MOSH-2174: Expose gradient_accumulation_steps in the API

openapi.yaml

@@ -2569,6 +2569,10 @@ paths:
                         - max
                   default: 'max'
                   description: Number of training examples processed together (larger batches use more memory but may train faster). Defaults to "max". We use training optimizations like packing, so the effective batch size may be different than the value you set.
+                gradient_accumulation_steps:
+                  type: integer
+                  default: 1
+                  description: Number of steps to accumulate gradients before performing a weight update. Effectively increases the batch size without requiring more memory. For example, with batch_size=4 and gradient_accumulation_steps=8, the effective batch size is 32.
                 learning_rate:
                   type: number
                   format: float