Skip to content

Commit

Permalink
Merge pull request #113 from cohere-ai/chat-rag-upd
Browse files Browse the repository at this point in the history 
Update chat rag notebooks
  • Loading branch information
@mrmer1
mrmer1 authored Jan 30, 2024
2 parents 07a6135 + 09c9854 commit 0b40861
Show file tree
Hide file tree
Showing 3 changed files with 161 additions and 121 deletions.
89 changes: 54 additions & 35 deletions examples/chat_rag_connector/RAG_Chatbot_with_Connectors.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
"cells": [
{
"cell_type": "code",
"execution_count": 2,
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
Expand All @@ -11,8 +11,7 @@
"import uuid\n",
"from typing import List, Dict\n",
"\n",
"COHERE_API_KEY = os.getenv(\"COHERE_API_KEY\")\n",
"co = cohere.Client(COHERE_API_KEY)"
"co = cohere.Client(\"COHERE_API_KEY\")"
]
},
{
Expand Down Expand Up @@ -79,7 +78,7 @@
},
{
"cell_type": "code",
"execution_count": 10,
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
Expand All @@ -98,7 +97,8 @@
" )\n",
"\n",
" for event in response:\n",
" yield event"
" yield event\n",
" yield response"
]
},
{
Expand All @@ -110,7 +110,7 @@
},
{
"cell_type": "code",
"execution_count": 11,
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
Expand All @@ -136,31 +136,33 @@
" # Print the chatbot response\n",
" print(\"Chatbot:\")\n",
" \n",
" documents = []\n",
" documents_flag = False\n",
" citations_flag = False\n",
" \n",
" for event in response:\n",
" # Documents\n",
" if event.event_type == \"search-results\":\n",
" documents_flag = True\n",
" documents = event.documents\n",
" \n",
" stream_type = type(event).__name__\n",
" \n",
" # Text\n",
" if event.event_type == \"text-generation\":\n",
" print(event.text, end=\"\") \n",
" if stream_type == \"StreamTextGeneration\":\n",
" print(event.text, end=\"\")\n",
"\n",
" # Citations\n",
" if event.event_type == \"citation-generation\":\n",
" if stream_type == \"StreamCitationGeneration\":\n",
" if not citations_flag:\n",
" print(\"\\n\\nCITATIONS:\")\n",
" citations_flag = True\n",
" print(event.citations)\n",
" \n",
" if documents_flag:\n",
" print(\"\\n\\nDOCUMENTS:\")\n",
" for d in documents:\n",
" print(f'{d[\"title\"]} ({d[\"id\"]}). URL: {d[\"url\"]}')\n",
" print(event.citations[0])\n",
" \n",
" # Documents\n",
" if citations_flag:\n",
" if stream_type == \"StreamingChat\":\n",
" print(\"\\n\\nDOCUMENTS:\")\n",
" documents = [{'id': doc['id'],\n",
" 'text': doc['text'][:50] + '...',\n",
" 'title': doc['title'],\n",
" 'url': doc['url']} \n",
" for doc in event.documents]\n",
" for doc in documents:\n",
" print(doc)\n",
"\n",
" print(f\"\\n{'-'*100}\\n\")"
]
Expand All @@ -174,30 +176,47 @@
},
{
"cell_type": "code",
"execution_count": 13,
"execution_count": 10,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"User: What is attention\n",
"User: What are sentence embeddings\n",
"Chatbot:\n",
"Sentence embeddings are the building blocks of language models. They associate each sentence with a vector (list of numbers) in a way that similar sentences are assigned similar vectors. These vectors are composed of numbers and carry important properties of the sentence. The embeddings act as a form of translation between languages as well, as they provide a relatable vector for similar sentences in different languages.\n",
"\n",
"CITATIONS:\n",
"{'start': 69, 'end': 124, 'text': 'associate each sentence with a vector (list of numbers)', 'document_ids': ['demo-conn-e5y5ps_0', 'demo-conn-e5y5ps_1', 'demo-conn-e5y5ps_2']}\n",
"{'start': 139, 'end': 186, 'text': 'similar sentences are assigned similar vectors.', 'document_ids': ['demo-conn-e5y5ps_0', 'demo-conn-e5y5ps_1']}\n",
"{'start': 235, 'end': 272, 'text': 'important properties of the sentence.', 'document_ids': ['demo-conn-e5y5ps_1', 'demo-conn-e5y5ps_2']}\n",
"\n",
"\n",
"DOCUMENTS:\n",
"{'id': 'demo-conn-e5y5ps_0', 'text': 'In the previous chapter, we learned that sentence ...', 'title': 'Similarity Between Words and Sentences', 'url': 'https://docs.cohere.com/docs/similarity-between-words-and-sentences'}\n",
"{'id': 'demo-conn-e5y5ps_1', 'text': 'This is where sentence embeddings come into play. ...', 'title': 'Text Embeddings', 'url': 'https://docs.cohere.com/docs/text-embeddings'}\n",
"{'id': 'demo-conn-e5y5ps_2', 'text': 'Sentence embeddings are even more powerful, as the...', 'title': 'Similarity Between Words and Sentences', 'url': 'https://docs.cohere.com/docs/similarity-between-words-and-sentences'}\n",
"\n",
"----------------------------------------------------------------------------------------------------\n",
"\n",
"User: How is it different from word embeddings\n",
"Chatbot:\n",
"Attention is a technique used in language models to provide context to each word in a sentence or text, based on the other words. Attention plays a crucial role in transformer models, which can help improve large language models.\n",
"The primary distinction between word embeddings and sentence embeddings is that the latter assigns a vector to every sentence whereas the former does the same thing but for individual words. \n",
"\n",
"Both embeddings are similar in the sense that they associate vectors in a way that similar items (words or sentences) are mapped to similar vectors. Word embeddings are a subset of sentence embeddings.\n",
"\n",
"CITATIONS:\n",
"[{'start': 60, 'end': 67, 'text': 'context', 'document_ids': ['demo-conn-tm17qr_0', 'demo-conn-tm17qr_1', 'demo-conn-tm17qr_2']}]\n",
"[{'start': 68, 'end': 102, 'text': 'to each word in a sentence or text', 'document_ids': ['demo-conn-tm17qr_1', 'demo-conn-tm17qr_2']}]\n",
"[{'start': 117, 'end': 129, 'text': 'other words.', 'document_ids': ['demo-conn-tm17qr_1']}]\n",
"[{'start': 148, 'end': 160, 'text': 'crucial role', 'document_ids': ['demo-conn-tm17qr_2']}]\n",
"[{'start': 164, 'end': 182, 'text': 'transformer models', 'document_ids': ['demo-conn-tm17qr_2']}]\n",
"[{'start': 199, 'end': 229, 'text': 'improve large language models.', 'document_ids': ['demo-conn-tm17qr_2']}]\n",
"{'start': 91, 'end': 125, 'text': 'assigns a vector to every sentence', 'document_ids': ['demo-conn-e5y5ps_0', 'demo-conn-e5y5ps_1']}\n",
"{'start': 165, 'end': 190, 'text': 'but for individual words.', 'document_ids': ['demo-conn-e5y5ps_0']}\n",
"{'start': 244, 'end': 261, 'text': 'associate vectors', 'document_ids': ['demo-conn-e5y5ps_0', 'demo-conn-e5y5ps_1']}\n",
"{'start': 315, 'end': 341, 'text': 'mapped to similar vectors.', 'document_ids': ['demo-conn-e5y5ps_0', 'demo-conn-e5y5ps_1']}\n",
"{'start': 342, 'end': 394, 'text': 'Word embeddings are a subset of sentence embeddings.', 'document_ids': ['demo-conn-e5y5ps_1']}\n",
"\n",
"\n",
"DOCUMENTS:\n",
"Transformer Models (demo-conn-tm17qr_0). URL: https://docs.cohere.com/docs/transformer-models\n",
"Transformer Models (demo-conn-tm17qr_1). URL: https://docs.cohere.com/docs/transformer-models\n",
"Transformer Models (demo-conn-tm17qr_2). URL: https://docs.cohere.com/docs/transformer-models\n",
"{'id': 'demo-conn-e5y5ps_0', 'text': 'In the previous chapters, you learned about word a...', 'title': 'The Attention Mechanism', 'url': 'https://docs.cohere.com/docs/the-attention-mechanism'}\n",
"{'id': 'demo-conn-e5y5ps_1', 'text': 'This is where sentence embeddings come into play. ...', 'title': 'Text Embeddings', 'url': 'https://docs.cohere.com/docs/text-embeddings'}\n",
"\n",
"----------------------------------------------------------------------------------------------------\n",
"\n",
Expand All @@ -207,7 +226,7 @@
],
"source": [
"# Define connectors\n",
"connectors = [\"demo-conn-tm17qr\"]\n",
"connectors = [\"demo-conn-e5y5ps\"]\n",
"\n",
"# Create an instance of the Chatbot class by supplying the connectors\n",
"chatbot = Chatbot(connectors)\n",
Expand Down
Loading

0 comments on commit 0b40861

Please sign in to comment.