dynamo: Count number of opcodes processes (#147149)

Summary:
This gives us a decent proxy for how big of a graph we functionally had to parse.

Note that this is a cummulative counter. If people feel strongly, I can either write into the dynamo_timed datasets with metrics contexts, or clear the counters / write a counter per frame id as well.

X-link: pytorch/pytorch#147149
Approved by: https://github.com/jansel

Reviewed By: BoyuanFeng

Differential Revision: D70918753

fbshipit-source-id: 7f3e68ebe2e01b6c93ac075d62069cad6c709308

Author: c00w

userbenchmark/dynamo/dynamobench/_dynamo/utils.py

@@ -1230,6 +1230,7 @@ class CompilationMetrics:
     recompile_reason: Optional[str] = None
     num_graph_breaks: Optional[int] = None
     triton_kernel_compile_times_us: Optional[str] = None
+    ir_count: Optional[int] = None
 
     @classmethod
     def create(cls, metrics: dict[str, Any]):