facebookresearch/param issues and pull requests

#93 - Support for SLURM in commsTraceReplay

Pull Request - State: open - Opened by TaekyungHeo 6 months ago - 4 comments
Labels: CLA Signed

#92 - Fix incorrect path construction in commsTraceReplayBench

Pull Request - State: open - Opened by TaekyungHeo 6 months ago - 5 comments
Labels: CLA Signed

#91 - Simplify ET replay and add Integration testing

Issue - State: open - Opened by briancoutinho 6 months ago - 2 comments
Labels: bug, enhancement

#90 - Refactor and Optimize trace_link.py for Enhanced Readability and Performance

Pull Request - State: open - Opened by TaekyungHeo 7 months ago - 1 comment
Labels: CLA Signed

#89 - Support Chakra traces in execution_trace.py

Pull Request - State: closed - Opened by TaekyungHeo 9 months ago - 3 comments
Labels: CLA Signed, Merged, fb-exported

#88 - Update trace_link.py to merge Kineto trace with Pytorch execution trace.

Pull Request - State: closed - Opened by JoongunPark 10 months ago - 48 comments
Labels: CLA Signed, Merged

#87 - Add IB device filtering

Pull Request - State: closed - Opened by TiagoMAntunes 11 months ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#86 - Add execution trace trimming by iteration

Pull Request - State: closed - Opened by briancoutinho 12 months ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#85 - ImportError: cannot import name 'ExecutionTraceObserver' from 'torch.profiler'

Issue - State: closed - Opened by tangyuelm about 1 year ago - 1 comment

#84 - OOM for all gather comms tests

Issue - State: open - Opened by roywei about 1 year ago

#83 - Add warning for trace_link script when annotation is not present, fix write for gzip files

Pull Request - State: closed - Opened by briancoutinho about 1 year ago - 3 comments
Labels: CLA Signed, Merged, fb-exported

#82 - Enabling Kineto dumping on Pytorch

Issue - State: closed - Opened by rashidi1saeed about 1 year ago - 2 comments

#81 - Cloud Data Pull Request : true+ 1

Issue - State: closed - Opened by MrTuneUp101 about 1 year ago - 4 comments

#80 - Add record function ID parsing

Pull Request - State: closed - Opened by briancoutinho about 1 year ago - 3 comments
Labels: CLA Signed, Merged, fb-exported

#79 - Upload scripts to link execution trace and kineto trace

Pull Request - State: closed - Opened by Mingyu-Liang about 1 year ago - 18 comments
Labels: CLA Signed, Merged

#78 - Fix left over code calling ExecutionGraphObserver

Pull Request - State: closed - Opened by louisfeng about 1 year ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#76 - commsTraceReplay failing due to missing import

Issue - State: closed - Opened by harisj-xlnx about 1 year ago - 3 comments

#75 - Remove CPU and GPU support for approx_sgd

Pull Request - State: closed - Opened by q10 about 1 year ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#74 - Remove the dependency of the pre-registered op

Pull Request - State: closed - Opened by sunghlin about 1 year ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#73 - Update code owners

Pull Request - State: closed - Opened by louisfeng over 1 year ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#72 - Refactor EG to ET (#99694)

Pull Request - State: closed - Opened by louisfeng over 1 year ago - 24 comments
Labels: CLA Signed, Merged, fb-exported

#71 - Improve readTrace code

Pull Request - State: closed - Opened by Pavani-Panakanti over 1 year ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#70 - Enable setting CPU events along with CUPTI Range profiler

Pull Request - State: closed - Opened by briancoutinho over 1 year ago - 4 comments
Labels: CLA Signed, Merged, fb-exported

#69 - Add cpu operators to CUPTI profiler

Pull Request - State: closed - Opened by briancoutinho over 1 year ago - 8 comments
Labels: CLA Signed, fb-exported

#68 - refactor test_utils and bench_utils (#1478)

Pull Request - State: closed - Opened by jianyuh over 1 year ago - 1 comment
Labels: CLA Signed, fb-exported

#67 - Add CUPTI Profiler interface to measure performance events

Pull Request - State: closed - Opened by briancoutinho over 1 year ago - 1 comment
Labels: CLA Signed, fb-exported

#66 - move fbgemm_gpu import

Pull Request - State: closed - Opened by kingchc almost 2 years ago - 2 comments
Labels: CLA Signed, fb-exported

#65 - "Unrecognised argument(s): force" for comms/pt/comms_utils.py

Issue - State: closed - Opened by amathews-amd almost 2 years ago - 1 comment
Labels: bug

#64 - print arguments for comms-compute and compute modes

Pull Request - State: closed - Opened by minsii almost 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#63 - fix runtime errors for `dlrm.py`

Pull Request - State: closed - Opened by kingchc almost 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#62 - Fix TRT use in pytorch_linear inference PARAM benchmark

Pull Request - State: closed - Opened by nrsatish almost 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#61 - fix pt2pt support

Pull Request - State: closed - Opened by minsii about 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#60 - AttributeError

Issue - State: closed - Opened by PhdShi about 2 years ago - 1 comment

#59 - Request for BFloat16 support in comms.py

Issue - State: open - Opened by pruthvistony about 2 years ago

#58 - Improve compute benchmark framework test coverage

Pull Request - State: closed - Opened by louisfeng about 2 years ago - 2 comments
Labels: CLA Signed, fb-exported

#57 - Changing torch2trt to torch_tensorrt

Pull Request - State: closed - Opened by nrsatish about 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#56 - Supporting bf16 in param linear using torch.cuda.amp.autocast

Pull Request - State: closed - Opened by nrsatish about 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#55 - Fixes to param linear to hold FP32 master weights

Pull Request - State: closed - Opened by nrsatish about 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#54 - Add TF32 option explicitly to param_gemm and param_linear

Pull Request - State: closed - Opened by nrsatish about 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#53 - added bfloat16 support

Pull Request - State: closed - Opened by amathews-amd about 2 years ago - 3 comments
Labels: CLA Signed

#52 - Add option to collect execution graph

Pull Request - State: closed - Opened by louisfeng about 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#51 - execution time breakdown

Pull Request - State: open - Opened by ehsanardestani over 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#50 - Updating the arguments to accept DLRM like MLP config.

Pull Request - State: closed - Opened by ehsanardestani over 2 years ago - 2 comments
Labels: CLA Signed, fb-exported

#49 - All int and float dtype support for data_impl.py

Pull Request - State: closed - Opened by louisfeng over 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#48 - NSight Systems Collection, Analysis, and TorchScript Operator

Pull Request - State: closed - Opened by louisfeng over 2 years ago - 4 comments
Labels: CLA Signed, fb-exported

#47 - GPU memory utilization, operator execution modes

Pull Request - State: closed - Opened by louisfeng over 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#46 - Remove extra setTensorVal operation.

Pull Request - State: closed - Opened by caogao over 2 years ago - 2 comments
Labels: CLA Signed, fb-exported

#45 - Fixed PARAM CPU inference linear benchmark

Pull Request - State: closed - Opened by nrsatish over 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#44 - Inference GEMM benchmark

Pull Request - State: closed - Opened by nrsatish over 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#43 - remove num-batches from pytorch_linear

Pull Request - State: closed - Opened by jspark1105 over 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#42 - change default hidden_dim to 1024

Pull Request - State: closed - Opened by jspark1105 over 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#41 - do not share params among MLP hidden layers

Pull Request - State: closed - Opened by jspark1105 over 2 years ago - 2 comments
Labels: CLA Signed, fb-exported

#40 - recommended NCCL versions?

Issue - State: closed - Opened by QiaofengQin over 2 years ago - 1 comment

#39 - Improve benchmark runtime with data check

Pull Request - State: closed - Opened by Sergei-Lebedev over 2 years ago - 1 comment
Labels: CLA Signed

#38 - Change benchmark time unit to millisecond

Pull Request - State: closed - Opened by louisfeng over 2 years ago - 1 comment
Labels: CLA Signed, fb-exported

#37 - fix MASTER_ADDR and MASTER_PORT setting

Pull Request - State: closed - Opened by kingchc almost 3 years ago - 3 comments
Labels: CLA Signed, Merged, fb-exported

#36 - Support NSight Compute metric collection and resume benchmark on op run ID

Pull Request - State: closed - Opened by louisfeng almost 3 years ago - 11 comments
Labels: CLA Signed, Merged, fb-exported

#35 - Bug fix for pytorch linear non float datatype

Pull Request - State: closed - Opened by nrsatish almost 3 years ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#34 - New PARAM compute Python benchmarks framework

Pull Request - State: closed - Opened by louisfeng almost 3 years ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#33 - New compute benchmark framework

Pull Request - State: closed - Opened by louisfeng almost 3 years ago
Labels: CLA Signed

#32 - New compute microbenchmarks.

Pull Request - State: closed - Opened by louisfeng almost 3 years ago
Labels: CLA Signed

#31 - Restructure train/compute/pt benchmarks for new development

Pull Request - State: closed - Opened by louisfeng almost 3 years ago - 3 comments
Labels: CLA Signed, fb-exported

#30 - Reorganize and refactor new compute benchmarks

Pull Request - State: closed - Opened by louisfeng almost 3 years ago
Labels: CLA Signed

#29 - Added LAMB optimizer to param linear benchmark

Pull Request - State: closed - Opened by nrsatish almost 3 years ago - 1 comment
Labels: fb-exported

#28 - reproduce UCC error at exit

Pull Request - State: closed - Opened by kingchc almost 3 years ago - 1 comment
Labels: bug, CLA Signed, fb-exported, DO NOT MERGE

#27 - ata validation check for reduce_scatter and comms replay

Pull Request - State: closed - Opened by kingchc almost 3 years ago - 4 comments
Labels: CLA Signed, Merged, fb-exported

#26 - fix collective overlap test

Pull Request - State: closed - Opened by Sergei-Lebedev about 3 years ago - 2 comments
Labels: CLA Signed

#25 - Fix all_to_allv shrink for equal split

Pull Request - State: closed - Opened by Sergei-Lebedev about 3 years ago - 2 comments
Labels: CLA Signed, Merged

#24 - Auto-shrink for alltoallv

Issue - State: closed - Opened by mshiryaev about 3 years ago - 1 comment

#23 - How do I generate traces for replay?

Issue - State: closed - Opened by ChenYuHo about 3 years ago - 1 comment

#22 - Fix the timing issues in PT Linear benchmark

Pull Request - State: closed - Opened by jianyuh over 3 years ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#21 - pt2pt one-to-one benchmark implementation

Pull Request - State: closed - Opened by kshiteejm over 3 years ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#20 - update trace replay mode for post analysis

Pull Request - State: closed - Opened by kingchc over 3 years ago - 5 comments
Labels: CLA Signed, Merged, fb-exported

#19 - Adding some new computation benchmarks and utilities

Pull Request - State: closed - Opened by louisfeng over 3 years ago - 1 comment
Labels: CLA Signed

#18 - Create README.txt

Pull Request - State: closed - Opened by naghshineh over 3 years ago - 2 comments
Labels: CLA Signed

#17 - Add comms trace replay mode

Pull Request - State: closed - Opened by kingchc over 3 years ago - 1 comment
Labels: CLA Signed, fb-exported

#16 - add ucc plugin to comms

Pull Request - State: closed - Opened by pallab-zz over 3 years ago - 1 comment
Labels: CLA Signed, Merged, fb-exported

#15 - Fix issue of device option for DLRM bench in PARAM-Comms

Pull Request - State: closed - Opened by kingchc almost 4 years ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#14 - Support benchmarking UCC backend through PyTorch in PARAM-Comms

Pull Request - State: closed - Opened by kingchc almost 4 years ago - 2 comments
Labels: CLA Signed, Merged, fb-exported

#13 - ProcessGroupNCCL alltoall error

Issue - State: closed - Opened by amrragab8080 almost 4 years ago - 3 comments

#12 - DLRM-comms benchmark doesn't have device attribute

Issue - State: closed - Opened by Sergei-Lebedev almost 4 years ago - 4 comments

#11 - add zipf initializer for emb

Pull Request - State: closed - Opened by shz0116 almost 4 years ago - 3 comments
Labels: CLA Signed, Merged, fb-exported

#10 - Update README.md

Pull Request - State: closed - Opened by naghshineh almost 4 years ago
Labels: CLA Signed

#9 - add README.md to refer PR

Pull Request - State: closed - Opened by shz0116 almost 4 years ago - 2 comments
Labels: CLA Signed, fb-exported

#8 - Fix import paths

Pull Request - State: closed - Opened by kingchc almost 4 years ago
Labels: CLA Signed

#7 - Update README

Pull Request - State: closed - Opened by kingchc almost 4 years ago - 1 comment
Labels: CLA Signed, Merged

#6 - Update README and fix paths

Pull Request - State: closed - Opened by kingchc almost 4 years ago
Labels: CLA Signed

#5 - Add DLRM as submodule - exp_dist branch

Pull Request - State: closed - Opened by srinivas212 almost 4 years ago - 1 comment
Labels: CLA Signed, Merged

#4 - Srinivas/dlrm

Pull Request - State: closed - Opened by srinivas212 almost 4 years ago
Labels: CLA Signed

#3 - test.txt

Pull Request - State: closed - Opened by srinivas212 almost 4 years ago
Labels: CLA Signed

#2 - Adding Code of Conduct file

Pull Request - State: closed - Opened by facebook-github-bot about 4 years ago
Labels: CLA Signed

#1 - Adding Contributing file

Pull Request - State: closed - Opened by facebook-github-bot about 4 years ago
Labels: CLA Signed

GitHub / facebookresearch/param issues and pull requests