Templates (#12294)

Co-authored-by: Harrison Chase <hw.chase.17@gmail.com>
Co-authored-by: Lance Martin <lance@langchain.dev>
Co-authored-by: Jacob Lee <jacoblee93@gmail.com>

templates/sql-llamacpp/README.md

@@ -0,0 +1,36 @@
+# SQL with LLaMA2 using Ollama
+
+This template allows you to chat with a SQL database in natural language in private, using an open source LLM.
+
+## LLama.cpp
+ 
+### Enviorment
+
+From [here](https://python.langchain.com/docs/guides/local_llms) and [here](https://github.com/abetlen/llama-cpp-python/blob/main/docs/install/macos.md).
+
+```
+wget https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-MacOSX-arm64.sh
+bash Miniforge3-MacOSX-arm64.sh
+conda create -n llama python=3.9.16
+conda activate /Users/rlm/miniforge3/envs/llama
+CMAKE_ARGS="-DLLAMA_METAL=on" FORCE_CMAKE=1 pip install -U llama-cpp-python --no-cache-dir
+```
+
+### LLM
+
+It will download Mistral-7b model from [here](https://huggingface.co/TheBloke/Mistral-7B-Instruct-v0.1-GGUF).
+
+You can select other files and specify their download path (browse [here](https://huggingface.co/TheBloke)).
+
+## Set up SQL DB
+
+This template includes an example DB of 2023 NBA rosters.
+
+You can see instructions to build this DB [here](https://github.com/facebookresearch/llama-recipes/blob/main/demo_apps/StructuredLlama.ipynb).
+
+## Installation
+
+```bash
+# from inside your LangServe instance
+poe add sql/llama2-ollama
+```

templates/sql-llamacpp/pyproject.toml

@@ -0,0 +1,19 @@
+[tool.poetry]
+name = "sql-llamacpp"
+version = "0.1.0"
+description = ""
+authors = ["Lance Martin <lance@langchain.dev>"]
+readme = "README.md"
+
+[tool.poetry.dependencies]
+python = ">=3.8.1,<4.0"
+langchain = ">=0.0.313"
+llama-cpp-python = ">=0.1.79"
+
+[tool.langserve]
+export_module = "sql_llamacpp"
+export_attr = "chain"
+
+[build-system]
+requires = ["poetry-core"]
+build-backend = "poetry.core.masonry.api"

templates/sql-llamacpp/sql_llamacpp/__init__.py

@@ -0,0 +1,3 @@
+from llamacpp.chain import chain
+
+__all__ = ["chain"]

templates/sql-llamacpp/sql_llamacpp/chain.py

@@ -0,0 +1,106 @@
+from langchain.llms import LlamaCpp
+from langchain.prompts import ChatPromptTemplate
+from langchain.schema.output_parser import StrOutputParser
+from langchain.schema.runnable import RunnablePassthrough
+
+# Get LLM
+import os
+import requests
+# File name and URL
+file_name = "mistral-7b-instruct-v0.1.Q4_K_M.gguf"
+url = "https://huggingface.co/TheBloke/Mistral-7B-Instruct-v0.1-GGUF/resolve/main/mistral-7b-instruct-v0.1.Q4_K_M.gguf"
+# Check if file is present in the current directory
+if not os.path.exists(file_name):
+    print(f"'{file_name}' not found. Downloading...")
+    # Download the file
+    response = requests.get(url)
+    response.raise_for_status()  # Raise an exception for HTTP errors
+    with open(file_name, 'wb') as f:
+        f.write(response.content)
+    print(f"'{file_name}' has been downloaded.")
+else:
+    print(f"'{file_name}' already exists in the current directory.")
+
+# Add the LLM downloaded from HF
+model_path = file_name
+n_gpu_layers = 1  # Metal set to 1 is enough.
+n_batch = 512  # Should be between 1 and n_ctx, consider the amount of RAM of your Apple Silicon Chip.
+llm = LlamaCpp(
+    model_path=model_path,
+    n_gpu_layers=n_gpu_layers,
+    n_batch=n_batch,
+    n_ctx=2048,
+    f16_kv=True,  # MUST set to True, otherwise you will run into problem after a couple of calls
+    verbose=True,
+)
+
+from pathlib import Path
+from langchain.utilities import SQLDatabase
+db_path = Path(__file__).parent / "nba_roster.db"
+rel = db_path.relative_to(Path.cwd())
+db_string = f"sqlite:///{rel}"
+db = SQLDatabase.from_uri(db_string, sample_rows_in_table_info=0)
+
+def get_schema(_):
+    return db.get_table_info()
+
+
+def run_query(query):
+    return db.run(query)
+
+# Prompt
+from langchain.memory import ConversationBufferMemory
+from langchain.prompts import ChatPromptTemplate, MessagesPlaceholder
+template = """Based on the table schema below, write a SQL query that would answer the user's question:
+{schema}
+
+Question: {question}
+SQL Query:"""
+prompt = ChatPromptTemplate.from_messages([
+    ("system", "Given an input question, convert it to a SQL query. No pre-amble."),
+    MessagesPlaceholder(variable_name="history"),
+    ("human", template)
+])
+
+memory = ConversationBufferMemory(return_messages=True)
+
+# Chain to query with memory 
+from langchain.schema.runnable import RunnableLambda
+
+sql_chain = (
+    RunnablePassthrough.assign(
+       schema=get_schema,
+       history=RunnableLambda(lambda x: memory.load_memory_variables(x)["history"])
+    )| prompt
+    | llm.bind(stop=["\nSQLResult:"])
+    | StrOutputParser()
+)
+
+def save(input_output):
+    output = {"output": input_output.pop("output")}
+    memory.save_context(input_output, output)
+    return output['output']
+    
+sql_response_memory = RunnablePassthrough.assign(output=sql_chain) | save
+
+# Chain to answer
+template = """Based on the table schema below, question, sql query, and sql response, write a natural language response:
+{schema}
+
+Question: {question}
+SQL Query: {query}
+SQL Response: {response}"""
+prompt_response = ChatPromptTemplate.from_messages([
+    ("system", "Given an input question and SQL response, convert it to a natural language answer. No pre-amble."),
+    ("human", template)
+])
+
+chain = (
+    RunnablePassthrough.assign(query=sql_response_memory) 
+    | RunnablePassthrough.assign(
+        schema=get_schema,
+        response=lambda x: db.run(x["query"]),
+    )
+    | prompt_response 
+    | llm
+)