Jupiterghy commented on issue #16870:
URL: https://github.com/apache/tvm/issues/16870#issuecomment-2102059120
> If you call a pass directly (instead of using `Sequential`, it will bypass
the check for `opt_level`, `required_pass`, etc.
Thank you for your response. However, I'm curious
This is an automated email from the ASF dual-hosted git repository.
github-bot pushed a change to branch nightly
in repository https://gitbox.apache.org/repos/asf/tvm.git
from 819b0023e4 [Relax] Support nested ModuleList in nn.Module (#16971)
add 02c4c55eaa [SVE] Add codegen support for
krishnaraj36 commented on PR #16973:
URL: https://github.com/apache/tvm/pull/16973#issuecomment-2101932485
@srkreddy1238 @tqchen : Can you please take a look to this PR. let me know
your advise.
--
This is an automated message from the Apache Git Service.
To respond to the message, please
rickzx commented on PR #16982:
URL: https://github.com/apache/tvm/pull/16982#issuecomment-2101405369
cc: @MasterJH5574
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To u
rickzx opened a new pull request, #16982:
URL: https://github.com/apache/tvm/pull/16982
For workloads with a mixture of symbolic shape and concrete shape as batch
sizes, we cannot directly use `int()` to obtain the batch size. Instead, we can
use `arith.Analyzer` to check equality.
F
lhutton1 opened a new pull request, #16981:
URL: https://github.com/apache/tvm/pull/16981
This commit extends the functionality of the SME dense and matmul schedules
to support operations with fp16 inputs and an fp32 output, where
`transpose_a=False` and `transpose_b=True`.
For conve
masahi commented on PR #16955:
URL: https://github.com/apache/tvm/pull/16955#issuecomment-2101267069
@tvm-bot rerun
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsub
Lunderberg commented on code in PR #16966:
URL: https://github.com/apache/tvm/pull/16966#discussion_r1594285519
##
include/tvm/script/ir_builder/tir/ir.h:
##
@@ -411,8 +411,10 @@ Var EnvThread(String thread_tag, DataType dtype =
DataType::Int(32));
* \param buffer The buffer.
agoston-mc commented on PR #108:
URL: https://github.com/apache/tvm-rfcs/pull/108#issuecomment-2100973263
We have updated the PR with Relax frontend, but we have also kept the Relay,
as an option, thinking it could be useful to have both, because we noticed
performance differences during te
Lunderberg commented on code in PR #16955:
URL: https://github.com/apache/tvm/pull/16955#discussion_r1594281102
##
python/tvm/script/parser/relax/entry.py:
##
@@ -296,13 +298,17 @@ class CallableProxy(StructInfoProxy):
purity : bool
Whether the callable is pure.
Lunderberg commented on code in PR #16955:
URL: https://github.com/apache/tvm/pull/16955#discussion_r1594267879
##
python/tvm/relax/op/memory/view.py:
##
@@ -0,0 +1,77 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements. Se
Lunderberg opened a new pull request, #16980:
URL: https://github.com/apache/tvm/pull/16980
Prior to this commit, the RAII handler in `NDArray` would always attempt to
free a cuda memory allocation on destruction. However, the call to `cudaFree`
may throw an exception. If this happens dur
Lunderberg opened a new pull request, #16978:
URL: https://github.com/apache/tvm/pull/16978
Prior to this commit, while the `num_workers` argument was provided to the
`disco.Session` object, it could not be determined from an existing
`disco.Session` object. As a result, functions that int
Hzfengsy opened a new pull request, #16977:
URL: https://github.com/apache/tvm/pull/16977
As `blockIdx.z` is not allowed in WebGPU, this PR split `blockIdx.z` into
`blockIdx.y` to support WebGPU
--
This is an automated message from the Apache Git Service.
To respond to the message, please
tqchen commented on code in PR #16972:
URL: https://github.com/apache/tvm/pull/16972#discussion_r1593921530
##
src/target/target_kind.cc:
##
@@ -340,9 +340,9 @@ TVM_REGISTER_TARGET_KIND("rocm", kDLROCM)
.set_target_parser(UpdateROCmAttrs);
TVM_REGISTER_TARGET_KIND("openc
krishnaraj36 commented on code in PR #16972:
URL: https://github.com/apache/tvm/pull/16972#discussion_r1593921051
##
src/target/target_kind.cc:
##
@@ -340,9 +340,9 @@ TVM_REGISTER_TARGET_KIND("rocm", kDLROCM)
.set_target_parser(UpdateROCmAttrs);
TVM_REGISTER_TARGET_KIND(
tqchen commented on code in PR #16972:
URL: https://github.com/apache/tvm/pull/16972#discussion_r1593920403
##
src/target/target_kind.cc:
##
@@ -340,9 +340,9 @@ TVM_REGISTER_TARGET_KIND("rocm", kDLROCM)
.set_target_parser(UpdateROCmAttrs);
TVM_REGISTER_TARGET_KIND("openc
tqchen commented on code in PR #16972:
URL: https://github.com/apache/tvm/pull/16972#discussion_r1593918265
##
src/target/target_kind.cc:
##
@@ -340,9 +340,9 @@ TVM_REGISTER_TARGET_KIND("rocm", kDLROCM)
.set_target_parser(UpdateROCmAttrs);
TVM_REGISTER_TARGET_KIND("openc
tqchen commented on code in PR #16976:
URL: https://github.com/apache/tvm/pull/16976#discussion_r1593914420
##
src/target/source/codegen_webgpu.cc:
##
@@ -405,6 +410,19 @@ void CodeGenWebGPU::VisitExpr_(const CallNode* op,
std::ostream& os) { // NOLIN
this->EndScope(els
krishnaraj36 commented on code in PR #16972:
URL: https://github.com/apache/tvm/pull/16972#discussion_r1593911129
##
src/target/target_kind.cc:
##
@@ -340,9 +340,9 @@ TVM_REGISTER_TARGET_KIND("rocm", kDLROCM)
.set_target_parser(UpdateROCmAttrs);
TVM_REGISTER_TARGET_KIND(
tqchen commented on code in PR #16972:
URL: https://github.com/apache/tvm/pull/16972#discussion_r1593903178
##
src/target/target_kind.cc:
##
@@ -340,9 +340,9 @@ TVM_REGISTER_TARGET_KIND("rocm", kDLROCM)
.set_target_parser(UpdateROCmAttrs);
TVM_REGISTER_TARGET_KIND("openc
This is an automated email from the ASF dual-hosted git repository.
masahi pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git
The following commit(s) were added to refs/heads/main by this push:
new c0a47ed139 [CUBLAS][FP8] Enable R.matmul + R.multiply o
masahi merged PR #16974:
URL: https://github.com/apache/tvm/pull/16974
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@tvm.apache.or
lhutton1 merged PR #16962:
URL: https://github.com/apache/tvm/pull/16962
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@tvm.apache.
This is an automated email from the ASF dual-hosted git repository.
lukhut pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git
The following commit(s) were added to refs/heads/main by this push:
new 02c4c55eaa [SVE] Add codegen support for `vscale_range(
lhutton1 commented on PR #16962:
URL: https://github.com/apache/tvm/pull/16962#issuecomment-2100060824
Thanks @Anndrey24!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To
Jiawei-Shao opened a new pull request, #16976:
URL: https://github.com/apache/tvm/pull/16976
This patch adds the support of `__dp4a(int8x4, int8x4)` as a pure extern
method of WebGPU target. In the generated WGSL shader, `int8x4` will be
translated into `u32`, and `__dp4a(int8x4, int8x4)` w
27 matches
Mail list logo