model: Add support for Nanbeige4.2 (#25994)
* support nanbeige4.2 model * fix * fix flake8 Lint check * fix loop bound check and drop redundant head_dim --------- Co-authored-by: root <lizongqiang@kanzhun.com>
This commit is contained in:
@@ -148,6 +148,7 @@ enum llm_arch {
|
||||
LLM_ARCH_EAGLE3,
|
||||
LLM_ARCH_MINIMAX_M3,
|
||||
LLM_ARCH_DFLASH,
|
||||
LLM_ARCH_NANBEIGE,
|
||||
LLM_ARCH_UNKNOWN,
|
||||
};
|
||||
|
||||
@@ -226,6 +227,8 @@ enum llm_kv {
|
||||
LLM_KV_TOKEN_SHIFT_COUNT,
|
||||
LLM_KV_INTERLEAVE_MOE_LAYER_STEP,
|
||||
LLM_KV_FULL_ATTENTION_INTERVAL,
|
||||
LLM_KV_NUM_LOOPS,
|
||||
LLM_KV_SKIP_LOOP_FINAL_NORM,
|
||||
|
||||
LLM_KV_ATTENTION_HEAD_COUNT,
|
||||
LLM_KV_ATTENTION_HEAD_COUNT_KV,
|
||||
|
||||
Reference in New Issue
Block a user