Paged attention support for vision models

[EricLBuehler]

@haixuanTao this is that last thing to do before we can merge OpenVLA!

[github-actions]

Code Metrics Report

  ===============================================================================
 Language            Files        Lines         Code     Comments       Blanks
===============================================================================
 C Header                2           35           28            0            7
 Dockerfile              1           34           25            0            9
 Happy                   1          442          369            0           73
 JSON                   11          102          101            0            1
 Python                 40         1562         1346           46          170
 TOML                   19          557          491           11           55
-------------------------------------------------------------------------------
 Jupyter Notebooks       2            0            0            0            0
 |- Markdown             2           77           32           31           14
 |- Python               2          196          169            1           26
 (Total)                            273          201           32           40
-------------------------------------------------------------------------------
 Markdown               23         1808            0         1365          443
 |- BASH                 5          101           98            0            3
 |- JSON                 1           12           12            0            0
 |- Python               5           92           82            0           10
 |- Rust                 6          407          364           19           24
 |- TOML                 2           75           63            0           12
 (Total)                           2495          619         1384          492
-------------------------------------------------------------------------------
 Rust                  169        53838        48875          952         4011
 |- Markdown            91          827           13          768           46
 (Total)                          54665        48888         1720         4057
===============================================================================
 Total                 269        58378        51235         2374         4769
===============================================================================