Upload folder using huggingface_hub
Browse files- all_results.json +9 -0
- checkpoint-448/config.json +44 -0
- checkpoint-448/model.safetensors +3 -0
- checkpoint-448/optimizer.pt +3 -0
- checkpoint-448/rng_state.pth +3 -0
- checkpoint-448/scheduler.pt +3 -0
- checkpoint-448/tokenizer.json +0 -0
- checkpoint-448/tokenizer_config.json +15 -0
- checkpoint-448/trainer_state.json +193 -0
- checkpoint-448/training_args.bin +3 -0
- checkpoint-560/config.json +44 -0
- checkpoint-560/model.safetensors +3 -0
- checkpoint-560/optimizer.pt +3 -0
- checkpoint-560/rng_state.pth +3 -0
- checkpoint-560/scheduler.pt +3 -0
- checkpoint-560/tokenizer.json +0 -0
- checkpoint-560/tokenizer_config.json +15 -0
- checkpoint-560/trainer_state.json +238 -0
- checkpoint-560/training_args.bin +3 -0
- config.json +44 -0
- model.safetensors +3 -0
- test_results.json +9 -0
- tokenizer.json +0 -0
- tokenizer_config.json +15 -0
- trainer_state.json +257 -0
- training_args.bin +3 -0
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"test_loss": 0.17423087358474731,
|
| 4 |
+
"test_macro_f1": 0.8512159238545282,
|
| 5 |
+
"test_micro_f1": 0.896807720861173,
|
| 6 |
+
"test_runtime": 12.366,
|
| 7 |
+
"test_samples_per_second": 30.972,
|
| 8 |
+
"test_steps_per_second": 0.97
|
| 9 |
+
}
|
checkpoint-448/config.json
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation": "gelu",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"DistilBertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_dim": 3072,
|
| 13 |
+
"id2label": {
|
| 14 |
+
"0": "prompt_injection",
|
| 15 |
+
"1": "sensitive_data_exposure",
|
| 16 |
+
"2": "unauthorized_tool_use",
|
| 17 |
+
"3": "policy_violation",
|
| 18 |
+
"4": "high_impact_action",
|
| 19 |
+
"5": "safe_request"
|
| 20 |
+
},
|
| 21 |
+
"initializer_range": 0.02,
|
| 22 |
+
"label2id": {
|
| 23 |
+
"high_impact_action": 4,
|
| 24 |
+
"policy_violation": 3,
|
| 25 |
+
"prompt_injection": 0,
|
| 26 |
+
"safe_request": 5,
|
| 27 |
+
"sensitive_data_exposure": 1,
|
| 28 |
+
"unauthorized_tool_use": 2
|
| 29 |
+
},
|
| 30 |
+
"max_position_embeddings": 512,
|
| 31 |
+
"model_type": "distilbert",
|
| 32 |
+
"n_heads": 12,
|
| 33 |
+
"n_layers": 6,
|
| 34 |
+
"pad_token_id": 0,
|
| 35 |
+
"problem_type": "multi_label_classification",
|
| 36 |
+
"qa_dropout": 0.1,
|
| 37 |
+
"seq_classif_dropout": 0.2,
|
| 38 |
+
"sinusoidal_pos_embds": false,
|
| 39 |
+
"tie_weights_": true,
|
| 40 |
+
"tie_word_embeddings": true,
|
| 41 |
+
"transformers_version": "5.8.1",
|
| 42 |
+
"use_cache": false,
|
| 43 |
+
"vocab_size": 30522
|
| 44 |
+
}
|
checkpoint-448/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3957158c3a1a1b9a666e3b59e1e444104402a043eba7952ebd8d711f3efdb8db
|
| 3 |
+
size 267844872
|
checkpoint-448/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bbc0e30ab3cac8b0ec278dad2743ed8b3c873cd7cc2c23ff0b901acf9d4f3036
|
| 3 |
+
size 535749451
|
checkpoint-448/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81b014eb753f7439e11a68de2d95f4381eac6810ab5e3abc64b0fe861ec5843d
|
| 3 |
+
size 14455
|
checkpoint-448/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4830eadc6d5513a211beacafc3bf8c060ba403373db0781e4eea00a2254b8312
|
| 3 |
+
size 1465
|
checkpoint-448/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-448/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"do_lower_case": true,
|
| 5 |
+
"is_local": false,
|
| 6 |
+
"local_files_only": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
checkpoint-448/trainer_state.json
ADDED
|
@@ -0,0 +1,193 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 448,
|
| 3 |
+
"best_metric": 0.8342735133305208,
|
| 4 |
+
"best_model_checkpoint": "outputs/models/risk_model\\checkpoint-448",
|
| 5 |
+
"epoch": 4.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 448,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.22321428571428573,
|
| 14 |
+
"grad_norm": 0.7896187901496887,
|
| 15 |
+
"learning_rate": 8.571428571428571e-06,
|
| 16 |
+
"loss": 0.6760891723632813,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.44642857142857145,
|
| 21 |
+
"grad_norm": 0.5187107920646667,
|
| 22 |
+
"learning_rate": 1.7500000000000002e-05,
|
| 23 |
+
"loss": 0.5672797775268554,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.6696428571428571,
|
| 28 |
+
"grad_norm": 0.9706220626831055,
|
| 29 |
+
"learning_rate": 1.928571428571429e-05,
|
| 30 |
+
"loss": 0.49257232666015627,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.8928571428571429,
|
| 35 |
+
"grad_norm": 1.0596867799758911,
|
| 36 |
+
"learning_rate": 1.8293650793650794e-05,
|
| 37 |
+
"loss": 0.40760456085205077,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.0,
|
| 42 |
+
"eval_loss": 0.33985912799835205,
|
| 43 |
+
"eval_macro_f1": 0.5031273386176547,
|
| 44 |
+
"eval_micro_f1": 0.7580645161290323,
|
| 45 |
+
"eval_runtime": 12.5726,
|
| 46 |
+
"eval_samples_per_second": 30.463,
|
| 47 |
+
"eval_steps_per_second": 0.954,
|
| 48 |
+
"step": 112
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 1.1160714285714286,
|
| 52 |
+
"grad_norm": 0.873076856136322,
|
| 53 |
+
"learning_rate": 1.7301587301587302e-05,
|
| 54 |
+
"loss": 0.367144889831543,
|
| 55 |
+
"step": 125
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 1.3392857142857144,
|
| 59 |
+
"grad_norm": 1.0305687189102173,
|
| 60 |
+
"learning_rate": 1.630952380952381e-05,
|
| 61 |
+
"loss": 0.29081342697143553,
|
| 62 |
+
"step": 150
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.5625,
|
| 66 |
+
"grad_norm": 1.1709494590759277,
|
| 67 |
+
"learning_rate": 1.531746031746032e-05,
|
| 68 |
+
"loss": 0.27378103256225583,
|
| 69 |
+
"step": 175
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.7857142857142856,
|
| 73 |
+
"grad_norm": 0.8700473308563232,
|
| 74 |
+
"learning_rate": 1.4325396825396826e-05,
|
| 75 |
+
"loss": 0.2479074478149414,
|
| 76 |
+
"step": 200
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 2.0,
|
| 80 |
+
"eval_loss": 0.2255277782678604,
|
| 81 |
+
"eval_macro_f1": 0.7545455310224004,
|
| 82 |
+
"eval_micro_f1": 0.8553940321346595,
|
| 83 |
+
"eval_runtime": 12.437,
|
| 84 |
+
"eval_samples_per_second": 30.795,
|
| 85 |
+
"eval_steps_per_second": 0.965,
|
| 86 |
+
"step": 224
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.0089285714285716,
|
| 90 |
+
"grad_norm": 0.8799751400947571,
|
| 91 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 92 |
+
"loss": 0.2470851707458496,
|
| 93 |
+
"step": 225
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 2.232142857142857,
|
| 97 |
+
"grad_norm": 0.8738771677017212,
|
| 98 |
+
"learning_rate": 1.2341269841269844e-05,
|
| 99 |
+
"loss": 0.21093284606933593,
|
| 100 |
+
"step": 250
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 2.455357142857143,
|
| 104 |
+
"grad_norm": 1.715822696685791,
|
| 105 |
+
"learning_rate": 1.134920634920635e-05,
|
| 106 |
+
"loss": 0.21485193252563475,
|
| 107 |
+
"step": 275
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 2.678571428571429,
|
| 111 |
+
"grad_norm": 0.783058762550354,
|
| 112 |
+
"learning_rate": 1.0357142857142859e-05,
|
| 113 |
+
"loss": 0.20812702178955078,
|
| 114 |
+
"step": 300
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 2.9017857142857144,
|
| 118 |
+
"grad_norm": 1.1336286067962646,
|
| 119 |
+
"learning_rate": 9.365079365079366e-06,
|
| 120 |
+
"loss": 0.19541482925415038,
|
| 121 |
+
"step": 325
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 3.0,
|
| 125 |
+
"eval_loss": 0.19746434688568115,
|
| 126 |
+
"eval_macro_f1": 0.7849450206606753,
|
| 127 |
+
"eval_micro_f1": 0.8660578386605784,
|
| 128 |
+
"eval_runtime": 12.4428,
|
| 129 |
+
"eval_samples_per_second": 30.781,
|
| 130 |
+
"eval_steps_per_second": 0.964,
|
| 131 |
+
"step": 336
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 3.125,
|
| 135 |
+
"grad_norm": 1.2524009943008423,
|
| 136 |
+
"learning_rate": 8.373015873015875e-06,
|
| 137 |
+
"loss": 0.1961466598510742,
|
| 138 |
+
"step": 350
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 3.3482142857142856,
|
| 142 |
+
"grad_norm": 1.4880211353302002,
|
| 143 |
+
"learning_rate": 7.380952380952382e-06,
|
| 144 |
+
"loss": 0.1762843894958496,
|
| 145 |
+
"step": 375
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 3.571428571428571,
|
| 149 |
+
"grad_norm": 1.419501781463623,
|
| 150 |
+
"learning_rate": 6.3888888888888885e-06,
|
| 151 |
+
"loss": 0.1805323028564453,
|
| 152 |
+
"step": 400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.794642857142857,
|
| 156 |
+
"grad_norm": 1.1280697584152222,
|
| 157 |
+
"learning_rate": 5.396825396825397e-06,
|
| 158 |
+
"loss": 0.16648523330688478,
|
| 159 |
+
"step": 425
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 4.0,
|
| 163 |
+
"eval_loss": 0.1824188232421875,
|
| 164 |
+
"eval_macro_f1": 0.8342735133305208,
|
| 165 |
+
"eval_micro_f1": 0.8823082763857252,
|
| 166 |
+
"eval_runtime": 14.2332,
|
| 167 |
+
"eval_samples_per_second": 26.909,
|
| 168 |
+
"eval_steps_per_second": 0.843,
|
| 169 |
+
"step": 448
|
| 170 |
+
}
|
| 171 |
+
],
|
| 172 |
+
"logging_steps": 25,
|
| 173 |
+
"max_steps": 560,
|
| 174 |
+
"num_input_tokens_seen": 0,
|
| 175 |
+
"num_train_epochs": 5,
|
| 176 |
+
"save_steps": 500,
|
| 177 |
+
"stateful_callbacks": {
|
| 178 |
+
"TrainerControl": {
|
| 179 |
+
"args": {
|
| 180 |
+
"should_epoch_stop": false,
|
| 181 |
+
"should_evaluate": false,
|
| 182 |
+
"should_log": false,
|
| 183 |
+
"should_save": true,
|
| 184 |
+
"should_training_stop": false
|
| 185 |
+
},
|
| 186 |
+
"attributes": {}
|
| 187 |
+
}
|
| 188 |
+
},
|
| 189 |
+
"total_flos": 182184645486528.0,
|
| 190 |
+
"train_batch_size": 16,
|
| 191 |
+
"trial_name": null,
|
| 192 |
+
"trial_params": null
|
| 193 |
+
}
|
checkpoint-448/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea11d82e0b8b24b4999106b3c615c65439c3ec832161929cf2a960e32a0e5b12
|
| 3 |
+
size 5265
|
checkpoint-560/config.json
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation": "gelu",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"DistilBertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_dim": 3072,
|
| 13 |
+
"id2label": {
|
| 14 |
+
"0": "prompt_injection",
|
| 15 |
+
"1": "sensitive_data_exposure",
|
| 16 |
+
"2": "unauthorized_tool_use",
|
| 17 |
+
"3": "policy_violation",
|
| 18 |
+
"4": "high_impact_action",
|
| 19 |
+
"5": "safe_request"
|
| 20 |
+
},
|
| 21 |
+
"initializer_range": 0.02,
|
| 22 |
+
"label2id": {
|
| 23 |
+
"high_impact_action": 4,
|
| 24 |
+
"policy_violation": 3,
|
| 25 |
+
"prompt_injection": 0,
|
| 26 |
+
"safe_request": 5,
|
| 27 |
+
"sensitive_data_exposure": 1,
|
| 28 |
+
"unauthorized_tool_use": 2
|
| 29 |
+
},
|
| 30 |
+
"max_position_embeddings": 512,
|
| 31 |
+
"model_type": "distilbert",
|
| 32 |
+
"n_heads": 12,
|
| 33 |
+
"n_layers": 6,
|
| 34 |
+
"pad_token_id": 0,
|
| 35 |
+
"problem_type": "multi_label_classification",
|
| 36 |
+
"qa_dropout": 0.1,
|
| 37 |
+
"seq_classif_dropout": 0.2,
|
| 38 |
+
"sinusoidal_pos_embds": false,
|
| 39 |
+
"tie_weights_": true,
|
| 40 |
+
"tie_word_embeddings": true,
|
| 41 |
+
"transformers_version": "5.8.1",
|
| 42 |
+
"use_cache": false,
|
| 43 |
+
"vocab_size": 30522
|
| 44 |
+
}
|
checkpoint-560/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e0e4e651b2a5185101fe50852a441c3f7f831a5938a1b5d524bb814f645d392c
|
| 3 |
+
size 267844872
|
checkpoint-560/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b4c652b4f29e8c06b6df7b2127dfb93d0b87d5e5a3d50ff1eb54ec95f193f447
|
| 3 |
+
size 535749451
|
checkpoint-560/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b07bd7d9e42639ecf3ad67693bef88a7de91c0464fe9b70c3e684e68445a1a48
|
| 3 |
+
size 14455
|
checkpoint-560/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:40476b5d6f3134c47ddd0ac698e7c83ed4d75f838612b706b508a9d236e7d282
|
| 3 |
+
size 1465
|
checkpoint-560/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-560/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"do_lower_case": true,
|
| 5 |
+
"is_local": false,
|
| 6 |
+
"local_files_only": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
checkpoint-560/trainer_state.json
ADDED
|
@@ -0,0 +1,238 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 448,
|
| 3 |
+
"best_metric": 0.8342735133305208,
|
| 4 |
+
"best_model_checkpoint": "outputs/models/risk_model\\checkpoint-448",
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 560,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.22321428571428573,
|
| 14 |
+
"grad_norm": 0.7896187901496887,
|
| 15 |
+
"learning_rate": 8.571428571428571e-06,
|
| 16 |
+
"loss": 0.6760891723632813,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.44642857142857145,
|
| 21 |
+
"grad_norm": 0.5187107920646667,
|
| 22 |
+
"learning_rate": 1.7500000000000002e-05,
|
| 23 |
+
"loss": 0.5672797775268554,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.6696428571428571,
|
| 28 |
+
"grad_norm": 0.9706220626831055,
|
| 29 |
+
"learning_rate": 1.928571428571429e-05,
|
| 30 |
+
"loss": 0.49257232666015627,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.8928571428571429,
|
| 35 |
+
"grad_norm": 1.0596867799758911,
|
| 36 |
+
"learning_rate": 1.8293650793650794e-05,
|
| 37 |
+
"loss": 0.40760456085205077,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.0,
|
| 42 |
+
"eval_loss": 0.33985912799835205,
|
| 43 |
+
"eval_macro_f1": 0.5031273386176547,
|
| 44 |
+
"eval_micro_f1": 0.7580645161290323,
|
| 45 |
+
"eval_runtime": 12.5726,
|
| 46 |
+
"eval_samples_per_second": 30.463,
|
| 47 |
+
"eval_steps_per_second": 0.954,
|
| 48 |
+
"step": 112
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 1.1160714285714286,
|
| 52 |
+
"grad_norm": 0.873076856136322,
|
| 53 |
+
"learning_rate": 1.7301587301587302e-05,
|
| 54 |
+
"loss": 0.367144889831543,
|
| 55 |
+
"step": 125
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 1.3392857142857144,
|
| 59 |
+
"grad_norm": 1.0305687189102173,
|
| 60 |
+
"learning_rate": 1.630952380952381e-05,
|
| 61 |
+
"loss": 0.29081342697143553,
|
| 62 |
+
"step": 150
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.5625,
|
| 66 |
+
"grad_norm": 1.1709494590759277,
|
| 67 |
+
"learning_rate": 1.531746031746032e-05,
|
| 68 |
+
"loss": 0.27378103256225583,
|
| 69 |
+
"step": 175
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.7857142857142856,
|
| 73 |
+
"grad_norm": 0.8700473308563232,
|
| 74 |
+
"learning_rate": 1.4325396825396826e-05,
|
| 75 |
+
"loss": 0.2479074478149414,
|
| 76 |
+
"step": 200
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 2.0,
|
| 80 |
+
"eval_loss": 0.2255277782678604,
|
| 81 |
+
"eval_macro_f1": 0.7545455310224004,
|
| 82 |
+
"eval_micro_f1": 0.8553940321346595,
|
| 83 |
+
"eval_runtime": 12.437,
|
| 84 |
+
"eval_samples_per_second": 30.795,
|
| 85 |
+
"eval_steps_per_second": 0.965,
|
| 86 |
+
"step": 224
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.0089285714285716,
|
| 90 |
+
"grad_norm": 0.8799751400947571,
|
| 91 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 92 |
+
"loss": 0.2470851707458496,
|
| 93 |
+
"step": 225
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 2.232142857142857,
|
| 97 |
+
"grad_norm": 0.8738771677017212,
|
| 98 |
+
"learning_rate": 1.2341269841269844e-05,
|
| 99 |
+
"loss": 0.21093284606933593,
|
| 100 |
+
"step": 250
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 2.455357142857143,
|
| 104 |
+
"grad_norm": 1.715822696685791,
|
| 105 |
+
"learning_rate": 1.134920634920635e-05,
|
| 106 |
+
"loss": 0.21485193252563475,
|
| 107 |
+
"step": 275
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 2.678571428571429,
|
| 111 |
+
"grad_norm": 0.783058762550354,
|
| 112 |
+
"learning_rate": 1.0357142857142859e-05,
|
| 113 |
+
"loss": 0.20812702178955078,
|
| 114 |
+
"step": 300
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 2.9017857142857144,
|
| 118 |
+
"grad_norm": 1.1336286067962646,
|
| 119 |
+
"learning_rate": 9.365079365079366e-06,
|
| 120 |
+
"loss": 0.19541482925415038,
|
| 121 |
+
"step": 325
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 3.0,
|
| 125 |
+
"eval_loss": 0.19746434688568115,
|
| 126 |
+
"eval_macro_f1": 0.7849450206606753,
|
| 127 |
+
"eval_micro_f1": 0.8660578386605784,
|
| 128 |
+
"eval_runtime": 12.4428,
|
| 129 |
+
"eval_samples_per_second": 30.781,
|
| 130 |
+
"eval_steps_per_second": 0.964,
|
| 131 |
+
"step": 336
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 3.125,
|
| 135 |
+
"grad_norm": 1.2524009943008423,
|
| 136 |
+
"learning_rate": 8.373015873015875e-06,
|
| 137 |
+
"loss": 0.1961466598510742,
|
| 138 |
+
"step": 350
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 3.3482142857142856,
|
| 142 |
+
"grad_norm": 1.4880211353302002,
|
| 143 |
+
"learning_rate": 7.380952380952382e-06,
|
| 144 |
+
"loss": 0.1762843894958496,
|
| 145 |
+
"step": 375
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 3.571428571428571,
|
| 149 |
+
"grad_norm": 1.419501781463623,
|
| 150 |
+
"learning_rate": 6.3888888888888885e-06,
|
| 151 |
+
"loss": 0.1805323028564453,
|
| 152 |
+
"step": 400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.794642857142857,
|
| 156 |
+
"grad_norm": 1.1280697584152222,
|
| 157 |
+
"learning_rate": 5.396825396825397e-06,
|
| 158 |
+
"loss": 0.16648523330688478,
|
| 159 |
+
"step": 425
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 4.0,
|
| 163 |
+
"eval_loss": 0.1824188232421875,
|
| 164 |
+
"eval_macro_f1": 0.8342735133305208,
|
| 165 |
+
"eval_micro_f1": 0.8823082763857252,
|
| 166 |
+
"eval_runtime": 14.2332,
|
| 167 |
+
"eval_samples_per_second": 26.909,
|
| 168 |
+
"eval_steps_per_second": 0.843,
|
| 169 |
+
"step": 448
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 4.017857142857143,
|
| 173 |
+
"grad_norm": 0.8061633706092834,
|
| 174 |
+
"learning_rate": 4.404761904761905e-06,
|
| 175 |
+
"loss": 0.17099233627319335,
|
| 176 |
+
"step": 450
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.241071428571429,
|
| 180 |
+
"grad_norm": 1.1085864305496216,
|
| 181 |
+
"learning_rate": 3.412698412698413e-06,
|
| 182 |
+
"loss": 0.1619624137878418,
|
| 183 |
+
"step": 475
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 4.464285714285714,
|
| 187 |
+
"grad_norm": 0.9997268915176392,
|
| 188 |
+
"learning_rate": 2.420634920634921e-06,
|
| 189 |
+
"loss": 0.1607685089111328,
|
| 190 |
+
"step": 500
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 4.6875,
|
| 194 |
+
"grad_norm": 1.134731650352478,
|
| 195 |
+
"learning_rate": 1.4285714285714286e-06,
|
| 196 |
+
"loss": 0.16787288665771485,
|
| 197 |
+
"step": 525
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 4.910714285714286,
|
| 201 |
+
"grad_norm": 1.194025993347168,
|
| 202 |
+
"learning_rate": 4.365079365079365e-07,
|
| 203 |
+
"loss": 0.1533247184753418,
|
| 204 |
+
"step": 550
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 5.0,
|
| 208 |
+
"eval_loss": 0.17899587750434875,
|
| 209 |
+
"eval_macro_f1": 0.8317158215809735,
|
| 210 |
+
"eval_micro_f1": 0.8801213960546282,
|
| 211 |
+
"eval_runtime": 13.6418,
|
| 212 |
+
"eval_samples_per_second": 28.075,
|
| 213 |
+
"eval_steps_per_second": 0.88,
|
| 214 |
+
"step": 560
|
| 215 |
+
}
|
| 216 |
+
],
|
| 217 |
+
"logging_steps": 25,
|
| 218 |
+
"max_steps": 560,
|
| 219 |
+
"num_input_tokens_seen": 0,
|
| 220 |
+
"num_train_epochs": 5,
|
| 221 |
+
"save_steps": 500,
|
| 222 |
+
"stateful_callbacks": {
|
| 223 |
+
"TrainerControl": {
|
| 224 |
+
"args": {
|
| 225 |
+
"should_epoch_stop": false,
|
| 226 |
+
"should_evaluate": false,
|
| 227 |
+
"should_log": false,
|
| 228 |
+
"should_save": true,
|
| 229 |
+
"should_training_stop": true
|
| 230 |
+
},
|
| 231 |
+
"attributes": {}
|
| 232 |
+
}
|
| 233 |
+
},
|
| 234 |
+
"total_flos": 227810499962112.0,
|
| 235 |
+
"train_batch_size": 16,
|
| 236 |
+
"trial_name": null,
|
| 237 |
+
"trial_params": null
|
| 238 |
+
}
|
checkpoint-560/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea11d82e0b8b24b4999106b3c615c65439c3ec832161929cf2a960e32a0e5b12
|
| 3 |
+
size 5265
|
config.json
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation": "gelu",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"DistilBertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
+
"hidden_dim": 3072,
|
| 13 |
+
"id2label": {
|
| 14 |
+
"0": "prompt_injection",
|
| 15 |
+
"1": "sensitive_data_exposure",
|
| 16 |
+
"2": "unauthorized_tool_use",
|
| 17 |
+
"3": "policy_violation",
|
| 18 |
+
"4": "high_impact_action",
|
| 19 |
+
"5": "safe_request"
|
| 20 |
+
},
|
| 21 |
+
"initializer_range": 0.02,
|
| 22 |
+
"label2id": {
|
| 23 |
+
"high_impact_action": 4,
|
| 24 |
+
"policy_violation": 3,
|
| 25 |
+
"prompt_injection": 0,
|
| 26 |
+
"safe_request": 5,
|
| 27 |
+
"sensitive_data_exposure": 1,
|
| 28 |
+
"unauthorized_tool_use": 2
|
| 29 |
+
},
|
| 30 |
+
"max_position_embeddings": 512,
|
| 31 |
+
"model_type": "distilbert",
|
| 32 |
+
"n_heads": 12,
|
| 33 |
+
"n_layers": 6,
|
| 34 |
+
"pad_token_id": 0,
|
| 35 |
+
"problem_type": "multi_label_classification",
|
| 36 |
+
"qa_dropout": 0.1,
|
| 37 |
+
"seq_classif_dropout": 0.2,
|
| 38 |
+
"sinusoidal_pos_embds": false,
|
| 39 |
+
"tie_weights_": true,
|
| 40 |
+
"tie_word_embeddings": true,
|
| 41 |
+
"transformers_version": "5.8.1",
|
| 42 |
+
"use_cache": false,
|
| 43 |
+
"vocab_size": 30522
|
| 44 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3957158c3a1a1b9a666e3b59e1e444104402a043eba7952ebd8d711f3efdb8db
|
| 3 |
+
size 267844872
|
test_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"test_loss": 0.17423087358474731,
|
| 4 |
+
"test_macro_f1": 0.8512159238545282,
|
| 5 |
+
"test_micro_f1": 0.896807720861173,
|
| 6 |
+
"test_runtime": 12.366,
|
| 7 |
+
"test_samples_per_second": 30.972,
|
| 8 |
+
"test_steps_per_second": 0.97
|
| 9 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"do_lower_case": true,
|
| 5 |
+
"is_local": false,
|
| 6 |
+
"local_files_only": false,
|
| 7 |
+
"mask_token": "[MASK]",
|
| 8 |
+
"model_max_length": 512,
|
| 9 |
+
"pad_token": "[PAD]",
|
| 10 |
+
"sep_token": "[SEP]",
|
| 11 |
+
"strip_accents": null,
|
| 12 |
+
"tokenize_chinese_chars": true,
|
| 13 |
+
"tokenizer_class": "BertTokenizer",
|
| 14 |
+
"unk_token": "[UNK]"
|
| 15 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,257 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 448,
|
| 3 |
+
"best_metric": 0.8342735133305208,
|
| 4 |
+
"best_model_checkpoint": "outputs/models/risk_model\\checkpoint-448",
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 560,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.22321428571428573,
|
| 14 |
+
"grad_norm": 0.7896187901496887,
|
| 15 |
+
"learning_rate": 8.571428571428571e-06,
|
| 16 |
+
"loss": 0.6760891723632813,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.44642857142857145,
|
| 21 |
+
"grad_norm": 0.5187107920646667,
|
| 22 |
+
"learning_rate": 1.7500000000000002e-05,
|
| 23 |
+
"loss": 0.5672797775268554,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.6696428571428571,
|
| 28 |
+
"grad_norm": 0.9706220626831055,
|
| 29 |
+
"learning_rate": 1.928571428571429e-05,
|
| 30 |
+
"loss": 0.49257232666015627,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.8928571428571429,
|
| 35 |
+
"grad_norm": 1.0596867799758911,
|
| 36 |
+
"learning_rate": 1.8293650793650794e-05,
|
| 37 |
+
"loss": 0.40760456085205077,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.0,
|
| 42 |
+
"eval_loss": 0.33985912799835205,
|
| 43 |
+
"eval_macro_f1": 0.5031273386176547,
|
| 44 |
+
"eval_micro_f1": 0.7580645161290323,
|
| 45 |
+
"eval_runtime": 12.5726,
|
| 46 |
+
"eval_samples_per_second": 30.463,
|
| 47 |
+
"eval_steps_per_second": 0.954,
|
| 48 |
+
"step": 112
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 1.1160714285714286,
|
| 52 |
+
"grad_norm": 0.873076856136322,
|
| 53 |
+
"learning_rate": 1.7301587301587302e-05,
|
| 54 |
+
"loss": 0.367144889831543,
|
| 55 |
+
"step": 125
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 1.3392857142857144,
|
| 59 |
+
"grad_norm": 1.0305687189102173,
|
| 60 |
+
"learning_rate": 1.630952380952381e-05,
|
| 61 |
+
"loss": 0.29081342697143553,
|
| 62 |
+
"step": 150
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.5625,
|
| 66 |
+
"grad_norm": 1.1709494590759277,
|
| 67 |
+
"learning_rate": 1.531746031746032e-05,
|
| 68 |
+
"loss": 0.27378103256225583,
|
| 69 |
+
"step": 175
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.7857142857142856,
|
| 73 |
+
"grad_norm": 0.8700473308563232,
|
| 74 |
+
"learning_rate": 1.4325396825396826e-05,
|
| 75 |
+
"loss": 0.2479074478149414,
|
| 76 |
+
"step": 200
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 2.0,
|
| 80 |
+
"eval_loss": 0.2255277782678604,
|
| 81 |
+
"eval_macro_f1": 0.7545455310224004,
|
| 82 |
+
"eval_micro_f1": 0.8553940321346595,
|
| 83 |
+
"eval_runtime": 12.437,
|
| 84 |
+
"eval_samples_per_second": 30.795,
|
| 85 |
+
"eval_steps_per_second": 0.965,
|
| 86 |
+
"step": 224
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.0089285714285716,
|
| 90 |
+
"grad_norm": 0.8799751400947571,
|
| 91 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 92 |
+
"loss": 0.2470851707458496,
|
| 93 |
+
"step": 225
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 2.232142857142857,
|
| 97 |
+
"grad_norm": 0.8738771677017212,
|
| 98 |
+
"learning_rate": 1.2341269841269844e-05,
|
| 99 |
+
"loss": 0.21093284606933593,
|
| 100 |
+
"step": 250
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 2.455357142857143,
|
| 104 |
+
"grad_norm": 1.715822696685791,
|
| 105 |
+
"learning_rate": 1.134920634920635e-05,
|
| 106 |
+
"loss": 0.21485193252563475,
|
| 107 |
+
"step": 275
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 2.678571428571429,
|
| 111 |
+
"grad_norm": 0.783058762550354,
|
| 112 |
+
"learning_rate": 1.0357142857142859e-05,
|
| 113 |
+
"loss": 0.20812702178955078,
|
| 114 |
+
"step": 300
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 2.9017857142857144,
|
| 118 |
+
"grad_norm": 1.1336286067962646,
|
| 119 |
+
"learning_rate": 9.365079365079366e-06,
|
| 120 |
+
"loss": 0.19541482925415038,
|
| 121 |
+
"step": 325
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 3.0,
|
| 125 |
+
"eval_loss": 0.19746434688568115,
|
| 126 |
+
"eval_macro_f1": 0.7849450206606753,
|
| 127 |
+
"eval_micro_f1": 0.8660578386605784,
|
| 128 |
+
"eval_runtime": 12.4428,
|
| 129 |
+
"eval_samples_per_second": 30.781,
|
| 130 |
+
"eval_steps_per_second": 0.964,
|
| 131 |
+
"step": 336
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 3.125,
|
| 135 |
+
"grad_norm": 1.2524009943008423,
|
| 136 |
+
"learning_rate": 8.373015873015875e-06,
|
| 137 |
+
"loss": 0.1961466598510742,
|
| 138 |
+
"step": 350
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 3.3482142857142856,
|
| 142 |
+
"grad_norm": 1.4880211353302002,
|
| 143 |
+
"learning_rate": 7.380952380952382e-06,
|
| 144 |
+
"loss": 0.1762843894958496,
|
| 145 |
+
"step": 375
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 3.571428571428571,
|
| 149 |
+
"grad_norm": 1.419501781463623,
|
| 150 |
+
"learning_rate": 6.3888888888888885e-06,
|
| 151 |
+
"loss": 0.1805323028564453,
|
| 152 |
+
"step": 400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.794642857142857,
|
| 156 |
+
"grad_norm": 1.1280697584152222,
|
| 157 |
+
"learning_rate": 5.396825396825397e-06,
|
| 158 |
+
"loss": 0.16648523330688478,
|
| 159 |
+
"step": 425
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 4.0,
|
| 163 |
+
"eval_loss": 0.1824188232421875,
|
| 164 |
+
"eval_macro_f1": 0.8342735133305208,
|
| 165 |
+
"eval_micro_f1": 0.8823082763857252,
|
| 166 |
+
"eval_runtime": 14.2332,
|
| 167 |
+
"eval_samples_per_second": 26.909,
|
| 168 |
+
"eval_steps_per_second": 0.843,
|
| 169 |
+
"step": 448
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 4.017857142857143,
|
| 173 |
+
"grad_norm": 0.8061633706092834,
|
| 174 |
+
"learning_rate": 4.404761904761905e-06,
|
| 175 |
+
"loss": 0.17099233627319335,
|
| 176 |
+
"step": 450
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 4.241071428571429,
|
| 180 |
+
"grad_norm": 1.1085864305496216,
|
| 181 |
+
"learning_rate": 3.412698412698413e-06,
|
| 182 |
+
"loss": 0.1619624137878418,
|
| 183 |
+
"step": 475
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 4.464285714285714,
|
| 187 |
+
"grad_norm": 0.9997268915176392,
|
| 188 |
+
"learning_rate": 2.420634920634921e-06,
|
| 189 |
+
"loss": 0.1607685089111328,
|
| 190 |
+
"step": 500
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 4.6875,
|
| 194 |
+
"grad_norm": 1.134731650352478,
|
| 195 |
+
"learning_rate": 1.4285714285714286e-06,
|
| 196 |
+
"loss": 0.16787288665771485,
|
| 197 |
+
"step": 525
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 4.910714285714286,
|
| 201 |
+
"grad_norm": 1.194025993347168,
|
| 202 |
+
"learning_rate": 4.365079365079365e-07,
|
| 203 |
+
"loss": 0.1533247184753418,
|
| 204 |
+
"step": 550
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 5.0,
|
| 208 |
+
"eval_loss": 0.17899587750434875,
|
| 209 |
+
"eval_macro_f1": 0.8317158215809735,
|
| 210 |
+
"eval_micro_f1": 0.8801213960546282,
|
| 211 |
+
"eval_runtime": 13.6418,
|
| 212 |
+
"eval_samples_per_second": 28.075,
|
| 213 |
+
"eval_steps_per_second": 0.88,
|
| 214 |
+
"step": 560
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 5.0,
|
| 218 |
+
"step": 560,
|
| 219 |
+
"total_flos": 227810499962112.0,
|
| 220 |
+
"train_loss": 0.26780529745987486,
|
| 221 |
+
"train_runtime": 1216.99,
|
| 222 |
+
"train_samples_per_second": 7.33,
|
| 223 |
+
"train_steps_per_second": 0.46
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 5.0,
|
| 227 |
+
"step": 560,
|
| 228 |
+
"test_loss": 0.17423087358474731,
|
| 229 |
+
"test_macro_f1": 0.8512159238545282,
|
| 230 |
+
"test_micro_f1": 0.896807720861173,
|
| 231 |
+
"test_runtime": 12.366,
|
| 232 |
+
"test_samples_per_second": 30.972,
|
| 233 |
+
"test_steps_per_second": 0.97
|
| 234 |
+
}
|
| 235 |
+
],
|
| 236 |
+
"logging_steps": 25,
|
| 237 |
+
"max_steps": 560,
|
| 238 |
+
"num_input_tokens_seen": 0,
|
| 239 |
+
"num_train_epochs": 5,
|
| 240 |
+
"save_steps": 500,
|
| 241 |
+
"stateful_callbacks": {
|
| 242 |
+
"TrainerControl": {
|
| 243 |
+
"args": {
|
| 244 |
+
"should_epoch_stop": false,
|
| 245 |
+
"should_evaluate": false,
|
| 246 |
+
"should_log": false,
|
| 247 |
+
"should_save": true,
|
| 248 |
+
"should_training_stop": true
|
| 249 |
+
},
|
| 250 |
+
"attributes": {}
|
| 251 |
+
}
|
| 252 |
+
},
|
| 253 |
+
"total_flos": 227810499962112.0,
|
| 254 |
+
"train_batch_size": 16,
|
| 255 |
+
"trial_name": null,
|
| 256 |
+
"trial_params": null
|
| 257 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea11d82e0b8b24b4999106b3c615c65439c3ec832161929cf2a960e32a0e5b12
|
| 3 |
+
size 5265
|