Instructions to use Nelathan/Qwen2-7B-FocusMix with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Nelathan/Qwen2-7B-FocusMix with Transformers:

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

pipe = pipeline("text-generation", model="Nelathan/Qwen2-7B-FocusMix")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Nelathan/Qwen2-7B-FocusMix")
model = AutoModelForCausalLM.from_pretrained("Nelathan/Qwen2-7B-FocusMix")
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

vLLM

How to use Nelathan/Qwen2-7B-FocusMix with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Nelathan/Qwen2-7B-FocusMix"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Nelathan/Qwen2-7B-FocusMix",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Nelathan/Qwen2-7B-FocusMix

SGLang

How to use Nelathan/Qwen2-7B-FocusMix 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 "Nelathan/Qwen2-7B-FocusMix" \
    --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": "Nelathan/Qwen2-7B-FocusMix",
		"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 "Nelathan/Qwen2-7B-FocusMix" \
        --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": "Nelathan/Qwen2-7B-FocusMix",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Nelathan/Qwen2-7B-FocusMix with Docker Model Runner:
```
docker model run hf.co/Nelathan/Qwen2-7B-FocusMix
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

FocusMix 7B

This is a model created by merging several powerful language models:

Base Model: Qwen/Qwen2-7B
Merge Stock:

FocusMix inherits the strengths of its component models, resulting in a model with:

Enhanced Focus: FocusMix leverages the fine-tuning and instruction-following capabilities of Replete-LLM, Arcee-Spark, and Einstein-v7, leading to improved accuracy and coherence in task-specific responses.
Broader Knowledge Base: The diverse training datasets of the merged models provide FocusMix with a wider range of knowledge and abilities, making it more versatile and capable of handling a wider variety of prompts and tasks.
Improved Reasoning and Problem-Solving: The inclusion of Calme-2.8, known for its reasoning and problem-solving abilities, enhances FocusMix's capacity for logical deduction and complex task execution.

Purpose: aims to provide a powerful and versatile language model that excels in:

Task-Specific Instructions: FocusMix can effectively follow specific instructions and complete tasks with high accuracy.
Complex Reasoning: The model can handle intricate prompts requiring logical deduction and problem-solving.
Diverse Knowledge Domains: FocusMix can engage in conversations and provide information across a wide range of topics.

Configuration

The following YAML configuration was used to produce this model:

merge_method: model_stock
base_model: Qwen/Qwen2-7B
models:
  - model: Replete-AI/Replete-LLM-Qwen2-7b
  - model: arcee-ai/Arcee-Spark
  - model: Weyaxi/Einstein-v7-Qwen2-7B
  - model: MaziyarPanahi/calme-2.8-qwen2-7b
dtype: bfloat16
tokenizer_source: base

Downloads last month: 12

Safetensors

Model size

8B params

Tensor type

BF16

Model tree for Nelathan/Qwen2-7B-FocusMix

MaziyarPanahi/calme-2.8-qwen2-7b

Weyaxi/Einstein-v7-Qwen2-7B

arcee-ai/Arcee-Spark

Merge model

this model

Quantizations

3 models