Instructions to use enterprise-explorers/Llama-2-7b-chat-coreml with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use enterprise-explorers/Llama-2-7b-chat-coreml with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="enterprise-explorers/Llama-2-7b-chat-coreml")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("enterprise-explorers/Llama-2-7b-chat-coreml")
model = AutoModelForCausalLM.from_pretrained("enterprise-explorers/Llama-2-7b-chat-coreml")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use enterprise-explorers/Llama-2-7b-chat-coreml with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "enterprise-explorers/Llama-2-7b-chat-coreml"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "enterprise-explorers/Llama-2-7b-chat-coreml",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/enterprise-explorers/Llama-2-7b-chat-coreml

SGLang

How to use enterprise-explorers/Llama-2-7b-chat-coreml 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 "enterprise-explorers/Llama-2-7b-chat-coreml" \
    --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": "enterprise-explorers/Llama-2-7b-chat-coreml",
		"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 "enterprise-explorers/Llama-2-7b-chat-coreml" \
        --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": "enterprise-explorers/Llama-2-7b-chat-coreml",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use enterprise-explorers/Llama-2-7b-chat-coreml with Docker Model Runner:
```
docker model run hf.co/enterprise-explorers/Llama-2-7b-chat-coreml
```

M1 8G RAM macbook air run at 0.01token/second, something is wrong right?

by DanielCL - opened Aug 14, 2023

Discussion

DanielCL

Aug 14, 2023

•

edited Aug 14, 2023

According to the demo video at https://huggingface.co/blog/swift-coreml-llm, the speed of this model was 600token/second (pretty cool).

So, I have cloned https://github.com/huggingface/swift-chat, and downloaded this model successfully at coreml-projects/Llama-2-7b-chat-coreml, and it can run without error.

Although I don't know which chip M2 or M1 or something even more powerful is used in the demo, but my M1 8G RAM macbook air can only run at 0.01token/s seems very wrong.

What am I doing wrong?

Thank you so much!

Xenova

Aug 14, 2023

•

edited Aug 14, 2023

The demo video you refer to indicates a generation speed of 6.42 tokens per second (as opposed to 600), but 0.01 tokens per second definitely seems wrong.

It might simply be an issue with timing, as if this was indeed the case, it would take 100 seconds per token (which would mean the text shown would take ~30 minutes (which I doubt is the case). Can you take a recording, or maybe just answer whether it is actually that slow?

Xenova changed discussion status to closed Aug 14, 2023

Xenova changed discussion status to open Aug 14, 2023

DanielCL

Aug 16, 2023

Thanks @Xenova for your reply. It is very very slow in deed about 20 minutes or more per answer, every time I run it and every question on each run.

nifta

Aug 19, 2023

Not sure if a "me too" is useful, but it's the same on my 2022 m2 16gb air. i get ~0.02 tokens/s. I've tried release builds as well as debug. Happy to provide more details/test stuff.

chrispie

Aug 21, 2023

I made an issue @ https://github.com/huggingface/swift-chat/issues/3 and get 0.02 tokens/s too.

pcuenq

Enterprise Explorers org Aug 23, 2023

My computer is an M1 Max with 64 GB. I believe the main factor that affects performance in this case is the amount of RAM installed. Because the model is so large, the system will swap if there's not enough memory to hold everything in RAM, and that will kill performance.

I'll run some tests and answer in the github issue!

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment