Single model config: remove model arg from all action functions

Model is configured ONCE via configure_dspy() at notebook startup.
Action functions (run_baseline, run_optimization, etc.) use the
already-configured LM. No more threading issues from widget callbacks.

Fixed broken prompt_workshop() calls (missing commas from regex cleanup).

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Author: abossard

notebooks/01_evaluation_and_tuning.ipynb

@@ -71,16 +71,14 @@
    "outputs": [],
    "source": [
     "from dspy_tasks.visualize import prompt_workshop\n",
-    "import ipywidgets as widgets\n",
     "\n",
     "# Dein manueller Versuch — editiere den Prompt!\n",
     "manual_workshop = prompt_workshop(\n",
     "    task_id=\"multihop_qa\",\n",
-    "    model_widget=model_dd,\n",
-    "    default_instructions=\"Read the context carefully and answer the question. Think step by step. If the answer requires combining multiple facts, do so explicitly.\",\n",
+    "    default_instructions=\"Read the context carefully and answer the question. Think step by step.\",\n",
     "    max_eval=6,\n",
     ")\n",
-    "display(manual_workshop)"
+    "display(manual_workshop)\n"
    ]
   },
   {
@@ -123,7 +121,7 @@
     "        print(f\"⏳ Optimizing {task.name} with {opt_dd.value} on {model_dd.value}...\")\n",
     "        print(f\"   This may take 10-60 seconds...\\n\")\n",
     "\n",
-    "        result = run_optimization(\"multihop_qa\", model_dd.value, opt_dd.value, max_eval=8)\n",
+    "        result = run_optimization(\"multihop_qa\", opt_dd.value, max_eval=8)\n",
     "\n",
     "        display_improvement(result.baseline_score, result.optimized_score)\n",
     "        print(f\"⏱️  Optimization took {result.elapsed_seconds}s | {result.llm_calls} LLM calls\")\n",
@@ -166,10 +164,10 @@
     "        task = get_task(\"ticket_routing\")\n",
     "        print(f\"⏳ Comparing optimizers on {task.name}...\\n\")\n",
     "\n",
-    "        r_bs = run_optimization(\"ticket_routing\", model_dd.value, \"BootstrapFewShot\", max_eval=8)\n",
+    "        r_bs = run_optimization(\"ticket_routing\", \"BootstrapFewShot\", max_eval=8)\n",
     "        print(f\"BootstrapFewShot: {r_bs.baseline_score:.0%} → {r_bs.optimized_score:.0%} ({r_bs.elapsed_seconds}s)\")\n",
     "\n",
-    "        r_mipro = run_optimization(\"ticket_routing\", model_dd.value, \"MIPROv2\", max_eval=8)\n",
+    "        r_mipro = run_optimization(\"ticket_routing\", \"MIPROv2\", max_eval=8)\n",
     "        print(f\"MIPROv2:          {r_mipro.baseline_score:.0%} → {r_mipro.optimized_score:.0%} ({r_mipro.elapsed_seconds}s)\")\n",
     "\n",
     "        scores = {\n",
@@ -197,7 +195,7 @@
     "def on_any_optimize(b):\n",
     "    with optimize_out:\n",
     "        optimize_out.clear_output()\n",
-    "        result = run_optimization(task_dd.value, model_dd.value, opt_dd.value, max_eval=8)\n",
+    "        result = run_optimization(task_dd.value, opt_dd.value, max_eval=8)\n",
     "        display_improvement(result.baseline_score, result.optimized_score)\n",
     "        display_prompt_diff(result.prompt_before, result.prompt_after)\n",
     "\n",
@@ -258,4 +256,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}

notebooks/02_optimization.ipynb

@@ -74,8 +74,7 @@
     "diagram_compare(\n",
     "    left={\"title\": \"Generischer Prompt\", \"items\": [\"Keine Domain-Daten\", \"Gemietetes Modell\", \"Ergebnis: Mittelmässig\"], \"icon\": \"📝\", \"color\": \"#a4262c\"},\n",
     "    right={\"title\": \"Getuned mit deinen Daten\", \"items\": [\"Echte Tickets als Training\", \"Gleiches Modell\", \"Ergebnis: Deutlich besser! 🎯\"], \"icon\": \"📊\", \"color\": \"#107c10\"},\n",
-    "    title=\"Vorher vs. Nachher: Deine Daten machen den Unterschied\",\n",
-    ")\n"
+    "    title=\"Vorher vs. Nachher: Deine Daten machen den Unterschied\")\n"
    ]
   },
   {
@@ -95,7 +94,7 @@
     "        out.clear_output()\n",
     "        print(f\"⏳ Running ticket routing: generic vs. domain-tuned on {model_dd.value}...\")\n",
     "\n",
-    "        result = run_optimization(\"ticket_routing\", model_dd.value, \"BootstrapFewShot\", max_eval=10)\n",
+    "        result = run_optimization(\"ticket_routing\", \"BootstrapFewShot\", max_eval=10)\n",
     "\n",
     "        display_improvement(result.baseline_score, result.optimized_score)\n",
     "        display_prompt_diff(result.prompt_before, result.prompt_after,\n",
@@ -164,7 +163,7 @@
     "def on_run_extra(b):\n",
     "    with run_out:\n",
     "        run_out.clear_output()\n",
-    "        result = run_optimization(task_dd.value, model_dd.value, max_eval=8)\n",
+    "        result = run_optimization(task_dd.value, max_eval=8)\n",
     "        display_improvement(result.baseline_score, result.optimized_score)\n",
     "        display_results_table(result.individual_scores if hasattr(result, 'individual_scores') else [])\n",
     "\n",
@@ -207,4 +206,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}

notebooks/03_domain_tuning.ipynb

@@ -102,7 +102,7 @@
     "    with out:\n",
     "        out.clear_output()\n",
     "        print(f\"⏳ Running Calculator Agent on {model_dd.value}...\")\n",
-    "        result = run_baseline(\"calculator_agent\", model_dd.value, max_eval=5)\n",
+    "        result = run_baseline(\"calculator_agent\", max_eval=5)\n",
     "        display_score(\"Calculator Agent\", result.score)\n",
     "        display_results_table(result.individual_scores)\n",
     "\n",
@@ -133,7 +133,7 @@
     "        out2.clear_output()\n",
     "        print(f\"⏳ Running Search Agent on {model_dd.value}...\")\n",
     "        print(\"   The agent will search your ticket database and synthesize answers...\\n\")\n",
-    "        result = run_baseline(\"search_agent\", model_dd.value, max_eval=5)\n",
+    "        result = run_baseline(\"search_agent\", max_eval=5)\n",
     "        display_score(\"Search Agent\", result.score)\n",
     "        display_results_table(result.individual_scores)\n",
     "\n",
@@ -172,7 +172,7 @@
     "        print(f\"⏳ Optimizing {task.name} on {model_dd.value}...\")\n",
     "        print(f\"   Using {opt_dd.value} optimizer. This may take a minute...\\n\")\n",
     "\n",
-    "        result = run_optimization(task_dd.value, model_dd.value, opt_dd.value, max_eval=5)\n",
+    "        result = run_optimization(task_dd.value, opt_dd.value, max_eval=5)\n",
     "\n",
     "        display_improvement(result.baseline_score, result.optimized_score)\n",
     "        display_prompt_diff(result.prompt_before, result.prompt_after,\n",
@@ -227,4 +227,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}

notebooks/04_agents.ipynb

@@ -88,7 +88,7 @@
     "\n",
     "            for model in MODELS:\n",
     "                print(f\"  ⏳ {task.name} on {model.split('/')[-1]}...\", end=\" \")\n",
-    "                result = run_baseline(task_id, model, max_eval=5)\n",
+    "                result = run_baseline(task_id, max_eval=5)\n",
     "                row.append(result.score)\n",
     "                print(f\"{result.score:.0%}\")\n",
     "\n",
@@ -355,4 +355,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}

notebooks/05_full_picture.ipynb

@@ -224,7 +224,7 @@
     "    with out:\n",
     "        out.clear_output()\n",
     "        print(f\"⏳ Running sentiment task on {model_dd.value}...\")\n",
-    "        result = run_baseline(\"sentiment\", model_dd.value, max_eval=10)\n",
+    "        result = run_baseline(\"sentiment\", max_eval=10)\n",
     "        display_score(\"Baseline Score\", result.score)\n",
     "        print(f\"⏱️  {result.elapsed_seconds}s | {result.llm_calls} LLM calls\")\n",
     "        display_results_table(result.individual_scores)\n",
@@ -281,7 +281,7 @@
     "        out2.clear_output()\n",
     "        task = get_task(task_dd.value)\n",
     "        print(f\"⏳ Running {task.name} on {model_dd.value}...\")\n",
-    "        result = run_baseline(task_dd.value, model_dd.value, max_eval=8)\n",
+    "        result = run_baseline(task_dd.value, max_eval=8)\n",
     "        display_score(task.name, result.score)\n",
     "        print(f\"⏱️  {result.elapsed_seconds}s | 💡 {task.teaching_point}\")\n",
     "        display_results_table(result.individual_scores[:5])\n",
@@ -312,8 +312,7 @@
     "from dspy_tasks.data import ClassifySentiment\n",
     "import dspy\n",
     "\n",
-    "lm = configure_dspy(model_dd.value)\n",
-    "\n",
+    "lm = \n",
     "tricky_reviews = [\n",
     "    (\"Super, schon wieder ein Produkt das nach einer Woche kaputt geht. Genau was ich brauchte.\", \"negative\"),\n",
     "    (\"Die Kamera ist ok, aber für den Preis hätte ich mehr erwartet.\", \"neutral\"),\n",
@@ -470,4 +469,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}

notebooks/appendix_a_grokking_simplicity.ipynb

@@ -85,7 +85,6 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "configure_dspy(model=model_dd.value)\n",
     "\n",
     "# SHALLOW: Just predict\n",
     "shallow = dspy.Predict(TranslateEnDe)\n",
@@ -167,7 +166,6 @@
     "    with out:\n",
     "        out.clear_output()\n",
     "        task = get_task(task_dd.value)\n",
-    "        configure_dspy(model=model_dd.value)\n",
     "        if module_dd.value == \"ChainOfThought\":\n",
     "            module = dspy.ChainOfThought(task.signature_class)\n",
     "        else:\n",
@@ -213,4 +211,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}