TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Colab Pro

225 点作者 rahidz超过 5 年前

19 条评论

fibrennan超过 5 年前
Just wanted to share a Colab alternative I work on called Gradient[0] (also includes a free GPU).<p>Some of the key differences:<p>- Faster storage. Colab uses Google Drive which is convenient to use but very slow. For example, training datasets often contain a large amount of small files (eg 50k images in the sample TensorFlow and PyTorch datasets). Colab will start to crawl when it tries to ingest these files which is a really standard workflow for ML&#x2F;DL. It&#x27;s great for toy projects eg training MNIST but not for training more interesting models that are popular in the research&#x2F;professional communities today.<p>- Notebooks are fully persistent. With Colab, you need to re-install everything every time you start your Notebook.<p>- Colab instances can be shutdown (preempted) in the middle of a session leading to potential loss of work. Gradient will guarantee the entire session.<p>- Gradient offers the ability to add more storage and higher-end dedicated GPUs from the same environment. If you want to train a more sophisticated model that requires say a day or two of training and maybe a 1TB dataset, that&#x27;s all possible. You could even use the 1-click deploy option to make your model available as an API endpoint. The free GPU tier is just an entrypoint into a full production-ready ML pipeline. With Colab, you would need to take your model somewhere else to accomplish these more advanced tasks.<p>- A large repository of ML templates that include all the major frameworks eg the obvious TensorFlow and PyTorch but also MXNet, Chainer, CNTK, etc. Gradient also includes a public datasets repository with a growing list of common datasets freely available to use in your projects.<p>Those are the main pieces but happy to elaborate on any of this or other questions!<p>[0] <a href="https:&#x2F;&#x2F;gradient.paperspace.com" rel="nofollow">https:&#x2F;&#x2F;gradient.paperspace.com</a>
评论 #22279588 未加载
评论 #22281752 未加载
评论 #22283219 未加载
freediver超过 5 年前
Good idea, but it&#x27;s the first premium product that I&#x27;ve seen where the pitch is &#x27;you <i>may</i> get certain features if you subscribe&#x27;. In another words there is no guarantee and a premium subscriber may still end up with same GPU as a free user. You may end up with a high-end V100 (not available to free) might be a better pitch.
评论 #22280228 未加载
评论 #22278849 未加载
评论 #22283693 未加载
iddan超过 5 年前
Colab is the best notebook I&#x27;ve ever used. It is a real game-changer and I can totally understand why people who use daily would pay for it.
评论 #22281213 未加载
评论 #22278637 未加载
评论 #22281332 未加载
mark_l_watson超过 5 年前
I am tempted to sign up. Colab is very usable on Safari for iOS&#x2F;iPad.<p>I invested 18 months ago in a GPU setup for home. Really convenient but I somewhat regret the purchase. I used to spin up GCP GPU instanced when needed and that was not convenient. Colab is very convenient.<p>$10&#x2F;month for better GPUs and longer sessions seems like a good deal.
hobofan超过 5 年前
I hate to be the boy who cries &quot;Google will cancel this service&quot;, but this offering just seems strange.<p>With a very low price point coupled with not that huge of a user base, this will end up making how much for Google? $1MM&#x2F;month? $10MM&#x2F;month? Either would be negligible for them.
评论 #22281780 未加载
评论 #22283415 未加载
评论 #22281216 未加载
评论 #22281374 未加载
评论 #22282841 未加载
ludwigschubert超过 5 年前
Can anyone see a reason why they wouldn’t just allow you to provision (and pay for) a persistent Google Cloud VM instead? (I currently do that manually and need port forwarding to a machine that runs Jupyter.)<p>It’s hard for me to understand why Colab would build such a vague pro tier instead of the simplest possible solution: let me pay for my compute.<p>There’s so much more potential, too; they could offer whole clusters on demand, with really simple Python integrations say using dask, or ray.
评论 #22279604 未加载
评论 #22279434 未加载
评论 #22279005 未加载
评论 #22279038 未加载
评论 #22278994 未加载
评论 #22279015 未加载
评论 #22283456 未加载
评论 #22279106 未加载
m0zg超过 5 年前
Not sure who this is geared towards. People mostly use Colab to share GPU-dependent work from what I can tell. How would that work on a paid subscription? Do others need to pay to run the notebooks you shared? Can they use their &quot;free&quot; account?<p>As far as utility for research, as a researcher, I _already have_ several local GPUs at my disposal, and I only use notebooks to kick the tires on things and visualize. The moment something starts to look like it&#x27;s useful, I move it to a real *.py file where it&#x27;s more maintainable and diffable.<p>Edit: actually I now think I know who this is geared towards. It&#x27;s geared towards people who aren&#x27;t going to really use it, and don&#x27;t mind to pay $120&#x2F;yr (+tax) for something they don&#x27;t use. Which, IMO, is pretty smart.
minimaxir超过 5 年前
A preemptible P100 + VM on Google Compute Engine is about ~$0.45&#x2F;hr, so to exceed that value with Colaboratory Pro (ignoring conveience factors) you&#x27;d need to train for more than 22 hours in a month. Which, for deep learning, is not too unreasonable.<p>Reading between the lines of both the signup page and up-to-date FAQ, it seems like the free TPU in Colab notebooks will be depreciated, which isn&#x27;t too surprising.
评论 #22279242 未加载
评论 #22278614 未加载
dankle超过 5 年前
&gt; For now, Colab Pro is only available in the US.
评论 #22280557 未加载
bhl超过 5 年前
I wonder who made the decision to spin this out into a commercial product; maybe it has to do with Google&#x27;s push into the cloud further? I always thought Colab was just an experimental tool; it&#x27;s still under the research.google domain.
jonbaer超过 5 年前
I wish they would connect Colab under <a href="https:&#x2F;&#x2F;script.google.com" rel="nofollow">https:&#x2F;&#x2F;script.google.com</a> so you can run a notebook at interval times, something akin to what <a href="https:&#x2F;&#x2F;github.com&#x2F;TensorTom&#x2F;colabctl" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;TensorTom&#x2F;colabctl</a> does.
ccarpenterg超过 5 年前
I&#x27;ve been using Colab for over a year now. I train deep learning models on NLP and medical imaging datasets.<p>It&#x27;s a great tool and it lets you focus on the code and the models, instead of the hardware and OS. But $9.99&#x2F;month is a little expensive for my taste.<p>You can&#x27;t customize it and if they change something you have to install software by hand sometimes. It should be $1.99&#x2F;month, that&#x27;s the kind of price I&#x27;d pay for this basic cloud computing service.<p>edit: I use Colab to play with ML models. I really don&#x27;t think it&#x27;s possible, for instance, to train a model on Imagenet using Colab. So Colab is similar to the microwave, if you want to cook a serious recipe you should use a real kitchen.
评论 #22279339 未加载
评论 #22279275 未加载
评论 #22279172 未加载
评论 #22279299 未加载
评论 #22279279 未加载
wildermuthn超过 5 年前
A lot of comments are missing the value here: cheap and easy TPU access for hobbyist use of deep learning models that need TPUs for fine-tuning and&#x2F;or inference (GPT-2, I’m looking at you).
bitxbit超过 5 年前
This is so much better than buying your own hardwares.
agluszak超过 5 年前
I wonder how long it takes until Google shuts down the free service. It&#x27;s so easy to abuse it. And what Google gets in return?
评论 #22281981 未加载
gizmodo59超过 5 年前
Some content in their github repo: <a href="https:&#x2F;&#x2F;github.com&#x2F;googlecolab" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;googlecolab</a>
fulafel超过 5 年前
This seems to be a hosted Jupyter service, right mybinder, is that right?
评论 #22275064 未加载
lupire超过 5 年前
--
评论 #22279768 未加载
zapf超过 5 年前
There&#x27;s so much data in this universe, people don&#x27;t know what to do with it. When people don&#x27;t know what to do, an industry grows to let them &quot;feel&quot; they are doing something useful.