mt5.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c7a27d0a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 1;\n",
       "                var nbb_unformatted_code = \"%load_ext nb_black\";\n",
       "                var nbb_formatted_code = \"%load_ext nb_black\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%load_ext nb_black"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e507e35b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Found cached dataset generator (C:/tempjeka/cache/huggingface/generator/default-f628ff2f2d994b84/0.0.0)\n",
      "Loading cached split indices for dataset at C:\\tempjeka\\cache\\huggingface\\generator\\default-f628ff2f2d994b84\\0.0.0\\cache-a84cfc0a0e11ed54.arrow and C:\\tempjeka\\cache\\huggingface\\generator\\default-f628ff2f2d994b84\\0.0.0\\cache-665145fe0a5a6a51.arrow\n",
      "Loading cached split indices for dataset at C:\\tempjeka\\cache\\huggingface\\generator\\default-f628ff2f2d994b84\\0.0.0\\cache-e104a2bffe25c7fa.arrow and C:\\tempjeka\\cache\\huggingface\\generator\\default-f628ff2f2d994b84\\0.0.0\\cache-4096519e2393e19f.arrow\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 2;\n",
       "                var nbb_unformatted_code = \"from src.env import *\\nfrom src.imports import *\\nfrom src.config import *\\nfrom src.dataset import *\\nfrom src.metric import *\";\n",
       "                var nbb_formatted_code = \"from src.env import *\\nfrom src.imports import *\\nfrom src.config import *\\nfrom src.dataset import *\\nfrom src.metric import *\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from src.env import *\n",
    "from src.imports import *\n",
    "from src.config import *\n",
    "from src.dataset import *\n",
    "from src.metric import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "83d3db35",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 3;\n",
       "                var nbb_unformatted_code = \"enc_name = \\\"google/mt5-small\\\"\\nmodel_name = \\\"mt5_small\\\"\";\n",
       "                var nbb_formatted_code = \"enc_name = \\\"google/mt5-small\\\"\\nmodel_name = \\\"mt5_small\\\"\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "enc_name = \"google/mt5-small\"\n",
    "model_name = \"mt5_small\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "56ccf276",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\tempjeka\\anaconda3\\lib\\site-packages\\transformers\\convert_slow_tokenizer.py:454: UserWarning: The sentencepiece tokenizer that you are converting to a fast tokenizer uses the byte fallback option which is not implemented in the fast tokenizers. In practice this means that the fast version of the tokenizer can produce unknown tokens whereas the sentencepiece version would have converted these unknown tokens into a sequence of byte tokens matching the original piece of text.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 4;\n",
       "                var nbb_unformatted_code = \"enc_tokenizer = AutoTokenizer.from_pretrained(enc_name)\";\n",
       "                var nbb_formatted_code = \"enc_tokenizer = AutoTokenizer.from_pretrained(enc_name)\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "enc_tokenizer = AutoTokenizer.from_pretrained(enc_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "a088a0ee",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 5;\n",
       "                var nbb_unformatted_code = \"dec_tokenizer = enc_tokenizer\";\n",
       "                var nbb_formatted_code = \"dec_tokenizer = enc_tokenizer\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dec_tokenizer = enc_tokenizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "532b58e3",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Loading cached processed dataset at C:\\tempjeka\\cache\\huggingface\\generator\\default-f628ff2f2d994b84\\0.0.0\\cache-e8e63b2a8a7e4ad3.arrow\n",
      "Loading cached processed dataset at C:\\tempjeka\\cache\\huggingface\\generator\\default-f628ff2f2d994b84\\0.0.0\\cache-d6ceae0a37a67c62.arrow\n",
      "Loading cached processed dataset at C:\\tempjeka\\cache\\huggingface\\generator\\default-f628ff2f2d994b84\\0.0.0\\cache-0c3f31464f8f7989.arrow\n",
      "Loading cached sorted indices for dataset at C:\\tempjeka\\cache\\huggingface\\generator\\default-f628ff2f2d994b84\\0.0.0\\cache-bdb4805bd3c0c42c.arrow\n",
      "Loading cached sorted indices for dataset at C:\\tempjeka\\cache\\huggingface\\generator\\default-f628ff2f2d994b84\\0.0.0\\cache-fc809c3039eedd94.arrow\n",
      "Loading cached sorted indices for dataset at C:\\tempjeka\\cache\\huggingface\\generator\\default-f628ff2f2d994b84\\0.0.0\\cache-733901f9d738dbc2.arrow\n"
     ]
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 6;\n",
       "                var nbb_unformatted_code = \"dataset = create_dataset(enc_tokenizer, dec_tokenizer).sort(\\\"input_ids_length\\\")\";\n",
       "                var nbb_formatted_code = \"dataset = create_dataset(enc_tokenizer, dec_tokenizer).sort(\\\"input_ids_length\\\")\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dataset = create_dataset(enc_tokenizer, dec_tokenizer).sort(\"input_ids_length\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c222954f",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 7;\n",
       "                var nbb_unformatted_code = \"from transformers import AutoModel\\n\\nmodel = AutoModelForSeq2SeqLM.from_pretrained(enc_name)\";\n",
       "                var nbb_formatted_code = \"from transformers import AutoModel\\n\\nmodel = AutoModelForSeq2SeqLM.from_pretrained(enc_name)\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from transformers import AutoModel\n",
    "\n",
    "model = AutoModelForSeq2SeqLM.from_pretrained(enc_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "478ee05e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 8;\n",
       "                var nbb_unformatted_code = \"model.config.max_length = max_input_length\\nmodel.config.min_length = 10\\nmodel.config.no_repeat_ngram_size = 3\\nmodel.config.early_stopping = True\\nmodel.config.num_beams = 10\";\n",
       "                var nbb_formatted_code = \"model.config.max_length = max_input_length\\nmodel.config.min_length = 10\\nmodel.config.no_repeat_ngram_size = 3\\nmodel.config.early_stopping = True\\nmodel.config.num_beams = 10\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.config.max_length = max_input_length\n",
    "model.config.min_length = 10\n",
    "model.config.no_repeat_ngram_size = 3\n",
    "model.config.early_stopping = True\n",
    "model.config.num_beams = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "5ce6805a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 9;\n",
       "                var nbb_unformatted_code = \"version = 1\";\n",
       "                var nbb_formatted_code = \"version = 1\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "version = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "6adb2d8a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 10;\n",
       "                var nbb_unformatted_code = \"version += 1\\nargs = Seq2SeqTrainingArguments(\\n    output_dir=Path(\\\"models\\\") / model_name,\\n    evaluation_strategy=\\\"steps\\\",\\n    learning_rate=5e-5,\\n    #     lr_scheduler_type=\\\"constant_with_warmup\\\",\\n    per_device_train_batch_size=batch_size,\\n    per_device_eval_batch_size=batch_size,\\n    save_total_limit=1,\\n    num_train_epochs=4,\\n    predict_with_generate=True,\\n    warmup_steps=3000,\\n    fp16=False,\\n    logging_dir=f\\\"../made/logs/lightning_logs/{model_name}/{version}\\\",\\n    logging_first_step=False,\\n    eval_steps=1000,\\n    logging_steps=500,\\n    overwrite_output_dir=True,\\n    group_by_length=True,\\n)\";\n",
       "                var nbb_formatted_code = \"version += 1\\nargs = Seq2SeqTrainingArguments(\\n    output_dir=Path(\\\"models\\\") / model_name,\\n    evaluation_strategy=\\\"steps\\\",\\n    learning_rate=5e-5,\\n    #     lr_scheduler_type=\\\"constant_with_warmup\\\",\\n    per_device_train_batch_size=batch_size,\\n    per_device_eval_batch_size=batch_size,\\n    save_total_limit=1,\\n    num_train_epochs=4,\\n    predict_with_generate=True,\\n    warmup_steps=3000,\\n    fp16=False,\\n    logging_dir=f\\\"../made/logs/lightning_logs/{model_name}/{version}\\\",\\n    logging_first_step=False,\\n    eval_steps=1000,\\n    logging_steps=500,\\n    overwrite_output_dir=True,\\n    group_by_length=True,\\n)\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "version += 1\n",
    "args = Seq2SeqTrainingArguments(\n",
    "    output_dir=Path(\"models\") / model_name,\n",
    "    evaluation_strategy=\"steps\",\n",
    "    learning_rate=5e-5,\n",
    "    #     lr_scheduler_type=\"constant_with_warmup\",\n",
    "    per_device_train_batch_size=batch_size,\n",
    "    per_device_eval_batch_size=batch_size,\n",
    "    save_total_limit=1,\n",
    "    num_train_epochs=4,\n",
    "    predict_with_generate=True,\n",
    "    warmup_steps=3000,\n",
    "    fp16=False,\n",
    "    logging_dir=f\"../made/logs/lightning_logs/{model_name}/{version}\",\n",
    "    logging_first_step=False,\n",
    "    eval_steps=1000,\n",
    "    logging_steps=500,\n",
    "    overwrite_output_dir=True,\n",
    "    group_by_length=True,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "d5738285",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 11;\n",
       "                var nbb_unformatted_code = \"trainer = Seq2SeqTrainer(\\n    model=model,\\n    args=args,\\n    compute_metrics=partial(compute_metrics, dec_tokenizer, metric),\\n    train_dataset=dataset[\\\"train\\\"],\\n    eval_dataset=dataset[\\\"val\\\"],\\n    data_collator=DataCollatorForSeq2Seq(dec_tokenizer, model=model),\\n)\";\n",
       "                var nbb_formatted_code = \"trainer = Seq2SeqTrainer(\\n    model=model,\\n    args=args,\\n    compute_metrics=partial(compute_metrics, dec_tokenizer, metric),\\n    train_dataset=dataset[\\\"train\\\"],\\n    eval_dataset=dataset[\\\"val\\\"],\\n    data_collator=DataCollatorForSeq2Seq(dec_tokenizer, model=model),\\n)\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "trainer = Seq2SeqTrainer(\n",
    "    model=model,\n",
    "    args=args,\n",
    "    compute_metrics=partial(compute_metrics, dec_tokenizer, metric),\n",
    "    train_dataset=dataset[\"train\"],\n",
    "    eval_dataset=dataset[\"val\"],\n",
    "    data_collator=DataCollatorForSeq2Seq(dec_tokenizer, model=model),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "520bc29c",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\tempjeka\\anaconda3\\lib\\site-packages\\transformers\\optimization.py:391: FutureWarning: This implementation of AdamW is deprecated and will be removed in a future version. Use the PyTorch implementation torch.optim.AdamW instead, or set `no_deprecation_warning=True` to disable this warning\n",
      "  warnings.warn(\n",
      "You're using a T5TokenizerFast tokenizer. Please note that with a fast tokenizer, using the `__call__` method is faster than using a method to encode the text followed by a call to the `pad` method to get a padded encoding.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "    <div>\n",
       "      \n",
       "      <progress value='45000' max='45000' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
       "      [45000/45000 2:56:20, Epoch 4/4]\n",
       "    </div>\n",
       "    <table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       " <tr style=\"text-align: left;\">\n",
       "      <th>Step</th>\n",
       "      <th>Training Loss</th>\n",
       "      <th>Validation Loss</th>\n",
       "      <th>Bleu</th>\n",
       "      <th>Gen Len</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>1000</td>\n",
       "      <td>16.670000</td>\n",
       "      <td>7.664755</td>\n",
       "      <td>0.014361</td>\n",
       "      <td>43.220000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2000</td>\n",
       "      <td>4.359400</td>\n",
       "      <td>2.134417</td>\n",
       "      <td>0.022299</td>\n",
       "      <td>47.840000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3000</td>\n",
       "      <td>2.913600</td>\n",
       "      <td>1.695495</td>\n",
       "      <td>0.080851</td>\n",
       "      <td>63.060000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4000</td>\n",
       "      <td>2.497700</td>\n",
       "      <td>1.483175</td>\n",
       "      <td>0.131925</td>\n",
       "      <td>63.340000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5000</td>\n",
       "      <td>2.245500</td>\n",
       "      <td>1.354940</td>\n",
       "      <td>0.134872</td>\n",
       "      <td>65.520000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6000</td>\n",
       "      <td>2.121300</td>\n",
       "      <td>1.282575</td>\n",
       "      <td>0.140392</td>\n",
       "      <td>69.660000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7000</td>\n",
       "      <td>1.965500</td>\n",
       "      <td>1.199819</td>\n",
       "      <td>0.157548</td>\n",
       "      <td>74.600000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8000</td>\n",
       "      <td>1.886700</td>\n",
       "      <td>1.144614</td>\n",
       "      <td>0.153701</td>\n",
       "      <td>70.920000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9000</td>\n",
       "      <td>1.834500</td>\n",
       "      <td>1.083573</td>\n",
       "      <td>0.172849</td>\n",
       "      <td>72.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10000</td>\n",
       "      <td>1.781900</td>\n",
       "      <td>1.058200</td>\n",
       "      <td>0.214244</td>\n",
       "      <td>75.180000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>11000</td>\n",
       "      <td>1.743800</td>\n",
       "      <td>1.038370</td>\n",
       "      <td>0.216212</td>\n",
       "      <td>77.420000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>12000</td>\n",
       "      <td>1.653600</td>\n",
       "      <td>1.011461</td>\n",
       "      <td>0.202337</td>\n",
       "      <td>74.120000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>13000</td>\n",
       "      <td>1.623500</td>\n",
       "      <td>0.970160</td>\n",
       "      <td>0.239447</td>\n",
       "      <td>79.020000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>14000</td>\n",
       "      <td>1.641500</td>\n",
       "      <td>0.965437</td>\n",
       "      <td>0.233701</td>\n",
       "      <td>76.180000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>15000</td>\n",
       "      <td>1.576800</td>\n",
       "      <td>0.948105</td>\n",
       "      <td>0.251259</td>\n",
       "      <td>80.760000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>16000</td>\n",
       "      <td>1.530100</td>\n",
       "      <td>0.935343</td>\n",
       "      <td>0.251600</td>\n",
       "      <td>75.980000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17000</td>\n",
       "      <td>1.516100</td>\n",
       "      <td>0.923528</td>\n",
       "      <td>0.272303</td>\n",
       "      <td>80.520000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>18000</td>\n",
       "      <td>1.467300</td>\n",
       "      <td>0.911642</td>\n",
       "      <td>0.275436</td>\n",
       "      <td>78.080000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>19000</td>\n",
       "      <td>1.484100</td>\n",
       "      <td>0.897832</td>\n",
       "      <td>0.281630</td>\n",
       "      <td>79.560000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>20000</td>\n",
       "      <td>1.443600</td>\n",
       "      <td>0.888290</td>\n",
       "      <td>0.284184</td>\n",
       "      <td>83.300000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>21000</td>\n",
       "      <td>1.407500</td>\n",
       "      <td>0.876297</td>\n",
       "      <td>0.280895</td>\n",
       "      <td>79.920000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>22000</td>\n",
       "      <td>1.411800</td>\n",
       "      <td>0.864952</td>\n",
       "      <td>0.317760</td>\n",
       "      <td>84.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>23000</td>\n",
       "      <td>1.386400</td>\n",
       "      <td>0.858435</td>\n",
       "      <td>0.314591</td>\n",
       "      <td>83.580000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>24000</td>\n",
       "      <td>1.353600</td>\n",
       "      <td>0.854307</td>\n",
       "      <td>0.289974</td>\n",
       "      <td>82.600000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>25000</td>\n",
       "      <td>1.386100</td>\n",
       "      <td>0.852438</td>\n",
       "      <td>0.290594</td>\n",
       "      <td>81.760000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>26000</td>\n",
       "      <td>1.353500</td>\n",
       "      <td>0.841216</td>\n",
       "      <td>0.317286</td>\n",
       "      <td>84.160000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>27000</td>\n",
       "      <td>1.344400</td>\n",
       "      <td>0.838917</td>\n",
       "      <td>0.310416</td>\n",
       "      <td>81.720000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>28000</td>\n",
       "      <td>1.317500</td>\n",
       "      <td>0.834349</td>\n",
       "      <td>0.346344</td>\n",
       "      <td>86.320000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>29000</td>\n",
       "      <td>1.323300</td>\n",
       "      <td>0.823217</td>\n",
       "      <td>0.336400</td>\n",
       "      <td>84.860000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>30000</td>\n",
       "      <td>1.311200</td>\n",
       "      <td>0.830775</td>\n",
       "      <td>0.303155</td>\n",
       "      <td>84.620000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>31000</td>\n",
       "      <td>1.328700</td>\n",
       "      <td>0.820964</td>\n",
       "      <td>0.309328</td>\n",
       "      <td>83.660000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>32000</td>\n",
       "      <td>1.304000</td>\n",
       "      <td>0.817610</td>\n",
       "      <td>0.311445</td>\n",
       "      <td>83.740000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>33000</td>\n",
       "      <td>1.307000</td>\n",
       "      <td>0.811262</td>\n",
       "      <td>0.307775</td>\n",
       "      <td>83.200000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>34000</td>\n",
       "      <td>1.286000</td>\n",
       "      <td>0.807731</td>\n",
       "      <td>0.306075</td>\n",
       "      <td>84.080000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>35000</td>\n",
       "      <td>1.272600</td>\n",
       "      <td>0.805498</td>\n",
       "      <td>0.295825</td>\n",
       "      <td>83.800000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>36000</td>\n",
       "      <td>1.271500</td>\n",
       "      <td>0.805610</td>\n",
       "      <td>0.312034</td>\n",
       "      <td>86.180000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>37000</td>\n",
       "      <td>1.274900</td>\n",
       "      <td>0.804061</td>\n",
       "      <td>0.305637</td>\n",
       "      <td>84.520000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>38000</td>\n",
       "      <td>1.256600</td>\n",
       "      <td>0.798698</td>\n",
       "      <td>0.310537</td>\n",
       "      <td>86.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>39000</td>\n",
       "      <td>1.264400</td>\n",
       "      <td>0.795822</td>\n",
       "      <td>0.311806</td>\n",
       "      <td>84.220000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>40000</td>\n",
       "      <td>1.235900</td>\n",
       "      <td>0.796523</td>\n",
       "      <td>0.318747</td>\n",
       "      <td>85.240000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>41000</td>\n",
       "      <td>1.243300</td>\n",
       "      <td>0.796051</td>\n",
       "      <td>0.315118</td>\n",
       "      <td>84.580000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>42000</td>\n",
       "      <td>1.260000</td>\n",
       "      <td>0.794373</td>\n",
       "      <td>0.310994</td>\n",
       "      <td>86.160000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>43000</td>\n",
       "      <td>1.272900</td>\n",
       "      <td>0.792673</td>\n",
       "      <td>0.312895</td>\n",
       "      <td>84.100000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>44000</td>\n",
       "      <td>1.261600</td>\n",
       "      <td>0.791305</td>\n",
       "      <td>0.312813</td>\n",
       "      <td>84.920000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>45000</td>\n",
       "      <td>1.248200</td>\n",
       "      <td>0.791375</td>\n",
       "      <td>0.312108</td>\n",
       "      <td>85.600000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table><p>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\tempjeka\\anaconda3\\lib\\site-packages\\transformers\\generation\\utils.py:1219: UserWarning: You have modified the pretrained model configuration to control generation. This is a deprecated strategy to control generation and will be removed soon, in a future version. Please use a generation configuration file (see https://huggingface.co/docs/transformers/main_classes/text_generation)\n",
      "  warnings.warn(\n",
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n",
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n",
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "TrainOutput(global_step=45000, training_loss=2.0930444566514756, metrics={'train_runtime': 10583.3098, 'train_samples_per_second': 17.008, 'train_steps_per_second': 4.252, 'total_flos': 4253584720281600.0, 'train_loss': 2.0930444566514756, 'epoch': 4.0})"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 12;\n",
       "                var nbb_unformatted_code = \"trainer.train()\";\n",
       "                var nbb_formatted_code = \"trainer.train()\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "trainer.train()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "077a277c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 13;\n",
       "                var nbb_unformatted_code = \"model.save_pretrained(model_name)\";\n",
       "                var nbb_formatted_code = \"model.save_pretrained(model_name)\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.save_pretrained(model_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "6bd470bf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 7;\n",
       "                var nbb_unformatted_code = \"model = EncoderDecoderModel.from_pretrained(model_name)\";\n",
       "                var nbb_formatted_code = \"model = EncoderDecoderModel.from_pretrained(model_name)\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model = EncoderDecoderModel.from_pretrained(model_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "7c9c1727",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/4950 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 15;\n",
       "                var nbb_unformatted_code = \"metric = evaluate.load(\\\"bleu\\\")\\nmodel.cuda()\\nmodel.eval()\\nwith torch.no_grad():\\n    dataset[\\\"test\\\"].map(\\n        partial(\\n            compute_metrics_test,\\n            model,\\n            DataCollatorWithPadding(enc_tokenizer),\\n            dec_tokenizer,\\n            metric,\\n            log=False,\\n        ),\\n        batched=True,\\n        batch_size=batch_size,\\n    )\";\n",
       "                var nbb_formatted_code = \"metric = evaluate.load(\\\"bleu\\\")\\nmodel.cuda()\\nmodel.eval()\\nwith torch.no_grad():\\n    dataset[\\\"test\\\"].map(\\n        partial(\\n            compute_metrics_test,\\n            model,\\n            DataCollatorWithPadding(enc_tokenizer),\\n            dec_tokenizer,\\n            metric,\\n            log=False,\\n        ),\\n        batched=True,\\n        batch_size=batch_size,\\n    )\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "metric = evaluate.load(\"bleu\")\n",
    "model.cuda()\n",
    "model.eval()\n",
    "with torch.no_grad():\n",
    "    dataset[\"test\"].map(\n",
    "        partial(\n",
    "            compute_metrics_test,\n",
    "            model,\n",
    "            DataCollatorWithPadding(enc_tokenizer),\n",
    "            dec_tokenizer,\n",
    "            metric,\n",
    "            log=False,\n",
    "        ),\n",
    "        batched=True,\n",
    "        batch_size=batch_size,\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "02cdeb75",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'bleu': 0.2802308261876972,\n",
       " 'precisions': [0.5835117292175033,\n",
       "  0.37066080292324854,\n",
       "  0.26366016977451273,\n",
       "  0.18787657125257465],\n",
       " 'brevity_penalty': 0.8710238680265673,\n",
       " 'length_ratio': 0.8786682977519961,\n",
       " 'translation_length': 66799,\n",
       " 'reference_length': 76023}"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 16;\n",
       "                var nbb_unformatted_code = \"metric.compute()\";\n",
       "                var nbb_formatted_code = \"metric.compute()\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "metric.compute()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "996e4891",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/16 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\tempjeka\\anaconda3\\lib\\site-packages\\transformers\\generation\\utils.py:1313: UserWarning: Using `max_length`'s default (512) to control the generation length. This behaviour is deprecated and will be removed from the config in v5 of Transformers -- we recommend using `max_new_tokens` to control the maximum length of the generation.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[{'input': 'Bed linen is featured.',\n",
       "  'target': ['Предоставляется постельное белье.'],\n",
       "  'preds': 'Предоставляется постельное белье.'},\n",
       " {'input': 'A fitness room is also available.',\n",
       "  'target': ['Кроме того, к услугам фитнес-зал.'],\n",
       "  'preds': 'В распоряжении гостей фитнес-зал.'},\n",
       " {'input': 'Free parking is available in the garden',\n",
       "  'target': ['Гостям предоставляется бесплатная парковка в саду.'],\n",
       "  'preds': 'На территории отеля обустроена бесплатная парковка.'},\n",
       " {'input': 'The property offers free parking.',\n",
       "  'target': ['На территории дома для отпуска обустроена бесплатная парковка.'],\n",
       "  'preds': 'На территории обустроена бесплатная парковка.'}]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "[{'input': 'Guests benefit from balcony.',\n",
       "  'target': ['К услугам гостей балкон.'],\n",
       "  'preds': 'Гости могут воспользоваться балконом.'},\n",
       " {'input': 'A business centre is also available.',\n",
       "  'target': ['На территории работает бизнес-центр.'],\n",
       "  'preds': 'Кроме того, в распоряжении гостей бизнес-центр.'},\n",
       " {'input': 'The property offers free parking.',\n",
       "  'target': ['Гостям апартаментов предоставляется бесплатная парковка.'],\n",
       "  'preds': 'На территории обустроена бесплатная парковка.'},\n",
       " {'input': 'The property offers free parking.',\n",
       "  'target': ['Гостям доступна бесплатная парковка.'],\n",
       "  'preds': 'На территории обустроена бесплатная парковка.'}]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "[{'input': 'Some units are split level.',\n",
       "  'target': ['Некоторые апартаменты занимают 2 этажа.'],\n",
       "  'preds': 'В некоторых апартаментах имеется отдельная зона.'},\n",
       " {'input': 'The property offers free parking.',\n",
       "  'target': ['Предоставляется бесплатная парковка.'],\n",
       "  'preds': 'На территории обустроена бесплатная парковка.'},\n",
       " {'input': 'It offers free WiFi access.',\n",
       "  'target': ['К услугам гостей апартаменты с 2 спальнями и бесплатным WiFi.'],\n",
       "  'preds': 'К услугам гостей бесплатный Wi-Fi.'},\n",
       " {'input': 'The lobby is open 24 hours.',\n",
       "  'target': ['Вестибюль отеля открыт круглосуточно.'],\n",
       "  'preds': 'На круглосуточной стойке регистрации работает лобби.'}]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "[{'input': 'Massages are available on request.',\n",
       "  'target': ['Также можно заказать различные виды массажа.'],\n",
       "  'preds': 'По запросу предоставляются массажные процедуры.'},\n",
       " {'input': 'The property offers free parking.',\n",
       "  'target': ['На территории дома предоставляется бесплатная парковка.'],\n",
       "  'preds': 'На территории обустроена бесплатная парковка.'},\n",
       " {'input': 'The property offers free parking.',\n",
       "  'target': ['Гостям предоставляется бесплатная парковка.'],\n",
       "  'preds': 'На территории обустроена бесплатная парковка.'},\n",
       " {'input': 'The property offers free parking.',\n",
       "  'target': ['На территории апартаментов обустроена бесплатная парковка.'],\n",
       "  'preds': 'На территории обустроена бесплатная парковка.'}]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "{'bleu': 0.2014668239611586,\n",
       " 'precisions': [0.4948453608247423,\n",
       "  0.3333333333333333,\n",
       "  0.18461538461538463,\n",
       "  0.061224489795918366],\n",
       " 'brevity_penalty': 0.9695455377552523,\n",
       " 'length_ratio': 0.97,\n",
       " 'translation_length': 97,\n",
       " 'reference_length': 100}"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "            setTimeout(function() {\n",
       "                var nbb_cell_id = 14;\n",
       "                var nbb_unformatted_code = \"metric = evaluate.load(\\\"bleu\\\")\\nmodel.cuda()\\nmodel.eval()\\n_idx = 100\\nwith torch.no_grad():\\n    dataset[\\\"test\\\"].select(range(_idx, _idx + 16)).map(\\n        partial(\\n            compute_metrics_test,\\n            model,\\n            DataCollatorWithPadding(enc_tokenizer),\\n            dec_tokenizer,\\n            metric,\\n            log=True,\\n        ),\\n        batched=True,\\n        batch_size=batch_size,\\n    )\\nmetric.compute()\";\n",
       "                var nbb_formatted_code = \"metric = evaluate.load(\\\"bleu\\\")\\nmodel.cuda()\\nmodel.eval()\\n_idx = 100\\nwith torch.no_grad():\\n    dataset[\\\"test\\\"].select(range(_idx, _idx + 16)).map(\\n        partial(\\n            compute_metrics_test,\\n            model,\\n            DataCollatorWithPadding(enc_tokenizer),\\n            dec_tokenizer,\\n            metric,\\n            log=True,\\n        ),\\n        batched=True,\\n        batch_size=batch_size,\\n    )\\nmetric.compute()\";\n",
       "                var nbb_cells = Jupyter.notebook.get_cells();\n",
       "                for (var i = 0; i < nbb_cells.length; ++i) {\n",
       "                    if (nbb_cells[i].input_prompt_number == nbb_cell_id) {\n",
       "                        if (nbb_cells[i].get_text() == nbb_unformatted_code) {\n",
       "                             nbb_cells[i].set_text(nbb_formatted_code);\n",
       "                        }\n",
       "                        break;\n",
       "                    }\n",
       "                }\n",
       "            }, 500);\n",
       "            "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "metric = evaluate.load(\"bleu\")\n",
    "model.cuda()\n",
    "model.eval()\n",
    "_idx = 100\n",
    "with torch.no_grad():\n",
    "    dataset[\"test\"].select(range(_idx, _idx + 16)).map(\n",
    "        partial(\n",
    "            compute_metrics_test,\n",
    "            model,\n",
    "            DataCollatorWithPadding(enc_tokenizer),\n",
    "            dec_tokenizer,\n",
    "            metric,\n",
    "            log=True,\n",
    "        ),\n",
    "        batched=True,\n",
    "        batch_size=batch_size,\n",
    "    )\n",
    "metric.compute()"
   ]
  },
  {
   "attachments": {
    "image-13.png": {
     "image/png": "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"
    },
    "image-14.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "676aa8e4",
   "metadata": {},
   "source": [
    "![image-13.png](attachment:image-13.png) ![image-14.png](attachment:image-14.png)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}