Skip to content

Commit 6158125

Browse files
author
brian-tecton-ai
authored
Add Tecton example to the "Connecting to a Feature Store" example notebook (langchain-ai#3626)
This PR adds a similar example to the Feast example, using the [Tecton Feature Platform](https://www.tecton.ai/) and features from the [Tecton Fundamentals Tutorial](https://docs.tecton.ai/docs/tutorials/tecton-fundamentals).
1 parent 3b7d27d commit 6158125

File tree

1 file changed

+239
-2
lines changed

1 file changed

+239
-2
lines changed

‎docs/modules/prompts/prompt_templates/examples/connecting_to_a_feature_store.ipynb

Lines changed: 239 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -17,7 +17,9 @@
1717
{
1818
"cell_type": "markdown",
1919
"id": "ad0b5edf",
20-
"metadata": {},
20+
"metadata": {
21+
"tags": []
22+
},
2123
"source": [
2224
"## Feast\n",
2325
"\n",
@@ -211,6 +213,241 @@
211213
"metadata": {},
212214
"outputs": [],
213215
"source": []
216+
},
217+
{
218+
"cell_type": "markdown",
219+
"id": "c4049990-651d-44d3-82b1-0cd122da55c1",
220+
"metadata": {},
221+
"source": [
222+
"## Tecton\n",
223+
"\n",
224+
"Above, we showed how you could use Feast, a popular open source and self-managed feature store, with LangChain. Our examples below will show a similar integration using Tecton. Tecton is a fully managed feature platform built to orchestrate the complete ML feature lifecycle, from transformation to online serving, with enterprise-grade SLAs."
225+
]
226+
},
227+
{
228+
"cell_type": "markdown",
229+
"id": "7bb4dba1-0678-4ea4-be0a-d353c0b13fc2",
230+
"metadata": {
231+
"tags": []
232+
},
233+
"source": [
234+
"### Prerequisites\n",
235+
"\n",
236+
"* Tecton Deployment (sign up at [https://tecton.ai](https://tecton.ai))\n",
237+
"* `TECTON_API_KEY` environment variable set to a valid Service Account key"
238+
]
239+
},
240+
{
241+
"cell_type": "markdown",
242+
"id": "ac9eb618-8c52-4cd6-bb8e-9c99a150dfa6",
243+
"metadata": {
244+
"tags": []
245+
},
246+
"source": [
247+
"### Define and Load Features\n",
248+
"\n",
249+
"We will use the user_transaction_counts Feature View from the [Tecton tutorial](https://docs.tecton.ai/docs/tutorials/tecton-fundamentals) as part of a Feature Service. For simplicity, we are only using a single Feature View; however, more sophisticated applications may require more feature views to retrieve the features needed for its prompt.\n",
250+
"\n",
251+
"```python\n",
252+
"user_transaction_metrics = FeatureService(\n",
253+
" name = \"user_transaction_metrics\",\n",
254+
" features = [user_transaction_counts]\n",
255+
")\n",
256+
"```\n",
257+
"\n",
258+
"The above Feature Service is expected to be [applied to a live workspace](https://docs.tecton.ai/docs/applying-feature-repository-changes-to-a-workspace). For this example, we will be using the \"prod\" workspace."
259+
]
260+
},
261+
{
262+
"cell_type": "code",
263+
"execution_count": 60,
264+
"id": "32e9675d-a7e5-429f-906f-2260294d3e46",
265+
"metadata": {
266+
"tags": []
267+
},
268+
"outputs": [],
269+
"source": [
270+
"import tecton\n",
271+
"\n",
272+
"workspace = tecton.get_workspace(\"prod\")\n",
273+
"feature_service = workspace.get_feature_service(\"user_transaction_metrics\")"
274+
]
275+
},
276+
{
277+
"cell_type": "markdown",
278+
"id": "29b7550c-0eb4-4bd1-a501-1c63fb77aa56",
279+
"metadata": {},
280+
"source": [
281+
"### Prompts\n",
282+
"\n",
283+
"Here we will set up a custom TectonPromptTemplate. This prompt template will take in a user_id , look up their stats, and format those stats into a prompt.\n",
284+
"\n",
285+
"Note that the input to this prompt template is just `user_id`, since that is the only user defined piece (all other variables are looked up inside the prompt template)."
286+
]
287+
},
288+
{
289+
"cell_type": "code",
290+
"execution_count": 61,
291+
"id": "6fb77ea4-64c6-4e48-a783-bd1ece021b82",
292+
"metadata": {
293+
"tags": []
294+
},
295+
"outputs": [],
296+
"source": [
297+
"from langchain.prompts import PromptTemplate, StringPromptTemplate"
298+
]
299+
},
300+
{
301+
"cell_type": "code",
302+
"execution_count": 77,
303+
"id": "02a98fbc-8135-4b11-bf60-85d28e426667",
304+
"metadata": {
305+
"tags": []
306+
},
307+
"outputs": [],
308+
"source": [
309+
"template = \"\"\"Given the vendor's up to date transaction stats, write them a note based on the following rules:\n",
310+
"\n",
311+
"1. If they had a transaction in the last day, write a short congratulations message on their recent sales\n",
312+
"2. If no transaction in the last day, but they had a transaction in the last 30 days, playfully encourage them to sell more.\n",
313+
"3. Always add a silly joke about chickens at the end\n",
314+
"\n",
315+
"Here are the vendor's stats:\n",
316+
"Number of Transactions Last Day: {transaction_count_1d}\n",
317+
"Number of Transactions Last 30 Days: {transaction_count_30d}\n",
318+
"\n",
319+
"Your response:\"\"\"\n",
320+
"prompt = PromptTemplate.from_template(template)"
321+
]
322+
},
323+
{
324+
"cell_type": "code",
325+
"execution_count": 78,
326+
"id": "a35cdfd5-6ccc-4394-acfe-60d53804be51",
327+
"metadata": {
328+
"tags": []
329+
},
330+
"outputs": [],
331+
"source": [
332+
"class TectonPromptTemplate(StringPromptTemplate):\n",
333+
" \n",
334+
" def format(self, **kwargs) -> str:\n",
335+
" user_id = kwargs.pop(\"user_id\")\n",
336+
" feature_vector = feature_service.get_online_features(join_keys={\"user_id\": user_id}).to_dict()\n",
337+
" kwargs[\"transaction_count_1d\"] = feature_vector[\"user_transaction_counts.transaction_count_1d_1d\"]\n",
338+
" kwargs[\"transaction_count_30d\"] = feature_vector[\"user_transaction_counts.transaction_count_30d_1d\"]\n",
339+
" return prompt.format(**kwargs)"
340+
]
341+
},
342+
{
343+
"cell_type": "code",
344+
"execution_count": 79,
345+
"id": "d5915df0-fb16-4770-8a82-22f885b74d1a",
346+
"metadata": {
347+
"tags": []
348+
},
349+
"outputs": [],
350+
"source": [
351+
"prompt_template = TectonPromptTemplate(input_variables=[\"user_id\"])"
352+
]
353+
},
354+
{
355+
"cell_type": "code",
356+
"execution_count": 80,
357+
"id": "a36abfc8-ea60-4ae0-a36d-d7b639c7307c",
358+
"metadata": {
359+
"tags": []
360+
},
361+
"outputs": [
362+
{
363+
"name": "stdout",
364+
"output_type": "stream",
365+
"text": [
366+
"Given the vendor's up to date transaction stats, write them a note based on the following rules:\n",
367+
"\n",
368+
"1. If they had a transaction in the last day, write a short congratulations message on their recent sales\n",
369+
"2. If no transaction in the last day, but they had a transaction in the last 30 days, playfully encourage them to sell more.\n",
370+
"3. Always add a silly joke about chickens at the end\n",
371+
"\n",
372+
"Here are the vendor's stats:\n",
373+
"Number of Transactions Last Day: 657\n",
374+
"Number of Transactions Last 30 Days: 20326\n",
375+
"\n",
376+
"Your response:\n"
377+
]
378+
}
379+
],
380+
"source": [
381+
"print(prompt_template.format(user_id=\"user_469998441571\"))"
382+
]
383+
},
384+
{
385+
"cell_type": "markdown",
386+
"id": "f8d4b905-1051-4303-9c33-8eddb65c1274",
387+
"metadata": {
388+
"tags": []
389+
},
390+
"source": [
391+
"### Use in a chain\n",
392+
"\n",
393+
"We can now use this in a chain, successfully creating a chain that achieves personalization backed by the Tecton Feature Platform"
394+
]
395+
},
396+
{
397+
"cell_type": "code",
398+
"execution_count": 81,
399+
"id": "ffb60cd0-8e3c-4c9d-b639-43d766e12c4c",
400+
"metadata": {
401+
"tags": []
402+
},
403+
"outputs": [],
404+
"source": [
405+
"from langchain.chat_models import ChatOpenAI\n",
406+
"from langchain.chains import LLMChain"
407+
]
408+
},
409+
{
410+
"cell_type": "code",
411+
"execution_count": 82,
412+
"id": "3918abc7-00b5-466f-bdfc-ab046cd282da",
413+
"metadata": {
414+
"tags": []
415+
},
416+
"outputs": [],
417+
"source": [
418+
"chain = LLMChain(llm=ChatOpenAI(), prompt=prompt_template)"
419+
]
420+
},
421+
{
422+
"cell_type": "code",
423+
"execution_count": 83,
424+
"id": "e7d91c4b-3e99-40cc-b3e9-a004c8c9193e",
425+
"metadata": {
426+
"tags": []
427+
},
428+
"outputs": [
429+
{
430+
"data": {
431+
"text/plain": [
432+
"'Wow, congratulations on your recent sales! Your business is really soaring like a chicken on a hot air balloon! Keep up the great work!'"
433+
]
434+
},
435+
"execution_count": 83,
436+
"metadata": {},
437+
"output_type": "execute_result"
438+
}
439+
],
440+
"source": [
441+
"chain.run(\"user_469998441571\")"
442+
]
443+
},
444+
{
445+
"cell_type": "code",
446+
"execution_count": null,
447+
"id": "f752b924-caf9-4f7a-b78b-cb8c8ada8c2e",
448+
"metadata": {},
449+
"outputs": [],
450+
"source": []
214451
}
215452
],
216453
"metadata": {
@@ -229,7 +466,7 @@
229466
"name": "python",
230467
"nbconvert_exporter": "python",
231468
"pygments_lexer": "ipython3",
232-
"version": "3.9.1"
469+
"version": "3.9.13"
233470
}
234471
},
235472
"nbformat": 4,

0 commit comments

Comments
 (0)