Enhancements/Fixes to HF Benchmark Runtime

[raikonenfnu]

HF Benchmarker is a module within SHARK that enable easy testing of HF models with ONNX, Torch, TF, and SHARK-RT of course. this work is based of SharkBenchmarker for MLIR part and Microsoft Transformer Benchmark.
EDIT: nightly ORT did not fix GPU nor did it fix TF.

Some issues/Enhancements that need fixing

1. Integrate running of TF in HF-Benchmarker.

Has some Runtime issues wrt RuntimeError: Intra op parallelism cannot be modified after initialization. and RuntimeError: Visible devices cannot be modified after being initialized. See https://github.com/microsoft/onnxruntime/issues/ 11751 for more details.

2. Fix up HF Benchmark Runtime with GPU

Currently the only supported device is CPU, since we will get OOM with GPU. The problem lies within importing of onnxruntime causes to load 39GB of data into the GPU, this leaves very little space for us to load our model and even run anything.

[powderluv]

3: Also prevent saving into /tmp/

Saved vmfb in /tmp/torch_forward_cpu.vmfb.

4: Print timer per iteration in ms (currently we print only iterations / sec)

5: Save results in a results.csv (or append to the Onnx results.csv file)

[raikonenfnu]

For 3. I think we then need to change the default of --repro_dir in shark/parser.py