[TVM Discuss] [Questions] BYOC: Runtime access Tensor values

Hi, I am working on a BYOC backend and finally reached the runtime level. But I've looked at the DNNL JSON example (I am using JSON as well) and have not found a way to access the real tensor values to serialize them for my target. >From the JSON serialization I can access the shapes, etc, but n

[TVM Discuss] [Questions] [BYOC] Codegen, JSON, non-composite

Thank you, but I have got another question: Is there a way to actually see the raw JSON output, that has been generated by the backend easily? --- [Visit Topic](https://discuss.tvm.ai/t/byoc-codegen-json-non-composite/7719/3) to respond. You are receiving this because you enabled mailing

[TVM Discuss] [Questions] [BYOC] Codegen, JSON, non-composite

Hi, I am in the process of building a codegeneration for my own (currently abstract) accelerator. And finally got the graph annotation working correctly. My goal was to directly generate the instructions from its ISA during the codegeneration phase in TVMs BYOC flow, and adapt the runtime acco

[TVM Discuss] [Questions] Post-Processing Ops and Quantization

Hi, I am currently exploring relay with the BYOC infrastructure and realized that pooling, relu and a number of other support operations are still done using float32. As my target accelerator supports pool, relu and activations only in the int8 range, I want to quantize all operations. Am I

[TVM Discuss] [Questions] Codegeneration for own DLA Instruction Set

I have another problem: The graph annotation has been defined (by adding my own version at python/tvm/relay/op/contrib/test_dla.py) But it seems like that is not enough to get the annotation going, as `mod_t = transform.AnnotateTarget("test_dla")(mod)` followed by `mod_t = transform.Parti