Merge pull request #1 from lightbroker/init

initial port from 'lightbroker/llm' private repo

.github/workflows/llmsecops-cicd.yml

@@ -0,0 +1,39 @@
+name: REST Server
+
+on:
+  # push:
+  #   branches: [ "main" ]
+  # pull_request:
+  #   branches: [ "main" ]
+  workflow_dispatch:
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683    
+
+    - name: Set up git LFS
+      run: git lfs install
+
+    - name: Set up Python
+      uses: actions/setup-python@v3
+      with:
+        python-version: '3.12'
+
+    - name: Download Huggingface CLI
+      run: |
+        pip install huggingface-hub[cli]
+        huggingface-cli download microsoft/Phi-3-mini-4k-instruct-onnx --include cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4/* --local-dir .
+        pip install onnxruntime-genai
+        curl https://raw.githubusercontent.com/microsoft/onnxruntime-genai/main/examples/python/phi3-qa.py -o phi3-qa.py
+        python phi3-qa.py -m cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4 -e cpu -v
+
+    - name: Run REST API server
+      run: |
+        python ${{ github.workspace }}/tests/api/server.py &
+        
+    - name: Test API call
+      run: |
+        curl -i localhost:9999/hello

.gitignore

@@ -0,0 +1,187 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# UV
+#   Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#uv.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/latest/usage/project/#working-with-version-control
+.pdm.toml
+.pdm-python
+.pdm-build/
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/
+
+# Ruff stuff:
+.ruff_cache/
+
+# PyPI configuration file
+.pypirc
+
+# HuggingFace / Microsoft LLM supporting files
+# (these are downloaded for local development via bash script, or inside GH Action workflow context)
+tests/llm/cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4/added_tokens.json
+tests/llm/cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4/config.json
+tests/llm/cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4/configuration_phi3.py
+tests/llm/cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4/genai_config.json
+tests/llm/cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4/phi3-mini-4k-instruct-cpu-int4-rtn-block-32-acc-level-4.onnx
+tests/llm/cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4/phi3-mini-4k-instruct-cpu-int4-rtn-block-32-acc-level-4.onnx.data
+tests/llm/cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4/special_tokens_map.json
+tests/llm/cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4/tokenizer_config.json
+tests/llm/cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4/tokenizer.json
+tests/llm/cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4/tokenizer.model

LICENSE

@@ -0,0 +1,53 @@
+MIT License
+
+Copyright (c) 2025 Adam Wilson
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+
+
+------------------------------------------------------------------------------
+
+Additional source code is based on work by Microsoft:
+https://github.com/microsoft/onnxruntime-genai
+
+That repos is made available under the MIT License, referenced below:
+https://github.com/microsoft/onnxruntime-genai/blob/main/LICENSE
+
+
+    MIT License
+
+    Copyright (c) Microsoft Corporation.
+
+    Permission is hereby granted, free of charge, to any person obtaining a copy
+    of this software and associated documentation files (the "Software"), to deal
+    in the Software without restriction, including without limitation the rights
+    to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+    copies of the Software, and to permit persons to whom the Software is
+    furnished to do so, subject to the following conditions:
+
+    The above copyright notice and this permission notice shall be included in all
+    copies or substantial portions of the Software.
+
+    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+    AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+    OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+    SOFTWARE

README.md

@@ -1 +1,11 @@
-# llmsecops-research
+# LLMSecOps Research
+
+## Overview 
+
+This repo supports graduate research conducted by Adam Wilson for the M.Sc., Information Security Engineering program at SANS Technology Institute.
+
+## Local setup (Linux Ubuntu)
+
+```sh
+$ sudo ./llm_setup.sh
+```

llm_setup.sh

@@ -0,0 +1,21 @@
+#!/usr/bin/bash
+
+# create Python virtual environment
+virtualenv --python="/usr/bin/python3.12" .env
+source .env/bin/activate
+
+# the ONNX model/data require git Large File System support
+git lfs install
+
+# get the system-under-test LLM dependencies from HuggingFace / Microsoft
+pip3.12 install huggingface-hub[cli]
+cd ./tests/llm
+huggingface-cli download microsoft/Phi-3-mini-4k-instruct-onnx --include cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4/* --local-dir .
+pip3.12 install onnxruntime-genai
+
+if ! [[ -e ./phi3-qa.py ]]
+then
+    curl https://raw.githubusercontent.com/microsoft/onnxruntime-genai/main/examples/python/phi3-qa.py -o ./phi3-qa.py
+fi
+
+python3.12 ./phi3-qa.py -m ./cpu_and_mobile/cpu-int4-rtn-block-32-acc-level-4 -e cpu -v

tests/api/PathDispatcher.py

@@ -0,0 +1,22 @@
+import cgi
+
+
+class PathDispatcher:
+    def __init__(self):
+        self.routes = {}
+
+    def notfound_404(self, env, start_response):
+        start_response('404 Not Found', [ ('Content-Type', 'text/plain') ])
+        return [b'Not Found']
+
+    def __call__(self, env, start_response):
+        path = env.get('PATH_INFO')
+        params = cgi.FieldStorage(env.get('wsgi.output'), environ=env)
+        method = env.get('REQUEST_METHOD').lower()
+        env['params'] = { key: params.getvalue(key) for key in params }
+        handler = self.routes.get((method,path), self.notfound_404)
+        return handler(env, start_response)
+    
+    def register(self, method, path, function):
+        self.routes[method.lower(), path] = function
+        return function

tests/api/server.py

@@ -0,0 +1,24 @@
+from PathDispatcher import PathDispatcher
+from wsgiref.simple_server import make_server
+
+
+class RestApiServer:
+    def __init__(self):
+        pass
+
+    def response_function(self, environ, start_response):
+        start_response('200 OK', [('Content-Type','text/html')])
+        yield str(f'testing...\n').encode('utf-8')
+
+    def listen(self):
+        port = 9999
+        dispatcher = PathDispatcher()
+        dispatcher.register('GET', '/hello', self.response_function)
+        wsgi_srv = make_server('', port, dispatcher)
+        print(f'listening on port {port}...')
+        wsgi_srv.serve_forever()
+
+
+if __name__ == '__main__':
+    srv = RestApiServer()
+    srv.listen()

tests/llm/phi3-qa.py

@@ -0,0 +1,98 @@
+import onnxruntime_genai as og
+import argparse
+import time
+
+def main(args):
+    if args.verbose: print("Loading model...")
+    if args.timings:
+        started_timestamp = 0
+        first_token_timestamp = 0
+
+    config = og.Config(args.model_path)
+    config.clear_providers()
+    if args.execution_provider != "cpu":
+        if args.verbose: print(f"Setting model to {args.execution_provider}")
+        config.append_provider(args.execution_provider)
+    model = og.Model(config)
+
+    if args.verbose: print("Model loaded")
+    
+    tokenizer = og.Tokenizer(model)
+    tokenizer_stream = tokenizer.create_stream()
+    if args.verbose: print("Tokenizer created")
+    if args.verbose: print()
+    search_options = {name:getattr(args, name) for name in ['do_sample', 'max_length', 'min_length', 'top_p', 'top_k', 'temperature', 'repetition_penalty'] if name in args}
+    
+    # Set the max length to something sensible by default, unless it is specified by the user,
+    # since otherwise it will be set to the entire context length
+    if 'max_length' not in search_options:
+        search_options['max_length'] = 2048
+
+    chat_template = '<|user|>\n{input} <|end|>\n<|assistant|>'
+
+    params = og.GeneratorParams(model)
+    params.set_search_options(**search_options)
+    generator = og.Generator(model, params)
+
+    # Keep asking for input prompts in a loop
+    while True:
+        text = input("Input: ")
+        if not text:
+            print("Error, input cannot be empty")
+            continue
+
+        if args.timings: started_timestamp = time.time()
+
+        # If there is a chat template, use it
+        prompt = f'{chat_template.format(input=text)}'
+
+        input_tokens = tokenizer.encode(prompt)
+
+        generator.append_tokens(input_tokens)
+        if args.verbose: print("Generator created")
+
+        if args.verbose: print("Running generation loop ...")
+        if args.timings:
+            first = True
+            new_tokens = []
+
+        print()
+        print("Output: ", end='', flush=True)
+
+        try:
+            while not generator.is_done():
+                generator.generate_next_token()
+                if args.timings:
+                    if first:
+                        first_token_timestamp = time.time()
+                        first = False
+
+                new_token = generator.get_next_tokens()[0]
+                print(tokenizer_stream.decode(new_token), end='', flush=True)
+                if args.timings: new_tokens.append(new_token)
+        except KeyboardInterrupt:
+            print("  --control+c pressed, aborting generation--")
+        print()
+        print()
+
+        if args.timings:
+            prompt_time = first_token_timestamp - started_timestamp
+            run_time = time.time() - first_token_timestamp
+            print(f"Prompt length: {len(input_tokens)}, New tokens: {len(new_tokens)}, Time to first: {(prompt_time):.2f}s, Prompt tokens per second: {len(input_tokens)/prompt_time:.2f} tps, New tokens per second: {len(new_tokens)/run_time:.2f} tps")
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(argument_default=argparse.SUPPRESS, description="End-to-end AI Question/Answer example for gen-ai")
+    parser.add_argument('-m', '--model_path', type=str, required=True, help='Onnx model folder path (must contain genai_config.json and model.onnx)')
+    parser.add_argument('-e', '--execution_provider', type=str, required=True, choices=["cpu", "cuda", "dml"], help="Execution provider to run ONNX model with")
+    parser.add_argument('-i', '--min_length', type=int, help='Min number of tokens to generate including the prompt')
+    parser.add_argument('-l', '--max_length', type=int, help='Max number of tokens to generate including the prompt')
+    parser.add_argument('-ds', '--do_sample', action='store_true', default=False, help='Do random sampling. When false, greedy or beam search are used to generate the output. Defaults to false')
+    parser.add_argument('-p', '--top_p', type=float, help='Top p probability to sample with')
+    parser.add_argument('-k', '--top_k', type=int, help='Top k tokens to sample from')
+    parser.add_argument('-t', '--temperature', type=float, help='Temperature to sample with')
+    parser.add_argument('-r', '--repetition_penalty', type=float, help='Repetition penalty to sample with')
+    parser.add_argument('-v', '--verbose', action='store_true', default=False, help='Print verbose output and timing information. Defaults to false')
+    parser.add_argument('-g', '--timings', action='store_true', default=False, help='Print timing information for each generation step. Defaults to false')
+    args = parser.parse_args()
+    main(args)

tests/llm/phi3_interface.py

@@ -0,0 +1 @@
+# TODO: business logic for REST API interaction w/ LLM via prompt input