Update spaCy pipeline
Browse files- .gitattributes +5 -0
- README.md +56 -0
- config.cfg +147 -0
- entity_punctuation_removal.py +34 -0
- fr_arches_ner-any-py3-none-any.whl +3 -0
- meta.json +140 -0
- ner/cfg +13 -0
- ner/model +0 -0
- ner/moves +1 -0
- repartition.txt +31 -0
- resultats_post_ponct.json +87 -0
- tok2vec/cfg +3 -0
- tok2vec/model +3 -0
- tokenizer +0 -0
- vocab/key2row +3 -0
- vocab/lookups.bin +3 -0
- vocab/strings.json +3 -0
- vocab/vectors +3 -0
- vocab/vectors.cfg +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,8 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
fr_arches_ner-any-py3-none-any.whl filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
tok2vec/model filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
vocab/key2row filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
vocab/strings.json filter=lfs diff=lfs merge=lfs -text
|
| 40 |
+
vocab/vectors filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
tags:
|
| 3 |
+
- spacy
|
| 4 |
+
- token-classification
|
| 5 |
+
language:
|
| 6 |
+
- fr
|
| 7 |
+
model-index:
|
| 8 |
+
- name: fr_arches_ner
|
| 9 |
+
results:
|
| 10 |
+
- task:
|
| 11 |
+
name: NER
|
| 12 |
+
type: token-classification
|
| 13 |
+
metrics:
|
| 14 |
+
- name: NER Precision
|
| 15 |
+
type: precision
|
| 16 |
+
value: 0.6778376222
|
| 17 |
+
- name: NER Recall
|
| 18 |
+
type: recall
|
| 19 |
+
value: 0.7156697557
|
| 20 |
+
- name: NER F Score
|
| 21 |
+
type: f_score
|
| 22 |
+
value: 0.6962401393
|
| 23 |
+
---
|
| 24 |
+
| Feature | Description |
|
| 25 |
+
| --- | --- |
|
| 26 |
+
| **Name** | `fr_arches_ner` |
|
| 27 |
+
| **Version** | `0.0.0` |
|
| 28 |
+
| **spaCy** | `>=3.6.1,<3.7.0` |
|
| 29 |
+
| **Default Pipeline** | `tok2vec`, `ner`, `entity_punctuation_removal` |
|
| 30 |
+
| **Components** | `tok2vec`, `ner`, `entity_punctuation_removal` |
|
| 31 |
+
| **Vectors** | 500000 keys, 500000 unique vectors (300 dimensions) |
|
| 32 |
+
| **Sources** | n/a |
|
| 33 |
+
| **License** | n/a |
|
| 34 |
+
| **Author** | [n/a]() |
|
| 35 |
+
|
| 36 |
+
### Label Scheme
|
| 37 |
+
|
| 38 |
+
<details>
|
| 39 |
+
|
| 40 |
+
<summary>View label scheme (15 labels for 1 components)</summary>
|
| 41 |
+
|
| 42 |
+
| Component | Labels |
|
| 43 |
+
| --- | --- |
|
| 44 |
+
| **`ner`** | `CHRONOLOGIE`, `DECOR`, `EDIFICE`, `ESPECE`, `GPE`, `ID`, `LIEUDIT_SITE`, `LOC`, `MATERIAU`, `MOBILIER`, `ORG`, `PERSONNE`, `PEUPLE_CULTURE`, `STRUCTURE`, `TECHNIQUE_STYLE` |
|
| 45 |
+
|
| 46 |
+
</details>
|
| 47 |
+
|
| 48 |
+
### Accuracy
|
| 49 |
+
|
| 50 |
+
| Type | Score |
|
| 51 |
+
| --- | --- |
|
| 52 |
+
| `ENTS_F` | 69.62 |
|
| 53 |
+
| `ENTS_P` | 67.78 |
|
| 54 |
+
| `ENTS_R` | 71.57 |
|
| 55 |
+
| `TOK2VEC_LOSS` | 63436.09 |
|
| 56 |
+
| `NER_LOSS` | 246059.83 |
|
config.cfg
ADDED
|
@@ -0,0 +1,147 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[paths]
|
| 2 |
+
train = "train.spacy"
|
| 3 |
+
dev = "dev.spacy"
|
| 4 |
+
vectors = "fr_core_news_lg"
|
| 5 |
+
init_tok2vec = null
|
| 6 |
+
|
| 7 |
+
[system]
|
| 8 |
+
gpu_allocator = null
|
| 9 |
+
seed = 0
|
| 10 |
+
|
| 11 |
+
[nlp]
|
| 12 |
+
lang = "fr"
|
| 13 |
+
pipeline = ["tok2vec","ner","entity_punctuation_removal"]
|
| 14 |
+
batch_size = 1000
|
| 15 |
+
disabled = []
|
| 16 |
+
before_creation = null
|
| 17 |
+
after_creation = null
|
| 18 |
+
after_pipeline_creation = null
|
| 19 |
+
tokenizer = {"@tokenizers":"spacy.Tokenizer.v1"}
|
| 20 |
+
|
| 21 |
+
[components]
|
| 22 |
+
|
| 23 |
+
[components.entity_punctuation_removal]
|
| 24 |
+
factory = "entity_punctuation_removal"
|
| 25 |
+
|
| 26 |
+
[components.ner]
|
| 27 |
+
factory = "ner"
|
| 28 |
+
incorrect_spans_key = null
|
| 29 |
+
moves = null
|
| 30 |
+
scorer = {"@scorers":"spacy.ner_scorer.v1"}
|
| 31 |
+
update_with_oracle_cut_size = 100
|
| 32 |
+
|
| 33 |
+
[components.ner.model]
|
| 34 |
+
@architectures = "spacy.TransitionBasedParser.v2"
|
| 35 |
+
state_type = "ner"
|
| 36 |
+
extra_state_tokens = false
|
| 37 |
+
hidden_width = 64
|
| 38 |
+
maxout_pieces = 2
|
| 39 |
+
use_upper = true
|
| 40 |
+
nO = null
|
| 41 |
+
|
| 42 |
+
[components.ner.model.tok2vec]
|
| 43 |
+
@architectures = "spacy.Tok2VecListener.v1"
|
| 44 |
+
width = ${components.tok2vec.model.encode.width}
|
| 45 |
+
upstream = "*"
|
| 46 |
+
|
| 47 |
+
[components.tok2vec]
|
| 48 |
+
factory = "tok2vec"
|
| 49 |
+
|
| 50 |
+
[components.tok2vec.model]
|
| 51 |
+
@architectures = "spacy.Tok2Vec.v2"
|
| 52 |
+
|
| 53 |
+
[components.tok2vec.model.embed]
|
| 54 |
+
@architectures = "spacy.MultiHashEmbed.v2"
|
| 55 |
+
width = ${components.tok2vec.model.encode.width}
|
| 56 |
+
attrs = ["NORM","PREFIX","SUFFIX","SHAPE"]
|
| 57 |
+
rows = [5000,1000,2500,2500]
|
| 58 |
+
include_static_vectors = true
|
| 59 |
+
|
| 60 |
+
[components.tok2vec.model.encode]
|
| 61 |
+
@architectures = "spacy.MaxoutWindowEncoder.v2"
|
| 62 |
+
width = 256
|
| 63 |
+
depth = 8
|
| 64 |
+
window_size = 1
|
| 65 |
+
maxout_pieces = 3
|
| 66 |
+
|
| 67 |
+
[corpora]
|
| 68 |
+
|
| 69 |
+
[corpora.dev]
|
| 70 |
+
@readers = "spacy.Corpus.v1"
|
| 71 |
+
path = ${paths.dev}
|
| 72 |
+
max_length = 0
|
| 73 |
+
gold_preproc = false
|
| 74 |
+
limit = 0
|
| 75 |
+
augmenter = null
|
| 76 |
+
|
| 77 |
+
[corpora.train]
|
| 78 |
+
@readers = "spacy.Corpus.v1"
|
| 79 |
+
path = ${paths.train}
|
| 80 |
+
max_length = 0
|
| 81 |
+
gold_preproc = false
|
| 82 |
+
limit = 0
|
| 83 |
+
augmenter = null
|
| 84 |
+
|
| 85 |
+
[training]
|
| 86 |
+
dev_corpus = "corpora.dev"
|
| 87 |
+
train_corpus = "corpora.train"
|
| 88 |
+
seed = ${system.seed}
|
| 89 |
+
gpu_allocator = ${system.gpu_allocator}
|
| 90 |
+
dropout = 0.1
|
| 91 |
+
accumulate_gradient = 1
|
| 92 |
+
patience = 1600
|
| 93 |
+
max_epochs = 0
|
| 94 |
+
max_steps = 20000
|
| 95 |
+
eval_frequency = 200
|
| 96 |
+
frozen_components = []
|
| 97 |
+
annotating_components = []
|
| 98 |
+
before_to_disk = null
|
| 99 |
+
before_update = null
|
| 100 |
+
|
| 101 |
+
[training.batcher]
|
| 102 |
+
@batchers = "spacy.batch_by_words.v1"
|
| 103 |
+
discard_oversize = false
|
| 104 |
+
tolerance = 0.2
|
| 105 |
+
get_length = null
|
| 106 |
+
|
| 107 |
+
[training.batcher.size]
|
| 108 |
+
@schedules = "compounding.v1"
|
| 109 |
+
start = 100
|
| 110 |
+
stop = 1000
|
| 111 |
+
compound = 1.001
|
| 112 |
+
t = 0.0
|
| 113 |
+
|
| 114 |
+
[training.logger]
|
| 115 |
+
@loggers = "spacy.ConsoleLogger.v1"
|
| 116 |
+
progress_bar = false
|
| 117 |
+
|
| 118 |
+
[training.optimizer]
|
| 119 |
+
@optimizers = "Adam.v1"
|
| 120 |
+
beta1 = 0.9
|
| 121 |
+
beta2 = 0.999
|
| 122 |
+
L2_is_weight_decay = true
|
| 123 |
+
L2 = 0.01
|
| 124 |
+
grad_clip = 1.0
|
| 125 |
+
use_averages = false
|
| 126 |
+
eps = 0.00000001
|
| 127 |
+
learn_rate = 0.001
|
| 128 |
+
|
| 129 |
+
[training.score_weights]
|
| 130 |
+
ents_f = 1.0
|
| 131 |
+
ents_p = 0.0
|
| 132 |
+
ents_r = 0.0
|
| 133 |
+
ents_per_type = null
|
| 134 |
+
|
| 135 |
+
[pretraining]
|
| 136 |
+
|
| 137 |
+
[initialize]
|
| 138 |
+
vectors = ${paths.vectors}
|
| 139 |
+
init_tok2vec = ${paths.init_tok2vec}
|
| 140 |
+
vocab_data = null
|
| 141 |
+
lookups = null
|
| 142 |
+
before_init = null
|
| 143 |
+
after_init = null
|
| 144 |
+
|
| 145 |
+
[initialize.components]
|
| 146 |
+
|
| 147 |
+
[initialize.tokenizer]
|
entity_punctuation_removal.py
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import spacy
|
| 2 |
+
from spacy.language import Language
|
| 3 |
+
import regex
|
| 4 |
+
|
| 5 |
+
|
| 6 |
+
@Language.component("entity_punctuation_removal")
|
| 7 |
+
|
| 8 |
+
|
| 9 |
+
def entity_punctuation_removal(doc):
|
| 10 |
+
# On liste nos entités
|
| 11 |
+
ents = list(doc.ents)
|
| 12 |
+
|
| 13 |
+
i = 0
|
| 14 |
+
while i < len(ents):
|
| 15 |
+
current_ent = ents[i]
|
| 16 |
+
|
| 17 |
+
# On attrape l'entité si :
|
| 18 |
+
# - Elle consiste en un unique signe de ponctuation
|
| 19 |
+
# - Et que son tag IOB est B (début d'entité)
|
| 20 |
+
# - Et qu'elle n'est pas suivie par une entité de tag IOB I (suite d'entité) OU qu'elle n'est suivie par aucun token
|
| 21 |
+
# Si l'entité est attrapée, elle est supprimée du doc
|
| 22 |
+
if i + 1 < len(ents) and regex.match(r'^\p{P}$', current_ent.text) and current_ent.root.ent_iob_ == "B" :
|
| 23 |
+
ents.pop(i)
|
| 24 |
+
elif i == len(ents) - 1 and regex.match(r'^\p{P}$', current_ent.text) and current_ent.root.ent_iob_ == "B" :
|
| 25 |
+
ents.pop(i)
|
| 26 |
+
else:
|
| 27 |
+
i += 1
|
| 28 |
+
|
| 29 |
+
# On met à jour le doc avec les entités modifiées
|
| 30 |
+
doc.ents = tuple(ents)
|
| 31 |
+
|
| 32 |
+
return doc
|
| 33 |
+
|
| 34 |
+
Language.component("entity_punctuation_removal", func=entity_punctuation_removal)
|
fr_arches_ner-any-py3-none-any.whl
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e42c101b766cc100c041e3e898d6565703687e9f2a20a14809f77aef3a11ca8
|
| 3 |
+
size 587844207
|
meta.json
ADDED
|
@@ -0,0 +1,140 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"lang":"fr",
|
| 3 |
+
"name":"arches_ner",
|
| 4 |
+
"version":"0.0.0",
|
| 5 |
+
"description":"",
|
| 6 |
+
"author":"",
|
| 7 |
+
"email":"",
|
| 8 |
+
"url":"",
|
| 9 |
+
"license":"",
|
| 10 |
+
"spacy_version":">=3.6.1,<3.7.0",
|
| 11 |
+
"spacy_git_version":"458bc5f45",
|
| 12 |
+
"vectors":{
|
| 13 |
+
"width":300,
|
| 14 |
+
"vectors":500000,
|
| 15 |
+
"keys":500000,
|
| 16 |
+
"name":"fr_vectors"
|
| 17 |
+
},
|
| 18 |
+
"labels":{
|
| 19 |
+
"tok2vec":[
|
| 20 |
+
|
| 21 |
+
],
|
| 22 |
+
"ner":[
|
| 23 |
+
"CHRONOLOGIE",
|
| 24 |
+
"DECOR",
|
| 25 |
+
"EDIFICE",
|
| 26 |
+
"ESPECE",
|
| 27 |
+
"GPE",
|
| 28 |
+
"ID",
|
| 29 |
+
"LIEUDIT_SITE",
|
| 30 |
+
"LOC",
|
| 31 |
+
"MATERIAU",
|
| 32 |
+
"MOBILIER",
|
| 33 |
+
"ORG",
|
| 34 |
+
"PERSONNE",
|
| 35 |
+
"PEUPLE_CULTURE",
|
| 36 |
+
"STRUCTURE",
|
| 37 |
+
"TECHNIQUE_STYLE"
|
| 38 |
+
]
|
| 39 |
+
},
|
| 40 |
+
"pipeline":[
|
| 41 |
+
"tok2vec",
|
| 42 |
+
"ner",
|
| 43 |
+
"entity_punctuation_removal"
|
| 44 |
+
],
|
| 45 |
+
"components":[
|
| 46 |
+
"tok2vec",
|
| 47 |
+
"ner",
|
| 48 |
+
"entity_punctuation_removal"
|
| 49 |
+
],
|
| 50 |
+
"disabled":[
|
| 51 |
+
|
| 52 |
+
],
|
| 53 |
+
"performance":{
|
| 54 |
+
"ents_f":0.6962401393,
|
| 55 |
+
"ents_p":0.6778376222,
|
| 56 |
+
"ents_r":0.7156697557,
|
| 57 |
+
"ents_per_type":{
|
| 58 |
+
"STRUCTURE":{
|
| 59 |
+
"p":0.5348226018,
|
| 60 |
+
"r":0.6840336134,
|
| 61 |
+
"f":0.6002949853
|
| 62 |
+
},
|
| 63 |
+
"GPE":{
|
| 64 |
+
"p":0.6465201465,
|
| 65 |
+
"r":0.8609756098,
|
| 66 |
+
"f":0.7384937238
|
| 67 |
+
},
|
| 68 |
+
"ORG":{
|
| 69 |
+
"p":0.6923076923,
|
| 70 |
+
"r":0.5142857143,
|
| 71 |
+
"f":0.5901639344
|
| 72 |
+
},
|
| 73 |
+
"CHRONOLOGIE":{
|
| 74 |
+
"p":0.6940966011,
|
| 75 |
+
"r":0.7548638132,
|
| 76 |
+
"f":0.7232059646
|
| 77 |
+
},
|
| 78 |
+
"LOC":{
|
| 79 |
+
"p":0.7228915663,
|
| 80 |
+
"r":0.447761194,
|
| 81 |
+
"f":0.5529953917
|
| 82 |
+
},
|
| 83 |
+
"LIEUDIT_SITE":{
|
| 84 |
+
"p":0.5875,
|
| 85 |
+
"r":0.4585365854,
|
| 86 |
+
"f":0.5150684932
|
| 87 |
+
},
|
| 88 |
+
"PEUPLE_CULTURE":{
|
| 89 |
+
"p":0.2692307692,
|
| 90 |
+
"r":0.4666666667,
|
| 91 |
+
"f":0.3414634146
|
| 92 |
+
},
|
| 93 |
+
"MATERIAU":{
|
| 94 |
+
"p":0.6987951807,
|
| 95 |
+
"r":0.7341772152,
|
| 96 |
+
"f":0.7160493827
|
| 97 |
+
},
|
| 98 |
+
"TECHNIQUE_STYLE":{
|
| 99 |
+
"p":0.6875,
|
| 100 |
+
"r":0.5770491803,
|
| 101 |
+
"f":0.6274509804
|
| 102 |
+
},
|
| 103 |
+
"MOBILIER":{
|
| 104 |
+
"p":0.7508939213,
|
| 105 |
+
"r":0.6992230855,
|
| 106 |
+
"f":0.724137931
|
| 107 |
+
},
|
| 108 |
+
"ID":{
|
| 109 |
+
"p":0.7455673759,
|
| 110 |
+
"r":0.8368159204,
|
| 111 |
+
"f":0.7885607126
|
| 112 |
+
},
|
| 113 |
+
"EDIFICE":{
|
| 114 |
+
"p":0.4090909091,
|
| 115 |
+
"r":0.4090909091,
|
| 116 |
+
"f":0.4090909091
|
| 117 |
+
},
|
| 118 |
+
"ESPECE":{
|
| 119 |
+
"p":0.8,
|
| 120 |
+
"r":0.4285714286,
|
| 121 |
+
"f":0.5581395349
|
| 122 |
+
},
|
| 123 |
+
"PERSONNE":{
|
| 124 |
+
"p":0.7832167832,
|
| 125 |
+
"r":0.7943262411,
|
| 126 |
+
"f":0.7887323944
|
| 127 |
+
},
|
| 128 |
+
"DECOR":{
|
| 129 |
+
"p":0.3538461538,
|
| 130 |
+
"r":0.4423076923,
|
| 131 |
+
"f":0.3931623932
|
| 132 |
+
}
|
| 133 |
+
},
|
| 134 |
+
"tok2vec_loss":634.3608600097,
|
| 135 |
+
"ner_loss":2460.5982992177
|
| 136 |
+
},
|
| 137 |
+
"requirements":[
|
| 138 |
+
|
| 139 |
+
]
|
| 140 |
+
}
|
ner/cfg
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"moves":null,
|
| 3 |
+
"update_with_oracle_cut_size":100,
|
| 4 |
+
"multitasks":[
|
| 5 |
+
|
| 6 |
+
],
|
| 7 |
+
"min_action_freq":1,
|
| 8 |
+
"learn_tokens":false,
|
| 9 |
+
"beam_width":1,
|
| 10 |
+
"beam_density":0.0,
|
| 11 |
+
"beam_update_prob":0.0,
|
| 12 |
+
"incorrect_spans_key":null
|
| 13 |
+
}
|
ner/model
ADDED
|
Binary file (183 kB). View file
|
|
|
ner/moves
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
��moves��{"0":{},"1":{"MOBILIER":24870,"STRUCTURE":23884,"CHRONOLOGIE":21745,"ID":20721,"MATERIAU":11489,"GPE":4094,"TECHNIQUE_STYLE":3857,"LIEUDIT_SITE":3672,"PERSONNE":2710,"DECOR":2038,"LOC":1686,"ORG":1596,"ESPECE":1310,"EDIFICE":1216,"PEUPLE_CULTURE":757},"2":{"MOBILIER":24870,"STRUCTURE":23884,"CHRONOLOGIE":21745,"ID":20721,"MATERIAU":11489,"GPE":4094,"TECHNIQUE_STYLE":3857,"LIEUDIT_SITE":3672,"PERSONNE":2710,"DECOR":2038,"LOC":1686,"ORG":1596,"ESPECE":1310,"EDIFICE":1216,"PEUPLE_CULTURE":757},"3":{"MOBILIER":24870,"STRUCTURE":23884,"CHRONOLOGIE":21745,"ID":20721,"MATERIAU":11489,"GPE":4094,"TECHNIQUE_STYLE":3857,"LIEUDIT_SITE":3672,"PERSONNE":2710,"DECOR":2038,"LOC":1686,"ORG":1596,"ESPECE":1310,"EDIFICE":1216,"PEUPLE_CULTURE":757},"4":{"MOBILIER":24870,"STRUCTURE":23884,"CHRONOLOGIE":21745,"ID":20721,"MATERIAU":11489,"GPE":4094,"TECHNIQUE_STYLE":3857,"LIEUDIT_SITE":3672,"PERSONNE":2710,"DECOR":2038,"LOC":1686,"ORG":1596,"ESPECE":1310,"EDIFICE":1216,"PEUPLE_CULTURE":757,"":1},"5":{"":1}}�cfg��neg_key�
|
repartition.txt
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
Corpus train:
|
| 2 |
+
|
| 3 |
+
CIF_antique_C001185_BD.conll
|
| 4 |
+
BFC_neo_F200449601_BD.conll
|
| 5 |
+
BFC_proto_F101604_01.conll
|
| 6 |
+
BFC_paleo_CB09001802_01.conll
|
| 7 |
+
BFC_moderne_F114238.conll
|
| 8 |
+
ARA_neo_F102890_01.conll
|
| 9 |
+
ARA_paleo_F125492_01.conll
|
| 10 |
+
ARA_contempo_F110732.conll
|
| 11 |
+
ARA_med_F108171.conll
|
| 12 |
+
BFC_med_F106159.conll
|
| 13 |
+
ARA_antique_F120714.conll
|
| 14 |
+
CIF_proto_F004820.conll
|
| 15 |
+
CIF_neo_C102845.conll
|
| 16 |
+
CIF_contempo_F115347.conll
|
| 17 |
+
CIF_moderne_C106097.conll
|
| 18 |
+
CIF_med_F148315_BD.conll
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
Corpus dev:
|
| 22 |
+
|
| 23 |
+
CIF_paleo_F110854.conll
|
| 24 |
+
BFC_contempo_F123602.conll
|
| 25 |
+
ARA_proto_F128240.conll
|
| 26 |
+
|
| 27 |
+
|
| 28 |
+
Corpus test:
|
| 29 |
+
|
| 30 |
+
BFC_antique_CB09003002.conll
|
| 31 |
+
ARA_moderne_F114296.conll
|
resultats_post_ponct.json
ADDED
|
@@ -0,0 +1,87 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"token_acc":1.0,
|
| 3 |
+
"token_p":1.0,
|
| 4 |
+
"token_r":1.0,
|
| 5 |
+
"token_f":1.0,
|
| 6 |
+
"ents_p":0.7385135135,
|
| 7 |
+
"ents_r":0.7481177276,
|
| 8 |
+
"ents_f":0.7432845971,
|
| 9 |
+
"ents_per_type":{
|
| 10 |
+
"PERSONNE":{
|
| 11 |
+
"p":0.8863636364,
|
| 12 |
+
"r":0.8387096774,
|
| 13 |
+
"f":0.861878453
|
| 14 |
+
},
|
| 15 |
+
"GPE":{
|
| 16 |
+
"p":0.4850299401,
|
| 17 |
+
"r":0.7941176471,
|
| 18 |
+
"f":0.6022304833
|
| 19 |
+
},
|
| 20 |
+
"LIEUDIT_SITE":{
|
| 21 |
+
"p":0.32,
|
| 22 |
+
"r":0.2424242424,
|
| 23 |
+
"f":0.275862069
|
| 24 |
+
},
|
| 25 |
+
"CHRONOLOGIE":{
|
| 26 |
+
"p":0.8034188034,
|
| 27 |
+
"r":0.752,
|
| 28 |
+
"f":0.7768595041
|
| 29 |
+
},
|
| 30 |
+
"STRUCTURE":{
|
| 31 |
+
"p":0.7475649351,
|
| 32 |
+
"r":0.7150621118,
|
| 33 |
+
"f":0.730952381
|
| 34 |
+
},
|
| 35 |
+
"ORG":{
|
| 36 |
+
"p":0.65,
|
| 37 |
+
"r":0.4482758621,
|
| 38 |
+
"f":0.5306122449
|
| 39 |
+
},
|
| 40 |
+
"LOC":{
|
| 41 |
+
"p":0.6875,
|
| 42 |
+
"r":0.431372549,
|
| 43 |
+
"f":0.5301204819
|
| 44 |
+
},
|
| 45 |
+
"TECHNIQUE_STYLE":{
|
| 46 |
+
"p":0.7205882353,
|
| 47 |
+
"r":0.6805555556,
|
| 48 |
+
"f":0.7
|
| 49 |
+
},
|
| 50 |
+
"ID":{
|
| 51 |
+
"p":0.8164291702,
|
| 52 |
+
"r":0.855887522,
|
| 53 |
+
"f":0.8356928357
|
| 54 |
+
},
|
| 55 |
+
"MATERIAU":{
|
| 56 |
+
"p":0.6729131175,
|
| 57 |
+
"r":0.7274401473,
|
| 58 |
+
"f":0.6991150442
|
| 59 |
+
},
|
| 60 |
+
"EDIFICE":{
|
| 61 |
+
"p":0.3913043478,
|
| 62 |
+
"r":0.75,
|
| 63 |
+
"f":0.5142857143
|
| 64 |
+
},
|
| 65 |
+
"PEUPLE_CULTURE":{
|
| 66 |
+
"p":0.0,
|
| 67 |
+
"r":0.0,
|
| 68 |
+
"f":0.0
|
| 69 |
+
},
|
| 70 |
+
"MOBILIER":{
|
| 71 |
+
"p":0.7248322148,
|
| 72 |
+
"r":0.7081967213,
|
| 73 |
+
"f":0.7164179104
|
| 74 |
+
},
|
| 75 |
+
"DECOR":{
|
| 76 |
+
"p":0.2040816327,
|
| 77 |
+
"r":0.3846153846,
|
| 78 |
+
"f":0.2666666667
|
| 79 |
+
},
|
| 80 |
+
"ESPECE":{
|
| 81 |
+
"p":0.625,
|
| 82 |
+
"r":0.5555555556,
|
| 83 |
+
"f":0.5882352941
|
| 84 |
+
}
|
| 85 |
+
},
|
| 86 |
+
"speed":17108.4619219715
|
| 87 |
+
}
|
tok2vec/cfg
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
|
| 3 |
+
}
|
tok2vec/model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d9da4a07dd47786095e49afeabd31cbfff7f79a0019aeaa6f6f48dbc38a068c8
|
| 3 |
+
size 34434008
|
tokenizer
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
vocab/key2row
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d310125304022a2ff941d0224a09dc113048bc06a6d2bc1d551a163cb789c85c
|
| 3 |
+
size 6868181
|
vocab/lookups.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:76be8b528d0075f7aae98d6fa57a6d3c83ae480a8469e668d7b0af968995ac71
|
| 3 |
+
size 1
|
vocab/strings.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e9a2c495d7976654419f316e3ec3cab68e1a0a9dab9bd96d001f10eac940f6cb
|
| 3 |
+
size 10647622
|
vocab/vectors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:055541ce7b97e6d7bc98e75678cf1e58c71c962c1e77ebc674b7247e3ffb75b8
|
| 3 |
+
size 600000128
|
vocab/vectors.cfg
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"mode":"default"
|
| 3 |
+
}
|