Thomas
Thomas
@luisquintanilla, I assemble several float32 arrays into one and depending on the model being trained, I have different lengths (the code is quite generic). In this case: ``` type DataRow...
In this case, it won't compile:
It’s a float32 array with all the training data for that row
Ah yes, that compiles! I didn't know that overload was there. Thanks a lot! As a quick side question, can I also load a model from memory? I'm currently loading...
same problem here, setting DOCKER_API_VERSION doesn't work for me ``` DOCKER_API_VERSION=1.40 dry INFO[0000] Dry could not start: retrieve container list: Error response from daemon: client version 1.44 is too new....
I also tried different version numbers and got the same result. FYI, running Dry through docker works on the same machine.
I'm going to check if it's not an issue with f#/async where the thread switches and the call is not finished yet
Thanks Nicola! I'll be able to check this right after a holidays and report back.
@bartelink we've totally dropped the use of DynamoDB in our system, so I haven't touched that system in many months now.
I am on Intel x64.