Weekly pr metrics report - 2025-02-01..2025-02-07

[github-actions]

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