Open
Description
Issue Metrics
| Metric | Average | Median | 90th percentile |
|---|---|---|---|
| Time to first response | 13:09:15 | 0:04:35 | 1 day, 23:01:44 |
| Time to close | 1 day, 12:18:41 | 14:33:58 | 4 days, 4:38:25 |
| Time to answer | None | None | None |
| Metric | Count |
|---|---|
| Number of items that remain open | 15 |
| Number of items closed | 103 |
| Number of most active mentors | 0 |
| Total number of items created | 118 |
| Title | URL | Author | Time to first response | Time to close | Time to answer |
|---|---|---|---|---|---|
| Expand dict API to allow dtype strings | pytorch/executorch#8323 | lucylq | 0:00:11 | 7:57:28 | None |
| Reuse GELU implementation from PyTorch core | pytorch/executorch#8322 | pytorchbot | 3 days, 20:08:10 | 4 days, 21:31:20 | None |
| Integrate torchgen exception boundary with ExecuTorch | pytorch/executorch#8321 | pytorchbot | 3 days, 20:15:12 | 3 days, 20:15:34 | None |
| Use c10 version of half/bfloat16 in executorch | pytorch/executorch#8320 | pytorchbot | 1:17:56 | 3 days, 20:14:32 | None |
| Update HuggingFace model checkpointer import | pytorch/executorch#8319 | jackzhxng | 3 days, 22:20:37 | 3 days, 23:01:06 | None |
| roll out s/IF_FALSE/UNLESS/ for ET_LOG macros | pytorch/executorch#8318 | swolchok | 5 days, 1:23:58 | None | None |
| s/IF_FALSE/UNLESS/ in ET_LOG macro names | pytorch/executorch#8317 | swolchok | 6:21:24 | None | None |
| Use tensor_shape_to_c_string for error in check_glu_args | pytorch/executorch#8316 | swolchok | 2 days, 21:20:23 | 5 days, 20:18:20 | None |
| [BE] Move submodule check out from setup.py | pytorch/executorch#8315 | larryliu0820 | 1:01:25 | 1:41:19 | None |
| Use tensor_shape_to_c_string for error in check_mask_indices | pytorch/executorch#8314 | swolchok | 7:12:43 | 4 days, 5:01:20 | None |
| Add lift scalar to constant tensor pass | pytorch/executorch#8313 | cccclai | 0:00:21 | None | None |
| Convolution 1D optimized axis map | pytorch/executorch#8312 | Abhi-hpp | 0:00:15 | None | None |
| Static attention implementation | pytorch/executorch#8310 | sxu | 0:00:11 | 7:22:16 | None |
| Qualcomm AI Engine Direct - Fix Argmin | pytorch/executorch#8308 | winskuo-quic | 1:40:43 | 4 days, 3:06:47 | None |
| [ArmBackend] Fix in setup.sh for MacOS | pytorch/executorch#8307 | freddan80 | 2:56:39 | 3:32:06 | None |
| Qualcomm AI Engine Direct - op support | pytorch/executorch#8306 | haowhsu-quic | 8:18:26 | 3 days, 19:29:58 | None |
| [ET-VK] fix index error bug in ViewCopyToSqueezeUnsqueezePass | pytorch/executorch#8305 | pytorchbot | None | None | None |
| oss qnn bug fix | pytorch/executorch#8304 | billmguo | 0:00:09 | 3 days, 17:34:52 | None |
| update emitter default evalue to have dim order match size | pytorch/executorch#8303 | JacobSzwejbka | 0:00:10 | 12:45:54 | None |
| Replace pytree_assert with production pytree_check. Remove pytree_unreachable | pytorch/executorch#8302 | pytorchbot | 0:34:24 | 0:34:36 | None |
| add a bunch of bounds checking to pytree | pytorch/executorch#8301 | pytorchbot | 0:33:27 | 0:33:43 | None |
| add begin/end to ExecuTorch pytree::arr | pytorch/executorch#8300 | pytorchbot | 0:00:45 | 0:01:23 | None |
| Qualcomm AI Engine Direct - XR model enablement pipe_clean | pytorch/executorch#8299 | haowhsu-quic | 4 days, 2:49:05 | 7 days, 16:04:15 | None |
| Add a pass to convert rank-0 tensor to rank-1 tensor | pytorch/executorch#8298 | cccclai | 0:00:08 | 4 days, 21:26:12 | None |
| Add BUCK files for llava python and C++ libs | pytorch/executorch#8297 | larryliu0820 | 0:00:08 | 20:18:07 | None |
| Move tensor_shape_to_c_string back to runtime/core | pytorch/executorch#8296 | swolchok | 18:00:30 | 18:57:37 | None |
| skip op in partitioning if there are bool input tensors | pytorch/executorch#8295 | nathanaelsee | 0:00:07 | 4:41:44 | None |
| [ET-VK] support biases in buffer-based linear shader | pytorch/executorch#8291 | pytorchbot | 0:07:58 | 0:09:13 | None |
| [Not for Commit] Example use of new dim_order api | pytorch/executorch#8289 | digantdesai | None | None | None |
| Update documentation flow and add placeholders | pytorch/executorch#8287 | GregoryComer | 1:15:08 | 15:13:29 | None |
| Fix build | pytorch/executorch#8286 | kirklandsign | 0:00:32 | 4:07:42 | None |
| Dont error check unused tensors in convolution_backward.out | pytorch/executorch#8285 | JacobSzwejbka | 0:00:06 | 6:49:17 | None |
| [ET-VK] support biases in buffer-based linear shader | pytorch/executorch#8284 | SS-JIA | 0:00:13 | 2:55:36 | None |
| [ET-VK] fix index error bug in ViewCopyToSqueezeUnsqueezePass | pytorch/executorch#8281 | nathanaelsee | 0:00:20 | 14:11:31 | None |
| [executorch][flat_tensor] DataMap implementation | pytorch/executorch#8280 | kirklandsign | 0:02:16 | 0:02:32 | None |
| Fix unittests failures introduced from D69212362 | pytorch/executorch#8278 | zhxchen17 | 0:00:07 | 2:34:16 | None |
| Implement _fft_r2c core ATen op | pytorch/executorch#8277 | swolchok | 10:42:08 | 10:42:17 | None |
| fix memory planning not skipping None values | pytorch/executorch#8276 | JacobSzwejbka | 0:00:10 | 3:40:16 | None |
| Add visualization debug pass | pytorch/executorch#8275 | AdrianLundell | 6:20:22 | 6:21:31 | None |
| Use STDOUT fallback after FileNotFoundError in gather benchmarks | pytorch/executorch#8258 | GregoryComer | 9:53:39 | 10:01:13 | None |
| [ExecuTorch] Add broadcasting support to optimized op_div | pytorch/executorch#8257 | kimishpatel | 5 days, 15:20:01 | None | None |
| [Executorch] Add broadcasting support to optimized op_sub | pytorch/executorch#8256 | kimishpatel | 5 days, 15:18:41 | None | None |
| [Executorch] Refactor op_add to support op_sub broadcasting | pytorch/executorch#8255 | kimishpatel | 13:04:32 | None | None |
| [ET-VK][int4] Wrap int4 linear calls with view_copy nodes to squeeze/unsqueeze inputs | pytorch/executorch#8254 | pytorchbot | 14:34:10 | 14:34:28 | None |
| [ET-VK][int4] patch 4-bit linear op for ensuring w-packed in/out | pytorch/executorch#8253 | pytorchbot | 14:33:44 | 14:33:58 | None |
| [ET-VK][int4] patch 4-bit source transformation quantizer to support linear modules with biases | pytorch/executorch#8252 | pytorchbot | 2:39:51 | 2:40:21 | None |
| Install mkl for PyTorch and libjpeg/libpng for TorchVision | pytorch/executorch#8248 | huydhn | 19:53:45 | 6:28:18 | None |
| Actually run op_linear_test in OSS | pytorch/executorch#8238 | swolchok | 4 days, 16:14:49 | 4 days, 17:39:28 | None |
| Add missing Pyre mode headers] [batch:12/1577] [shard:13/N] | pytorch/executorch#8236 | kirklandsign | 0:00:09 | 0:03:20 | None |
| Update verifier to handle None Tensor outputs | pytorch/executorch#8235 | JacobSzwejbka | 0:00:10 | 4:28:54 | None |
| Fix allocation-size-too-big crash in prepare_input_tensors | pytorch/executorch#8233 | dbort | 0:00:11 | 2:53:26 | None |
| Delete the ramaining empty folder | pytorch/executorch#8232 | cccclai | 0:00:08 | 4:55:40 | None |
| [BE] Remind users to update submodule | pytorch/executorch#8229 | pytorchbot | 0:51:22 | 0:51:45 | None |
| aten.leakyrelu.default Op registery | pytorch/executorch#8227 | hossein1387 | 0:00:12 | None | None |
| [ET-VK][int4] Wrap int4 linear calls with view_copy nodes to squeeze/unsqueeze inputs | pytorch/executorch#8226 | nathanaelsee | 0:00:24 | 7:42:07 | None |
| [ET-VK][int4] patch 4-bit linear op for ensuring w-packed in/out | pytorch/executorch#8225 | nathanaelsee | 0:00:13 | 7:42:04 | None |
| [ET-VK][int4] patch 4-bit source transformation quantizer to support linear modules with biases | pytorch/executorch#8224 | nathanaelsee | 0:00:15 | 7:42:01 | None |
| [executorch] Suppress wdeprecated warnings on tensor_layout.h and named_data_map.h | pytorch/executorch#8223 | pytorchbot | 0:01:28 | 0:01:51 | None |
| [ET-VK] Add util to extract VkGraph from Program delegate data | pytorch/executorch#8219 | pytorchbot | 0:31:19 | 1:36:33 | None |
| Enable buck2 daemon restarter | pytorch/executorch#8212 | swolchok | 2:03:33 | 2:28:02 | None |
| Qualcomm AI Engine Direct - Enable AR-N model for prompt processing in hybrid mode | pytorch/executorch#8210 | shewu-quic | None | None | None |
| Arm backend: Add is_node_supported checks for 4 ops | pytorch/executorch#8209 | Erik-Lundell | 22:20:53 | 1 day, 5:59:59 | None |
| Increase string buffer size for tensor size errors | pytorch/executorch#8208 | GregoryComer | 0:00:15 | 9:39:46 | None |
| Move print_delegate_info to devtools | pytorch/executorch#8207 | limintang | 0:00:07 | 1 day, 18:06:51 | None |
| [ExecuTorch] Add broadcast support for optimized add op | pytorch/executorch#8205 | kimishpatel | 1 day, 18:19:52 | 10 days, 3:17:34 | None |
| [Executorch] Refactor op_mul's broadcasting utils | pytorch/executorch#8204 | kimishpatel | 1 day, 17:42:20 | 8 days, 13:30:58 | None |
| [BE] Remind users to update submodule | pytorch/executorch#8203 | larryliu0820 | 5:00:00 | 19:39:51 | None |
| [cadece][g3][facto] enable facto generation on G3 operators | pytorch/executorch#8202 | zonglinpeng | 0:01:50 | 21:10:23 | None |
| [executorch] Suppress wdeprecated warnings on tensor_layout.h and named_data_map.h | pytorch/executorch#8201 | lucylq | 0:00:21 | 18:51:10 | None |
| [executorch][runtime] Add NamedDataMap to method load | pytorch/executorch#8200 | lucylq | 0:00:11 | 6 days, 18:25:22 | None |
| [ET-VK] Add util to extract VkGraph from Program delegate data | pytorch/executorch#8199 | jorgep31415 | 0:00:08 | 17:58:57 | None |
| method_meta add uses_backend function | pytorch/executorch#8198 | cmt0 | 0:00:09 | 1 day, 3:35:34 | None |
| Elide static view copies | pytorch/executorch#8197 | metascroy | 0:00:08 | 2 days, 22:05:28 | None |
| Warn if wrong type is given for Llama export for XNNPACK | pytorch/executorch#8195 | mergennachin | 0:08:40 | None | None |
| Trial revert of --use-pt-pinned-commit for llava | pytorch/executorch#8193 | swolchok | None | None | None |
| NOCOMMIT: log peak mem usage and swapouts to debug test-llava-runner | pytorch/executorch#8192 | swolchok | None | None | None |
Add strict=True explicit flag to Executorch invocations to torch.export.export |
pytorch/executorch#8190 | gmagogsfm | 0:00:10 | 1 day, 0:44:15 | None |
| [ET-VK] Add utility functions to get pipeline executable properties | pytorch/executorch#8189 | pytorchbot | 0:01:49 | 0:02:29 | None |
| Fix missing canonicalize_program in qnn | pytorch/executorch#8188 | metascroy | 2:38:14 | 3:11:33 | None |
| Pass ForwardOptions from top level module and also return any relevant state as output | pytorch/executorch#8186 | sxu | 0:00:09 | 2 days, 4:17:04 | None |
| Implement portable abs for complex input | pytorch/executorch#8183 | pytorchbot | 0:49:09 | 0:49:42 | None |
| [ET-VK] Add utility functions to get pipeline executable properties | pytorch/executorch#8182 | SS-JIA | 0:00:13 | 2:09:23 | None |
| try to use 48xlarge runners for llava-runner and eval_llama-mmlu | pytorch/executorch#8181 | swolchok | None | None | None |
| New Getting Started documentation | pytorch/executorch#8179 | GregoryComer | 17:58:11 | 19:02:03 | None |
| Fix runner_utils path in Arm backend | pytorch/executorch#8176 | AdrianLundell | 0:18:16 | 0:41:15 | None |
| [Do Not Merge] multi partitioned graph in one qnn delegate | pytorch/executorch#8175 | chunit-quic | None | None | None |
| Add a parameter to output delegate summary in llama export | pytorch/executorch#8174 | limintang | 15:27:49 | 2 days, 20:11:24 | None |
| Increase CI timeout | pytorch/executorch#8173 | metascroy | 1:13:15 | 14:07:17 | None |
| Error out if registering prim ops multiple times | pytorch/executorch#8172 | larryliu0820 | 0:00:08 | 21:38:16 | None |
| patch rms norm recompose pass | pytorch/executorch#8170 | cccclai | 0:00:09 | 19:43:24 | None |
| [ExecuTorch][to_backend] Enable passing Delegation Spec to to_backend | pytorch/executorch#8165 | mcr229 | 0:00:08 | None | None |
| open source quantized_fully_connected | pytorch/executorch#8164 | zonglinpeng | 0:00:07 | 1 day, 20:29:11 | None |
| [data_loader] Fix a log message that didn't use all of its args | pytorch/executorch#8161 | dbort | 0:09:40 | 1:44:08 | None |
| use _upload_docs job | pytorch/executorch#8160 | metascroy | 1:36:14 | None | None |
| create MetaETDumpGen class to abstract copy_tensor_to_debug_buffer function | pytorch/executorch#8157 | Gasoonjia | 0:00:11 | None | None |
| Add EXECUTORCH_CLIENTS for visibility in extension/export_util:export_util | pytorch/executorch#8155 | YIWENX14 | 0:00:07 | 20:14:17 | None |
| .gitignores for ios demo artifacts | pytorch/executorch#8153 | dustinfreeman | 2 days, 19:22:58 | None | None |
| add aot op checks for new hifi ops, add pow_Tensor_Scalar test in HiFi | pytorch/executorch#8152 | zonglinpeng | 0:00:08 | 1 day, 23:04:56 | None |
| Adding instructions on how to run the Deepseek R1 Distill Llama 8B model | pytorch/executorch#8151 | tarun292 | 0:04:54 | 0:20:22 | None |
| Draft version of sliding cache model for coreml | pytorch/executorch#8150 | YIWENX14 | 0:00:09 | None | None |
| Add SoC SXR2250P | pytorch/executorch#8148 | cccclai | 0:01:02 | 8 days, 0:51:43 | None |
| [executorch] Create target for named_data_map | pytorch/executorch#8147 | kirklandsign | 0:02:02 | 0:04:23 | None |
| Implement portable abs for complex input | pytorch/executorch#8146 | swolchok | 0:00:11 | 23:21:47 | None |
| Implement portable abs for complex input | pytorch/executorch#8145 | swolchok | None | None | None |
| Implement portable abs for complex input | pytorch/executorch#8144 | swolchok | None | None | None |
| [cadence][hifi] update quantized_relu_per_tensor_out signature to match internal flow | pytorch/executorch#8143 | zonglinpeng | 0:00:07 | 2 days, 3:37:22 | None |
| don't double-install executorch for one-off linux pull-time jobs | pytorch/executorch#8142 | swolchok | 1 day, 4:34:34 | 1 day, 5:49:40 | None |
| Implement portable abs for complex input | pytorch/executorch#8141 | swolchok | None | None | None |
| Prioritise FuseBatchnorm2D over DecomposeBatchNormPass | pytorch/executorch#8140 | tom-arm | 0:17:15 | 2 days, 15:50:59 | None |
| Bugfix in TosaSpecification representation | pytorch/executorch#8137 | per | 1:41:36 | 2 days, 1:06:21 | None |
| Add hardswish operator to Arm backend | pytorch/executorch#8136 | tom-arm | 2:03:30 | 21:17:56 | None |
| Qualcomm AI Engine Direct - documentation for KV cache update | pytorch/executorch#8134 | haowhsu-quic | 2 days, 9:59:27 | 2 days, 9:59:48 | None |
| Add tosa serializer to .mypy.ini | pytorch/executorch#8133 | perheld | 0:05:21 | 9:15:16 | None |
| Update README.md with refined list of products and backends | pytorch/executorch#8131 | orionr | 14:43:17 | 22:49:18 | None |
| Add a pass to replace nodes with empty tensors with full. | pytorch/executorch#8130 | hsharma35 | 0:00:12 | 2 days, 15:41:42 | None |
| Single location to update optional args for all attentions | pytorch/executorch#8128 | iseeyuan | 0:00:09 | 14:26:51 | None |
| support input_pos > 0 for prefill model | pytorch/executorch#8127 | billmguo | 0:00:07 | None | None |
| Update the Readme with better copy and details on how to find us on Discord | pytorch/executorch#8126 | byjlw | 0:04:16 | 0:09:13 | None |
This report was generated with the Issue Metrics Action
Search query used to find these items: repo:pytorch/executorch is:pr created:2025-02-01..2025-02-07
Metadata
Metadata
Assignees
Labels
No labels