{"id":14712,"date":"2026-07-16T13:02:42","date_gmt":"2026-07-16T17:02:42","guid":{"rendered":"https:\/\/labs.icahn.mssm.edu\/minervalab\/?page_id=14712"},"modified":"2026-07-19T15:37:08","modified_gmt":"2026-07-19T19:37:08","slug":"llm-module","status":"publish","type":"page","link":"https:\/\/labs.icahn.mssm.edu\/minervalab\/documentation\/llm-module\/","title":{"rendered":"LLM Module"},"content":{"rendered":"<p>[et_pb_section fb_built=&#8221;1&#8243; _builder_version=&#8221;4.16&#8243; global_colors_info=&#8221;{}&#8221;][et_pb_row _builder_version=&#8221;4.27.4&#8243; background_size=&#8221;initial&#8221; background_position=&#8221;top_left&#8221; background_repeat=&#8221;repeat&#8221; custom_padding=&#8221;||1px||false|false&#8221; global_colors_info=&#8221;{}&#8221;][et_pb_column type=&#8221;4_4&#8243; _builder_version=&#8221;4.27.4&#8243; _module_preset=&#8221;default&#8221; global_colors_info=&#8221;{}&#8221;][et_pb_text _builder_version=&#8221;4.27.4&#8243; _module_preset=&#8221;default&#8221; hover_enabled=&#8221;0&#8243; global_colors_info=&#8221;{}&#8221; sticky_enabled=&#8221;0&#8243;]<\/p>\n<p><a href=\"https:\/\/labs.icahn.mssm.edu\/minervalab\/scientific-computing-and-data\/\">Scientific Computing and Data<\/a> \/ <a href=\"https:\/\/labs.icahn.mssm.edu\/minervalab\/\">High Performance Computing<\/a> \/ <a title=\"Documentation\" href=\"https:\/\/labs.icahn.mssm.edu\/minervalab\/documentation\/\">Documentation<\/a> \/ The minerva-llm Module<\/p>\n<p>[\/et_pb_text][et_pb_text _builder_version=&#8221;4.27.4&#8243; _module_preset=&#8221;default&#8221; custom_css_free_form=&#8221;.minerva-doc {||  &#8211;ink-900: #212070;||  &#8211;ink-700: #3a3a52;||  &#8211;ink-500: #6f7086;||  &#8211;line: #e4e4ee;||  &#8211;line-soft: #efeff6;||  &#8211;bg-page: #fbfbfd;||  &#8211;bg-card: #ffffff;||  &#8211;bg-subtle: #f5f5fa;||  &#8211;accent-600: #dc298d;||  &#8211;accent-500: #dc298d;||  &#8211;accent-100: #fbdeee;||  &#8211;accent-050: #fdf0f7;||  &#8211;indigo-100: #e4e4f2;||  &#8211;indigo-050: #f2f2f9;||  &#8211;code-bg: #1b1b3a;||  &#8211;code-text: #e6e6f0;||  &#8211;code-line: rgba(255,255,255,0.08);||  &#8211;font-body: %22Open Sans%22, -apple-system, BlinkMacSystemFont, %22Segoe UI%22, Helvetica, Arial, sans-serif;||  &#8211;font-mono: %22SFMono-Regular%22, %22JetBrains Mono%22, Consolas, Menlo, monospace;||  &#8211;radius-lg: 14px;||  &#8211;radius-md: 8px;||  width: 100%;||  padding: 0 8px;||  background: var(&#8211;bg-page);||  color: var(&#8211;ink-700);||  font-family: var(&#8211;font-body);||  font-size: 16px;||  line-height: 1.65;||  -webkit-font-smoothing: antialiased;||}||||.minerva-doc *,||.minerva-doc *::before,||.minerva-doc *::after {||  box-sizing: border-box;||}||||.minerva-doc .site-header {||  padding: 40px 0 8px;||}||||.minerva-doc .brand-tag {||  display: inline-flex;||  align-items: center;||  font-size: 12.5px;||  font-weight: 600;||  color: #212070;||  background: var(&#8211;indigo-100);||  padding: 5px 12px;||  border-radius: 999px;||  letter-spacing: 0.2px;||  margin-bottom: 14px;||}||||.minerva-doc .brand-title {||  margin: 0;||  color: var(&#8211;ink-900);||  font-size: 32px;||  font-weight: 700;||  letter-spacing: -0.6px;||  line-height: 1.2;||}||||.minerva-doc .intro {||  padding: 14px 0 30px;||  border-bottom: 1px solid var(&#8211;line);||}||||.minerva-doc .doc-section {||  border-bottom: 1px solid var(&#8211;line);||}||||.minerva-doc .doc-section:last-of-type {||  border-bottom: none;||}||||.minerva-doc .doc-section summary {||  list-style: none;||  cursor: pointer;||  padding: 22px 0;||  display: flex;||  align-items: baseline;||  gap: 8px;||  font-size: 26px;||  font-weight: 700;||  color: var(&#8211;ink-900);||  position: relative;||}||||.minerva-doc .doc-section summary::-webkit-details-marker {||  display: none;||}||||.minerva-doc .doc-section summary::after {||  content: %22%22;||  margin-left: auto;||  align-self: center;||  width: 10px;||  height: 10px;||  border-right: 2px solid var(&#8211;ink-500);||  border-bottom: 2px solid var(&#8211;ink-500);||  transform: rotate(45deg);||  transition: transform 0.2s ease;||  flex: none;||}||||.minerva-doc .doc-section%91open%93 summary::after {||  transform: rotate(-135deg);||  border-color: var(&#8211;accent-600);||}||||.minerva-doc .doc-section summary:hover {||  color: var(&#8211;accent-600);||}||||.minerva-doc .section-body {||  padding: 0 0 28px;||}||||.minerva-doc .eyebrow {||  display: inline-block;||  font-family: var(&#8211;font-body);||  font-size: 26px;||  font-weight: 700;||  color: var(&#8211;ink-900);||  letter-spacing: 0;||}||||.minerva-doc .intro-eyebrow {||  display: block;||  font-family: var(&#8211;font-body);||  font-size: 26px;||  font-weight: 700;||  color: var(&#8211;ink-900);||  letter-spacing: 0;||  margin-bottom: 10px;||}||||.minerva-doc h3 {||  color: var(&#8211;ink-900);||  font-size: 15px;||  font-weight: 600;||  margin: 24px 0 10px;||  line-height: 1.3;||}||||.minerva-doc p {||  margin: 0 0 16px;||  color: var(&#8211;ink-700);||}||||.minerva-doc .section-body &gt; p:last-child {||  margin-bottom: 0;||}||||.minerva-doc a {||  color: var(&#8211;accent-600);||  text-decoration: none;||  background-image: linear-gradient(var(&#8211;accent-500), var(&#8211;accent-500));||  background-repeat: no-repeat;||  background-size: 100% 1px;||  background-position: 0 100%;||}||||.minerva-doc a:hover {||  color: var(&#8211;ink-900);||  background-image: linear-gradient(#212070, #212070);||}||||.minerva-doc strong {||  font-weight: 600;||  color: var(&#8211;ink-900);||}||||.minerva-doc ul,||.minerva-doc ol {||  margin: 0 0 16px;||  padding-left: 22px;||}||||.minerva-doc li {||  margin-bottom: 9px;||}||||.minerva-doc li ul,||.minerva-doc li ol {||  margin-top: 9px;||  margin-bottom: 0;||}||||.minerva-doc ul &gt; li::marker {||  color: var(&#8211;accent-500);||}||||.minerva-doc ol &gt; li::marker {||  color: var(&#8211;accent-500);||  font-weight: 600;||}||||.minerva-doc .callout {||  margin: 20px 0 4px;||  padding: 18px 22px;||  border-radius: var(&#8211;radius-md);||}||||.minerva-doc .callout p:last-child {||  margin-bottom: 0;||}||||.minerva-doc .callout-accent {||  background: var(&#8211;indigo-050);||  border: 1px solid var(&#8211;indigo-100);||  border-left: 3px solid var(&#8211;accent-500);||}||||.minerva-doc .callout-accent strong {||  color: var(&#8211;ink-900);||}||||.minerva-doc .note {||  background: var(&#8211;bg-subtle);||  border: 1px solid var(&#8211;line);||  border-left: 3px solid #212070;||  padding: 14px 18px;||  margin: 0 0 16px;||  border-radius: 0 var(&#8211;radius-md) var(&#8211;radius-md) 0;||  color: var(&#8211;ink-700);||}||||.minerva-doc code {||  font-family: var(&#8211;font-mono);||  background: var(&#8211;bg-subtle);||  color: #b81f77;||  padding: 2px 6px;||  border-radius: 4px;||  font-size: 0.87em;||  border: 1px solid var(&#8211;line);||}||||.minerva-doc .code-block {||  background: var(&#8211;code-bg);||  border-radius: var(&#8211;radius-md);||  margin: 14px 0 22px;||  overflow: hidden;||  box-shadow: 0 8px 24px rgba(33, 32, 112, 0.18);||}||||.minerva-doc .code-block-header {||  display: flex;||  align-items: center;||  gap: 6px;||  padding: 8px 10px 8px 14px;||  border-bottom: 1px solid var(&#8211;code-line);||}||||.minerva-doc .code-lang {||  font-family: var(&#8211;font-mono);||  font-size: 11px;||  color: #9b9bc0;||  text-transform: uppercase;||  letter-spacing: 1px;||  padding: 2px 6px;||  border-radius: 3px;||}||||\/* Code indicators customization options *\/||.minerva-doc .code-lang.code-python {||  color: #38bdf8; \/* Soft Python Blue *\/||  background: rgba(56, 189, 248, 0.1);||}||||.minerva-doc .code-lang.code-output {||  color: #34d399; \/* Console green *\/||  background: rgba(52, 211, 153, 0.1);||}||||.minerva-doc .code-block pre {||  margin: 0;||  padding: 16px 18px 18px;||  font-family: var(&#8211;font-mono);||  font-size: 13.5px;||  line-height: 1.7;||  color: var(&#8211;code-text);||  white-space: pre-wrap;||  word-break: break-word;||  overflow-x: auto;||}||||.minerva-doc .code-block pre code {||  background: none;||  border: none;||  padding: 0;||  color: inherit;||}||||.minerva-doc .code-block a {||  color: #f6a3d0;||  background-image: linear-gradient(#f6a3d0, #f6a3d0);||}||||.minerva-doc .table-wrap {||  overflow-x: auto;||  margin: 14px 0 4px;||  border-radius: var(&#8211;radius-md);||  border: 1px solid var(&#8211;line);||}||||.minerva-doc table {||  width: 100%;||  border-collapse: collapse;||  font-size: 14px;||}||||.minerva-doc th {||  background: var(&#8211;indigo-050);||  color: var(&#8211;ink-900);||  font-weight: 600;||  text-align: left;||  padding: 12px 16px;||  font-size: 12px;||  text-transform: uppercase;||  letter-spacing: 0.5px;||  border-bottom: 1px solid var(&#8211;indigo-100);||}||||.minerva-doc td {||  padding: 12px 16px;||  border-top: 1px solid var(&#8211;line-soft);||  vertical-align: top;||  color: var(&#8211;ink-700);||}||||.minerva-doc tbody tr:hover {||  background: var(&#8211;bg-subtle);||}||||.minerva-doc .doc-footer {||  padding: 24px 0 8px;||  text-align: center;||  color: var(&#8211;ink-500);||  font-size: 13px;||}||||@media (max-width: 520px) {||  .minerva-doc .brand-title { font-size: 24px; }||  .minerva-doc .doc-section summary { font-size: 21px; }||  .minerva-doc .eyebrow,||  .minerva-doc .intro-eyebrow { font-size: 21px; }||}&#8221; global_colors_info=&#8221;{}&#8221;]<\/p>\n<div class=\"minerva-doc\">\n<header class=\"site-header\">\n    <span class=\"brand-tag\">HPC Documentation<\/span><\/p>\n<h1 class=\"brand-title\">User Guide for the minerva-llm Module<\/h1>\n<\/header>\n<p>  <!-- Overview &amp; NIH Guidelines --><\/p>\n<section class=\"intro\" id=\"overview\">\n    <span class=\"intro-eyebrow\">Overview<\/span><\/p>\n<p>The <code>minerva-llm<\/code> module provides access to centrally managed, locally hosted, and approved AI models for researchers to use in Python and other applications on Minerva. These models are hosted within the Minerva environment, allowing users to perform inference without downloading models from external repositories.<\/p>\n<p>    <!-- Data Security Callout --><\/p>\n<div class=\"callout callout-accent\">\n      <strong>\u26a0\ufe0f Data Security and NIH Guidance<\/strong><\/p>\n<p>When using AI models with controlled-access human genomic data, researchers are responsible for complying with all applicable institutional and funding agency policies. The NIH Guide Notice, <em>Protecting Human Genomic Data when Developing Generative Artificial Intelligence Tools and Applications<\/em>, includes:<\/p>\n<ul>\n<li>A reminder that the GDS Policy does not allow users to share controlled-access data or their data derivatives with unauthorized users, including public generative AI tools.<\/li>\n<li>A prohibition against sharing generative AI models and their parameters trained on human genomic data obtained from NIH controlled-access repositories.<\/li>\n<\/ul>\n<p>NIH intends to release future guidance outlining the responsible use and sharing of generative AI models based on controlled-access data.<\/p>\n<p><a href=\"https:\/\/grants.nih.gov\/news-events\/nih-extramural-nexus-news\/2025\/05\/protecting-human-genomic-data-when-developing-generative-artificial-intelligence-tools-and-applications\" target=\"_blank\" rel=\"noopener\">Read the Complete NIH Guide Notice &rarr;<\/a><\/p>\n<\/p><\/div>\n<\/section>\n<p>  <!-- Section 1: Conda Environment --><\/p>\n<details class=\"doc-section\" id=\"conda-env\" open>\n<summary><span class=\"eyebrow\">1<\/span> Conda Environment Setup<\/summary>\n<div class=\"section-body\">\n<p>We recommend using a conda environment for running LLM applications.<\/p>\n<h3>1. Load Proxies and Miniforge<\/h3>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang\">Shell<\/span><\/div>\n<pre><code># Load standard proxy\r\nmodule load proxies\/1\r\n\r\n# OR on the Restricted Cluster:\r\nmodule load res_proxies\/1\r\n\r\n# Load Miniforge\r\nmodule load miniforge3\/26.1.1-3<\/code><\/pre>\n<\/p><\/div>\n<h3>2. Create and Activate Environment<\/h3>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang\">Shell<\/span><\/div>\n<pre><code># Create environment\r\nconda create -n llm-test python=3.12\r\n\r\n# Activate environment\r\nconda activate llm-test\r\n\r\n# Prevent Python from loading packages from ~\/.local directory\r\nexport PYTHONNOUSERSITE=1<\/code><\/pre>\n<\/p><\/div>\n<h3>3. Install Packages<\/h3>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang\">Shell<\/span><\/div>\n<pre><code>conda install pytorch transformers accelerate sentencepiece protobuf wrapt<\/code><\/pre>\n<\/p><\/div>\n<div class=\"note\">\n        <strong>Note:<\/strong> You only need to create the environment once. For future sessions, simply load Miniforge and activate the environment.\n      <\/div>\n<\/p><\/div>\n<\/details>\n<p>  <!-- Section 2: Load Module --><\/p>\n<details class=\"doc-section\" id=\"load-module\">\n<summary><span class=\"eyebrow\">2<\/span> Load the Minerva LLM Module<\/summary>\n<div class=\"section-body\">\n<p>Load the centrally managed module to configure your runtime environment variables automatically:<\/p>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang\">Shell<\/span><\/div>\n<pre><code>module load minerva-llm\/1.0<\/code><\/pre>\n<\/p><\/div>\n<p>Once loaded, you can verify your environment directories using:<\/p>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang\">Shell<\/span><\/div>\n<pre><code>echo $MINERVA_LLM_ROOT\r\necho $MINERVA_LLM_MODELS<\/code><\/pre>\n<\/p><\/div>\n<\/p><\/div>\n<\/details>\n<p>  <!-- Section 3: List Models --><\/p>\n<details class=\"doc-section\" id=\"list-models\">\n<summary><span class=\"eyebrow\">3<\/span> List Available Models<\/summary>\n<div class=\"section-body\">\n<p>Display all available shared models hosted locally on Minerva:<\/p>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang\">Shell<\/span><\/div>\n<pre><code>minerva-llm-list<\/code><\/pre>\n<\/p><\/div>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang code-output\">Output<\/span><\/div>\n<pre><code>Llama-3.2-1B-Instruct\r\nLlama-3.2-3B-Instruct\r\nTinyLlama-1.1B-Chat-v1.0<\/code><\/pre>\n<\/p><\/div>\n<\/p><\/div>\n<\/details>\n<p>  <!-- Section 4: Search for a Model --><\/p>\n<details class=\"doc-section\" id=\"search-models\">\n<summary><span class=\"eyebrow\">4<\/span> Search for a Model<\/summary>\n<div class=\"section-body\">\n<p>Search for models using a keyword. For example, to find &#8220;tiny&#8221; models:<\/p>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang\">Shell<\/span><\/div>\n<pre><code>minerva-llm-list tiny<\/code><\/pre>\n<\/p><\/div>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang code-output\">Output<\/span><\/div>\n<pre><code>TinyLlama-1.1B-Chat-v1.0<\/code><\/pre>\n<\/p><\/div>\n<p>To search for &#8220;llama&#8221; models:<\/p>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang\">Shell<\/span><\/div>\n<pre><code>minerva-llm-list llama<\/code><\/pre>\n<\/p><\/div>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang code-output\">Output<\/span><\/div>\n<pre><code>Llama-3.2-1B-Instruct\r\nLlama-3.2-3B-Instruct\r\nLlama-3.3-70B-Instruct<\/code><\/pre>\n<\/p><\/div>\n<\/p><\/div>\n<\/details>\n<p>  <!-- Section 5: Shared Model in Python --><\/p>\n<details class=\"doc-section\" id=\"python-usage\">\n<summary><span class=\"eyebrow\">5<\/span> Using a Shared Model in Python<\/summary>\n<div class=\"section-body\">\n<p>Create an execution script (e.g., <code>example.py<\/code>) to run local model inference:<\/p>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang code-python\">Python<\/span><\/div>\n<pre><code>import os\r\nfrom transformers import AutoTokenizer, AutoModelForCausalLM\r\n\r\n# Access models folder via environment variable\r\nmodel = os.path.join(os.environ[\"MINERVA_LLM_MODELS\"], \"TinyLlama-1.1B-Chat-v1.0\")\r\nprint(f\"Loading model: {model}\")\r\n\r\ntokenizer = AutoTokenizer.from_pretrained(model)\r\nllm = AutoModelForCausalLM.from_pretrained(model)\r\n\r\ninputs = tokenizer.apply_chat_template(\r\n    [{\"role\": \"user\", \"content\": \"What are the main causes of cardiovascular disease?\"}],\r\n    tokenize=True,\r\n    return_tensors=\"pt\",\r\n    return_dict=True,\r\n    add_generation_prompt=True,\r\n)\r\n\r\noutput = llm.generate(**inputs, max_new_tokens=100)\r\nprint(tokenizer.decode(output[0][inputs[\"input_ids\"].shape[1]:], skip_special_tokens=True))<\/code><\/pre>\n<\/p><\/div>\n<p>Execute the Python script:<\/p>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang\">Shell<\/span><\/div>\n<pre><code>python example.py<\/code><\/pre>\n<\/p><\/div>\n<p>Expected Script Output:<\/p>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang code-output\">Output<\/span><\/div>\n<pre><code>Loading model: \/hpc\/packages\/minerva-rocky9\/minerva-llm\/huggingface\/TinyLlama-1.1B-Chat-v1.0\r\nLoading weights: 100%|\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588\u2588| 201\/201 [00:00&lt;00:00, 1721.67it\/s]\r\n\r\nThere are several main causes of cardiovascular disease (CVD), including:\r\n1. Smoking: Smoking is the leading cause of CVD. It causes the blood vessels to narrow, making it harder for blood to flow through them.\r\n2. High blood pressure: High blood pressure is a major risk factor for CVD. It can cause the arteries to become narrow and hard to open, which can lead to heart attacks and strokes.<\/code><\/pre>\n<\/p><\/div>\n<\/p><\/div>\n<\/details>\n<p>  <!-- Section 6: Login Nodes --><\/p>\n<details class=\"doc-section\" id=\"login-nodes\">\n<summary><span class=\"eyebrow\">6<\/span> Running on a Login Node<\/summary>\n<div class=\"section-body\">\n<p>Running multi-threaded model inference directly on login nodes may trigger thread limits:<\/p>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang code-output\">Output<\/span><\/div>\n<pre><code>libgomp: Thread creation failed: Resource temporarily unavailable<\/code><\/pre>\n<\/p><\/div>\n<p>To avoid this, set environment limits to restrict CPU usage before launching your script on login nodes:<\/p>\n<div class=\"code-block\">\n<div class=\"code-block-header\"><span class=\"code-lang\">Shell<\/span><\/div>\n<pre><code>export OMP_NUM_THREADS=1\r\nexport MKL_NUM_THREADS=1\r\nexport OPENBLAS_NUM_THREADS=1\r\nexport NUMEXPR_NUM_THREADS=1<\/code><\/pre>\n<\/p><\/div>\n<div class=\"note\">\n        <strong>Note:<\/strong> These thread restriction settings are generally not required when executing workloads on dedicated compute nodes.\n      <\/div>\n<\/p><\/div>\n<\/details>\n<footer class=\"doc-footer\">Minerva High Performance Computing<\/footer>\n<\/div>\n<p>[\/et_pb_text][\/et_pb_column][\/et_pb_row][\/et_pb_section][et_pb_section fb_built=&#8221;1&#8243; _builder_version=&#8221;4.27.4&#8243; _module_preset=&#8221;default&#8221; global_colors_info=&#8221;{}&#8221;][et_pb_row _builder_version=&#8221;4.27.4&#8243; _module_preset=&#8221;default&#8221; global_colors_info=&#8221;{}&#8221;][et_pb_column type=&#8221;4_4&#8243; _builder_version=&#8221;4.27.4&#8243; _module_preset=&#8221;default&#8221; global_colors_info=&#8221;{}&#8221;][\/et_pb_column][\/et_pb_row][\/et_pb_section]<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Scientific Computing and Data \/ High Performance Computing \/ Documentation \/ The minerva-llm Module HPC Documentation User Guide for the minerva-llm Module Overview The minerva-llm module provides access to centrally managed, locally hosted, and approved AI models for researchers to use in Python and other applications on Minerva. These models are hosted within the Minerva [&hellip;]<\/p>\n","protected":false},"author":624,"featured_media":0,"parent":35,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_et_pb_use_builder":"on","_et_pb_old_content":"","_et_gb_content_width":"","footnotes":""},"class_list":["post-14712","page","type-page","status-publish","hentry"],"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/labs.icahn.mssm.edu\/minervalab\/wp-json\/wp\/v2\/pages\/14712","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/labs.icahn.mssm.edu\/minervalab\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/labs.icahn.mssm.edu\/minervalab\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/labs.icahn.mssm.edu\/minervalab\/wp-json\/wp\/v2\/users\/624"}],"replies":[{"embeddable":true,"href":"https:\/\/labs.icahn.mssm.edu\/minervalab\/wp-json\/wp\/v2\/comments?post=14712"}],"version-history":[{"count":6,"href":"https:\/\/labs.icahn.mssm.edu\/minervalab\/wp-json\/wp\/v2\/pages\/14712\/revisions"}],"predecessor-version":[{"id":14775,"href":"https:\/\/labs.icahn.mssm.edu\/minervalab\/wp-json\/wp\/v2\/pages\/14712\/revisions\/14775"}],"up":[{"embeddable":true,"href":"https:\/\/labs.icahn.mssm.edu\/minervalab\/wp-json\/wp\/v2\/pages\/35"}],"wp:attachment":[{"href":"https:\/\/labs.icahn.mssm.edu\/minervalab\/wp-json\/wp\/v2\/media?parent=14712"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}