1254 lines
262 KiB
Plaintext
1254 lines
262 KiB
Plaintext
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 27,
|
|
"id": "9aeaf08b-26da-4300-b4d4-e20e9b835876",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
|
|
"To disable this warning, you can either:\n",
|
|
"\t- Avoid using `tokenizers` before the fork if possible\n",
|
|
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Note: you may need to restart the kernel to use updated packages.\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"%pip install datasets matplotlib numpy scikit-learn sentence_transformers wordcloud -q"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 12,
|
|
"id": "07d7e876-b3ba-40ed-b6d6-81683e7ee513",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"from datasets import load_dataset\n",
|
|
"import matplotlib.pyplot as plt\n",
|
|
"import random\n",
|
|
"import numpy as np\n",
|
|
"import re\n",
|
|
"from sklearn.model_selection import train_test_split\n",
|
|
"from sklearn.feature_extraction.text import TfidfVectorizer\n",
|
|
"import gc\n",
|
|
"from sentence_transformers import SentenceTransformer\n",
|
|
"from sklearn.linear_model import SGDClassifier\n",
|
|
"from sklearn.metrics import log_loss, classification_report, accuracy_score\n",
|
|
"\n",
|
|
"ds = load_dataset(\"PleIAs/French-PD-Books\")"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 13,
|
|
"id": "1c5400ed-494e-4bae-a540-71655b000319",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"image/png": "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",
|
|
"text/plain": [
|
|
"<Figure size 1000x600 with 1 Axes>"
|
|
]
|
|
},
|
|
"metadata": {},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"# Extraire les années de publication\n",
|
|
"years = ds['train']['date']\n",
|
|
"\n",
|
|
"# Nettoyer et convertir en entiers (certaines dates peuvent être manquantes ou non numériques)\n",
|
|
"years_clean = [int(y) for y in years if str(y).isdigit()]\n",
|
|
"\n",
|
|
"# Créer l'histogramme\n",
|
|
"plt.figure(figsize=(10, 6))\n",
|
|
"plt.hist(years_clean, bins=50, color='skyblue', edgecolor='black')\n",
|
|
"plt.xlabel('Année de publication')\n",
|
|
"plt.ylabel('Nombre de textes')\n",
|
|
"plt.title('Répartition des textes dans le temps')\n",
|
|
"plt.show()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 30,
|
|
"id": "f1064659",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"french_stopwords = set([\n",
|
|
" 'a', 'ai', 'aie', 'aient', 'aies', 'ait', 'alors', 'as', 'au', 'aucun', 'aura', 'aurai', 'auraient', 'aurais', 'aurait', 'auras', 'aurez', 'auriez', 'aurions', 'aurons', 'auront', 'aussi', 'autre', 'aux', 'avaient', 'avais', 'avait', 'avant', 'avec', 'avez', 'aviez', 'avions', 'avoir', 'avons', 'ayant', 'ayez', 'ayons',\n",
|
|
" 'bon',\n",
|
|
" 'c', 'ce', 'ceci', 'cela', 'ces', 'cet', 'cette', 'chaque', 'comme', 'comment',\n",
|
|
" 'd', 'dans', 'de', 'des', 'deux', 'donc', 'dont', 'du',\n",
|
|
" 'elle', 'en', 'encore', 'es', 'est', 'et', 'etaient', 'etais', 'etait', 'etant', 'ete', 'etes', 'etiez', 'etions', 'etre', 'eu', 'eue', 'eues', 'eurent', 'eus', 'eusse', 'eussent', 'eusses', 'eussiez', 'eussions', 'eut', 'eux', 'eûmes', 'eût', 'eûtes',\n",
|
|
" 'fait', 'fais', 'faisaient', 'faisais', 'faisait', 'faisant', 'faire', 'faites', 'fasse', 'fassent', 'fasses', 'fassiez', 'fassions', 'faut', 'fi', 'font', 'force', 'furent', 'fus', 'fusse', 'fussent', 'fusses', 'fussiez', 'fussions', 'fut', 'fûmes', 'fût', 'fûtes',\n",
|
|
" 'hors',\n",
|
|
" 'i', 'ici', 'il', 'ils',\n",
|
|
" 'j', 'je',\n",
|
|
" 'l', 'la', 'le', 'les', 'leur', 'leurs', 'lui',\n",
|
|
" 'm', 'ma', 'mais', 'me', 'mes', 'moi', 'mon',\n",
|
|
" 'n', 'ne', 'ni', 'nos', 'notre', 'nous',\n",
|
|
" 'on', 'ont', 'ou', 'où',\n",
|
|
" 'par', 'pas', 'pendant', 'peu', 'peut', 'peux', 'plus', 'point', 'pour', 'pourquoi',\n",
|
|
" 'qu', 'quand', 'que', 'quel', 'quelle', 'quelles', 'quels', 'qui',\n",
|
|
" 's', 'sa', 'sans', 'se', 'sera', 'serai', 'seraient', 'serais', 'serait', 'seras', 'serez', 'seriez', 'serions', 'serons', 'seront', 'ses', 'soi', 'soient', 'sois', 'soit', 'sommes', 'son', 'sont', 'soyez', 'soyons', 'suis', 'sur',\n",
|
|
" 't', 'ta', 'te', 'tes', 'toi', 'ton', 'tous', 'tout', 'tu', 'un', 'une',\n",
|
|
" 'va', 'vers', 'voici', 'voilà', 'vos', 'votre', 'vous',\n",
|
|
" 'y', 'à'\n",
|
|
"])"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 31,
|
|
"id": "91c483b8-ae5b-4f1d-9f8a-dffc8eea16c9",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"def clean_text(example):\n",
|
|
" text = example[\"complete_text\"]\n",
|
|
" date = example.get(\"date\", None)\n",
|
|
"\n",
|
|
" # --- Nettoyage de la date\n",
|
|
" if \"-\" in str(date) and date is not None:\n",
|
|
" parts = str(date).split(\"-\")\n",
|
|
" if (parts[1].isdigit() and len(parts[1]) == 4) and parts[0] == \"????\":\n",
|
|
" date = str(parts[1])\n",
|
|
" else:\n",
|
|
" date = str(parts[0])\n",
|
|
"\n",
|
|
" # --- Nettoyage de texte\n",
|
|
" text = (text.replace(\"\\\\n\", \" \")\n",
|
|
" .replace(\"\\\\r\", \" \")\n",
|
|
" .replace(\"\\\\t\", \" \"))\n",
|
|
" \n",
|
|
" text = text.lower()\n",
|
|
"\n",
|
|
" text = re.sub(r\"[^a-zàâäæçéèêëïîôùûüœ\\s]\", \" \", text)\n",
|
|
" \n",
|
|
" text = re.sub(r\"\\s+\", \" \", text).strip()\n",
|
|
"\n",
|
|
" words = text.split()\n",
|
|
" filtered_words = [word for word in words if word not in french_stopwords]\n",
|
|
" text = \" \".join(filtered_words)\n",
|
|
"\n",
|
|
" return {\"text\": text, \"date\": str(date)}"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 32,
|
|
"id": "39e38bbf",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Map: 100%|██████████| 5000/5000 [03:51<00:00, 21.59 examples/s]\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"reduced_ds = ds['train'].shuffle(seed=42).select(range(5000))\n",
|
|
"\n",
|
|
"cleaned_ds = reduced_ds.map(clean_text, remove_columns=reduced_ds.column_names)\n",
|
|
"# cleaned_ds = ds['train'].map(clean_text, remove_columns=ds['train'].column_names, num_proc=4)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 33,
|
|
"id": "0474c7fd",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Filter: 100%|██████████| 5000/5000 [00:02<00:00, 2351.58 examples/s]\n",
|
|
"Map: 100%|██████████| 5000/5000 [00:02<00:00, 2033.97 examples/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Colonnes du DataFrame:\n",
|
|
"Index(['date', 'text', 'period'], dtype='object')\n",
|
|
"\\Aperçu des données avec périodes:\n",
|
|
" date text period\n",
|
|
"0 1888 v congres international sociétés chahitables s... 1850-1899\n",
|
|
"1 1716 ordonnance roy reformer quatre compagnies ze r... 1700-1749\n",
|
|
"2 1856 souvenir sme année janvier respects double all... 1850-1899\n",
|
|
"3 1782 flore bourgogne seconde partie flore bourgogne... 1750-1799\n",
|
|
"4 1863 approvisionnement ville lyon lt gt lt gt houil... 1850-1899\n",
|
|
"\n",
|
|
"Périodes et nombre de textes:\n",
|
|
"period\n",
|
|
"1550-1599 5\n",
|
|
"1600-1649 62\n",
|
|
"1650-1699 44\n",
|
|
"1700-1749 122\n",
|
|
"1750-1799 381\n",
|
|
"1800-1849 1066\n",
|
|
"1850-1899 2361\n",
|
|
"1900-1949 910\n",
|
|
"1950-1999 43\n",
|
|
"2000-2049 6\n",
|
|
"Name: text, dtype: int64\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"import pandas as pd\n",
|
|
"from sklearn.feature_extraction.text import TfidfVectorizer\n",
|
|
"from wordcloud import WordCloud\n",
|
|
"import matplotlib.pyplot as plt\n",
|
|
"import math\n",
|
|
"\n",
|
|
"def create_period_label(example, period_length=50):\n",
|
|
" \"\"\"Crée une étiquette textuelle (ex: '1880-1899') pour chaque période.\"\"\"\n",
|
|
" try:\n",
|
|
" # Gérer les dates potentiellement non numériques ou vides\n",
|
|
" year_str = str(example['date']).strip()\n",
|
|
" if not year_str.isdigit():\n",
|
|
" # Retourne une étiquette spéciale ou None si la date n'est pas valide\n",
|
|
" # Ou essayez d'extraire l'année si c'est une plage comme \"188?-1890\"\n",
|
|
" # Pour simplifier, on ignore les dates non valides ici.\n",
|
|
" return {\"period\": None}\n",
|
|
" year = int(year_str)\n",
|
|
" start_year = (year // period_length) * period_length\n",
|
|
" end_year = start_year + period_length - 1\n",
|
|
" # Gérer les années avant 0 si nécessaire, bien que peu probable pour ce dataset\n",
|
|
" if start_year < 0:\n",
|
|
" start_year = 0\n",
|
|
" end_year = period_length -1\n",
|
|
" return {\"period\": f\"{start_year}-{end_year}\"}\n",
|
|
" except (ValueError, TypeError):\n",
|
|
" # Gère les erreurs de conversion ou les types inattendus\n",
|
|
" return {\"period\": None}\n",
|
|
"\n",
|
|
"# Création des labels de période (ex: 20 ans)\n",
|
|
"# Filtrer les exemples sans date valide AVANT le map\n",
|
|
"dataset_with_dates = cleaned_ds.filter(lambda x: str(x.get('date', '')).isdigit())\n",
|
|
"dataset_with_labels = dataset_with_dates.map(create_period_label)\n",
|
|
"\n",
|
|
"# --- Groupement des textes par période ---\n",
|
|
"# Convertir en DataFrame pandas pour faciliter le groupement\n",
|
|
"df = pd.DataFrame(dataset_with_labels)\n",
|
|
"\n",
|
|
"# Vérifier les colonnes après conversion en DataFrame\n",
|
|
"print(\"Colonnes du DataFrame:\")\n",
|
|
"print(df.columns)\n",
|
|
"print(\"\\Aperçu des données avec périodes:\")\n",
|
|
"print(df.head())\n",
|
|
"\n",
|
|
"# Filtrer les lignes où 'period' est None si le filtrage précédent n'a pas tout attrapé\n",
|
|
"df = df.dropna(subset=['period'])\n",
|
|
"\n",
|
|
"# Grouper les textes par période\n",
|
|
"grouped_texts = df.groupby('period')['text'].apply(list)\n",
|
|
"\n",
|
|
"# Afficher les périodes trouvées et le nombre de textes par période\n",
|
|
"print(\"\\nPériodes et nombre de textes:\")\n",
|
|
"print(grouped_texts.apply(len))"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 34,
|
|
"id": "4acca337",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"\n",
|
|
"--- Traitement de la période : 1550-1599 ---\n",
|
|
"Nuage de mots sauvegardé : wordclouds_by_period/wordcloud_1550-1599.png\n",
|
|
"\n",
|
|
"--- Traitement de la période : 1600-1649 ---\n",
|
|
"Nuage de mots sauvegardé : wordclouds_by_period/wordcloud_1600-1649.png\n",
|
|
"\n",
|
|
"--- Traitement de la période : 1650-1699 ---\n",
|
|
"Nuage de mots sauvegardé : wordclouds_by_period/wordcloud_1650-1699.png\n",
|
|
"\n",
|
|
"--- Traitement de la période : 1700-1749 ---\n",
|
|
"Nuage de mots sauvegardé : wordclouds_by_period/wordcloud_1700-1749.png\n",
|
|
"\n",
|
|
"--- Traitement de la période : 1750-1799 ---\n",
|
|
"Nuage de mots sauvegardé : wordclouds_by_period/wordcloud_1750-1799.png\n",
|
|
"\n",
|
|
"--- Traitement de la période : 1800-1849 ---\n",
|
|
"Nuage de mots sauvegardé : wordclouds_by_period/wordcloud_1800-1849.png\n",
|
|
"\n",
|
|
"--- Traitement de la période : 1850-1899 ---\n",
|
|
"Nuage de mots sauvegardé : wordclouds_by_period/wordcloud_1850-1899.png\n",
|
|
"\n",
|
|
"--- Traitement de la période : 1900-1949 ---\n",
|
|
"Nuage de mots sauvegardé : wordclouds_by_period/wordcloud_1900-1949.png\n",
|
|
"\n",
|
|
"--- Traitement de la période : 1950-1999 ---\n",
|
|
"Nuage de mots sauvegardé : wordclouds_by_period/wordcloud_1950-1999.png\n",
|
|
"\n",
|
|
"--- Traitement de la période : 2000-2049 ---\n",
|
|
"Nuage de mots sauvegardé : wordclouds_by_period/wordcloud_2000-2049.png\n",
|
|
"\n",
|
|
"--- Analyse terminée ---\n",
|
|
"10 nuages de mots ont été générés dans le dossier 'wordclouds_by_period'.\n",
|
|
"Fichiers générés :\n",
|
|
"- wordclouds_by_period/wordcloud_1550-1599.png\n",
|
|
"- wordclouds_by_period/wordcloud_1600-1649.png\n",
|
|
"- wordclouds_by_period/wordcloud_1650-1699.png\n",
|
|
"- wordclouds_by_period/wordcloud_1700-1749.png\n",
|
|
"- wordclouds_by_period/wordcloud_1750-1799.png\n",
|
|
"- wordclouds_by_period/wordcloud_1800-1849.png\n",
|
|
"- wordclouds_by_period/wordcloud_1850-1899.png\n",
|
|
"- wordclouds_by_period/wordcloud_1900-1949.png\n",
|
|
"- wordclouds_by_period/wordcloud_1950-1999.png\n",
|
|
"- wordclouds_by_period/wordcloud_2000-2049.png\n",
|
|
"\n",
|
|
"Exemple des mots les plus pertinents (TF-IDF) pour 1550-1599:\n",
|
|
"[('roy', np.float64(0.6274568158951948)), ('auoit', np.float64(0.6253854286446879)), ('ceux', np.float64(0.599779325806743)), ('ladite', np.float64(0.5234890491823647)), ('lt', np.float64(0.5067755595792786)), ('vostre', np.float64(0.47243370797333845)), ('esté', np.float64(0.46174504904120917)), ('lequel', np.float64(0.41994470660442945)), ('pere', np.float64(0.3969215294336842)), ('ceste', np.float64(0.38365711348985176))]\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"# Paramètres pour TF-IDF et Word Cloud\n",
|
|
"max_features_tfidf = 5000 # Limiter le vocabulaire pour TF-IDF\n",
|
|
"min_df_tfidf = 3 # Ignorer les mots trop rares\n",
|
|
"max_df_tfidf = 0.85 # Ignorer les mots trop fréquents\n",
|
|
"top_n_words = 100 # Nombre de mots à inclure dans le nuage\n",
|
|
"\n",
|
|
"# S'assurer que le répertoire existe (si ce n'est pas déjà fait)\n",
|
|
"import os\n",
|
|
"output_dir = \"wordclouds_by_period\"\n",
|
|
"os.makedirs(output_dir, exist_ok=True)\n",
|
|
"\n",
|
|
"tfidf_results = {}\n",
|
|
"wordcloud_files = []\n",
|
|
"\n",
|
|
"# Calculer le TF-IDF et générer les nuages pour chaque période\n",
|
|
"for period, texts in grouped_texts.items():\n",
|
|
" print(f\"\\n--- Traitement de la période : {period} ---\")\n",
|
|
" if not texts:\n",
|
|
" print(\"Aucun texte pour cette période.\")\n",
|
|
" continue\n",
|
|
"\n",
|
|
" try:\n",
|
|
" # Initialiser et appliquer TF-IDF pour la période actuelle\n",
|
|
" tfidf_vectorizer_period = TfidfVectorizer(\n",
|
|
" max_features=max_features_tfidf,\n",
|
|
" min_df=min_df_tfidf,\n",
|
|
" max_df=max_df_tfidf,\n",
|
|
" stop_words='english' # Utiliser 'french' si disponible, sinon None ou une liste personnalisée\n",
|
|
" # Note: Le VM pourrait ne pas avoir les stopwords français par défaut.\n",
|
|
" # Si 'french' cause une erreur, utiliser None.\n",
|
|
" )\n",
|
|
" tfidf_matrix = tfidf_vectorizer_period.fit_transform(texts)\n",
|
|
"\n",
|
|
" # Obtenir les noms des features (mots)\n",
|
|
" feature_names = tfidf_vectorizer_period.get_feature_names_out()\n",
|
|
"\n",
|
|
" # Calculer le score TF-IDF total pour chaque mot sur l'ensemble des textes de la période\n",
|
|
" # Somme des scores TF-IDF pour chaque terme sur tous les documents de la période\n",
|
|
" sum_tfidf = tfidf_matrix.sum(axis=0)\n",
|
|
" tfidf_scores = [(feature_names[col], sum_tfidf[0, col]) for col in range(sum_tfidf.shape[1])]\n",
|
|
"\n",
|
|
" # Trier les mots par score TF-IDF décroissant\n",
|
|
" tfidf_scores.sort(key=lambda x: x[1], reverse=True)\n",
|
|
"\n",
|
|
" # Garder les top N mots pour le nuage\n",
|
|
" top_words_scores = dict(tfidf_scores[:top_n_words])\n",
|
|
" tfidf_results[period] = top_words_scores # Stocker les scores pour info\n",
|
|
"\n",
|
|
" if not top_words_scores:\n",
|
|
" print(\"Aucun mot pertinent trouvé après TF-IDF pour cette période.\")\n",
|
|
" continue\n",
|
|
"\n",
|
|
" # Générer le nuage de mots\n",
|
|
" wordcloud = WordCloud(width=800, height=400, background_color='white').generate_from_frequencies(top_words_scores)\n",
|
|
"\n",
|
|
" # Sauvegarder l'image du nuage de mots\n",
|
|
" filename = os.path.join(output_dir, f\"wordcloud_{period}.png\")\n",
|
|
" wordcloud.to_file(filename)\n",
|
|
" wordcloud_files.append(filename)\n",
|
|
" print(f\"Nuage de mots sauvegardé : {filename}\")\n",
|
|
"\n",
|
|
" # Afficher le nuage de mots (optionnel, peut être redondant si sauvegardé)\n",
|
|
" # plt.figure(figsize=(10, 5))\n",
|
|
" # plt.imshow(wordcloud, interpolation='bilinear')\n",
|
|
" # plt.axis('off')\n",
|
|
" # plt.title(f\"Nuage de mots pour la période {period}\")\n",
|
|
" # plt.show() # Attention: plt.show() n'est pas idéal dans le VM, préférer savefig\n",
|
|
"\n",
|
|
" except ValueError as e:\n",
|
|
" # Gérer le cas où il n'y a pas assez de vocabulaire après filtrage\n",
|
|
" print(f\"Erreur lors du traitement de la période {period}: {e}\")\n",
|
|
" except Exception as e:\n",
|
|
" # Gérer d'autres erreurs potentielles\n",
|
|
" print(f\"Erreur inattendue pour la période {period}: {e}\")\n",
|
|
"\n",
|
|
"\n",
|
|
"print(\"\\n--- Analyse terminée ---\")\n",
|
|
"print(f\"{len(wordcloud_files)} nuages de mots ont été générés dans le dossier '{output_dir}'.\")\n",
|
|
"\n",
|
|
"# Afficher les noms des fichiers générés pour que l'utilisateur sache ce qui est disponible\n",
|
|
"print(\"Fichiers générés :\")\n",
|
|
"for fname in wordcloud_files:\n",
|
|
" print(f\"- {fname}\")\n",
|
|
"\n",
|
|
"# Optionnel: Afficher un exemple de mots/scores TF-IDF pour une période\n",
|
|
"# (Choisir une période existante dans tfidf_results.keys())\n",
|
|
"example_period = list(tfidf_results.keys())[0] if tfidf_results else None\n",
|
|
"if example_period:\n",
|
|
" print(f\"\\nExemple des mots les plus pertinents (TF-IDF) pour {example_period}:\")\n",
|
|
" # Afficher les 10 premiers mots pour l'exemple\n",
|
|
" print(list(tfidf_results[example_period].items())[:10])"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 35,
|
|
"id": "2f38c125",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Création des embeddings pour 4000 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 250/250 [08:35<00:00, 2.06s/it]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Création des embeddings pour 1000 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 63/63 [02:13<00:00, 2.12s/it]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Shape des embeddings (entraînement) : (4000, 768)\n"
|
|
]
|
|
},
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"9132"
|
|
]
|
|
},
|
|
"execution_count": 35,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"def get_text_sample(text, selection_method=\"start_tokens\", n_words=500):\n",
|
|
" \"\"\"\n",
|
|
" Extrait un échantillon de texte en sélectionnant des mots (tokens)\n",
|
|
" du début, du milieu, de la fin, et de manière aléatoire.\n",
|
|
" \"\"\"\n",
|
|
" tokens = text.split()\n",
|
|
" \n",
|
|
" # Si le texte est trop court pour un échantillonnage significatif,\n",
|
|
" # on retourne le texte entier.\n",
|
|
" if len(tokens) < n_words * 4:\n",
|
|
" return text\n",
|
|
"\n",
|
|
" if selection_method == \"start_tokens\":\n",
|
|
" # 1. 500 mots du début\n",
|
|
" sorted_tokens = tokens[:n_words]\n",
|
|
"\n",
|
|
" if selection_method == \"end_tokens\":\n",
|
|
" # 2. 500 mots de la fin\n",
|
|
" sorted_tokens = tokens[-n_words:]\n",
|
|
"\n",
|
|
" if selection_method == \"middle_tokens\":\n",
|
|
" # 3. 500 mots du milieu\n",
|
|
" middle_index = len(tokens) // 2\n",
|
|
" middle_start = max(0, middle_index - (n_words // 2))\n",
|
|
" sorted_tokens = tokens[middle_start : middle_start + n_words]\n",
|
|
"\n",
|
|
" if selection_method == \"random_tokens\":\n",
|
|
" # 4. 500 mots aléatoires (en évitant les doublons avec les sections déjà extraites)\n",
|
|
" # On construit la liste des indices déjà utilisés\n",
|
|
" middle_index = len(tokens) // 2\n",
|
|
" middle_start = max(0, middle_index - (n_words // 2))\n",
|
|
" used_indices = set(range(n_words)) | set(range(len(tokens) - n_words, len(tokens))) | set(range(middle_start, middle_start + n_words))\n",
|
|
" # On crée la liste des indices restants disponibles\n",
|
|
" remaining_indices = [i for i in range(len(tokens)) if i not in used_indices]\n",
|
|
" \n",
|
|
" if len(remaining_indices) < n_words:\n",
|
|
" random_indices = remaining_indices\n",
|
|
" else:\n",
|
|
" random_indices = random.sample(remaining_indices, n_words)\n",
|
|
" \n",
|
|
" sorted_tokens = [tokens[i] for i in sorted(random_indices)] # trier pour garder un semblant d'ordre\n",
|
|
"\n",
|
|
" return \" \".join(sorted_tokens)\n",
|
|
"\n",
|
|
"# Charger un modèle CamemBERT léger, spécialisé dans la création d'embeddings\n",
|
|
"embedding_model = SentenceTransformer('dangvantuan/sentence-camembert-base')\n",
|
|
"\n",
|
|
"# Fonction pour créer les embeddings par lots (pour optimiser l'usage de la RAM)\n",
|
|
"def create_embeddings(texts, selection_method=\"start_tokens\", batch_size=32):\n",
|
|
" # Créer des échantillons de texte représentatifs\n",
|
|
" sampled_texts = [get_text_sample(text, selection_method, n_words=500) for text in texts]\n",
|
|
" \n",
|
|
" print(f\"Création des embeddings pour {len(sampled_texts)} textes...\")\n",
|
|
" return embedding_model.encode(\n",
|
|
" sampled_texts,\n",
|
|
" batch_size=batch_size,\n",
|
|
" show_progress_bar=True,\n",
|
|
" convert_to_numpy=True,\n",
|
|
" normalize_embeddings=True\n",
|
|
" )\n",
|
|
"\n",
|
|
"# Recréer les listes de textes car elles ont été supprimées\n",
|
|
"# Assurez-vous que les variables `train_dataset` et `test_dataset` sont bien définies en exécutant les cellules précédentes.\n",
|
|
"train_texts = [ex['text'] for ex in train_dataset]\n",
|
|
"test_texts = [ex['text'] for ex in test_dataset]\n",
|
|
"\n",
|
|
"# start_tokens, end_tokens, middle_tokens, random_tokens\n",
|
|
"selection_method = \"start_tokens\"\n",
|
|
"X_train_embedding = create_embeddings(train_texts, selection_method, batch_size=16)\n",
|
|
"X_test_embedding = create_embeddings(test_texts, selection_method, batch_size=16)\n",
|
|
"\n",
|
|
"print(f\"Shape des embeddings (entraînement) : {X_train_embedding.shape}\")\n",
|
|
"\n",
|
|
"# Libérer de la mémoire\n",
|
|
"del train_texts\n",
|
|
"del test_texts\n",
|
|
"gc.collect()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 36,
|
|
"id": "dfafa732",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"def train_with_loss_tracking(X_train, X_test, y_train, y_test, method_name, n_epochs=20):\n",
|
|
" \"\"\"\n",
|
|
" Entraîne un classifieur SGD, suit la perte à chaque époque,\n",
|
|
" et affiche les résultats ainsi qu'un graphique de la perte.\n",
|
|
" \"\"\"\n",
|
|
" print(f\"\\\\n{'='*50}\")\n",
|
|
" print(f\"Entraînement avec suivi de la perte ({method_name})\")\n",
|
|
" print(f\"{'='*50}\\\\n\")\n",
|
|
"\n",
|
|
" classifier = SGDClassifier(\n",
|
|
" loss='log_loss', \n",
|
|
" random_state=42, \n",
|
|
" n_jobs=-1, \n",
|
|
" #learning_rate='constant',\n",
|
|
" #eta0=0.001\n",
|
|
" )\n",
|
|
" \n",
|
|
" train_loss_history = []\n",
|
|
" test_loss_history = []\n",
|
|
" \n",
|
|
" # --- CORRECTION 1 : Définir la liste complète de TOUS les labels ---\n",
|
|
" # On combine les labels du train et du test pour être sûr de n'en oublier aucun.\n",
|
|
" all_classes = np.unique(np.concatenate((y_train, y_test)))\n",
|
|
"\n",
|
|
" for epoch in range(n_epochs):\n",
|
|
" classifier.partial_fit(X_train, y_train, classes=all_classes)\n",
|
|
" \n",
|
|
" y_train_pred_proba = classifier.predict_proba(X_train)\n",
|
|
" y_test_pred_proba = classifier.predict_proba(X_test)\n",
|
|
" \n",
|
|
" train_loss = log_loss(y_train, y_train_pred_proba, labels=all_classes)\n",
|
|
" test_loss = log_loss(y_test, y_test_pred_proba, labels=all_classes)\n",
|
|
" \n",
|
|
" train_loss_history.append(train_loss)\n",
|
|
" test_loss_history.append(test_loss)\n",
|
|
" \n",
|
|
" print(f\"Époque {epoch+1}/{n_epochs} - Perte entraînement: {train_loss:.4f} - Perte test: {test_loss:.4f}\")\n",
|
|
"\n",
|
|
" # Visualisation de la fonction de perte\n",
|
|
" plt.figure(figsize=(10, 5))\n",
|
|
" plt.plot(train_loss_history, label=\"Perte d'entraînement (Train Loss)\")\n",
|
|
" plt.plot(test_loss_history, label=\"Perte de validation (Test Loss)\")\n",
|
|
" plt.title(f\"Évolution de la fonction de perte ({method_name})\")\n",
|
|
" plt.xlabel(\"Époques\")\n",
|
|
" plt.ylabel(\"Log Loss\")\n",
|
|
" plt.legend()\n",
|
|
" plt.grid(True)\n",
|
|
" plt.show()\n",
|
|
"\n",
|
|
" # Evaluation de l'entrainement\n",
|
|
" y_pred = classifier.predict(X_test)\n",
|
|
" mean_absolute_error = np.mean(np.abs(y_test - y_pred))\n",
|
|
" print(f\"📊 Mean Absolute Error (MAE) finale: {mean_absolute_error:.4f}\\\\n\")\n",
|
|
" \n",
|
|
" return classifier"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 37,
|
|
"id": "c366fc0d",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"\\n==================================================\n",
|
|
"Entraînement avec suivi de la perte (Embeddings (CamemBERT))\n",
|
|
"==================================================\\n\n",
|
|
"Époque 1/50 - Perte entraînement: 1.1450 - Perte test: 1.2324\n",
|
|
"Époque 2/50 - Perte entraînement: 1.0274 - Perte test: 1.1181\n",
|
|
"Époque 3/50 - Perte entraînement: 0.9842 - Perte test: 1.0743\n",
|
|
"Époque 4/50 - Perte entraînement: 0.9606 - Perte test: 1.0501\n",
|
|
"Époque 5/50 - Perte entraînement: 0.9463 - Perte test: 1.0355\n",
|
|
"Époque 6/50 - Perte entraînement: 0.9372 - Perte test: 1.0262\n",
|
|
"Époque 7/50 - Perte entraînement: 0.9312 - Perte test: 1.0200\n",
|
|
"Époque 8/50 - Perte entraînement: 0.9269 - Perte test: 1.0156\n",
|
|
"Époque 9/50 - Perte entraînement: 0.9238 - Perte test: 1.0125\n",
|
|
"Époque 10/50 - Perte entraînement: 0.9214 - Perte test: 1.0101\n",
|
|
"Époque 11/50 - Perte entraînement: 0.9195 - Perte test: 1.0082\n",
|
|
"Époque 12/50 - Perte entraînement: 0.9179 - Perte test: 1.0067\n",
|
|
"Époque 13/50 - Perte entraînement: 0.9166 - Perte test: 1.0055\n",
|
|
"Époque 14/50 - Perte entraînement: 0.9155 - Perte test: 1.0044\n",
|
|
"Époque 15/50 - Perte entraînement: 0.9145 - Perte test: 1.0035\n",
|
|
"Époque 16/50 - Perte entraînement: 0.9136 - Perte test: 1.0027\n",
|
|
"Époque 17/50 - Perte entraînement: 0.9128 - Perte test: 1.0019\n",
|
|
"Époque 18/50 - Perte entraînement: 0.9121 - Perte test: 1.0013\n",
|
|
"Époque 19/50 - Perte entraînement: 0.9114 - Perte test: 1.0007\n",
|
|
"Époque 20/50 - Perte entraînement: 0.9108 - Perte test: 1.0002\n",
|
|
"Époque 21/50 - Perte entraînement: 0.9102 - Perte test: 0.9997\n",
|
|
"Époque 22/50 - Perte entraînement: 0.9097 - Perte test: 0.9993\n",
|
|
"Époque 23/50 - Perte entraînement: 0.9092 - Perte test: 0.9989\n",
|
|
"Époque 24/50 - Perte entraînement: 0.9088 - Perte test: 0.9985\n",
|
|
"Époque 25/50 - Perte entraînement: 0.9084 - Perte test: 0.9981\n",
|
|
"Époque 26/50 - Perte entraînement: 0.9080 - Perte test: 0.9978\n",
|
|
"Époque 27/50 - Perte entraînement: 0.9076 - Perte test: 0.9975\n",
|
|
"Époque 28/50 - Perte entraînement: 0.9073 - Perte test: 0.9972\n",
|
|
"Époque 29/50 - Perte entraînement: 0.9070 - Perte test: 0.9970\n",
|
|
"Époque 30/50 - Perte entraînement: 0.9067 - Perte test: 0.9968\n",
|
|
"Époque 31/50 - Perte entraînement: 0.9064 - Perte test: 0.9966\n",
|
|
"Époque 32/50 - Perte entraînement: 0.9061 - Perte test: 0.9964\n",
|
|
"Époque 33/50 - Perte entraînement: 0.9059 - Perte test: 0.9962\n",
|
|
"Époque 34/50 - Perte entraînement: 0.9057 - Perte test: 0.9960\n",
|
|
"Époque 35/50 - Perte entraînement: 0.9055 - Perte test: 0.9958\n",
|
|
"Époque 36/50 - Perte entraînement: 0.9053 - Perte test: 0.9957\n",
|
|
"Époque 37/50 - Perte entraînement: 0.9051 - Perte test: 0.9956\n",
|
|
"Époque 38/50 - Perte entraînement: 0.9049 - Perte test: 0.9954\n",
|
|
"Époque 39/50 - Perte entraînement: 0.9048 - Perte test: 0.9953\n",
|
|
"Époque 40/50 - Perte entraînement: 0.9046 - Perte test: 0.9952\n",
|
|
"Époque 41/50 - Perte entraînement: 0.9045 - Perte test: 0.9951\n",
|
|
"Époque 42/50 - Perte entraînement: 0.9044 - Perte test: 0.9950\n",
|
|
"Époque 43/50 - Perte entraînement: 0.9042 - Perte test: 0.9949\n",
|
|
"Époque 44/50 - Perte entraînement: 0.9041 - Perte test: 0.9949\n",
|
|
"Époque 45/50 - Perte entraînement: 0.9040 - Perte test: 0.9948\n",
|
|
"Époque 46/50 - Perte entraînement: 0.9039 - Perte test: 0.9947\n",
|
|
"Époque 47/50 - Perte entraînement: 0.9038 - Perte test: 0.9946\n",
|
|
"Époque 48/50 - Perte entraînement: 0.9037 - Perte test: 0.9946\n",
|
|
"Époque 49/50 - Perte entraînement: 0.9036 - Perte test: 0.9945\n",
|
|
"Époque 50/50 - Perte entraînement: 0.9035 - Perte test: 0.9945\n"
|
|
]
|
|
},
|
|
{
|
|
"data": {
|
|
"image/png": "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",
|
|
"text/plain": [
|
|
"<Figure size 1000x500 with 1 Axes>"
|
|
]
|
|
},
|
|
"metadata": {},
|
|
"output_type": "display_data"
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"📊 Mean Absolute Error (MAE) finale: 24.9000\\n\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"# Entraînement TF-IDF\n",
|
|
"#model_tfidf_sgd = train_with_loss_tracking(\n",
|
|
"# X_train_tfidf, X_test_tfidf, \n",
|
|
"# train_labels, test_labels,\n",
|
|
"# \"TF-IDF\",\n",
|
|
"# n_epochs=30\n",
|
|
"#)\n",
|
|
"\n",
|
|
"# Entraînement Embedings\n",
|
|
"model_embedding_sgd = train_with_loss_tracking(\n",
|
|
" X_train_embedding, X_test_embedding,\n",
|
|
" train_labels, test_labels,\n",
|
|
" \"Embeddings (CamemBERT)\",\n",
|
|
" n_epochs=50\n",
|
|
")"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 38,
|
|
"id": "02f2c811",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"def predict_random_text(test_dataset):\n",
|
|
" \"\"\"\n",
|
|
" Sélectionne un texte aléatoire dans le jeu de données, le prédit avec les deux modèles\n",
|
|
" et affiche une comparaison.\n",
|
|
" \"\"\"\n",
|
|
" # 1. Sélectionner un exemple aléatoire\n",
|
|
" random_index = random.randint(0, len(test_dataset) - 1)\n",
|
|
" random_example = test_dataset[random_index]\n",
|
|
" \n",
|
|
" title = random_example.get('title', 'Titre non disponible')\n",
|
|
" true_date = random_example.get('date', 'Date non disponible')\n",
|
|
" full_text = random_example.get('complete_text', '')\n",
|
|
"\n",
|
|
" print(f\"--- Interrogation d'un texte aléatoire ---\")\n",
|
|
" print(f\"📚 Titre : {title}\")\n",
|
|
" print(f\"🗓️ Date réelle : {true_date}\")\n",
|
|
" print(\"-\" * 40)\n",
|
|
"\n",
|
|
" # 2. Nettoyer le texte comme pour l'entraînement\n",
|
|
" # La fonction clean_text renvoie un dictionnaire, nous extrayons le texte\n",
|
|
" cleaned_text_dict = clean_text(random_example)\n",
|
|
" cleaned_text = cleaned_text_dict['text']\n",
|
|
" \n",
|
|
" # Le texte doit être dans une liste pour les transformateurs\n",
|
|
" text_to_predict = [cleaned_text]\n",
|
|
"\n",
|
|
" # 3. Prédiction avec le modèle TF-IDF\n",
|
|
" #tfidf_features = tfidf_vectorizer.transform(text_to_predict)\n",
|
|
" #prediction_tfidf = model_tfidf_sgd.predict(tfidf_features)[0]\n",
|
|
" #print(f\"🤖 Prédiction (TF-IDF) : {prediction_tfidf}s\")\n",
|
|
"\n",
|
|
" # 4. Prédiction avec le modèle Embeddings\n",
|
|
" embedding_features = create_embeddings(text_to_predict, batch_size=1)\n",
|
|
" prediction_embedding = model_embedding_sgd.predict(embedding_features)[0]\n",
|
|
" print(f\"🧠 Prédiction (Embeddings) : {prediction_embedding}s\")\n",
|
|
" print(f\"----------------------------------------\\\\n\")"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 39,
|
|
"id": "38ab11b8",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"--- Interrogation d'un texte aléatoire ---\n",
|
|
"📚 Titre : Oeuvres de Charles Nodier. T. 12\n",
|
|
"🗓️ Date réelle : 1832-1837\n",
|
|
"----------------------------------------\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 3.72it/s]"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"🧠 Prédiction (Embeddings) : 1850s\n",
|
|
"----------------------------------------\\n\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"predict_random_text(ds['train'])"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 40,
|
|
"id": "e5a99b09-681b-4a44-82f7-7c1ffe1ac84d",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"\n",
|
|
"def test_and_plot_predictions(test_dataset, num_trials=15):\n",
|
|
" \"\"\"\n",
|
|
" Sélectionne plusieurs textes aléatoires, prédit leur date avec deux modèles,\n",
|
|
" et affiche un graphique comparatif des résultats.\n",
|
|
" \n",
|
|
" Args:\n",
|
|
" test_dataset (list): La liste d'exemples à tester.\n",
|
|
" num_trials (int): Le nombre d'essais à effectuer.\n",
|
|
" \"\"\"\n",
|
|
" real_dates = []\n",
|
|
" tfidf_predictions = []\n",
|
|
" embedding_predictions = []\n",
|
|
"\n",
|
|
" print(f\"--- Lancement de {num_trials} essais de prédiction ---\")\n",
|
|
"\n",
|
|
" for i in range(num_trials):\n",
|
|
" # 1. Sélectionner un exemple aléatoire\n",
|
|
" random_example = random.choice(test_dataset)\n",
|
|
" \n",
|
|
" # Extrait la date et la convertit en entier pour le graphique\n",
|
|
" # On suppose que la date est au format 'YYYYs' (ex: '1990s')\n",
|
|
" try:\n",
|
|
" true_date_str = random_example.get('date', '0s')\n",
|
|
" true_date_int = int(true_date_str[:4]) # Convertit '1990s' en 1990\n",
|
|
" except (ValueError, IndexError):\n",
|
|
" print(f\"Avertissement : impossible de parser la date '{true_date_str}'. Essai ignoré.\")\n",
|
|
" continue\n",
|
|
"\n",
|
|
" # 2. Nettoyer le texte\n",
|
|
" cleaned_text = clean_text(random_example)['text']\n",
|
|
" text_to_predict = [cleaned_text]\n",
|
|
"\n",
|
|
" # 3. Prédiction avec le modèle TF-IDF\n",
|
|
" #tfidf_features = tfidf_vectorizer.transform(text_to_predict)\n",
|
|
" #prediction_tfidf = model_tfidf_sgd.predict(tfidf_features)[0]\n",
|
|
"\n",
|
|
" # 4. Prédiction avec le modèle Embeddings\n",
|
|
" embedding_features = create_embeddings(text_to_predict, batch_size=1)\n",
|
|
" prediction_embedding = model_embedding_sgd.predict(embedding_features)[0]\n",
|
|
" \n",
|
|
" # 5. Stocker les résultats\n",
|
|
" real_dates.append(true_date_int)\n",
|
|
" #tfidf_predictions.append(prediction_tfidf)\n",
|
|
" embedding_predictions.append(prediction_embedding)\n",
|
|
" \n",
|
|
" #print(f\"Essai {i+1}/{num_trials} | Réel : {true_date_int} | TF-IDF : {prediction_tfidf} | Embeddings : {prediction_embedding}\")\n",
|
|
" print(f\"Essai {i+1}/{num_trials} | Réel : {true_date_int} | Embeddings : {prediction_embedding}\")\n",
|
|
"\n",
|
|
" # 6. Générer le graphique\n",
|
|
" if not real_dates:\n",
|
|
" print(\"Aucune donnée à afficher.\")\n",
|
|
" return\n",
|
|
"\n",
|
|
" trials = range(1, len(real_dates) + 1)\n",
|
|
" \n",
|
|
" plt.style.use('seaborn-v0_8-whitegrid') # Style visuel agréable\n",
|
|
" plt.figure(figsize=(14, 7)) # Taille de la figure\n",
|
|
"\n",
|
|
" # Tracer chaque série de données\n",
|
|
" plt.plot(trials, real_dates, 'o-', color='green', label='Date Réelle', markersize=8)\n",
|
|
" #plt.plot(trials, tfidf_predictions, 's--', color='blue', label='Prédiction (TF-IDF)')\n",
|
|
" plt.plot(trials, embedding_predictions, '^:', color='red', label='Prédiction (Embeddings)')\n",
|
|
"\n",
|
|
" # Ajouter les titres et les légendes\n",
|
|
" plt.title('Comparaison des Prédictions de Date sur Plusieurs Essais', fontsize=16)\n",
|
|
" plt.xlabel(\"Numéro de l'essai\", fontsize=12)\n",
|
|
" plt.ylabel(\"Date (Année)\", fontsize=12)\n",
|
|
" plt.xticks(trials) # Assure que chaque essai a une graduation sur l'axe X\n",
|
|
" plt.legend(fontsize=10)\n",
|
|
" plt.tight_layout() # Ajuste le graphique pour qu'il s'affiche bien\n",
|
|
"\n",
|
|
" # Afficher le graphique\n",
|
|
" plt.show()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 41,
|
|
"id": "2df4280e",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"--- Lancement de 20 essais de prédiction ---\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 5.28it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 1/20 | Réel : 1896 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 4.52it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 2/20 | Réel : 1939 | Embeddings : 1900\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 5.07it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 3/20 | Réel : 1783 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 5.92it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 4/20 | Réel : 1885 | Embeddings : 1900\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 4.30it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 5/20 | Réel : 1913 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 1.56it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 6/20 | Réel : 1852 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 4.23it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 7/20 | Réel : 1887 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 4.98it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 8/20 | Réel : 1829 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 4.83it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 9/20 | Réel : 1811 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 4.89it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 10/20 | Réel : 1866 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 2.63it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 11/20 | Réel : 1871 | Embeddings : 1900\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 5.05it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 12/20 | Réel : 1843 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 3.99it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 13/20 | Réel : 1912 | Embeddings : 1900\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 4.27it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 14/20 | Réel : 1975 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 5.31it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 15/20 | Réel : 1890 | Embeddings : 1900\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 4.27it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 16/20 | Réel : 1865 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 3.56it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 17/20 | Réel : 1868 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 3.32it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 18/20 | Réel : 1863 | Embeddings : 1900\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 4.89it/s]\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 19/20 | Réel : 1791 | Embeddings : 1850\n",
|
|
"Création des embeddings pour 1 textes...\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Batches: 100%|██████████| 1/1 [00:00<00:00, 1.48it/s]"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Essai 20/20 | Réel : 1821 | Embeddings : 1850\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"\n"
|
|
]
|
|
},
|
|
{
|
|
"data": {
|
|
"image/png": "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",
|
|
"text/plain": [
|
|
"<Figure size 1400x700 with 1 Axes>"
|
|
]
|
|
},
|
|
"metadata": {},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"test_and_plot_predictions(ds['train'], num_trials=20)"
|
|
]
|
|
}
|
|
],
|
|
"metadata": {
|
|
"kernelspec": {
|
|
"display_name": "projet_ia5",
|
|
"language": "python",
|
|
"name": "python3"
|
|
},
|
|
"language_info": {
|
|
"codemirror_mode": {
|
|
"name": "ipython",
|
|
"version": 3
|
|
},
|
|
"file_extension": ".py",
|
|
"mimetype": "text/x-python",
|
|
"name": "python",
|
|
"nbconvert_exporter": "python",
|
|
"pygments_lexer": "ipython3",
|
|
"version": "3.11.14"
|
|
}
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 5
|
|
}
|