[VLM] Merged multi-modal processor for InternVL-based models (#12553)
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk> Signed-off-by: Isotr0py <2037008807@qq.com> Co-authored-by: Isotr0py <2037008807@qq.com>
This commit is contained in:
@@ -151,7 +151,8 @@ class MultiModalProfiler(Generic[_I]):
|
||||
mm_counts = self.get_mm_limits()
|
||||
|
||||
info = self.processing_info
|
||||
mm_max_tokens_per_item = info.get_mm_max_tokens_per_item(seq_len)
|
||||
mm_max_tokens_per_item = info.get_mm_max_tokens_per_item(
|
||||
seq_len, mm_counts)
|
||||
|
||||
if mm_counts.keys() != mm_max_tokens_per_item.keys():
|
||||
raise AssertionError(
|
||||
|
||||
Reference in New Issue
Block a user