Allow models to force a new batch

This is useful for a few things:
 - Work around bugs, such as having 2 images in one batch
 - Keep the image in a single batch for fully connected attention
 - Improve performance by not evaluating embeddings multiple times

runner/ollamarunner/runner.go

@@ -363,7 +363,7 @@ func (s *Server) processBatch() error {
 				}
 			}
 
-			if j >= s.batchSize {
+			if j >= s.batchSize || (inp.BatchBreak && len(seq.pendingInputs) != 0) {
 				break
 			}