Upload eval_standard_benchmarks.py with huggingface_hub

eval_standard_benchmarks.py

@@ -12,6 +12,8 @@ benchmarks (MMLU, HellaSwag, ARC, etc.) using lm-evaluation-harness.
 #     "peft>=0.7.0",
 #     "huggingface-hub>=0.20.0",
 #     "accelerate>=0.20.0",
+#     "protobuf>=3.20.0",
+#     "sentencepiece>=0.1.99",
 # ]
 # ///