Reviews from AWS Marketplace
0 AWS reviews
-
5 star0
-
4 star0
-
3 star0
-
2 star0
-
1 star0
External reviews
External reviews are not included in the AWS star rating for the product.
Top-tier service for bargain bin prices
What do you like best about the product?
The hosted aspect is great and if you avoid storing large artifacts, too many histograms, or too many images, it's very cheap or free even for heavy usage. I often run training on remote systems and checking the eval outputs remotely can be a pain with other software like TensorFlow. You need to be sure to run the server on your instance to view the dashboard. With W&B you get comparably powerful features, and you don't need to do anything but run your train script. In addition to hosting, the Sweeps functionality is excellent for hyper-parameter sweeping and pre-determined groups of runs in multi-task settings.
What do you dislike about the product?
Some UI elements, like the Runs data table, can be laggy, and sometimes the auto-refresh seems finicky.
What problems is the product solving and how is that benefiting you?
W&B helps track and debug my experiments. I use it for developing new models, observing loss scales and adjusting them so they're in compatible ranges, detecting and addressing gradient instabilities such as vanishing or exploding. I also use it for hyper-parameter searching to find the best values for my training runs. Additionally, I use the Sweeps functionality for coordinating runs that involve many related models that are deployed together.
- Leave a Comment |
- Mark review as helpful
Pretty good and unique solution
What do you like best about the product?
Managing runs that are lauched across different machines
What do you dislike about the product?
All new accounts have to be a team not personal
What problems is the product solving and how is that benefiting you?
Keep tracking of hyperparameter searches and evaluation metrics across different runs
Perfect for seemless experiment tracking
What do you like best about the product?
I use it for every project, it became a key tool for me to keep an overview about experiments no matter if or when using LLMs
What do you dislike about the product?
Working in teams is limited in the free version
What problems is the product solving and how is that benefiting you?
Experiment tracking which helps to keep track of trials
Smooth workflow for AI benchmarks
What do you like best about the product?
The online API is really helpful to organize results and projects.
What do you dislike about the product?
In the online website, the graphs showing evolution of a metric for example can't be dezoomed easily with the scroll wheel and we can't move in the graph after zooming.
What problems is the product solving and how is that benefiting you?
It stores online my results and shows in live the logs. (without the need to run another command to visualize the results or refresh a page)
LLM training tracking
What do you like best about the product?
I like how it seemlessly integrates into the workflow and reports all relevant information.
What do you dislike about the product?
I did not like how it remains logged in even without the login command, and how it starts a different log for the same 'run name'.
What problems is the product solving and how is that benefiting you?
Monitoring my multiple training logs, keeps all the plots in one place for easy access.
Wandb works well out of the box
What do you like best about the product?
Easy to plug in and use. Works well with PyTorch and lightning and easy to compare models when away from local network.
What do you dislike about the product?
Can slow down training speed and there are some minor bugs sometimes when using on a less well supported device/framework. Mobile UI also slow.
What problems is the product solving and how is that benefiting you?
Helping me compare models quickly on the go. A lot easier to use than other solutions and cloud syncing to share with others.
have recommended to others before, would again
What do you like best about the product?
I first learned about w&b 5 years ago in high school, and used it for a few projects. I now use it every day and have convinced several colleagues in and out of the lab to use w&b.
What do you dislike about the product?
i wish there was a way to locally/offline view the graphs, or at least be able to view graphs with latency.
What problems is the product solving and how is that benefiting you?
easy system agnostic logging, and composting across runs
wandb works great and is very easy to setup and use supporting wide variety of media types
What do you like best about the product?
Ease of setting it up, the dashboard is pretty much each to use and allows to visualize a wide range of features at once. The integration is pretty much solid and works out of box for any script that I worked on
What do you dislike about the product?
I used to pre train certain checkpoints in a sequential manner or sometimes my runs used to break in between due to memory/connection issues from there on it was quite difficult to visualize all the previous run in a single curve using the dashboard, setting the x axis as wall time helped but the curve was still not a single continuous graph
What problems is the product solving and how is that benefiting you?
In some of my experiments I wanted to visualize how the grad norm varies with train iter or how does the learning rate scheduler affect the perplexity of the model that I am working on, wandb makes doing these experiments much easier
A no-brainer tool to assist in model training/evaluation/comparison
What do you like best about the product?
Its easy and seamless integration with PyTorch lightening and simple API usage. The model (artifact) and logs logging also help me trace back a model that was training months ago.
What do you dislike about the product?
Nothing as of now. Would appreciate more dark modes and API control to give experiment names, rather than having my experiment named 'Tasty-Aadvark'.
What problems is the product solving and how is that benefiting you?
Having all my training sessions accessble from single site, is the biggest benefit. Also, saving logs, training meta is quite helpful too.
Very quick and easy to start online logging
What do you like best about the product?
This tool is perfect for logging information from programs, especially during training. It allows you to see how your model is training from anywhere in the world. I like how you can just dump some raw data onto the platform, and then you can make your graphs and manipulate the data separately from your training loop.
What do you dislike about the product?
Some features are missing but I am sure they would come if I did a feature request.
What problems is the product solving and how is that benefiting you?
Logging data from machine learning training loops that may be running headless.
showing 11 - 20