Fix correctness issue for triton decoding kernel #3065
Set up job
6s
6s
Error:
This step has been truncated due to its large size. Download the full logs from the menu
once the workflow run has completed.
Checkout code
6s
6s
Error:
This step has been truncated due to its large size. Download the full logs from the menu
once the workflow run has completed.
Install dependencies
48s
48s
Error:
This step has been truncated due to its large size. Download the full logs from the menu
once the workflow run has completed.
Benchmark offline throughput (w/o RadixAttention)
3m 32s
3m 32s
Error:
This step has been truncated due to its large size. Download the full logs from the menu
once the workflow run has completed.
Benchmark offline throughput (w/ Triton)
3m 18s
3m 18s
Error:
This step has been truncated due to its large size. Download the full logs from the menu
once the workflow run has completed.
Benchmark offline throughput (w/ FP8)
2m 53s
2m 53s
Error:
This step has been truncated due to its large size. Download the full logs from the menu
once the workflow run has completed.
Post Checkout code
0s
0s
Error:
This step has been truncated due to its large size. Download the full logs from the menu
once the workflow run has completed.
Complete job
0s
0s
Error:
This step has been truncated due to its large size. Download the full logs from the menu
once the workflow run has completed.
Loading