Update spaCy pipeline

.gitattributes

@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+fr_arches_ner_trf-any-py3-none-any.whl filter=lfs diff=lfs merge=lfs -text
+transformer/model filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,59 @@
+---
+tags:
+- spacy
+- token-classification
+language:
+- fr
+license: cc-by-nc-2.0
+model-index:
+- name: fr_arches_ner_trf
+  results:
+  - task:
+      name: NER
+      type: token-classification
+    metrics:
+    - name: NER Precision
+      type: precision
+      value: 0.8033279872
+    - name: NER Recall
+      type: recall
+      value: 0.8439342881
+    - name: NER F Score
+      type: f_score
+      value: 0.8231306491
+---
+French model trained to recognize named entities from archaeological reports.
+
+| Feature | Description |
+| --- | --- |
+| **Name** | `fr_arches_ner_trf` |
+| **Version** | `0.0.0` |
+| **spaCy** | `>=3.6.1,<3.7.0` |
+| **Default Pipeline** | `transformer`, `ner`, `entity_punctuation_removal` |
+| **Components** | `transformer`, `ner`, `entity_punctuation_removal` |
+| **Vectors** | 0 keys, 0 unique vectors (0 dimensions) |
+| **Sources** | n/a |
+| **License** | `cc-by-nc 2.0` |
+| **Author** | [n/a]() |
+
+### Label Scheme
+
+<details>
+
+<summary>View label scheme (15 labels for 1 components)</summary>
+
+| Component | Labels |
+| --- | --- |
+| **`ner`** | `CHRONOLOGIE`, `DECOR`, `EDIFICE`, `ESPECE`, `GPE`, `ID`, `LIEUDIT_SITE`, `LOC`, `MATERIAU`, `MOBILIER`, `ORG`, `PERSONNE`, `PEUPLE_CULTURE`, `STRUCTURE`, `TECHNIQUE_STYLE` |
+
+</details>
+
+### Accuracy
+
+| Type | Score |
+| --- | --- |
+| `ENTS_F` | 82.31 |
+| `ENTS_P` | 80.33 |
+| `ENTS_R` | 84.39 |
+| `TRANSFORMER_LOSS` | 218923.98 |
+| `NER_LOSS` | 51779.36 |

config.cfg

@@ -0,0 +1,149 @@
+[paths]
+train = "drive/MyDrive/ARCHES/TAL/donnees/ner/corpus_artificiel/train_dev_test_spacy/train.spacy"
+dev = "drive/MyDrive/ARCHES/TAL/donnees/ner/corpus_artificiel/train_dev_test_spacy/dev.spacy"
+vectors = null
+init_tok2vec = null
+
+[system]
+gpu_allocator = "pytorch"
+seed = 0
+
+[nlp]
+lang = "fr"
+pipeline = ["transformer","ner","entity_punctuation_removal"]
+batch_size = 128
+disabled = []
+before_creation = null
+after_creation = null
+after_pipeline_creation = null
+tokenizer = {"@tokenizers":"spacy.Tokenizer.v1"}
+
+[components]
+
+[components.entity_punctuation_removal]
+factory = "entity_punctuation_removal"
+
+[components.ner]
+factory = "ner"
+incorrect_spans_key = null
+moves = null
+scorer = {"@scorers":"spacy.ner_scorer.v1"}
+update_with_oracle_cut_size = 100
+
+[components.ner.model]
+@architectures = "spacy.TransitionBasedParser.v2"
+state_type = "ner"
+extra_state_tokens = false
+hidden_width = 64
+maxout_pieces = 2
+use_upper = false
+nO = null
+
+[components.ner.model.tok2vec]
+@architectures = "spacy-transformers.TransformerListener.v1"
+grad_factor = 1.0
+pooling = {"@layers":"reduce_mean.v1"}
+upstream = "*"
+
+[components.transformer]
+factory = "transformer"
+max_batch_items = 4096
+set_extra_annotations = {"@annotation_setters":"spacy-transformers.null_annotation_setter.v1"}
+
+[components.transformer.model]
+@architectures = "spacy-transformers.TransformerModel.v3"
+name = "camembert-base"
+mixed_precision = false
+
+[components.transformer.model.get_spans]
+@span_getters = "spacy-transformers.strided_spans.v1"
+window = 128
+stride = 96
+
+[components.transformer.model.grad_scaler_config]
+
+[components.transformer.model.tokenizer_config]
+use_fast = true
+
+[components.transformer.model.transformer_config]
+
+[corpora]
+
+[corpora.dev]
+@readers = "spacy.Corpus.v1"
+path = ${paths.dev}
+max_length = 0
+gold_preproc = false
+limit = 0
+augmenter = null
+
+[corpora.train]
+@readers = "spacy.Corpus.v1"
+path = ${paths.train}
+max_length = 0
+gold_preproc = false
+limit = 0
+augmenter = null
+
+[training]
+accumulate_gradient = 3
+dev_corpus = "corpora.dev"
+train_corpus = "corpora.train"
+seed = ${system.seed}
+gpu_allocator = ${system.gpu_allocator}
+dropout = 0.1
+patience = 1600
+max_epochs = 0
+max_steps = 20000
+eval_frequency = 200
+frozen_components = []
+annotating_components = []
+before_to_disk = null
+before_update = null
+
+[training.batcher]
+@batchers = "spacy.batch_by_padded.v1"
+discard_oversize = true
+size = 2000
+buffer = 256
+get_length = null
+
+[training.logger]
+@loggers = "spacy.ConsoleLogger.v1"
+progress_bar = false
+
+[training.optimizer]
+@optimizers = "Adam.v1"
+beta1 = 0.9
+beta2 = 0.999
+L2_is_weight_decay = true
+L2 = 0.01
+grad_clip = 1.0
+use_averages = false
+eps = 0.00000001
+
+[training.optimizer.learn_rate]
+@schedules = "warmup_linear.v1"
+warmup_steps = 250
+total_steps = 20000
+initial_rate = 0.00005
+
+[training.score_weights]
+ents_f = 1.0
+ents_p = 0.0
+ents_r = 0.0
+ents_per_type = null
+
+[pretraining]
+
+[initialize]
+vectors = ${paths.vectors}
+init_tok2vec = ${paths.init_tok2vec}
+vocab_data = null
+lookups = null
+before_init = null
+after_init = null
+
+[initialize.components]
+
+[initialize.tokenizer]

entity_punctuation_removal.py

@@ -0,0 +1,34 @@
+import spacy
+from spacy.language import Language
+import regex
+
+
+@Language.component("entity_punctuation_removal")
+
+
+def entity_punctuation_removal(doc):
+    # On liste nos entités
+    ents = list(doc.ents)
+
+    i = 0
+    while i < len(ents):
+        current_ent = ents[i]
+
+        # On attrape l'entité si :
+        # - Elle consiste en un unique signe de ponctuation
+        # - Et que son tag IOB est B (début d'entité)
+        # - 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
+        # Si l'entité est attrapée, elle est supprimée du doc
+        if i + 1 < len(ents) and regex.match(r'^\p{P}$', current_ent.text) and current_ent.root.ent_iob_ == "B" :
+          ents.pop(i)
+        elif i == len(ents) - 1 and regex.match(r'^\p{P}$', current_ent.text) and current_ent.root.ent_iob_ == "B" :
+          ents.pop(i)
+        else:
+            i += 1
+
+    # On met à jour le doc avec les entités modifiées
+    doc.ents = tuple(ents)
+
+    return doc
+
+Language.component("entity_punctuation_removal", func=entity_punctuation_removal)

fr_arches_ner_trf-any-py3-none-any.whl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:952cfb3ea4a3a2822b19e14ced55ed54fa9a88c4642252c2872e726d19649de6
+size 404718336

meta.json

@@ -0,0 +1,140 @@
+{
+  "lang":"fr",
+  "name":"arches_ner_trf",
+  "version":"0.0.0",
+  "description":"French model trained to recognize named entities from archaeological reports.",
+  "author":"",
+  "email":"",
+  "url":"",
+  "license":"cc-by-nc 2.0",
+  "spacy_version":">=3.6.1,<3.7.0",
+  "spacy_git_version":"458bc5f45",
+  "vectors":{
+    "width":0,
+    "vectors":0,
+    "keys":0,
+    "name":null
+  },
+  "labels":{
+    "transformer":[
+
+    ],
+    "ner":[
+      "CHRONOLOGIE",
+      "DECOR",
+      "EDIFICE",
+      "ESPECE",
+      "GPE",
+      "ID",
+      "LIEUDIT_SITE",
+      "LOC",
+      "MATERIAU",
+      "MOBILIER",
+      "ORG",
+      "PERSONNE",
+      "PEUPLE_CULTURE",
+      "STRUCTURE",
+      "TECHNIQUE_STYLE"
+    ]
+  },
+  "pipeline":[
+    "transformer",
+    "ner",
+    "entity_punctuation_removal"
+  ],
+  "components":[
+    "transformer",
+    "ner",
+    "entity_punctuation_removal"
+  ],
+  "disabled":[
+
+  ],
+  "performance":{
+    "ents_f":0.8231306491,
+    "ents_p":0.8033279872,
+    "ents_r":0.8439342881,
+    "ents_per_type":{
+      "ORG":{
+        "p":0.6967213115,
+        "r":0.8095238095,
+        "f":0.7488986784
+      },
+      "GPE":{
+        "p":0.869047619,
+        "r":0.8902439024,
+        "f":0.8795180723
+      },
+      "CHRONOLOGIE":{
+        "p":0.8232044199,
+        "r":0.8696498054,
+        "f":0.8457899716
+      },
+      "LOC":{
+        "p":0.8062015504,
+        "r":0.776119403,
+        "f":0.7908745247
+      },
+      "LIEUDIT_SITE":{
+        "p":0.7112299465,
+        "r":0.6487804878,
+        "f":0.6785714286
+      },
+      "PEUPLE_CULTURE":{
+        "p":0.2708333333,
+        "r":0.8666666667,
+        "f":0.4126984127
+      },
+      "MATERIAU":{
+        "p":0.8187702265,
+        "r":0.8006329114,
+        "f":0.8096
+      },
+      "TECHNIQUE_STYLE":{
+        "p":0.785467128,
+        "r":0.7442622951,
+        "f":0.7643097643
+      },
+      "MOBILIER":{
+        "p":0.8251308901,
+        "r":0.8745837958,
+        "f":0.849137931
+      },
+      "STRUCTURE":{
+        "p":0.7746031746,
+        "r":0.8201680672,
+        "f":0.7967346939
+      },
+      "ID":{
+        "p":0.8271155596,
+        "r":0.9044776119,
+        "f":0.8640684411
+      },
+      "ESPECE":{
+        "p":0.96,
+        "r":0.8571428571,
+        "f":0.9056603774
+      },
+      "PERSONNE":{
+        "p":0.875,
+        "r":0.8936170213,
+        "f":0.8842105263
+      },
+      "DECOR":{
+        "p":0.5223880597,
+        "r":0.6730769231,
+        "f":0.5882352941
+      },
+      "EDIFICE":{
+        "p":0.4761904762,
+        "r":0.4545454545,
+        "f":0.4651162791
+      }
+    },
+    "transformer_loss":2189.2397952173,
+    "ner_loss":517.7935899099
+  },
+  "requirements":[
+    "spacy-transformers>=1.3.4,<1.4.0"
+  ]
+}

ner/cfg

@@ -0,0 +1,13 @@
+{
+  "moves":null,
+  "update_with_oracle_cut_size":100,
+  "multitasks":[
+
+  ],
+  "min_action_freq":1,
+  "learn_tokens":false,
+  "beam_width":1,
+  "beam_density":0.0,
+  "beam_update_prob":0.0,
+  "incorrect_spans_key":null
+}

ner/moves

@@ -0,0 +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

@@ -0,0 +1,31 @@
+Corpus train:
+
+ARA_paleo_F125492_01.conll
+CIF_med_F148315_BD.conll
+BFC_moderne_F114238.conll
+CIF_moderne_C106097.conll
+CIF_proto_F004820.conll
+CIF_neo_C102845.conll
+CIF_contempo_F115347.conll
+BFC_neo_F200449601_BD.conll
+BFC_paleo_CB09001802_01.conll
+CIF_antique_C001185_BD.conll
+BFC_proto_F101604_01.conll
+BFC_med_F106159.conll
+ARA_neo_F102890_01.conll
+ARA_antique_F120714.conll
+ARA_med_F108171.conll
+ARA_contempo_F110732.conll
+
+
+Corpus dev:
+
+CIF_paleo_F110854.conll
+ARA_proto_F128240.conll
+BFC_contempo_F123602.conll
+
+
+Corpus test:
+
+ARA_moderne_F114296.conll
+BFC_antique_CB09003002.conll

resultats_post_ponct.json

@@ -0,0 +1,87 @@
+{
+  "token_acc":1.0,
+  "token_p":1.0,
+  "token_r":1.0,
+  "token_f":1.0,
+  "ents_p":0.8308413301,
+  "ents_r":0.8494182067,
+  "ents_f":0.8400270758,
+  "ents_per_type":{
+    "PERSONNE":{
+      "p":0.9523809524,
+      "r":0.8602150538,
+      "f":0.9039548023
+    },
+    "GPE":{
+      "p":0.8152173913,
+      "r":0.7352941176,
+      "f":0.7731958763
+    },
+    "LIEUDIT_SITE":{
+      "p":0.5666666667,
+      "r":0.5151515152,
+      "f":0.5396825397
+    },
+    "CHRONOLOGIE":{
+      "p":0.8973684211,
+      "r":0.9093333333,
+      "f":0.9033112583
+    },
+    "ORG":{
+      "p":0.7586206897,
+      "r":0.7586206897,
+      "f":0.7586206897
+    },
+    "STRUCTURE":{
+      "p":0.8469945355,
+      "r":0.8423913043,
+      "f":0.8446866485
+    },
+    "LOC":{
+      "p":0.7058823529,
+      "r":0.7058823529,
+      "f":0.7058823529
+    },
+    "EDIFICE":{
+      "p":0.5238095238,
+      "r":0.9166666667,
+      "f":0.6666666667
+    },
+    "PEUPLE_CULTURE":{
+      "p":0.0,
+      "r":0.0,
+      "f":0.0
+    },
+    "ID":{
+      "p":0.8692240628,
+      "r":0.8760984183,
+      "f":0.8726477024
+    },
+    "MOBILIER":{
+      "p":0.8184663537,
+      "r":0.8573770492,
+      "f":0.837469976
+    },
+    "MATERIAU":{
+      "p":0.7838765009,
+      "r":0.8416206262,
+      "f":0.811722913
+    },
+    "TECHNIQUE_STYLE":{
+      "p":0.6829268293,
+      "r":0.7777777778,
+      "f":0.7272727273
+    },
+    "DECOR":{
+      "p":0.3947368421,
+      "r":0.5769230769,
+      "f":0.46875
+    },
+    "ESPECE":{
+      "p":0.8888888889,
+      "r":0.8888888889,
+      "f":0.8888888889
+    }
+  },
+  "speed":4940.2096129626
+}

transformer/cfg

@@ -0,0 +1,3 @@
+{
+  "max_batch_items":4096
+}

transformer/model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff62ae3f6360aef680dc111d6e64f7c55be57461ced8eeb9d22e0edaf3567909
+size 445800276

vocab/key2row

@@ -0,0 +1 @@
+�

vocab/lookups.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76be8b528d0075f7aae98d6fa57a6d3c83ae480a8469e668d7b0af968995ac71
+size 1

vocab/vectors.cfg

@@ -0,0 +1,3 @@
+{
+  "mode":"default"
+}