update dependencies - liger + trl (#2987)

* update dependencies

* set dataset processes for tests

* add support for GSPO

src/axolotl/core/trainers/grpo/__init__.py

@@ -82,6 +82,11 @@ class GRPOStrategy:
         grpo_args_kwargs["log_completions"] = trl.log_completions
         grpo_args_kwargs["num_completions_to_print"] = trl.num_completions_to_print
 
+        if trl.importance_sampling_level is not None:
+            grpo_args_kwargs["importance_sampling_level"] = (
+                trl.importance_sampling_level
+            )
+
         if cfg.sequence_parallel_degree > 1:
             grpo_args_kwargs["sequence_parallel_degree"] = cfg.sequence_parallel_degree
 

src/axolotl/utils/schemas/trl.py

@@ -80,6 +80,14 @@ class TRLConfig(BaseModel):
             "description": "Number of completions to print when log_completions is True."
         },
     )
+    importance_sampling_level: Literal["sequence", "token"] | None = Field(
+        default=None,
+        json_schema_extra={
+            "description": "Controls whether importance sampling ratios are computed at the `'token'` or `'sequence'` level. "
+            "For GSPO, use `sequence`, default is None which corresponds to the original GRPO paper."
+        },
+    )
+
     sync_ref_model: bool | None = Field(
         default=False,
         json_schema_extra={"description": "Whether to sync the reference model."},