[Fix/Inference] Fix GQA Triton and Support Llama3 (#5624)

* [fix] GQA calling of flash decoding triton

* fix kv cache alloc shape

* fix rotary triton - GQA

* fix sequence max length assigning

* Sequence max length logic

* fix scheduling and spec-dec

* skip without import error

* fix pytest - skip without ImportError

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

colossalai/inference/struct.py

@@ -117,6 +117,14 @@ class Sequence:
 
         return False
 
+    def revoke_finished_status(self) -> None:
+        """
+        Revoke the finished status of the sequence.
+        This is only used by speculative decoding for now.
+        """
+        if RequestStatus.is_finished(self.status):
+            self.status = RequestStatus.RUNNING
+
     def __hash__(self):
         return hash(self.request_id)