[GitHub] [tvm] tqchen edited a comment on issue #9022: [Bug] BuiltinLower does not use alloca for storage on kDLCPU target devices

tqchen edited a comment on issue #9022: URL: https://github.com/apache/tvm/issues/9022#issuecomment-920896447 So in the above post I tried to summarize the state. Now let me try to share some of my thoughts based on the summary. First of all, R0 and R1 are not that different in

[GitHub] [tvm] tqchen commented on issue #9022: [Bug] BuiltinLower does not use alloca for storage on kDLCPU target devices

tqchen commented on issue #9022: URL: https://github.com/apache/tvm/issues/9022#issuecomment-920896447 So in the above post I tried to summarize the state. Now let me try to share some of my thoughts based on the summary. First of all, R0 and R1 are not that different in nature.

[GitHub] [tvm-rfcs] manupa-arm commented on pull request #22: [RFC][TIR] TIR Non-scalar Constants

manupa-arm commented on pull request #22: URL: https://github.com/apache/tvm-rfcs/pull/22#issuecomment-920894021 @junrushao1994 has followed up with individual comments and its pending my response :) I ll do it next once I get some free cycles. -- This is an automated message from

[GitHub] [tvm] ashutosh-arm commented on a change in pull request #8951: [3/10] Moved TIR generation from Python to C++ for CMSIS-NN

ashutosh-arm commented on a change in pull request #8951: URL: https://github.com/apache/tvm/pull/8951#discussion_r710106598 ## File path: src/relay/backend/contrib/cmsisnn/codegen_cmsisnn.cc ## @@ -16,190 +16,36 @@ * specific language governing permissions and limitations

[GitHub] [tvm] Mousius commented on a change in pull request #8951: [3/10] Moved TIR generation from Python to C++ for CMSIS-NN

Mousius commented on a change in pull request #8951: URL: https://github.com/apache/tvm/pull/8951#discussion_r710096209 ## File path: src/relay/backend/contrib/cmsisnn/codegen_cmsisnn.cc ## @@ -16,190 +16,36 @@ * specific language governing permissions and limitations *

[GitHub] [tvm] tqchen edited a comment on issue #9022: [Bug] BuiltinLower does not use alloca for storage on kDLCPU target devices

tqchen edited a comment on issue #9022: URL: https://github.com/apache/tvm/issues/9022#issuecomment-920882592 Thanks for the discussions. Before we suggest a resolution, it would be helpful to summarize the discussions so far. ## Semantics of Allocate and storage_scope

[GitHub] [tvm] tqchen edited a comment on issue #9022: [Bug] BuiltinLower does not use alloca for storage on kDLCPU target devices

tqchen edited a comment on issue #9022: URL: https://github.com/apache/tvm/issues/9022#issuecomment-920882592 Thanks for the discussions. Before we suggest a resolution, it would be helpful to summarize the discussions so far. ## Semantics of Allocate and storage_scope

[GitHub] [tvm] tqchen edited a comment on issue #9022: [Bug] BuiltinLower does not use alloca for storage on kDLCPU target devices

tqchen edited a comment on issue #9022: URL: https://github.com/apache/tvm/issues/9022#issuecomment-920882592 Thanks for the discussions. Before we suggest a resolution, it would be helpful to summarize the discussions so far. ## Semantics of Allocate and storage_scope

[GitHub] [tvm] tqchen commented on issue #9022: [Bug] BuiltinLower does not use alloca for storage on kDLCPU target devices

tqchen commented on issue #9022: URL: https://github.com/apache/tvm/issues/9022#issuecomment-920882592 Thanks for the discussions. Before we suggest a resolution, it would be helpful to summarize the discussions so far. # Semantics of Allocate and storage_scope Allocate

[GitHub] [tvm] manupa-arm commented on issue #9022: [Bug] BuiltinLower does not use alloca for storage on kDLCPU target devices

manupa-arm commented on issue #9022: URL: https://github.com/apache/tvm/issues/9022#issuecomment-920880447 I feel that is a workaround for an optimization that certain set of CPUs require it. Moreover, TVMBAWs allows an abstraction to use an Arena that could be used in any memory

[GitHub] [tvm-rfcs] manupa-arm commented on pull request #23: [RFC][TIR] TIR Pinned Memory Representation

manupa-arm commented on pull request #23: URL: https://github.com/apache/tvm-rfcs/pull/23#issuecomment-920873541 @tqchen, can you take a look? I can modify the USMP RFC to reflect the decisions made here, if we could finalize the discussion here. -- This is an automated message

[GitHub] [tvm] mikepapadim opened a new pull request #9031: [CI] Prevent the complete Jenkins pipeline to run when files commited only to `/docs`

mikepapadim opened a new pull request #9031: URL: https://github.com/apache/tvm/pull/9031 This is an updated version for #8992 and fixes #8673. With this PR when changes affect files in `docs/` then only the following stages will run: * Sanity Check * BUILD: GPU * docs:

[GitHub] [tvm] arangasa commented on issue #8838: [Tracking Issue][ONNX] Quantized operator support in ONNX importer

arangasa commented on issue #8838: URL: https://github.com/apache/tvm/issues/8838#issuecomment-920865864 Hi @tmoreau89, I've placed a PR [#9028](https://github.com/apache/tvm/pull/9028) for QLinearSigmoid. Just FYI. -- This is an automated message from the Apache Git Service. To

[GitHub] [tvm] mikepapadim closed pull request #8992: Prevent non-docs stages when a PR only modifies docs/

mikepapadim closed pull request #8992: URL: https://github.com/apache/tvm/pull/8992 -- 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:

[GitHub] [tvm] tqchen edited a comment on issue #9022: [Bug] BuiltinLower does not use alloca for storage on kDLCPU target devices

tqchen edited a comment on issue #9022: URL: https://github.com/apache/tvm/issues/9022#issuecomment-920841982 @manupa-arm in cpu we do not necessarily differentiate local from global for now as they are from the same namespace. I can understand the need from the micro side, and I

[GitHub] [tvm] tqchen commented on issue #9022: [Bug] BuiltinLower does not use alloca for storage on kDLCPU target devices

tqchen commented on issue #9022: URL: https://github.com/apache/tvm/issues/9022#issuecomment-920841982 @manupa-arm in cpu we do not necessarily differentiate local from global for now as they are from the same namespace. I can understand the need from the micro side, and I believe

[GitHub] [tvm] quic-sanirudh commented on a change in pull request #9017: [ONNX] QLinearAveragePool and QLinearGlobalAveragePool contrib op

quic-sanirudh commented on a change in pull request #9017: URL: https://github.com/apache/tvm/pull/9017#discussion_r710050671 ## File path: tests/python/frontend/onnx/test_forward.py ## @@ -3056,6 +3056,152 @@ def verify_global_pooling(x_shape, mode):

[GitHub] [tvm] quic-sanirudh commented on a change in pull request #9017: [ONNX] QLinearAveragePool and QLinearGlobalAveragePool contrib op

quic-sanirudh commented on a change in pull request #9017: URL: https://github.com/apache/tvm/pull/9017#discussion_r710050405 ## File path: tests/python/frontend/onnx/test_forward.py ## @@ -3056,6 +3056,152 @@ def verify_global_pooling(x_shape, mode):

[GitHub] [tvm] quic-sanirudh commented on a change in pull request #9017: [ONNX] QLinearAveragePool and QLinearGlobalAveragePool contrib op

quic-sanirudh commented on a change in pull request #9017: URL: https://github.com/apache/tvm/pull/9017#discussion_r710043945 ## File path: python/tvm/relay/frontend/onnx.py ## @@ -351,6 +366,13 @@ class AveragePool(Pool): name = "avg_pool" +class

[GitHub] [tvm] quic-sanirudh commented on a change in pull request #9017: [ONNX] QLinearAveragePool and QLinearGlobalAveragePool contrib op

quic-sanirudh commented on a change in pull request #9017: URL: https://github.com/apache/tvm/pull/9017#discussion_r710043241 ## File path: python/tvm/relay/frontend/onnx.py ## @@ -654,6 +676,40 @@ def _impl_v1(cls, inputs, attr, params): ) +class

[GitHub] [tvm] quic-sanirudh commented on a change in pull request #9017: [ONNX] QLinearAveragePool and QLinearGlobalAveragePool contrib op

quic-sanirudh commented on a change in pull request #9017: URL: https://github.com/apache/tvm/pull/9017#discussion_r710042850 ## File path: python/tvm/relay/frontend/onnx.py ## @@ -3794,12 +3850,14 @@ def _get_convert_map(opset): "Xor": Renamer("logical_xor"),

[GitHub] [tvm] cheimu commented on pull request #8447: [Frontend, pytorch] Vc/pytorch lstm

cheimu commented on pull request #8447: URL: https://github.com/apache/tvm/pull/8447#issuecomment-920759037 Hi there. I found this error still exist. I just clone this repo yesterday

[GitHub] [tvm] leandron opened a new pull request #9029: Fix line break in `setup.py`

leandron opened a new pull request #9029: URL: https://github.com/apache/tvm/pull/9029 Fix one issue caused by #9005 in the packaging system. cc @tqchen @manupa-arm @areusch for reviews -- This is an automated message from the Apache Git Service. To respond to the message, please

[GitHub] [tvm] FrozenGene commented on pull request #6023: [CI][Caffe Frontend] add caffe environment

FrozenGene commented on pull request #6023: URL: https://github.com/apache/tvm/pull/6023#issuecomment-920740949 > Since BVLC distribution support has successfully completed, it would be great if we could replace BVLC distribution with SSD distribution. > > Anyone working on that?

[GitHub] [tvm] masahi closed issue #8964: [CI Problem][FLAKY] Possible flaky test: test_onnx_nodes[llvm-test_nllloss_NCd1d2d3_sum_weight_high_ii]

masahi closed issue #8964: URL: https://github.com/apache/tvm/issues/8964 -- 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:

[GitHub] [tvm] masahi closed issue #8918: [FLAKY] test_onnx_nodes[llvm-test_nllloss_NCd1d2d3_none_no_weight_negative_ii]

masahi closed issue #8918: URL: https://github.com/apache/tvm/issues/8918 -- 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:

[tvm] branch main updated (6dc54aa -> 02fbaf0)

This is an automated email from the ASF dual-hosted git repository. masahi pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git. from 6dc54aa [Bugfix] Fix other div zero errors also in rewrite_simplify (#8983) add 02fbaf0 [Onnx] Fix NLL Loss tests

[GitHub] [tvm] masahi merged pull request #8971: [Onnx] Fix NLL Loss tests

masahi merged pull request #8971: URL: https://github.com/apache/tvm/pull/8971 -- 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:

[GitHub] [tvm] mshr-h commented on pull request #6023: [CI][Caffe Frontend] add caffe environment

mshr-h commented on pull request #6023: URL: https://github.com/apache/tvm/pull/6023#issuecomment-920705954 Since BVLC distribution support has successfully completed, it would be great if we could replace BVLC distribution with SSD distribution. Anyone working on that? If no,

[GitHub] [tvm-rfcs] u99127 commented on pull request #22: [RFC][TIR] TIR Non-scalar Constants

u99127 commented on pull request #22: URL: https://github.com/apache/tvm-rfcs/pull/22#issuecomment-920704274 How can we move this forward - this appears to be getting stalled ? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to

[GitHub] [tvm] Alex-grovety commented on issue #8717: Fusion of operations and cast in mobilenet v1 conv2d causing large feature maps

Alex-grovety commented on issue #8717: URL: https://github.com/apache/tvm/issues/8717#issuecomment-920703220 To get this result: ``` PrimFunc([placeholder, placeholder, placeholder, T_cast]) attrs={"from_legacy_te_schedule": (bool)1, "global_symbol":

[tvm] branch main updated (2711229 -> 6dc54aa)

This is an automated email from the ASF dual-hosted git repository. jroesch pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git. from 2711229 [ONNX] enable the onnx tests after PR #8274 merged (#9019) add 6dc54aa [Bugfix] Fix other div zero errors

[GitHub] [tvm] jroesch merged pull request #8983: [Bugfix] Fix other div zero errors also in rewrite_simplify

jroesch merged pull request #8983: URL: https://github.com/apache/tvm/pull/8983 -- 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:

[GitHub] [tvm] jroesch commented on pull request #8971: [Onnx] Fix NLL Loss tests

jroesch commented on pull request #8971: URL: https://github.com/apache/tvm/pull/8971#issuecomment-920701813 @masahi can you land this one if you are OK? -- 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

[GitHub] [tvm-rfcs] u99127 commented on pull request #30: Command Line Configuration Files

u99127 commented on pull request #30: URL: https://github.com/apache/tvm-rfcs/pull/30#issuecomment-920701756 > Can you address the common usage of environment variables for configuration, why that wasn't considered, and the potential for future work? I think the usage of environment

[tvm] branch main updated (fd27aab -> 2711229)

This is an automated email from the ASF dual-hosted git repository. jroesch pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git. from fd27aab [Hexagon] Disable `thread_local` on Hexagon (#9025) add 2711229 [ONNX] enable the onnx tests after PR

[tvm] branch main updated (a579cea -> fd27aab)

This is an automated email from the ASF dual-hosted git repository. jroesch pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git. from a579cea [Hexagon] Allow undefined symbols in libtvm_runtime.so on Hexagon (#9024) add fd27aab [Hexagon] Disable

[GitHub] [tvm] jroesch merged pull request #9019: [ONNX] enable the onnx tests after PR #8274 merged

jroesch merged pull request #9019: URL: https://github.com/apache/tvm/pull/9019 -- 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:

[GitHub] [tvm] arangasa opened a new pull request #9028: [ONNX][#8838] QLinearSigmoid contrib op and Bug Fix for DequantizeLinear

arangasa opened a new pull request #9028: URL: https://github.com/apache/tvm/pull/9028 This PR implements com.microsoft.QLinearSigmoid (tracked here). This is implemented as dequantize->sigmoid->quantize for now. This also fixes an issue with DequantizeLinear for tensors with

[GitHub] [tvm] jroesch merged pull request #9025: [Hexagon] Disable `thread_local` on Hexagon

jroesch merged pull request #9025: URL: https://github.com/apache/tvm/pull/9025 -- 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:

[tvm] branch main updated (4c77bae -> a579cea)

This is an automated email from the ASF dual-hosted git repository. jroesch pushed a change to branch main in repository https://gitbox.apache.org/repos/asf/tvm.git. from 4c77bae [Onnx] Add momentum (#9000) add a579cea [Hexagon] Allow undefined symbols in libtvm_runtime.so on

[GitHub] [tvm] jroesch merged pull request #9024: [Hexagon] Allow undefined symbols in libtvm_runtime.so on Hexagon

jroesch merged pull request #9024: URL: https://github.com/apache/tvm/pull/9024 -- 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:

[GitHub] [tvm] arangasa closed pull request #9027: [ONNX][#8838] QLinearSigmoid contrib op and Bug Fix for DequantizeLinear

arangasa closed pull request #9027: URL: https://github.com/apache/tvm/pull/9027 -- 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:

[GitHub] [tvm] arangasa opened a new pull request #9027: [ONNX][#8838] QLinearSigmoid contrib op and Bug Fix for DequantizeLinear

arangasa opened a new pull request #9027: URL: https://github.com/apache/tvm/pull/9027 This PR implements com.microsoft.QLinearSigmoid (tracked [here](https://github.com/apache/tvm/issues/8838)). This is implemented as dequantize->sigmoid->quantize for now. This also fixes an issue

[GitHub] [tvm] corehalt commented on pull request #8668: Visualization Relay IR on terminal

corehalt commented on pull request #8668: URL: https://github.com/apache/tvm/pull/8668#issuecomment-920608150 Hi @kueitang love this feature, thank you! This will also allow to visualize a BYOC-partitioned Relay graph? Reading partitioned Relay IR is a bit tedious in some cases and it

<    1   2