_sport_f1": 0.6947368421052632,
|
| 207 |
+
"eval_sport_loss": 0.6867804527282715,
|
| 208 |
+
"eval_sport_macro_f1": 0.7629024016351559,
|
| 209 |
+
"eval_sport_micro_f1": 0.7825,
|
| 210 |
+
"eval_sport_precision": 0.7734375,
|
| 211 |
+
"eval_sport_recall": 0.6305732484076433,
|
| 212 |
+
"eval_sport_runtime": 1.1444,
|
| 213 |
+
"eval_sport_samples_per_second": 349.54,
|
| 214 |
+
"eval_sport_specificity": 0.8806584362139918,
|
| 215 |
+
"eval_sport_steps_per_second": 0.874,
|
| 216 |
+
"step": 450
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 5.0,
|
| 220 |
+
"grad_norm": 3.9464383125305176,
|
| 221 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 222 |
+
"loss": 0.3167,
|
| 223 |
+
"step": 500
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 5.0,
|
| 227 |
+
"eval_sport_AUC": 0.8293622709758591,
|
| 228 |
+
"eval_sport_accuracy": 0.7725,
|
| 229 |
+
"eval_sport_f1": 0.6956521739130435,
|
| 230 |
+
"eval_sport_loss": 0.7660211324691772,
|
| 231 |
+
"eval_sport_macro_f1": 0.7570077236830686,
|
| 232 |
+
"eval_sport_micro_f1": 0.7725,
|
| 233 |
+
"eval_sport_precision": 0.7323943661971831,
|
| 234 |
+
"eval_sport_recall": 0.6624203821656051,
|
| 235 |
+
"eval_sport_runtime": 1.1648,
|
| 236 |
+
"eval_sport_samples_per_second": 343.4,
|
| 237 |
+
"eval_sport_specificity": 0.8436213991769548,
|
| 238 |
+
"eval_sport_steps_per_second": 0.859,
|
| 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": 678669432573600.0,
|
| 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 sha256:65d0e85882425e1f397469e25cd0fb2f8d00c20c484516af4471b0ee816d77f8
|
| 3 |
+
size 5432
|