Review pass over the merged diffusion phases: seven correctness fixes

Re-reviewed each merged phase PR against this branch's tip and fixed what is
still real:

- A superseded background load no longer cancels the current model's in-flight
  generation: the load-token check now runs BEFORE the cancel signal, with a
  re-check under the generate lock (Phase 1 review).
- enable_model_cpu_offload / enable_sequential_cpu_offload now forward the
  resolved target device; diffusers defaults to CUDA, which broke offloaded
  loads on non-CUDA accelerators such as Intel XPU (Phase 2 review).
- build_sd_cpp_command rejects a None prompt (str(None) previously slipped
  into argv as the literal "None") and a mask without an init image, which
  is an invalid sd-cli inpaint invocation (Phase 4/6 review).
- The dense-quant OOM fallback drops the caught exception before
  clear_gpu_cache(): the traceback pinned the partially built dense
  transformer, so the VRAM this cleanup exists to reclaim stayed allocated
  through the GGUF rebuild (Phase 8 review).
- Pre-quantized transformers (built via from_config) are eval()'d to match
  the from_pretrained paths, so train-mode layers cannot make prequant
  inference nondeterministic (Phase 9 review).
- FBCache state is reset before each generation when a step cache is engaged:
  diffusers never clears the stateful first-block residuals on the resident
  transformer, so a resolution or batch change on the next request hit a
  shape mismatch, and an unchanged request could reuse stale residuals
  (Phase 12 review).

Each fix carries a regression test; the full diffusion battery passes.

studio/backend/core/inference/diffusion_prequant.py

@@ -193,6 +193,15 @@ def load_prequantized_transformer(
             transformer.load_state_dict(state_dict, strict = True, assign = True)
 
         transformer = transformer.to(device)
+        # Built via from_config (not from_pretrained), so it starts in TRAIN mode; the
+        # dense and GGUF paths load through from_pretrained, which diffusers documents as
+        # returning an eval()'d module. Match that here so any train/eval-sensitive layer
+        # (e.g. dropout) can't make prequant inference nondeterministic or diverge from
+        # the other load paths.
+        try:
+            transformer.eval()
+        except Exception:  # noqa: BLE001 — eval() is best-effort
+            pass
         try:  # diagnostic marker, mirrors the runtime-quant path
             transformer._unsloth_runtime_quant = scheme
         except Exception:  # noqa: BLE001 — marker is best-effort