merge in setup refactor

CODE_OF_CONDUCT.md

@@ -0,0 +1,128 @@
+# Contributor Covenant Code of Conduct
+
+## Our Pledge
+
+We as members, contributors, and leaders pledge to make participation in our
+community a harassment-free experience for everyone, regardless of age, body
+size, visible or invisible disability, ethnicity, sex characteristics, gender
+identity and expression, level of experience, education, socio-economic status,
+nationality, personal appearance, race, religion, or sexual identity
+and orientation.
+
+We pledge to act and interact in ways that contribute to an open, welcoming,
+diverse, inclusive, and healthy community.
+
+## Our Standards
+
+Examples of behavior that contributes to a positive environment for our
+community include:
+
+* Demonstrating empathy and kindness toward other people
+* Being respectful of differing opinions, viewpoints, and experiences
+* Giving and gracefully accepting constructive feedback
+* Accepting responsibility and apologizing to those affected by our mistakes,
+  and learning from the experience
+* Focusing on what is best not just for us as individuals, but for the
+  overall community
+
+Examples of unacceptable behavior include:
+
+* The use of sexualized language or imagery, and sexual attention or
+  advances of any kind
+* Trolling, insulting or derogatory comments, and personal or political attacks
+* Public or private harassment
+* Publishing others' private information, such as a physical or email
+  address, without their explicit permission
+* Other conduct which could reasonably be considered inappropriate in a
+  professional setting
+
+## Enforcement Responsibilities
+
+Community leaders are responsible for clarifying and enforcing our standards of
+acceptable behavior and will take appropriate and fair corrective action in
+response to any behavior that they deem inappropriate, threatening, offensive,
+or harmful.
+
+Community leaders have the right and responsibility to remove, edit, or reject
+comments, commits, code, wiki edits, issues, and other contributions that are
+not aligned to this Code of Conduct, and will communicate reasons for moderation
+decisions when appropriate.
+
+## Scope
+
+This Code of Conduct applies within all community spaces, and also applies when
+an individual is officially representing the community in public spaces.
+Examples of representing our community include using an official e-mail address,
+posting via an official social media account, or acting as an appointed
+representative at an online or offline event.
+
+## Enforcement
+
+Instances of abusive, harassing, or otherwise unacceptable behavior may be
+reported to the community leaders responsible for enforcement at
+[email protected].
+All complaints will be reviewed and investigated promptly and fairly.
+
+All community leaders are obligated to respect the privacy and security of the
+reporter of any incident.
+
+## Enforcement Guidelines
+
+Community leaders will follow these Community Impact Guidelines in determining
+the consequences for any action they deem in violation of this Code of Conduct:
+
+### 1. Correction
+
+**Community Impact**: Use of inappropriate language or other behavior deemed
+unprofessional or unwelcome in the community.
+
+**Consequence**: A private, written warning from community leaders, providing
+clarity around the nature of the violation and an explanation of why the
+behavior was inappropriate. A public apology may be requested.
+
+### 2. Warning
+
+**Community Impact**: A violation through a single incident or series
+of actions.
+
+**Consequence**: A warning with consequences for continued behavior. No
+interaction with the people involved, including unsolicited interaction with
+those enforcing the Code of Conduct, for a specified period of time. This
+includes avoiding interactions in community spaces as well as external channels
+like social media. Violating these terms may lead to a temporary or
+permanent ban.
+
+### 3. Temporary Ban
+
+**Community Impact**: A serious violation of community standards, including
+sustained inappropriate behavior.
+
+**Consequence**: A temporary ban from any sort of interaction or public
+communication with the community for a specified period of time. No public or
+private interaction with the people involved, including unsolicited interaction
+with those enforcing the Code of Conduct, is allowed during this period.
+Violating these terms may lead to a permanent ban.
+
+### 4. Permanent Ban
+
+**Community Impact**: Demonstrating a pattern of violation of community
+standards, including sustained inappropriate behavior,  harassment of an
+individual, or aggression toward or disparagement of classes of individuals.
+
+**Consequence**: A permanent ban from any sort of public interaction within
+the community.
+
+## Attribution
+
+This Code of Conduct is adapted from the [Contributor Covenant][homepage],
+version 2.0, available at
+https://www.contributor-covenant.org/version/2/0/code_of_conduct.html.
+
+Community Impact Guidelines were inspired by [Mozilla's code of conduct
+enforcement ladder](https://github.com/mozilla/diversity).
+
+[homepage]: https://www.contributor-covenant.org
+
+For answers to common questions about this code of conduct, see the FAQ at
+https://www.contributor-covenant.org/faq. Translations are available at
+https://www.contributor-covenant.org/translations.

CONTRIBUTION.md

@@ -0,0 +1,30 @@
+# Contributing to Latent Scope
+
+Thank you for considering contributing to Latent Scope! We welcome contributions from the community and are grateful for your support.
+
+## Case Studies
+
+We are looking for case studies of how Latent Scope has been used to solve real-world problems. If you have a case study to share, please open an issue or send an email to [[email protected]](mailto:[email protected]).
+
+## Development Guide
+
+For detailed information on setting up the development environment and the structure of the project, please refer to the [Development Guide](DEVELOPMENT.md).
+
+## Help Wanted
+
+We have a list of issues that are good entry points for new contributors. Check out the [Help Wanted Issues](https://github.com/enjalot/latent-scope/issues?q=is%3Aissue+is%3Aopen+label%3A%22help+wanted%22) to get started.
+
+## Roadmap
+
+We have organized the roadmap into a list of features and improvements that we would like to make to Latent Scope in [Milestones](https://github.com/enjalot/latent-scope/milestones).
+The goal is to launch [1.0 by the end of 2024](https://github.com/enjalot/latent-scope/milestone/1), which will be a polished and refined version of the current Latent Scope process. This work is being done with the support of the [Mozilla Builders Accelerator](https://future.mozilla.org/builders/2024_cohort/).
+
+We have a set of features planned to advance the UX around clustering and add more methods slated for [2.0](https://github.com/enjalot/latent-scope/milestone/3).
+
+Additionally there are some more speculative research ideas captured in [Research Ideas](https://github.com/enjalot/latent-scope/milestone/2) that focus more on the meta process of comparing embeddings, umaps and clusters.
+
+## Code of Conduct
+
+Please note that this project is released with a [Contributor Code of Conduct](CODE_OF_CONDUCT.md). By participating in this project you agree to abide by its terms.
+
+Thank you for your contributions!

documentation/DEVELOPMENT.md → DEVELOPMENT.md

@@ -3,7 +3,7 @@
 If you are interested in customizing or contributing to latent-scope this document explains the ways to develop it.
 
 ## Python module
-The `latentscope` directory contains the python source code for the [latent scope pip module](). There are three primary parts to the module:
+The `latentscope` directory contains the python source code for the [latent scope pip module](https://pypi.org/project/latentscope/). There are three primary parts to the module:
 
 1. server - contains the flask app and API routes
 2. scripts - 
@@ -15,11 +15,12 @@ If you modify the python code and want to try out your changes, you can run loca
 
 ```
 python -m venv testenv
-source venv/bin/activate
+source testenv/bin/activate
 pip install -e .
 ls-serve ~/latent-scope-data
 ```
 
+We recommend python 3.12 as that's what the project is developed with.
 
 ## Web client
 The `web` directory contains the JavaScript React source code for the web interface. Node.js is required to be installed on your system to run the development server or build a new version of the module.
@@ -36,17 +37,11 @@ This will call the local API at http://localhost:5001 as set in `web/.env.develo
 ## Building for distribution
 You can build a new version of the module, this will package the latest version of the web interface as well.
 
-```
-python setup.py sdist bdist_wheel
-```
-This builds the package, including the React app, and bundles it all up. 
-
-```
-deactivate
-python setup.py sdist bdist_wheel
-source testenv/bin/activate
-pip install dist/latentscope-0.1.0-py3-none-any.whl
-
+You can use the included `build.sh` script to build and install the wheel in a test environment:
+```bash
+./build.sh 0.4.4
+source testenv-whl/bin/activate
+ls-serve
 ```
 
 
@@ -78,7 +73,7 @@ There is a `get_embedding_model(id)` function which will load the appropriate cl
 
 ### Chat models
 Chat models for summarization of clusters are prepared in [latentscope/models/chat_models.json](latentscope/models/chat_models.json). 
-There is a `get_chat_model(id)` function which will load the appropriate class based on the model provider. Each provider can support a chat model if the interface is implemented. Adding more chat providers should be relatively straightforward and is still a TODO item.
+There is a `get_chat_model(id)` function which will load the appropriate class based on the model provider. Each provider can support a chat model if the interface is implemented. Adding more chat models is captured in [this issue](https://github.com/enjalot/latent-scope/issues/2).
 
 ## Scripts
 The scripts are outlined in the [README.md](README.md). Each one provides a python interface as well as a command line interface for running its part of the process. They all use ids to read relevant data from disk and then output any relevant information to disk.

README.md

@@ -2,6 +2,8 @@
 [![](https://dcbadge.vercel.app/api/server/x7NvpnM4pY?style=flat)](https://discord.gg/x7NvpnM4pY)
 [![PyPI version](https://img.shields.io/pypi/v/latentscope.svg)](https://pypi.org/project/latentscope/)
 
+[Docs](https://enjalot.github.io/latent-scope/)
+
 Quickly embed, project, cluster and explore a dataset. This project is a new kind of workflow + tool for visualizing and exploring datasets through the lens of latent spaces. 
 
 <img src="https://github.com/enjalot/latent-scope/blob/main/documentation/end.png?raw=true" width="45%"  alt="Setup your scope"><img src="https://github.com/enjalot/latent-scope/blob/main/documentation/curate-combined-cluster.png?raw=true" width="45%"  alt="Explore and Curate your data">
@@ -27,18 +29,21 @@ What can you do with Latent Scope? The following examples demonstrate the kinds
 
 
 ### Quick Start
-To get started, install the [latent-scope module](https://pypi.org/project/latentscope/) and run the server via the Command Line:
+Latent Scope works on Mac, Linux and Windows. Python 3.12 is the recommended python version.
+
+To get started, install the [latent-scope python module](https://pypi.org/project/latentscope/) and run the server via the Command Line:
 
 ```bash
 python -m venv venv
 source venv/bin/activate
 pip install latentscope
-ls-init ~/local-scope-data --openai_key=XXX --mistral_key=YYY # optional api keys to enable API models 
+ls-init ~/latent-scope-data --openai_key=XXX --mistral_key=YYY # optional api keys to enable API models 
 ls-serve 
 ```
 
 Then open your browser to http://localhost:5001 and start processing your first dataset!  
 
+
 See the [Your First Scope](https://enjalot.github.io/latent-scope/your-first-scope) guide for a detailed walk-through of the process.
 
 ### Python interface
@@ -92,7 +97,7 @@ ls-serve
 ### Repository overview
 This repository is currently meant to run locally, with a React frontend that communicates with a python server backend. We support several popular open source embedding models that can run locally as well as proprietary API embedding services. Adding new models and services should be quick and easy.
 
-To learn more about customizing, extending and contributing see [DEVELOPMENT.md](documentation/DEVELOPMENT.md)
+To learn more about contributing and the project roadmap see [CONTRIBUTION.md](CONTRIBUTION.md), for technical details see [DEVELOPMENT.md](DEVELOPMENT.md).
 
 
 ### Design principles

latentscope/__version__.py

@@ -1 +1 @@
-__version__ = '0.4.1'
+__version__ = '0.4.4'

latentscope/models/chat_models.json

@@ -24,14 +24,6 @@
             "max_tokens": 128000
         }
     },
-    {
-        "provider": "openai",
-        "name": "gpt-4-turbo-preview",
-        "id": "openai-gpt-4-turbo-preview",
-        "params": {
-            "max_tokens": 128000
-        }
-    },
     {
         "provider": "openai",
         "name": "gpt-4-turbo",
@@ -47,38 +39,14 @@
         "params": {
             "max_tokens": 8192
         }
-    },
-    {
-      "provider": "mistralai",
-      "name": "mistral-tiny",
-      "id": "mistralai-mistral-tiny",
-      "params": {
-          "max_tokens": 8192
-      }
-    },
-    {
-      "provider": "mistralai",
-      "name": "mistral-small",
-      "id": "mistralai-mistral-small",
-      "params": {
-          "max_tokens": 8192
-      }
-    },
-    {
-      "provider": "mistralai",
-      "name": "mistral-medium",
-      "id": "mistralai-mistral-medium",
-      "params": {
-          "max_tokens": 8192
-      }
-    },
+    },    
     {
         "provider": "transformers",
-        "name": "meta-llama/Meta-Llama-3-8B-Instruct",
-        "sanitized_name": "meta-llama___Meta-Llama-3-8B-Instruct",
-        "id": "transformers-meta-llama___Meta-Llama-3-8B-Instruct",
+        "name": "meta-llama/Meta-Llama-3.1-8B-Instruct",
+        "sanitized_name": "meta-llama___Meta-Llama-3.1-8B-Instruct",
+        "id": "transformers-meta-llama___Meta-Llama-3.1-8B-Instruct",
         "params": {
-            "max_tokens": 8192
+            "max_tokens": 128000
         }
     }
 ]

latentscope/models/providers/nltk.py

@@ -22,7 +22,13 @@ def decode(self, tokens):
 
     def chat(self, messages):
         # TODO: this is kind of hacky, since we aren't really using a chat model
-        tokens = self.encoder.encode(messages[1]["content"])
+        # We are slicing off the first 86 characters because it contains the prompt for LLMs
+        text = messages[1]["content"][86:]
+        text = text.replace("<ListItem>", "")
+        text = text.replace("< ListItem >", "")
+        text = text.replace("</ListItem>", "")
+        text = text.replace("</ ListItem >", "")
+        tokens = self.encoder.encode(text)
         # Remove stopwords
         tokens = [word for word in tokens if word.isalpha() and word.lower() not in self.stopwords.words('english')]
         # Get the top 3 words

latentscope/models/providers/transformers.py

@@ -24,7 +24,7 @@ def load_model(self):
         # from transformers import AutoTokenizer, AutoModel
         from sentence_transformers import SentenceTransformer
 
-        self.model = SentenceTransformer(self.name, trust_remote_code=True, device=self.device)
+        self.model = SentenceTransformer(self.name, trust_remote_code=True, device=self.device)#, backend="onnx")
         self.tokenizer = self.model.tokenizer
 
         # if "rps" in self.params and self.params["rps"]:
@@ -74,12 +74,11 @@ def __init__(self, name, params):
         from transformers import pipeline
         self.torch = torch
         self.pipeline = pipeline
+        self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
     def load_model(self):
-        # self.pipe = pipeline("text-generation", model="TinyLlama/TinyLlama-1.1B-Chat-v1.0", torch_dtype=torch.bfloat16, device_map="auto")
-        # self.pipe = pipeline("text-generation", model="TinyLlama/TinyLlama-1.1B-Chat-v1.0", torch_dtype=torch.bfloat16, device_map="cpu")
         # TODO: support bfloat16 for non mac environmentss
-        self.pipe = self.pipeline("text-generation", model=self.name, torch_dtype=self.torch.float16, device_map="auto")
+        self.pipe = self.pipeline("text-generation", model=self.name, torch_dtype=self.torch.float16, device=self.device, trust_remote_code=True)
         self.encoder = self.pipe.tokenizer
 
     def chat(self, messages, max_new_tokens=24):

latentscope/scripts/export_plot.py

@@ -90,15 +90,19 @@ def get_next_plot_number():
     label_sizes.reset_index()
     print(label_sizes)
 
+
+    threshold = -1
+    if plot_config:
+        plot_config = json.loads(plot_config)
+        threshold = plot_config.get("threshold", -1)
+        del plot_config["threshold"]
+        plot_config["noise_label"] = "No Topic"
+
     simplified_labels = scope_parquet["label"].copy()
-    # TODO: optional cut clusters with less than threshold
-    threshold = 100
-    # simplified_labels[np.in1d(simplified_labels, label_sizes[label_sizes < threshold].index)] = "No Topic"
+    if threshold >= 0:
+        simplified_labels[np.in1d(simplified_labels, label_sizes[label_sizes < threshold].index)] = "No Topic"
 
     highlight_labels = np.unique(simplified_labels)
-
-    if plot_config:
-        plot_config = json.loads(plot_config)
     if plot_config is None:
         plot_config = {
             "label_over_points": True,