[fp8] support gemini plugin (#5978)

* [fp8] refactor hook

* [fp8] support gemini plugin

* [example] add fp8 option for llama benchmark

colossalai/quantization/fp8_hook.py

@@ -0,0 +1,23 @@
+import torch.nn.functional as F
+
+from colossalai.quantization.fp8 import linear_fp8
+from colossalai.tensor.param_op_hook import ColoParamOpHook
+
+
+class FP8Hook(ColoParamOpHook):
+    def pre_forward(self, params) -> None:
+        pass
+
+    def post_forward(self, params) -> None:
+        pass
+
+    def pre_backward(self, params) -> None:
+        pass
+
+    def post_backward(self, params) -> None:
+        pass
+
+    def rewrite_op(self, func):
+        if func is F.linear:
+            return linear_fp8
+        return func