I updated the blog with the reference. Basically it crashes to compile the model with https://github.com/NodLabs/shark-samples/blob/main/examples/.... The coremltools converter is very version specific (like all vendor conversion kits) and still on a version of TF I couldn't get on conda. Also it doesn't allow for training and only FP16 for inference with ANE. All our tests were with FP32.
In my experience there are many lesser significant hyperparameters that can impact performance when going from the released code to your personal favorite framework.
Nothing you can't figure out by reading source code of the two frameworks or by reading the documentation closely.
Generally, people don't seem to care about reproducing exact metrics - as long as it is close enough they're happy. You need to dig a bit deeper if you want the full quality.
There can be hidden variables that the PID system failed to account for. Workload distribution over time, temperature at the time of day, heat leaking into the datacenter, etc. I'm not sure if a PID based controller would account for such things.
What remains to be seen, however, is whether these manipulations keep the machines working at similar efficiencies and if they fail at similar rates as before turning on the control system.
Peter Watts's Blindsight - good fiction on the one hand, and exhaustively referenced on the other. (Watts is the only sf author I've run across whose novels include discussions and bibliographies of the research he used to inform and shape the narrative. In addition to earning him a new high score on the hardness scale, this seems counterintuitively to produce better stories, not worse ones.)
The fact that this file format doesn't enclose the AdobeRGB vs sRGB information is a deal breaker.
A lot of computer vision research datasets uses a similar file format - so I can see its use there. But things like movie creation, publishing, etc will probably never adopt this.
Indeed. I'll reiterate - it's vital that images have a known color space. Even if OP just states that farbfield is sRGB, it'll help. Otherwise, you've only got a 3D matrix, rather than an image.
