Adam Gibson
Adam Gibson
Ok @fasahath can you provide the model so we can reproduce this? Thanks @saudet
@mjpt777 Anywhere we can track this? We are using aeron already in our ML stack and trying to figure out a viable path to this.
This looks like a great start. I took the following approach: https://github.com/deeplearning4j/deeplearning4j/blob/master/deeplearning4j-scaleout/spark/dl4j-spark/src/main/java/org/deeplearning4j/spark/util/MLLibUtil.java#L112 I assumed a directory structure like mentioned above. In deep learning I typically see images as well as...
@devenyantis I would try to make the metadata agnostic of the real file on the system. Since you have a local database, I would consider having a local path or...
@CompilerCrash actually these interfaces are all generated. I just had to convert the data type to batchMmul. Good catch! Please see: https://github.com/deeplearning4j/deeplearning4j/blob/2668876811fbebf1fafc09f954a34dfcd8c95780/contrib/codegen-tools/codegen/src/main/ops/org/nd4j/codegen/ops/SDBaseOps.kt#L2143 and data types here: https://github.com/deeplearning4j/deeplearning4j/blob/46dbd0b2035fab86f5d500508af46690b97af7c8/contrib/codegen-tools/codegen/src/main/kotlin/org/nd4j/codegen/api/DataType.kt#L25
@CompilerCrash you can find automatically generated definitions for ops as well: https://github.com/deeplearning4j/deeplearning4j/blob/master/nd4j/nd4j-backends/nd4j-api-parent/nd4j-api/src/main/resources/nd4j-op-def.pbtxt - that's generated by parsing the java and c++ code bases: https://github.com/deeplearning4j/deeplearning4j/tree/master/contrib/codegen-tools/libnd4j-gen That's what we use for model...
@agentgt @remkop I haven't been able to look at this in a while but haven't bumped in to this in a while. I would go ahead and just close this....
Seems like this has been fixed in subsequent versions.
Checked this with the latest version. Doesn't appear to be a problem as of M2.1. Closing.
Closing. We support quite a few models now. For anyone finding this in the future file an issue if we're missing something.