Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / aws/sagemaker-inference-toolkit issues and pull requests

#137 - Is streaming supported?

Issue - State: open - Opened by inf3rnus 8 months ago

#135 - (WIP) Feature: fix `ENABLE_MULTI_MODEL` setting

Pull Request - State: open - Opened by taylorfturner about 1 year ago - 1 comment

#134 - Triton container documentation implies py38

Issue - State: open - Opened by david-waterworth about 1 year ago

#133 - Fix zombie process exception

Pull Request - State: closed - Opened by sachanub about 1 year ago - 5 comments

#130 - feat: support codeartifact for installing requirements.txt packages

Pull Request - State: closed - Opened by humanzz over 1 year ago - 7 comments

#129 - Add support for SAGEMAKER_MODEL_SERVER_TIMEOUT_SECONDS variable

Pull Request - State: closed - Opened by davidthomas426 over 1 year ago - 1 comment

#128 - Tensorflow Inference toolkit

Issue - State: open - Opened by vincentvic over 1 year ago

#127 - Support for parquet encoder and decoder

Issue - State: open - Opened by lorenzwalthert over 1 year ago

#126 - chore: translate readme to turkish

Pull Request - State: open - Opened by CaglarTaha over 1 year ago

#125 - fix: transform function to support proper batch inference

Pull Request - State: open - Opened by taepd over 1 year ago - 3 comments

#124 - OOM errors creating an endpoint for LLMs

Issue - State: closed - Opened by jsleight over 1 year ago - 1 comment

#123 - Support user-defined batch inference logic

Issue - State: open - Opened by dgcnz over 1 year ago

#121 - handle max_request_size param set by SM platofrm

Pull Request - State: closed - Opened by rohithkrn over 1 year ago - 1 comment

#120 - feature: Add support for py38, py39 and py310

Pull Request - State: closed - Opened by maaquib over 1 year ago - 2 comments

#119 - relax retrying dependency

Pull Request - State: closed - Opened by jakob-keller almost 2 years ago - 7 comments

#118 - Add environment variable VMARGS

Pull Request - State: closed - Opened by nskool almost 2 years ago - 8 comments

#117 - Document default_pre_model_fn and default_model_warmup_fn

Issue - State: open - Opened by l3ku about 2 years ago

#116 - stop_server function

Issue - State: open - Opened by Duncan-Haywood about 2 years ago

#115 - feature: Write handler to config file and serve model directly from o…

Pull Request - State: closed - Opened by davidthomas426 about 2 years ago - 3 comments

#113 - Long Model Loading times in Multimodel Server

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

#112 - fix: fix loading user custom script

Pull Request - State: open - Opened by SuperBo about 2 years ago - 3 comments

#111 - [DRAFT] Support CustomAttributes / request context in handler function overrides

Pull Request - State: open - Opened by athewsey over 2 years ago - 1 comment

#110 - Support CustomAttributes in script mode (for PyTorch, HuggingFace, etc)

Issue - State: open - Opened by athewsey over 2 years ago - 1 comment

#109 - Pass context from handler service to handler function

Pull Request - State: closed - Opened by waytrue17 over 2 years ago - 25 comments

#108 - Modify transform function to support batch inference

Pull Request - State: closed - Opened by nskool over 2 years ago - 11 comments

#107 - Add (and test) Support for Python 3.8

Issue - State: open - Opened by ddluke over 2 years ago

#106 - Include "Requires-Python" in source distributions

Issue - State: open - Opened by ddluke over 2 years ago

#105 - fix: inference handler issue

Pull Request - State: closed - Opened by Qingzi-Lan over 2 years ago - 5 comments

#104 - feature: preModel and warmup function support

Pull Request - State: closed - Opened by Qingzi-Lan over 2 years ago - 10 comments

#102 - Launch MMS without repackaging model contents

Issue - State: closed - Opened by fm1ch4 over 2 years ago

#99 - Server Timeout Unit is Minutes in MMS, but docstring says Seconds

Issue - State: closed - Opened by kastman almost 3 years ago - 2 comments

#98 - fix: Add configurable startup timeout

Pull Request - State: closed - Opened by davidthomas426 almost 3 years ago - 9 comments

#97 - fix: log4j migration from 1 to 2. Moving properties file to xml

Pull Request - State: closed - Opened by maaquib almost 3 years ago - 3 comments

#96 - Dummy pull request to check CI build

Pull Request - State: closed - Opened by davidthomas426 almost 3 years ago - 1 comment

#95 - fix: Add formatter to logger with a timestamp.

Pull Request - State: closed - Opened by davidthomas426 almost 3 years ago - 2 comments

#94 - Model Server fails to start with multi-model-server version 1.1.5

Issue - State: closed - Opened by joseproura almost 3 years ago - 1 comment

#93 - fix: Re-enable output capturing

Pull Request - State: closed - Opened by nskool almost 3 years ago - 3 comments

#92 - fix: Increase timeout for starting model server to 10 minutes

Pull Request - State: closed - Opened by nskool almost 3 years ago - 11 comments

#91 - fix: Increase timeout for starting model server to 10 minutes

Pull Request - State: closed - Opened by davidthomas426 almost 3 years ago - 2 comments

#90 - feature: support configuration of extra mms parameters via env

Pull Request - State: closed - Opened by henryhu666 about 3 years ago - 3 comments

#89 - fix: fix whl installation error from inference requirements

Pull Request - State: open - Opened by yifeim about 3 years ago - 4 comments

#88 - fix: Add NOTICE to MANIFEST

Pull Request - State: closed - Opened by BastianZim about 3 years ago - 12 comments

#87 - Add VMARGS to disable the container support.

Pull Request - State: closed - Opened by vdantu over 3 years ago - 2 comments

#87 - Add VMARGS to disable the container support.

Pull Request - State: closed - Opened by vdantu over 3 years ago - 2 comments

#85 - Support CodeArtifact repositories for installing Python packages

Issue - State: closed - Opened by setu4993 over 3 years ago - 4 comments

#83 - fix: Fixing issue #82

Pull Request - State: closed - Opened by amaharek over 3 years ago - 3 comments

#82 - JVM detect the CPU count as 1 when more CPUs are available for the container.

Issue - State: closed - Opened by amaharek over 3 years ago - 2 comments

#81 - fix: remove conflicting dependencies

Pull Request - State: closed - Opened by ajaykarpur almost 4 years ago - 1 comment
Labels: priority: high, type: bug

#80 - fix: remove conflicting dependencies

Pull Request - State: closed - Opened by schenqian almost 4 years ago - 1 comment
Labels: priority: high, type: bug

#79 - fix: remove conflicting dependencies

Pull Request - State: closed - Opened by schenqian almost 4 years ago - 2 comments

#77 - In a multi-model scenario pass model name as argument to input_fn() and output_fn()

Issue - State: open - Opened by RajeshRamchander almost 4 years ago
Labels: type: enhancement

#76 - fix: add SIGCHILD Handler for MMS

Pull Request - State: closed - Opened by dhanainme almost 4 years ago - 6 comments

#75 - fix: noop change

Pull Request - State: closed - Opened by ahakanbaba about 4 years ago - 2 comments

#75 - fix: noop change

Pull Request - State: closed - Opened by ahakanbaba about 4 years ago - 2 comments

#74 - fix: write the mms config file to a subdir of base_dir instead of /etc

Pull Request - State: open - Opened by ahakanbaba about 4 years ago - 4 comments

#73 - fix: add model-archiver and multi-model-server to required packages

Pull Request - State: closed - Opened by ajaykarpur about 4 years ago - 4 comments
Labels: type: bug

#72 - SageMaker inference should be able to run as non-root user.

Issue - State: open - Opened by akulk314 about 4 years ago - 3 comments
Labels: type: enhancement

#71 - change: upgrade MMS version and update command

Pull Request - State: closed - Opened by akulk314 about 4 years ago - 5 comments

#70 - Be able to change SageMaker endpoint log level

Issue - State: open - Opened by oonisim about 4 years ago - 12 comments
Labels: type: enhancement, contribution welcomed

#69 - Warning: "Calling MMS with mxnet-model-server. Please move to multi-model-server."

Issue - State: open - Opened by mb-dev about 4 years ago - 2 comments
Labels: type: enhancement, contribution welcomed

#68 - How to overwrite batch transform output in S3

Issue - State: open - Opened by BaoshengHeTR about 4 years ago - 5 comments
Labels: type: documentation

#68 - How to overwrite batch transform output in S3

Issue - State: open - Opened by BaoshengHeTR about 4 years ago - 5 comments
Labels: type: documentation

#67 - Update README.md

Pull Request - State: closed - Opened by yonghyeokrhee about 4 years ago - 1 comment

#66 - fix: add model-archiver and multi-model-server to listed dependencies

Pull Request - State: closed - Opened by ankurtri over 4 years ago - 3 comments

#66 - fix: add model-archiver and multi-model-server to listed dependencies

Pull Request - State: closed - Opened by ankurtri over 4 years ago - 3 comments

#65 - Allow model_fn to get more arguments

Issue - State: open - Opened by ehsanmok over 4 years ago - 2 comments
Labels: type: enhancement

#64 - Enhance UX for inference

Issue - State: open - Opened by ehsanmok over 4 years ago
Labels: type: UX, type: enhancement

#63 - fix: clean up stack trace and sanitize only response phrase

Pull Request - State: closed - Opened by ajaykarpur over 4 years ago - 1 comment

#62 - fix: remove prohibited characters from error response

Pull Request - State: closed - Opened by ajaykarpur over 4 years ago - 7 comments

#61 - feature: Support multiple accept types

Pull Request - State: closed - Opened by bveeramani over 4 years ago - 2 comments

#60 - feature: Decode application/x-npz content type

Pull Request - State: closed - Opened by bveeramani over 4 years ago - 2 comments

#60 - feature: Decode application/x-npz content type

Pull Request - State: closed - Opened by bveeramani over 4 years ago - 2 comments

#59 - logging in JSON format

Issue - State: open - Opened by alext234 over 4 years ago - 2 comments
Labels: type: question

#58 - Torchserve support for PyTorch Inference

Pull Request - State: closed - Opened by dhanainme over 4 years ago - 1 comment

#58 - Torchserve support for PyTorch Inference

Pull Request - State: closed - Opened by dhanainme over 4 years ago - 1 comment

#57 - infra: clarify feature request issue template

Pull Request - State: closed - Opened by ajaykarpur over 4 years ago - 1 comment
Labels: priority: low, type: maintenance

#56 - Request encoding function for application/jsonlines output format

Issue - State: open - Opened by BaoshengHeTR over 4 years ago - 5 comments
Labels: type: question, status: pending

#55 - Deploying multiple model artifacts, each having their own inference handler

Issue - State: open - Opened by dean-cpi over 4 years ago - 5 comments
Labels: type: enhancement

#54 - documentation: fix package name in pip install command

Pull Request - State: closed - Opened by obrep over 4 years ago - 2 comments

#53 - End-to-End example for *NON* multi model deployment

Issue - State: open - Opened by nvs-abhilash over 4 years ago - 7 comments
Labels: type: question, type: documentation

#52 - Ability to define custom log4j properties

Issue - State: open - Opened by arvarik over 4 years ago - 1 comment
Labels: type: enhancement

#51 - fix: include stacktrace when handling an error

Pull Request - State: closed - Opened by laurenyu over 4 years ago

#51 - fix: include stacktrace when handling an error

Pull Request - State: closed - Opened by laurenyu over 4 years ago

#50 - Cloudwatch does not log stack trace when error occurs during invocation.

Issue - State: closed - Opened by Olamyy over 4 years ago - 3 comments
Labels: status: pending, type: bug

#49 - fix: Remove typing

Pull Request - State: closed - Opened by ajaykarpur over 4 years ago
Labels: priority: low, type: bug

#49 - fix: Remove typing

Pull Request - State: closed - Opened by ajaykarpur over 4 years ago
Labels: priority: low, type: bug

#48 - infra: Do not parallelize unit tests in buildspecs

Pull Request - State: closed - Opened by ajaykarpur over 4 years ago

#48 - infra: Do not parallelize unit tests in buildspecs

Pull Request - State: closed - Opened by ajaykarpur over 4 years ago

#47 - feature: Add Python 3.7 support

Pull Request - State: closed - Opened by ajaykarpur over 4 years ago

#46 - Update README example

Issue - State: open - Opened by ajaykarpur over 4 years ago
Labels: type: documentation

#45 - providing gpu_id to the custom inference script's model_fn

Pull Request - State: closed - Opened by muhyun over 4 years ago - 7 comments