Instructions to use jondurbin/bagel-dpo-34b-v0.5 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use jondurbin/bagel-dpo-34b-v0.5 with Transformers:

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

pipe = pipeline("text-generation", model="jondurbin/bagel-dpo-34b-v0.5")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("jondurbin/bagel-dpo-34b-v0.5")
model = AutoModelForMultimodalLM.from_pretrained("jondurbin/bagel-dpo-34b-v0.5")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use jondurbin/bagel-dpo-34b-v0.5 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "jondurbin/bagel-dpo-34b-v0.5"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jondurbin/bagel-dpo-34b-v0.5",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/jondurbin/bagel-dpo-34b-v0.5

SGLang

How to use jondurbin/bagel-dpo-34b-v0.5 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 "jondurbin/bagel-dpo-34b-v0.5" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jondurbin/bagel-dpo-34b-v0.5",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "jondurbin/bagel-dpo-34b-v0.5" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jondurbin/bagel-dpo-34b-v0.5",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use jondurbin/bagel-dpo-34b-v0.5 with Docker Model Runner:
```
docker model run hf.co/jondurbin/bagel-dpo-34b-v0.5
```

Can't wait for Exl2 or GGUF

by LLLLMAfficionado - opened Apr 8, 2024

Discussion

LLLLMAfficionado

Apr 8, 2024

I'll love to be able to run this bad boy locally, without waiting a minute for every letter to pop :(

Love bagel, that is YOUR bagel, breaded... not as much :D

LLLLMAfficionado

Apr 9, 2024

•

edited Apr 9, 2024

I found one here: https://huggingface.co/mradermacher/bagel-dpo-34b-v0.5-GGUF. It runs even better than 0.2, and I'm really enjoying your new model. It's very verbose when prompted to be, but sometimes it mixes context when the prompt describes more than one actor. Perhaps I still need to adjust the settings a bit; I don't think it's the model's fault.

Furthermore, I've noticed it output very intriguing narratives from shorter examples, presenting unique ways of expressing the same thing that flow much better in a given story excerpt. I'm really rocking it, I hope it only gets better.

@Undi95 , I encourage you, if I'm not overstepping, to consider this model for mixing with one of yours, such as Noromaid, not just for role-playing but also for its potential in short story telling.

Suparious

Apr 12, 2024

•

edited Apr 12, 2024

I am working on AWQ now

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