Update CLI arg handling #860

Job	Run time
test-tinystories-executorch (16-core-ubuntu)	14m 16s
test-tinystories-executorch (macos-14-xlarge)	13m 45s
test-tinystories-eager (macos-12)	6m 21s
torchchat-command-load-test (macos-14)	1m 41s
gather-models-gpu	5s
compile-gguf (macos-14)	4m 4s
gather-models-cpu	5s
runner-et (macos-14-xlarge)	10m 37s
test-gguf-util (macos-14)	7m 16s
runner-et (8-core-ubuntu)	10m 50s
test-mps / macos-job	3m 31s
test-build-runner-et-android / linux-job	13m 59s
test-mps-dtype / macos-job	4m 4s
test-cpu-eval-sanity-check (x86_64, stories15M)	49m 57s
test-runner-aoti (x86_64, stories15M)	2m 40s
test-cpu-compile (x86_64, stories15M)	13m 17s
test-cpu-eval-sanity-check (aarch64, stories15M)	52m 46s
test-runner-aoti (aarch64, stories15M)	1m 51s
test-cpu-aoti (x86_64, stories15M)	20m 32s
test-cpu-compile (aarch64, stories15M)	8m 28s
test-cpu-aoti (aarch64, stories15M)	7m 21s
test-gpu-compile (cuda, stories15M) / linux-job	21m 12s
test-gpu-aoti (cuda, stories15M) / linux-job	30m 11s
test-gpu-eval-sanity-check (cuda, stories15M) / linux-job	12m 24s
	5h 11m 13s

Provide feedback