Compare runs
cohort sha256:403a1fe25cdd8b16… · source-native profile
ranking-v1
no ranks: not comparable
Comparison identity
operation
liger-fused-linear-orpo-loss
workload
seq = 1024 · batch = 8 · vocab = 128256 · hidden = 4096 · bf16 · 280a97cf
protocol
Liger-Kernel benchmark scripts · a79eaf4d
environment
NVIDIA A100 · 90c9e276
correctness policy
52ff5e65
metric
latency median (ns)
Context · 10 fields›
architecture
sm_80
CUDA
unknown
driver
unknown
framework
unknown
samples
unknown
evidence
reported
license
Apache-2.0
source
available
status
passed
observed
2024-11-13
No winner: these runs didn't measure the same thing. The rows above show what differs. Why comparable?
Add run ›
Export ›
Markdown CSV JSON
Transformers · sha256:e66a35d1719b8… · observed 2024-11-13