Remove stale Qwen references from minimal-baseline docstrings

Runtime defaults already point to CohereLabs/aya-expanse-32b; only
docstrings / usage examples still named Qwen. Cosmetic cleanup only.

app_minimal.py

@@ -1,6 +1,6 @@
 """Minimal baseline Gradio entry point for the Month 1-3 rebuild.
 
-Wires the simplest possible slice: Whisper (zero-shot) -> Qwen -> MMS-TTS.
+Wires the simplest possible slice: Whisper (zero-shot) -> Aya-Expanse -> MMS-TTS.
 No LoRA adapters, no memory loop, no speaker ID, no voice cloning, no IoT,
 no phrase matcher. Used for field testing and building a real-user eval set.
 
@@ -10,7 +10,7 @@ Run locally:
     HF_TOKEN=hf_xxx python app_minimal.py
 
 Environment variables (all optional except HF_TOKEN, which is needed for the
-Qwen HF Serverless call):
+HF Serverless LLM call):
     HF_TOKEN       — HuggingFace token with read access
     LLM_MODEL_ID   — default "CohereLabs/aya-expanse-32b"
                      (23-language multilingual, strong African-language coverage)

src/llm/minimal_client.py

@@ -126,7 +126,7 @@ class MinimalClient:
     """Dialect-anchored plain-text LLM client over HF Serverless Inference.
 
     Usage:
-        client = MinimalClient(model_id="Qwen/Qwen2.5-7B-Instruct", hf_token=TOK)
+        client = MinimalClient(model_id="CohereLabs/aya-expanse-32b", hf_token=TOK)
         reply  = client.chat("Good morning", target_lang="bam")
         # → "I ni sɔgɔma. I ka kɛnɛ wa?"
     """