llama-4-scout

best known per operation on the selected GPU

7 operations · 3 families · 0 eligible runs

Coverage gaps

Operation

Family

Status

GQA paged prefill causal h5 kv1 d128 ps64

gqa-paged-attention

0 eligible runs · gap

Trtllm fp4 block scale MoE topk1 e16 h5120 i8192

moe

0 eligible runs · gap

Trtllm fp4 block scale routed MoE topk1 e16 h5120 i8192

moe

0 eligible runs · gap

Trtllm FP8 per tensor scale MoE topk1 e16 h5120 i8192

moe

0 eligible runs · gap

Top k sampling from probs v202048

sampling

0 eligible runs · gap

Top k top p sampling from probs v202048

sampling

0 eligible runs · gap

Top p sampling from probs v202048

sampling

0 eligible runs · gap

No eligible evidence yet for these operations. Challenges →

Related tags: llama-4-scout-17b-16e

No source imports yet.How comparison works JSON