{"metadata":{"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"name":"python","version":"3.12.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"nvidiaTeslaT4","dataSources":[{"sourceType":"datasetVersion","sourceId":24919,"datasetId":1036,"databundleVersionId":24958,"isSourceIdPinned":false}],"dockerImageVersionId":31329,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":true}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"## Я выборал кейс с классификацией статей\n\n\nДля классификации статей будем использовать distilbert-base-uncased, тк это c одной стороны очень мощная модель, с другой стороны ее достаточно легко инферить на проде на cpu без дополнительных приседаний с инференс серверами\n\nВ качестве датасета буду использовать ccdv/arxiv-classification, отдельно рассматривал и другие варианты (например https://huggingface.co/datasets/ccdv/arxiv-classification, но там было сильное смещение в сs тематику)\nПри этом все еще далеко на все тематики раскрыты, нужно больше данных для более робастной модели\n\n- Отдельно обучил модель для классификации \"мусора\", те если нам на вход передают не статью, а какой-то абстрактный бред, то мы не будем его классифицировать на тематики\n\n- Сделал возможность передавать ссылку на статью, чтобы руками не копировать тест\n\n- Есть замер таймингов работы\n\n- Достаточно красивый интерфейс","metadata":{}},{"cell_type":"code","source":"!pip3 install -q transformers datasets evaluate accelerate scikit-learn matplotlib joblib","metadata":{"execution":{"iopub.status.busy":"2026-04-07T18:16:47.422747Z","iopub.execute_input":"2026-04-07T18:16:47.423046Z","iopub.status.idle":"2026-04-07T18:16:54.242001Z","shell.execute_reply.started":"2026-04-07T18:16:47.423008Z","shell.execute_reply":"2026-04-07T18:16:54.241240Z"},"trusted":true},"outputs":[{"name":"stdout","text":"\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m84.1/84.1 kB\u001b[0m \u001b[31m4.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n\u001b[?25h","output_type":"stream"}],"execution_count":1},{"cell_type":"code","source":"# ── Data ──────────────────────────────────────────────────────────────────────\nDATASET_NAME: str = \"ccdv/arxiv-classification\"\n# https://huggingface.co/datasets/ccdv/arxiv-classification\nRANDOM_SEED: int = 42\n\n# ── Model ─────────────────────────────────────────────────────────────────────\nMODEL_NAME: str = \"distilbert-base-uncased\"\nMAX_SEQ_LENGTH: int = 512\n\n# ── Training ──────────────────────────────────────────────────────────────────\nLEARNING_RATE: float = 2e-5\nTRAIN_BATCH_SIZE: int = 16\nEVAL_BATCH_SIZE: int = 64\nNUM_EPOCHS: int = 3\nWEIGHT_DECAY: float = 0.01\n\n# ── Paths ─────────────────────────────────────────────────────────────────────\nOUTPUT_DIR: str = \"./results\"\nSAVE_PATH: str = \"./my_arxiv_classifier\"\nARCHIVE_NAME: str = \"my_arxiv_classifier\"","metadata":{"execution":{"iopub.status.busy":"2026-04-07T18:16:54.244027Z","iopub.execute_input":"2026-04-07T18:16:54.244374Z","iopub.status.idle":"2026-04-07T18:16:54.250244Z","shell.execute_reply.started":"2026-04-07T18:16:54.244345Z","shell.execute_reply":"2026-04-07T18:16:54.249666Z"},"trusted":true},"outputs":[],"execution_count":2},{"cell_type":"code","source":"from datasets import load_dataset\n\nraw_dataset = load_dataset(DATASET_NAME)\n\nprint(raw_dataset)","metadata":{"execution":{"iopub.status.busy":"2026-04-07T19:17:43.948372Z","iopub.execute_input":"2026-04-07T19:17:43.949188Z","iopub.status.idle":"2026-04-07T19:17:44.094425Z","shell.execute_reply.started":"2026-04-07T19:17:43.949154Z","shell.execute_reply":"2026-04-07T19:17:44.093653Z"},"trusted":true},"outputs":[{"name":"stdout","text":"DatasetDict({\n    train: Dataset({\n        features: ['text', 'label'],\n        num_rows: 28388\n    })\n    validation: Dataset({\n        features: ['text', 'label'],\n        num_rows: 2500\n    })\n    test: Dataset({\n        features: ['text', 'label'],\n        num_rows: 2500\n    })\n})\n","output_type":"stream"}],"execution_count":12},{"cell_type":"code","source":"from datasets import DatasetDict\n\nlabel_names: list[str] = raw_dataset[\"train\"].features[\"label\"].names\nNUM_CLASSES: int = len(label_names)\nlabel2id: dict[str, int] = {name: i for i, name in enumerate(label_names)}\nid2label: dict[int, str] = {i: name for i, name in enumerate(label_names)}\n\nhf_dataset = DatasetDict({\n    \"train\": raw_dataset[\"train\"],\n    \"test\": raw_dataset[\"validation\"],\n})\n\nprint(f\"Categories ({NUM_CLASSES}): {label_names}\")\nprint(hf_dataset)","metadata":{"execution":{"iopub.status.busy":"2026-04-07T18:17:48.226767Z","iopub.execute_input":"2026-04-07T18:17:48.227295Z","iopub.status.idle":"2026-04-07T18:17:48.233453Z","shell.execute_reply.started":"2026-04-07T18:17:48.227267Z","shell.execute_reply":"2026-04-07T18:17:48.232729Z"},"trusted":true},"outputs":[{"name":"stdout","text":"Categories (11): ['math.AC', 'cs.CV', 'cs.AI', 'cs.SY', 'math.GR', 'cs.CE', 'cs.PL', 'cs.IT', 'cs.DS', 'cs.NE', 'math.ST']\nDatasetDict({\n    train: Dataset({\n        features: ['text', 'label'],\n        num_rows: 28388\n    })\n    test: Dataset({\n        features: ['text', 'label'],\n        num_rows: 2500\n    })\n})\n","output_type":"stream"}],"execution_count":4},{"cell_type":"code","source":"import matplotlib.pyplot as plt\nfrom collections import Counter\n\ntrain_labels = raw_dataset[\"train\"][\"label\"]\ncounts = Counter(train_labels)\ncategory_counts = {id2label[k]: v for k, v in sorted(counts.items())}\n\nfig, axes = plt.subplots(1, 2, figsize=(14, 5))\nfig.suptitle(\"ArXiv Dataset — Category Distribution\", fontsize=14, fontweight=\"bold\")\n\nbars = axes[0].bar(category_counts.keys(), category_counts.values(), color=\"steelblue\", edgecolor=\"white\")\naxes[0].set_title(\"Papers per Category\")\naxes[0].set_xlabel(\"Category\")\naxes[0].set_ylabel(\"Count\")\naxes[0].tick_params(axis=\"x\", rotation=45)\nfor bar in bars:\n    axes[0].text(\n        bar.get_x() + bar.get_width() / 2,\n        bar.get_height() + 80,\n        f\"{int(bar.get_height()):,}\",\n        ha=\"center\", va=\"bottom\", fontsize=8,\n    )\n\naxes[1].pie(\n    category_counts.values(),\n    labels=category_counts.keys(),\n    autopct=\"%1.1f%%\",\n    startangle=140,\n    wedgeprops={\"edgecolor\": \"white\", \"linewidth\": 1.2},\n)\naxes[1].set_title(\"Share per Category\")\n\nplt.tight_layout()\nplt.show()\n\ntotal = sum(category_counts.values())\nfor name, count in sorted(category_counts.items(), key=lambda x: -x[1]):\n    print(f\"  {name}: {count:,} ({count/total*100:.1f}%)\")","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2026-04-07T18:17:48.234550Z","iopub.execute_input":"2026-04-07T18:17:48.234844Z","iopub.status.idle":"2026-04-07T18:17:49.594817Z","shell.execute_reply.started":"2026-04-07T18:17:48.234811Z","shell.execute_reply":"2026-04-07T18:17:49.594164Z"}},"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 1400x500 with 2 Axes>","image/png":"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\n"},"metadata":{}},{"name":"stdout","text":"  cs.DS: 3,527 (12.4%)\n  cs.IT: 2,768 (9.8%)\n  cs.SY: 2,631 (9.3%)\n  math.GR: 2,599 (9.2%)\n  math.ST: 2,581 (9.1%)\n  cs.AI: 2,569 (9.0%)\n  cs.NE: 2,560 (9.0%)\n  math.AC: 2,456 (8.7%)\n  cs.PL: 2,443 (8.6%)\n  cs.CV: 2,137 (7.5%)\n  cs.CE: 2,117 (7.5%)\n","output_type":"stream"}],"execution_count":5},{"cell_type":"code","source":"import numpy as np\nimport evaluate\nfrom transformers import AutoTokenizer\n\ntokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)\n\ndef tokenize_function(examples: dict) -> dict:\n    return tokenizer(\n        examples[\"text\"],\n        padding=\"max_length\",\n        truncation=True,\n        max_length=MAX_SEQ_LENGTH,\n    )\n\n\ntokenized_datasets = hf_dataset.map(tokenize_function, batched=True)\n\naccuracy_metric = evaluate.load(\"accuracy\")\nf1_metric = evaluate.load(\"f1\")\n\ndef compute_metrics(eval_pred: tuple) -> dict[str, float]:\n    logits, labels = eval_pred\n    predictions = np.argmax(logits, axis=-1)\n    acc = accuracy_metric.compute(predictions=predictions, references=labels)[\"accuracy\"]\n    f1 = f1_metric.compute(predictions=predictions, references=labels, average=\"macro\")[\"f1\"]\n    return {\"accuracy\": acc, \"f1_macro\": f1}\n","metadata":{"execution":{"iopub.status.busy":"2026-04-07T18:19:19.365465Z","iopub.execute_input":"2026-04-07T18:19:19.365832Z","iopub.status.idle":"2026-04-07T18:29:29.536891Z","shell.execute_reply.started":"2026-04-07T18:19:19.365804Z","shell.execute_reply":"2026-04-07T18:29:29.535887Z"},"trusted":true},"outputs":[{"output_type":"display_data","data":{"text/plain":"vocab.txt: 0.00B [00:00, ?B/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"360f51c4064744518bd1062b6a4fe0d0"}},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"tokenizer.json: 0.00B [00:00, ?B/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"352e664c5c2c43128e62d89991c1a976"}},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"Map:   0%|          | 0/28388 [00:00<?, ? examples/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"ab82532bd28b4103b29b59ffab0f372b"}},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"Map:   0%|          | 0/2500 [00:00<?, ? examples/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"831ce8abf8dd4fde893e03c0dc6ffb9c"}},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"Downloading builder script: 0.00B [00:00, ?B/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"ca17635e76a247099b730a5a80745839"}},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"Downloading builder script: 0.00B [00:00, ?B/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"3431d0ae79e74ceaaffbdb2421653454"}},"metadata":{}}],"execution_count":7},{"cell_type":"code","source":"import shutil\nimport torch\nfrom transformers import AutoModelForSequenceClassification, DataCollatorWithPadding, Trainer, TrainingArguments\n\n\ndef get_device() -> str:\n    if torch.cuda.is_available():\n        return \"cuda\"\n    if torch.backends.mps.is_available():\n        return \"mps\"\n    return \"cpu\"\n\n\nDEVICE = get_device()\nprint(f\"Training device: {DEVICE}\")\n\nmodel = AutoModelForSequenceClassification.from_pretrained(\n    MODEL_NAME,\n    num_labels=NUM_CLASSES,\n    id2label=id2label,\n    label2id=label2id,\n)\n\ntraining_args = TrainingArguments(\n    output_dir=OUTPUT_DIR,\n    learning_rate=LEARNING_RATE,\n    per_device_train_batch_size=TRAIN_BATCH_SIZE,\n    per_device_eval_batch_size=EVAL_BATCH_SIZE,\n    num_train_epochs=NUM_EPOCHS,\n    eval_strategy=\"epoch\",\n    save_strategy=\"epoch\",\n    load_best_model_at_end=True,\n    weight_decay=WEIGHT_DECAY,\n    report_to=\"none\",\n    use_cpu=(DEVICE == \"cpu\"),\n)\n\ntrainer = Trainer(\n    model=model,\n    args=training_args,\n    train_dataset=tokenized_datasets[\"train\"],\n    eval_dataset=tokenized_datasets[\"test\"],\n    compute_metrics=compute_metrics,\n    data_collator=DataCollatorWithPadding(tokenizer=tokenizer),\n    processing_class=tokenizer,\n)\n\ntrainer.train()\n\nresults = trainer.evaluate()\nprint(f\"Accuracy : {results['eval_accuracy']:.4f}\")\nprint(f\"F1 Macro : {results['eval_f1_macro']:.4f}\")\n\ntrainer.save_model(SAVE_PATH)\ntokenizer.save_pretrained(SAVE_PATH)\nshutil.make_archive(ARCHIVE_NAME, \"zip\", SAVE_PATH)\nprint(f\"\\nModel saved to {SAVE_PATH}/ and archived as {ARCHIVE_NAME}.zip\")\n","metadata":{"execution":{"iopub.status.busy":"2026-04-07T18:33:41.890369Z","iopub.execute_input":"2026-04-07T18:33:41.890880Z","iopub.status.idle":"2026-04-07T19:12:36.074300Z","shell.execute_reply.started":"2026-04-07T18:33:41.890847Z","shell.execute_reply":"2026-04-07T19:12:36.073359Z"},"trusted":true},"outputs":[{"name":"stdout","text":"Training device: cuda\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"model.safetensors:   0%|          | 0.00/268M [00:00<?, ?B/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"d5bbd8e7714d47f496e174cb62e9a95c"}},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"Loading weights:   0%|          | 0/100 [00:00<?, ?it/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"e6a03d4655864749822ce87b4b9f1ae9"}},"metadata":{}},{"name":"stderr","text":"DistilBertForSequenceClassification LOAD REPORT from: distilbert-base-uncased\nKey                     | Status     | \n------------------------+------------+-\nvocab_projector.bias    | UNEXPECTED | \nvocab_layer_norm.weight | UNEXPECTED | \nvocab_transform.weight  | UNEXPECTED | \nvocab_transform.bias    | UNEXPECTED | \nvocab_layer_norm.bias   | UNEXPECTED | \nclassifier.weight       | MISSING    | \nclassifier.bias         | MISSING    | \npre_classifier.bias     | MISSING    | \npre_classifier.weight   | MISSING    | \n\nNotes:\n- UNEXPECTED\t:can be ignored when loading from different task/architecture; not ok if you expect identical arch.\n- MISSING\t:those params were newly initialized because missing from the checkpoint. Consider training on your downstream task.\n/usr/local/lib/python3.12/dist-packages/torch/autograd/function.py:583: UserWarning: Was asked to gather along dimension 0, but all input tensors were scalars; will instead unsqueeze and return a vector.\n  return super().apply(*args, **kwargs)  # type: ignore[misc]\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<IPython.core.display.HTML object>","text/html":"\n    <div>\n      \n      <progress value='2664' max='2664' style='width:300px; height:20px; vertical-align: middle;'></progress>\n      [2664/2664 38:10, Epoch 3/3]\n    </div>\n    <table border=\"1\" class=\"dataframe\">\n  <thead>\n <tr style=\"text-align: left;\">\n      <th>Epoch</th>\n      <th>Training Loss</th>\n      <th>Validation Loss</th>\n      <th>Accuracy</th>\n      <th>F1 Macro</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <td>1</td>\n      <td>2.050387</td>\n      <td>1.034079</td>\n      <td>0.840800</td>\n      <td>0.836284</td>\n    </tr>\n    <tr>\n      <td>2</td>\n      <td>0.872269</td>\n      <td>0.874096</td>\n      <td>0.858400</td>\n      <td>0.855084</td>\n    </tr>\n    <tr>\n      <td>3</td>\n      <td>0.703794</td>\n      <td>0.862993</td>\n      <td>0.867600</td>\n      <td>0.864008</td>\n    </tr>\n  </tbody>\n</table><p>"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"d4bd6f1ab9ec40048f3aea82a16f8d12"}},"metadata":{}},{"name":"stderr","text":"/usr/local/lib/python3.12/dist-packages/torch/autograd/function.py:583: UserWarning: Was asked to gather along dimension 0, but all input tensors were scalars; will instead unsqueeze and return a vector.\n  return super().apply(*args, **kwargs)  # type: ignore[misc]\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"7705954cf5384c63a3dfa995d6e9c5f7"}},"metadata":{}},{"name":"stderr","text":"/usr/local/lib/python3.12/dist-packages/torch/autograd/function.py:583: UserWarning: Was asked to gather along dimension 0, but all input tensors were scalars; will instead unsqueeze and return a vector.\n  return super().apply(*args, **kwargs)  # type: ignore[misc]\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"ea81996b72874fc383289e50f087256f"}},"metadata":{}},{"name":"stderr","text":"There were missing keys in the checkpoint model loaded: ['distilbert.embeddings.LayerNorm.weight', 'distilbert.embeddings.LayerNorm.bias'].\nThere were unexpected keys in the checkpoint model loaded: ['distilbert.embeddings.LayerNorm.beta', 'distilbert.embeddings.LayerNorm.gamma'].\n/usr/local/lib/python3.12/dist-packages/torch/autograd/function.py:583: UserWarning: Was asked to gather along dimension 0, but all input tensors were scalars; will instead unsqueeze and return a vector.\n  return super().apply(*args, **kwargs)  # type: ignore[misc]\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<IPython.core.display.HTML object>","text/html":"\n    <div>\n      \n      <progress value='20' max='20' style='width:300px; height:20px; vertical-align: middle;'></progress>\n      [20/20 00:19]\n    </div>\n    "},"metadata":{}},{"name":"stdout","text":"Accuracy : 0.8676\nF1 Macro : 0.8640\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"Writing model shards:   0%|          | 0/1 [00:00<?, ?it/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"d3e4933c98ef4130892c7f9d1c97402e"}},"metadata":{}},{"name":"stdout","text":"\nModel saved to ./my_arxiv_classifier/ and archived as my_arxiv_classifier.zip\n","output_type":"stream"}],"execution_count":8},{"cell_type":"code","source":"from transformers import pipeline\n\nclassifier = pipeline(\"text-classification\", model=SAVE_PATH, top_k=None)\n\ntest_articles: list[dict[str, str]] = [\n    {\n        \"title\": \"Attention Is All You Need\",\n        \"abstract\": \"The dominant sequence transduction models are based on complex recurrent or convolutional neural networks that include an encoder and a decoder. We propose a new simple network architecture, the Transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely.\",\n    },\n    {\n        \"title\": \"Quantum Supremacy Using a Programmable Superconducting Processor\",\n        \"abstract\": \"The promise of quantum computers is that certain computational tasks might be executed exponentially faster on a quantum processor than on a classical processor. We report the use of a processor with programmable superconducting qubits to create quantum states on 53 qubits.\",\n    },\n    {\n        \"title\": \"chemistry of water\",\n        \"abstract\": \"chemistry of water.\",\n    },\n]\n\n\ndef get_top95_predictions(text: str) -> list[tuple[str, float]]:\n    predictions = sorted(classifier(text)[0], key=lambda x: x[\"score\"], reverse=True)\n    result: list[tuple[str, float]] = []\n    cumulative = 0.0\n    for pred in predictions:\n        result.append((pred[\"label\"], pred[\"score\"]))\n        cumulative += pred[\"score\"]\n        if cumulative >= 0.95:\n            break\n    return result\n\n\nfor i, article in enumerate(test_articles, 1):\n    input_text = f\"{article['title']}. {article['abstract']}\"\n    top_preds = get_top95_predictions(input_text)\n    print(f\"Article {i}: {article['title']}\")\n    for label, score in top_preds:\n        print(f\"  {label.upper()}: {score * 100:.2f}%\")\n    print(\"-\" * 50)","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2026-04-07T19:12:36.075736Z","iopub.execute_input":"2026-04-07T19:12:36.076075Z","iopub.status.idle":"2026-04-07T19:12:36.691920Z","shell.execute_reply.started":"2026-04-07T19:12:36.076047Z","shell.execute_reply":"2026-04-07T19:12:36.691146Z"}},"outputs":[{"output_type":"display_data","data":{"text/plain":"Loading weights:   0%|          | 0/104 [00:00<?, ?it/s]","application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"7a0049241d9f424f8a6f42ffd099747f"}},"metadata":{}},{"name":"stdout","text":"Article 1: Attention Is All You Need\n  CS.NE: 94.11%\n  CS.AI: 3.42%\n--------------------------------------------------\nArticle 2: Quantum Supremacy Using a Programmable Superconducting Processor\n  CS.PL: 48.76%\n  CS.DS: 40.72%\n  CS.AI: 4.53%\n  CS.CE: 1.85%\n--------------------------------------------------\nArticle 3: chemistry of water\n  CS.CE: 77.54%\n  CS.SY: 5.05%\n  MATH.GR: 3.79%\n  MATH.AC: 2.75%\n  MATH.ST: 2.36%\n  CS.DS: 1.75%\n  CS.NE: 1.65%\n  CS.PL: 1.49%\n--------------------------------------------------\n","output_type":"stream"}],"execution_count":9},{"cell_type":"code","source":"import pandas as pd\nfrom datasets import load_dataset\nfrom sklearn.feature_extraction.text import TfidfVectorizer\nfrom sklearn.linear_model import LogisticRegression\nfrom sklearn.pipeline import Pipeline\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import classification_report\nimport joblib\n\n\narxiv = load_dataset(\"ccdv/arxiv-classification\", split=\"train[:10000]\")\ndf_science = pd.DataFrame({\"text\": arxiv[\"text\"], \"label\": 1})\n\nimdb = load_dataset(\"imdb\", split=\"train[:5000]\")       # Отзывы\n\ndf_imdb = pd.DataFrame({\"text\": imdb[\"text\"], \"label\": 0})\n\ndf = pd.concat([df_science, df_imdb]).sample(frac=1.0, random_state=42).reset_index(drop=True)\n\nX_train, X_test, y_train, y_test = train_test_split(df['text'], df['label'], test_size=0.1, random_state=42)\n\nprint(\"Обучаем TF-IDF + Logistic Regression\")\npipeline = Pipeline([\n    ('tfidf', TfidfVectorizer(max_features=15000, ngram_range=(1, 2))), # максимум 15к слов\n    ('clf', LogisticRegression(max_iter=1000, C=1.0))\n])\n\npipeline.fit(X_train, y_train)\n\nprint(\"\\nОтчет о классификации (Test Data):\")\nprint(classification_report(y_test, pipeline.predict(X_test)))\n\n\ntests = [\n    \"We propose a novel attention mechanism for deep neural networks that reduces complexity.\", # Наука (1)\n    \"I went to buy some milk and the store was closed! Can you believe this?\",                # Жизнь (0)\n    \"The new iPhone features an amazing camera and long lasting battery life.\"                # Новости (0)\n]\n\nfor text, pred in zip(tests, pipeline.predict(tests)):\n    print(f\"[{'НАУКА' if pred==1 else 'МУСОР'}] - {text}\")\n\njoblib.dump(pipeline, 'ood_detector.pkl')\nprint(\"сохранен в файл 'ood_detector.pkl'!\")\n","metadata":{"trusted":true,"execution":{"iopub.status.busy":"2026-04-07T19:23:06.940557Z","iopub.execute_input":"2026-04-07T19:23:06.941274Z","iopub.status.idle":"2026-04-07T19:26:16.689086Z","shell.execute_reply.started":"2026-04-07T19:23:06.941239Z","shell.execute_reply":"2026-04-07T19:26:16.688193Z"}},"outputs":[{"name":"stdout","text":"Обучаем TF-IDF + Logistic Regression\n\nОтчет о классификации (Test Data):\n              precision    recall  f1-score   support\n\n           0       1.00      1.00      1.00       510\n           1       1.00      1.00      1.00       990\n\n    accuracy                           1.00      1500\n   macro avg       1.00      1.00      1.00      1500\nweighted avg       1.00      1.00      1.00      1500\n\n[НАУКА] - We propose a novel attention mechanism for deep neural networks that reduces complexity.\n[МУСОР] - I went to buy some milk and the store was closed! Can you believe this?\n[МУСОР] - The new iPhone features an amazing camera and long lasting battery life.\nсохранен в файл 'ood_detector.pkl'!\n","output_type":"stream"}],"execution_count":13},{"cell_type":"code","source":"","metadata":{"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"code","source":"","metadata":{"trusted":true},"outputs":[],"execution_count":null}]}