qunluo opened a new pull request #6123:
URL: https://github.com/apache/incubator-tvm/pull/6123
add attr opthion mfloat-abi for arm32
referenceļ¼
https://github.com/apache/incubator-tvm/issues/6019
https://discuss.tvm.ai/t/build-with-arm-toolchain-32bit/92
icemelon9 opened a new pull request #6122:
URL: https://github.com/apache/incubator-tvm/pull/6122
Thanks for contributing to TVM! Please refer to guideline
https://tvm.apache.org/docs/contribute/ for useful information and tips. After
the pull request is submitted, please request code
anijain2305 commented on pull request #6115:
URL: https://github.com/apache/incubator-tvm/pull/6115#issuecomment-662819127
@eric-haibin-lin Yes, the MXNet+MKLDNN baseline is also in `int8`.
This is an automated message from
anijain2305 commented on pull request #6084:
URL: https://github.com/apache/incubator-tvm/pull/6084#issuecomment-662818914
@tqchen This is ready to go in!
This is an automated message from the Apache Git Service.
To respond
jroesch commented on pull request #6105:
URL: https://github.com/apache/incubator-tvm/pull/6105#issuecomment-662816593
@icemelon9 it looks like something is wrong with CI can you try
retriggering?
This is an automated
jroesch commented on a change in pull request #6105:
URL: https://github.com/apache/incubator-tvm/pull/6105#discussion_r459212277
##
File path: python/tvm/runtime/vm.py
##
@@ -273,29 +272,58 @@ def get_function_params(self, func_name):
class VirtualMachine(object):
-
hatsunexym closed issue #6120:
URL: https://github.com/apache/incubator-tvm/issues/6120
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
roastduck commented on a change in pull request #6066:
URL: https://github.com/apache/incubator-tvm/pull/6066#discussion_r459188187
##
File path: src/tir/transforms/hoist_if_then_else.cc
##
@@ -0,0 +1,376 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+
lsy643 commented on a change in pull request #5955:
URL: https://github.com/apache/incubator-tvm/pull/5955#discussion_r459186976
##
File path: tests/python/relay/test_any.py
##
@@ -816,6 +816,66 @@ def test_mixed_input_type():
assert result.asnumpy().shape ==
lsy643 commented on a change in pull request #5955:
URL: https://github.com/apache/incubator-tvm/pull/5955#discussion_r459186931
##
File path: tests/python/relay/test_any.py
##
@@ -816,6 +816,66 @@ def test_mixed_input_type():
assert result.asnumpy().shape ==
GaryYuyjl opened a new pull request #6121:
URL: https://github.com/apache/incubator-tvm/pull/6121
This pr supports int4/int8 tensor core with HWNC layout.
Part of the layer-wise results of HWNC and NHWC are here. HWNC layout runs
faster in most of the workloads.
workload
junrushao1994 commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r459183926
##
File path: src/auto_scheduler/compute_dag.cc
##
@@ -126,6 +555,7 @@ class FlopEstimator : public ExprFunctor {
fail_ = true;
hatsunexym opened a new issue #6120:
URL: https://github.com/apache/incubator-tvm/issues/6120
I successfully used assert command by Python to make sure the **tvm_output**
is similar to the **torch_output** with 1e-3 when using TVM to do model
transform.
Then I deploy the **param, graph
junrushao1994 commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r459182796
##
File path: include/tvm/auto_scheduler/compute_dag.h
##
@@ -0,0 +1,237 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under
jcf94 commented on a change in pull request #6107:
URL: https://github.com/apache/incubator-tvm/pull/6107#discussion_r459180895
##
File path: src/auto_scheduler/transform_step.cc
##
@@ -923,5 +958,275 @@ String
ComputeRootStepNode::PrintAsPythonAPI(Array* stages,
return
lsy643 commented on a change in pull request #6108:
URL: https://github.com/apache/incubator-tvm/pull/6108#discussion_r459178559
##
File path: topi/python/topi/cuda/nms.py
##
@@ -93,44 +93,41 @@ def get_valid_counts_ir(data, valid_count, out, out_indices,
valid_count =
lsy643 commented on a change in pull request #6108:
URL: https://github.com/apache/incubator-tvm/pull/6108#discussion_r459178638
##
File path: tests/python/relay/test_op_level5.py
##
@@ -270,9 +270,9 @@ def verify_get_valid_counts(dshape, score_threshold,
id_index,
anijain2305 opened a new pull request #6119:
URL: https://github.com/apache/incubator-tvm/pull/6119
After going through multiple frameworks, I found that it is common to ignore
-128 for int8 quantization.
TFLite - from Pete Warden's blog and TFLite paper -
jcf94 commented on a change in pull request #6107:
URL: https://github.com/apache/incubator-tvm/pull/6107#discussion_r459166066
##
File path: src/auto_scheduler/transform_step.cc
##
@@ -923,5 +958,275 @@ String
ComputeRootStepNode::PrintAsPythonAPI(Array* stages,
return
gussmith23 commented on a change in pull request #5812:
URL: https://github.com/apache/incubator-tvm/pull/5812#discussion_r459160110
##
File path: tests/python/unittest/test_target_custom_datatypes.py
##
@@ -85,6 +85,7 @@ def test_bfloat_add_and_cast_1():
built_cast(x,
merrymercy commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r459143595
##
File path: src/auto_scheduler/compute_dag.cc
##
@@ -126,6 +555,7 @@ class FlopEstimator : public ExprFunctor {
fail_ = true;
merrymercy commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r459143595
##
File path: src/auto_scheduler/compute_dag.cc
##
@@ -126,6 +555,7 @@ class FlopEstimator : public ExprFunctor {
fail_ = true;
trevor-m commented on a change in pull request #6108:
URL: https://github.com/apache/incubator-tvm/pull/6108#discussion_r459138745
##
File path: tests/python/relay/test_op_level5.py
##
@@ -270,9 +270,9 @@ def verify_get_valid_counts(dshape, score_threshold,
id_index,
trevor-m commented on a change in pull request #6108:
URL: https://github.com/apache/incubator-tvm/pull/6108#discussion_r459138745
##
File path: tests/python/relay/test_op_level5.py
##
@@ -270,9 +270,9 @@ def verify_get_valid_counts(dshape, score_threshold,
id_index,
binarybana commented on pull request #6116:
URL: https://github.com/apache/incubator-tvm/pull/6116#issuecomment-662743251
Okay, builds and tests pass now.
This is an automated message from the Apache Git Service.
To respond
binarybana commented on a change in pull request #6116:
URL: https://github.com/apache/incubator-tvm/pull/6116#discussion_r459134821
##
File path: rust/tvm-sys/Cargo.toml
##
@@ -32,4 +32,4 @@ ndarray = "0.12"
enumn = "^0.1"
[build-dependencies]
-bindgen = "0.51"
+bindgen =
merrymercy commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r459132211
##
File path: include/tvm/auto_scheduler/compute_dag.h
##
@@ -0,0 +1,237 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+
merrymercy commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r459132211
##
File path: include/tvm/auto_scheduler/compute_dag.h
##
@@ -0,0 +1,237 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+
merrymercy commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r459132211
##
File path: include/tvm/auto_scheduler/compute_dag.h
##
@@ -0,0 +1,237 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+
merrymercy commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r459132211
##
File path: include/tvm/auto_scheduler/compute_dag.h
##
@@ -0,0 +1,237 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+
merrymercy commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r459132211
##
File path: include/tvm/auto_scheduler/compute_dag.h
##
@@ -0,0 +1,237 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+
giuseros commented on pull request #6117:
URL: https://github.com/apache/incubator-tvm/pull/6117#issuecomment-662737470
cc @FrozenGene @anijain2305 @u99127
Please note that since I will be off from Friday (for 15 days), I might turn
this into a draft and pick it up when I come back
alexbooth opened a new pull request #6118:
URL: https://github.com/apache/incubator-tvm/pull/6118
I have an app level use case that this information is useful for.
Implementation mirrors the existing 'get_num_outputs' API.
giuseros commented on pull request #6110:
URL: https://github.com/apache/incubator-tvm/pull/6110#issuecomment-662736167
Hi @anijain2305 ,
Thanks for the approval! I did as you said and re-triggered the CI
This is an
giuseros opened a new pull request #6117:
URL: https://github.com/apache/incubator-tvm/pull/6117
## High level description of this contribution
The following tuning entities have been introduced:
- Unrolling and vectorizing input matrix transform
- Reordering gemm to exploit
junrushao1994 commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r459046943
##
File path: src/auto_scheduler/compute_dag.cc
##
@@ -114,7 +118,432 @@ Array TopoSortOps(const Array&
tensors) {
return ops;
}
-//
merrymercy edited a comment on pull request #6040:
URL: https://github.com/apache/incubator-tvm/pull/6040#issuecomment-662727589
I close this PR because after more testing I found the current python
version in the master works well. This pr is trying to fix the c++ version,
which is only
merrymercy edited a comment on pull request #6040:
URL: https://github.com/apache/incubator-tvm/pull/6040#issuecomment-662727589
I close this PR because after more testing I found the current python
version in the master works well. This pr is trying to fix the c++ version,
which is only
merrymercy edited a comment on pull request #6040:
URL: https://github.com/apache/incubator-tvm/pull/6040#issuecomment-662727589
I close this PR because after more testing I found the current python
version in the master works well. This pr is trying to fix the c++ version,
which is only
merrymercy edited a comment on pull request #6040:
URL: https://github.com/apache/incubator-tvm/pull/6040#issuecomment-662727589
I close this PR because after more testing I found the current python
version in the master works well. This pr is trying to fix the c++ version,
which is only
merrymercy edited a comment on pull request #6040:
URL: https://github.com/apache/incubator-tvm/pull/6040#issuecomment-662727589
I close this PR because after more testing I found the current python
version in the master works well. This pr is trying to fix the c++ version,
which is only
merrymercy edited a comment on pull request #6040:
URL: https://github.com/apache/incubator-tvm/pull/6040#issuecomment-662727589
This is an automated message from the Apache Git Service.
To respond to the message, please log
merrymercy commented on pull request #6040:
URL: https://github.com/apache/incubator-tvm/pull/6040#issuecomment-662727589
I close this PR because after more test I found the current python version
in tvm master works well. This pr is trying to fix the c++ version, which is
only used in
junrushao1994 commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r458574335
##
File path: src/auto_scheduler/compute_dag.cc
##
@@ -114,7 +118,432 @@ Array TopoSortOps(const Array&
tensors) {
return ops;
}
-//
binarybana commented on pull request #6116:
URL: https://github.com/apache/incubator-tvm/pull/6116#issuecomment-662712240
I noticed a small build error I introduced in the last rebase. Will fix and
push when I get a second, but wanted to flag it first.
jroesch commented on pull request #6116:
URL: https://github.com/apache/incubator-tvm/pull/6116#issuecomment-662693206
LGTM, @mwillsey can you review as well?
This is an automated message from the Apache Git Service.
To
jroesch commented on a change in pull request #6116:
URL: https://github.com/apache/incubator-tvm/pull/6116#discussion_r459078752
##
File path: rust/tvm-sys/Cargo.toml
##
@@ -32,4 +32,4 @@ ndarray = "0.12"
enumn = "^0.1"
[build-dependencies]
-bindgen = "0.51"
+bindgen = {
binarybana opened a new pull request #6116:
URL: https://github.com/apache/incubator-tvm/pull/6116
* Turn off default features for bindgen
* Upgrade some deps for smaller total dep tree
* Switch (/complete switch) to thiserror
* Remove unnecessary transmutes
Thanks for
anijain2305 commented on pull request #6110:
URL: https://github.com/apache/incubator-tvm/pull/6110#issuecomment-662674836
Can you please retrigger the CI. It is failing due to a known issue -
https://github.com/apache/incubator-tvm/pull/6084
anijain2305 opened a new pull request #6115:
URL: https://github.com/apache/incubator-tvm/pull/6115
Using MKL for quantized dense, following the MKL fallback for FP32 dense.
On C5.12x large cascade lake with VNNI support, results for BERT base are as
follows (latency in ms)
junrushao1994 commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r459036902
##
File path: src/auto_scheduler/compute_dag.cc
##
@@ -114,7 +118,432 @@ Array TopoSortOps(const Array&
tensors) {
return ops;
}
-//
comaniac commented on a change in pull request #6109:
URL: https://github.com/apache/incubator-tvm/pull/6109#discussion_r459011905
##
File path: src/relay/backend/contrib/arm_compute_lib/codegen.cc
##
@@ -49,6 +49,18 @@ class ACLJSONSerializer : public
icemelon9 commented on a change in pull request #5955:
URL: https://github.com/apache/incubator-tvm/pull/5955#discussion_r459019580
##
File path: tests/python/relay/test_any.py
##
@@ -816,6 +816,66 @@ def test_mixed_input_type():
assert result.asnumpy().shape ==
mwillsey opened a new pull request #6114:
URL: https://github.com/apache/incubator-tvm/pull/6114
This PR cleans up conversions between TVM's `Function` and Rust functions.
The `ToBoxedFn` trait is removed entirely, replaced by standard `From`
implementations.
The `ToFunction` and
comaniac commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r458995260
##
File path: include/tvm/auto_scheduler/compute_dag.h
##
@@ -0,0 +1,242 @@
+/*r
+ * Licensed to the Apache Software Foundation (ASF) under one
+
csullivan opened a new pull request #6113:
URL: https://github.com/apache/incubator-tvm/pull/6113
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL
junrushao1994 commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r458573185
##
File path: src/auto_scheduler/compute_dag.cc
##
@@ -114,7 +118,432 @@ Array TopoSortOps(const Array&
tensors) {
return ops;
}
-//
comaniac commented on a change in pull request #6107:
URL: https://github.com/apache/incubator-tvm/pull/6107#discussion_r458955101
##
File path: src/auto_scheduler/transform_step.cc
##
@@ -923,5 +958,275 @@ String
ComputeRootStepNode::PrintAsPythonAPI(Array* stages,
return
tmoreau89 commented on issue #6090:
URL: https://github.com/apache/incubator-tvm/issues/6090#issuecomment-662585519
Having a small script to reproduce the bug would be helpful, thanks!
This is an automated message from the
icemelon9 commented on a change in pull request #6105:
URL: https://github.com/apache/incubator-tvm/pull/6105#discussion_r458961935
##
File path: python/tvm/runtime/vm.py
##
@@ -273,29 +272,58 @@ def get_function_params(self, func_name):
class VirtualMachine(object):
-
icemelon9 commented on a change in pull request #6105:
URL: https://github.com/apache/incubator-tvm/pull/6105#discussion_r458957518
##
File path: python/tvm/runtime/vm.py
##
@@ -273,29 +272,58 @@ def get_function_params(self, func_name):
class VirtualMachine(object):
-
comaniac commented on a change in pull request #6107:
URL: https://github.com/apache/incubator-tvm/pull/6107#discussion_r458946554
##
File path: python/tvm/auto_scheduler/loop_state.py
##
@@ -371,6 +433,23 @@ def _update_stage_id_map(self):
for index, stage in
ANSHUMAN87 commented on pull request #6066:
URL: https://github.com/apache/incubator-tvm/pull/6066#issuecomment-662572200
All the CI issues are resolved now. Also my internal tests shows good
result. I think we can start review now. TIA!
comaniac commented on a change in pull request #6107:
URL: https://github.com/apache/incubator-tvm/pull/6107#discussion_r458944675
##
File path: python/tvm/auto_scheduler/loop_state.py
##
@@ -371,6 +433,23 @@ def _update_stage_id_map(self):
for index, stage in
anijain2305 edited a comment on issue #6111:
URL: https://github.com/apache/incubator-tvm/issues/6111#issuecomment-662568366
Thanks @tqchen for diligently adding issues. For this one, I looked at the
logs and this is the duplicate of
https://github.com/apache/incubator-tvm/issues/6064
anijain2305 commented on issue #6111:
URL: https://github.com/apache/incubator-tvm/issues/6111#issuecomment-662568366
Thanks @tqchen for diligently adding issues. For this one, I looked at the
logs and this is the duplicate of
https://github.com/apache/incubator-tvm/issues/6064
The
tqchen commented on a change in pull request #6112:
URL: https://github.com/apache/incubator-tvm/pull/6112#discussion_r458934696
##
File path: Jenkinsfile
##
@@ -167,6 +167,7 @@ stage('Build') {
sh "${docker_run} ${ci_cpu} ./tests/scripts/task_python_vta_tsim.sh"
leandron commented on pull request #6112:
URL: https://github.com/apache/incubator-tvm/pull/6112#issuecomment-662561133
If possible, I'll be presenting a live demo today on the "Online Compiler
Chat".
This is an automated
tqchen commented on issue #6093:
URL: https://github.com/apache/incubator-tvm/issues/6093#issuecomment-662561068
Sounds good, feel free to send a PR
This is an automated message from the Apache Git Service.
To respond to the
zhiics commented on a change in pull request #6105:
URL: https://github.com/apache/incubator-tvm/pull/6105#discussion_r458925888
##
File path: python/tvm/runtime/vm.py
##
@@ -273,29 +272,58 @@ def get_function_params(self, func_name):
class VirtualMachine(object):
-
hanzz2007 commented on issue #6093:
URL: https://github.com/apache/incubator-tvm/issues/6093#issuecomment-662554484
Yes, move can solve. And we should add a default constructor and a rvref
constructor to AttrInitEntry
```c++
template
struct AttrInitEntry {
// The attributes
leandron opened a new pull request #6112:
URL: https://github.com/apache/incubator-tvm/pull/6112
TVMC - a command line driver for TVM
* Introduce a command line driver to compile, run and tune models, using
TVM graph runtime
* tvmc tests and integrate tvmc with regular linting,
zhiics commented on pull request #5235:
URL: https://github.com/apache/incubator-tvm/pull/5235#issuecomment-662548353
@jwfromm please take another look and approve explicitly if everything looks
good to you.
This is an
tqchen edited a comment on pull request #6107:
URL: https://github.com/apache/incubator-tvm/pull/6107#issuecomment-662542322
cc @jroesch @jwfromm @junrushao1994
This is an automated message from the Apache Git Service.
To
tqchen commented on pull request #6107:
URL: https://github.com/apache/incubator-tvm/pull/6107#issuecomment-662542322
cc @jroesch @jwfromm
This is an automated message from the Apache Git Service.
To respond to the message,
giuseros commented on a change in pull request #6095:
URL: https://github.com/apache/incubator-tvm/pull/6095#discussion_r458905799
##
File path: topi/python/topi/arm_cpu/depthwise_conv2d.py
##
@@ -181,6 +181,154 @@ def depthwise_conv2d_nchw_spatial_pack(cfg, data, kernel,
tqchen commented on pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#issuecomment-662541150
cc @jwfromm @mbrookhart @jroesch
This is an automated message from the Apache Git Service.
To respond to
tqchen commented on issue #6111:
URL: https://github.com/apache/incubator-tvm/issues/6111#issuecomment-662540683
cc @anijain2305
This is an automated message from the Apache Git Service.
To respond to the message, please
tqchen opened a new issue #6111:
URL: https://github.com/apache/incubator-tvm/issues/6111
https://ci.tvm.ai/job/tvm/job/master/1322/execution/node/354/log/
This is an automated message from the Apache Git Service.
To
giuseros commented on a change in pull request #6095:
URL: https://github.com/apache/incubator-tvm/pull/6095#discussion_r458893850
##
File path: topi/python/topi/arm_cpu/depthwise_conv2d.py
##
@@ -181,6 +181,154 @@ def depthwise_conv2d_nchw_spatial_pack(cfg, data, kernel,
FrozenGene commented on a change in pull request #6095:
URL: https://github.com/apache/incubator-tvm/pull/6095#discussion_r458874360
##
File path: topi/python/topi/arm_cpu/depthwise_conv2d.py
##
@@ -181,6 +181,154 @@ def depthwise_conv2d_nchw_spatial_pack(cfg, data, kernel,
giuseros opened a new pull request #6110:
URL: https://github.com/apache/incubator-tvm/pull/6110
This improves reduction by ~10x using x86 schedules. While for `float32`
reduction is not common, it is widely used in quantized operations (`int8`,
`uint8`) in order to compute the offset
giuseros commented on a change in pull request #6095:
URL: https://github.com/apache/incubator-tvm/pull/6095#discussion_r458842593
##
File path: topi/python/topi/arm_cpu/depthwise_conv2d.py
##
@@ -181,6 +181,154 @@ def depthwise_conv2d_nchw_spatial_pack(cfg, data, kernel,
FrozenGene commented on a change in pull request #6095:
URL: https://github.com/apache/incubator-tvm/pull/6095#discussion_r458834973
##
File path: topi/python/topi/arm_cpu/depthwise_conv2d.py
##
@@ -181,6 +181,154 @@ def depthwise_conv2d_nchw_spatial_pack(cfg, data, kernel,
FrozenGene commented on a change in pull request #6095:
URL: https://github.com/apache/incubator-tvm/pull/6095#discussion_r458834973
##
File path: topi/python/topi/arm_cpu/depthwise_conv2d.py
##
@@ -181,6 +181,154 @@ def depthwise_conv2d_nchw_spatial_pack(cfg, data, kernel,
giuseros commented on a change in pull request #6095:
URL: https://github.com/apache/incubator-tvm/pull/6095#discussion_r458814745
##
File path: topi/python/topi/arm_cpu/depthwise_conv2d.py
##
@@ -181,6 +181,154 @@ def depthwise_conv2d_nchw_spatial_pack(cfg, data, kernel,
giuseros commented on a change in pull request #6095:
URL: https://github.com/apache/incubator-tvm/pull/6095#discussion_r458814745
##
File path: topi/python/topi/arm_cpu/depthwise_conv2d.py
##
@@ -181,6 +181,154 @@ def depthwise_conv2d_nchw_spatial_pack(cfg, data, kernel,
giuseros commented on a change in pull request #6095:
URL: https://github.com/apache/incubator-tvm/pull/6095#discussion_r458814745
##
File path: topi/python/topi/arm_cpu/depthwise_conv2d.py
##
@@ -181,6 +181,154 @@ def depthwise_conv2d_nchw_spatial_pack(cfg, data, kernel,
giuseros commented on a change in pull request #6095:
URL: https://github.com/apache/incubator-tvm/pull/6095#discussion_r458814745
##
File path: topi/python/topi/arm_cpu/depthwise_conv2d.py
##
@@ -181,6 +181,154 @@ def depthwise_conv2d_nchw_spatial_pack(cfg, data, kernel,
lhutton1 opened a new pull request #6109:
URL: https://github.com/apache/incubator-tvm/pull/6109
Adds support for asymmetric per-layer quantization in the Arm Compute
Library runtime module. This includes support for qnn.conv2d, nn.maxpool2d and
reshape.
* Adds support for
lsy643 commented on pull request #6033:
URL: https://github.com/apache/incubator-tvm/pull/6033#issuecomment-662307838
cc @Laurawly @Huyuwei
This is an automated message from the Apache Git Service.
To respond to the
merrymercy commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r458605272
##
File path: include/tvm/auto_scheduler/auto_schedule.h
##
@@ -18,29 +18,27 @@
*/
/*!
- * \file auto_scheduler/auto_schedule.h
Review
merrymercy closed pull request #6040:
URL: https://github.com/apache/incubator-tvm/pull/6040
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
jcf94 commented on a change in pull request #6107:
URL: https://github.com/apache/incubator-tvm/pull/6107#discussion_r458576077
##
File path: python/tvm/auto_scheduler/loop_state.py
##
@@ -371,6 +433,23 @@ def _update_stage_id_map(self):
for index, stage in
lsy643 commented on pull request #6108:
URL: https://github.com/apache/incubator-tvm/pull/6108#issuecomment-662279185
cc @yongwww @Laurawly
This is an automated message from the Apache Git Service.
To respond to the
lsy643 opened a new pull request #6108:
URL: https://github.com/apache/incubator-tvm/pull/6108
In this PR, the CUDA compute funtions of `get_valid_counts` and `nms` are
changed to make them work as expected.
1. For `get_valid_counts`, only one thread is used for one image. I am not
jcf94 commented on a change in pull request #6107:
URL: https://github.com/apache/incubator-tvm/pull/6107#discussion_r458569182
##
File path: python/tvm/auto_scheduler/loop_state.py
##
@@ -351,6 +351,68 @@ def compute_root(self, stage):
self.state_object =
jcf94 commented on a change in pull request #6107:
URL: https://github.com/apache/incubator-tvm/pull/6107#discussion_r458568998
##
File path: python/tvm/auto_scheduler/loop_state.py
##
@@ -351,6 +351,68 @@ def compute_root(self, stage):
self.state_object =
junrushao1994 commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r458567027
##
File path: include/tvm/auto_scheduler/compute_dag.h
##
@@ -0,0 +1,242 @@
+/*r
+ * Licensed to the Apache Software Foundation (ASF) under
junrushao1994 commented on a change in pull request #6103:
URL: https://github.com/apache/incubator-tvm/pull/6103#discussion_r458504576
##
File path: include/tvm/auto_scheduler/loop_state.h
##
@@ -159,10 +159,16 @@ using IterKey = std::pair;
*/
class AttachMapNode : public
1 - 100 of 103 matches
Mail list logo
