model data

README.md

@@ -0,0 +1,49 @@
+---
+tags:
+- generated_from_trainer
+model-index:
+- name: CamemBERT pretrained on french trade directories from the XIXth century
+  results: []
+---
+
+# CamemBERT trained and fine-tuned for NER on french trade directories from the XIXth century [PERO-OCR training set]
+
+This mdoel is part of the material of the paper
+> Abadie, N., Carlinet, E., Chazalon, J., Duménieu, B. (2022). A
+> Benchmark of Named Entity Recognition Approaches in Historical
+> Documents Application to 19𝑡ℎ Century French Directories. In: Uchida,
+> S., Barney, E., Eglin, V. (eds) Document Analysis Systems. DAS 2022.
+> Lecture Notes in Computer Science, vol 13237. Springer, Cham.
+> https://doi.org/10.1007/978-3-031-06555-2_30
+
+The source code to train this model is available on the  [GitHub repository](https://github.com/soduco/paper-ner-bench-das22) of the paper as a Jupyter notebook in  `src/ner/40_experiment_2.ipynb`.
+
+
+## Model description
+This model adapts the model [Jean-Baptiste/camembert-ner](https://huggingface.co/Jean-Baptiste/camembert-ner) for NER on 6004 manually annotated directory entries referred as the "reference dataset" in the paper.
+
+Trade directory entries are short and strongly structured texts that giving the name, activity and location of a person or business, e.g: 
+```
+Peynaud, R. de la Vieille Bouclerie, 18. Richard, Joullain et comp., (commission- —Phéâtre Français. naire, (entrepôt), au port de la Rapée-
+```
+
+## Intended uses & limitations
+This model is intended for reproducibility of the NER evaluation published in the DAS2022 paper.
+Several derived models trained for NER on trade directories are available on HuggingFace, each trained on a different dataset :
+- [das22-10-camembert_pretrained_finetuned_ref](): trained for NER on ~6000 directory entries manually corrected.
+- [das22-10-camembert_pretrained_finetuned_pero](): trained for NER on ~6000 directory entries extracted with PERO-OCR.
+- [das22-10-camembert_pretrained_finetuned_tess](): trained for NER on ~6000 directory entries extracted with Tesseract.
+
+
+
+### Training hyperparameters
+
+### Training results
+
+### Framework versions
+
+- Transformers 4.16.0.dev0
+- Pytorch 1.10.1+cu102
+- Datasets 1.17.0
+- Tokenizers 0.10.3
+

config.json

@@ -0,0 +1,51 @@
+{
+  "_name_or_path": "HueyNemud/das22-10-camembert_pretrained",
+  "architectures": [
+    "CamembertForTokenClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 5,
+  "classifier_dropout": null,
+  "eos_token_id": 6,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "O",
+    "1": "I-LOC",
+    "2": "I-PER",
+    "3": "I-MISC",
+    "4": "I-ORG",
+    "5": "I-CARDINAL",
+    "6": "I-ACT",
+    "7": "I-TITRE",
+    "8": "I-FT"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "I-ACT": 6,
+    "I-CARDINAL": 5,
+    "I-FT": 8,
+    "I-LOC": 1,
+    "I-MISC": 3,
+    "I-ORG": 4,
+    "I-PER": 2,
+    "I-TITRE": 7,
+    "O": 0
+  },
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "camembert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "output_past": true,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.15.0",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 32005
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fcf98e1b00eeaf32429e0d3ce91bb56f8d315b53cc601a6c980093af31312048
+size 880421605

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10b4c449f459a1dd150bde11f8cf172d41be99a9f3b58800012ae433cc402cc2
+size 440237809

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de31c8caf9890f76aa2de98c914e1aa6dfdd878c7859cced157280250484890c
+size 14503

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e026fb7cb5555507d8791aa654af1e3922b1146f7bb21228d191cc76d86c51e1
+size 623

sentencepiece.bpe.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:988bc5a00281c6d210a5d34bd143d0363741a432fefe741bf71e61b1869d4314
+size 810912

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": "<mask>", "additional_special_tokens": ["<s>NOTUSED", "</s>NOTUSED"]}

tokenizer_config.json

@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "additional_special_tokens": ["<s>NOTUSED", "</s>NOTUSED"], "model_max_length": 512, "name_or_path": "HueyNemud/das22-10-camembert_pretrained", "special_tokens_map_file": "/lrde/home2/jchazalo/.cache/huggingface/transformers/fe0e213c44079a9ee091098f81fff2941484006e9ba3001a9bf1ee9f87537599.cb3ec3a6c1200d181228d8825ae9767572abca54efa1bbb37fd83d721b2ef323", "sp_model_kwargs": {}, "tokenizer_class": "CamembertTokenizer"}

trainer_state.json

@@ -0,0 +1,214 @@
+{
+  "best_metric": 0.9595484477892756,
+  "best_model_checkpoint": "/content/drive/MyDrive/SODUCO/article_das_2022/43-camembert_pretrained_finetuned_pero/checkpoint-1500",
+  "epoch": 3.9893617021276597,
+  "global_step": 1500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.27,
+      "eval_accuracy": 0.9711394302848576,
+      "eval_f1": 0.9418582912693962,
+      "eval_loss": 0.24740596115589142,
+      "eval_precision": 0.9343471810089021,
+      "eval_recall": 0.9494911421032793,
+      "eval_runtime": 2.7876,
+      "eval_samples_per_second": 239.631,
+      "eval_steps_per_second": 15.067,
+      "step": 100
+    },
+    {
+      "epoch": 0.53,
+      "eval_accuracy": 0.9741379310344828,
+      "eval_f1": 0.950751879699248,
+      "eval_loss": 0.15050481259822845,
+      "eval_precision": 0.9482564679415073,
+      "eval_recall": 0.9532604598567659,
+      "eval_runtime": 2.5926,
+      "eval_samples_per_second": 257.656,
+      "eval_steps_per_second": 16.2,
+      "step": 200
+    },
+    {
+      "epoch": 0.8,
+      "eval_accuracy": 0.9730134932533733,
+      "eval_f1": 0.9475262368815591,
+      "eval_loss": 0.12827223539352417,
+      "eval_precision": 0.9422288483041371,
+      "eval_recall": 0.9528835280814173,
+      "eval_runtime": 2.73,
+      "eval_samples_per_second": 244.686,
+      "eval_steps_per_second": 15.384,
+      "step": 300
+    },
+    {
+      "epoch": 1.06,
+      "eval_accuracy": 0.9710144927536232,
+      "eval_f1": 0.9431689875799775,
+      "eval_loss": 0.1230660155415535,
+      "eval_precision": 0.9417512213453589,
+      "eval_recall": 0.9445910290237467,
+      "eval_runtime": 2.7678,
+      "eval_samples_per_second": 241.348,
+      "eval_steps_per_second": 15.175,
+      "step": 400
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 9e-05,
+      "loss": 0.2505,
+      "step": 500
+    },
+    {
+      "epoch": 1.33,
+      "eval_accuracy": 0.9786356821589205,
+      "eval_f1": 0.9521306551529942,
+      "eval_loss": 0.08949972689151764,
+      "eval_precision": 0.9483919222139118,
+      "eval_recall": 0.9558989822842066,
+      "eval_runtime": 2.9407,
+      "eval_samples_per_second": 227.154,
+      "eval_steps_per_second": 14.282,
+      "step": 500
+    },
+    {
+      "epoch": 1.6,
+      "eval_accuracy": 0.9768865567216392,
+      "eval_f1": 0.9537420082737871,
+      "eval_loss": 0.09133146703243256,
+      "eval_precision": 0.951594746716698,
+      "eval_recall": 0.9558989822842066,
+      "eval_runtime": 2.8849,
+      "eval_samples_per_second": 231.55,
+      "eval_steps_per_second": 14.559,
+      "step": 600
+    },
+    {
+      "epoch": 1.86,
+      "eval_accuracy": 0.9765117441279361,
+      "eval_f1": 0.9579800263802524,
+      "eval_loss": 0.09044157713651657,
+      "eval_precision": 0.9577995478522984,
+      "eval_recall": 0.9581605729362985,
+      "eval_runtime": 2.8752,
+      "eval_samples_per_second": 232.332,
+      "eval_steps_per_second": 14.608,
+      "step": 700
+    },
+    {
+      "epoch": 2.13,
+      "eval_accuracy": 0.9801349325337332,
+      "eval_f1": 0.9586621570838031,
+      "eval_loss": 0.07938132435083389,
+      "eval_precision": 0.9557886849007119,
+      "eval_recall": 0.9615529589144365,
+      "eval_runtime": 2.8692,
+      "eval_samples_per_second": 232.816,
+      "eval_steps_per_second": 14.638,
+      "step": 800
+    },
+    {
+      "epoch": 2.39,
+      "eval_accuracy": 0.9771364317841079,
+      "eval_f1": 0.9548872180451128,
+      "eval_loss": 0.09603337943553925,
+      "eval_precision": 0.9523809523809523,
+      "eval_recall": 0.9574067093856012,
+      "eval_runtime": 2.8538,
+      "eval_samples_per_second": 234.072,
+      "eval_steps_per_second": 14.717,
+      "step": 900
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 8e-05,
+      "loss": 0.0748,
+      "step": 1000
+    },
+    {
+      "epoch": 2.66,
+      "eval_accuracy": 0.9761369315342329,
+      "eval_f1": 0.9575871819038644,
+      "eval_loss": 0.09003584831953049,
+      "eval_precision": 0.9577677224736049,
+      "eval_recall": 0.9574067093856012,
+      "eval_runtime": 2.9061,
+      "eval_samples_per_second": 229.858,
+      "eval_steps_per_second": 14.452,
+      "step": 1000
+    },
+    {
+      "epoch": 2.93,
+      "eval_accuracy": 0.9730134932533733,
+      "eval_f1": 0.954383330204618,
+      "eval_loss": 0.10545289516448975,
+      "eval_precision": 0.9506357516828721,
+      "eval_recall": 0.9581605729362985,
+      "eval_runtime": 2.9384,
+      "eval_samples_per_second": 227.334,
+      "eval_steps_per_second": 14.293,
+      "step": 1100
+    },
+    {
+      "epoch": 3.19,
+      "eval_accuracy": 0.9761369315342329,
+      "eval_f1": 0.9593220338983051,
+      "eval_loss": 0.099028080701828,
+      "eval_precision": 0.9585999247271358,
+      "eval_recall": 0.9600452318130418,
+      "eval_runtime": 3.0033,
+      "eval_samples_per_second": 222.419,
+      "eval_steps_per_second": 13.984,
+      "step": 1200
+    },
+    {
+      "epoch": 3.46,
+      "eval_accuracy": 0.9752623688155923,
+      "eval_f1": 0.9561782960315968,
+      "eval_loss": 0.09300831705331802,
+      "eval_precision": 0.9542042042042042,
+      "eval_recall": 0.9581605729362985,
+      "eval_runtime": 2.8431,
+      "eval_samples_per_second": 234.956,
+      "eval_steps_per_second": 14.773,
+      "step": 1300
+    },
+    {
+      "epoch": 3.72,
+      "eval_accuracy": 0.9775112443778111,
+      "eval_f1": 0.9578471960858111,
+      "eval_loss": 0.0967017412185669,
+      "eval_precision": 0.9564073656520106,
+      "eval_recall": 0.9592913682623445,
+      "eval_runtime": 2.8669,
+      "eval_samples_per_second": 233.003,
+      "eval_steps_per_second": 14.65,
+      "step": 1400
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 7e-05,
+      "loss": 0.0497,
+      "step": 1500
+    },
+    {
+      "epoch": 3.99,
+      "eval_accuracy": 0.9776361819090454,
+      "eval_f1": 0.9595484477892756,
+      "eval_loss": 0.0839652568101883,
+      "eval_precision": 0.9579263711495116,
+      "eval_recall": 0.9611760271390878,
+      "eval_runtime": 2.8628,
+      "eval_samples_per_second": 233.335,
+      "eval_steps_per_second": 14.671,
+      "step": 1500
+    }
+  ],
+  "max_steps": 5000,
+  "num_train_epochs": 14,
+  "total_flos": 592464605771736.0,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31aeb7cf016b0e47597e331fa262b44873f81073521049d8a85dc46bd2004681
+size 3055