Ecosyste.ms: Issues

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

GitHub / ELS-RD/kernl issues and pull requests

#265 - fix: update dep + fix

Pull Request - State: closed - Opened by pommedeterresautee almost 2 years ago
Labels: bug, dependencies, fix

#265 - fix: update dep + fix

Pull Request - State: closed - Opened by pommedeterresautee almost 2 years ago
Labels: bug, dependencies, fix

#263 - bug: when executed in a specific order, tests crash

Issue - State: closed - Opened by pommedeterresautee almost 2 years ago
Labels: bug

#263 - bug: when executed in a specific order, tests crash

Issue - State: closed - Opened by pommedeterresautee almost 2 years ago
Labels: bug

#262 - feature: replace Whisper script by a notebook

Issue - State: closed - Opened by pommedeterresautee almost 2 years ago
Labels: feature

#262 - feature: replace Whisper script by a notebook

Issue - State: closed - Opened by pommedeterresautee almost 2 years ago
Labels: feature

#258 - [M4M Insider] Upgrade version (kernl + blog) to 4.30.2.

Issue - State: closed - Opened by white-gorilla almost 2 years ago - 2 comments

#258 - [M4M Insider] Upgrade version (kernl + blog) to 4.30.2.

Issue - State: closed - Opened by white-gorilla almost 2 years ago - 2 comments

#253 - Where can I install the software (on which hosting I can try it)

Issue - State: closed - Opened by Oxi84 almost 2 years ago - 5 comments
Labels: documentation, question, stale

#253 - Where can I install the software (on which hosting I can try it)

Issue - State: closed - Opened by Oxi84 almost 2 years ago - 5 comments
Labels: documentation, question, stale

#247 - [M4M | BLOG] Updated kernl/docs CI to support M4M Insiders.

Issue - State: closed - Opened by white-gorilla almost 2 years ago
Labels: documentation

#247 - [M4M | BLOG] Updated kernl/docs CI to support M4M Insiders.

Issue - State: closed - Opened by white-gorilla almost 2 years ago
Labels: documentation

#245 - ci: update m4mkdocs with insiders version.

Pull Request - State: closed - Opened by white-gorilla almost 2 years ago

#245 - ci: update m4mkdocs with insiders version.

Pull Request - State: closed - Opened by white-gorilla almost 2 years ago

#242 - Accelerate warmup IRL?

Issue - State: closed - Opened by JaheimLee almost 2 years ago - 4 comments
Labels: question, stale

#242 - Accelerate warmup IRL?

Issue - State: closed - Opened by JaheimLee almost 2 years ago - 4 comments
Labels: question, stale

#233 - feat: add linear layer train support

Pull Request - State: open - Opened by ayoub-louati almost 2 years ago - 1 comment
Labels: feature

#232 - Torchdynamo + Inductor faster than kernl in t5 e2e example?

Issue - State: closed - Opened by wangjunhaoumich almost 2 years ago - 5 comments
Labels: question, stale

#232 - Torchdynamo + Inductor faster than kernl in t5 e2e example?

Issue - State: closed - Opened by wangjunhaoumich almost 2 years ago - 5 comments
Labels: question, stale

#228 - linear layer backward

Issue - State: closed - Opened by ayoub-louati almost 2 years ago - 1 comment
Labels: enhancement, feature

#228 - linear layer backward

Issue - State: closed - Opened by ayoub-louati almost 2 years ago - 1 comment
Labels: enhancement, feature

#227 - Possible mistake in linear implementation

Issue - State: closed - Opened by gaetansnl almost 2 years ago
Labels: bug

#227 - Possible mistake in linear implementation

Issue - State: closed - Opened by gaetansnl almost 2 years ago
Labels: bug

#226 - Blank output in the Inference while using a customize trained T5 model

Issue - State: open - Opened by gaurav21s almost 2 years ago - 10 comments
Labels: bug

#226 - Blank output in the Inference while using a customize trained T5 model

Issue - State: open - Opened by gaurav21s almost 2 years ago - 10 comments
Labels: bug

#224 - Is there a way to export a warmed up graph so that it is ready to go from cold boot?

Issue - State: closed - Opened by amazingvince almost 2 years ago - 4 comments
Labels: question, stale

#224 - Is there a way to export a warmed up graph so that it is ready to go from cold boot?

Issue - State: closed - Opened by amazingvince almost 2 years ago - 4 comments
Labels: question, stale

#223 - Update setup.py

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

#223 - Update setup.py

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

#213 - optimized whisper.generate() hangs forever, including in tests

Issue - State: closed - Opened by TheExGenesis almost 2 years ago - 30 comments
Labels: question

#213 - optimized whisper.generate() hangs forever, including in tests

Issue - State: closed - Opened by TheExGenesis almost 2 years ago - 30 comments
Labels: question

#204 - feat: update optimize call

Pull Request - State: closed - Opened by jonathlela almost 2 years ago - 1 comment
Labels: feature

#204 - feat: update optimize call

Pull Request - State: closed - Opened by jonathlela almost 2 years ago - 1 comment
Labels: feature

#200 - Backward Kernel Implementation

Issue - State: closed - Opened by conceptofmind almost 2 years ago - 3 comments
Labels: triton

#197 - Comparison with pytorch 2.0

Issue - State: closed - Opened by acostin1 almost 2 years ago - 11 comments
Labels: question

#197 - Comparison with pytorch 2.0

Issue - State: closed - Opened by acostin1 almost 2 years ago - 11 comments
Labels: question

#196 - feat: new debugger

Pull Request - State: closed - Opened by gaetansnl almost 2 years ago
Labels: feature

#195 - Debugger 1: Refactoring and improve reliability and DX

Issue - State: closed - Opened by gaetansnl almost 2 years ago
Labels: enhancement, feature

#195 - Debugger 1: Refactoring and improve reliability and DX

Issue - State: closed - Opened by gaetansnl almost 2 years ago
Labels: enhancement, feature

#194 - CUDA: Error- illegal address when run bert e2e.ipynb on A30 gpu

Issue - State: closed - Opened by chenxu001 almost 2 years ago - 4 comments
Labels: bug

#194 - CUDA: Error- illegal address when run bert e2e.ipynb on A30 gpu

Issue - State: closed - Opened by chenxu001 almost 2 years ago - 4 comments
Labels: bug

#188 - Question regarding larger T5 model support

Issue - State: closed - Opened by caffeinetoomuch almost 2 years ago - 14 comments
Labels: question, stale

#188 - Question regarding larger T5 model support

Issue - State: closed - Opened by caffeinetoomuch almost 2 years ago - 14 comments
Labels: question, stale

#187 - feat: add vector addition kernel example to test triton compile api

Pull Request - State: closed - Opened by ayoub-louati almost 2 years ago
Labels: feature

#181 - docs: how to support a new model tutorial

Pull Request - State: closed - Opened by jonathlela almost 2 years ago - 3 comments
Labels: feature

#181 - docs: how to support a new model tutorial

Pull Request - State: closed - Opened by jonathlela almost 2 years ago - 3 comments
Labels: feature

#179 - docs: new kernel design

Pull Request - State: open - Opened by pommedeterresautee almost 2 years ago

#179 - docs: new kernel design

Pull Request - State: open - Opened by pommedeterresautee almost 2 years ago

#178 - Add new kernel design documentation

Issue - State: closed - Opened by pommedeterresautee almost 2 years ago
Labels: documentation

#164 - feat: add bw pass on layernorm/rmsnorm

Pull Request - State: open - Opened by pommedeterresautee about 2 years ago - 3 comments
Labels: feature

#164 - feat: add bw pass on layernorm/rmsnorm

Pull Request - State: open - Opened by pommedeterresautee about 2 years ago - 3 comments
Labels: feature

#154 - Add tutorial "how to support a new model with kernl"

Issue - State: closed - Opened by jonathlela about 2 years ago

#154 - Add tutorial "how to support a new model with kernl"

Issue - State: closed - Opened by jonathlela about 2 years ago

#146 - GPT-J Internal Triton PTX codegen error:

Issue - State: closed - Opened by zelcookie about 2 years ago - 4 comments
Labels: question, model, stale

#146 - GPT-J Internal Triton PTX codegen error:

Issue - State: closed - Opened by zelcookie about 2 years ago - 4 comments
Labels: question, model, stale

#145 - why is model.generate repeated 10 times in t5 generate example?

Issue - State: closed - Opened by GongCQ about 2 years ago - 13 comments
Labels: question, stale

#141 - Would it be possible to use kernl to speed up Stable Diffusion?

Issue - State: closed - Opened by 0xdevalias about 2 years ago - 3 comments
Labels: enhancement, feature

#141 - Would it be possible to use kernl to speed up Stable Diffusion?

Issue - State: closed - Opened by 0xdevalias about 2 years ago - 3 comments
Labels: enhancement, feature

#135 - docs: tutorial table added for easy access of examples.

Pull Request - State: closed - Opened by ashishpatel26 about 2 years ago - 1 comment
Labels: documentation

#135 - docs: tutorial table added for easy access of examples.

Pull Request - State: closed - Opened by ashishpatel26 about 2 years ago - 1 comment
Labels: documentation

#125 - install problem

Issue - State: closed - Opened by DeqianBai about 2 years ago - 7 comments
Labels: bug, question, stale

#125 - install problem

Issue - State: closed - Opened by DeqianBai about 2 years ago - 7 comments
Labels: bug, question, stale

#106 - Add new annotation based on triton.compile to replace the use of jit

Issue - State: open - Opened by pommedeterresautee about 2 years ago - 3 comments
Labels: performance

#106 - Add new annotation based on triton.compile to replace the use of jit

Issue - State: open - Opened by pommedeterresautee about 2 years ago - 3 comments
Labels: performance

#84 - Flash attention crashs on turing hardware.

Issue - State: closed - Opened by jonathlela about 2 years ago - 5 comments

#84 - Flash attention crashs on turing hardware.

Issue - State: closed - Opened by jonathlela about 2 years ago - 5 comments

#76 - feat: test filtering command line

Pull Request - State: closed - Opened by pommedeterresautee about 2 years ago - 1 comment
Labels: benchmark

#76 - feat: test filtering command line

Pull Request - State: closed - Opened by pommedeterresautee about 2 years ago - 1 comment
Labels: benchmark

#68 - T5 mask generation

Issue - State: closed - Opened by pommedeterresautee about 2 years ago
Labels: model, performance