transformers
3340ccbd - Fix gpt-oss router_indices in EP (#40545)

Commit
121 days ago
Fix gpt-oss router_indices in EP (#40545) * fix out shape Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * fix router indice Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * fix mod Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * fix masking Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * fix typo Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * fix typo Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * fix format Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * add safety cheking Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * fix checking Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * enable 1 expert per rank Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * fix skip Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * add ep plan in config Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * add update ep plan Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * fix typo Signed-off-by: jiqing-feng <jiqing.feng@intel.com> * rm ep_plan and add comments Signed-off-by: jiqing-feng <jiqing.feng@intel.com> --------- Signed-off-by: jiqing-feng <jiqing.feng@intel.com>
Author
Parents
Loading