Llama-3.1-8B-Instruct-abliterated_via_adapter

Instructions to use jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter with Transformers:

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

pipe = pipeline("text-generation", model="jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter")
model = AutoModelForCausalLM.from_pretrained("jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter", device_map="auto")
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 jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter

SGLang

How to use jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter 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 "jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter" \
    --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": "jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter",
		"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 "jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter" \
        --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": "jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter with Docker Model Runner:
```
docker model run hf.co/jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter
```

Llama-3.1-8B-Instruct-abliterated_via_adapter

This model is a merge of pre-trained language models created using mergekit.

A LoRA was applied to "abliterate" refusals in meta-llama/Meta-Llama-3.1-8B-Instruct. The result appears to work despite the LoRA having been derived from Llama 3 instead of Llama 3.1, which implies that there is significant feature commonality between the 3 and 3.1 models.

The LoRA was extracted from failspy/Meta-Llama-3-8B-Instruct-abliterated-v3 using meta-llama/Meta-Llama-3-8B-Instruct as a base.

Built with Llama.

Merge Details

Merge Method

This model was merged using the task arithmetic merge method using meta-llama/Meta-Llama-3.1-8B-Instruct + grimjim/Llama-3-Instruct-abliteration-LoRA-8B as a base.

Configuration

The following YAML configuration was used to produce this model:

base_model: meta-llama/Meta-Llama-3.1-8B-Instruct+grimjim/Llama-3-Instruct-abliteration-LoRA-8B
dtype: bfloat16
merge_method: task_arithmetic
parameters:
  normalize: false
slices:
- sources:
  - layer_range: [0, 32]
    model: meta-llama/Meta-Llama-3.1-8B-Instruct+grimjim/Llama-3-Instruct-abliteration-LoRA-8B
    parameters:
      weight: 1.0

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	22.95
IFEval (0-Shot)	48.70
BBH (3-Shot)	29.42
MATH Lvl 5 (4-Shot)	12.39
GPQA (0-shot)	8.50
MuSR (0-shot)	9.26
MMLU-PRO (5-shot)	29.46

Downloads last month: 9

Safetensors

Model size

8B params

Tensor type

BF16

Model tree for jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter

grimjim/Llama-3-Instruct-abliteration-LoRA-8B

meta-llama/Llama-3.1-8B-Instruct

Merge model

this model

Paper for jinmrong/Llama-3.1-8B-Instruct-abliterated_via_adapter

Editing Models with Task Arithmetic

Paper • 2212.04089 • Published Dec 8, 2022 • 9

Evaluation results

strict accuracy on IFEval (0-Shot)
Open LLM Leaderboard

48.700
normalized accuracy on BBH (3-Shot)
Open LLM Leaderboard

29.420
exact match on MATH Lvl 5 (4-Shot)
Open LLM Leaderboard

12.390
acc_norm on GPQA (0-shot)
Open LLM Leaderboard

8.500
acc_norm on MuSR (0-shot)
Open LLM Leaderboard

9.260
accuracy on MMLU-PRO (5-shot)
test set Open LLM Leaderboard

29.460