Token Classification
Transformers
PyTorch
English
roberta
HHansi commited on
Commit
6ce7404
·
1 Parent(s): 4b1b595
config.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "models/lm/roberta-large-sentence/best_model",
3
+ "architectures": [
4
+ "RobertaForTokenClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "gelu",
11
+ "hidden_dropout_prob": 0.1,
12
+ "hidden_size": 1024,
13
+ "id2label": {
14
+ "0": "LABEL_0",
15
+ "1": "LABEL_1",
16
+ "2": "LABEL_2",
17
+ "3": "LABEL_3",
18
+ "4": "LABEL_4",
19
+ "5": "LABEL_5",
20
+ "6": "LABEL_6",
21
+ "7": "LABEL_7",
22
+ "8": "LABEL_8"
23
+ },
24
+ "initializer_range": 0.02,
25
+ "intermediate_size": 4096,
26
+ "label2id": {
27
+ "LABEL_0": 0,
28
+ "LABEL_1": 1,
29
+ "LABEL_2": 2,
30
+ "LABEL_3": 3,
31
+ "LABEL_4": 4,
32
+ "LABEL_5": 5,
33
+ "LABEL_6": 6,
34
+ "LABEL_7": 7,
35
+ "LABEL_8": 8
36
+ },
37
+ "layer_norm_eps": 1e-05,
38
+ "max_position_embeddings": 514,
39
+ "model_type": "roberta",
40
+ "num_attention_heads": 16,
41
+ "num_hidden_layers": 24,
42
+ "pad_token_id": 1,
43
+ "position_embedding_type": "absolute",
44
+ "torch_dtype": "float32",
45
+ "transformers_version": "4.16.2",
46
+ "type_vocab_size": 1,
47
+ "use_cache": true,
48
+ "vocab_size": 50265
49
+ }
eval_results.txt ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ cls_report = precision recall f1-score support
2
+
3
+ object 0.7910 0.7260 0.7571 146
4
+ property 0.5593 0.6875 0.6168 48
5
+ quality 0.5822 0.6115 0.5965 139
6
+ value 0.8889 0.8571 0.8727 28
7
+
8
+ micro avg 0.6776 0.6870 0.6823 361
9
+ macro avg 0.7054 0.7205 0.7108 361
10
+ weighted avg 0.6874 0.6870 0.6856 361
11
+
12
+ eval_loss = 0.4966140389442444
13
+ f1_score = 0.7107959469617118
14
+ f1_score_strict = 0.7351027636388913
15
+ precision = 0.7053618699321287
16
+ precision_strict = 0.7654965648386702
17
+ recall = 0.7205452614425094
18
+ recall_strict = 0.7100274770398944
19
+ strict_cls_report = precision recall f1-score support
20
+
21
+ object 0.7955 0.7192 0.7554 146
22
+ property 0.6154 0.6667 0.6400 48
23
+ quality 0.7281 0.5971 0.6561 139
24
+ value 0.9231 0.8571 0.8889 28
25
+
26
+ micro avg 0.7531 0.6759 0.7124 361
27
+ macro avg 0.7655 0.7100 0.7351 361
28
+ weighted avg 0.7555 0.6759 0.7122 361
29
+
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model_args.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"adam_epsilon": 1e-08, "best_model_dir": "outputs/best_model", "cache_dir": "temp/cache_dir", "config": {}, "custom_layer_parameters": [], "custom_parameter_groups": [], "dataloader_num_workers": 1, "do_lower_case": false, "dynamic_quantize": false, "early_stopping_consider_epochs": false, "early_stopping_delta": 0, "early_stopping_metric": "eval_loss", "early_stopping_metric_minimize": true, "early_stopping_patience": 10, "encoding": null, "eval_batch_size": 512, "evaluate_during_training": true, "evaluate_during_training_silent": true, "evaluate_during_training_steps": 8, "evaluate_during_training_verbose": true, "evaluate_each_epoch": true, "fp16": false, "gradient_accumulation_steps": 1, "learning_rate": 0.0001, "local_rank": -1, "logging_steps": 8, "manual_seed": 157, "max_grad_norm": 1.0, "max_seq_length": 128, "model_name": "models/lm/roberta-large-sentence/best_model", "model_type": "roberta", "multiprocessing_chunksize": 500, "n_gpu": 1, "no_cache": false, "no_save": false, "not_saved_args": [], "num_train_epochs": 5, "output_dir": "temp/outputs", "overwrite_output_dir": true, "process_count": 70, "quantized_model": false, "reprocess_input_data": true, "save_best_model": true, "save_eval_checkpoints": false, "save_model_every_epoch": false, "save_optimizer_and_scheduler": true, "save_recent_only": true, "save_steps": 8, "silent": false, "tensorboard_dir": null, "thread_count": null, "train_batch_size": 16, "train_custom_parameters_only": false, "use_cached_eval_features": false, "use_early_stopping": true, "use_multiprocessing": false, "wandb_kwargs": {"name": "roberta-large-1e-4-lms"}, "wandb_project": "ner", "warmup_ratio": 0.06, "warmup_steps": 12, "weight_decay": 0, "skip_special_tokens": true, "model_class": "LanguageModelingModel", "classification_report": true, "labels_list": ["O", "B-quality", "B-property", "I-property", "I-quality", "B-object", "I-object", "B-value", "I-value"], "lazy_loading": false, "lazy_loading_start_line": 0, "onnx": false}
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17b28279ef43261f61690f09c5345c498afe99d8d22b9aaa741a87133e4306d3
3
+ size 2834784197
predictions.csv ADDED
The diff for this file is too large to render. See raw diff
 
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:658eda7d14db6b96c8796b9a0c4b619f87dff4d58e8729552fcaafd81fc6592a
3
+ size 1417443693
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f57420b1f167a7ba6f460977978bff268e299d1bc2c9f8665e430954332489a5
3
+ size 627
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"bos_token": {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "eos_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "unk_token": {"content": "<unk>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "sep_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "pad_token": {"content": "<pad>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "cls_token": {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true}}
test_eval.txt ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Default classification report:
2
+ precision recall f1-score support
3
+
4
+ object 0.5031 0.5110 0.5070 317
5
+ property 0.4348 0.4082 0.4211 98
6
+ quality 0.3089 0.3849 0.3427 317
7
+ value 0.4773 0.4375 0.4565 48
8
+
9
+ micro avg 0.4045 0.4423 0.4225 780
10
+ macro avg 0.4310 0.4354 0.4318 780
11
+ weighted avg 0.4140 0.4423 0.4263 780
12
+
13
+ Strict classification report:
14
+ precision recall f1-score support
15
+
16
+ object 0.5212 0.5047 0.5128 317
17
+ property 0.4750 0.3878 0.4270 98
18
+ quality 0.3805 0.3817 0.3811 317
19
+ value 0.4773 0.4375 0.4565 48
20
+
21
+ micro avg 0.4539 0.4359 0.4447 780
22
+ macro avg 0.4635 0.4279 0.4444 780
23
+ weighted avg 0.4555 0.4359 0.4450 780
24
+
25
+ precision(macro) = 0.43100542138104053
26
+ recall(macro) = 0.4353905797334707
27
+ f1_score(macro) = 0.43182831336099803
28
+ precision_strict(macro) = 0.4634871275908244
29
+ recall_strict(macro) = 0.42792260831777507
30
+ f1_score_strict(macro) = 0.4443527265726259
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"errors": "replace", "unk_token": {"content": "<unk>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "bos_token": {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "eos_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "add_prefix_space": false, "sep_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "cls_token": {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "pad_token": {"content": "<pad>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "model_max_length": 512, "special_tokens_map_file": null, "tokenizer_file": "cache_dir/e16a2590deb9e6d73711d6e05bf27d832fa8c1162d807222e043ca650a556964.fc9576039592f026ad76a1c231b89aee8668488c671dfbe6616bab2ed298d730", "name_or_path": "models/lm/roberta-large-sentence/best_model", "do_lower_case": false, "tokenizer_class": "RobertaTokenizer"}
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e0b295df5aa7b57c6078ef5626a6bd6641ab8c4cd3cf70d05112d246ced0328
3
+ size 3323
training_progress_scores.csv ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ global_step,precision,recall,f1_score,train_loss,eval_loss,precision_strict,recall_strict,f1_score_strict
2
+ 8,0.1793594306049822,0.06448950428698137,0.04921948912015137,1.5140964984893799,1.365554928779602,0.24375,0.015928353207844685,0.02928861354411451
3
+ 16,0.27487823717467336,0.23606166436713644,0.24159583510412241,0.7765843868255615,0.8628915548324585,0.3884543178973717,0.2128677195230117,0.2649989616502838
4
+ 24,0.4900657731447009,0.5622217979041425,0.507872039793361,1.0288734436035156,0.770804226398468,0.6073207468384634,0.5308557948162018,0.5523266964009107
5
+ 32,0.5216763001037195,0.5887565876848424,0.5508478272530071,0.928479790687561,0.6206213235855103,0.5859954478798086,0.551849127351739,0.5639903065561519
6
+ 39,0.5562304848230443,0.6141191010009996,0.5804447581043326,0.7451213002204895,0.5828236937522888,0.6197552447552448,0.5999030765602782,0.6072766354122735
7
+ 40,0.5714834028692723,0.6181051489532444,0.591830765156568,0.4681490659713745,0.5672286748886108,0.6406362932369644,0.5952561336449431,0.6163127664183048
8
+ 48,0.6170478957133092,0.6390835961367892,0.6276718504300983,0.4824497103691101,0.6826385855674744,0.6841313887454827,0.6196741237147269,0.6477696773129918
9
+ 56,0.6497260955105653,0.614773836508093,0.6303793233791719,0.542269766330719,0.6344236135482788,0.7032428248523183,0.5814070367971768,0.6361820588878431
10
+ 64,0.6868466464153796,0.703113927644577,0.6944071540259829,0.312347948551178,0.5910742878913879,0.7363698331743445,0.6906102319709789,0.7117328058311433
11
+ 72,0.6757663604727363,0.7030919295874455,0.6863179255285954,0.38219600915908813,0.5123444199562073,0.7352152014652015,0.6925592598328383,0.7120016699261982
12
+ 78,0.6914870556174904,0.7015300675311023,0.6950345884620526,0.2568081021308899,0.5217363238334656,0.7341726007650682,0.6821919421410034,0.7055625863750838
13
+ 80,0.6502093538991129,0.7043659837296503,0.6743122693942367,0.15562495589256287,0.508587658405304,0.6944144529745935,0.6921358705599121,0.6926649276649276
14
+ 88,0.7043623714994682,0.7243493121354188,0.7125199535913822,0.241601824760437,0.607262909412384,0.7473760986864311,0.7103206378146016,0.7272649572649572
15
+ 96,0.6847560140669942,0.7054078850422597,0.6926461501154404,0.160390242934227,0.5868411660194397,0.7281798830940053,0.6910193958336266,0.707217882974078
16
+ 104,0.6420330503932787,0.6867401871066747,0.6627260639504462,0.29913395643234253,0.5807334184646606,0.703917837626854,0.6867401871066747,0.6931002670487625
17
+ 112,0.7053618699321287,0.7205452614425094,0.7107959469617118,0.23882241547107697,0.4966140389442444,0.7654965648386702,0.7100274770398944,0.7351027636388913
18
+ 117,0.7251608851278264,0.7204025673785824,0.7221109995369133,0.06621954590082169,0.5260559916496277,0.7841402842053546,0.7186040062275033,0.7484020810107767
19
+ 120,0.7313743604066185,0.7312801666690132,0.7302708579794063,0.17493538558483124,0.5740806460380554,0.7754898906867411,0.7258844832157757,0.7492376775271512
20
+ 128,0.6806934760498726,0.7329847227892833,0.7043812512044878,0.03242364525794983,0.6142690777778625,0.7282000309038179,0.729387600487125,0.7273024468837411
21
+ 136,0.6915230920748747,0.7260640606888267,0.706787603930461,0.16539587080478668,0.6960902214050293,0.733842223667805,0.7137477151351326,0.7229931274041175
22
+ 144,0.6915522800552741,0.7310285822222847,0.7088708991232194,0.0528966449201107,0.6215918660163879,0.7383108735892272,0.7276039246880383,0.7325356207685868
23
+ 152,0.7252524630541872,0.7297105319658168,0.7257349295549856,0.08879994601011276,0.609069287776947,0.762671568627451,0.720803958594377,0.7391175524562513
24
+ 156,0.7168708945380323,0.7388978005462558,0.7254317332213958,0.1931147575378418,0.6626339554786682,0.7466899518032379,0.7318760207098509,0.7371025994939039
25
+ 160,0.718127635576348,0.7371854717791325,0.7250104618331302,0.08354368060827255,0.6829016804695129,0.7426523014684789,0.7335883494769742,0.7359229474887837
26
+ 168,0.7270477016717618,0.7508692165641263,0.7373424194480808,0.010109204798936844,0.6239851117134094,0.7647082963984373,0.7472720942619681,0.7552509566762334
27
+ 176,0.7316153069999484,0.7546388036341963,0.7412630672638264,0.08013226836919785,0.6381521821022034,0.7634949491727228,0.7493293525649148,0.7553204907694072
28
+ 184,0.7284983399792717,0.7546388036341963,0.7395185152890282,0.01630091853439808,0.6346596479415894,0.7589877289123732,0.7528402424831172,0.7547108032028385
29
+ 192,0.7092296728400774,0.7564373647852755,0.7301968716381126,0.05021103471517563,0.625979483127594,0.7455346931511916,0.7510416813320381,0.7465768143494647
30
+ 195,0.7075740437010045,0.7546388036341963,0.7284727337070782,0.025623565539717674,0.6265556216239929,0.7450078077558852,0.749243120180959,0.7453829788816981
vocab.json ADDED
The diff for this file is too large to render. See raw diff