Compare runs
cohort sha256:5230a49500e020a3… · source-native profile
ranking-v1
no ranks: not comparable
Comparison identity
operation
liger-fused-linear-kto-loss
workload
seq = 512 · batch = 8 · vocab = 128256 · hidden = 1024 · bf16 · bc08758b
protocol
Liger-Kernel benchmark scripts · 5b02ee9a
environment
NVIDIA H100 · 2750ae38
correctness policy
52ff5e65
metric
latency median (ns)
Context · 10 fields›
architecture
sm_90
CUDA
unknown
driver
unknown
framework
unknown
samples
unknown
evidence
reported
license
BSD-2-Clause
source
available
status
passed
observed
2025-03-03
No winner: these runs didn't measure the same thing. The rows above show what differs. Why comparable?
Add run ›
Export ›
Markdown CSV JSON
Liger · sha256:dfb8fd11a76e8… · observed 2025-03-03