Upload 11 files
Browse files- config.json +86 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +7 -0
- tokenizer.json +0 -0
- tokenizer_config.json +13 -0
- trainer_state.json +186 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
config.json
ADDED
|
@@ -0,0 +1,86 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "distilbert-base-uncased",
|
| 3 |
+
"activation": "gelu",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"DistilBertForMultilabelSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_dropout": 0.1,
|
| 8 |
+
"dim": 768,
|
| 9 |
+
"dropout": 0.1,
|
| 10 |
+
"hidden_dim": 3072,
|
| 11 |
+
"id2label": [
|
| 12 |
+
{
|
| 13 |
+
"0": "admiration",
|
| 14 |
+
"1": "amusement",
|
| 15 |
+
"10": "disapproval",
|
| 16 |
+
"11": "disgust",
|
| 17 |
+
"12": "embarrassment",
|
| 18 |
+
"13": "excitement",
|
| 19 |
+
"14": "fear",
|
| 20 |
+
"15": "gratitude",
|
| 21 |
+
"16": "grief",
|
| 22 |
+
"17": "joy",
|
| 23 |
+
"18": "love",
|
| 24 |
+
"19": "nervousness",
|
| 25 |
+
"2": "anger",
|
| 26 |
+
"20": "optimism",
|
| 27 |
+
"21": "pride",
|
| 28 |
+
"22": "realization",
|
| 29 |
+
"23": "relief",
|
| 30 |
+
"24": "remorse",
|
| 31 |
+
"25": "sadness",
|
| 32 |
+
"26": "surprise",
|
| 33 |
+
"27": "neutral",
|
| 34 |
+
"3": "annoyance",
|
| 35 |
+
"4": "approval",
|
| 36 |
+
"5": "caring",
|
| 37 |
+
"6": "confusion",
|
| 38 |
+
"7": "curiosity",
|
| 39 |
+
"8": "desire",
|
| 40 |
+
"9": "disappointment"
|
| 41 |
+
}
|
| 42 |
+
],
|
| 43 |
+
"initializer_range": 0.02,
|
| 44 |
+
"label2id": {
|
| 45 |
+
"admiration": 0,
|
| 46 |
+
"amusement": 1,
|
| 47 |
+
"anger": 2,
|
| 48 |
+
"annoyance": 3,
|
| 49 |
+
"approval": 4,
|
| 50 |
+
"caring": 5,
|
| 51 |
+
"confusion": 6,
|
| 52 |
+
"curiosity": 7,
|
| 53 |
+
"desire": 8,
|
| 54 |
+
"disappointment": 9,
|
| 55 |
+
"disapproval": 10,
|
| 56 |
+
"disgust": 11,
|
| 57 |
+
"embarrassment": 12,
|
| 58 |
+
"excitement": 13,
|
| 59 |
+
"fear": 14,
|
| 60 |
+
"gratitude": 15,
|
| 61 |
+
"grief": 16,
|
| 62 |
+
"joy": 17,
|
| 63 |
+
"love": 18,
|
| 64 |
+
"nervousness": 19,
|
| 65 |
+
"neutral": 27,
|
| 66 |
+
"optimism": 20,
|
| 67 |
+
"pride": 21,
|
| 68 |
+
"realization": 22,
|
| 69 |
+
"relief": 23,
|
| 70 |
+
"remorse": 24,
|
| 71 |
+
"sadness": 25,
|
| 72 |
+
"surprise": 26
|
| 73 |
+
},
|
| 74 |
+
"max_position_embeddings": 512,
|
| 75 |
+
"model_type": "distilbert",
|
| 76 |
+
"n_heads": 12,
|
| 77 |
+
"n_layers": 6,
|
| 78 |
+
"pad_token_id": 0,
|
| 79 |
+
"qa_dropout": 0.1,
|
| 80 |
+
"seq_classif_dropout": 0.2,
|
| 81 |
+
"sinusoidal_pos_embds": false,
|
| 82 |
+
"tie_weights_": true,
|
| 83 |
+
"torch_dtype": "float32",
|
| 84 |
+
"transformers_version": "4.28.1",
|
| 85 |
+
"vocab_size": 30522
|
| 86 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea3bbe2d8bdd5565454dd7c68ecceb347f4e157ba01389f5f6e5a5eb093c0061
|
| 3 |
+
size 531134981
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2fefe2e4cf610b43c385b84909dafdd6fb41005b1c3e0719f4dd3a28f8c5009e
|
| 3 |
+
size 267935469
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac1ffb7924c460fb76131293e9e9d7a0be0e93f63374133233bc4d73ecfe1d96
|
| 3 |
+
size 14575
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cee5f0ff00ee32da37f27264092fac385a34bb5a3552401f846f12b729cd05fe
|
| 3 |
+
size 627
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": "[CLS]",
|
| 3 |
+
"mask_token": "[MASK]",
|
| 4 |
+
"pad_token": "[PAD]",
|
| 5 |
+
"sep_token": "[SEP]",
|
| 6 |
+
"unk_token": "[UNK]"
|
| 7 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"clean_up_tokenization_spaces": true,
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"do_lower_case": true,
|
| 5 |
+
"mask_token": "[MASK]",
|
| 6 |
+
"model_max_length": 512,
|
| 7 |
+
"pad_token": "[PAD]",
|
| 8 |
+
"sep_token": "[SEP]",
|
| 9 |
+
"strip_accents": null,
|
| 10 |
+
"tokenize_chinese_chars": true,
|
| 11 |
+
"tokenizer_class": "DistilBertTokenizer",
|
| 12 |
+
"unk_token": "[UNK]"
|
| 13 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,186 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"global_step": 52940,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 1.0,
|
| 12 |
+
"learning_rate": 1.8000377786173025e-05,
|
| 13 |
+
"loss": 0.1263,
|
| 14 |
+
"step": 5293
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.0,
|
| 18 |
+
"eval_accuracy_thresh": 0.9616630673408508,
|
| 19 |
+
"eval_f1_macro": 0.16946832551734195,
|
| 20 |
+
"eval_f1_micro": 0.23566588370481772,
|
| 21 |
+
"eval_loss": 0.11356716603040695,
|
| 22 |
+
"eval_runtime": 48.6873,
|
| 23 |
+
"eval_samples_per_second": 859.506,
|
| 24 |
+
"eval_steps_per_second": 26.865,
|
| 25 |
+
"step": 5294
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 2.0,
|
| 29 |
+
"learning_rate": 1.6000755572346052e-05,
|
| 30 |
+
"loss": 0.1102,
|
| 31 |
+
"step": 10586
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 2.0,
|
| 35 |
+
"eval_accuracy_thresh": 0.9620087146759033,
|
| 36 |
+
"eval_f1_macro": 0.17900886975226568,
|
| 37 |
+
"eval_f1_micro": 0.2526954290189805,
|
| 38 |
+
"eval_loss": 0.11111615598201752,
|
| 39 |
+
"eval_runtime": 47.7242,
|
| 40 |
+
"eval_samples_per_second": 876.85,
|
| 41 |
+
"eval_steps_per_second": 27.407,
|
| 42 |
+
"step": 10588
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 3.0,
|
| 46 |
+
"learning_rate": 1.400113335851908e-05,
|
| 47 |
+
"loss": 0.1051,
|
| 48 |
+
"step": 15879
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 3.0,
|
| 52 |
+
"eval_accuracy_thresh": 0.9615273475646973,
|
| 53 |
+
"eval_f1_macro": 0.2072184242961422,
|
| 54 |
+
"eval_f1_micro": 0.27365762031219637,
|
| 55 |
+
"eval_loss": 0.1117364689707756,
|
| 56 |
+
"eval_runtime": 47.6559,
|
| 57 |
+
"eval_samples_per_second": 878.107,
|
| 58 |
+
"eval_steps_per_second": 27.447,
|
| 59 |
+
"step": 15882
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 4.0,
|
| 63 |
+
"learning_rate": 1.2001511144692106e-05,
|
| 64 |
+
"loss": 0.1008,
|
| 65 |
+
"step": 21172
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 4.0,
|
| 69 |
+
"eval_accuracy_thresh": 0.9611817002296448,
|
| 70 |
+
"eval_f1_macro": 0.22506513860997251,
|
| 71 |
+
"eval_f1_micro": 0.2893829659382017,
|
| 72 |
+
"eval_loss": 0.11295813322067261,
|
| 73 |
+
"eval_runtime": 48.0019,
|
| 74 |
+
"eval_samples_per_second": 871.779,
|
| 75 |
+
"eval_steps_per_second": 27.249,
|
| 76 |
+
"step": 21176
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 5.0,
|
| 80 |
+
"learning_rate": 1.0001888930865132e-05,
|
| 81 |
+
"loss": 0.0969,
|
| 82 |
+
"step": 26465
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 5.0,
|
| 86 |
+
"eval_accuracy_thresh": 0.9608395099639893,
|
| 87 |
+
"eval_f1_macro": 0.2055009373845579,
|
| 88 |
+
"eval_f1_micro": 0.28528452452579245,
|
| 89 |
+
"eval_loss": 0.11423922330141068,
|
| 90 |
+
"eval_runtime": 47.8236,
|
| 91 |
+
"eval_samples_per_second": 875.029,
|
| 92 |
+
"eval_steps_per_second": 27.351,
|
| 93 |
+
"step": 26470
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 6.0,
|
| 97 |
+
"learning_rate": 8.002266717038157e-06,
|
| 98 |
+
"loss": 0.0935,
|
| 99 |
+
"step": 31758
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 6.0,
|
| 103 |
+
"eval_accuracy_thresh": 0.9599553346633911,
|
| 104 |
+
"eval_f1_macro": 0.24715534791751095,
|
| 105 |
+
"eval_f1_micro": 0.31626149162404954,
|
| 106 |
+
"eval_loss": 0.11684752255678177,
|
| 107 |
+
"eval_runtime": 47.548,
|
| 108 |
+
"eval_samples_per_second": 880.101,
|
| 109 |
+
"eval_steps_per_second": 27.509,
|
| 110 |
+
"step": 31764
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 7.0,
|
| 114 |
+
"learning_rate": 6.002644503211182e-06,
|
| 115 |
+
"loss": 0.0905,
|
| 116 |
+
"step": 37051
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 7.0,
|
| 120 |
+
"eval_accuracy_thresh": 0.9595038294792175,
|
| 121 |
+
"eval_f1_macro": 0.2441714312173638,
|
| 122 |
+
"eval_f1_micro": 0.3219438566376134,
|
| 123 |
+
"eval_loss": 0.11869944632053375,
|
| 124 |
+
"eval_runtime": 47.7399,
|
| 125 |
+
"eval_samples_per_second": 876.562,
|
| 126 |
+
"eval_steps_per_second": 27.398,
|
| 127 |
+
"step": 37058
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 8.0,
|
| 131 |
+
"learning_rate": 4.003022289384209e-06,
|
| 132 |
+
"loss": 0.088,
|
| 133 |
+
"step": 42344
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 8.0,
|
| 137 |
+
"eval_accuracy_thresh": 0.9587323069572449,
|
| 138 |
+
"eval_f1_macro": 0.2548664525024525,
|
| 139 |
+
"eval_f1_micro": 0.32230417164679476,
|
| 140 |
+
"eval_loss": 0.12079060822725296,
|
| 141 |
+
"eval_runtime": 47.4927,
|
| 142 |
+
"eval_samples_per_second": 881.125,
|
| 143 |
+
"eval_steps_per_second": 27.541,
|
| 144 |
+
"step": 42352
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 9.0,
|
| 148 |
+
"learning_rate": 2.0034000755572347e-06,
|
| 149 |
+
"loss": 0.0858,
|
| 150 |
+
"step": 47637
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 9.0,
|
| 154 |
+
"eval_accuracy_thresh": 0.958522379398346,
|
| 155 |
+
"eval_f1_macro": 0.25593002791116004,
|
| 156 |
+
"eval_f1_micro": 0.326565974528547,
|
| 157 |
+
"eval_loss": 0.12228038907051086,
|
| 158 |
+
"eval_runtime": 47.6614,
|
| 159 |
+
"eval_samples_per_second": 878.006,
|
| 160 |
+
"eval_steps_per_second": 27.444,
|
| 161 |
+
"step": 47646
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 10.0,
|
| 165 |
+
"learning_rate": 3.777861730260673e-09,
|
| 166 |
+
"loss": 0.0841,
|
| 167 |
+
"step": 52930
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 10.0,
|
| 171 |
+
"eval_accuracy_thresh": 0.9580487012863159,
|
| 172 |
+
"eval_f1_macro": 0.26093237174433664,
|
| 173 |
+
"eval_f1_micro": 0.3314736272807492,
|
| 174 |
+
"eval_loss": 0.12347622960805893,
|
| 175 |
+
"eval_runtime": 47.9519,
|
| 176 |
+
"eval_samples_per_second": 872.686,
|
| 177 |
+
"eval_steps_per_second": 27.277,
|
| 178 |
+
"step": 52940
|
| 179 |
+
}
|
| 180 |
+
],
|
| 181 |
+
"max_steps": 52940,
|
| 182 |
+
"num_train_epochs": 10,
|
| 183 |
+
"total_flos": 1.603873359628872e+16,
|
| 184 |
+
"trial_name": null,
|
| 185 |
+
"trial_params": null
|
| 186 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19ab0059a48620b920f02d2501937385c3d4088206925978e56d5643d06f249b
|
| 3 |
+
size 3579
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|