fix

examples/notebooks/FLARE.ipynb

@@ -53,7 +53,14 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "nbmake": {
+     "post_cell_execute": [
+      "from conftest import before_notebook",
+      "before_notebook()"
+     ]
+    }
+   },
    "outputs": [],
    "source": [
     "! pip install ragstack-ai"
@@ -699,4 +706,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}

examples/notebooks/QA_with_cassio.ipynb

@@ -19,11 +19,11 @@
     "This notebook guides you through setting up [RAGStack](https://www.datastax.com/products/ragstack) using [Astra Vector Search](https://docs.datastax.com/en/astra-serverless/docs/vector-search/overview.html), [OpenAI](https://openai.com/about), and [CassIO](https://cassio.org/) to implement a generative Q&A over your own documentation.\n",
     "\n",
     "## Astra Vector Search\n",
-    "Astra vector search enables developers to search a database by context or meaning rather than keywords or literal values. This is done by using “embeddings”. Embeddings are a type of representation used in machine learning where high-dimensional or complex data is mapped onto vectors in a lower-dimensional space. These vectors capture the semantic properties of the input data, meaning that similar data points have similar embeddings.\n",
+    "Astra vector search enables developers to search a database by context or meaning rather than keywords or literal values. This is done by using \u201cembeddings\u201d. Embeddings are a type of representation used in machine learning where high-dimensional or complex data is mapped onto vectors in a lower-dimensional space. These vectors capture the semantic properties of the input data, meaning that similar data points have similar embeddings.\n",
     "Reference: [Astra Vector Search](https://docs.datastax.com/en/astra-serverless/docs/vector-search/overview.html)\n",
     "\n",
     "## CassIO\n",
-    "CassIO is the ultimate solution for seamlessly integrating Apache Cassandra® with generative artificial intelligence and other machine learning workloads. This powerful Python library simplifies the complicated process of accessing the advanced features of the Cassandra database, including vector search capabilities. With CassIO, developers can fully concentrate on designing and perfecting their AI systems without any concerns regarding the complexities of integration with Cassandra.\n",
+    "CassIO is the ultimate solution for seamlessly integrating Apache Cassandra\u00ae with generative artificial intelligence and other machine learning workloads. This powerful Python library simplifies the complicated process of accessing the advanced features of the Cassandra database, including vector search capabilities. With CassIO, developers can fully concentrate on designing and perfecting their AI systems without any concerns regarding the complexities of integration with Cassandra.\n",
     "Reference: [CassIO](https://cassio.org/)\n",
     "\n",
     "## OpenAI\n",
@@ -72,6 +72,12 @@
     },
     "editable": true,
     "id": "a6d88d66",
+    "nbmake": {
+     "post_cell_execute": [
+      "from conftest import before_notebook",
+      "before_notebook()"
+     ]
+    },
     "outputId": "dc543d17-3fb2-4362-cc4e-0050bd7787ba",
     "slideshow": {
      "slide_type": ""
@@ -543,4 +549,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

examples/notebooks/RAG_with_cassio.ipynb

@@ -67,19 +67,24 @@
      "base_uri": "https://localhost:8080/"
     },
     "id": "2953d95b",
-    "outputId": "f1d1a4dc-9984-405d-bf28-74697815ea12", 
     "nbmake": {
-     "post_cell_execute": ["from conftest import before_notebook", "before_notebook()"]
-    }
+     "post_cell_execute": [
+      "from conftest import before_notebook",
+      "before_notebook()"
+     ]
+    },
+    "outputId": "f1d1a4dc-9984-405d-bf28-74697815ea12"
+   },
+   "nbmake": {
+    "post_cell_execute": [
+     "before_notebook()"
+    ]
    },
    "outputs": [],
    "source": [
     "# install required dependencies\n",
     "! pip install -qU ragstack-ai datasets google-cloud-aiplatform "
-   ],
-   "nbmake": {
-     "post_cell_execute": ["before_notebook()"]
-    }
+   ]
   },
   {
    "cell_type": "markdown",
@@ -478,4 +483,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

examples/notebooks/advancedRAG.ipynb

@@ -55,7 +55,14 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "nbmake": {
+     "post_cell_execute": [
+      "from conftest import before_notebook",
+      "before_notebook()"
+     ]
+    }
+   },
    "outputs": [],
    "source": [
     "%pip install ragstack-ai"
@@ -1035,4 +1042,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}

examples/notebooks/astradb.ipynb

@@ -13,7 +13,7 @@
    "source": [
     "# Astra DB\n",
     "\n",
-    "This page provides a quickstart for using [Astra DB](https://docs.datastax.com/en/astra/home/astra.html) and [Apache Cassandra®](https://cassandra.apache.org/) as a Vector Store.\n",
+    "This page provides a quickstart for using [Astra DB](https://docs.datastax.com/en/astra/home/astra.html) and [Apache Cassandra\u00ae](https://cassandra.apache.org/) as a Vector Store.\n",
     "\n",
     "_Note: in addition to access to the database, an OpenAI API Key is required to run the full example._"
    ]
@@ -39,7 +39,14 @@
    "cell_type": "code",
    "execution_count": null,
    "id": "8d00fcf4-9798-4289-9214-d9734690adfc",
-   "metadata": {},
+   "metadata": {
+    "nbmake": {
+     "post_cell_execute": [
+      "from conftest import before_notebook",
+      "before_notebook()"
+     ]
+    }
+   },
    "outputs": [],
    "source": [
     "!pip install --quiet datasets pypdf"
@@ -902,4 +909,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

examples/notebooks/langchain-unstructured-astra.ipynb

@@ -32,6 +32,12 @@
      "start_time": "2024-02-13T11:47:41.568768Z"
     },
     "editable": true,
+    "nbmake": {
+     "post_cell_execute": [
+      "from conftest import before_notebook",
+      "before_notebook()"
+     ]
+    },
     "slideshow": {
      "slide_type": ""
     },
@@ -254,7 +260,7 @@
     {
      "data": {
       "text/html": [
-       "<table><thead><th></th><th>N</th><th>dwss</th><th>dn</th><th>b</th><th>di</th><th>Pug</th><th>as</th><th>gon</th><th>| 08</th><th>BORT</th><th>PR</th></thead><tr><td>base</td><td>| 6</td><td>512</td><td>2048</td><td>8</td><td>64</td><td>0.1</td><td>01</td><td>100K</td><td>| 492</td><td>258</td><td>65</td></tr><tr><td></td><td></td><td></td><td></td><td>1</td><td>512</td><td></td><td></td><td></td><td>529</td><td>249</td><td></td></tr><tr><td></td><td></td><td></td><td></td><td>4</td><td>128</td><td></td><td></td><td></td><td>500</td><td>255</td><td></td></tr><tr><td>(A)</td><td></td><td></td><td></td><td>16</td><td>32</td><td></td><td></td><td></td><td>491</td><td>258</td><td></td></tr><tr><td></td><td></td><td></td><td></td><td>32</td><td>16</td><td></td><td></td><td></td><td>501</td><td>254</td><td></td></tr><tr><td></td><td></td><td></td><td></td><td></td><td>16</td><td></td><td></td><td></td><td>516</td><td>251</td><td>58</td></tr><tr><td>®)</td><td></td><td></td><td></td><td></td><td>32</td><td></td><td></td><td></td><td>501</td><td>254</td><td>60</td></tr><tr><td rowspan=\"7\">©)</td><td>2</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td>6.11</td><td>237</td><td>36</td></tr><tr><td></td><td>4</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td>519</td><td>253</td><td>50</td></tr><tr><td></td><td>8</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td>488</td><td>255</td><td>80</td></tr><tr><td></td><td></td><td>256</td><td></td><td></td><td>32</td><td></td><td></td><td></td><td>575</td><td>245</td><td>28</td></tr><tr><td></td><td></td><td>1024</td><td></td><td></td><td>128</td><td></td><td></td><td></td><td>4.66</td><td>26.0</td><td>168</td></tr><tr><td></td><td></td><td></td><td>1024</td><td></td><td></td><td></td><td></td><td></td><td>512</td><td>254</td><td>53</td></tr><tr><td></td><td></td><td></td><td>4096</td><td></td><td></td><td></td><td></td><td></td><td>475</td><td>262</td><td>90</td></tr><tr><td></td><td></td><td></td><td></td><td></td><td></td><td>0.0</td><td></td><td></td><td>577</td><td>246</td><td></td></tr><tr><td></td><td></td><td></td><td></td><td></td><td></td><td>0.2</td><td></td><td></td><td>495</td><td>255</td><td></td></tr><tr><td>D</td><td></td><td></td><td></td><td></td><td></td><td></td><td>0.0</td><td></td><td>461</td><td>253</td><td></td></tr><tr><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td>0.2</td><td></td><td>547</td><td>257</td><td></td></tr><tr><td>(E)</td><td></td><td></td><td>positional</td><td>embedding</td><td>instead</td><td>of sinusoids</td><td></td><td></td><td>4.92</td><td>25.7</td><td></td></tr><tr><td>big</td><td>| 6</td><td>1024</td><td>4096</td><td>16</td><td></td><td>0.3</td><td></td><td>300K</td><td>| 433</td><td>264</td><td>213</td></tr></table>"
+       "<table><thead><th></th><th>N</th><th>dwss</th><th>dn</th><th>b</th><th>di</th><th>Pug</th><th>as</th><th>gon</th><th>| 08</th><th>BORT</th><th>PR</th></thead><tr><td>base</td><td>| 6</td><td>512</td><td>2048</td><td>8</td><td>64</td><td>0.1</td><td>01</td><td>100K</td><td>| 492</td><td>258</td><td>65</td></tr><tr><td></td><td></td><td></td><td></td><td>1</td><td>512</td><td></td><td></td><td></td><td>529</td><td>249</td><td></td></tr><tr><td></td><td></td><td></td><td></td><td>4</td><td>128</td><td></td><td></td><td></td><td>500</td><td>255</td><td></td></tr><tr><td>(A)</td><td></td><td></td><td></td><td>16</td><td>32</td><td></td><td></td><td></td><td>491</td><td>258</td><td></td></tr><tr><td></td><td></td><td></td><td></td><td>32</td><td>16</td><td></td><td></td><td></td><td>501</td><td>254</td><td></td></tr><tr><td></td><td></td><td></td><td></td><td></td><td>16</td><td></td><td></td><td></td><td>516</td><td>251</td><td>58</td></tr><tr><td>\u00ae)</td><td></td><td></td><td></td><td></td><td>32</td><td></td><td></td><td></td><td>501</td><td>254</td><td>60</td></tr><tr><td rowspan=\"7\">\u00a9)</td><td>2</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td>6.11</td><td>237</td><td>36</td></tr><tr><td></td><td>4</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td>519</td><td>253</td><td>50</td></tr><tr><td></td><td>8</td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td>488</td><td>255</td><td>80</td></tr><tr><td></td><td></td><td>256</td><td></td><td></td><td>32</td><td></td><td></td><td></td><td>575</td><td>245</td><td>28</td></tr><tr><td></td><td></td><td>1024</td><td></td><td></td><td>128</td><td></td><td></td><td></td><td>4.66</td><td>26.0</td><td>168</td></tr><tr><td></td><td></td><td></td><td>1024</td><td></td><td></td><td></td><td></td><td></td><td>512</td><td>254</td><td>53</td></tr><tr><td></td><td></td><td></td><td>4096</td><td></td><td></td><td></td><td></td><td></td><td>475</td><td>262</td><td>90</td></tr><tr><td></td><td></td><td></td><td></td><td></td><td></td><td>0.0</td><td></td><td></td><td>577</td><td>246</td><td></td></tr><tr><td></td><td></td><td></td><td></td><td></td><td></td><td>0.2</td><td></td><td></td><td>495</td><td>255</td><td></td></tr><tr><td>D</td><td></td><td></td><td></td><td></td><td></td><td></td><td>0.0</td><td></td><td>461</td><td>253</td><td></td></tr><tr><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td>0.2</td><td></td><td>547</td><td>257</td><td></td></tr><tr><td>(E)</td><td></td><td></td><td>positional</td><td>embedding</td><td>instead</td><td>of sinusoids</td><td></td><td></td><td>4.92</td><td>25.7</td><td></td></tr><tr><td>big</td><td>| 6</td><td>1024</td><td>4096</td><td>16</td><td></td><td>0.3</td><td></td><td>300K</td><td>| 433</td><td>264</td><td>213</td></tr></table>"
       ],
       "text/plain": [
        "<IPython.core.display.HTML object>"
@@ -269,10 +275,10 @@
      "text": [
       "development set, newstest2013. We used beam search as described in the previous section, but no checkpoint averaging. We present these results in Table 3.\n",
       "In Table 3 rows (A), we vary the number of attention heads and the attention key and value dimensions, keeping the amount of computation constant, as described in Section 3.2.2. While single-head attention is 0.9 BLEU worse than the best setting, quality also drops off with too many heads.\n",
-      "In Table 3 rows (B), we observe that reducing the attention key size dk hurts model quality. This suggests that determining compatibility is not easy and that a more sophisticated compatibility function than dot product may be beneficial. We further observe in rows (C) and (D) that, as expected, bigger models are better, and dropout is very helpful in avoiding over-fitting. In row (E) we replace our sinusoidal positional encoding with learned positional embeddings [9], and observe nearly identical results to the base model.\n",
+      "In Table 3 rows (B), we observe that reducing the attention key size dk hurts model quality. This suggests that determining compatibility is not easy and that a more sophisticated compatibility function than dot product may be bene\ufb01cial. We further observe in rows (C) and (D) that, as expected, bigger models are better, and dropout is very helpful in avoiding over-\ufb01tting. In row (E) we replace our sinusoidal positional encoding with learned positional embeddings [9], and observe nearly identical results to the base model.\n",
       "6.3 English Constituency Parsing\n",
-      "parent: '6.3 English Constituency Parsing' content: To evaluate if the Transformer can generalize to other tasks we performed experiments on English constituency parsing. This task presents specific challenges: the output is subject to strong structural constraints and is significantly longer than the input. Furthermore, RNN sequence-to-sequence models have not been able to attain state-of-the-art results in small-data regimes [37].\n",
-      "parent: '6.3 English Constituency Parsing' content: We trained a 4-layer transformer with dmodel = 1024 on the Wall Street Journal (WSJ) portion of the Penn Treebank [25], about 40K training sentences. We also trained it in a semi-supervised setting, using the larger high-confidence and BerkleyParser corpora from with approximately 17M sentences [37]. We used a vocabulary of 16K tokens for the WSJ only setting and a vocabulary of 32K tokens for the semi-supervised setting.\n",
+      "parent: '6.3 English Constituency Parsing' content: To evaluate if the Transformer can generalize to other tasks we performed experiments on English constituency parsing. This task presents speci\ufb01c challenges: the output is subject to strong structural constraints and is signi\ufb01cantly longer than the input. Furthermore, RNN sequence-to-sequence models have not been able to attain state-of-the-art results in small-data regimes [37].\n",
+      "parent: '6.3 English Constituency Parsing' content: We trained a 4-layer transformer with dmodel = 1024 on the Wall Street Journal (WSJ) portion of the Penn Treebank [25], about 40K training sentences. We also trained it in a semi-supervised setting, using the larger high-con\ufb01dence and BerkleyParser corpora from with approximately 17M sentences [37]. We used a vocabulary of 16K tokens for the WSJ only setting and a vocabulary of 32K tokens for the semi-supervised setting.\n",
       "parent: '6.3 English Constituency Parsing' content: We performed only a small number of experiments to select the dropout, both attention and residual (section 5.4), learning rates and beam size on the Section 22 development set, all other parameters remained unchanged from the English-to-German base translation model. During inference, we\n",
       "9\n",
       "Table 4: The Transformer generalizes well to English constituency parsing (Results are on Section 23 of WSJ)\n"
@@ -294,13 +300,13 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "increased the maximum output length to input length + 300. We used a beam size of 21 and ↵ = 0.3 for both WSJ only and the semi-supervised setting.\n",
-      "Our results in Table 4 show that despite the lack of task-specific tuning our model performs sur- prisingly well, yielding better results than all previously reported models with the exception of the Recurrent Neural Network Grammar [8].\n",
+      "increased the maximum output length to input length + 300. We used a beam size of 21 and \u21b5 = 0.3 for both WSJ only and the semi-supervised setting.\n",
+      "Our results in Table 4 show that despite the lack of task-speci\ufb01c tuning our model performs sur- prisingly well, yielding better results than all previously reported models with the exception of the Recurrent Neural Network Grammar [8].\n",
       "In contrast to RNN sequence-to-sequence models [37], the Transformer outperforms the Berkeley- Parser [29] even when training only on the WSJ training set of 40K sentences.\n",
       "7 Conclusion\n",
-      "parent: '7 Conclusion' content: In this work, we presented the Transformer, the first sequence transduction model based entirely on attention, replacing the recurrent layers most commonly used in encoder-decoder architectures with multi-headed self-attention.\n",
-      "parent: '7 Conclusion' content: For translation tasks, the Transformer can be trained significantly faster than architectures based on recurrent or convolutional layers. On both WMT 2014 English-to-German and WMT 2014 English-to-French translation tasks, we achieve a new state of the art. In the former task our best model outperforms even all previously reported ensembles.\n",
-      "parent: '7 Conclusion' content: We are excited about the future of attention-based models and plan to apply them to other tasks. We plan to extend the Transformer to problems involving input and output modalities other than text and to investigate local, restricted attention mechanisms to efficiently handle large inputs and outputs such as images, audio and video. Making generation less sequential is another research goals of ours.\n",
+      "parent: '7 Conclusion' content: In this work, we presented the Transformer, the \ufb01rst sequence transduction model based entirely on attention, replacing the recurrent layers most commonly used in encoder-decoder architectures with multi-headed self-attention.\n",
+      "parent: '7 Conclusion' content: For translation tasks, the Transformer can be trained signi\ufb01cantly faster than architectures based on recurrent or convolutional layers. On both WMT 2014 English-to-German and WMT 2014 English-to-French translation tasks, we achieve a new state of the art. In the former task our best model outperforms even all previously reported ensembles.\n",
+      "parent: '7 Conclusion' content: We are excited about the future of attention-based models and plan to apply them to other tasks. We plan to extend the Transformer to problems involving input and output modalities other than text and to investigate local, restricted attention mechanisms to ef\ufb01ciently handle large inputs and outputs such as images, audio and video. Making generation less sequential is another research goals of ours.\n",
       "parent: '7 Conclusion' content: The code we used to train and evaluate our models is available at https://github.com/ tensorflow/tensor2tensor.\n",
       "parent: '7 Conclusion' content: Acknowledgements We are grateful to Nal Kalchbrenner and Stephan Gouws for their fruitful comments, corrections and inspiration.\n",
       "parent: '7 Conclusion' content: References [1] Jimmy Lei Ba, Jamie Ryan Kiros, and Geoffrey E Hinton. Layer normalization. arXiv preprint\n",
@@ -557,4 +563,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}

examples/notebooks/langchain_evaluation.ipynb

@@ -47,7 +47,14 @@
   {
    "cell_type": "code",
    "execution_count": 3,
-   "metadata": {},
+   "metadata": {
+    "nbmake": {
+     "post_cell_execute": [
+      "from conftest import before_notebook",
+      "before_notebook()"
+     ]
+    }
+   },
    "outputs": [],
    "source": [
     "! pip install -q ragstack-ai"
@@ -648,4 +655,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}