Upload model 0.8
Browse files- config.json +31 -0
- merges.txt +0 -0
- params.json +1 -0
- pdn2_config.yml +191 -0
- pdn2_metadata.yml +6 -0
- pdn2_model.pt +3 -0
- special_tokens_map.json +1 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- train_args.json +41 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "allegro/herbert-large-cased",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertModel"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"directionality": "bidi",
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 1024,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 4096,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"max_position_embeddings": 514,
|
| 16 |
+
"model_type": "bert",
|
| 17 |
+
"num_attention_heads": 16,
|
| 18 |
+
"num_hidden_layers": 24,
|
| 19 |
+
"pad_token_id": 1,
|
| 20 |
+
"pooler_fc_size": 768,
|
| 21 |
+
"pooler_num_attention_heads": 12,
|
| 22 |
+
"pooler_num_fc_layers": 3,
|
| 23 |
+
"pooler_size_per_head": 128,
|
| 24 |
+
"pooler_type": "first_token_transform",
|
| 25 |
+
"position_embedding_type": "absolute",
|
| 26 |
+
"tokenizer_class": "HerbertTokenizerFast",
|
| 27 |
+
"transformers_version": "4.17.0",
|
| 28 |
+
"type_vocab_size": 2,
|
| 29 |
+
"use_cache": true,
|
| 30 |
+
"vocab_size": 50000
|
| 31 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
params.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"dropout": 0.3, "num_labels": 181, "label_list": ["B-date", "B-date#B-geogName", "B-date#B-orgName", "B-date#B-time", "B-geogName", "B-geogName#B-orgName", "B-geogName#B-persName", "B-geogName#B-persName#B-persName-addName", "B-geogName#B-persName#B-persName-forename", "B-geogName#B-persName#B-persName-surname", "B-geogName#B-persName-surname", "B-geogName#B-placeName-country", "B-geogName#B-placeName-district", "B-geogName#B-placeName-region", "B-geogName#B-placeName-settlement", "B-orgName", "B-orgName#B-geogName", "B-orgName#B-persName", "B-orgName#B-persName#B-persName-addName", "B-orgName#B-persName#B-persName-forename", "B-orgName#B-persName#B-persName-surname", "B-orgName#B-persName-surname", "B-orgName#B-placeName-bloc", "B-orgName#B-placeName-country", "B-orgName#B-placeName-district", "B-orgName#B-placeName-region", "B-orgName#B-placeName-settlement", "B-persName", "B-persName#B-geogName#B-persName-addName", "B-persName#B-persName-addName", "B-persName#B-persName-addName#B-persName-forename", "B-persName#B-persName-addName#B-placeName-settlement", "B-persName#B-persName-forename", "B-persName#B-persName-surname", "B-persName#B-persName-surname#B-placeName-settlement", "B-persName#B-placeName-settlement", "B-persName-addName", "B-persName-forename", "B-persName-surname", "B-persName-surname#B-persName-forename", "B-placeName", "B-placeName-bloc", "B-placeName-country", "B-placeName-district", "B-placeName-district#B-placeName-country", "B-placeName-district#B-placeName-settlement", "B-placeName-region", "B-placeName-region#B-geogName", "B-placeName-region#B-placeName-country", "B-placeName-region#B-placeName-settlement", "B-placeName-settlement", "B-placeName-settlement#B-placeName-country", "B-time", "I-date", "I-date#B-persName#B-persName-forename", "I-date#I-geogName", "I-date#I-orgName", "I-date#I-time", "I-geogName", "I-geogName#B-date", "I-geogName#B-orgName", "I-geogName#B-persName", "I-geogName#B-persName#B-persName-addName", "I-geogName#B-persName#B-persName-forename", "I-geogName#B-persName#B-persName-surname", "I-geogName#B-persName-forename", "I-geogName#B-persName-surname", "I-geogName#B-placeName", "I-geogName#B-placeName-country", "I-geogName#B-placeName-district", "I-geogName#B-placeName-settlement", "I-geogName#I-date", "I-geogName#I-orgName", "I-geogName#I-orgName#B-persName#B-persName-forename", "I-geogName#I-orgName#B-persName-addName", "I-geogName#I-orgName#B-placeName-country", "I-geogName#I-orgName#B-placeName-settlement", "I-geogName#I-orgName#I-persName#B-persName-surname", "I-geogName#I-persName", "I-geogName#I-persName#B-persName-addName", "I-geogName#I-persName#B-persName-forename", "I-geogName#I-persName#B-persName-surname", "I-geogName#I-persName#B-placeName-settlement", "I-geogName#I-persName#I-persName-forename", "I-geogName#I-placeName-country", "I-geogName#I-placeName-district", "I-geogName#I-placeName-settlement", "I-orgName", "I-orgName#B-date", "I-orgName#B-geogName", "I-orgName#B-persName", "I-orgName#B-persName#B-persName-addName", "I-orgName#B-persName#B-persName-forename", "I-orgName#B-persName#B-persName-surname", "I-orgName#B-persName-addName", "I-orgName#B-persName-forename", "I-orgName#B-persName-surname", "I-orgName#B-placeName", "I-orgName#B-placeName-bloc", "I-orgName#B-placeName-country", "I-orgName#B-placeName-district", "I-orgName#B-placeName-region", "I-orgName#B-placeName-settlement", "I-orgName#I-date", "I-orgName#I-geogName", "I-orgName#I-geogName#B-placeName-settlement", "I-orgName#I-persName", "I-orgName#I-persName#B-persName-addName", "I-orgName#I-persName#B-persName-forename", "I-orgName#I-persName#B-persName-surname", "I-orgName#I-persName#I-persName-addName", "I-orgName#I-persName#I-persName-forename", "I-orgName#I-placeName-bloc", "I-orgName#I-placeName-country", "I-orgName#I-placeName-district", "I-orgName#I-placeName-region", "I-orgName#I-placeName-region#B-geogName", "I-orgName#I-placeName-region#B-placeName-settlement", "I-orgName#I-placeName-settlement", "I-orgName#I-placeName-settlement#B-geogName", "I-persName", "I-persName#B-geogName", "I-persName#B-geogName#B-persName-addName", "I-persName#B-orgName", "I-persName#B-persName-addName", "I-persName#B-persName-addName#B-placeName-country", "I-persName#B-persName-addName#B-placeName-settlement", "I-persName#B-persName-forename", "I-persName#B-persName-surname", "I-persName#B-persName-surname#B-placeName-settlement", "I-persName#B-placeName-country", "I-persName#B-placeName-settlement", "I-persName#I-orgName", "I-persName#I-persName-addName", "I-persName#I-persName-addName#B-geogName", "I-persName#I-persName-addName#B-persName-forename", "I-persName#I-persName-addName#B-persName-surname", "I-persName#I-persName-addName#B-placeName-settlement", "I-persName#I-persName-forename", "I-persName#I-persName-surname", "I-persName#I-placeName-settlement", "I-persName#I-placeName-settlement#B-persName-forename", "I-persName-forename", "I-persName-surname", "I-placeName", "I-placeName#B-geogName", "I-placeName-bloc", "I-placeName-bloc#B-geogName", "I-placeName-bloc#B-placeName", "I-placeName-bloc#B-placeName-country", "I-placeName-bloc#B-placeName-settlement", "I-placeName-bloc#I-geogName", "I-placeName-country", "I-placeName-country#B-geogName", "I-placeName-country#B-orgName", "I-placeName-country#B-placeName-settlement", "I-placeName-district", "I-placeName-district#B-persName", "I-placeName-district#B-persName#B-persName-forename", "I-placeName-district#B-persName#B-persName-surname", "I-placeName-district#B-placeName-settlement", "I-placeName-district#I-persName", "I-placeName-district#I-persName#B-persName-addName", "I-placeName-district#I-persName#B-persName-forename", "I-placeName-region", "I-placeName-region#B-geogName", "I-placeName-region#B-persName", "I-placeName-region#B-placeName-district", "I-placeName-region#B-placeName-settlement", "I-placeName-region#I-persName#B-persName-forename", "I-placeName-region#I-placeName-settlement", "I-placeName-region#I-placeName-settlement#B-geogName", "I-placeName-region#I-placeName-settlement#B-placeName-country", "I-placeName-settlement", "I-placeName-settlement#B-geogName", "I-placeName-settlement#B-persName#B-persName-forename", "I-placeName-settlement#B-placeName-country", "I-placeName-settlement#B-placeName-region", "I-time", "O"]}
|
pdn2_config.yml
ADDED
|
@@ -0,0 +1,191 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
!!python/object:poldeepner2.model.hf_for_token_calssification.Pdn2ModelConfiguration
|
| 2 |
+
device: cuda:0
|
| 3 |
+
dropout_p: 0.2
|
| 4 |
+
head_init_range: 0.04
|
| 5 |
+
hidden_size: 1024
|
| 6 |
+
label_ignore_idx: 0
|
| 7 |
+
labels:
|
| 8 |
+
- B-date
|
| 9 |
+
- B-date#B-geogName
|
| 10 |
+
- B-date#B-orgName
|
| 11 |
+
- B-date#B-time
|
| 12 |
+
- B-geogName
|
| 13 |
+
- B-geogName#B-orgName
|
| 14 |
+
- B-geogName#B-persName
|
| 15 |
+
- B-geogName#B-persName#B-persName-addName
|
| 16 |
+
- B-geogName#B-persName#B-persName-forename
|
| 17 |
+
- B-geogName#B-persName#B-persName-surname
|
| 18 |
+
- B-geogName#B-persName-surname
|
| 19 |
+
- B-geogName#B-placeName-country
|
| 20 |
+
- B-geogName#B-placeName-district
|
| 21 |
+
- B-geogName#B-placeName-region
|
| 22 |
+
- B-geogName#B-placeName-settlement
|
| 23 |
+
- B-orgName
|
| 24 |
+
- B-orgName#B-geogName
|
| 25 |
+
- B-orgName#B-persName
|
| 26 |
+
- B-orgName#B-persName#B-persName-addName
|
| 27 |
+
- B-orgName#B-persName#B-persName-forename
|
| 28 |
+
- B-orgName#B-persName#B-persName-surname
|
| 29 |
+
- B-orgName#B-persName-surname
|
| 30 |
+
- B-orgName#B-placeName-bloc
|
| 31 |
+
- B-orgName#B-placeName-country
|
| 32 |
+
- B-orgName#B-placeName-district
|
| 33 |
+
- B-orgName#B-placeName-region
|
| 34 |
+
- B-orgName#B-placeName-settlement
|
| 35 |
+
- B-persName
|
| 36 |
+
- B-persName#B-geogName#B-persName-addName
|
| 37 |
+
- B-persName#B-persName-addName
|
| 38 |
+
- B-persName#B-persName-addName#B-persName-forename
|
| 39 |
+
- B-persName#B-persName-addName#B-placeName-settlement
|
| 40 |
+
- B-persName#B-persName-forename
|
| 41 |
+
- B-persName#B-persName-surname
|
| 42 |
+
- B-persName#B-persName-surname#B-placeName-settlement
|
| 43 |
+
- B-persName#B-placeName-settlement
|
| 44 |
+
- B-persName-addName
|
| 45 |
+
- B-persName-forename
|
| 46 |
+
- B-persName-surname
|
| 47 |
+
- B-persName-surname#B-persName-forename
|
| 48 |
+
- B-placeName
|
| 49 |
+
- B-placeName-bloc
|
| 50 |
+
- B-placeName-country
|
| 51 |
+
- B-placeName-district
|
| 52 |
+
- B-placeName-district#B-placeName-country
|
| 53 |
+
- B-placeName-district#B-placeName-settlement
|
| 54 |
+
- B-placeName-region
|
| 55 |
+
- B-placeName-region#B-geogName
|
| 56 |
+
- B-placeName-region#B-placeName-country
|
| 57 |
+
- B-placeName-region#B-placeName-settlement
|
| 58 |
+
- B-placeName-settlement
|
| 59 |
+
- B-placeName-settlement#B-placeName-country
|
| 60 |
+
- B-time
|
| 61 |
+
- I-date
|
| 62 |
+
- I-date#B-persName#B-persName-forename
|
| 63 |
+
- I-date#I-geogName
|
| 64 |
+
- I-date#I-orgName
|
| 65 |
+
- I-date#I-time
|
| 66 |
+
- I-geogName
|
| 67 |
+
- I-geogName#B-date
|
| 68 |
+
- I-geogName#B-orgName
|
| 69 |
+
- I-geogName#B-persName
|
| 70 |
+
- I-geogName#B-persName#B-persName-addName
|
| 71 |
+
- I-geogName#B-persName#B-persName-forename
|
| 72 |
+
- I-geogName#B-persName#B-persName-surname
|
| 73 |
+
- I-geogName#B-persName-forename
|
| 74 |
+
- I-geogName#B-persName-surname
|
| 75 |
+
- I-geogName#B-placeName
|
| 76 |
+
- I-geogName#B-placeName-country
|
| 77 |
+
- I-geogName#B-placeName-district
|
| 78 |
+
- I-geogName#B-placeName-settlement
|
| 79 |
+
- I-geogName#I-date
|
| 80 |
+
- I-geogName#I-orgName
|
| 81 |
+
- I-geogName#I-orgName#B-persName#B-persName-forename
|
| 82 |
+
- I-geogName#I-orgName#B-persName-addName
|
| 83 |
+
- I-geogName#I-orgName#B-placeName-country
|
| 84 |
+
- I-geogName#I-orgName#B-placeName-settlement
|
| 85 |
+
- I-geogName#I-orgName#I-persName#B-persName-surname
|
| 86 |
+
- I-geogName#I-persName
|
| 87 |
+
- I-geogName#I-persName#B-persName-addName
|
| 88 |
+
- I-geogName#I-persName#B-persName-forename
|
| 89 |
+
- I-geogName#I-persName#B-persName-surname
|
| 90 |
+
- I-geogName#I-persName#B-placeName-settlement
|
| 91 |
+
- I-geogName#I-persName#I-persName-forename
|
| 92 |
+
- I-geogName#I-placeName-country
|
| 93 |
+
- I-geogName#I-placeName-district
|
| 94 |
+
- I-geogName#I-placeName-settlement
|
| 95 |
+
- I-orgName
|
| 96 |
+
- I-orgName#B-date
|
| 97 |
+
- I-orgName#B-geogName
|
| 98 |
+
- I-orgName#B-persName
|
| 99 |
+
- I-orgName#B-persName#B-persName-addName
|
| 100 |
+
- I-orgName#B-persName#B-persName-forename
|
| 101 |
+
- I-orgName#B-persName#B-persName-surname
|
| 102 |
+
- I-orgName#B-persName-addName
|
| 103 |
+
- I-orgName#B-persName-forename
|
| 104 |
+
- I-orgName#B-persName-surname
|
| 105 |
+
- I-orgName#B-placeName
|
| 106 |
+
- I-orgName#B-placeName-bloc
|
| 107 |
+
- I-orgName#B-placeName-country
|
| 108 |
+
- I-orgName#B-placeName-district
|
| 109 |
+
- I-orgName#B-placeName-region
|
| 110 |
+
- I-orgName#B-placeName-settlement
|
| 111 |
+
- I-orgName#I-date
|
| 112 |
+
- I-orgName#I-geogName
|
| 113 |
+
- I-orgName#I-geogName#B-placeName-settlement
|
| 114 |
+
- I-orgName#I-persName
|
| 115 |
+
- I-orgName#I-persName#B-persName-addName
|
| 116 |
+
- I-orgName#I-persName#B-persName-forename
|
| 117 |
+
- I-orgName#I-persName#B-persName-surname
|
| 118 |
+
- I-orgName#I-persName#I-persName-addName
|
| 119 |
+
- I-orgName#I-persName#I-persName-forename
|
| 120 |
+
- I-orgName#I-placeName-bloc
|
| 121 |
+
- I-orgName#I-placeName-country
|
| 122 |
+
- I-orgName#I-placeName-district
|
| 123 |
+
- I-orgName#I-placeName-region
|
| 124 |
+
- I-orgName#I-placeName-region#B-geogName
|
| 125 |
+
- I-orgName#I-placeName-region#B-placeName-settlement
|
| 126 |
+
- I-orgName#I-placeName-settlement
|
| 127 |
+
- I-orgName#I-placeName-settlement#B-geogName
|
| 128 |
+
- I-persName
|
| 129 |
+
- I-persName#B-geogName
|
| 130 |
+
- I-persName#B-geogName#B-persName-addName
|
| 131 |
+
- I-persName#B-orgName
|
| 132 |
+
- I-persName#B-persName-addName
|
| 133 |
+
- I-persName#B-persName-addName#B-placeName-country
|
| 134 |
+
- I-persName#B-persName-addName#B-placeName-settlement
|
| 135 |
+
- I-persName#B-persName-forename
|
| 136 |
+
- I-persName#B-persName-surname
|
| 137 |
+
- I-persName#B-persName-surname#B-placeName-settlement
|
| 138 |
+
- I-persName#B-placeName-country
|
| 139 |
+
- I-persName#B-placeName-settlement
|
| 140 |
+
- I-persName#I-orgName
|
| 141 |
+
- I-persName#I-persName-addName
|
| 142 |
+
- I-persName#I-persName-addName#B-geogName
|
| 143 |
+
- I-persName#I-persName-addName#B-persName-forename
|
| 144 |
+
- I-persName#I-persName-addName#B-persName-surname
|
| 145 |
+
- I-persName#I-persName-addName#B-placeName-settlement
|
| 146 |
+
- I-persName#I-persName-forename
|
| 147 |
+
- I-persName#I-persName-surname
|
| 148 |
+
- I-persName#I-placeName-settlement
|
| 149 |
+
- I-persName#I-placeName-settlement#B-persName-forename
|
| 150 |
+
- I-persName-forename
|
| 151 |
+
- I-persName-surname
|
| 152 |
+
- I-placeName
|
| 153 |
+
- I-placeName#B-geogName
|
| 154 |
+
- I-placeName-bloc
|
| 155 |
+
- I-placeName-bloc#B-geogName
|
| 156 |
+
- I-placeName-bloc#B-placeName
|
| 157 |
+
- I-placeName-bloc#B-placeName-country
|
| 158 |
+
- I-placeName-bloc#B-placeName-settlement
|
| 159 |
+
- I-placeName-bloc#I-geogName
|
| 160 |
+
- I-placeName-country
|
| 161 |
+
- I-placeName-country#B-geogName
|
| 162 |
+
- I-placeName-country#B-orgName
|
| 163 |
+
- I-placeName-country#B-placeName-settlement
|
| 164 |
+
- I-placeName-district
|
| 165 |
+
- I-placeName-district#B-persName
|
| 166 |
+
- I-placeName-district#B-persName#B-persName-forename
|
| 167 |
+
- I-placeName-district#B-persName#B-persName-surname
|
| 168 |
+
- I-placeName-district#B-placeName-settlement
|
| 169 |
+
- I-placeName-district#I-persName
|
| 170 |
+
- I-placeName-district#I-persName#B-persName-addName
|
| 171 |
+
- I-placeName-district#I-persName#B-persName-forename
|
| 172 |
+
- I-placeName-region
|
| 173 |
+
- I-placeName-region#B-geogName
|
| 174 |
+
- I-placeName-region#B-persName
|
| 175 |
+
- I-placeName-region#B-placeName-district
|
| 176 |
+
- I-placeName-region#B-placeName-settlement
|
| 177 |
+
- I-placeName-region#I-persName#B-persName-forename
|
| 178 |
+
- I-placeName-region#I-placeName-settlement
|
| 179 |
+
- I-placeName-region#I-placeName-settlement#B-geogName
|
| 180 |
+
- I-placeName-region#I-placeName-settlement#B-placeName-country
|
| 181 |
+
- I-placeName-settlement
|
| 182 |
+
- I-placeName-settlement#B-geogName
|
| 183 |
+
- I-placeName-settlement#B-persName#B-persName-forename
|
| 184 |
+
- I-placeName-settlement#B-placeName-country
|
| 185 |
+
- I-placeName-settlement#B-placeName-region
|
| 186 |
+
- I-time
|
| 187 |
+
- O
|
| 188 |
+
max_seq_length: 256
|
| 189 |
+
output_top_k: 1
|
| 190 |
+
seed: 101
|
| 191 |
+
sequence_generator: merged
|
pdn2_metadata.yml
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
author: Michał Marcińczuk
|
| 2 |
+
contact: [email protected]
|
| 3 |
+
description: Model trained on the NKJP corpus recognizes 12 types of nested entities.
|
| 4 |
+
reference: https://www.sciencedirect.com/science/article/pii/S1877050921015179
|
| 5 |
+
score: 92.1 on the PolEval 2018 NER dataset
|
| 6 |
+
|
pdn2_model.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b11ab8d10db1b7288e822f4520378722d260605f9b4eb4298cff1f9d197cba5
|
| 3 |
+
size 1425398669
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"bos_token": "<s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": "<mask>"}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "sep_token": "</s>", "do_lowercase_and_remove_accent": false, "bos_token": "<s>", "additional_special_tokens": [], "model_max_length": 512, "special_tokens_map_file": "/home/czuk/.cache/huggingface/transformers/7e8fe8852a1ff7e03195cb41fac16af837f8c14a34a61850b02a7395eb294f00.b8e113717eb1828d09e47de853cf49c8fad05ebdce24df2614cd942dc23e2a77", "name_or_path": "allegro/herbert-large-cased", "lang2id": null, "id2lang": null, "tokenizer_class": "HerbertTokenizer"}
|
train_args.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"data_train": [
|
| 3 |
+
"data/nkjp-nested-ttt/train.txt",
|
| 4 |
+
"data/nkjp-nested-ttt/valid.txt",
|
| 5 |
+
"data/nkjp-nested-ttt/test.txt"
|
| 6 |
+
],
|
| 7 |
+
"data_tune": [
|
| 8 |
+
"data/nkjp-nested-ttt/valid.txt"
|
| 9 |
+
],
|
| 10 |
+
"data_test": [
|
| 11 |
+
"data/nkjp-nested-ttt/test.txt"
|
| 12 |
+
],
|
| 13 |
+
"pretrained_path": "allegro/herbert-large-cased",
|
| 14 |
+
"output_dir": "../poldeepner2_models/dev/nkjp_full/model_nkjp_full_union_256_101_v_003",
|
| 15 |
+
"cache_dir": "",
|
| 16 |
+
"device": "cuda:0",
|
| 17 |
+
"max_seq_length": 256,
|
| 18 |
+
"do_eval": false,
|
| 19 |
+
"do_lower_case": false,
|
| 20 |
+
"train_batch_size": 16,
|
| 21 |
+
"eval_batch_size": 16,
|
| 22 |
+
"learning_rate": 5e-06,
|
| 23 |
+
"num_train_epochs": 20,
|
| 24 |
+
"warmup_proportion": 0.0,
|
| 25 |
+
"weight_decay": 0.01,
|
| 26 |
+
"adam_epsilon": 1e-08,
|
| 27 |
+
"max_grad_norm": 1.0,
|
| 28 |
+
"seed": 101,
|
| 29 |
+
"gradient_accumulation_steps": 1,
|
| 30 |
+
"fp16": false,
|
| 31 |
+
"fp16_opt_level": "O1",
|
| 32 |
+
"loss_scale": 0,
|
| 33 |
+
"dropout": 0.2,
|
| 34 |
+
"freeze_model": false,
|
| 35 |
+
"epoch_save_model": true,
|
| 36 |
+
"sequence_generator": "union",
|
| 37 |
+
"sequence_generator_for_eval": "context-window",
|
| 38 |
+
"training_mix": false,
|
| 39 |
+
"wandb": "nkjp_full",
|
| 40 |
+
"hidden_size": 1024
|
| 41 |
+
}
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|