The 2nd PR for #4532. This PR supports loading FP8 kv-cache scaling factors from a FP8 checkpoint (with .kv_scale parameter). |
||
|---|---|---|
| .. | ||
| benchmark_aqlm.py | ||
| benchmark_marlin.py | ||
| benchmark_mixtral_moe.py | ||
| benchmark_paged_attention.py | ||
| benchmark_rope.py | ||
| benchmark_shapes.py | ||