Interpreter v1.0.0

[itemkelvin]

All ops in file NeuraOps.dt have been implemented, and test cases have been saved in path ./test/neura/interpreter/ .

[tancheng]

Thanks for the update @itemkelvin, it seems we need to fix the Github Action by revising the CHECK inside the .mlir (as the dump is removed?). Maybe we can just CHECK the final output?

Thanks!

[itemkelvin]

Thanks for the update @itemkelvin, it seems we need to fix the Github Action by revising the CHECK inside the .mlir (as the dump is removed?). Maybe we can just CHECK the final output?

Thanks!

Correct CHECKs have been added to the test cases, and the workflow in Git Action has run successfully.

[tancheng]

Excellent job. I just have a concern about the dump's condition. I believe deepseek/gpt can help you set up the flag for interpreter to enable the conditional dump. Plz try, thanks @itemkelvin~!

[tancheng]

Hi @itemkelvin, any blockers to resolve the comments? We can discuss alternatives if they are hard to fix?

[itemkelvin]

Hi @itemkelvin, any blockers to resolve the comments? We can discuss alternatives if they are hard to fix?

I sincerely apologize for the delay in code development. I've been swamped with other tasks these days, which has affected my progress. Rest assured, I'll prioritize this project and catch up as soon as possible. Now I've completed all the code modifications, and all op execution logs support verbose mode. In addition, I've adjusted the logging format to make it easier to monitor the execution status as follows.

[tancheng]

Hi @itemkelvin, any blockers to resolve the comments? We can discuss alternatives if they are hard to fix?

I sincerely apologize for the delay in code development. I've been swamped with other tasks these days, which has affected my progress. Rest assured, I'll prioritize this project and catch up as soon as possible. Now I've completed all the code modifications, and all op execution logs support verbose mode. In addition, I've adjusted the logging format to make it easier to monitor the execution status as follows.

Thanks @itemkelvin~! Great work~!

I just left a few more comments, which can be easily fixed by you.

The next step could be interpreter for dataflow execution #42, plz let @ShangkunLi know if you have any questions when you enable/implement it for https://github.com/coredac/dataflow/blob/main/test/neura/ctrl/branch_for.mlir. And feel free to schedule 1:1 with me if needed :-)

[tancheng]

Thanks~! Let's see whether the test passes in Github Action so then we can merge this PR :-)