[Transform] Fix QuIP targets (#1770)

## Purpose ##
* Fix QuIP transform targets

## Testing ##
* Used `test_correctness.py` test and validated that previously, the
loss was zero, but now that the targets are correct, the loss matches
what was previously measured
*
https://huggingface.co/nm-testing/Llama-3.1-8B-Instruct-quip-w4a16/tree/main/
* TODO: will do basic lm_eval, but good to land now

Signed-off-by: Kyle Sayers <[email protected]>
Co-authored-by: Dipika Sikka <[email protected]>

Author: kylesayrs

src/llmcompressor/modifiers/transform/quip/base.py

@@ -61,7 +61,7 @@ class QuIPModifier(Modifier):
     transform_type: Literal["hadamard", "random-hadamard", "random-matrix"] = Field(
         default="random-hadamard"
     )
-    targets: Union[List[str], str] = Field(default="str")
+    targets: Union[List[str], str] = Field(default="Linear")
     randomize: bool = Field(default=False)
     learnable: bool = Field(default=False)
     precision: TorchDtype = Field(default=torch.float64)