[bugfix]: keep text-encoder CPU offload on by default

Revert the default flips from 40c8b7c1 (OffloadConfig.text_encoder and
FastVideoArgs.text_encoder_cpu_offload False -> True). Flipping the
global default changes the baseline VRAM footprint for every model
family, not just Kandinsky, and breaks the unit lane:
fastvideo/tests/api/test_parser.py::test_load_run_config_supports_yaml_roundtrip
still golden-checks text_encoder offload as True.

Nothing in this PR needs the flip: the Kandinsky examples pass
text_encoder_cpu_offload explicitly, and the offload-aware
TextEncodingStage (kept) makes offload work for Kandinsky's CLIP
either way. Offload stays opt-out rather than opt-in.

fastvideo/api/schema.py

@@ -25,7 +25,7 @@ class ParallelismConfig:
 class OffloadConfig:
     dit: bool = True
     dit_layerwise: bool = True
-    text_encoder: bool = False
+    text_encoder: bool = True
     image_encoder: bool = True
     vae: bool = True
     pin_cpu_memory: bool = True

fastvideo/fastvideo_args.py

@@ -127,7 +127,7 @@ class FastVideoArgs:
     dit_cpu_offload: bool = True
     use_fsdp_inference: bool = False
     dit_layerwise_offload: bool = True
-    text_encoder_cpu_offload: bool = False
+    text_encoder_cpu_offload: bool = True
     image_encoder_cpu_offload: bool = True
     vae_cpu_offload: bool = True
     pin_cpu_memory: bool = True