sammo

Name: sammo
Author: microsoft

759stars45forks

How to install sammo?

microsoft/sammo

$git clone https://github.com/microsoft/sammo

Installs into the current project.

Summary

A flexible, easy-to-use library for running and optimizing prompts for Large Language Models (LLMs).

Files · 1

View on GitHub

README.md

1	# SAMMO ([📘User Guide](https://microsoft.github.io/sammo/))
2
3	[![Latest PyPI version](https://img.shields.io/pypi/v/sammo.svg)](https://pypi.python.org/pypi/sammo)
4	[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
5	[![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/microsoft/sammo/main?urlpath=tree/docs/tutorials/quickstart.ipynb)
6
7	A flexible, easy-to-use library for running and optimizing prompts for Large Language Models (LLMs).
8
9	## 🎉 News
10	- Nov 13, 2024: Turn Markdown into prompt programs: First version of SAMMO express released
11	- Nov 1, 2024: Use CSS selectors to query and modify prompt programs!
12	- Oct 15, 2024: SAMMO now supports structured outputs!
13
14	## How to Get Started
15	Go to the [user guide](https://microsoft.github.io/sammo/) for examples, how-tos, and API reference.
16
17	Just want to have a quick look? Try the [live demo on Binder](https://mybinder.org/v2/gh/microsoft/sammo/main?urlpath=tree/docs/tutorials/quickstart.ipynb).
18
19	<!--start-->
20	### Install library only
21
22	```bash
23	pip install sammo
24	```
25
26	### Install and run tutorials
27
28	*Prerequisites*
29	* Python 3.9+
30
31	The following commands will install sammo and jupyter and launch jupyter notebook. It's recommended that you create and activate a virtualenv prior to installing packages.
32
33	```bash
34	pip install sammo jupyter
35
36	# clone sammo to a local directory
37	git clone https://github.com/microsoft/sammo.git
38	cd sammo
39
40	# launch jupyter notebook and open tutorials directory
41	jupyter notebook --notebook-dir docs/tutorials
42	```
43
44	## Example
45	This example shows how easy it is to optimize a prompt with SAMMO. The full example is in the [user guide](https://microsoft.github.io/sammo/).
46	```python
47	runner = OpenAIChat(model_id="gpt-3.5-turbo", api_config=API_CONFIG)
48	PROMPT_IN_MARKDOWN = """
49	# Instructions <!-- #instr -->
50	Convert the following user queries into a SQL query.
51
52	# Table
53	Users:
54	- user_id (INTEGER, PRIMARY KEY)
55	- name (TEXT)
56	- age (INTEGER)
57	- city (TEXT)
58
59	# Complete this
60	Input: {{{input}}}
61	Output:
62	"""
63
64	spp = MarkdownParser(PROMPT_IN_MARKDOWN).get_sammo_program()
65	mutation_operators = BagOfMutators(
66	Output(GenerateText(spp)),
67	Paraphrase("#instr"),
68	Rewrite("#instr", "Make this more verbose.\n\n {{{{text}}}}")
69	)
70	prompt_optimizer = BeamSearch(runner, mutation_operators, accuracy)
71	prompt_optimizer.fit(d_train)
72	prompt_optimizer.show_report()
73	```
74
75	## Use Cases
76	![Overview](https://microsoft.github.io/sammo/_images/overview.png)
77
78	SAMMO is designed to support
79	- Efficient data labeling: Supports minibatching by packing and parsing multiple datapoints into a single prompt.
80	- Prompt prototyping and engineering: Re-usable components and prompt structures to quickly build and test new prompts.
81	- Instruction optimization: Optimize instructions to do better on a given task.
82	- Prompt compression: Compress prompts while maintaining performance.
83	- Large-scale prompt execution: parallelization
84	and rate-limiting out-of-the-box so you can run many queries in parallel and at scale without overwhelming the LLM API.
85
86	It is less useful if you want to build
87	- Interactive, agent-based LLM applications (→ check out [AutoGen](https://microsoft.github.io/autogen/))
88	- Interactive, production-ready LLM applications (→ check out [LangChain](https://www.langchain.com/))
89
90	<!--end-->
91
92	## Licence
93
94	This project is licensed under [MIT](https://choosealicense.com/licenses/mit/).
95
96	To cite this paper, you can use the following BibTeX entry:
97
98	```bibtex
99	@inproceedings{schnabel-neville-2024-symbolic,
100	title = "Symbolic Prompt Program Search: A Structure-Aware Approach to Efficient Compile-Time Prompt Optimization",
101	author = "Schnabel, Tobias and Neville, Jennifer",
102	booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
103	year = "2024",
104	url = "https://aclanthology.org/2024.findings-emnlp.37",
105	pages = "670--686"
106	}
107	```
108
109	## Authors
110
111	`SAMMO` was written by [Tobias Schnabel](ma

People who liked this also viewed

6 picks

Type

tunellm chat in text filePromptsJul 202653328
snippetsSupercharge your Snippets in VS Code — Manage your code snippets & AI prompts without quitting your editor.PromptsJul 2026302169
prompt-optimizerAn AI prompt optimizer for writing better prompts and getting better AI results.PromptsJul 2026—32k
chatgpt_system_promptA collection of GPT system prompts and various prompt injection/leaking knowledge.PromptsJul 2026—11k
prompt-masterA Claude skill that writes the accurate prompts for any AI tool. Zero tokens or credits wasted. Full context and memory retentionPromptsJun 2026—11k
python-prompt-toolkitLibrary for building powerful interactive command line applications in PythonPromptsMay 2026—11k

sammo

759stars45forks

View on GitHub

How to install sammo?

microsoft/sammo

$git clone https://github.com/microsoft/sammo

Installs into the current project.

Summary

A flexible, easy-to-use library for running and optimizing prompts for Large Language Models (LLMs).

Files · 1

View on GitHub

README.md

1	# SAMMO ([📘User Guide](https://microsoft.github.io/sammo/))
2
3	[![Latest PyPI version](https://img.shields.io/pypi/v/sammo.svg)](https://pypi.python.org/pypi/sammo)
4	[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
5	[![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/microsoft/sammo/main?urlpath=tree/docs/tutorials/quickstart.ipynb)
6
7	A flexible, easy-to-use library for running and optimizing prompts for Large Language Models (LLMs).
8
9	## 🎉 News
10	- Nov 13, 2024: Turn Markdown into prompt programs: First version of SAMMO express released
11	- Nov 1, 2024: Use CSS selectors to query and modify prompt programs!
12	- Oct 15, 2024: SAMMO now supports structured outputs!
13
14	## How to Get Started
15	Go to the [user guide](https://microsoft.github.io/sammo/) for examples, how-tos, and API reference.
16
17	Just want to have a quick look? Try the [live demo on Binder](https://mybinder.org/v2/gh/microsoft/sammo/main?urlpath=tree/docs/tutorials/quickstart.ipynb).
18
19	<!--start-->
20	### Install library only
21
22	```bash
23	pip install sammo
24	```
25
26	### Install and run tutorials
27
28	*Prerequisites*
29	* Python 3.9+
30
31	The following commands will install sammo and jupyter and launch jupyter notebook. It's recommended that you create and activate a virtualenv prior to installing packages.
32
33	```bash
34	pip install sammo jupyter
35
36	# clone sammo to a local directory
37	git clone https://github.com/microsoft/sammo.git
38	cd sammo
39
40	# launch jupyter notebook and open tutorials directory
41	jupyter notebook --notebook-dir docs/tutorials
42	```
43
44	## Example
45	This example shows how easy it is to optimize a prompt with SAMMO. The full example is in the [user guide](https://microsoft.github.io/sammo/).
46	```python
47	runner = OpenAIChat(model_id="gpt-3.5-turbo", api_config=API_CONFIG)
48	PROMPT_IN_MARKDOWN = """
49	# Instructions <!-- #instr -->
50	Convert the following user queries into a SQL query.
51
52	# Table
53	Users:
54	- user_id (INTEGER, PRIMARY KEY)
55	- name (TEXT)
56	- age (INTEGER)
57	- city (TEXT)
58
59	# Complete this
60	Input: {{{input}}}
61	Output:
62	"""
63
64	spp = MarkdownParser(PROMPT_IN_MARKDOWN).get_sammo_program()
65	mutation_operators = BagOfMutators(
66	Output(GenerateText(spp)),
67	Paraphrase("#instr"),
68	Rewrite("#instr", "Make this more verbose.\n\n {{{{text}}}}")
69	)
70	prompt_optimizer = BeamSearch(runner, mutation_operators, accuracy)
71	prompt_optimizer.fit(d_train)
72	prompt_optimizer.show_report()
73	```
74
75	## Use Cases
76	![Overview](https://microsoft.github.io/sammo/_images/overview.png)
77
78	SAMMO is designed to support
79	- Efficient data labeling: Supports minibatching by packing and parsing multiple datapoints into a single prompt.
80	- Prompt prototyping and engineering: Re-usable components and prompt structures to quickly build and test new prompts.
81	- Instruction optimization: Optimize instructions to do better on a given task.
82	- Prompt compression: Compress prompts while maintaining performance.
83	- Large-scale prompt execution: parallelization
84	and rate-limiting out-of-the-box so you can run many queries in parallel and at scale without overwhelming the LLM API.
85
86	It is less useful if you want to build
87	- Interactive, agent-based LLM applications (→ check out [AutoGen](https://microsoft.github.io/autogen/))
88	- Interactive, production-ready LLM applications (→ check out [LangChain](https://www.langchain.com/))
89
90	<!--end-->
91
92	## Licence
93
94	This project is licensed under [MIT](https://choosealicense.com/licenses/mit/).
95
96	To cite this paper, you can use the following BibTeX entry:
97
98	```bibtex
99	@inproceedings{schnabel-neville-2024-symbolic,
100	title = "Symbolic Prompt Program Search: A Structure-Aware Approach to Efficient Compile-Time Prompt Optimization",
101	author = "Schnabel, Tobias and Neville, Jennifer",
102	booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
103	year = "2024",
104	url = "https://aclanthology.org/2024.findings-emnlp.37",
105	pages = "670--686"
106	}
107	```
108
109	## Authors
110
111	`SAMMO` was written by [Tobias Schnabel](ma

People who liked this also viewed

6 picks

Type