Instructions to use MetaIX/OpenAssistant-Llama-30b-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MetaIX/OpenAssistant-Llama-30b-4bit with Transformers:

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

pipe = pipeline("text-generation", model="MetaIX/OpenAssistant-Llama-30b-4bit")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MetaIX/OpenAssistant-Llama-30b-4bit")
model = AutoModelForCausalLM.from_pretrained("MetaIX/OpenAssistant-Llama-30b-4bit")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use MetaIX/OpenAssistant-Llama-30b-4bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MetaIX/OpenAssistant-Llama-30b-4bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MetaIX/OpenAssistant-Llama-30b-4bit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/MetaIX/OpenAssistant-Llama-30b-4bit

SGLang

How to use MetaIX/OpenAssistant-Llama-30b-4bit 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 "MetaIX/OpenAssistant-Llama-30b-4bit" \
    --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": "MetaIX/OpenAssistant-Llama-30b-4bit",
		"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 "MetaIX/OpenAssistant-Llama-30b-4bit" \
        --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": "MetaIX/OpenAssistant-Llama-30b-4bit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use MetaIX/OpenAssistant-Llama-30b-4bit with Docker Model Runner:
```
docker model run hf.co/MetaIX/OpenAssistant-Llama-30b-4bit
```

MetaIX commited on May 19, 2023

Commit

76e9f07

1 Parent(s): c8612e2

Update README.md

Browse files

Files changed (1) hide show

README.md +7 -2

README.md CHANGED Viewed

@@ -1,7 +1,12 @@
 <p><strong><font size="5">Information</font></strong></p>
 OpenAssistant-Llama-30B-4-bit working with GPTQ versions used in Oobabooga's Text Generation Webui and KoboldAI.
-<p>There are 3 quantized versions, one is quantized using GPTQ's <i>--true-sequential</i> and <i>--act-order</i> optimizations, the second is quantized using GPTQ's <i>--true-sequential</i> and <i>--groupsize 128</i> optimization, and the third one is quantized for GGML using q4_1</p>
-This was made using <a href="https://huggingface.co/OpenAssistant/oasst-sft-7-llama-30b-xor">Open Assistant's native fine-tune</a> of Llama 30b on their dataset.</p>
 <p><strong><font size="5">Update 05.19.2023</font></strong></p>
 <p>Updated the ggml quantizations to be compatible with the latest version of llamacpp.</p>

 <p><strong><font size="5">Information</font></strong></p>
 OpenAssistant-Llama-30B-4-bit working with GPTQ versions used in Oobabooga's Text Generation Webui and KoboldAI.
+<p>This was made using <a href="https://huggingface.co/OpenAssistant/oasst-sft-7-llama-30b-xor">Open Assistant's native fine-tune</a> of Llama 30b on their dataset.</p>
+<p><strong>What's included</strong></p>
+<P>GPTQ: 2 quantized versions. One quantized --true-sequential and act-order optimizations, and the other was quantized using --true-sequential --groupsize 128 optimizations</P>
+<P>GGML: 3 quantized versions. One quantized using q4_1, another one was quantized using q5_0, and the last one was quantized using q5_1.</P>
 <p><strong><font size="5">Update 05.19.2023</font></strong></p>
 <p>Updated the ggml quantizations to be compatible with the latest version of llamacpp.</p>