Fix starters for Weight Dequantization Problem

[zinchse]

The starter code for the Weight Dequantization problem has an incorrect function signature that doesn't match the test runner's expectations

# current
def solve(X: torch.Tensor, S: torch.Tensor, Y: torch.Tensor, TILE_SIZE: int):
    pass 

# vs expected
def solve(X: torch.Tensor, S: torch.Tensor, Y: torch.Tensor, M: int, N: int, TILE_SIZE: int):
    pass

To reproduce the issue (e.g. in Pytorch) submit the default starter code to see the error "TypeError: solve() takes 4 positional arguments but 6 were given. Test case failed"

Users must manually fix the function signature before being able to test their solution. While this is a minor fix, it creates confusion and breaks the out-of-box experience.

What I've done: I've updated starter code for Pytorch, Cute, Mojo, and Triton to include the missing M and N parameters. After quickly reviewing the eval code, I believe they share the same call signature. I also verified from the user side and testing confirmed the runner does pass 6 arguments in all cases. However, I have not thoroughly audited the entire codebase - please verify no other dependencies are affected by this change.

---

P.S. Submitting the unmodified starter explicitly crashes in Pytorch, Cute, and Triton. Mojo doesn't crash with the wrong signature, so I verified it separately by printing the arguments:

@export
fn solve(X: UnsafePointer[Float32], S: UnsafePointer[Float32], Y: UnsafePointer[Float32], M: Int32, N: Int32, TILE_SIZE: Int32) raises:
    print("M=", M, "N=", N, "TILE=", TILE_SIZE)