Instructions to use ethzanalytics/open_llama_13b-sharded-8bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ethzanalytics/open_llama_13b-sharded-8bit with Transformers:

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

pipe = pipeline("text-generation", model="ethzanalytics/open_llama_13b-sharded-8bit")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("ethzanalytics/open_llama_13b-sharded-8bit")
model = AutoModelForCausalLM.from_pretrained("ethzanalytics/open_llama_13b-sharded-8bit", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ethzanalytics/open_llama_13b-sharded-8bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ethzanalytics/open_llama_13b-sharded-8bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ethzanalytics/open_llama_13b-sharded-8bit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/ethzanalytics/open_llama_13b-sharded-8bit

SGLang

How to use ethzanalytics/open_llama_13b-sharded-8bit 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 "ethzanalytics/open_llama_13b-sharded-8bit" \
    --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": "ethzanalytics/open_llama_13b-sharded-8bit",
		"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 "ethzanalytics/open_llama_13b-sharded-8bit" \
        --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": "ethzanalytics/open_llama_13b-sharded-8bit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use ethzanalytics/open_llama_13b-sharded-8bit with Docker Model Runner:
```
docker model run hf.co/ethzanalytics/open_llama_13b-sharded-8bit
```

open_llama_13b-sharded-8bit / README.md

pszemraj

Update README.md

39a3355 about 3 years ago

preview code

Raw

History Blame Contribute Delete

1.06 kB

	---
	license: apache-2.0
	library_name: transformers
	pipeline_tag: text-generation
	tags:
	- 8bit
	- sharded
	- open_llama
	inference: False
	---

	# open_llama_13b-sharded-8bit

	<a href="https://colab.research.google.com/gist/pszemraj/166ad661c6af1e024d4e2897621fc886/open_llama_13b-sharded-8bit-example.ipynb">
	<img src="https://colab.research.google.com/assets/colab-badge.svg" alt="Open In Colab"/>
	</a>

	This is [open_llama_13b](https://huggingface.co/openlm-research/open_llama_13b) sharded into 2 GB shards, and in 8-bit precision using `bitsandbytes==0.38.0`. Please refer to the original model card for details.



	## loading

	```sh
	pip install -U -q sentencepiece transformers accelerate bitsandbytes
	```

	load the model and tokenizer:

	```python
	import torch
	from transformers import LlamaTokenizer, LlamaForCausalLM

	model_name = "ethzanalytics/open_llama_13b-sharded-8bit"
	tokenizer = LlamaTokenizer.from_pretrained(model_name, use_fast=False)
	model = LlamaForCausalLM.from_pretrained(
	model_name,
	load_in_8bit=True,
	device_map="auto",
	)
	```