issues Search Results · repo:microsoft/Olive language:Python
Filter by
272 results
(85 ms)272 results
inmicrosoft/Olive (press backspace or delete to remove)Proposal Summary
Olive provides many CLI such as olive quantize, olive auto-opt.
This is a feature request to add new CLI olive optimize. It will make following command line options available
• model_name_or ...
enhancement
devang-ml
- Opened 7 days ago
- #1995
Describe the bug I encountered an issue while converting phi3.5-mini-instruct and phi-4--mini-instruct using olive run
-- config qnn.config. json
To Reproduce olive run -- config qnn.config. json
Olive ...
qihui-liu
- 24
- Opened 18 days ago
- #1962
Describe the bug Based on this sample Phi3.5 mini instruct recipe:
1) https://github.com/microsoft/Olive/tree/main/examples/phi3_5
2) https://github.com/microsoft/Olive/tree/main/examples/llama3
3) ...
chapsiru
- Opened 26 days ago
- #1946
Hi there! I am trying to quantize and run a Llama 3.2 1B Instruct model on the AMD RYZEN AI MAX+ 395 on the integrated
GPU through DirectML.
I created a python venv, and followed these steps:
pip install ...
Kohltrane
- 1
- Opened 27 days ago
- #1942
Describe the bug A clear and concise description of what the bug is.
Cache is not calculated correctly for some models.
i.e. two different models will use the same cache which will lead to wrong results ...
xieofxie
- 1
- Opened on Jun 26
- #1936
OnnxIODataTypeConverter pass converts inputs and outputs types. It d be a good idea to have a surgery as part of the
GraphSurgeries pass to do the same and remove OnnxIODataTypeConverter pass.
- Opened on Jun 23
- #1928
On Windows 11 pro 24h2, using uv with python 3.12. Deps:
dependencies = [
olive-ai[auto-opt] =0.9.1 ,
onnxruntime-genai =0.8.2 ,
transformers =4.52.4 ,
]
Running the command from the quickstart ...
doless-dev
- 3
- Opened on Jun 15
- #1916
Update olive/passes/onnx/hqq_quantization.py to use onnx_ir for graph manipulation. You can use
olive/passes/onnx/rtn_quantization.py as a reference. Be sure to create sufficient tests.
justinchuby
- Opened on Jun 14
- #1914
Describe the bug Olive EPContextBinaryGenerator pass fails when creating the AOT Compilation Python Environment with
latest Olive (main).
To Reproduce Follow Phi 3.5 mini example:
https://github.com/microsoft/Olive/tree/main/examples/phi3_5#aot-compilation-python-environment-setup ...
jake-leland-dell
- 4
- Opened on Jun 13
- #1913
Describe the bug Olive ModelBuilder pass fails when installing packages according to readme. The pip install
onnxruntime-gpu =1.21.0 onnxruntime-genai-cuda =0.6.0 instruction will install onnxruntime-gpu==1.22.0 ...
- Opened on Jun 13
- #1912
Learn how you can use GitHub Issues to plan and track your work.
Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub IssuesProTip!
Press the /
key to activate the search input again and adjust your query.Learn how you can use GitHub Issues to plan and track your work.
Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub IssuesProTip!
Restrict your search to the title by using the in:title qualifier.