The CLI only supports 6 GPU types while the website supports 8.
Supported GPUs:
Website (https://github.com/tensara/tensara/blob/main/src/constants/gpu.ts#L1-L11):
export const GPU_DISPLAY_NAMES: Record<string, string> = {
all: "All GPUs",
T4: "Tesla T4",
H100: "NVIDIA H100",
H200: "NVIDIA H200",
B200: "NVIDIA B200",
"A100-80GB": "NVIDIA A100",
A10G: "NVIDIA A10G",
L40S: "NVIDIA L40S",
L4: "NVIDIA L4",
}CLI (https://github.com/tensara/cli/blob/main/src/parser.rs#L7-L14):
pub enum GPU {
T4,
A100,
A100_80GB,
H100,
L4,
L40s,
}By the way, the CLI uses L40s (lowercase 's') while the website uses L40S (uppercase 'S'). Seems like this might cause some issues potentially, although I haven't tested it to be honest.
The fix is obvious and straightforward. Add the missing GPU variants to the enum in https://github.com/tensara/cli/blob/main/src/parser.rs#L7-L14. Also update the Display implementation (https://github.com/tensara/cli/blob/main/src/parser.rs#L25-L36) and parser (https://github.com/tensara/cli/blob/main/src/parser.rs#L71-L77) to handle these new cases.
The CLI only supports 6 GPU types while the website supports 8.
Supported GPUs:
Website (https://github.com/tensara/tensara/blob/main/src/constants/gpu.ts#L1-L11):
CLI (https://github.com/tensara/cli/blob/main/src/parser.rs#L7-L14):
By the way, the CLI uses L40s (lowercase 's') while the website uses L40S (uppercase 'S'). Seems like this might cause some issues potentially, although I haven't tested it to be honest.
The fix is obvious and straightforward. Add the missing GPU variants to the enum in https://github.com/tensara/cli/blob/main/src/parser.rs#L7-L14. Also update the Display implementation (https://github.com/tensara/cli/blob/main/src/parser.rs#L25-L36) and parser (https://github.com/tensara/cli/blob/main/src/parser.rs#L71-L77) to handle these new cases.