Add train/eval loss plot after training

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

notebooks/ch_trader_finetune.ipynb

@@ -1235,6 +1235,13 @@
    "outputs": [],
    "execution_count": null
   },
+  {
+   "cell_type": "code",
+   "source": "import matplotlib.pyplot as plt\n\n# Extract losses from trainer log history\ntrain_steps, train_loss = [], []\neval_steps, eval_loss = [], []\n\nfor entry in trainer.state.log_history:\n    if \"loss\" in entry and \"eval_loss\" not in entry:\n        train_steps.append(entry[\"step\"])\n        train_loss.append(entry[\"loss\"])\n    if \"eval_loss\" in entry:\n        eval_steps.append(entry[\"step\"])\n        eval_loss.append(entry[\"eval_loss\"])\n\n# Plot\nfig, ax = plt.subplots(figsize=(10, 5))\nax.plot(train_steps, train_loss, label=\"Train Loss\", alpha=0.8)\nax.plot(eval_steps, eval_loss, label=\"Eval Loss\", marker=\"o\", markersize=4)\nax.set_xlabel(\"Step\")\nax.set_ylabel(\"Loss\")\nax.set_title(\"Training & Validation Loss\")\nax.legend()\nax.grid(True, alpha=0.3)\nplt.tight_layout()\nplt.show()\n\nprint(f\"Final train loss: {train_loss[-1]:.4f}\")\nprint(f\"Final eval loss:  {eval_loss[-1]:.4f}\")",
+   "metadata": {},
+   "execution_count": null,
+   "outputs": []
+  },
   {
    "cell_type": "markdown",
    "source": "## 5. Save & Push to HuggingFace Hub\n\nMerges LoRA adapters into the base model weights and pushes the full model.",

notebooks/stockex-clearing-house-llm-fine-tuning.ipynb

@@ -1235,6 +1235,13 @@
    "outputs": [],
    "execution_count": null
   },
+  {
+   "cell_type": "code",
+   "source": "import matplotlib.pyplot as plt\n\n# Extract losses from trainer log history\ntrain_steps, train_loss = [], []\neval_steps, eval_loss = [], []\n\nfor entry in trainer.state.log_history:\n    if \"loss\" in entry and \"eval_loss\" not in entry:\n        train_steps.append(entry[\"step\"])\n        train_loss.append(entry[\"loss\"])\n    if \"eval_loss\" in entry:\n        eval_steps.append(entry[\"step\"])\n        eval_loss.append(entry[\"eval_loss\"])\n\n# Plot\nfig, ax = plt.subplots(figsize=(10, 5))\nax.plot(train_steps, train_loss, label=\"Train Loss\", alpha=0.8)\nax.plot(eval_steps, eval_loss, label=\"Eval Loss\", marker=\"o\", markersize=4)\nax.set_xlabel(\"Step\")\nax.set_ylabel(\"Loss\")\nax.set_title(\"Training & Validation Loss\")\nax.legend()\nax.grid(True, alpha=0.3)\nplt.tight_layout()\nplt.show()\n\nprint(f\"Final train loss: {train_loss[-1]:.4f}\")\nprint(f\"Final eval loss:  {eval_loss[-1]:.4f}\")",
+   "metadata": {},
+   "execution_count": null,
+   "outputs": []
+  },
   {
    "cell_type": "markdown",
    "source": "## 5. Save & Push to HuggingFace Hub\n\nMerges LoRA adapters into the base model weights and pushes the full model.",