Add files

.pre-commit-config.yaml

@@ -0,0 +1,55 @@
+repos:
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v4.5.0
+    hooks:
+      - id: check-executables-have-shebangs
+      - id: check-json
+      - id: check-merge-conflict
+      - id: check-shebang-scripts-are-executable
+      - id: check-toml
+      - id: check-yaml
+      - id: end-of-file-fixer
+      - id: mixed-line-ending
+        args: ["--fix=lf"]
+      - id: requirements-txt-fixer
+      - id: trailing-whitespace
+  - repo: https://github.com/myint/docformatter
+    rev: v1.7.5
+    hooks:
+      - id: docformatter
+        args: ["--in-place"]
+  - repo: https://github.com/pycqa/isort
+    rev: 5.13.2
+    hooks:
+      - id: isort
+        args: ["--profile", "black"]
+  - repo: https://github.com/pre-commit/mirrors-mypy
+    rev: v1.7.1
+    hooks:
+      - id: mypy
+        args: ["--ignore-missing-imports"]
+        additional_dependencies:
+          ["types-python-slugify", "types-requests", "types-PyYAML"]
+  - repo: https://github.com/psf/black
+    rev: 23.12.0
+    hooks:
+      - id: black
+        language_version: python3.10
+        args: ["--line-length", "119"]
+  - repo: https://github.com/kynan/nbstripout
+    rev: 0.6.1
+    hooks:
+      - id: nbstripout
+        args:
+          [
+            "--extra-keys",
+            "metadata.interpreter metadata.kernelspec cell.metadata.pycharm",
+          ]
+  - repo: https://github.com/nbQA-dev/nbQA
+    rev: 1.7.1
+    hooks:
+      - id: nbqa-black
+      - id: nbqa-pyupgrade
+        args: ["--py37-plus"]
+      - id: nbqa-isort
+        args: ["--float-to-top"]

.vscode/settings.json

@@ -0,0 +1,26 @@
+{
+    "editor.formatOnSave": true,
+    "files.insertFinalNewline": false,
+    "[python]": {
+        "editor.defaultFormatter": "ms-python.black-formatter",
+        "editor.formatOnType": true,
+        "editor.codeActionsOnSave": {
+            "source.organizeImports": "explicit"
+        }
+    },
+    "[jupyter]": {
+        "files.insertFinalNewline": false
+    },
+    "black-formatter.args": [
+        "--line-length=119"
+    ],
+    "isort.args": ["--profile", "black"],
+    "flake8.args": [
+        "--max-line-length=119"
+    ],
+    "ruff.lint.args": [
+        "--line-length=119"
+    ],
+    "notebook.output.scrolling": true,
+    "notebook.formatOnCellExecution": true
+}

README.md

@@ -1,5 +1,5 @@
 ---
-title: Swallow 13B
+title: Swallow-13B instruct
 emoji: 🐢
 colorFrom: purple
 colorTo: purple

app.py

@@ -0,0 +1,140 @@
+#!/usr/bin/env python
+
+import os
+from threading import Thread
+from typing import Iterator
+
+import gradio as gr
+import spaces
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
+
+DESCRIPTION = """# Swallow-13B instruct"""
+
+if not torch.cuda.is_available():
+    DESCRIPTION += "\n<p>Running on CPU 🥶 This demo does not work on CPU.</p>"
+
+if torch.cuda.is_available():
+    model_name = "tokyotech-llm/Swallow-13b-instruct-hf"
+    tokenizer = AutoTokenizer.from_pretrained(model_name)
+    model = AutoModelForCausalLM.from_pretrained(
+        model_name, load_in_8bit=True, low_cpu_mem_usage=True, device_map="auto"
+    )
+
+MAX_INPUT_TOKENS = 2048
+
+PROMPT_DICT = {
+    "prompt_input": (
+        "以下に、あるタスクを説明する指示があり、それに付随する入力が更なる文脈を提供しています。"
+        "リクエストを適切に完了するための回答を記述してください。\n\n"
+        "### 指示:\n{instruction}\n\n### 入力:\n{input}\n\n### 応答:"
+    ),
+    "prompt_no_input": (
+        "以下に、あるタスクを説明する指示があります。" "リクエストを適切に完了するための回答を記述してください。\n\n" "### 指示:\n{instruction}\n\n### 応答:"
+    ),
+}
+
+
+def create_prompt(instruction: str, input_text: str | None = None) -> str:
+    """Generates a prompt based on the given instruction and an optional input.
+    If input is provided, it uses the 'prompt_input' template from PROMPT_DICT.
+    If no input is provided, it uses the 'prompt_no_input' template.
+
+    Args:
+        instruction (str): The instruction describing the task.
+        input_text (str, optional): Additional input providing context for the task. Default is None.
+
+    Returns:
+        str: The generated prompt.
+    """
+    if input_text:
+        # Use the 'prompt_input' template when additional input is provided
+        return PROMPT_DICT["prompt_input"].format(instruction=instruction, input=input_text)
+    else:
+        # Use the 'prompt_no_input' template when no additional input is provided
+        return PROMPT_DICT["prompt_no_input"].format(instruction=instruction)
+
+
+@spaces.GPU
+@torch.inference_mode()
+def run(
+    instruction: str,
+    input_text: str | None = None,
+    max_new_tokens: int = 256,
+    temperature: float = 0.99,
+    top_p: float = 0.95,
+) -> Iterator[str]:
+    if input_text == "":
+        input_text = None
+
+    prompt = create_prompt(instruction, input_text)
+    input_ids = tokenizer.encode(prompt, add_special_tokens=False, return_tensors="pt")
+    if input_ids.shape[-1] > MAX_INPUT_TOKENS:
+        raise gr.Error(f"Input exceeds maximum number of tokens ({MAX_INPUT_TOKENS})")
+
+    streamer = TextIteratorStreamer(tokenizer, timeout=20.0, skip_prompt=True, skip_special_tokens=True)
+    generate_kwargs = dict(
+        {"input_ids": input_ids.to(model.device)},
+        streamer=streamer,
+        max_new_tokens=max_new_tokens,
+        temperature=temperature,
+        top_p=top_p,
+        do_sample=True,
+    )
+    t = Thread(target=model.generate, kwargs=generate_kwargs)
+    t.start()
+
+    outputs = []
+    for text in streamer:
+        outputs.append(text)
+        yield "".join(outputs)
+
+
+def process_example(instruction: str, input_text: str) -> Iterator[str]:
+    yield from run(instruction, input_text)
+
+
+with gr.Blocks(css="style.css") as demo:
+    gr.Markdown(DESCRIPTION)
+    gr.DuplicateButton(
+        value="Duplicate Space for private use",
+        elem_id="duplicate-button",
+        visible=os.getenv("SHOW_DUPLICATE_BUTTON") == "1",
+    )
+
+    with gr.Row():
+        with gr.Column():
+            instruction = gr.Textbox(label="Instruction", lines=5)
+            input_text = gr.Textbox(label="Input (optional)", lines=5)
+            run_button = gr.Button()
+
+            with gr.Accordion(label="Advanced Options", open=False):
+                max_new_tokens = gr.Slider(label="Max New Tokens", minimum=1, maximum=1024, step=1, value=256)
+                temperature = gr.Slider(label="Temperature", minimum=0.0, maximum=2.0, step=0.01, value=0.99)
+                top_p = gr.Slider(label="Top P", minimum=0.0, maximum=1.0, step=0.01, value=0.95)
+
+        with gr.Column():
+            output = gr.Textbox(label="Output", lines=10)
+
+        run_button.click(
+            fn=run,
+            inputs=[instruction, input_text, max_new_tokens, temperature, top_p],
+            outputs=output,
+            api_name="run",
+        )
+
+    gr.Examples(
+        examples=[
+            ["以下のトピックに関する詳細な情報を提供してください。", "東京工業大学の主なキャンパスについて教えてください。"],
+            ["以下のトピックに関する詳細な情報を提供してください。", "夢オチとは何かについて教えてください。"],
+            ["暴れん坊将軍って誰のことですか？", ""],
+        ],
+        inputs=[instruction, input_text, max_new_tokens, temperature, top_p],
+        outputs=output,
+        fn=process_example,
+        cache_examples=os.getenv("CACHE_EXAMPLES") == "1",
+        api_name=False,
+    )
+
+if __name__ == "__main__":
+    demo.queue(max_size=20).launch()

requirements.txt

@@ -0,0 +1,8 @@
+accelerate==0.25.0
+bitsandbytes==0.41.2.post2
+protobuf==4.25.1
+scipy==1.11.4
+sentencepiece==0.1.99
+spaces==0.19.2
+torch==2.0.0
+transformers==4.36.2

style.css

@@ -0,0 +1,10 @@
+h1 {
+  text-align: center;
+}
+
+#duplicate-button {
+  margin: auto;
+  color: #fff;
+  background: #1565c0;
+  border-radius: 100vh;
+}