include convert latency in bench_append_paged_kv_cache

[abcdabcd987]

model: l1b      seqlens: [1, 1, 1, 1, 1, 1, 1, 1]                 convert: 45us 1layer:  7us 16layers: 151us throughput:    4.936GB/s
model: l1b      seqlens: [4993, 1, 1, 1, 1, 1, 1, 1]              convert: 42us 1layer: 14us 16layers: 271us throughput: 1434.769GB/s
model: l1b      seqlens: [5000]                                   convert: 44us 1layer: 14us 16layers: 272us throughput: 1438.581GB/s
model: l1b      seqlens: [625, 625, 625, 625, 625, 625, 625, 625] convert: 46us 1layer: 14us 16layers: 274us throughput: 1440.357GB/s
---
model: l3b      seqlens: [1, 1, 1, 1, 1, 1, 1, 1]                 convert: 42us 1layer:  7us 28layers: 226us throughput:    9.946GB/s
model: l3b      seqlens: [4993, 1, 1, 1, 1, 1, 1, 1]              convert: 43us 1layer: 22us 28layers: 647us throughput: 1896.687GB/s
model: l3b      seqlens: [5000]                                   convert: 42us 1layer: 22us 28layers: 646us throughput: 1898.796GB/s
model: l3b      seqlens: [625, 625, 625, 625, 625, 625, 625, 625] convert: 41us 1layer: 22us 28layers: 648us throughput: 1890.115GB/s
---
model: l8b      seqlens: [1, 1, 1, 1, 1, 1, 1, 1]                 convert: 41us 1layer:  7us 32layers: 252us throughput:    9.940GB/s
model: l8b      seqlens: [4993, 1, 1, 1, 1, 1, 1, 1]              convert: 42us 1layer: 21us 32layers: 730us throughput: 1905.826GB/s
model: l8b      seqlens: [5000]                                   convert: 41us 1layer: 22us 32layers: 729us throughput: 1903.697GB/s
model: l8b      seqlens: [625, 625, 625, 625, 625, 625, 625, 625] convert: 47us 1layer: 22us 32layers: 737us throughput: 1899.630GB/s
---
model: l70b-tp8 seqlens: [1, 1, 1, 1, 1, 1, 1, 1]                 convert: 42us 1layer:  6us 80layers: 552us throughput:    1.283GB/s
model: l70b-tp8 seqlens: [4993, 1, 1, 1, 1, 1, 1, 1]              convert: 41us 1layer:  9us 80layers: 800us throughput:  539.484GB/s
model: l70b-tp8 seqlens: [5000]                                   convert: 41us 1layer:  9us 80layers: 788us throughput:  548.648GB/s
model: l70b-tp8 seqlens: [625, 625, 625, 625, 625, 625, 625, 625] convert: 41us 1layer: 10us 80layers: 803us throughput:  537.731GB/s