builds for torch 2.7.0 (#2552)

* builds for torch==2.7.0

* use xformers==0.0.29.post3

* no vllm support with torch 2.7

* update default, fix conditional

* no xformers for 270

* no vllm on 2.7.0 for multigpu test too

* remove deprecated verbose arg from scheduler

* 2.7.0 tests on cpu

src/axolotl/monkeypatch/relora.py

@@ -272,7 +272,7 @@ class ReLoRAScheduler(LRScheduler):
         self.warmup_steps = warmup_steps
         self.anneal_steps = anneal_steps
         self.min_lr_scale = min_lr_scale
-        super().__init__(optimizer, inner_schedule.last_epoch, inner_schedule.verbose)
+        super().__init__(optimizer, inner_schedule.last_epoch)
 
     def get_lr(self) -> float:
         self.inner_schedule.last_epoch = self.last_epoch