done

challange1.ipynb

@@ -0,0 +1,237 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from datasets import load_dataset, DatasetDict\n",
+    "from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, DataCollatorForSeq2Seq, Seq2SeqTrainingArguments, Seq2SeqTrainer\n",
+    "from sklearn.model_selection import train_test_split\n",
+    "from huggingface_hub import notebook_login\n",
+    "import torch\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "619aead440a04dbd9eafa156e3713251",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "VBox(children=(HTML(value='<center> <img\\nsrc=https://huggingface.co/front/assets/huggingface_logo-noborder.sv…"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "notebook_login()\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "dataset = load_dataset(\"SKNahin/bengali-transliteration-data\")\n",
+    "\n",
+    "dataset = dataset[\"train\"].train_test_split(test_size=0.2, seed=42)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "DatasetDict({\n",
+      "    train: Dataset({\n",
+      "        features: ['bn', 'rm'],\n",
+      "        num_rows: 4004\n",
+      "    })\n",
+      "    validation: Dataset({\n",
+      "        features: ['bn', 'rm'],\n",
+      "        num_rows: 1002\n",
+      "    })\n",
+      "})\n"
+     ]
+    }
+   ],
+   "source": [
+    "dataset = DatasetDict({\n",
+    "    \"train\": dataset[\"train\"],\n",
+    "    \"validation\": dataset[\"test\"]\n",
+    "})\n",
+    "print(dataset)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from transformers import MBartForConditionalGeneration, MBart50TokenizerFast\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "NameError",
+     "evalue": "name 'MBartForConditionalGeneration' is not defined",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
+      "Cell \u001b[1;32mIn[1], line 2\u001b[0m\n\u001b[0;32m      1\u001b[0m model_name\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mfacebook/mbart-large-50-many-to-many-mmt\u001b[39m\u001b[38;5;124m'\u001b[39m\n\u001b[1;32m----> 2\u001b[0m model \u001b[38;5;241m=\u001b[39m \u001b[43mMBartForConditionalGeneration\u001b[49m\u001b[38;5;241m.\u001b[39mfrom_pretrained(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfacebook/mbart-large-50-many-to-many-mmt\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m      3\u001b[0m tokenizer \u001b[38;5;241m=\u001b[39m MBart50TokenizerFast\u001b[38;5;241m.\u001b[39mfrom_pretrained(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfacebook/mbart-large-50-many-to-many-mmt\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m      4\u001b[0m \u001b[38;5;66;03m#this process took huge time as it is very big compared to the internet bandwith, so the next steps could not be run until its finishing\u001b[39;00m\n",
+      "\u001b[1;31mNameError\u001b[0m: name 'MBartForConditionalGeneration' is not defined"
+     ]
+    }
+   ],
+   "source": [
+    "model_name='facebook/mbart-large-50-many-to-many-mmt'\n",
+    "model = MBartForConditionalGeneration.from_pretrained(\"facebook/mbart-large-50-many-to-many-mmt\")\n",
+    "tokenizer = MBart50TokenizerFast.from_pretrained(\"facebook/mbart-large-50-many-to-many-mmt\")\n",
+    "#this process took huge time as it is very big compared to the internet bandwith, and it is still running so the next steps could not be run until its finishing\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def preprocess_function(examples):\n",
+    "    inputs = examples[\"banglish\"]\n",
+    "    targets = examples[\"bangla\"]\n",
+    "    model_inputs = tokenizer(inputs, max_length=128, truncation=True, padding=\"max_length\")\n",
+    "    with tokenizer.as_target_tokenizer():\n",
+    "        labels = tokenizer(targets, max_length=128, truncation=True, padding=\"max_length\")\n",
+    "    model_inputs[\"labels\"] = labels[\"input_ids\"]\n",
+    "    return model_inputs"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "tokenized_datasets = dataset.map(preprocess_function, batched=True, remove_columns=dataset[\"train\"].column_names)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "model = AutoModelForSeq2SeqLM.from_pretrained(model_name)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "data_collator = DataCollatorForSeq2Seq(tokenizer, model=model)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "training_args = Seq2SeqTrainingArguments(\n",
+    "    output_dir=\"./results\",\n",
+    "    evaluation_strategy=\"epoch\",\n",
+    "    learning_rate=2e-5,\n",
+    "    per_device_train_batch_size=16,\n",
+    "    per_device_eval_batch_size=16,\n",
+    "    weight_decay=0.01,\n",
+    "    save_total_limit=2,\n",
+    "    num_train_epochs=5,\n",
+    "    predict_with_generate=True,\n",
+    "    logging_dir=\"./logs\",\n",
+    "    logging_strategy=\"epoch\",\n",
+    "    save_strategy=\"epoch\"\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "trainer = Seq2SeqTrainer(\n",
+    "    model=model,\n",
+    "    args=training_args,\n",
+    "    train_dataset=tokenized_datasets[\"train\"],\n",
+    "    eval_dataset=tokenized_datasets[\"validation\"],\n",
+    "    tokenizer=tokenizer,\n",
+    "    data_collator=data_collator\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "#training start using the provided dataset for fine tuning\n",
+    "trainer.train()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "model.save_pretrained(\"banglish-to-bangla-model\")\n",
+    "tokenizer.save_pretrained(\"banglish-to-bangla-model\")"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": ".venv",
+   "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": "3.12.3"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}