What are expected response times on the medium plan?
On the Medium Plan, we deploy your models on our GPUs, so response times are much faster.
Note: This does not include network latencies.
What are the expected accuracies?
On the Medium Plan, our team of Deep Learning Engineers will run custom experiments for you. On average we're able to get most models' accuracies to >90%
Can I get some help with code integration?
You can check out https://app.nanonets.com/documentation for more info
If you have any adhoc request, feel free to ask. We'll try and help you here.
Can I deploy my models on-premise?
On our Medium plan, you can download fully managed lightweight Docker images that contain your models. These can be both online and completely offline.
Do you use my data to train other users' models?
We don't use your data anywhere else.
How many API calls?
100K API calls.
On the Medium plan, you can make 100K Prediction requests/month.
Will you label/annotate my images?
Yes. 500 images per month.
On our Medium plan, our in house team of annotators will label/annotate 500 images per month for you.
How much data do you need? ->
25-50 images per label.
Can the model run on edge devices?
On our medium plan, we provide lightweight fully managed docker images. These docker images can run on Rasberry PIs, CPUs, GPUs etc
Can the models work real-time?
Our models have been deployed in real time use cases. We recommend using our docker images to reduce network latencies if real-time is necessary.
Can I buy the model forever?
You can get in touch with our sales team to discuss more details at firstname.lastname@example.org
Will you give me the model file?
On our medium plan, we'll give you your model file to use with any deep learning framework of your choice.,
How many requests can i make concurrently?
You can make as many concurrent requests as you like.
What image sizes/ dimensions can be handled?
1000x1000 - Free Plan.
3000x3000 - Medium Plan.
We support larger image sizes and dimensions on the medium plan.
Maximum size of images used in training?
< 2MB - Free Plan.
< 10MB - Medium Plan.
Do you support video?
While we do support video, we'd recommend sampling the frames at your side and sending us the images instead. Happy to share the code that does this on a video stream and send the images for prediction.