Fix load_lora_weights crash on modular sub-pipelines without transformer

[ParamChordiya]

Summary

Fixes #13487

• load_lora_weights crashes with AttributeError: 'Flux2ModularPipeline' object has no attribute 'transformer' when called on a modular sub-pipeline that doesn't have the transformer component (e.g., a text-encoder-only sub-pipeline)
• Root cause: load_lora_weights unconditionally accesses self.transformer via unsafe getattr/hasattr patterns
• Fix: use safe attribute access (getattr with default None) and skip loading with a warning when the target component is not available

Changes

Flux2LoraLoaderMixin.load_lora_weights: checks if transformer exists; warns and returns early if missing.

FluxLoraLoaderMixin.load_lora_weights: independently guards transformer and text_encoder loading, so a sub-pipeline with only one of them still works. Only returns early if both are absent.

Test: added test_load_lora_weights_warns_when_transformer_missing to verify the fix.

Before

text_pipe.load_lora_weights("flux.2-turbo-lora.safetensors")
# AttributeError: 'Flux2ModularPipeline' object has no attribute 'transformer'

After

text_pipe.load_lora_weights("flux.2-turbo-lora.safetensors")
# WARNING: The `transformer` component is not available in this pipeline. Skipping LoRA weight loading.

Test plan

• Verify load_lora_weights on a full Flux2Pipeline still works normally
• Verify load_lora_weights on a text-encoder-only modular sub-pipeline warns and skips without crashing
• Verify load_lora_weights on a denoise-only modular sub-pipeline loads transformer LoRA correctly
• Run existing LoRA tests: pytest tests/lora/test_lora_layers_flux2.py