objective_performance.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d9f919c9-5338-4742-99b0-1f15268c7053",
   "metadata": {},
   "outputs": [],
   "source": [
    "from datasets import load_dataset\n",
    "from transformers import AutoTokenizer\n",
    "from vllm import LLM, SamplingParams\n",
    "from rouge_score import rouge_scorer\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "import spacy\n",
    "nlp = spacy.load(\"en_core_web_sm\")\n",
    "\n",
    "import pandas as pd\n",
    "pd.set_option('display.max_colwidth', None)\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "bcbee639-9022-49f0-a2cd-784fba999ef3",
   "metadata": {},
   "outputs": [],
   "source": [
    "summary_prefs_ds = load_dataset('amang1802/summary_train')['test']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "c31c5b2d-adcc-42ce-8678-8a0a38c63f13",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Dataset({\n",
       "    features: ['info', 'summaries', 'choice', 'worker', 'batch', 'split', 'extra', 'query', 'tokens'],\n",
       "    num_rows: 256\n",
       "})"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "summary_prefs_ds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e10a4cde-85a4-407c-ad2f-3e707e5941b2",
   "metadata": {},
   "outputs": [],
   "source": [
    "def df_from_id(i):\n",
    "    text = summary_prefs_ds[i]['info']['post']\n",
    "    choice = summary_prefs_ds[i]['choice']\n",
    "    chosen = summary_prefs_ds[i]['summaries'][choice]['text']\n",
    "    rejected = summary_prefs_ds[i]['summaries'][1-choice]['text']\n",
    "    return pd.DataFrame({\"text\": [text], \"chosen\": [chosen], \"rejected\": [rejected]})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "f2f7e561-27b4-486c-9df0-0bc675dbbbfa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>text</th>\n",
       "      <th>chosen</th>\n",
       "      <th>rejected</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>I've been on and off with this girl for the past 8 months. She lived with her ex boyfriend but insisted that the were broken up and she didn't want to be with him.  He moved away from her 3 months ago and we've been casually seeing each other and having sex. \\n\\nThis weekend I went on a vacation with her and friends and found out that the ex had her listed as in a relationship on Facebook. When I found this out I messaged the ex and told him of all the times we had sex. I took screen shots of our conversation and told her to never talk to me again. She's since blocked me on Facebook and won't reply to my texts even though she agreed to talk to me about it. \\n\\nShe was a major part of my support system and I haven't been able to sleep well for 3 days. Does anyone have any advise how to break free from this depression?</td>\n",
       "      <td>found out she was cheating on me messaged ex with proof that we've been seeing each other, now I'm depressed and need advise.</td>\n",
       "      <td>I'm super depressed and can't get a break from the relationship.</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                           text  \\\n",
       "0  I've been on and off with this girl for the past 8 months. She lived with her ex boyfriend but insisted that the were broken up and she didn't want to be with him.  He moved away from her 3 months ago and we've been casually seeing each other and having sex. \\n\\nThis weekend I went on a vacation with her and friends and found out that the ex had her listed as in a relationship on Facebook. When I found this out I messaged the ex and told him of all the times we had sex. I took screen shots of our conversation and told her to never talk to me again. She's since blocked me on Facebook and won't reply to my texts even though she agreed to talk to me about it. \\n\\nShe was a major part of my support system and I haven't been able to sleep well for 3 days. Does anyone have any advise how to break free from this depression?   \n",
       "\n",
       "                                                                                                                           chosen  \\\n",
       "0   found out she was cheating on me messaged ex with proof that we've been seeing each other, now I'm depressed and need advise.   \n",
       "\n",
       "                                                            rejected  \n",
       "0   I'm super depressed and can't get a break from the relationship.  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "row = df_from_id(140)\n",
    "row"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "43b2785b-fd04-4a35-b1c5-66e577a5969e",
   "metadata": {},
   "outputs": [],
   "source": [
    "scorer = rouge_scorer.RougeScorer(['rouge1', 'rougeL'], use_stemmer=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "e670b26b-36e8-4a63-9ad9-2695726a480b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'rouge1': Score(precision=0.8, recall=0.11560693641618497, fmeasure=0.20202020202020204),\n",
       " 'rougeL': Score(precision=0.48, recall=0.06936416184971098, fmeasure=0.12121212121212122)}"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scorer.score(row.iloc[0]['text'], row.iloc[0]['chosen'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "29810e1b-9527-40dc-a593-2c8ea1f10448",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'rouge1': Score(precision=0.6923076923076923, recall=0.05202312138728324, fmeasure=0.0967741935483871),\n",
       " 'rougeL': Score(precision=0.46153846153846156, recall=0.03468208092485549, fmeasure=0.06451612903225806)}"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scorer.score(row.iloc[0]['text'], row.iloc[0]['rejected'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc43c263-9414-4cca-9c03-ab99fc1dd844",
   "metadata": {},
   "source": [
    "### What's the expected summary length?\n",
    "\n",
    "Let's analyze the distribution of lengths of the chosen summaries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "7b2d4c33-56f3-498e-b49e-96f3e6a4f3bd",
   "metadata": {},
   "outputs": [],
   "source": [
    "def word_count(text):\n",
    "    doc = nlp(text)\n",
    "    return sum(1 for token in doc if token.is_alpha)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "3f5ce7b5-7058-44b6-aacb-1c371d493bdf",
   "metadata": {},
   "outputs": [],
   "source": [
    "def chosen_summary_metrics(row):\n",
    "    choice = row['choice']\n",
    "    chosen = row['summaries'][choice]['text']\n",
    "    return {\"word_count_chosen\": word_count(chosen), \"chosen_summary\": chosen}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "ddc76951-1fe0-4e40-919a-e7f3bab9b14a",
   "metadata": {},
   "outputs": [],
   "source": [
    "chosen_metrics_ds = summary_prefs_ds.map(chosen_summary_metrics)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "9edc5dfa-5594-42bb-93b6-68b3a00b5c89",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean:  26.61328125\n",
      "Std Dev:  10.508314201545291\n"
     ]
    }
   ],
   "source": [
    "summary_lengths = np.array(chosen_metrics_ds['word_count_chosen'])\n",
    "print(\"Mean: \", summary_lengths.mean())\n",
    "print(\"Std Dev: \", summary_lengths.std())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "603f62fb-39d5-4b3b-9338-f55f0cfc70e8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([14., 18., 45., 35., 30., 48., 24.,  7., 11.,  6.,  3.,  7.,  2.,\n",
       "         1.,  1.,  0.,  3.,  0.,  0.,  1.]),\n",
       " array([10.  , 13.35, 16.7 , 20.05, 23.4 , 26.75, 30.1 , 33.45, 36.8 ,\n",
       "        40.15, 43.5 , 46.85, 50.2 , 53.55, 56.9 , 60.25, 63.6 , 66.95,\n",
       "        70.3 , 73.65, 77.  ]),\n",
       " <BarContainer object of 20 artists>)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(chosen_metrics_ds['word_count_chosen'], color='lightblue', ec='black', bins=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "02eeb24d-b2d7-4874-a40f-532b123ff60b",
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "64c4a37f-c0e0-4161-8356-295fc067e368",
   "metadata": {},
   "source": [
    "### Model generated summary length distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "c0e15dd8-ebf4-4c51-8637-dd4961348fe7",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_id = \"meta-llama/Llama-3.2-1B-Instruct\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "08549421-e964-495a-9eb3-1f37463c9272",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO 12-06 16:00:24 config.py:350] This model supports multiple tasks: {'embedding', 'generate'}. Defaulting to 'generate'.\n",
      "INFO 12-06 16:00:24 llm_engine.py:249] Initializing an LLM engine (v0.6.4.post1) with config: model='meta-llama/Llama-3.2-1B-Instruct', speculative_config=None, tokenizer='meta-llama/Llama-3.2-1B-Instruct', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, override_neuron_config=None, tokenizer_revision=None, trust_remote_code=False, dtype=torch.bfloat16, max_seq_len=4096, download_dir=None, load_format=LoadFormat.AUTO, tensor_parallel_size=1, pipeline_parallel_size=1, disable_custom_all_reduce=False, quantization=None, enforce_eager=False, kv_cache_dtype=auto, quantization_param_path=None, device_config=cuda, decoding_config=DecodingConfig(guided_decoding_backend='outlines'), observability_config=ObservabilityConfig(otlp_traces_endpoint=None, collect_model_forward_time=False, collect_model_execute_time=False), seed=0, served_model_name=meta-llama/Llama-3.2-1B-Instruct, num_scheduler_steps=1, chunked_prefill_enabled=False multi_step_stream_outputs=True, enable_prefix_caching=False, use_async_output_proc=True, use_cached_outputs=False, chat_template_text_format=string, mm_processor_kwargs=None, pooler_config=None)\n",
      "INFO 12-06 16:00:24 selector.py:135] Using Flash Attention backend.\n",
      "INFO 12-06 16:00:25 model_runner.py:1072] Starting to load model meta-llama/Llama-3.2-1B-Instruct...\n",
      "INFO 12-06 16:00:25 weight_utils.py:243] Using model weights format ['*.safetensors']\n",
      "INFO 12-06 16:00:25 weight_utils.py:288] No model.safetensors.index.json found in remote.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "80facac79a2842139c04b272088b9bc3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Loading safetensors checkpoint shards:   0% Completed | 0/1 [00:00<?, ?it/s]\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO 12-06 16:00:26 model_runner.py:1077] Loading model weights took 2.3185 GB\n",
      "INFO 12-06 16:00:27 worker.py:232] Memory profiling results: total_gpu_memory=23.69GiB initial_memory_usage=3.26GiB peak_torch_memory=3.51GiB memory_usage_post_profile=3.27GiB non_torch_memory=0.94GiB kv_cache_size=16.87GiB gpu_memory_utilization=0.90\n",
      "INFO 12-06 16:00:27 gpu_executor.py:113] # GPU blocks: 34551, # CPU blocks: 8192\n",
      "INFO 12-06 16:00:27 gpu_executor.py:117] Maximum concurrency for 4096 tokens per request: 134.96x\n",
      "INFO 12-06 16:00:28 model_runner.py:1400] Capturing cudagraphs for decoding. This may lead to unexpected consequences if the model is not static. To run the model in eager mode, set 'enforce_eager=True' or use '--enforce-eager' in the CLI.\n",
      "INFO 12-06 16:00:28 model_runner.py:1404] If out-of-memory error occurs during cudagraph capture, consider decreasing `gpu_memory_utilization` or switching to eager mode. You can also reduce the `max_num_seqs` as needed to decrease memory usage.\n",
      "INFO 12-06 16:00:38 model_runner.py:1518] Graph capturing finished in 10 secs, took 0.46 GiB\n"
     ]
    }
   ],
   "source": [
    "tokenizer = AutoTokenizer.from_pretrained(model_id)\n",
    "llm = LLM(model=model_id, max_model_len=4096)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "ab314705-939a-42e3-a1ab-e7a303c12465",
   "metadata": {},
   "outputs": [],
   "source": [
    "def text_to_messages(info, target_words):\n",
    "    return [\n",
    "        {\"role\": \"system\", \"content\": f\"Summarize the text using exactly {target_words} words:\"},\n",
    "        {\"role\": \"user\", \"content\": info['post']}\n",
    "    ]\n",
    "    \n",
    "def summarize_batch(batch, llm, tokenizer, target_words):\n",
    "    prompts = [tokenizer.apply_chat_template(text_to_messages(text, target_words), add_generation_prompt=True, tokenize=False)\n",
    "               for text in batch]\n",
    "    sampling_params = SamplingParams(temperature=0.25, max_tokens=128)\n",
    "    outputs = llm.generate(prompts, sampling_params)\n",
    "\n",
    "    return {\"llm_summary\": [output.outputs[0].text.strip() for output in outputs]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "20966e8d-652d-47dc-b274-c4a650723085",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Parameter 'function'=<function <lambda> at 0x7f80dc760680> of the transform datasets.arrow_dataset.Dataset._map_single couldn't be hashed properly, a random hash was used instead. Make sure your transforms and parameters are serializable with pickle or dill for the dataset fingerprinting and caching to work. If you reuse this transform, the caching mechanism will consider it to be different from the previous calls and recompute everything. This warning is only showed once. Subsequent hashing failures won't be showed.\n",
      "WARNING:datasets.fingerprint:Parameter 'function'=<function <lambda> at 0x7f80dc760680> of the transform datasets.arrow_dataset.Dataset._map_single couldn't be hashed properly, a random hash was used instead. Make sure your transforms and parameters are serializable with pickle or dill for the dataset fingerprinting and caching to work. If you reuse this transform, the caching mechanism will consider it to be different from the previous calls and recompute everything. This warning is only showed once. Subsequent hashing failures won't be showed.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e7bd3c4c79924d40be2305c9664bd387",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/256 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "Processed prompts:   0%|                              | 0/256 [00:00<?, ?it/s, est. speed input: 0.00 toks/s, output: 0.00 toks/s]\u001b[A\n",
      "Processed prompts:   0%|                    | 1/256 [00:03<12:47,  3.01s/it, est. speed input: 103.00 toks/s, output: 2.66 toks/s]\u001b[A\n",
      "Processed prompts:   1%|▏                  | 3/256 [00:03<03:33,  1.18it/s, est. speed input: 216.85 toks/s, output: 10.36 toks/s]\u001b[A\n",
      "Processed prompts:   2%|▍                  | 6/256 [00:03<01:29,  2.80it/s, est. speed input: 396.26 toks/s, output: 26.10 toks/s]\u001b[A\n",
      "Processed prompts:   8%|█▎              | 21/256 [00:03<00:17, 13.72it/s, est. speed input: 1734.78 toks/s, output: 127.08 toks/s]\u001b[A\n",
      "Processed prompts:  21%|███▍            | 54/256 [00:03<00:04, 43.46it/s, est. speed input: 4837.32 toks/s, output: 384.25 toks/s]\u001b[A\n",
      "Processed prompts:  36%|█████▋          | 91/256 [00:03<00:02, 80.78it/s, est. speed input: 8270.19 toks/s, output: 708.14 toks/s]\u001b[A\n",
      "Processed prompts:  50%|██████      | 128/256 [00:03<00:01, 119.91it/s, est. speed input: 11155.45 toks/s, output: 1078.89 toks/s]\u001b[A\n",
      "Processed prompts:  65%|███████▊    | 167/256 [00:03<00:00, 162.24it/s, est. speed input: 14235.87 toks/s, output: 1525.62 toks/s]\u001b[A\n",
      "Processed prompts:  77%|█████████▏  | 196/256 [00:04<00:00, 183.97it/s, est. speed input: 16416.53 toks/s, output: 1902.26 toks/s]\u001b[A\n",
      "Processed prompts: 100%|█████████████| 256/256 [00:04<00:00, 56.96it/s, est. speed input: 19436.69 toks/s, output: 3249.75 toks/s]\u001b[A\n"
     ]
    }
   ],
   "source": [
    "gen_summaries = chosen_metrics_ds.map(lambda batch: summarize_batch(batch, llm, tokenizer, 31),\n",
    "                                     input_columns=['info'],\n",
    "                                     batched=True,\n",
    "                                     batch_size=256)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "267a541b-74fc-422a-827d-b2297ec53e17",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[\"Here's a summary of the text in exactly 31 words:\\n\\nYou're concerned about your friend's behavior, concerned about his skewed view of sex, and want to confront him about his recent hookups, but need a safe and effective way to approach the situation.\",\n",
       " \"The writer is unsure how to interpret the woman's openness, which may indicate a desire for a romantic or intimate connection, but also raises concerns about her intentions and boundaries.\",\n",
       " \"It sounds like you're feeling conflicted about attending a family gathering and a party. Here's a 31-word summary:\\n\\nYou're torn between feeling guilty for not attending family events and enjoying a party with friends. Consider having an open conversation with your mom to clear up the misunderstanding and find a solution that works for everyone.\",\n",
       " 'A person was repeatedly calling a landline phone number, breathing for 2 seconds before hanging up, and the caller claimed to be a patient from a special needs institute, causing frustration and annoyance.',\n",
       " \"The girl's statement about IV TAs being kicked out for sleeping with students and boosting grades suggests she may be hinting about her own behavior due to intoxication, implying she may have also engaged in infidelity.\"]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gen_summaries['llm_summary'][:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "a4c917ae-242f-4da3-8861-638f4db1882f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def rouge_score(chosen, generated):\n",
    "    score = scorer.score(chosen, generated)\n",
    "    return score['rouge1'].fmeasure, score['rougeL'].fmeasure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "35ebc2ce-d6ae-4afc-805e-713773526288",
   "metadata": {},
   "outputs": [],
   "source": [
    "def llm_summary_metrics(chosen, generated):\n",
    "    rouge1, rougeL = rouge_score(chosen, generated)\n",
    "    return {\"word_count_llm_summary\": word_count(generated), 'rouge1': rouge1, 'rougeL': rougeL}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "00a9b3fa-cefb-4898-a1c9-e69840ef8512",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2dc9336574b144359e55e7fa1e6a88b3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/256 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "llm_metrics_ds = gen_summaries.map(llm_summary_metrics, input_columns=['chosen_summary', 'llm_summary'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "0d2bbeba-0582-438d-bc9f-00c156b32641",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean:  44.59375\n",
      "Std Dev:  27.538733375692864\n"
     ]
    }
   ],
   "source": [
    "summary_lengths = np.array(llm_metrics_ds['word_count_llm_summary'])\n",
    "print(\"Mean: \", summary_lengths.mean())\n",
    "print(\"Std Dev: \", summary_lengths.std())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "f4a879eb-8b8d-43ea-8881-1c6dd086ae1a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([ 3.,  3., 25., 39., 43., 32., 26., 19., 11.,  5.,  1.,  1.,  1.,\n",
       "         0.,  1.,  7., 15., 18.,  4.,  2.]),\n",
       " array([  5. ,  10.5,  16. ,  21.5,  27. ,  32.5,  38. ,  43.5,  49. ,\n",
       "         54.5,  60. ,  65.5,  71. ,  76.5,  82. ,  87.5,  93. ,  98.5,\n",
       "        104. , 109.5, 115. ]),\n",
       " <BarContainer object of 20 artists>)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(summary_lengths, color='lightblue', ec='black', bins=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "f22c6c4e-3cec-4a16-8832-2e1f5d1e7421",
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "f0f71a97-b590-4fac-8671-e8174a4e310f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Rouge1.fmeasure Mean:  0.1859368127738743\n",
      "RougeL.fmeasure Mean:  0.131804314879524\n"
     ]
    }
   ],
   "source": [
    "rouge1_scores = np.array(llm_metrics_ds['rouge1'])\n",
    "rougeL_scores = np.array(llm_metrics_ds['rougeL'])\n",
    "print(\"Rouge1.fmeasure Mean: \", rouge1_scores.mean())\n",
    "print(\"RougeL.fmeasure Mean: \", rougeL_scores.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "bd91b41d-fe2d-4ca2-a21a-1adde1620055",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(range(llm_metrics_ds.num_rows), llm_metrics_ds['rouge1'], label=\"Rouge1.fmeasure\")\n",
    "plt.plot(range(llm_metrics_ds.num_rows), llm_metrics_ds['rougeL'], label=\"RougeL.fmeasure\")\n",
    "plt.legend(loc=\"upper left\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a9c620bb-480b-4164-913a-5c7b92687724",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.11.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}