{ "cells": [ { "cell_type": "code", "execution_count": null, "id": "b1b875fd", "metadata": {}, "outputs": [], "source": [ "# load all the english vtt files into a text corpus for searching" ] }, { "cell_type": "code", "execution_count": 26, "id": "e5154934", "metadata": {}, "outputs": [], "source": [ "import os\n", "import json\n", "import tqdm\n", "import pandas as pd\n", "import webvtt\n", "import re\n", "import string\n", "import html\n", "import numpy as np\n", "\n", "from suno_utils.utils.text import normalize_whitespace\n", "\n", "DATA_DIR_LG = \"/data/suno/data/harvest/youtube_lg\"\n", "DATA_DIR_MED = \"/data/suno/data/harvest/youtube_med\"" ] }, { "cell_type": "code", "execution_count": 28, "id": "dcddd067", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 469791/469791 [00:16<00:00, 27788.04it/s]\n" ] } ], "source": [ "def get_fns_in_dir(data_dir):\n", " fns = os.listdir(os.path.join(data_dir, \"audio\"))\n", " data = []\n", " for fn in fns:\n", " parts = fn.split(\".\")\n", " if len(parts) != 3:\n", " continue\n", " data.append((parts[0], parts[1]))\n", " df = pd.DataFrame(data)\n", " _df = df[df[1].str[:2] == \"en\"]\n", " _df = _df.drop_duplicates(subset=[0])\n", " en_vtt_fns = set()\n", " for _, row in tqdm.tqdm(_df.iterrows(), total=_df.shape[0]):\n", " en_vtt_fns.add(\".\".join([row[0], row[1], \"vtt\"]))\n", " en_vtt_fns = list(en_vtt_fns)\n", " return en_vtt_fns\n", "\n", "# en_vtt_fns_med = get_fns_in_dir(DATA_DIR_MED)\n", "# print(len(en_vtt_fns_med), \"found\")\n", "en_vtt_fns_lg = get_fns_in_dir(DATA_DIR_LG)\n", "print(len(en_vtt_fns_lg), \"found\")" ] }, { "cell_type": "code", "execution_count": 29, "id": "3ee402ef", "metadata": {}, "outputs": [], "source": [ "# import copy\n", "from suno_utils.utils.numbers import safe_round\n", "\n", "ts_match_ptn = r\"^(\\d{2}\\:\\d{2}\\:\\d{2}\\.\\d{3}) \\-\\-\\> (\\d{2}\\:\\d{2}\\:\\d{2}\\.\\d{3})(?:.+\\%)?$\"\n", "\n", "def _parse_ts(ts):\n", " # 00:00:10.610\n", " hours, mins, secs = ts.split(\":\")\n", " return safe_round(float(hours) * 60 * 60 + float(mins) * 60 + float(secs))\n", "\n", "\n", "def clean_vtt(text):\n", " # Remove header and tags\n", " text = re.sub(r\"\\<\\/c\\>\", \"\", text)\n", " text = re.sub(r\"\\\", \"\", text)\n", " text = re.sub(r\"\\<\\d{2}\\:\\d{2}\\:\\d{2}\\.\\d{3}\\>\", \"\", text)\n", " text = html.unescape(text)\n", " blocks = text.strip().split(\"\\n\\n\")\n", " assert(len(blocks) >= 1)\n", " header_block = blocks[0]\n", " header_lines = header_block.strip().split(\"\\n\")\n", " assert(len(header_lines) == 3)\n", " assert(header_lines[0].strip() == \"WEBVTT\")\n", " assert(header_lines[1].strip() == \"Kind: captions\")\n", " assert(header_lines[2].strip().startswith(\"Language:\"))\n", " # extract language\n", " lang_code = header_lines[2].strip()[9:].strip()\n", " # split into sequential lines\n", " lines = []\n", " for b in blocks[1:]:\n", " block_lines = b.split(\"\\n\")\n", " if (\n", " len(block_lines) >= 2 and \n", " re.search(ts_match_ptn, block_lines[1]) and \n", " re.match(r\"^\\s*\\d+\\s*$\", block_lines[0])\n", " ):\n", " block_lines = block_lines[1:]\n", " cleaned_block_lines = []\n", " for line in block_lines:\n", " line = normalize_whitespace(line)\n", " if len(line) == 0:\n", " continue\n", " cleaned_block_lines.append(line)\n", " assert(len(cleaned_block_lines) <= 4) # most likely parse mistake if more \n", " lines.extend(cleaned_block_lines)\n", " if len(lines) == 0:\n", " return lang_code, []\n", " # parse timestamps of chunks\n", " m = re.search(ts_match_ptn, lines[0])\n", " assert(m)\n", " lines_with_ts = []\n", " cur_ts = (_parse_ts(m.group(1)), _parse_ts(m.group(2)))\n", " for line in lines[1:]:\n", " line = line.strip()\n", " m = re.search(ts_match_ptn, line)\n", " if m:\n", " cur_ts = [_parse_ts(m.group(1)), _parse_ts(m.group(2))]\n", " else:\n", " if len(line) > 0:\n", " lines_with_ts.append([cur_ts, line])\n", " if len(lines_with_ts) == 0:\n", " return lang_code, []\n", " # do some sanity checks\n", " assert(not any([ts[0] is None for ts, _ in lines_with_ts]))\n", " tmp_full_str = \" \".join([line for _, line in lines_with_ts])\n", " assert(\"align:\" not in tmp_full_str)\n", " assert(\"-->\" not in tmp_full_str)\n", " assert(not re.search(r\"\\d+\\s*\\:\\s*\\d+\\s*\\:\\s*\\d+\\.\\d+\", tmp_full_str))\n", " # dedupe repeating lines and merge timestamps\n", " lines_with_ts_deduped = [\n", " [[lines_with_ts[0][0][0], lines_with_ts[0][0][1]], lines_with_ts[0][1]]\n", " ]\n", " for (start_s, end_s), line in lines_with_ts[1:]:\n", " if line == lines_with_ts_deduped[-1][-1] and start_s <= lines_with_ts_deduped[-1][0][-1]:\n", " lines_with_ts_deduped[-1][0][-1] = end_s\n", " continue\n", " else:\n", " lines_with_ts_deduped.append([[start_s, end_s], line]) \n", " lines_with_ts_deduped = [((start_s, end_s), line) for (start_s, end_s), line in lines_with_ts_deduped]\n", " # merge concurrent lines\n", " merged_lines = []\n", " tmp_ts = lines_with_ts_deduped[0][0]\n", " tmp_lines = [lines_with_ts_deduped[0][1]]\n", " for ts, line in lines_with_ts_deduped[1:]:\n", " if ts != tmp_ts:\n", " # if very short then continue\n", " if len(line) <= 20 and line[-1:] in (\".\", \",\", \"?\") and line[:1] in set(string.ascii_lowercase):\n", " tmp_ts = (tmp_ts[0], ts[1])\n", " tmp_lines.append(line)\n", " else:\n", " if len(tmp_lines) > 0:\n", " merged_lines.append((tmp_ts, \" \".join(tmp_lines)))\n", " tmp_lines = [line]\n", " tmp_ts = ts\n", " else:\n", " tmp_lines.append(line)\n", " if len(tmp_lines) > 0:\n", " merged_lines.append((tmp_ts, \" \".join(tmp_lines)))\n", " # assert increasing times\n", " assert((np.diff([ts[0] for ts, _ in merged_lines]) >= -1).all())\n", " # somhow the ends are often not consistent\n", "# assert((np.diff([ts[1] for ts, _ in merged_lines]) >= -1).all())\n", " # flatten timestamps\n", " safe_lines = [merged_lines[0]]\n", " for (ts, te), line in merged_lines[1:]:\n", " nts = max(ts, safe_lines[-1][0][0])\n", " nte = max(ts, safe_lines[-1][0][1])\n", " safe_lines.append(((nts, nte), line))\n", " return lang_code, merged_lines" ] }, { "cell_type": "code", "execution_count": 10, "id": "fc4a3ae2", "metadata": {}, "outputs": [], "source": [ "# failed_l = []\n", "# for n, fn in enumerate(en_vtt_fns[:50]):\n", "# fp = os.path.join(DATA_DIR, \"audio\", fn)\n", "# with open(fp) as f:\n", "# text = f.read()\n", "# try:\n", "# lang_code, lines_with_ts = clean_vtt(text)\n", "# except:\n", "# failed_l.append(n)\n", "# print(len(failed_l), \"failed\")" ] }, { "cell_type": "code", "execution_count": null, "id": "7d4cb85a", "metadata": {}, "outputs": [], "source": [ "# n_iter = 0\n", "# with open(\"/data/suno/data/harvest/nlp/captions_med.txt\", \"w\") as fw:\n", "# for fn in tqdm.tqdm(en_vtt_fns_med):\n", "# fp = os.path.join(DATA_DIR_MED, \"audio\", fn)\n", "# with open(fp) as f:\n", "# text = f.read()\n", "# try:\n", "# lang_code, lines_with_ts = clean_vtt(text)\n", "# except:\n", "# continue\n", "# if lang_code[:2] != \"en\" or len(lines_with_ts) == 0:\n", "# continue\n", "# for _, line in lines_with_ts:\n", "# fw.write(line + \"\\n\")\n", "# fw.write(\"\\n\")\n", "# n_iter += 1\n", "# # if n_iter == 1:\n", "# # break\n", "# print(\"{}/{} correctly parsed\".format(n_iter, len(en_vtt_fns_med)))" ] }, { "cell_type": "code", "execution_count": 31, "id": "98ba6027", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 469791/469791 [1:14:44<00:00, 104.75it/s]" ] }, { "name": "stdout", "output_type": "stream", "text": [ "456313/411129 correctly parsed\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "\n" ] } ], "source": [ "n_iter = 0\n", "with open(\"/data/suno/data/harvest/nlp/captions_lg.txt\", \"w\") as fw:\n", " for fn in tqdm.tqdm(en_vtt_fns_lg):\n", " fp = os.path.join(DATA_DIR_LG, \"audio\", fn)\n", " with open(fp) as f:\n", " text = f.read()\n", " try:\n", " lang_code, lines_with_ts = clean_vtt(text)\n", " except:\n", " continue\n", " if lang_code[:2] != \"en\" or len(lines_with_ts) == 0:\n", " continue\n", " for _, line in lines_with_ts:\n", " fw.write(line + \"\\n\")\n", " fw.write(\"\\n\")\n", " n_iter += 1\n", "# if n_iter == 1:\n", "# break\n", "print(\"{}/{} correctly parsed\".format(n_iter, len(en_vtt_fns_lg)))" ] }, { "cell_type": "code", "execution_count": null, "id": "463c3175", "metadata": {}, "outputs": [], "source": [ "# TODO: what about leading dashes for new speaker?\n", "\n", "# TODO: other weirdnesses\n", "# H: Hey, Ke names\n", "# present with the Micra. >>Dr. Oza says the Micra\n", "# ♪Music♪\n", "# “Mechtild [00:15:29] It's often called the\n", "# Consider if someone has IWMI ( Inferior wall myocardial ischemia )\n", "# transcatheter aortic valve implantation (TAVI/TAVR).\n", "# VACCINE. UPMC TELLS ME THEY HAVE A LOT\n", "# • Macro siphoned mycelium\n", "# THIS IS THE MICRA APPROVED BY 00:02 line:92% THE FDA JUST YESTERDAY.\n", "# An mRNA vaccine (Comirnaty (Pfizer) or Spikevax (Moderna)) is preferred to AstraZeneca for this third dose.\n", "# r\"(\\((?>[^()]+|(?1))*\\))\"\n", "# - [Gabby] Nice. - I got my Moderna.\n", "# والمتعة pleasant time\n", "# Vedolizumab (Entyvio®), their clinical\n", "# [*bwwww*] CIVVIE 11: “I need to cancel my Paramount+ subscription.”" ] }, { "cell_type": "code", "execution_count": 37, "id": "0115594f", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "I have Nvidia Titan RTX 3090 24G V-ram, SSD raid, and AMD Threadripper X3990,\n", "Now this is running on the GeForce RTX 3090,\n", "RTX 3090 Gundam graphics card , which will be released to celebrate the 40th anniversary of the\n", "It is based on the usual ASUS ROG Strix RTX 3090, which is\n", "managed to get hold of an RTX 3090 in this version.\n", "Naturally, we've gone with an RTX 3090 for our test bench,\n", "The lineup will include the RTX 3080 and RTX 3090, equipped with a\n", "and the RTX 3090 benches overall performance\n", "M1 Ultra's dye area is bigger than the RTX 3090 Ti.\n", "than the RTX 3090 Ti.\n", "Nvidia’s GeForce RTX 3080, RTX 3080 Ti, and RTX 3090.\n", "So we made a giant Ampere. Ladies and gentlemen, the RTX 3090.\n", "They became the first to experience the new GeForce RTX 3090 on an LG OLED TV\n", "Or you could buy an RTX 3090 Ti,\n", "RTX 3090 Founders Edition we saw an overall 3Dmark Time Spy Extreme score of 9814. Now the\n", "current most powerful gaming cart, up until today, was the RTX 3090TI with a nice jump up to 10,785.\n", "the centre part has so many stars that I can only reproduce a part of by using graphics card RTX 3090.\n", "capability that on the M1 Ultra will match an RTX 3090\n", "i.e. RTX 3090, This is an expensive build so most of the people won't build this setup\n", "My main motive is to show you the performance of i9 with which we have paired RTX 3090\n", "Then coming on to the graphics card, I have MSI Suprim X RTX 3090 which is a 24 GB graphics card\n", "This game is so unoptimized that it is providing only 70-80 FPS on ultra settings even with RTX 3090\n", "which in this case is the RTX 3090 TI\n", "But, on the other hand, if you're building an RTX 3090\n", "the new RTX 3090, here I come.\n", "air-cooled RTX 3090 Ti models at the moment, including the RTX 3090 Ti FTW3 Black\n", "for $1999, the RTX 3090 Ti FTW3 for $2149, and the RTX 3090 Ti\n", "And the vanilla RTX 3090 FTW3 Ultra Gaming is on sale for\n", "Images have leaked of MSI's RTX 3090 Ti SUPRIM X indicating\n", "RTX 3090ti.\n", "even to Sapphire RX 6900XT Toxic Extreme or EVGA RTX 3090 Kingpin Hybrid.\n", "NUC from Intel, an RTX 3090, and the Game Jam features on GeForce\n", "Don't forget that they just ship you random RTX 3090s.\n", "and we're not gonna swap in the RTX 3090s,\n", "were limited by our RTX 3090 Ti graphics card at 1440p.\n", "bit more about this graphics card here and it's the GALAX GeForce RTX 3090 HOF Limited Edition.\n", "Obviously, we have the GALAX RTX 3090 and we'll definitely come back to this one.\n" ] } ], "source": [ "!cat /data/suno/data/harvest/nlp/captions_lg.txt | grep 'RTX 3090'" ] }, { "cell_type": "code", "execution_count": 21, "id": "015f9d41", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "like Humira, like Cimzia, then instead of two weeks, you will go nt for 17 days,\n", "are signs of joint erosion. Humira can help stop the clock.”\n", "the average retail price in the US of one carton (containing 2 dosing pens) of Humira\n", "of my body, I could use something that inject called Humira and they are medications\n", "In advanced diseases you may need to get immune modulator or biologics, such as 6-MP, or Remicade, or Humira.\n", "the top-selling drug in the world Humira\n", "So, AbbVie obviously makes biological Humira, of course we've all heard of Humira, and some\n", "blind multicenter study with about 800 patients, but only 500 treated with Humira in that study.\n", "Now the drug got approved. Boom. More than 10,000 patients have been treated with Humira\n", "household names. Humira. Advair. Ambien CR. Viagra single packs.\n", "blood sugar and apparent histamine intolerance after six months of Humira, a medication used\n", "Our first commercial product was Humira.\n", "introduced so Infliximab, Humira I think they were just sort of coming into play\n", "they needed to knock back her immune system with Humira.\n", "Humira was the number one drug in the world\n", "and was charging us just under $30,000 for the Humira.\n", "to ship the Humira directly to the infusion center.\n", "This means there is no markup in the price of the Humira.\n", "a drug called Humira will start\n", "direct competition for Humira.\n", "for Humira starts to develop early next year,\n", "and specifically Humira.\n", "for a prescription of Humira?\n", "if you're taking Humira, Etanercept or Enbrel,\n", "Humira, Enbrel,\n", "Humira for arthritis.\n", "Humira or Adalimumab. ELISA Genie provided\n", "highly validated kits for Humira and anti-drug antibodies to Humira.\n", "The benefits were quantitative measurement of Humira\n", "as well as anti-drug antibodies to Humira in serum and plasma samples.\n", "Adalimumab or Humira therapeutic drug monitoring\n", "like Remicade and Humira.\n", "Humira?\n", "Okay, it turns out, what does Humira treat?\n", "Humira, the brand name form of adalimumab, is a prescription medication used to treat\n", "colitis in adults and children. Humira belongs to a group of drugs called tumor necrosis\n", "of Humira include reaction at the injection site, sinus infections, and headaches.\n", "this is Humira Davis he music's shit up\n", "Laboratory testing for Adalimumab, also known as Humira, is also available at ARUP using\n", "moisture that we use with Humira Laka was the more similar to the Burcham, the albacore\n", "Humira detainer cept in bro Guillermo mAb somepony infliximab\n", "couldn't take Humira because it caused a fake tumor in her brain,\n", "which is a side effect of Humira.\n", "40 teams and Humira with 40 teams.\n", "Today in the final of three presentations, we continue reviewing Humira, a medication\n", "Humira is a treatment option for patients with moderate to severe rheumatoid arthritis or Crohn’s disease.\n", "The most common Humira side effects include: upper respiratory tract infections or sinusitis,\n", "Humira may lower the function of the immune system, and there are several important precautions\n", "One of the most important warnings with Humira is that patients are at an increased risk\n", "If a patient is actively sick with an infection, they should not begin treatment with Humira\n", "while receiving Humira therapy, and even for several months after stopping treatment.\n", "Drug interactions with Humira are important as well\n", "Additionally, live vaccines should not be given to patients receiving Humira.\n", "Patients should be tested for tuberculosis, or TB, before starting Humira and monitored\n", "Lymphoma, skin and other types of cancers may occur with TNF blockers like Humira.\n", "been reported with Humira, although these are not frequent side effects.\n", "There are other warnings, side effects and drug interactions that may occur with Humira.\n", "Patients should closely review the medication guide that accompanies Humira each time they\n", "Thank you for joining us at Drugs.com for a brief review of Humira.\n", "Patients with a concern about the use of Humira should consult with their health care provider.\n", "Visit www.drugs.com/Humira for more information\n", "This drug is called Humira, This is for rheumatoid arthritis is taken as a oral drug, so the\n", "and inject an immune-suppressant medication called Humira.\n", "and home injections of an immune-suppressant drug, Humira,\n", "and the second one is called \"Adalimumab\" or \"Humira\". We do expect others in the future,\n", "Humira and there's lots of other names for it. Humira is the originator one and if you have a\n", "antibodies to Infliximab, you don't necessarily develop antibodies to Humira to Adalimumab.\n", "like Humira as long as they don't live in England. Otherwise, they could get hold of\n", "The brand names of TNF are Humira, Enbrel, Cimzia\n", "Humira, Enbrel, Simponi and Cimzia are given by an injection\n", "With Humira.\"\n", "That's Humira, a drug that won FDA approval more than 16 years ago, and yet here we are,\n", "So why is there no generic Humira?\n", "That's because there's no such thing as generics for drugs like Humira.\n", "But drugs like Humira are different.\n", "Humira is an antibody grown from bacteria to mimic an actual human protein.\n", "So if you wanted to make a Humira of your own, scientifically, you could do so, but\n", "So let's get back to Lipitor and Humira.\n", "Whereas for Humira, the main patent on the drug expired back in 2016, and yet as of this\n", "very moment, there are no biosimilars of Humira on the market in the United States.\n", "AbbVie, the inventor of Humira, has created what lawyers call a \"patent thicket.\"\n", "And because of that thicket, at least as things stand right now, it doesn't look like Humira\n", "The story of Humira illustrates how these new biologic medicines haven't been forced\n", "Which is a long way of saying that all those happy Humira patients you see on TV aren't\n", "\"Humira can lower your ability to fight infections.\"\n", "platinum so he's put in the ICU they stopped his Humira and his\n", "like it's called Humira\n", "You may know it as Humira.\n", "Some of these drugs include Orencia, Humira, and Trexall.\n", "Some of you may have heard of Humira from television commercials.\n", "Humira they are all very important drugs. They are all TNF alpha inhibitors, popularly\n", "which would include Enbrel, Humira,\n", "For instance, Humira, if anybody's heard of Humira?\n", "Enbrel and Humira, they in fact target the markers\n", "So you take the Enbrel and Humira and after a certain\n", "the drugs Humira and Enbrel.\n", "advertising, passing Humira.\n", "But Humira or Adalimumab, Humira's a brand name,\n", "- But Humira or Adalimumab,\n", "Humira, human, is the human form.\n", "He didn’t want to go on Humira.\n", "AbbVie for its pricing of Humira,\n", "Humira, it generated $3.9 billion\n", "drugs, Humira was one of them,\n", "Now Humira, a drug used to treat\n", "such as Adalimumab, which is also called Humira.\n", "and Humira. Can you guess why? I know it’s early in this video and you probably weren’t expecting\n", "There were no biologics like Humira and Remicade back then.\n", "Next Humira tops others for patients with ankylosing\n", "Compared with both Enbrel and Remicade, Humira led to many fewer office visits for anterior uveitis.\n", "a pen for medicine like Humira.\n", "I chose Abbvie at the time because they make Humira\n", "maybe you're on something like Humira\n", "Let me give you an example: Adalimumab, commercially known as Humira, is a TNF- alpha inhibitor\n", "The first three are without Humira.\n", "Humira is an anti-inflammatory drug\n", "So the first three are without Humira,\n", "and the second three are with Humira.\n", "in the Humira treated group,\n", "like Humira or Taltz or Xolair over here,\n", "So Humira is for knocking down\n", "So that's Abbvie's drug Humira.\n", "to put me on a medication called Humira. Now, you guys may have seen a lot of commercials for this.\n", "It is on all the time and I laugh every time I see it. I’m not currently on Humira anymore. I\n", "was on Humira for many years after my diagnosis. It was given to me though because it had just\n", "who makes Humira to see if I could qualify for their free medication program. We were working\n", "And then really, really good results with Humira.\n", "as well as nonheme iron. There be non-here, not Humira and that is a broccoli and the broccoli\n", "Adalimumab Humira injection\n", "A couple of the ones, you've heard of Enbrel or Humira.\n", "and Humira attacks something called interleukin-12.\n", "of called Remicade or Humira or Cimzia or Simponi.\n", "I tried Humira -- which worked for about a year and then suddenly stopped working and\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Today in the first of three presentations, we are reviewing Humira, a medication in the\n", "Humira, known by the generic name of adalimumab, is approved by the FDA for treatment of adults\n", "Humira is also approved for treatment of other immune-mediated conditions, including plaque\n", "Humira is a synthetic antibody and works by binding and blocking this TNF protein. By\n", "blocking this protein, Humira helps to keep the protein from attacking healthy cells,\n", "Although Humira is an injection that is given under the skin, most patients, or their caregivers\n", "Humira is supplied in a single-use prefilled glass syringe, which makes it easier for the\n", "patient to give their own doses. Humira should be stored in the refrigerator until needed.\n", "Thank you for joining us at Drugs.com for a brief review of Humira. Please refer to\n", "Patients with a concern about the use of Humira should consult with their health care provider.\n", "Visit www.drugs.com/Humira for more information\n", "They will not be able to achieve similar results to what Abbvie was able to accomplish with Humira and rheumatoid arthritis\n", "the top selling drugs, the top selling drug in 2017 was Humira which is used to treat various\n", "in 2017 of Humira were 18.4 billion dollars U.S worldwide; it's pretty big business.\n", "Some examples of TNF-alpha blocking agents include adalimumab (Humira), infliximab (Remicade),\n", "(such as etanercept [Enbrel], adalimumab [Humira], infliximab [Remicade],\n", "Again, my steroids, my Humira. I went home. Again, I ate entirely way too much food, and threw it up\n", "you all know Humira, Enbrel, Orencia\n", "Humira for rheumatoid arthritis or psoriasis, other drugs\n", "medication. So for almost two years now I've been on Humira which is a biologic.\n", "medication failed and then I was started along Humira. So every two weeks I\n", "give myself a Humira injection. You can give it to yourself from the stomach\n", "week I call it my Humira day and now that I'm done I put the little cap back\n", ">> So, we started the Humira which is an injectable medication.\n", "prescriptions for Lipitor, Toujeo, Enbrel, Humira,\n", "So a drug like Humira, which is commonly prescribed\n", "Side effects are known to occur in a small percentage of patients taking Humira.\n", "the most common side effects reported by patients taking Humira were upper respiratory infections,\n", "In general, these symptoms are temporary and go away as soon as you stop taking Humira.\n", "I accidentally adminstered a 40 mg shot of Humira to my wife 5 days early.\n", "the $35,000 a year drug REMICADE; or the $75,000 a year drug Humira.\n", "Aisha Humira\n", "Aisha Humira\n", "Aisha Humira\n", "Biologic agents for moderate to severe psoriasis include: etanercept (Enbrel), adalimumab (Humira)\n", "Medications used: Humira, Enbrel, Prednisone, Celebrex, Metotrexate, Voltaren, Gold.\"\n", "After adjusting for baseline HbA1c among patients without diabetes, those taking Humira had\n", "For example, Abbvie’s patent on the rheumatoid arthritis drug Humira expires in 2016, leaving\n", "for its lead rheumatoid joint inflammation drug Humira, which represented 58% of the\n", "\"Before we put you on treatment -- probably Humira, something like that -- let's do bloodwork again.\"\n", "and the working hypothesis is, instead of going on Humira or some other biologic,\n", "And then I was on Humira for a long time.\n", "I've since switched off of Humira and now I'm on Entyvio\n", "The results of all this Humira wasn't working for me in the end.\n", "like Erin mentioned Humira,\n", "So Erin mentioned Humira,\n", "So Humira actually would help with your joints, your eyes,\n", "sort of receptors, like say Humira, say Entyvio,\n", "Humira and then Entyvio you know,\n", "Erin mentioned like Humira and Entyvio or Remicade,\n", "and similar medications, you mean like Humira,\n", "your immune system, the biologics are Revellex, Enbrel, Humira, Stelara, Cosentyx and Tremfya\n", "So we delivered adalimumab, or Humira, insulin again,\n", "Humira… a great drug for rheumatoid arthritis… is it effective for osteoarthritis also?\n", "Treatment With Humira Unsuccessful In Providing Pain Relief For Hand OA.\n", "Diseases indicated that “treatment with adalimumab (Humira) was unsuccessful in providing\n", "Humira, the TNF monoclonal antibody\n", "Today in the second of three presentations, we continue reviewing Humira, a commonly used\n", "Humira is also often referred to as a biological response modifier.\n", "We will review some of the rheumatoid arthritis and Crohn’s disease clinical trial information for Humira.\n", "In clinical studies of Humira in early but aggressive rheumatoid arthritis, roughly 6\n", "More patients who combined methotrexate with Humira had a better response than those who\n", "Studies with Humira have also shown that joint damage, pain and overall daily function are\n", "Placebo-controlled studies with Humira showed that more than 50 percent of patients saw\n", "Thank you for joining us at Drugs.com for a brief review of Humira.\n", "Patients with a concern about the use of Humira should consult with their health care provider.\n", "Visit www.drugs.com/Humira for more information\n", "\"Any comments regarding Humira for RA?\" This is very interesting.\n", "use them for inflammatory bowel disease, for psoriasis, for RA like Humira, and for\n", "getting Humira. In fact, it may be helping the cavernous malformation. If you\n", "or antibody drugs like Humira or Enbrel.\n", "like Cymbalta, Prozac, Zoloft, you've got the immune modulators like Humira and Enbrel.\n", "Infliximab (Remicade®), Adalimumab (Humira®), Vedolizumab (Entyvio®), their clinical trials\n", "So drugs like Humira, Keytruda, Avastin, Herceptin, they’re well-known biopharmaceutical\n", "5. Humira\n", "Humira is a drug that can help cure rheumatoid arthritis and Crohn’s disease and it is very\n", "restrictions. Furthermore, Humira, in many cases, is placed in high-tier insurance formularies and\n", "additional complication of having ulcerative colitis, so I'm on Humira which\n", "Even try Humira. I was on\n", "So Adalimumab is Humira, Certolizumab is Cimzia, Golimumab is Simponi and Infliximab is Remicaid.\n", "Europeans who take the world's biggest selling drug, Humira will soon have the option of\n", "Why are the cheaper versions of Humira being rolled out in Europe?\n", "These are pretty much identical to the drugs that are already like Humira, biosimilar is why they call them.\n", "drugs, Mollye, I'm curious about what are the legal differences in patents for Humira\n", "Even though the main patent for Humira expired in the United States last year, there are\n", "In fact, the soonest a biosimilar or generic version of Humira could be available in the\n", "I'm curious about whether or not Humira and others, Humira and other drugs in the future,\n", "And Humira, which is all over the television advertising.\n", "The Humira injections, and I know personally, it's about seven thousand\n", "dollars a month that my insurance company pays for two injections for this Humira.\n", "and I still have to take Humira\n" ] } ], "source": [ "!cat /data/suno/data/harvest/nlp/captions_med.txt | grep Humira" ] }, { "cell_type": "code", "execution_count": null, "id": "74608fda", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "b654bb4f", "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "id": "2b8a580d", "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": "3.8.13" } }, "nbformat": 4, "nbformat_minor": 5 }