Update

.gitignore

@@ -0,0 +1,6 @@
+install-mine.sh
+install-sw.sh
+install-sw1.sh
+okteto-up.bat
+okteto.yml
+start-sshd.sh

.ruff.toml

@@ -0,0 +1,18 @@
+# Assume Python 3.10.
+target-version = "py310"
+# Decrease the maximum line length to 79 characters.
+line-length = 1000
+
+# pyflakes, pycodestyle, isort
+# flake8 YTT, pydocstyle D, pylint PLC
+select = ["F", "E", "W", "I001", "YTT", "D", "PLC"]
+# select = ["ALL"]
+
+# "D100" Missing docstring in public module
+# D103 Missing docstring in public function
+# D101 Missing docstring in public class
+# `multi-line-summary-first-line` (D212)
+# `one-blank-line-before-class` (D203)
+extend-ignore = ["D100", "D103", "D101", "D212", "D203"]
+
+exclude = [".venv"]

.stignore

@@ -0,0 +1,102 @@
+.ruuf_cache
+.git
+# 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
+pip-wheel-metadata
+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
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+
+# Flask stuff:
+instance
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build
+
+# PyBuilder
+target
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default
+ipython_config.py
+
+# pyenv
+.python-version
+
+# celery beat schedule file
+celerybeat-schedule
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env
+venv
+ENV
+env.bak
+venv.bak
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mypy
+.mypy_cache
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre

README.md

@@ -1,15 +1,13 @@
 ---
 title: falcon-mini
-emoji: 🦅💸
-colorFrom: red
+emoji: 🦅
+colorFrom: yellow
 colorTo: blue
-sdk: docker
-pinned: false
-app_port: 7860
+sdk: gradio
+sdk_version: 3.36.1
+app_file: demo.py
+pinned: true
 duplicated_from: matthoffner/falcon-mini
 ---
 
-# falcon-7b-instruct-q3
-
-## <a href="https://github.com/cmp-nct/ggllm.cpp" target="_blank">ggllm.cpp</a>
-## ctransformers
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

README.md-

@@ -0,0 +1,15 @@
+---
+title: falcon-mini
+emoji: 🦅💸
+colorFrom: red
+colorTo: blue
+sdk: docker
+pinned: false
+app_port: 7860
+duplicated_from: matthoffner/falcon-mini
+---
+
+# falcon-7b-instruct-q3
+
+## <a href="https://github.com/cmp-nct/ggllm.cpp" target="_blank">ggllm.cpp</a>
+## ctransformers

api.py → api.py-

@@ -1,14 +1,14 @@
-import fastapi
 import json
+from typing import Any, Generator, List
+
+import fastapi
 import uvicorn
+from ctransformers import AutoModelForCausalLM
 from fastapi import HTTPException
-from fastapi.responses import HTMLResponse
 from fastapi.middleware.cors import CORSMiddleware
+from pydantic import BaseModel
 from sse_starlette.sse import EventSourceResponse
 from starlette.responses import StreamingResponse
-from ctransformers import AutoModelForCausalLM
-from pydantic import BaseModel
-from typing import List, Dict, Any, Generator
 
 llm = AutoModelForCausalLM.from_pretrained("TheBloke/falcon-40b-instruct-GGML", model_file="falcon40b-instruct.ggmlv3.q2_K.bin",
                                            model_type="falcon", threads=8)
@@ -41,7 +41,7 @@ async def completion(request: ChatCompletionRequestV0, response_mode=None):
 async def chat(request: ChatCompletionRequest):
     combined_messages = ' '.join([message.content for message in request.messages])
     tokens = llm.tokenize(combined_messages)
-    
+
     try:
         chat_chunks = llm.generate(tokens)
     except Exception as e:
@@ -76,4 +76,4 @@ async def chat(request: ChatCompletionRequestV0, response_mode=None):
     return EventSourceResponse(server_sent_events(tokens, llm))
 
 if __name__ == "__main__":
-  uvicorn.run(app, host="0.0.0.0", port=8000)
+  uvicorn.run(app, host="0.0.0.0", port=8000)

demo.py

@@ -1,16 +1,43 @@
-import os
-from ctransformers import AutoModelForCausalLM
 import gradio as gr
+import psutil
+from ctransformers import AutoModelForCausalLM
+from huggingface_hub import hf_hub_download
+
+_ = """
+llm = AutoModelForCausalLM.from_pretrained(
+    "TheBloke/falcon-7b-instruct-GGML",
+    model_file="falcon7b-instruct.ggmlv3.q4_0.bin",
+    model_type="falcon",
+    threads=psutil.cpu_count(logical=False))
+"""
+
+URL = "https://huggingface.co/TheBloke/falcon-7b-instruct-GGML/blob/ggmlv3/falcon7b-instruct.ggmlv3.q4_0.bin"
+repo_id = "/".join(urlparse(URL).path.strip("/").split("/")[:2])
+
+model_file = Path(URL).name
 
+_ = hf_hub_download(
+    repo_id=repo_id,
+    revision="ggmlv3",
+    filename=model_file,
+    local_dir="models",
+    # local_dir_use_symlinks=True,
+)
 
-llm = AutoModelForCausalLM.from_pretrained("TheBloke/falcon-7b-instruct-GGML", model_file="falcon7b-instruct.ggmlv3.q4_0.bin",
-                                           model_type="falcon", threads=8)
+llm = AutoModelForCausalLM.from_pretrained(
+    # repo_id,  # "TheBloke/falcon-7b-instruct-GGML",
+    # model_file=model_file,
+    # model_file=_,
+    _,
+    model_type="falcon",
+    threads=psutil.cpu_count(logical=False),
+)
 
 
-TITLE = """<h1 align="center">🦅 falcon-mini (7b-q4_0) ggml demo 🦅</h1>"""
+TITLE = f"""<h1 align="center">🦅 falcon-mini ({model_file}) ggml demo 🦅</h1>"""
 USER_NAME = "User"
 BOT_NAME = "Falcon"
-DEFAULT_INSTRUCTIONS = f"""The following is a conversation between a highly knowledgeable and intelligent AI assistant, called Falcon, and a human user, called User. In the following interactions, User and Falcon will converse in natural language, and Falcon will answer User's questions. Falcon was built to be respectful, polite and inclusive. Falcon was built by the Technology Innovation Institute in Abu Dhabi. Falcon will never decline to answer a question, and always attempts to give an answer that User would be satisfied with. It knows a lot, and always tells the truth. The conversation begins.
+DEFAULT_INSTRUCTIONS = """The following is a conversation between a highly knowledgeable and intelligent AI assistant, called Falcon, and a human user, called User. In the following interactions, User and Falcon will converse in natural language, and Falcon will answer User's questions. Falcon was built to be respectful, polite and inclusive. Falcon was built by the Technology Innovation Institute in Abu Dhabi. Falcon will never decline to answer a question, and always attempts to give an answer that User would be satisfied with. It knows a lot, and always tells the truth. The conversation begins.
 """
 RETRY_COMMAND = "/retry"
 STOP_STR = f"\n{USER_NAME}:"
@@ -147,7 +174,7 @@ def chat():
         run_chat,
         [inputs, chatbot, instructions, temperature, top_p],
         outputs=[chatbot],
-        show_progress=False,
+        show_progress="minimal",
     )
     inputs.submit(lambda: "", inputs=None, outputs=inputs)
     delete_turn_button.click(delete_last_turn, inputs=[chatbot], outputs=[chatbot])
@@ -155,7 +182,7 @@ def chat():
         run_retry,
         [inputs, chatbot, instructions, temperature, top_p],
         outputs=[chatbot],
-        show_progress=False,
+        show_progress="minimal",
     )
     clear_chat_button.click(clear_chat, [], chatbot)
 
@@ -176,7 +203,7 @@ def get_demo():
         with gr.Row():
             with gr.Column():
                 gr.Markdown(
-                    """**Chat with [Falcon-7B-Instruct](https://huggingface.co/tiiuae/falcon-7b-instruct), brainstorm ideas, discuss your holiday plans, and more!** 
+                    """**Chat with [Falcon-7B-Instruct](https://huggingface.co/tiiuae/falcon-7b-instruct), brainstorm ideas, discuss your holiday plans, and more!**
 
                     🧪 This uses a quantized [ggml](https://github.com/ggerganov/ggml) optimized for CPU. Special thanks to [ggllm.cpp](https://github.com/cmp-nct/ggllm.cpp), [ctransformers](https://github.com/marella/ctransformers), and [TheBloke](https://huggingface.co/TheBloke).
                     """
@@ -189,4 +216,4 @@ def get_demo():
 if __name__ == "__main__":
   demo = get_demo()
   demo.queue(max_size=128, concurrency_count=8)
-  demo.launch(server_name="0.0.0.0", server_port=7860)
+  demo.launch(server_name="0.0.0.0", server_port=7860)

requirements.txt

@@ -1,11 +1,7 @@
-uvicorn
 markdown
-fastapi
 loguru
-torch
-numpy
-transformers
-accelerate
+ctransformers
 langchain
-sse_starlette
-gradio
+# gradio
+psutil
+huggingface-hub

requirements.txt-

@@ -0,0 +1,11 @@
+uvicorn
+markdown
+fastapi
+loguru
+torch
+numpy
+transformers
+accelerate
+langchain
+sse_starlette
+gradio