prompt-poet

Name: prompt-poet
Author: character-ai

1.2kstars95forks

How to install prompt-poet?

character-ai/prompt-poet

$git clone https://github.com/character-ai/prompt-poet

Installs into the current project.

Summary

Prompt Poet streamlines and simplifies prompt design for both developers and non-technical users with its low code approach. Using a mix of YAML and Jinja2, Prompt Poet allows for flexible, dynamic prompt creation, enhancing the efficiency and quality of interactions with AI mode

Files · 1

View on GitHub

README.md

1	[![Downloads](https://static.pepy.tech/badge/prompt-poet)](https://pepy.tech/project/prompt-poet)
2
3	# Prompt Poet
4
5	Prompt Poet streamlines and simplifies prompt design for both developers and non-technical users with its low code approach. Using a mix of YAML and Jinja2, Prompt Poet allows for flexible, dynamic prompt creation, enhancing the efficiency and quality of interactions with AI models. It saves time on engineering string manipulations, enabling everyone to focus more on crafting the optimal prompts for their users.
6
7	### Installation
8
9	```shell
10	pip install prompt-poet
11	```
12
13	### Basic Usage
14
15	```python
16	import os
17	import getpass
18
19	from prompt_poet import Prompt
20	from langchain import ChatOpenAI
21
22	# Uncomment if you need to set OPENAI_API_KEY.
23	# os.environ["OPENAI_API_KEY"] = getpass.getpass()
24
25	raw_template = """
26	- name: system instructions
27	role: system
28	content: \|
29	Your name is {{ character_name }} and you are meant to be helpful and never harmful to humans.
30
31	- name: user query
32	role: user
33	content: \|
34	{{ username}}: {{ user_query }}
35
36	- name: response
37	role: user
38	content: \|
39	{{ character_name }}:
40	"""
41
42	template_data = {
43	"character_name": "Character Assistant",
44	"username": "Jeff",
45	"user_query": "Can you help me with my homework?"
46	}
47
48	prompt = Prompt(
49	raw_template=raw_template,
50	template_data=template_data
51	)
52
53	model = ChatOpenAI(model="gpt-4o-mini")
54	response = model.invoke(prompt.messages)
55	```
56
57	### Prompt Templates
58	Prompt Poet templates use a mix of YAML and Jinja2. Template processing occurs in two primary stages:
59
60	- Rendering: Initially, Jinja2 processes the input data. During this phase, control flow logic is executed, data is validated and appropriately bound to variables, and functions within the template are appropriately evaluated.
61	- Loading: Post-rendering, the output is a structured YAML file. This YAML structure consists of repeated blocks or parts, each encapsulated into a Python data structure. These parts are characterized by several attributes:
62	- Name: A clear, human-readable identifier for the part.
63	- Content: The actual string payload that forms part of the prompt.
64	- Role (Optional): Specifies the role of the participant, aiding in distinguishing between different users or system components.
65	- Truncation Priority (Optional): Determines the order of truncation when necessary, with parts having the same priority being truncated in the order in which they appear.
66
67	#### Example: Basic Q&A Bot
68	```yaml
69	- name: system instructions
70	role: system
71	content: \|
72	Your name is {{ character_name }} and you are meant to be helpful and never harmful to humans.
73
74	- name: user query
75	role: user
76	content: \|
77	{{ username}}: {{ user_query }}
78
79	- name: reply_prompt
80	role: user
81	content: \|
82	{{ character_name }}:
83	```
84
85	#### Interpolating Lists
86	If you have elements (e.g. messages) in a list you can parse them into your template like so.
87	```yaml
88	{% for message in current_chat_messages %}
89	- name: chat_message
90	role: user
91	content: \|
92	{{ message.author }}: {{ message.content }}
93	{% endfor %}
94	```
95
96	#### Truncating Old Messages
97	Context length is limited and can’t always fit the entire chat history– so we can set a truncation priority on the message parts and Prompt Poet will truncate these parts in the order in which they appear (oldest to newest).
98	```yaml
99	{% for message in current_chat_messages %}
100	- name: chat_message
101	role: user
102	truncation_priority: 1
103	content: \|
104	{{ message.author }}: {{ message.content }}
105	{% endfor %}
106	```
107
108	#### Adapting to User Modality
109	To tailor instructions based on the user's current modality (audio or text).
110	```yaml
111	{% if modality == "audio" %}
112	- name: special audio instruction
113	role: system
114	content: \|
115	{{ username }} is currently using audio. Keep your answers succinct.
116	{% endif %}
117	```
118
119	#### Targeting Specific Queries
120	To include context-specific examples like homework help when needed.
121	```yaml
122	{% if extract_user_query_topic(user_query) == "homework_help" %}
123	{

People who liked this also viewed

6 picks

Type

prompt-poet

1.2kstars95forks

View on GitHub

How to install prompt-poet?

character-ai/prompt-poet

$git clone https://github.com/character-ai/prompt-poet

Installs into the current project.

Summary

Files · 1

View on GitHub

README.md

1	[![Downloads](https://static.pepy.tech/badge/prompt-poet)](https://pepy.tech/project/prompt-poet)
2
3	# Prompt Poet
4
5	Prompt Poet streamlines and simplifies prompt design for both developers and non-technical users with its low code approach. Using a mix of YAML and Jinja2, Prompt Poet allows for flexible, dynamic prompt creation, enhancing the efficiency and quality of interactions with AI models. It saves time on engineering string manipulations, enabling everyone to focus more on crafting the optimal prompts for their users.
6
7	### Installation
8
9	```shell
10	pip install prompt-poet
11	```
12
13	### Basic Usage
14
15	```python
16	import os
17	import getpass
18
19	from prompt_poet import Prompt
20	from langchain import ChatOpenAI
21
22	# Uncomment if you need to set OPENAI_API_KEY.
23	# os.environ["OPENAI_API_KEY"] = getpass.getpass()
24
25	raw_template = """
26	- name: system instructions
27	role: system
28	content: \|
29	Your name is {{ character_name }} and you are meant to be helpful and never harmful to humans.
30
31	- name: user query
32	role: user
33	content: \|
34	{{ username}}: {{ user_query }}
35
36	- name: response
37	role: user
38	content: \|
39	{{ character_name }}:
40	"""
41
42	template_data = {
43	"character_name": "Character Assistant",
44	"username": "Jeff",
45	"user_query": "Can you help me with my homework?"
46	}
47
48	prompt = Prompt(
49	raw_template=raw_template,
50	template_data=template_data
51	)
52
53	model = ChatOpenAI(model="gpt-4o-mini")
54	response = model.invoke(prompt.messages)
55	```
56
57	### Prompt Templates
58	Prompt Poet templates use a mix of YAML and Jinja2. Template processing occurs in two primary stages:
59
60	- Rendering: Initially, Jinja2 processes the input data. During this phase, control flow logic is executed, data is validated and appropriately bound to variables, and functions within the template are appropriately evaluated.
61	- Loading: Post-rendering, the output is a structured YAML file. This YAML structure consists of repeated blocks or parts, each encapsulated into a Python data structure. These parts are characterized by several attributes:
62	- Name: A clear, human-readable identifier for the part.
63	- Content: The actual string payload that forms part of the prompt.
64	- Role (Optional): Specifies the role of the participant, aiding in distinguishing between different users or system components.
65	- Truncation Priority (Optional): Determines the order of truncation when necessary, with parts having the same priority being truncated in the order in which they appear.
66
67	#### Example: Basic Q&A Bot
68	```yaml
69	- name: system instructions
70	role: system
71	content: \|
72	Your name is {{ character_name }} and you are meant to be helpful and never harmful to humans.
73
74	- name: user query
75	role: user
76	content: \|
77	{{ username}}: {{ user_query }}
78
79	- name: reply_prompt
80	role: user
81	content: \|
82	{{ character_name }}:
83	```
84
85	#### Interpolating Lists
86	If you have elements (e.g. messages) in a list you can parse them into your template like so.
87	```yaml
88	{% for message in current_chat_messages %}
89	- name: chat_message
90	role: user
91	content: \|
92	{{ message.author }}: {{ message.content }}
93	{% endfor %}
94	```
95
96	#### Truncating Old Messages
97	Context length is limited and can’t always fit the entire chat history– so we can set a truncation priority on the message parts and Prompt Poet will truncate these parts in the order in which they appear (oldest to newest).
98	```yaml
99	{% for message in current_chat_messages %}
100	- name: chat_message
101	role: user
102	truncation_priority: 1
103	content: \|
104	{{ message.author }}: {{ message.content }}
105	{% endfor %}
106	```
107
108	#### Adapting to User Modality
109	To tailor instructions based on the user's current modality (audio or text).
110	```yaml
111	{% if modality == "audio" %}
112	- name: special audio instruction
113	role: system
114	content: \|
115	{{ username }} is currently using audio. Keep your answers succinct.
116	{% endif %}
117	```
118
119	#### Targeting Specific Queries
120	To include context-specific examples like homework help when needed.
121	```yaml
122	{% if extract_user_query_topic(user_query) == "homework_help" %}
123	{

People who liked this also viewed

6 picks

Type