Buckets:

hf-doc-build
/

doc-dev

Files

xet

hf-doc-build/doc-dev / transformers /pr_33913 /en /perf_hardware.html

rtrm

29 days ago

download

raw

25.2 kB

	<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"Custom hardware for training","local":"custom-hardware-for-training","sections":[{"title":"GPU","local":"gpu","sections":[{"title":"Power and Cooling","local":"power-and-cooling","sections":[],"depth":3},{"title":"Multi-GPU Connectivity","local":"multi-gpu-connectivity","sections":[{"title":"NVlink","local":"nvlink","sections":[],"depth":4}],"depth":3}],"depth":2}],"depth":1}">
	<link href="/docs/transformers/pr_33913/en/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload">
	<link rel="modulepreload" href="/docs/transformers/pr_33913/en/_app/immutable/entry/start.b67f883f.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33913/en/_app/immutable/chunks/scheduler.25b97de1.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33913/en/_app/immutable/chunks/singletons.62a184e0.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33913/en/_app/immutable/chunks/index.e188933d.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33913/en/_app/immutable/chunks/paths.51881b9e.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33913/en/_app/immutable/entry/app.e436b1f2.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33913/en/_app/immutable/chunks/index.d9030fc9.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33913/en/_app/immutable/nodes/0.05e395f5.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33913/en/_app/immutable/chunks/each.e59479a4.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33913/en/_app/immutable/nodes/378.c8d73ca9.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33913/en/_app/immutable/chunks/CodeBlock.e6cd0d95.js">
	<link rel="modulepreload" href="/docs/transformers/pr_33913/en/_app/immutable/chunks/EditOnGithub.91d95064.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{"title":"Custom hardware for training","local":"custom-hardware-for-training","sections":[{"title":"GPU","local":"gpu","sections":[{"title":"Power and Cooling","local":"power-and-cooling","sections":[],"depth":3},{"title":"Multi-GPU Connectivity","local":"multi-gpu-connectivity","sections":[{"title":"NVlink","local":"nvlink","sections":[],"depth":4}],"depth":3}],"depth":2}],"depth":1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <h1 class="relative group"><a id="custom-hardware-for-training" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#custom-hardware-for-training"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Custom hardware for training</span></h1> <p data-svelte-h="svelte-iifysf">The hardware you use to run model training and inference can have a big effect on performance. For a deep dive into GPUs make sure to check out Tim Dettmer’s excellent <a href="https://timdettmers.com/2020/09/07/which-gpu-for-deep-learning/" rel="nofollow">blog post</a>.</p> <p data-svelte-h="svelte-1dshoiu">Let’s have a look at some practical advice for GPU setups.</p> <h2 class="relative group"><a id="gpu" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#gpu"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>GPU</span></h2> <p data-svelte-h="svelte-eleqzp">When you train bigger models you have essentially three options:</p> <ul data-svelte-h="svelte-pz3vuj"><li>bigger GPUs</li> <li>more GPUs</li> <li>more CPU and NVMe (offloaded to by <a href="main_classes/deepspeed#nvme-support">DeepSpeed-Infinity</a>)</li></ul> <p data-svelte-h="svelte-z1tcso">Let’s start at the case where you have a single GPU.</p> <h3 class="relative group"><a id="power-and-cooling" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#power-and-cooling"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Power and Cooling</span></h3> <p data-svelte-h="svelte-ppjikm">If you bought an expensive high end GPU make sure you give it the correct power and sufficient cooling.</p> <p data-svelte-h="svelte-11nr6dq"><strong>Power</strong>:</p> <p data-svelte-h="svelte-11nhv2k">Some high end consumer GPU cards have 2 and sometimes 3 PCI-E 8-Pin power sockets. Make sure you have as many independent 12V PCI-E 8-Pin cables plugged into the card as there are sockets. Do not use the 2 splits at one end of the same cable (also known as pigtail cable). That is if you have 2 sockets on the GPU, you want 2 PCI-E 8-Pin cables going from your PSU to the card and not one that has 2 PCI-E 8-Pin connectors at the end! You won’t get the full performance out of your card otherwise.</p> <p data-svelte-h="svelte-150xmew">Each PCI-E 8-Pin power cable needs to be plugged into a 12V rail on the PSU side and can supply up to 150W of power.</p> <p data-svelte-h="svelte-12gfejp">Some other cards may use a PCI-E 12-Pin connectors, and these can deliver up to 500-600W of power.</p> <p data-svelte-h="svelte-uh6lop">Low end cards may use 6-Pin connectors, which supply up to 75W of power.</p> <p data-svelte-h="svelte-s9j0io">Additionally you want the high-end PSU that has stable voltage. Some lower quality ones may not give the card the stable voltage it needs to function at its peak.</p> <p data-svelte-h="svelte-8g3p2u">And of course the PSU needs to have enough unused Watts to power the card.</p> <p data-svelte-h="svelte-1gvtzcc"><strong>Cooling</strong>:</p> <p data-svelte-h="svelte-hba1wh">When a GPU gets overheated it will start throttling down and will not deliver full performance and it can even shutdown if it gets too hot.</p> <p data-svelte-h="svelte-1ho6s7e">It’s hard to tell the exact best temperature to strive for when a GPU is heavily loaded, but probably anything under +80C is good, but lower is better - perhaps 70-75C is an excellent range to be in. The throttling down is likely to start at around 84-90C. But other than throttling performance a prolonged very high temperature is likely to reduce the lifespan of a GPU.</p> <p data-svelte-h="svelte-1c43r32">Next let’s have a look at one of the most important aspects when having multiple GPUs: connectivity.</p> <h3 class="relative group"><a id="multi-gpu-connectivity" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#multi-gpu-connectivity"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>Multi-GPU Connectivity</span></h3> <p data-svelte-h="svelte-1b7hx6s">If you use multiple GPUs the way cards are inter-connected can have a huge impact on the total training time. If the GPUs are on the same physical node, you can run:</p> <div class="code-block relative"><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->nvidia-smi topo -m<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1m7rqth">and it will tell you how the GPUs are inter-connected. On a machine with dual-GPU and which are connected with NVLink, you will most likely see something like:</p> <div class="code-block relative"><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --> <span class="hljs-attribute">GPU0</span> GPU1 CPU Affinity NUMA Affinity
	<span class="hljs-attribute">GPU0</span> X NV2 <span class="hljs-number">0</span>-<span class="hljs-number">23</span> N/A
	<span class="hljs-attribute">GPU1</span> NV2 X <span class="hljs-number">0</span>-<span class="hljs-number">23</span> N/A<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-hggxkn">on a different machine w/o NVLink we may see:</p> <div class="code-block relative"><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --> <span class="hljs-attribute">GPU0</span> GPU1 CPU Affinity NUMA Affinity
	<span class="hljs-attribute">GPU0</span> X PHB <span class="hljs-number">0</span>-<span class="hljs-number">11</span> N/A
	<span class="hljs-attribute">GPU1</span> PHB X <span class="hljs-number">0</span>-<span class="hljs-number">11</span> N/A<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-869n3z">The report includes this legend:</p> <div class="code-block relative"><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --> X = Self
	SYS = Connection traversing PCIe <span class="hljs-keyword">as</span> well <span class="hljs-keyword">as</span> <span class="hljs-keyword">the</span> SMP interconnect between NUMA nodes (e.g., QPI/UPI)
	NODE = Connection traversing PCIe <span class="hljs-keyword">as</span> well <span class="hljs-keyword">as</span> <span class="hljs-keyword">the</span> interconnect between PCIe Host Bridges <span class="hljs-keyword">within</span> <span class="hljs-keyword">a</span> NUMA node
	PHB = Connection traversing PCIe <span class="hljs-keyword">as</span> well <span class="hljs-keyword">as</span> <span class="hljs-keyword">a</span> PCIe Host Bridge (typically <span class="hljs-keyword">the</span> CPU)
	PXB = Connection traversing multiple PCIe bridges (<span class="hljs-keyword">without</span> traversing <span class="hljs-keyword">the</span> PCIe Host Bridge)
	PIX = Connection traversing <span class="hljs-keyword">at</span> most <span class="hljs-keyword">a</span> single PCIe bridge
	NV<span class="hljs-comment"># = Connection traversing a bonded set of # NVLinks</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-xobw9p">So the first report <code>NV2</code> tells us the GPUs are interconnected with 2 NVLinks, and the second report <code>PHB</code> we have a typical consumer-level PCIe+Bridge setup.</p> <p data-svelte-h="svelte-abx4jy">Check what type of connectivity you have on your setup. Some of these will make the communication between cards faster (e.g. NVLink), others slower (e.g. PHB).</p> <p data-svelte-h="svelte-1hcsw7p">Depending on the type of scalability solution used, the connectivity speed could have a major or a minor impact. If the GPUs need to sync rarely, as in DDP, the impact of a slower connection will be less significant. If the GPUs need to send messages to each other often, as in ZeRO-DP, then faster connectivity becomes super important to achieve faster training.</p> <h4 class="relative group"><a id="nvlink" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#nvlink"><span><svg class="" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg></span></a> <span>NVlink</span></h4> <p data-svelte-h="svelte-1gqi5fx"><a href="https://en.wikipedia.org/wiki/NVLink" rel="nofollow">NVLink</a> is a wire-based serial multi-lane near-range communications link developed by Nvidia.</p> <p data-svelte-h="svelte-gh3n8e">Each new generation provides a faster bandwidth, e.g. here is a quote from <a href="https://www.nvidia.com/content/dam/en-zz/Solutions/geforce/ampere/pdf/NVIDIA-ampere-GA102-GPU-Architecture-Whitepaper-V1.pdf" rel="nofollow">Nvidia Ampere GA102 GPU Architecture</a>:</p> <blockquote data-svelte-h="svelte-14khom1"><p>Third-Generation NVLink®
	GA102 GPUs utilize NVIDIA’s third-generation NVLink interface, which includes four x4 links,
	with each link providing 14.0625 GB/sec bandwidth in each direction between two GPUs. Four
	links provide 56.25 GB/sec bandwidth in each direction, and 112.5 GB/sec total bandwidth
	between two GPUs. Two RTX 3090 GPUs can be connected together for SLI using NVLink.
	(Note that 3-Way and 4-Way SLI configurations are not supported.)</p></blockquote> <p data-svelte-h="svelte-1naftur">So the higher <code>X</code> you get in the report of <code>NVX</code> in the output of <code>nvidia-smi topo -m</code> the better. The generation will depend on your GPU architecture.</p> <p data-svelte-h="svelte-1j0rfs7">Let’s compare the execution of an <code>openai-community/gpt2</code> language model training over a small sample of wikitext.</p> <p data-svelte-h="svelte-j6eczh">The results are:</p> <table data-svelte-h="svelte-1hvmx1i"><thead><tr><th>NVlink</th> <th align="right">Time</th></tr></thead> <tbody><tr><td>Y</td> <td align="right">101s</td></tr> <tr><td>N</td> <td align="right">131s</td></tr></tbody></table> <p data-svelte-h="svelte-192o6bz">You can see that NVLink completes the training ~23% faster. In the second benchmark we use <code>NCCL_P2P_DISABLE=1</code> to tell the GPUs not to use NVLink.</p> <p data-svelte-h="svelte-16jny0z">Here is the full benchmark code and outputs:</p> <div class="code-block relative"><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg class="" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg> <div class="absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0"><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent; "></div> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-comment"># DDP w/ NVLink</span>

	<span class="hljs-built_in">rm</span> -r /tmp/test-clm; CUDA_VISIBLE_DEVICES=0,1 torchrun \
	--nproc_per_node 2 examples/pytorch/language-modeling/run_clm.py --model_name_or_path openai-community/gpt2 \
	--dataset_name wikitext --dataset_config_name wikitext-2-raw-v1 --do_train \
	--output_dir /tmp/test-clm --per_device_train_batch_size 4 --max_steps 200

	{<span class="hljs-string">'train_runtime'</span>: 101.9003, <span class="hljs-string">'train_samples_per_second'</span>: 1.963, <span class="hljs-string">'epoch'</span>: 0.69}

	<span class="hljs-comment"># DDP w/o NVLink</span>

	<span class="hljs-built_in">rm</span> -r /tmp/test-clm; CUDA_VISIBLE_DEVICES=0,1 NCCL_P2P_DISABLE=1 torchrun \
	--nproc_per_node 2 examples/pytorch/language-modeling/run_clm.py --model_name_or_path openai-community/gpt2 \
	--dataset_name wikitext --dataset_config_name wikitext-2-raw-v1 --do_train
	--output_dir /tmp/test-clm --per_device_train_batch_size 4 --max_steps 200

	{<span class="hljs-string">'train_runtime'</span>: 131.4367, <span class="hljs-string">'train_samples_per_second'</span>: 1.522, <span class="hljs-string">'epoch'</span>: 0.69}<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-pittpg">Hardware: 2x TITAN RTX 24GB each + NVlink with 2 NVLinks (<code>NV2</code> in <code>nvidia-smi topo -m</code>)
	Software: <code>pytorch-1.8-to-be</code> + <code>cuda-11.0</code> / <code>transformers==4.3.0.dev0</code></p> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/transformers/blob/main/docs/source/en/perf_hardware.md" target="_blank"><span data-svelte-h="svelte-1kd6by1"><</span> <span data-svelte-h="svelte-x0xyl0">></span> <span data-svelte-h="svelte-1dajgef"><span class="underline ml-1.5">Update</span> on GitHub</span></a> <p></p>

	<script>
	{
	__sveltekit_z647wz = {
	assets: "/docs/transformers/pr_33913/en",
	base: "/docs/transformers/pr_33913/en",
	env: {}
	};

	const element = document.currentScript.parentElement;

	const data = [null,null];

	Promise.all([
	import("/docs/transformers/pr_33913/en/_app/immutable/entry/start.b67f883f.js"),
	import("/docs/transformers/pr_33913/en/_app/immutable/entry/app.e436b1f2.js")
	]).then(([kit, app]) => {
	kit.start(app, element, {
	node_ids: [0, 378],
	data,
	form: null,
	error: null
	});
	});
	}
	</script>

Xet Storage Details

Size:: 25.2 kB
Xet hash:: 62cec87d4c27a039544642b1e3bb8295baa69155c68b4628c08e0dd2e82fa9e1

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.