[BYOC] Switch TensorRT BYOC integration to IRModule-at-a-time using RelayToTIR hook #11979
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@masahi thanks, you are right. |
…elayToTIR hook This does for the TensorRT integration what apache#11631 did for the CUTLASS integration. - All compilation options are captured within the attributes of a Target of kind "tensorrt" (instead of the "relay.ext.tensorrt.options" attribute in PassContext). This means all BYOC configurations options needed by Collage can be captured uniformly by a list-of-Targets. It also means RPC boundaries (as used internally at OctoML) only need to worry about maintaining the fidelity of the Target instance(s) rather than reaching into the PassContext. - Compilation is switched from function-at-a-time (relying on the TECompiler) to IRModule-at-a-time (using the RelayToTIR target-specific hook mechanism). Though not strictly necessary for Collage I want to check the path is now clear to deprecate the support for BYOC in TEComplier. - Get all the TensorRT tests going again, except for a few I've disabled with x-link to a new issue apache#11765. CAUTION: The TensorRT runtime is not supported in CI so many of these tests are cosmetic. - While trying to track down a 'free(): invalid pointer' error in test_tensorrt_int8_exp.py made the TensorRT allocs/frees more robust, but turns out its also broken in main. No harm leaving these changes in though.
- can't use default Target("tensorrt") arg
mbs-octoml
force-pushed
the
mbs-collage-trt-refresh
branch
from
747cef3 to
8539ef4
Compare
- handle missing runtime in versioning - turn test_maskrcnn_resnet50 back on now that we have the import-torch-first workaround.
|
Pretty sure I've got all the ci failures and this is ready for review. |
Merged
masahi
approved these changes
Jul 1, 2022
|
Thank's @masahi! This can be merged. |
blackkker
pushed a commit
to blackkker/tvm
that referenced
this pull request
Jul 7, 2022
…elayToTIR hook (apache#11979) * [BYOC] Switch TensorRT BYOC integration to IRModule-at-a-time using RelayToTIR hook This does for the TensorRT integration what apache#11631 did for the CUTLASS integration. - All compilation options are captured within the attributes of a Target of kind "tensorrt" (instead of the "relay.ext.tensorrt.options" attribute in PassContext). This means all BYOC configurations options needed by Collage can be captured uniformly by a list-of-Targets. It also means RPC boundaries (as used internally at OctoML) only need to worry about maintaining the fidelity of the Target instance(s) rather than reaching into the PassContext. - Compilation is switched from function-at-a-time (relying on the TECompiler) to IRModule-at-a-time (using the RelayToTIR target-specific hook mechanism). Though not strictly necessary for Collage I want to check the path is now clear to deprecate the support for BYOC in TEComplier. - Get all the TensorRT tests going again, except for a few I've disabled with x-link to a new issue apache#11765. CAUTION: The TensorRT runtime is not supported in CI so many of these tests are cosmetic. - While trying to track down a 'free(): invalid pointer' error in test_tensorrt_int8_exp.py made the TensorRT allocs/frees more robust, but turns out its also broken in main. No harm leaving these changes in though. * - Lints * - Woops, fix test * - lints * - Use default tensorrt target if none given in targets list * - fix free error * - accidentally introduced 'transforms' namespace - can't use default Target("tensorrt") arg * - D'oh! Include ended up #if protected * - restore mark for test_dynamic_offload - handle missing runtime in versioning - turn test_maskrcnn_resnet50 back on now that we have the import-torch-first workaround. * - wibble
masahi
pushed a commit
to masahi/tvm
that referenced
this pull request
Jul 15, 2022
…elayToTIR hook (apache#11979) * [BYOC] Switch TensorRT BYOC integration to IRModule-at-a-time using RelayToTIR hook This does for the TensorRT integration what apache#11631 did for the CUTLASS integration. - All compilation options are captured within the attributes of a Target of kind "tensorrt" (instead of the "relay.ext.tensorrt.options" attribute in PassContext). This means all BYOC configurations options needed by Collage can be captured uniformly by a list-of-Targets. It also means RPC boundaries (as used internally at OctoML) only need to worry about maintaining the fidelity of the Target instance(s) rather than reaching into the PassContext. - Compilation is switched from function-at-a-time (relying on the TECompiler) to IRModule-at-a-time (using the RelayToTIR target-specific hook mechanism). Though not strictly necessary for Collage I want to check the path is now clear to deprecate the support for BYOC in TEComplier. - Get all the TensorRT tests going again, except for a few I've disabled with x-link to a new issue apache#11765. CAUTION: The TensorRT runtime is not supported in CI so many of these tests are cosmetic. - While trying to track down a 'free(): invalid pointer' error in test_tensorrt_int8_exp.py made the TensorRT allocs/frees more robust, but turns out its also broken in main. No harm leaving these changes in though. * - Lints * - Woops, fix test * - lints * - Use default tensorrt target if none given in targets list * - fix free error * - accidentally introduced 'transforms' namespace - can't use default Target("tensorrt") arg * - D'oh! Include ended up #if protected * - restore mark for test_dynamic_offload - handle missing runtime in versioning - turn test_maskrcnn_resnet50 back on now that we have the import-torch-first workaround. * - wibble
mikeseven
pushed a commit
to mikeseven/tvm
that referenced
this pull request
Sep 27, 2023
…elayToTIR hook (apache#11979) * [BYOC] Switch TensorRT BYOC integration to IRModule-at-a-time using RelayToTIR hook This does for the TensorRT integration what apache#11631 did for the CUTLASS integration. - All compilation options are captured within the attributes of a Target of kind "tensorrt" (instead of the "relay.ext.tensorrt.options" attribute in PassContext). This means all BYOC configurations options needed by Collage can be captured uniformly by a list-of-Targets. It also means RPC boundaries (as used internally at OctoML) only need to worry about maintaining the fidelity of the Target instance(s) rather than reaching into the PassContext. - Compilation is switched from function-at-a-time (relying on the TECompiler) to IRModule-at-a-time (using the RelayToTIR target-specific hook mechanism). Though not strictly necessary for Collage I want to check the path is now clear to deprecate the support for BYOC in TEComplier. - Get all the TensorRT tests going again, except for a few I've disabled with x-link to a new issue apache#11765. CAUTION: The TensorRT runtime is not supported in CI so many of these tests are cosmetic. - While trying to track down a 'free(): invalid pointer' error in test_tensorrt_int8_exp.py made the TensorRT allocs/frees more robust, but turns out its also broken in main. No harm leaving these changes in though. * - Lints * - Woops, fix test * - lints * - Use default tensorrt target if none given in targets list * - fix free error * - accidentally introduced 'transforms' namespace - can't use default Target("tensorrt") arg * - D'oh! Include ended up #if protected * - restore mark for test_dynamic_offload - handle missing runtime in versioning - turn test_maskrcnn_resnet50 back on now that we have the import-torch-first workaround. * - wibble
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This does for the TensorRT integration what #11631 did for the CUTLASS integration. See https://discuss.tvm.apache.org/t/byoc-how-backwards-compatible-does-the-tensorrt-partition-for-tensorrt-function-need-to-be/12957 for discussion.
All compilation options are captured within the attributes of a Target of kind "tensorrt" (instead of the "relay.ext.tensorrt.options" attribute in PassContext). This means all BYOC configurations options needed by Collage can be captured uniformly by a list-of-Targets. It also means RPC boundaries (as used internally at OctoML) only need to worry about maintaining the fidelity of the Target instance(s) rather than reaching into the PassContex.
The API is now:
Compilation is switched from function-at-a-time (relying on the TECompiler) to IRModule-at-a-time (using the RelayToTIR target-specific hook mechanism). Though not strictly necessary for Collage I want to check the path is now clear to deprecate the support for BYOC in TEComplier.
Get all the TensorRT tests going again, except for a few I've disabled with x-link to a new issue [BYOC] TensorRT testing issues #11765. CAUTION: The TensorRT runtime is not supported in CI so many of these tests are just cosmetic.
While trying to track down a 'free(): invalid pointer' error in test_tensorrt_int8_exp.py made the TensorRT allocs/frees more robust, but turns out it unrelated issue [Bug] PyTorch and TVM loading problem due to conflicting LLVM symbols #9362 (thanks Masa). No harm leaving these changes in though.