Project

General

Profile

Actions
Copy link

Task #7

open

Support Automotive benchmark in MLCFlow automation

Added by Anandhu Sooraj about 2 months ago. Updated about 2 months ago.

Status:
New
Priority:
High
Assignee:
Anandhu Sooraj
Start date:
03/08/2026
Due date:
03/16/2026 (45 days late)
% Done:

27%

Estimated time:
(Total: 4:00 h)

Description

We need to support the existing as well as new implementation in automation

Subtasks 11 (11 open — 0 closed)

Bug #8: Follow up on bug related to loadgenNewAnandhu Sooraj03/08/202603/09/2026

Actions
Task #9: Support docker GPU run for llama3.1-8b automotive benchmarkNewAnandhu Sooraj03/08/202603/09/2026

Actions
Task #10: Get valid run results for llama3.1-8b in GPUNewAnandhu Sooraj03/08/202603/10/2026

Actions
Task #11: Test submission checker based on the results from llama3.1-8b valid runNewAnandhu Sooraj03/08/202603/11/2026

Actions
Task #12: Cleanup the current automotive meta filesNewAnandhu Sooraj03/08/202603/11/2026

Actions
Task #13: Integrate and test UNIad benchmark on GPUNewAnandhu Sooraj03/08/202603/13/2026

Actions
Task #14: Download the Nuscenes dataset to MLC2 serverNewAnandhu Sooraj03/08/202603/10/2026

Actions
Task #15: Get valid results from UNIad runs for testing submission checkerNewAnandhu Sooraj03/08/202603/14/2026

Actions
Task #16: Test submission checker for both UNIad as well as LLama3NewAnandhu Sooraj03/08/202603/14/2026

Actions
Task #17: Test UI with the results taken from both Llama2 as well as UNIadNewAnandhu Sooraj03/08/202603/16/2026

Actions
Task #37: Fix docker run command to correctly map in GPUs for podman containersNewAnandhu Sooraj03/09/202603/09/2026

Actions
Actions
Copy link

Also available in: Atom PDF