Fix the failing Intel MLPerf inference Github actions

[anandhu-eng]

---

Migrated from mlcommons/mlperf-automations_archived#8
Originally created by @arjunsuresh on Thu, 14 Nov 2024 15:17:14 GMT

---

Intel MLPerf inference runs are failing for R50 and bert as shown here

[anandhu-eng]

---

Migrated from mlcommons/mlperf-automations_archived#8 (comment)
Originally created by @arjunsuresh on Thu, 14 Nov 2024 17:23:33 GMT

---

For intel implementation we need to use v4.0 version of the loadgen until we remove the --mlperf.conf argument from the implementation.