torchchat always prints help

[malfet]

For example:

 % python3 torchchat.py generate --quantize '{"embedding" : {"bitwidth": 8, "groupsize": 0}, "linear:int4" : {"groupsize": 32}}'  --checkpoint-path checkpoints/stories110M/model.pth       
usage: torchchat [-h]

Welcome to the torchchat CLI!

optional arguments:
  -h, --help  show this help message and exit

[malfet]

Another thing is that if one makes a typo in the arg, it's really hard to understand what is missing, for example

% python3 torchchat.py generate --quantze '{"embedding" : {"bitwidth": 8, "groupsize": 0}, "linear:int4" : {"groupsize": 32}}'  --checkpoint-path checkpoints/stories110M/model.pth 
usage: torchchat [-h]

Welcome to the torchchat CLI!

optional arguments:
  -h, --help  show this help message and exit
usage: torchchat [-h] [--chat] [--gui] [--prompt PROMPT] [--is-chat-model] [--seed SEED] [--num-samples NUM_SAMPLES] [--max-new-tokens MAX_NEW_TOKENS] [--top-k TOP_K]
                 [--temperature TEMPERATURE] [--compile] [--compile-prefill] [--parallel-prefill] [--profile PROFILE] [--speculate-k SPECULATE_K]
                 [--draft-checkpoint-path DRAFT_CHECKPOINT_PATH] [--checkpoint-path CHECKPOINT_PATH] [--params-path PARAMS_PATH] [--gguf-path GGUF_PATH] [--tokenizer-path TOKENIZER_PATH]
                 [--output-pte-path OUTPUT_PTE_PATH] [--output-dso-path OUTPUT_DSO_PATH] [--dso-path DSO_PATH] [--pte-path PTE_PATH] [-d {fp32,fp16,bf16,float,half,float32,float16,bfloat16}]
                 [-v] [--quantize QUANTIZE] [--draft-quantize DRAFT_QUANTIZE]
                 [--params-table {13B,70B,CodeLlama-7b-Python-hf,34B,stories42M,30B,stories110M,7B,stories15M,Mistral-7B,Meta-Llama-3-8B}] [--device {cpu,cuda,mps}] [--tasks TASKS [TASKS ...]]
                 [--limit LIMIT] [--max-seq-length MAX_SEQ_LENGTH] [--hf-token HF_TOKEN] [--model-directory MODEL_DIRECTORY] [--port PORT]
                 {chat,browser,download,generate,eval,export,list,remove} ...
torchchat: error: argument command: invalid choice: '{"embedding" : {"bitwidth": 8, "groupsize": 0}, "linear:int4" : {"groupsize": 32}}' (choose from 'chat', 'browser', 'download', 'generate', 'eval', 'export', 'list', 'remove')

[byjlw]

Yeah this could be better.
Can you articulate what you'd expect to see? maybe an example?

[GregoryComer]

Turns out it was bugged. Should be fixed in #488.