Instructions to use TheBloke/Llama-2-7B-Chat-GGML with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use TheBloke/Llama-2-7B-Chat-GGML with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="TheBloke/Llama-2-7B-Chat-GGML")# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("TheBloke/Llama-2-7B-Chat-GGML", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use TheBloke/Llama-2-7B-Chat-GGML with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "TheBloke/Llama-2-7B-Chat-GGML" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "TheBloke/Llama-2-7B-Chat-GGML", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/TheBloke/Llama-2-7B-Chat-GGML
- SGLang
How to use TheBloke/Llama-2-7B-Chat-GGML with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "TheBloke/Llama-2-7B-Chat-GGML" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "TheBloke/Llama-2-7B-Chat-GGML", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "TheBloke/Llama-2-7B-Chat-GGML" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "TheBloke/Llama-2-7B-Chat-GGML", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use TheBloke/Llama-2-7B-Chat-GGML with Docker Model Runner:
docker model run hf.co/TheBloke/Llama-2-7B-Chat-GGML
Prompts for Question Answering Assistant
I am building a question answering assistant using the model. What prompts can I use so that the responses generated by the model are brief, to the point and coherent?
An example using Langchain's prompts:
from langchain.llms import CTransformers
from langchain.prompts import PromptTemplate
from langchain.chains import LLMChain
model_path : str = "models/Llama-2-7B-Chat-GGML/llama-2-7b-chat.ggmlv3.q4_0.bin"
llm = CTransformers(
model=model_path,
model_type='llama',
)
prompt = PromptTemplate(
input_variables=["product"],
template="What is a good name for a company that makes {product}? Answer with a simple list only.",
)
llmchain = LLMChain(llm=llm, prompt=prompt)
print(llmchain.run("podcast player"))
The prompt template of Llama2 is <s>[INST]\n<<SYS>>\n{system_prompt}\n<</SYS>>\n\n{user_prompt}[/INST] (ref. https://huggingface.co/spaces/huggingface-projects/llama-2-7b-chat/blob/main/model.py). Does CTransformers library handles this automatically? Otherwise how can it work your example? Thanks, Vincenzo
The PR that dealing with this issue just got merged: https://github.com/huggingface/transformers/pull/25323