tag:github.com,2008:https://github.com/pytorch/serve/releases 2024-09-24T16:41:17Z tag:github.com,2008:Repository/212488700/v0.12.0 2024-09-30T22:46:39Z agunapal tag:github.com,2008:Repository/212488700/v0.11.1 2024-07-23T16:55:51Z <p>Bug fix for kserve build issue and fixing nightly tests (<a class="issue-link js-issue-link" href="https://github.com/pytorch/serve/pull/3251">#3251</a>)</p> <p>* Testing kserve build</p> <p>* Update kserve_cpu_tests.yml</p> agunapal tag:github.com,2008:Repository/212488700/v0.11.0 2024-05-17T16:38:42Z <p>IPEX LLM serving example (<a class="issue-link js-issue-link" href="https://github.com/pytorch/serve/pull/3068">#3068</a>)</p> <p>* adding the files for ipex int8 serving of llms</p> <p>* Update README.md</p> <p>Fixed some markdowns</p> <p>* Fix handler name</p> <p>* Adding default PyTorch support</p> <p>* Fixing some issues with handler, added test to verify smooth-quant</p> <p>* adding auto_mixed_precision flag to config</p> <p>* Removing min_new_tokens from generation config</p> <p>* fix lint</p> <p>* lint</p> <p>* lint</p> <p>* Fixing unit tests with different model that doesn't require license</p> <p>* Fix lint error</p> <p>* Fix lint error in test</p> <p>* Adding requirements.txt</p> <p>* adding datasets to the requirements</p> <p>* upgrading the ipex version to 2.3.0 to match that of pytorch</p> <p>* Skipping ipex llm tests if accelerate is not present</p> <p>---------</p> <p>Co-authored-by: Ubuntu &lt;ubuntu@ip-172-31-51-123.us-west-2.compute.internal&gt; <br />Co-authored-by: lxning &lt;23464292+lxning@users.noreply.github.com&gt; <br />Co-authored-by: lxning &lt;lninga@amazon.com&gt; <br />Co-authored-by: Matthias Reso &lt;13337103+mreso@users.noreply.github.com&gt;</p> lxning tag:github.com,2008:Repository/212488700/v0.10.0 2024-03-15T19:05:27Z lxning tag:github.com,2008:Repository/212488700/v0.9.0 2023-10-17T19:14:02Z lxning tag:github.com,2008:Repository/212488700/v0.8.2 2023-08-28T23:52:56Z <p>llama2 70b chat accelerate example (<a class="issue-link js-issue-link" href="https://github.com/pytorch/serve/pull/2494">#2494</a>)</p> <p>* llam2 accelerate example</p> <p>* add readme</p> <p>* fmt</p> <p>* fixing the padding and prompt</p> <p>* update steps</p> <p>* Updated readme with more details</p> <p>* changed to inheriting from basehandler</p> <p>* add model_path</p> <p>* change to int8</p> <p>* add download cmd</p> <p>* update download path</p> <p>* minor edit for model_path</p> <p>---------</p> <p>Co-authored-by: Geeta Chauhan &lt;4461127+chauhang@users.noreply.github.com&gt; <br />Co-authored-by: Mark Saroufim &lt;marksaroufim@fb.com&gt; <br />Co-authored-by: Ankith Gunapal &lt;agunapal@ischool.Berkeley.edu&gt; <br />Co-authored-by: Hamid Shojanazeri &lt;hamid.nazeri2010@gmail.com&gt;</p> lxning tag:github.com,2008:Repository/212488700/v0.8.1 2023-06-20T19:48:59Z <p>Revert "Update Telemetry env variable. (<a class="issue-link js-issue-link" href="https://github.com/pytorch/serve/pull/2356">#2356</a>)" (<a class="issue-link js-issue-link" href="https://github.com/pytorch/serve/pull/2413">#2413</a>)</p> <p>This reverts commit <a class="commit-link" href="https://github.com/pytorch/serve/commit/727044754523e80ea991adbbd8e15634842a8b2c"><tt>7270447</tt></a>.</p> <p>Co-authored-by: lxning &lt;23464292+lxning@users.noreply.github.com&gt; <br />Co-authored-by: Ankith Gunapal &lt;agunapal@ischool.Berkeley.edu&gt;</p> lxning tag:github.com,2008:Repository/212488700/v0.8.0 2023-05-17T22:46:36Z <p>Issues/fix docker dependencies (<a class="issue-link js-issue-link" href="https://github.com/pytorch/serve/pull/2340">#2340</a>)</p> <p>* re-run kserve build because of runner issue</p> <p>* Fixes missing dependencies</p> <p>* added depth 1 to reduce clone size by 50 MB</p> lxning tag:github.com,2008:Repository/212488700/v0.7.1 2023-02-09T00:00:52Z lxning tag:github.com,2008:Repository/212488700/v0.7.0 2023-07-18T21:36:21Z <p>Fix docker build script to build with CUDA 11.7 (<a class="issue-link js-issue-link" href="https://github.com/pytorch/serve/pull/2032">#2032</a>)</p> <p>* updated docker to build with CUDA 11.7 as default</p> <p>* temp workflow for building and pushing docker images</p> <p>* reverting tmp workflow</p> lxning