Click on the top " AI proper way ", select the "star" Public No.
Heavy dry goods, the first time served
Depth learning to learn, always not open around the GPU, running model. Even if force training model, the most appropriate or suitable for running on some servers depth study. I also used a lot of deep learning server. But often encounter two problems: First, the count is not enough, the second is more expensive. Relatively speaking, GPU cloud server is relatively expensive. Recent deep brain chain classmates let me try the GPU cloud server via DBC Token settlement, feeling quite cheap and easy to use! So, red stones to reach deep brain chain cooperation, the establishment of a dedicated GPU AI proper way cloud server rental platform at:
https://www.redstonewill-gpu.com
GPU Cloud Server Features
1. Deep Brain chain technology: safe and reliable
Deep brain chain provides a low-cost products for artificial intelligence, privacy, flexible, secure decentralized artificial intelligence computing platform. Deep brain chain aims to increase the use of artificial intelligence in the AI block chain three blocks: arithmetic, calculate power, the data to solve the bottleneck problem of the current industry.
Deep brain chain based GPU server technology platform is the use of digital currency DBC Token settlement. At first, I was hesitant for this, after all, is something new, and we do not want to pass this involves "currency speculation", just hope that we can be better and cheaper ways to use GPU cloud server, the form does not important. Also the first time I tried a little, very smooth process, but also easily ran down yolo v3 models.
Therefore, I think it is recommended for everyone to use. If your computer does not have GPU and wanted to run more complex deep learning model, or play with the machine temporarily need the GPU, the GPU can be rented server.
2. Type abundant, cheap
For most people (including me), the most important is the price. It can be said that the GPU cloud server lease is still very cheap. In the pages of "renting GPU -> Machine List" you can see all the currently available GPU types, very rich in species, including the common 1080TI and 2080TI.
Price GPU cloud server available denominated in US dollars, you can see, rented a single 2080TI GPU price is probably around 1 per hour:
The 1080TI is cheap, about 0.6 yuan per hour:
It can be said the price is very cheap, easy to use from the 2080TI!
3. exclusive benefits
To my readers to maximize the benefits, I let the students a deep brain chain to each [AI] proper way to provide users with new 2400 DBC, the equivalent of 30 dollars, you can use about 48 hours GPU cloud server, basically running a deep learning of the model!
At the same time, each GPU cloud servers are can apply for trial, the probation time is half an hour. Try fee: 0 DBC, probation conditions: the current holders of the purse than or equal to 1000 DBC.
In addition, in order to better serve you, we solve the problem in use, I deliberately built a seed exchange user group (group two-dimensional code, see end of text, or add assistant micro letter: AI_helper01 , marked " GPU ", invited the group), welcomed the students interested to join. For any problems during use of deep brain chain of students will timely answers in the group.
how to use?
Here are the most important elements, teach you how to rent GPU dedicated server. After familiar with the process, the future is very convenient to use.
1. Create a wallet
First, open the page: https: //www.redstonewill-gpu.com, click "Start renting GPU":
Then in the "My wallet" inside "Create wallet account":
Set the new password in the wallet, click "Create."
Download private key file private key file and the private key is the user ID code, please save a backup to prevent leakage (download the private key code has been registered and can not get back, be sure to keep).
After creating the wallet is complete, you can use the private key to open their wallet!
We note that the number of DBC is our "wealth value", the following is the US dollar equivalent.
2. Bind-mail
For the first time, in "my machine", click on "Binding mailbox"
Binding mailbox for receiving information about renting the machine, including login ID and password will be sent to you in the form of e-mail. Note that each requires a certain rental DBC for verifying the balance (approximately 0.5 DBC).
3. 购买 DBC
创建钱包之后，下一步我们就要购买合适的 DBC ，来租用 GPU 云服务器。
首先，在我的钱包里点击“如何购买 DBC”。
然后选择一种购买方式，例如我选择“直接购买DBC”。
接着输入购买金额，例如我想买 50 元，可以看到等值约 4200 DBC。
最后，点击继续，选择支付方式。可以直接使用支付宝支付，非常方便！
至此，我们就成功购买了 DBC 啦！下一步就可以来选择机器了。
4. 开始租用
在持有一定的 DBC 后，可以进入机器列表选择需要的 GPU 服务器，可以试用（免费半小时）或者租用：选择机器类型 --> GPU 卡数量和选择租用时间 --> 生成订单 --> 确认支付就可以了，这个环节中都需要一点验证时间，全部是通过邮件进行提示的，这是我最终收到的一封租用成功的机器信息（已经过期）：
值得注意的是，如果机器租用时间到期，机器所有数据将被清空，请注意保持租用时间充足。机器使用完成后，可以提前退租，只会收取实际产生的费用，剩余费用将会退还到钱包中。
然后，根据邮件中的 IP：111.44.254.133，端口号：53940，用户名：root，密码：jRU616aK，就可以通过 ssh 远程登录该服务器（推荐使用 Xshell 软件），剩下的就和平常使用 GPU 机器差不多了。
下面是我的登录界面：
在该服务器里，我们可以输入命令：nvidia-smi，查看 GPU 信息：
训练实例
下面，我来在该 GPU 服务器上跑一个简单的 yolo v3 的 darknet 网络实例。
1. 下载 darknet 网络
输入命令：
$ git clone https://github.com/pjreddie/darknet
2. 编译 darknet 网络
输入命令：
$ cd darknet
$ make
3. 下载预训练权重
输入命令：
$ wget https://pjreddie.com/media/files/yolov3.weights
4. 测试模型
输入命令：
$ ./darknet detect cfg/yolov3.cfg yolov3.weights data/dog.jpg
Test Results:
You can see, the results of target detection displayed.
Here just to make a simple test on the GPU cloud server. They can use their own data sets on the server rented retrain yolo v3 models.
Finally, with regard to the use of DBC GPU server, because I just use some of the problems is not very clear, welcome students interested in the following two-dimensional code sweep join [AI] proper way -DBC GPU user base, there will be the beginning of the deep the students responsible for answering brain chain, join now and use will receive 2400DBC Token, approximately equal to the duration of use for 48 hours. If the user than 100 people, can not join, or add assistant micro letter: AI_helper01 , marked " GPU ", invited into the group).
Note: DBC is an open source network operator force to support the AI community worldwide free GPU cloud deployment platform, GitHub address: https: //github.com/DeepBrainChain/DBChainWebsite.git
