Instructions to use Gille/StrangeMerges_46-7B-dare_ties with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Gille/StrangeMerges_46-7B-dare_ties with Transformers:

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

pipe = pipeline("text-generation", model="Gille/StrangeMerges_46-7B-dare_ties")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Gille/StrangeMerges_46-7B-dare_ties")
model = AutoModelForCausalLM.from_pretrained("Gille/StrangeMerges_46-7B-dare_ties")

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Gille/StrangeMerges_46-7B-dare_ties with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Gille/StrangeMerges_46-7B-dare_ties"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Gille/StrangeMerges_46-7B-dare_ties",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Gille/StrangeMerges_46-7B-dare_ties

SGLang

How to use Gille/StrangeMerges_46-7B-dare_ties 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 "Gille/StrangeMerges_46-7B-dare_ties" \
    --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": "Gille/StrangeMerges_46-7B-dare_ties",
		"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 "Gille/StrangeMerges_46-7B-dare_ties" \
        --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": "Gille/StrangeMerges_46-7B-dare_ties",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Gille/StrangeMerges_46-7B-dare_ties with Docker Model Runner:
```
docker model run hf.co/Gille/StrangeMerges_46-7B-dare_ties
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

StrangeMerges_46-7B-dare_ties

StrangeMerges_46-7B-dare_ties is a merge of the following models using LazyMergekit:

🧩 Configuration

models:
  - model: Gille/StrangeMerges_45-7B-dare_ties
    parameters:
      weight: 0.4
      density: 0.53
  - model: kettleguts/zephyr-7b-beta_sparse05
    parameters:
      weight: 0.4
      density: 0.53
  - model: chihoonlee10/T3Q-Mistral-Orca-Math-DPO
    parameters:
      weight: 0.2
      density: 0.53
base_model: liminerity/M7-7b
merge_method: dare_ties
dtype: bfloat16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Gille/StrangeMerges_46-7B-dare_ties"
messages = [{"role": "user", "content": "What is a large language model?"}]

tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    torch_dtype=torch.float16,
    device_map="auto",
)

outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	69.96
AI2 Reasoning Challenge (25-Shot)	67.24
HellaSwag (10-Shot)	86.40
MMLU (5-Shot)	62.17
TruthfulQA (0-shot)	65.17
Winogrande (5-shot)	79.48
GSM8k (5-shot)	59.29

Downloads last month: 62

Safetensors

Model size

7B params

Tensor type

BF16

Model tree for Gille/StrangeMerges_46-7B-dare_ties

Gille/StrangeMerges_45-7B-dare_ties

chihoonlee10/T3Q-Mistral-Orca-Math-DPO

kettleguts/zephyr-7b-beta_sparse05

Merge model

this model

Merges

2 models

Quantizations

1 model

Evaluation results

normalized accuracy on AI2 Reasoning Challenge (25-Shot)
test set Open LLM Leaderboard

67.240
normalized accuracy on HellaSwag (10-Shot)
validation set Open LLM Leaderboard

86.400
accuracy on MMLU (5-Shot)
test set Open LLM Leaderboard

62.170
mc2 on TruthfulQA (0-shot)
validation set Open LLM Leaderboard

65.170
accuracy on Winogrande (5-shot)
validation set Open LLM Leaderboard

79.480
accuracy on GSM8k (5-shot)
test set Open LLM Leaderboard

59.290