Skip to content

Commit b2d2b57

Browse files
cmurtz-msftcmurtz-msft
authored
[Azure] Chat completions example (openai#271)
* Added notebook for chat completions on Azure * Added pointer from README
1 parent 2eb3cc1 commit b2d2b57

File tree

3 files changed

+278
-5
lines changed

3 files changed

+278
-5
lines changed

‎README.md

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -52,6 +52,7 @@ Most code examples are written in Python, though the concepts can be applied in
5252
- DALL-E
5353
- [How to generate and edit images with DALL-E](examples/dalle/Image_generations_edits_and_variations_with_DALL-E.ipynb)
5454
- Azure OpenAI (alternative API from Microsoft Azure)
55+
- [How to use ChatGPT with Azure OpenAI](examples/azure/chat.ipynb)
5556
- [How to get completions from Azure OpenAI](examples/azure/completions.ipynb)
5657
- [How to get embeddings from Azure OpenAI](examples/azure/embeddings.ipynb)
5758
- [How to fine-tune GPT-3 with Azure OpenAI](examples/azure/finetuning.ipynb)

‎examples/azure/chat.ipynb

Lines changed: 273 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,273 @@
1+
{
2+
"cells": [
3+
{
4+
"attachments": {},
5+
"cell_type": "markdown",
6+
"metadata": {},
7+
"source": [
8+
"# Azure chat completions example (preview)\n",
9+
"In this example we'll try to go over all operations needed to get chat completions working using the Azure endpoints. \\\n",
10+
"This example focuses on chat completions but also touches on some other operations that are also available using the API. This example is meant to be a quick way of showing simple operations and is not meant as a tutorial."
11+
]
12+
},
13+
{
14+
"cell_type": "code",
15+
"execution_count": null,
16+
"metadata": {},
17+
"outputs": [],
18+
"source": [
19+
"import openai"
20+
]
21+
},
22+
{
23+
"attachments": {},
24+
"cell_type": "markdown",
25+
"metadata": {},
26+
"source": [
27+
"## Setup\n",
28+
"For the following sections to work properly we first have to setup some things. Let's start with the `api_base` and `api_version`. To find your `api_base` go to https://portal.azure.com, find your resource and then under \"Resource Management\" -> \"Keys and Endpoints\" look for the \"Endpoint\" value."
29+
]
30+
},
31+
{
32+
"cell_type": "code",
33+
"execution_count": null,
34+
"metadata": {},
35+
"outputs": [],
36+
"source": [
37+
"openai.api_version = '2023-03-15-preview'\n",
38+
"openai.api_base = '' # Please add your endpoint here"
39+
]
40+
},
41+
{
42+
"cell_type": "markdown",
43+
"metadata": {},
44+
"source": [
45+
"We next have to setup the `api_type` and `api_key`. We can either get the key from the portal or we can get it through Microsoft Active Directory Authentication. Depending on this the `api_type` is either `azure` or `azure_ad`."
46+
]
47+
},
48+
{
49+
"attachments": {},
50+
"cell_type": "markdown",
51+
"metadata": {},
52+
"source": [
53+
"### Setup: Portal\n",
54+
"Let's first look at getting the key from the portal. Go to https://portal.azure.com, find your resource and then under \"Resource Management\" -> \"Keys and Endpoints\" look for one of the \"Keys\" values."
55+
]
56+
},
57+
{
58+
"cell_type": "code",
59+
"execution_count": null,
60+
"metadata": {},
61+
"outputs": [],
62+
"source": [
63+
"openai.api_type = 'azure'\n",
64+
"openai.api_key = '' # Please add your api key here"
65+
]
66+
},
67+
{
68+
"attachments": {},
69+
"cell_type": "markdown",
70+
"metadata": {},
71+
"source": [
72+
"### (Optional) Setup: Microsoft Active Directory Authentication\n",
73+
"Let's now see how we can get a key via Microsoft Active Directory Authentication. Uncomment the following code if you want to use Active Directory Authentication instead of keys from the portal."
74+
]
75+
},
76+
{
77+
"cell_type": "code",
78+
"execution_count": null,
79+
"metadata": {},
80+
"outputs": [],
81+
"source": [
82+
"# from azure.identity import DefaultAzureCredential\n",
83+
"\n",
84+
"# default_credential = DefaultAzureCredential()\n",
85+
"# token = default_credential.get_token(\"https://cognitiveservices.azure.com/.default\")\n",
86+
"\n",
87+
"# openai.api_type = 'azure_ad'\n",
88+
"# openai.api_key = token.token"
89+
]
90+
},
91+
{
92+
"attachments": {},
93+
"cell_type": "markdown",
94+
"metadata": {},
95+
"source": [
96+
"## Deployments\n",
97+
"In this section we are going to create a deployment using the `gpt-35-turbo` model that we can then use to create chat completions."
98+
]
99+
},
100+
{
101+
"attachments": {},
102+
"cell_type": "markdown",
103+
"metadata": {},
104+
"source": [
105+
"### Deployments: Create manually\n",
106+
"Let's create a deployment using the `gpt-35-turbo` model. Go to https://portal.azure.com, find your resource and then under \"Resource Management\" -> \"Model deployments\" create a new `gpt-35-turbo` deployment. "
107+
]
108+
},
109+
{
110+
"cell_type": "code",
111+
"execution_count": null,
112+
"metadata": {},
113+
"outputs": [],
114+
"source": [
115+
"deployment_id = \"\" # Fill in the deployment id from the portal here"
116+
]
117+
},
118+
{
119+
"attachments": {},
120+
"cell_type": "markdown",
121+
"metadata": {},
122+
"source": [
123+
"### (Optional) Deployments: Create programatically\n",
124+
"We can also create the deployment using code. Note that you can only create one deployment per model."
125+
]
126+
},
127+
{
128+
"cell_type": "code",
129+
"execution_count": null,
130+
"metadata": {},
131+
"outputs": [],
132+
"source": [
133+
"model = \"gpt-35-turbo\"\n",
134+
"\n",
135+
"# Now let's create the deployment\n",
136+
"print(f'Creating a new deployment with model: {model}')\n",
137+
"result = openai.Deployment.create(model=model, scale_settings={\"scale_type\":\"standard\"})\n",
138+
"deployment_id = result[\"id\"]\n",
139+
"print(f'Successfully created deployment with id: {deployment_id}')"
140+
]
141+
},
142+
{
143+
"cell_type": "markdown",
144+
"metadata": {},
145+
"source": [
146+
"### (Optional) Deployments: Wait for deployment to succeed\n",
147+
"Now let's check the status of the newly created deployment and wait till it is succeeded."
148+
]
149+
},
150+
{
151+
"cell_type": "code",
152+
"execution_count": null,
153+
"metadata": {},
154+
"outputs": [],
155+
"source": [
156+
"print(f'Checking for deployment status.')\n",
157+
"resp = openai.Deployment.retrieve(id=deployment_id)\n",
158+
"status = resp[\"status\"]\n",
159+
"print(f'Deployment {deployment_id} has status: {status}')\n",
160+
"while status not in [\"succeeded\", \"failed\"]:\n",
161+
" resp = openai.Deployment.retrieve(id=deployment_id)\n",
162+
" status = resp[\"status\"]\n",
163+
" print(f'Deployment {deployment_id} has status: {status}')"
164+
]
165+
},
166+
{
167+
"attachments": {},
168+
"cell_type": "markdown",
169+
"metadata": {},
170+
"source": [
171+
"### Create chat completion\n",
172+
"Now let's send a sample chat completion to the deployment."
173+
]
174+
},
175+
{
176+
"cell_type": "code",
177+
"execution_count": null,
178+
"metadata": {},
179+
"outputs": [],
180+
"source": [
181+
"# For all possible arguments see https://platform.openai.com/docs/api-reference/chat-completions/create\n",
182+
"response = openai.ChatCompletion.create(\n",
183+
" deployment_id=deployment_id,\n",
184+
" messages=[\n",
185+
" {\"role\": \"system\", \"content\": \"You are a helpful assistant.\"},\n",
186+
" {\"role\": \"user\", \"content\": \"Knock knock.\"},\n",
187+
" {\"role\": \"assistant\", \"content\": \"Who's there?\"},\n",
188+
" {\"role\": \"user\", \"content\": \"Orange.\"},\n",
189+
" ],\n",
190+
" temperature=0,\n",
191+
")\n",
192+
"\n",
193+
"print(f\"{response.choices[0].message.role}: {response.choices[0].message.content}\")"
194+
]
195+
},
196+
{
197+
"attachments": {},
198+
"cell_type": "markdown",
199+
"metadata": {},
200+
"source": [
201+
"We can also stream the response.\n"
202+
]
203+
},
204+
{
205+
"cell_type": "code",
206+
"execution_count": null,
207+
"metadata": {},
208+
"outputs": [],
209+
"source": [
210+
"response = openai.ChatCompletion.create(\n",
211+
" deployment_id=deployment_id,\n",
212+
" messages=[\n",
213+
" {\"role\": \"system\", \"content\": \"You are a helpful assistant.\"},\n",
214+
" {\"role\": \"user\", \"content\": \"Knock knock.\"},\n",
215+
" {\"role\": \"assistant\", \"content\": \"Who's there?\"},\n",
216+
" {\"role\": \"user\", \"content\": \"Orange.\"},\n",
217+
" ],\n",
218+
" temperature=0,\n",
219+
" stream=True\n",
220+
")\n",
221+
"\n",
222+
"for chunk in response:\n",
223+
" delta = chunk.choices[0].delta\n",
224+
"\n",
225+
" if \"role\" in delta.keys():\n",
226+
" print(delta.role + \": \", end=\"\", flush=True)\n",
227+
" if \"content\" in delta.keys():\n",
228+
" print(delta.content, end=\"\", flush=True)"
229+
]
230+
},
231+
{
232+
"attachments": {},
233+
"cell_type": "markdown",
234+
"metadata": {},
235+
"source": [
236+
"### (Optional) Deployments: Delete\n",
237+
"Finally let's delete the deployment."
238+
]
239+
},
240+
{
241+
"cell_type": "code",
242+
"execution_count": null,
243+
"metadata": {},
244+
"outputs": [],
245+
"source": [
246+
"print(f'Deleting deployment: {deployment_id}')\n",
247+
"openai.Deployment.delete(sid=deployment_id)"
248+
]
249+
}
250+
],
251+
"metadata": {
252+
"kernelspec": {
253+
"display_name": "Python 3",
254+
"language": "python",
255+
"name": "python3"
256+
},
257+
"language_info": {
258+
"codemirror_mode": {
259+
"name": "ipython",
260+
"version": 3
261+
},
262+
"file_extension": ".py",
263+
"mimetype": "text/x-python",
264+
"name": "python",
265+
"nbconvert_exporter": "python",
266+
"pygments_lexer": "ipython3",
267+
"version": "3.8.10"
268+
},
269+
"orig_nbformat": 4
270+
},
271+
"nbformat": 4,
272+
"nbformat_minor": 2
273+
}

‎examples/azure/completions.ipynb

Lines changed: 4 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -11,12 +11,11 @@
1111
},
1212
{
1313
"cell_type": "code",
14-
"execution_count": null,
14+
"execution_count": 2,
1515
"metadata": {},
1616
"outputs": [],
1717
"source": [
18-
"import openai\n",
19-
"from openai import cli"
18+
"import openai"
2019
]
2120
},
2221
{
@@ -29,7 +28,7 @@
2928
},
3029
{
3130
"cell_type": "code",
32-
"execution_count": null,
31+
"execution_count": 3,
3332
"metadata": {},
3433
"outputs": [],
3534
"source": [
@@ -207,7 +206,7 @@
207206
"name": "python",
208207
"nbconvert_exporter": "python",
209208
"pygments_lexer": "ipython3",
210-
"version": "3.10.8"
209+
"version": "3.8.10"
211210
},
212211
"vscode": {
213212
"interpreter": {

0 commit comments

Comments
 (0)