training_loop.ipynb

{
 "metadata": {
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  },
  "orig_nbformat": 2,
  "kernelspec": {
   "name": "ml",
   "display_name": "ML",
   "language": "python"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2,
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "output_type": "stream",
     "name": "stderr",
     "text": [
      "Epoch 0: 100%|██████████| 12500/12500 [1:29:47<00:00,  2.32it/s, loss=0.358]\n",
      "Epoch 1: 100%|██████████| 12500/12500 [1:22:20<00:00,  2.53it/s, loss=0.31]\n"
     ]
    }
   ],
   "source": [
    "epochs = 2\n",
    "\n",
    "for epoch in range(epochs):\n",
    "    # setup loop with TQDM and dataloader\n",
    "    loop = tqdm(loader, leave=True)\n",
    "    for batch in loop:\n",
    "        # initialize calculated gradients (from prev step)\n",
    "        optim.zero_grad()\n",
    "        # pull all tensor batches required for training\n",
    "        input_ids = batch['input_ids'].to(device)\n",
    "        attention_mask = batch['attention_mask'].to(device)\n",
    "        labels = batch['labels'].to(device)\n",
    "        # process\n",
    "        outputs = model(input_ids, attention_mask=attention_mask,\n",
    "                        labels=labels)\n",
    "        # extract loss\n",
    "        loss = outputs.loss\n",
    "        # calculate loss for every parameter that needs grad update\n",
    "        loss.backward()\n",
    "        # update parameters\n",
    "        optim.step()\n",
    "        # print relevant info to progress bar\n",
    "        loop.set_description(f'Epoch {epoch}')\n",
    "        loop.set_postfix(loss=loss.item())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save_pretrained('./filiberto')  # and don't forget to save filiBERTo!"
   ]
  }
 ]
}