test: port session_latency and streaming_session_swa to NPU - #1017
test: port session_latency and streaming_session_swa to NPU#1017hhhh1252023 wants to merge 13 commits into
Conversation
CLA Signature Passhhhh1252023, thanks for your pull request. All authors of the commits have signed the CLA. 👍 |
CLA Signature Guide@hhhh1252023 , thanks for your pull request. The following commit(s) are not associated with a signed Contributor License Agreement (CLA).
To sign CLA, click here. To check if your email is configured correctly, refer to the FAQs. Once you've signed the CLA or updating your email, please comment |
CLA Signature Passhhhh1252023, thanks for your pull request. All authors of the commits have signed the CLA. 👍 |
CLA Signature Passhhhh1252023, thanks for your pull request. All authors of the commits have signed the CLA. 👍 |
CLA Signature Passhhhh1252023, thanks for your pull request. All authors of the commits have signed the CLA. 👍 |
…treaming_session_swa
CLA Signature Passhhhh1252023, thanks for your pull request. All authors of the commits have signed the CLA. 👍 |
CLA Signature Guide@hhhh1252023 , thanks for your pull request. The following commit(s) are not associated with a signed Contributor License Agreement (CLA).
To sign CLA, click here. To check if your email is configured correctly, refer to the FAQs. Once you've signed the CLA or updating your email, please comment |
CLA Signature Guide@hhhh1252023 , thanks for your pull request. The following commit(s) are not associated with a signed Contributor License Agreement (CLA).
To sign CLA, click here. To check if your email is configured correctly, refer to the FAQs. Once you've signed the CLA or updating your email, please comment |
CLA Signature Guide@hhhh1252023 , thanks for your pull request. The following commit(s) are not associated with a signed Contributor License Agreement (CLA).
To sign CLA, click here. To check if your email is configured correctly, refer to the FAQs. Once you've signed the CLA or updating your email, please comment |
CLA Signature Guide@hhhh1252023 , thanks for your pull request. The following commit(s) are not associated with a signed Contributor License Agreement (CLA).
To sign CLA, click here. To check if your email is configured correctly, refer to the FAQs. Once you've signed the CLA or updating your email, please comment |
CLA Signature Guide@hhhh1252023 , thanks for your pull request. The following commit(s) are not associated with a signed Contributor License Agreement (CLA).
To sign CLA, click here. To check if your email is configured correctly, refer to the FAQs. Once you've signed the CLA or updating your email, please comment |
CLA Signature Guide@hhhh1252023 , thanks for your pull request. The following commit(s) are not associated with a signed Contributor License Agreement (CLA).
To sign CLA, click here. To check if your email is configured correctly, refer to the FAQs. Once you've signed the CLA or updating your email, please comment |
CLA Signature Guide@hhhh1252023 , thanks for your pull request. The following commit(s) are not associated with a signed Contributor License Agreement (CLA).
To sign CLA, click here. To check if your email is configured correctly, refer to the FAQs. Once you've signed the CLA or updating your email, please comment |
2 participants
Port two high-difficulty session test files to Ascend NPU. Files: test_npu_session_latency.py (benchmark streaming session inter-turn latency on gpt-oss-20b), test_npu_streaming_session_swa.py (streaming session on hybrid-SWA model with 4 test classes). NPU adaptations: model openai/gpt-oss-20b -> openai-mirror/gpt-oss-20b, --cuda-graph-backend-prefill=disabled -> --disable-cuda-graph, --page-size 256 -> 128, added --attention-backend ascend, added GPT_OSS_20B_WEIGHTS_PATH constant. CI triggered via single-test-npu.yml.
CI States
Latest PR Test (Base): ❌ Missing
run-cilabel -- add it to run CI tests.Latest PR Test (Extra): ❌ Blocked --
run-ciis required first.