fix: log the actual (resolved) model in semantics step, not stale llm_model

The progress line printed the raw upstream llm_model (e.g. 'gpt-5.5') before
ModelSemanticUpdater ran it through resolve_model_name(), which currently
force-routes everything to claude-sonnet-4-6 (TEMPORARY_MODEL_OVERRIDE). Log
updater.llm_model so the message reflects the model actually used.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

thoughtspot_deployer.py

@@ -3679,9 +3679,9 @@ class ThoughtSpotDeployer:
                                     if company_research:
                                         try:
                                             from model_semantic_updater import ModelSemanticUpdater
-                                            log_progress(f"Generating model description, synonyms, and AI context with {llm_model}...")
                                             sem_start = time.time()
                                             updater = ModelSemanticUpdater(self, llm_model=llm_model)
+                                            log_progress(f"Generating model description, synonyms, and AI context with {updater.llm_model}...")
 
                                             model_description = updater.generate_model_description(
                                                 company_research=company_research,