-
Notifications
You must be signed in to change notification settings - Fork 4.3k
Issues: deepspeedai/DeepSpeed
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Author
Label
Projects
Milestones
Assignee
Sort
Issues list
[REQUEST] An option for SUM gradient allreduce instead of MEAN
enhancement
New feature or request
#7107
opened Mar 4, 2025 by
sfc-gh-lmerrick
[REQUEST] Proposal for Enhancing ChatGPT's Response Quality During Training
enhancement
New feature or request
#7097
opened Mar 1, 2025 by
sandyotic
LLama factory, quantized model and deepspeed compatibility
#7096
opened Mar 1, 2025 by
richardPang517
[BUG]Does MoQ mode deprecated in DeepSpeed? I run with MoQ config, but no quantization found in log
bug
Something isn't working
training
#7092
opened Feb 28, 2025 by
dujifeng
[BUG]error: the global scope has no "timespec_get" using ::timespec_get;
bug
Something isn't working
training
#7080
opened Feb 26, 2025 by
9050350
[BUG] Deepspeed does not update the model when using "Qwen/Qwen2.5-3B" and is fine with ""Qwen/Qwen2.5-1.%B""
bug
Something isn't working
training
#7077
opened Feb 25, 2025 by
MiladInk
[BUG?] zero++ init problem
bug
Something isn't working
training
#7066
opened Feb 21, 2025 by
cyr0930
Ascend 910B: attributeerror: 'deepspeedcpuadam' object has no attribute 'ds_opt_adam'
#7061
opened Feb 20, 2025 by
RyanOvO
[BUG] Question Regarding Weights After Reloading ZeroQuant Quantized W4A8 BERT Model
bug
Something isn't working
compression
#7060
opened Feb 20, 2025 by
RealJustinNi
[BUG] DS zero stage 1 or 2 communication uses reduce-scatter instead of All-reduce
bug
Something isn't working
training
#7059
opened Feb 20, 2025 by
Ind1x1
[REQUEST] Publish your Windows Wheels build workflow
enhancement
New feature or request
#7057
opened Feb 20, 2025 by
acidbubbles
[BUG] deepspeed zero2 training hangon and timeout after a fixed step
bug
Something isn't working
training
#7044
opened Feb 17, 2025 by
leeruibin
Getting requirements to build wheel: finished with status 'error'
windows
Questions or PRs relating to running DeepSpeed on Windows
#7043
opened Feb 17, 2025 by
Avroboros
[REQUEST] Runable solution of RTX 5090 GPU + Linux Driver version + Pytorch version + Deepspeed version for LLM finetuning?
enhancement
New feature or request
#7042
opened Feb 17, 2025 by
0781532
[REQUEST] activation checkpoint API should have parity with Pytorch, keywords arguments not supported
enhancement
New feature or request
#7038
opened Feb 15, 2025 by
AndreasMadsen
[REQUEST] Why is the column linear layer with all-gather not implemented in DeepSpeed Inference?
enhancement
New feature or request
#7037
opened Feb 14, 2025 by
zhangvia
Fix - Update DeepSpeed to be PEP517 compliant, update to Improvements to the build and testing systems.
install
Installation and package dependencies
pyproject.toml
build
#7031
opened Feb 13, 2025 by
loadams
[BUG] Something isn't working
training
import deepspeed
crashes on deepspeed==0.16.3
with triton==3.2.0
on CPU machine
bug
#7028
opened Feb 13, 2025 by
hongpeng-guo
[BUG]Issues with Running DeepSpeed Zero2 & Zero3 Not Taking Effect
bug
Something isn't working
training
#7026
opened Feb 12, 2025 by
fengdian8564
Previous Next
ProTip!
Exclude everything labeled
bug
with -label:bug.