Instructions to use athirdpath/Iambe-20b-DARE-v2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use athirdpath/Iambe-20b-DARE-v2 with Transformers:

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

pipe = pipeline("text-generation", model="athirdpath/Iambe-20b-DARE-v2")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("athirdpath/Iambe-20b-DARE-v2")
model = AutoModelForCausalLM.from_pretrained("athirdpath/Iambe-20b-DARE-v2")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use athirdpath/Iambe-20b-DARE-v2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "athirdpath/Iambe-20b-DARE-v2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "athirdpath/Iambe-20b-DARE-v2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/athirdpath/Iambe-20b-DARE-v2

SGLang

How to use athirdpath/Iambe-20b-DARE-v2 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 "athirdpath/Iambe-20b-DARE-v2" \
    --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": "athirdpath/Iambe-20b-DARE-v2",
		"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 "athirdpath/Iambe-20b-DARE-v2" \
        --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": "athirdpath/Iambe-20b-DARE-v2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use athirdpath/Iambe-20b-DARE-v2 with Docker Model Runner:
```
docker model run hf.co/athirdpath/Iambe-20b-DARE-v2
```

Discuss about DARE

by HoangHa - opened Dec 7, 2023

Discussion

HoangHa

Dec 7, 2023

I tested the model and it works perfectly for role-playing. I'm curious of how can we convert the model to DARE?

athirdpath

Owner Dec 7, 2023

Ah, I used MergeKit. The DARE paper talking about converting the model to a DARE format gave me quite a bit of confusion too, but the repo handles all of that and produces a normal model (like this one) again at the end.

HoangHa

Dec 8, 2023

Thank you for pointing out.

Wow I saw the branch “dare” in Mergekit repo. Is that what you did?

And one more stupid question. I tried merging models but I can’t see where is the tokenizer.model file? How can you get it?

athirdpath

Owner Dec 8, 2023

Yes, I used the dare branch, I also posted the yml file I used at the bottom of this model's card.

That's odd it didn't handle the tokenizer automatically, but if you copy the one from the base model into the new model's folder it should theoretically work.

HoangHa

Dec 8, 2023

wonderful, thank you. Really admire your work.

HoangHa changed discussion status to closed Dec 8, 2023

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