NLP

cayetanoguerra · Nov 5, 2024 · ac77027 · ac77027
1 parent 30c7677
commit ac77027
Show file tree

Hide file tree

Showing 29 changed files with 12,839 additions and 1,338 deletions.
diff --git a/.DS_Store b/.DS_Store
diff --git a/Adevinta-NLP/Tema-05/.DS_Store b/Adevinta-NLP/Tema-05/.DS_Store
diff --git a/Adevinta-NLP/Tema-06/.DS_Store b/Adevinta-NLP/Tema-06/.DS_Store
diff --git a/Adevinta-NLP/Tema-07/.DS_Store b/Adevinta-NLP/Tema-07/.DS_Store
diff --git a/Adevinta-NLP/Tema-08/.DS_Store b/Adevinta-NLP/Tema-08/.DS_Store
diff --git a/ia/.DS_Store b/ia/.DS_Store
diff --git a/ia/index.html b/ia/index.html
@@ -378,8 +378,9 @@ <h5>5. Modelos del lenguaje basados en redes neuronales artificiales</h5>
 </li>
 <li><a href="http://nbviewer.jupyter.org/url/cayetanoguerra.github.io/ia/nbpy/nlp/Tema-05/nb07.ipynb">Transformers</a></li>
 <li><a href="http://nbviewer.jupyter.org/url/cayetanoguerra.github.io/ia/nbpy/nlp/Tema-05/nb08.ipynb">🤗 Hugging Face</a></li>
-<li><a href="http://nbviewer.jupyter.org/url/cayetanoguerra.github.io/ia/nbpy/nlp/Tema-05/nb09.ipynb">BERT</a></li>
-<li><a href="http://nbviewer.jupyter.org/url/cayetanoguerra.github.io/ia/nbpy/nlp/Tema-05/nb10.ipynb">GPT</a></li>
+<li><a href="http://nbviewer.jupyter.org/url/cayetanoguerra.github.io/ia/nbpy/nlp/Tema-05/nb09.ipynb">T5 - Ejemplo de fine tuning</a></li>
+<li><a href="http://nbviewer.jupyter.org/url/cayetanoguerra.github.io/ia/nbpy/nlp/Tema-05/nb10.ipynb">BERT</a></li>
+<li><a href="http://nbviewer.jupyter.org/url/cayetanoguerra.github.io/ia/nbpy/nlp/Tema-05/nb11.ipynb">GPT</a></li>
 </ul>
 
 <h4>Exámenes</h4>

diff --git a/ia/nbpy/.DS_Store b/ia/nbpy/.DS_Store
diff --git a/ia/nbpy/nlp/.DS_Store b/ia/nbpy/nlp/.DS_Store
diff --git a/ia/nbpy/nlp/Tema-05/.DS_Store b/ia/nbpy/nlp/Tema-05/.DS_Store
diff --git a/ia/nbpy/nlp/Tema-05/data/.DS_Store b/ia/nbpy/nlp/Tema-05/data/.DS_Store
diff --git a/ia/nbpy/nlp/Tema-05/data/numbers.csv b/ia/nbpy/nlp/Tema-05/data/numbers.csv
diff --git a/ia/nbpy/nlp/Tema-05/imgs/T5.png b/ia/nbpy/nlp/Tema-05/imgs/T5.png
diff --git a/.../nlp/Tema-05/logs/events.out.tfevents.1730806454.MacBook-Pro-de-Cayetano-M1.local.59694.0 b/.../nlp/Tema-05/logs/events.out.tfevents.1730806454.MacBook-Pro-de-Cayetano-M1.local.59694.0
diff --git a/ia/nbpy/nlp/Tema-05/nb01-T5.ipynb b/ia/nbpy/nlp/Tema-05/nb01-T5.ipynb
diff --git a/ia/nbpy/nlp/Tema-05/nb08.ipynb b/ia/nbpy/nlp/Tema-05/nb08.ipynb
@@ -847,6 +847,31 @@
     "- **Uso de clases específicas**: Si sabes exactamente con qué modelo estás trabajando y necesitas un control fino sobre la configuración y el comportamiento del modelo, usar una clase específica puede ser beneficioso.\n",
     "- **Uso de clases \"Auto\"**: Si prefieres escribir código más genérico que pueda adaptarse fácilmente a diferentes modelos o si estás explorando diferentes modelos, usar una clase \"Auto\" puede ser más conveniente."
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2.2.1\n"
+     ]
+    }
+   ],
+   "source": [
+    "import torch\n",
+    "print(torch.__version__)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
   }
  ],
  "metadata": {

diff --git a/ia/nbpy/nlp/Tema-05/nb09.ipynb b/ia/nbpy/nlp/Tema-05/nb09.ipynb