Ecosyste.ms: Issues

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

GitHub / microsoft/DeepSpeed-MII issues and pull requests

#142 - Separate server and client

Pull Request - State: closed - Opened by tohtana about 2 years ago

#141 - Unable to execute mii-sd.py in benchmark

Issue - State: closed - Opened by Yanxing-Shi about 2 years ago - 2 comments

#140 - don't work with stable diffusion 2.1

Issue - State: open - Opened by 0xDing about 2 years ago - 1 comment

#138 - Decouple conversions for gRPC from server/client code

Pull Request - State: closed - Opened by tohtana about 2 years ago - 1 comment

#137 - How to deal with imbalanced GPU memories?

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

#136 - Allow `model_path` only to deploy "local" models

Issue - State: open - Opened by philschmid about 2 years ago - 3 comments

#135 - `AssertionError` when running examples from readme

Issue - State: closed - Opened by HeyangQin about 2 years ago - 3 comments

#133 - How to do the batch inference

Issue - State: open - Opened by Emerald01 about 2 years ago - 7 comments
Labels: enhancement

#132 - MII Text to Image Task Issue

Issue - State: open - Opened by gvijqb about 2 years ago - 3 comments

#131 - Question about a custom MII task

Issue - State: open - Opened by pablogranolabar about 2 years ago - 3 comments

#130 - Query DeepSpeed MII with FastAPI and GRPC crashing

Issue - State: closed - Opened by mallorbc about 2 years ago - 7 comments

#129 - Add max_tokens option to mii_config

Pull Request - State: closed - Opened by mallorbc about 2 years ago - 11 comments

#128 - Add release tag 0.05

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

#127 - Some generate parameters do not work for query

Issue - State: open - Opened by mallorbc about 2 years ago - 3 comments

#126 - Add distilbert support

Pull Request - State: closed - Opened by jeffra about 2 years ago

#125 - Change number of max_tokens from 1024

Issue - State: open - Opened by mallorbc about 2 years ago - 1 comment

#124 - Support text2text-generation pipelines and T5 models in particular

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

#123 - RuntimeError: server crashed for some reason, unable to proceed

Issue - State: closed - Opened by underlines about 2 years ago - 4 comments

#122 - Tests use transformers cache for model storage

Pull Request - State: closed - Opened by mrwyattii about 2 years ago

#121 - Free torch cache after loading model

Pull Request - State: closed - Opened by mrwyattii about 2 years ago

#120 - Is model split in OPT TP mode?

Issue - State: open - Opened by larry-fuy about 2 years ago - 2 comments

#119 - Support for Albert and Swin/ViT

Issue - State: open - Opened by larry-fuy about 2 years ago - 1 comment

#118 - Stable Diffusion | Multi-Pipeline Support (i.e. img2img)

Issue - State: open - Opened by ghunkins about 2 years ago - 1 comment

#117 - Memory issue when loading OPT

Issue - State: closed - Opened by larry-fuy about 2 years ago - 3 comments

#116 - Unable to run mii-sd.py for txt2img benchmark

Issue - State: closed - Opened by aponte411 about 2 years ago - 7 comments

#115 - AML deployment error due to missing az cli arguments

Issue - State: open - Opened by aponte411 about 2 years ago - 5 comments

#114 - Passing `stopping_criteria` to DeepSpeed MII

Issue - State: closed - Opened by tokestermw about 2 years ago - 2 comments

#113 - Add xlm-roberta

Pull Request - State: closed - Opened by jeffra about 2 years ago

#112 - 'DSUNet' object has no attribute 'config'

Issue - State: closed - Opened by entangledloops about 2 years ago - 3 comments

#109 - Stop Sequence

Issue - State: open - Opened by volkerha about 2 years ago - 4 comments

#108 - Use new DS-Inference config API

Pull Request - State: closed - Opened by mrwyattii about 2 years ago - 1 comment

#107 - Add Bloom model config checks for HF provider

Pull Request - State: closed - Opened by mrwyattii about 2 years ago

#106 - Support for FLAN-T5

Issue - State: open - Opened by jihan-yin about 2 years ago - 3 comments

#105 - Improvements to VRAM usage when loading HF models

Pull Request - State: closed - Opened by mrwyattii about 2 years ago

#104 - Add community tutorial

Pull Request - State: closed - Opened by jeffra over 2 years ago

#103 - OOM Error when deploying BLOOM-3B on 16GB GPU via MII

Issue - State: closed - Opened by marshmellow77 over 2 years ago - 6 comments

#102 - No text is shown when using MII in fp32 and greedy search

Issue - State: closed - Opened by marshmellow77 over 2 years ago - 2 comments

#101 - Error when trying MII with dtype=fp32 with sampling

Issue - State: closed - Opened by marshmellow77 over 2 years ago - 2 comments

#100 - RuntimeError: This event loop is already running

Issue - State: open - Opened by marshmellow77 over 2 years ago - 3 comments
Labels: enhancement

#99 - CUDA OOM when loading large models

Issue - State: closed - Opened by Tianwei-She over 2 years ago - 6 comments

#98 - feat(config): enable port dynamique assignment

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

#97 - Multi-gpu inference, the query gets stuck when using my own provider

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

#96 - enable cuda graph for diffusers models

Pull Request - State: closed - Opened by jeffra over 2 years ago

#94 - Question : How to query a remote DeepSpeed server?

Issue - State: open - Opened by Thytu over 2 years ago - 4 comments

#93 - DS-MII License query

Issue - State: closed - Opened by mayank31398 over 2 years ago - 3 comments

#92 - Samyam sd bench

Pull Request - State: closed - Opened by samyam over 2 years ago

#91 - Deactivate quantization?

Issue - State: closed - Opened by remi86 over 2 years ago - 1 comment

#90 - SD tutorial and benchmarks

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

#89 - Can DeepSpeed-MII also support models/libraries like YOLO / Detectron2 / FCOS / VitDet?

Issue - State: open - Opened by kartikeyporwal over 2 years ago - 1 comment
Labels: enhancement

#88 - Fix aio config dict in ZeRO text-gen example

Pull Request - State: closed - Opened by mrwyattii over 2 years ago

#87 - Errors running Zero-Inference text generation example

Issue - State: closed - Opened by joaopcm1996 over 2 years ago - 3 comments

#86 - Support for Salesforce/CodeGen

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

#85 - Socket timeouts in MII

Issue - State: closed - Opened by mayank31398 over 2 years ago - 4 comments

#84 - Error in inf/nan tensors

Issue - State: closed - Opened by mayank31398 over 2 years ago - 3 comments

#83 - feature request : Docker image for deepspeed-mii

Issue - State: open - Opened by Thytu over 2 years ago - 3 comments

#79 - refactor(example): conversational-query

Pull Request - State: closed - Opened by Thytu over 2 years ago - 6 comments

#75 - "error: cuda_runtime_api.h: No such file or directory"

Issue - State: closed - Opened by auwsom over 2 years ago - 15 comments

#73 - update txt2img example to sd 1.5

Pull Request - State: closed - Opened by lalalune over 2 years ago

#71 - OPT in TP or PP mode

Issue - State: closed - Opened by volkerha over 2 years ago - 7 comments
Labels: enhancement

#68 - Fix examples with wrong param names

Pull Request - State: closed - Opened by mrwyattii over 2 years ago

#55 - Enable generation of scripts and configs for AML deployments

Pull Request - State: closed - Opened by mrwyattii over 2 years ago

#49 - FileNotFoundError: [Errno 2] No such file or directory: 'deepspeed'

Issue - State: closed - Opened by C00reNUT over 2 years ago - 3 comments

#47 - Support multiple nodes deployment?

Issue - State: open - Opened by pohunghuang-nctu over 2 years ago - 3 comments
Labels: enhancement

#35 - Add support for HuggingFace GPT-NeoX implementation

Pull Request - State: closed - Opened by mrwyattii over 2 years ago

#31 - Add unit tests for local deployment

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

#23 - Enable zero inference as a separte code path from ds_optimize

Pull Request - State: closed - Opened by cli99 almost 3 years ago

#16 - Add huggingface models and tasks that use gpt2 and robert model type

Pull Request - State: closed - Opened by cli99 almost 3 years ago - 7 comments

#15 - Samyamr/aks examples

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

#12 - Samyamr/add features

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

#11 - Samyamr/model registration

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