Super-squash branch 'main' using huggingface_hub

.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,160 @@
+# 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
+
+# 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/#use-with-ide
+.pdm.toml
+
+# 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/

README.md

@@ -0,0 +1,46 @@
+---
+title: BeeCoder Demo
+emoji: 🐝
+colorFrom: gray
+colorTo: yellow
+sdk: gradio
+sdk_version: 3.28.3
+app_file: app.py
+pinned: true
+license: apache-2.0
+---
+
+# 🐝BeeCoder Demo🐝
+
+## Code-Completion Playground 💻 with 🐝[BeeCoder](https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA-python) Models
+
+This is a demo playground for generating Python code with the power of 🐝[BeeCoder](https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA-python), a **fine-tuned** version of the tiny [101M base model](https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA) on a dataset of pypi packages.
+
+ℹ️ This is not an instruction model but just a code completion tool.
+
+---
+
+**Intended Use**: This app and its [supporting model](https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA-python) are provided for demonstration purposes only; not to serve as a replacement for human expertise. For more details on the model, please refer to the [model card](https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA-python).
+
+In our country, we say _"To let 100M parameters model generate python script and not validate is like to let monkey fly a plane"_. So please be careful with the generated code.
+
+---
+
+## Base Model Information
+
+The base model, smol_llama-101M-GQA, has been pre-trained on a relatively small number of high quality tokens (less than ~20B). It has impressive performance despite its compact size of 101M parameters. Training data for this base model included:
+
+- [JeanKaddour/minipile](https://huggingface.co/datasets/JeanKaddour/minipile)
+- [pszemraj/simple_wikipedia_LM](https://huggingface.co/datasets/pszemraj/simple_wikipedia_LM)
+- [BEE-spoke-data/wikipedia-20230901.en-deduped](https://huggingface.co/datasets/BEE-spoke-data/wikipedia-20230901.en-deduped)
+- [mattymchen/refinedweb-3m](https://huggingface.co/datasets/mattymchen/refinedweb-3m)
+
+You can find more information about the base model [here](https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA).
+
+---
+
+### Credits
+
+This app is modified from a demo playground originally built for [StarCoder](https://huggingface.co/bigcode/starcoder) by [BigCode](https://huggingface.co/bigcode).  You can find the original demo [here](https://huggingface.co/spaces/bigcode/bigcode-playground).
+
+---

app.py

@@ -0,0 +1,234 @@
+import gradio as gr
+import torch
+from gradio.themes.utils import sizes
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+import utils
+from constants import END_OF_TEXT, MIN_TEMPERATURE
+
+# Load the tokenizer and model
+tokenizer = AutoTokenizer.from_pretrained(
+    "BEE-spoke-data/smol_llama-101M-GQA-python",
+    use_fast=False,
+)
+tokenizer.pad_token_id = tokenizer.eos_token_id
+tokenizer.pad_token = END_OF_TEXT
+model = AutoModelForCausalLM.from_pretrained(
+    "BEE-spoke-data/smol_llama-101M-GQA-python",
+    device_map="auto",
+)
+model = torch.compile(model, mode="reduce-overhead")
+
+# UI things
+
+_styles = utils.get_file_as_string("styles.css")
+
+# Loads ./README.md file & splits it into sections
+readme_file_content = utils.get_file_as_string("README.md", path="./")
+(
+    manifest,
+    description,
+    disclaimer,
+    base_model_info,
+    formats,
+) = utils.get_sections(readme_file_content, "---", up_to=5)
+
+theme = gr.themes.Soft(
+    primary_hue="yellow",
+    secondary_hue="orange",
+    neutral_hue="slate",
+    radius_size=sizes.radius_sm,
+    font=[
+        gr.themes.GoogleFont("IBM Plex Sans", [400, 600]),
+        "ui-sans-serif",
+        "system-ui",
+        "sans-serif",
+    ],
+    text_size=sizes.text_lg,
+)
+
+
+def run_inference(
+    prompt, temperature, max_new_tokens, top_p, repetition_penalty
+) -> str:
+    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+    outputs = model.generate(
+        **inputs,
+        do_sample=True,
+        epsilon_cutoff=1e-3,
+        max_new_tokens=max_new_tokens,
+        min_new_tokens=2,
+        no_repeat_ngram_size=6,
+        renormalize_logits=True,
+        repetition_penalty=repetition_penalty,
+        temperature=max(temperature, MIN_TEMPERATURE),
+        top_p=top_p,
+    )
+    text = tokenizer.batch_decode(
+        outputs,
+        skip_special_tokens=True,
+    )[0]
+    return text
+
+
+examples = [
+    [
+        'def greet(name: str) -> None:\n    """\n    Greets the user\n    """\n    print(f"Hello,',
+        0.2,
+        64,
+        0.9,
+        1.2,
+    ],
+    [
+        'for i in range(5):\n    """\n    Loop through 0 to 4\n    """\n    print(i,',
+        0.2,
+        64,
+        0.9,
+        1.2,
+    ],
+    ['x = 10\n"""Check if x is greater than 5"""\nif x > 5:', 0.2, 64, 0.9, 1.2],
+    ["def square(x: int) -> int:\n    return", 0.2, 64, 0.9, 1.2],
+    ['import math\n"""Math operations"""\nmath.', 0.2, 64, 0.9, 1.2],
+    [
+        'def is_even(n) -> bool:\n    """\n    Check if a number is even\n    """\n    if n % 2 == 0:',
+        0.2,
+        64,
+        0.9,
+        1.2,
+    ],
+    [
+        'while True:\n    """Infinite loop example"""\n    print("Infinite loop,',
+        0.2,
+        64,
+        0.9,
+        1.2,
+    ],
+    [
+        "def sum_list(lst: list[int]) -> int:\n    total = 0\n    for item in lst:",
+        0.2,
+        64,
+        0.9,
+        1.2,
+    ],
+    [
+        'try:\n    """\n    Exception handling\n    """\n    x = int(input("Enter a number: "))\nexcept ValueError:',
+        0.2,
+        64,
+        0.9,
+        1.2,
+    ],
+    [
+        'def divide(a: float, b: float) -> float:\n    """\n    Divide a by b\n    """\n    if b != 0:',
+        0.2,
+        64,
+        0.9,
+        1.2,
+    ],
+]
+
+
+# Define the Gradio Blocks interface
+with gr.Blocks(theme=theme, analytics_enabled=False, css=_styles) as demo:
+    with gr.Column():
+        gr.Markdown(description)
+        with gr.Row():
+            with gr.Column():
+                instruction = gr.Textbox(
+                    value=examples[0][0],
+                    placeholder="Enter your code here",
+                    label="Code",
+                    elem_id="q-input",
+                )
+                submit = gr.Button("Generate", variant="primary")
+                output = gr.Code(elem_id="q-output", language="python", lines=10)
+                with gr.Row():
+                    with gr.Column():
+                        with gr.Accordion("Advanced settings", open=False):
+                            with gr.Row():
+                                column_1, column_2 = gr.Column(), gr.Column()
+                                with column_1:
+                                    temperature = gr.Slider(
+                                        label="Temperature",
+                                        value=0.2,
+                                        minimum=0.0,
+                                        maximum=1.0,
+                                        step=0.05,
+                                        interactive=True,
+                                        info="Higher values produce more diverse outputs",
+                                    )
+                                    max_new_tokens = gr.Slider(
+                                        label="Max new tokens",
+                                        value=64,
+                                        minimum=32,
+                                        maximum=512,
+                                        step=32,
+                                        interactive=True,
+                                        info="Number of tokens to generate",
+                                    )
+                                with column_2:
+                                    top_p = gr.Slider(
+                                        label="Top-p (nucleus sampling)",
+                                        value=0.90,
+                                        minimum=0.0,
+                                        maximum=1,
+                                        step=0.05,
+                                        interactive=True,
+                                        info="Higher values sample more low-probability tokens",
+                                    )
+                                    repetition_penalty = gr.Slider(
+                                        label="Repetition penalty",
+                                        value=1.2,
+                                        minimum=1.0,
+                                        maximum=2.0,
+                                        step=0.05,
+                                        interactive=True,
+                                        info="Penalize repeated tokens",
+                                    )
+                    with gr.Column():
+                        version = gr.Dropdown(
+                            [
+                                "smol_llama-101M-GQA-python",
+                            ],
+                            value="smol_llama-101M-GQA-python",
+                            label="Version",
+                            info="",
+                        )
+                gr.Markdown(disclaimer)
+                gr.Examples(
+                    examples=examples,
+                    inputs=[
+                        instruction,
+                        temperature,
+                        max_new_tokens,
+                        top_p,
+                        repetition_penalty,
+                        version,
+                    ],
+                    cache_examples=False,
+                    fn=run_inference,
+                    outputs=[output],
+                )
+                gr.Markdown(base_model_info)
+                gr.Markdown(formats)
+
+    submit.click(
+        run_inference,
+        inputs=[
+            instruction,
+            temperature,
+            max_new_tokens,
+            top_p,
+            repetition_penalty,
+        ],
+        outputs=[output],
+        # preprocess=False,
+        # batch=False,
+        show_progress=True,
+    )
+
+# .queue(max_size=10, api_open=False)
+demo.launch(
+    debug=True,
+    show_api=False,
+    share=utils.is_google_colab(),
+)

constants.py

@@ -0,0 +1,4 @@
+END_OF_TEXT = "<|endoftext|>"
+
+# Near zero temperature to avoid division by zero
+MIN_TEMPERATURE = 1e-4

requirements.txt

@@ -0,0 +1,7 @@
+# Gradio
+gradio==3.28.3
+
+# HuggingFace
+accelerate
+sentencepiece
+transformers==4.33.3

settings.py

@@ -0,0 +1,16 @@
+# URLs for the StarCoder Models/APIs
+DEFAULT_HUGGINGFACE_MODELS_API_BASE_URL = "https://api-inference.huggingface.co/models/"
+DEFAULT_STARCODER_API_PATH = "bigcode/starcoder/"
+DEFAULT_STARCODER_BASE_API_PATH = "bigcode/starcoderbase/"
+FIM_INDICATOR = "<FILL_HERE>"
+DEFAULT_PORT = 7860
+
+STATIC_PATH = "static"
+
+DEFAULT_SETTINGS = dict(
+    temperature=0.9,
+    max_new_tokens=256,
+    top_p=0.95,
+    repetition_penalty=1.0,
+    version="StarCoder",
+)

static/styles.css

@@ -0,0 +1,78 @@
+@import url('https://fonts.googleapis.com/css2?family=IBM+Plex+Mono:wght@400;600;700&display=swap');
+
+h1, h2 {
+    font-family: 'IBM Plex Mono', sans-serif;
+}
+
+.generating {
+    visibility: hidden
+}
+
+.gradio-container {
+    color: black
+}
+
+/* monospace_css */
+#q-input textarea {
+    font-family: monospace, 'Consolas', Courier, monospace;
+}
+
+/* Share Button */
+
+/* it was hidden directly inside the svg xml content */
+#share-btn-loading-icon {
+    display: none;
+}
+
+a {
+    text-decoration-line: underline;
+    font-weight: 600;
+}
+
+.animate-spin {
+    animation: spin 1s linear infinite;
+}
+
+@keyframes spin {
+    from {
+        transform: rotate(0deg);
+    }
+    to {
+        transform: rotate(360deg);
+    }
+}
+
+#share-btn-container {
+    display: flex;
+    padding-left: 0.5rem !important;
+    padding-right: 0.5rem !important;
+    background-color: #000000;
+    justify-content: center;
+    align-items: center;
+    border-radius: 9999px !important;
+    width: 15rem;
+}
+
+#share-btn {
+    all: initial;
+    color: #ffffff;
+    font-weight: 600;
+    cursor: pointer;
+    font-family: 'IBM Plex Sans', sans-serif;
+    margin-left: 0.5rem !important;
+    padding-top: 0.25rem !important;
+    padding-bottom: 0.25rem !important;
+}
+
+#share-btn * {
+    all: unset;
+}
+
+#share-btn-container div:nth-child(-n+2) {
+    width: auto !important;
+    min-height: 0px !important;
+}
+
+#share-btn-container .wrap {
+    display: none !important;
+}

utils.py

@@ -0,0 +1,55 @@
+import os
+from typing import List
+
+from settings import STATIC_PATH
+
+
+def is_google_colab():
+    """Check if the environment is Google Colab."""
+    try:
+        from google.colab import drive
+
+        return True
+    except ImportError:
+        return False
+
+
+def get_file_as_string(file_name, path=STATIC_PATH) -> str:
+    """Loads the content of a file given its name
+    and returns all of its lines as a single string
+    if a file path is given, it will be used
+    instead of the default static path (from settings)
+
+    Args:
+        file_name (_type_): The name of the file to load.
+        path (str, optional): The path to the file. Defaults to the current directory.
+
+    Returns:
+        str: The content of the file as a single string
+    """
+    with open(os.path.join(path, file_name), mode="r", encoding="UTF-8") as f:
+        return f.read()
+
+
+def get_sections(string: str, delimiter: str, up_to: int = None) -> List[str]:
+    """Splits a string into sections given a delimiter
+
+    Args:
+        string (str): The string to split
+        delimiter (str): The delimiter to use
+        up_to (int, optional): The maximum number of sections to return.
+                Defaults to None (which means all sections)
+
+    Returns:
+        List[str]: The list of sections (up to the given limit, if any provided)
+    """
+    return [
+        section.strip()
+        for section in string.split(delimiter)
+        if (section and not section.isspace())
+    ][:up_to]
+
+
+def get_workers(safety: int = 4) -> int:
+    """Return the number of cores available on the current system, minus a safety margin."""
+    return max(1, os.cpu_count() - safety)