Instructions to use asyafiqe/Merak-7B-v2-GGML with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use asyafiqe/Merak-7B-v2-GGML with Transformers:

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

pipe = pipeline("text-generation", model="asyafiqe/Merak-7B-v2-GGML")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("asyafiqe/Merak-7B-v2-GGML")
model = AutoModelForCausalLM.from_pretrained("asyafiqe/Merak-7B-v2-GGML")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use asyafiqe/Merak-7B-v2-GGML with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "asyafiqe/Merak-7B-v2-GGML"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "asyafiqe/Merak-7B-v2-GGML",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/asyafiqe/Merak-7B-v2-GGML

SGLang

How to use asyafiqe/Merak-7B-v2-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 "asyafiqe/Merak-7B-v2-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": "asyafiqe/Merak-7B-v2-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 "asyafiqe/Merak-7B-v2-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": "asyafiqe/Merak-7B-v2-GGML",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use asyafiqe/Merak-7B-v2-GGML with Docker Model Runner:
```
docker model run hf.co/asyafiqe/Merak-7B-v2-GGML
```

Merak-7B-v2.ggmlv3.q4_0.bin

by Sandrian - opened Sep 7, 2023

Discussion

Sandrian

Sep 7, 2023

•

edited Sep 8, 2023

First of all thank you for this great enhancement to provide merak llm in ggml version.

I want to ask, did you test the Merak-7B-v2.ggmlv3.q4_0.bin yet? i already tried this model with Bahasa Indonesia prompt but no output to show. But it goes well when i tried with English prompt

Thank you

megaaziib

Sep 7, 2023

same me too, is there a trigger prompt for this model?

asyafiqe

Owner Sep 9, 2023

Did you use oobabooga's textgen? It's working fine either in chat mode or Default mode with QA prompt.

According to the Merak's model card the prompt format is following:
<|prompt|>{question}\n<|answer|>

Btw, try out further fine-tuned Merak: 🦚Merak-7B-v3-Mini-Orca-GGUF🐳
Compared to Merak-7B-v2, it gives more coherent and longer answer.

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