Instructions to use Cognitive-Machines-Labs/Dolus-14b-Mini with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Cognitive-Machines-Labs/Dolus-14b-Mini with Transformers:

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

pipe = pipeline("text-generation", model="Cognitive-Machines-Labs/Dolus-14b-Mini")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Cognitive-Machines-Labs/Dolus-14b-Mini")
model = AutoModelForCausalLM.from_pretrained("Cognitive-Machines-Labs/Dolus-14b-Mini")
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 Cognitive-Machines-Labs/Dolus-14b-Mini with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Cognitive-Machines-Labs/Dolus-14b-Mini"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Cognitive-Machines-Labs/Dolus-14b-Mini",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Cognitive-Machines-Labs/Dolus-14b-Mini

SGLang

How to use Cognitive-Machines-Labs/Dolus-14b-Mini 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 "Cognitive-Machines-Labs/Dolus-14b-Mini" \
    --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": "Cognitive-Machines-Labs/Dolus-14b-Mini",
		"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 "Cognitive-Machines-Labs/Dolus-14b-Mini" \
        --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": "Cognitive-Machines-Labs/Dolus-14b-Mini",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Cognitive-Machines-Labs/Dolus-14b-Mini with Docker Model Runner:
```
docker model run hf.co/Cognitive-Machines-Labs/Dolus-14b-Mini
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Dolus-14b-Mini

Ursidae-12b-Mini
The little sister to Ursidae-300b,Dolus 14b Mini has been developed with a focus on complex multi-step chain of thought problem solving while still being deployable on edge systems. A model focused complex multi-step chain of thought problem solving while still being deployable on edge systems. Now better at reasoning!

Main Goals:

Dolus was designed to address specific issues found in other chat models:

Overcome limitations in logical reasoning found in other chat models.
Efficiently solve complex, multi-step problems.
Provide better decision-making assistance by enhancing the model's ability to reason and think critically.
Removing restrictions and allowing the model to gain a true understanding of reality, greatly increasing overall results.

By focusing on these specific goals, the Ursidae-12b-Mini aims to provide a more sophisticated AI system that excels at critical thinking and problem-solving tasks requiring advanced logical reasoning skills. Its compact design makes it an efficient choice for applications where high cognitive abilities are necessary without occupying excessive computing resources.

Recommended Settings:

Defaults:

min_p: 0.074
top_k: 40
repetition_penalty: 1.12
temp: 1.18
context: 8192

Creative:

min_p: 0.062
top_k: 40
repetition_penalty: 1.11
temp: 1.24
context: 8192

Benchmarks:

PENDING FULL EVAL

Downloads last month: 8

Safetensors

Model size

12B params

Tensor type

BF16

Model tree for Cognitive-Machines-Labs/Dolus-14b-Mini

Merges

1 model

Quantizations

5 models