I think the idea is good, I remember it was raised on one of the code reviews
by @FrozenGene, to fulfill this use-case of users who are not interested in
`tvmc run` their modules, but instead, want to get an object file to link their
own apps.
The question now is how do we name optional thing
# [RFC] CSE
## Motivation
**C**ommon **S**ubexpression **E**limination is a compiler optimization that
effectively avoids repeated computation of the same expression. It is needed in
the AutoDiff submodule to improve its performance. As an example, consider a
compute definition that is as simpl
One thing to note is that the current TVM support this kind of behavior, by
generating a `library.tar`, so if the output file format is specified to be
`.tar` then it will output a tarball with a collection of object/c files that
can be statically linked to a target app.
---
[Visit
Topic
Hi @tqchen,
`tvmc` saves directly the `.so`, `.params` and `.json` in the the `.tar` file
it generates. This happens in `tvmc/compiler.py`. I might be wrong, but
probably this is because it doesn't want to store the `.c` files in the final
artifact (@leandron, can you confirm this?).
---
I was referring to the `irmodule.export_library()` API, where we can specify a
tar. Perhaps we can try to unify the API terminology a bit. Note that under the
new runtime packaging, we can directly package params json into dylib.
---
[Visit
Topic](https://discuss.tvm.apache.org/t/rfc-opti
>From what I see, in `tvmc.compiler`, `export_library()` is called with a
>`mod.so` input.
I agree we could generate directly the `tar` file, but I think this was done to
avoid storing the `.c` files (@leandron will know more than me on this).
As for storing directly in the dylib, I am not
Hi all,
The Apache TVM (incubating) community is happy to announce Apache TVM
(incubating) version 0.7.0!
Apache TVM (incubating) is an open deep learning compiler stack for
CPUs, GPUs, and specialized accelerators. It aims to close the gap
between the productivity-focused deep learning framework
My code review is what TQ said. When we call `export_library`, we could save
`a.tar` or `a.so`. If we save `a.tar`, which contains the object file (like
a.o), this is different with `tvmc`'s `tar` collections.
---
[Visit
Topic](https://discuss.tvm.apache.org/t/rfc-optionally-include-objec
Congrats to Apache TVM incubating community!
- Henry
On Fri, Oct 9, 2020 at 5:42 PM Ziheng Jiang wrote:
> Hi all,
>
> The Apache TVM (incubating) community is happy to announce Apache TVM
> (incubating) version 0.7.0!
>
> Apache TVM (incubating) is an open deep learning compiler stack for
> CPU
Congrats!
On 10 Oct 2020 Sat at 05:46 Henry Saputra wrote:
> Congrats to Apache TVM incubating community!
>
> - Henry
>
> On Fri, Oct 9, 2020 at 5:42 PM Ziheng Jiang wrote:
>
> > Hi all,
> >
> > The Apache TVM (incubating) community is happy to announce Apache TVM
> > (incubating) version 0.7.
10 matches
Mail list logo
