docs: minor changes to tutorials (NVIDIA-NeMo#1747)

minor updates

Signed-off-by: Krishna Kalyan <krkalyan@nvidia.com>

Author: krishnakalyan3

tutorials/README.md

@@ -3,4 +3,4 @@
 
 | Tutorial | Type | Brev Instance | Launch on Brev |
 | --- | --- | --- | --- |
-| [**Nemotron Parse fine-tune**](https://github.com/NVIDIA-NeMo/Automodel/blob/main/tutorials/nemotron-parse/finetune.ipynb) | Fine-tuning | L40S | [![Launch on Brev](https://brev-assets.s3.us-west-1.amazonaws.com/nv-lb-dark.svg)](https://brev.nvidia.com/launchable/deploy/now?launchableID=env-3BxDyGV2RLNqyftNxJO5rV5sp8x) |
+| [**Nemotron Parse fine-tune**](https://github.com/NVIDIA-NeMo/Automodel/blob/main/tutorials/nemotron-parse/finetune.ipynb) | Fine-tuning | L40S | [![Launch on Brev](https://brev-assets.s3.us-west-1.amazonaws.com/nv-lb-dark.svg)](https://brev.nvidia.com/launchable/deploy/now?launchableID=env-3C6LDKU2DfOvpVTFhjw3YQ4djPM) |

tutorials/nemotron-parse/finetune.ipynb

@@ -679,12 +679,18 @@
     "\n",
     "## Conclusion\n",
     "\n",
-    "This tutorial demonstrated how to fine-tune Nemotron Parse v1.1 for structured invoice extraction using NeMo Automodel. Starting from a base model that could only produce generic markdown, we trained on 425 invoices for 1 epoch and achieved near-perfect field extraction on held-out test data — bringing the average NED from ~0.81 down to ~0.05 and field-level accuracy from 0% to over 90%. The same approach generalizes to any document type where you need to extract structured fields from images.\n",
+    "This tutorial demonstrated how to fine-tune Nemotron Parse v1.1 for structured invoice extraction using NeMo Automodel. Starting from a base model that could only produce generic markdown, we trained on 425 invoices for 1 epoch and achieved near-perfect field extraction on held-out test data — bringing the average NED from ~0.81 down to ~0.1 and field-level accuracy from 0% to over 80%. The same approach generalizes to any document type where you need to extract structured fields from images.\n",
     "\n",
     "---\n",
     "\n",
     "📓 *This notebook was created by* **Aastha Jhunjhunwala** and **Huiying Li** from <b>NVIDIA</b>"
    ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "69f84387",
+   "metadata": {},
+   "source": []
   }
  ],
  "metadata": {