Fix package extras for watsonx support (EleutherAI#2426)

* Update pyproject.toml with watsonx package extra

Signed-off-by: kiersten-stokes <kierstenstokes@gmail.com>

* Remove unused function

Signed-off-by: kiersten-stokes <kierstenstokes@gmail.com>

---------

Signed-off-by: kiersten-stokes <kierstenstokes@gmail.com>

lm_eval/models/ibm_watsonx_ai.py

@@ -158,7 +158,7 @@ def __init__(
         project_id = watsonx_credentials.get("project_id", None)
         deployment_id = watsonx_credentials.get("deployment_id", None)
         client.set.default_project(project_id)
-        self.generate_params = generate_params
+        self.generate_params = generate_params or {}
         self.model = ModelInference(
             model_id=model_id,
             deployment_id=deployment_id,
@@ -167,12 +167,6 @@ def __init__(
         )
         self._model_id = model_id
 
-    def dump_parameters(self):
-        """
-        Dumps the model's parameters into a serializable format.
-        """
-        return self._parameters.model_dump()
-
     @staticmethod
     def _has_stop_token(response_tokens: List[str], context_tokens: List[str]) -> bool:
         """

pyproject.toml

@@ -62,6 +62,7 @@ dev = ["pytest", "pytest-cov", "pytest-xdist", "pre-commit", "mypy"]
 deepsparse = ["deepsparse-nightly[llm]>=1.8.0.20240404"]
 gptq = ["auto-gptq[triton]>=0.6.0"]
 hf_transfer = ["hf_transfer"]
+ibm_watsonx_ai = ["ibm_watsonx_ai"]
 ifeval = ["langdetect", "immutabledict", "nltk>=3.9.1"]
 neuronx = ["optimum[neuronx]"]
 mamba = ["mamba_ssm", "causal-conv1d==1.0.2"]
@@ -81,6 +82,7 @@ all = [
     "lm_eval[deepsparse]",
     "lm_eval[gptq]",
     "lm_eval[hf_transfer]",
+    "lm_eval[ibm_watsonx_ai]",
     "lm_eval[ifeval]",
     "lm_eval[mamba]",
     "lm_eval[math]",
@@ -93,7 +95,6 @@ all = [
     "lm_eval[vllm]",
     "lm_eval[zeno]",
     "lm_eval[wandb]",
-    "lm_eval[ibm_watsonx_ai]"
 ]
 
 [tool.ruff.lint]