Instructions to use Aeshp/deepseekR1_tunedchat with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Aeshp/deepseekR1_tunedchat with Transformers:

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

pipe = pipeline("text-generation", model="Aeshp/deepseekR1_tunedchat")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Aeshp/deepseekR1_tunedchat")
model = AutoModelForCausalLM.from_pretrained("Aeshp/deepseekR1_tunedchat")
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 Aeshp/deepseekR1_tunedchat with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Aeshp/deepseekR1_tunedchat"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Aeshp/deepseekR1_tunedchat",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Aeshp/deepseekR1_tunedchat

SGLang

How to use Aeshp/deepseekR1_tunedchat 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 "Aeshp/deepseekR1_tunedchat" \
    --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": "Aeshp/deepseekR1_tunedchat",
		"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 "Aeshp/deepseekR1_tunedchat" \
        --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": "Aeshp/deepseekR1_tunedchat",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio new

How to use Aeshp/deepseekR1_tunedchat with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Aeshp/deepseekR1_tunedchat to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Aeshp/deepseekR1_tunedchat to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Aeshp/deepseekR1_tunedchat to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="Aeshp/deepseekR1_tunedchat",
    max_seq_length=2048,
)

Docker Model Runner
How to use Aeshp/deepseekR1_tunedchat with Docker Model Runner:
```
docker model run hf.co/Aeshp/deepseekR1_tunedchat
```

Aeshp/deepseekR1_tunedchat

This model is a fine-tuned version of deepseek-ai/DeepSeek-R1-Distill-Llama-8B, loaded via Unsloth in 4-bit as unsloth/DeepSeek-R1-Distill-Llama-8B-unsloth-bnb-4bit. It has been trained on customer service and general chat datasets:

The training was performed in three steps, and the final weights were merged with the base model and pushed here. It is a light model.

📝 License

This model is released under the MIT license, allowing free use, modification, and further fine-tuning.

💡 How to Fine-Tune Further

All code and instructions for further fine-tuning, inference, and pushing to the Hugging Face Hub are available in the open-source GitHub repository:
https://github.com/Aeshp/deepseekR1finetune

You can fine-tune this model on your own domain-specific data.
Please adjust hyperparameters and dataset size as needed.
Example scripts and notebooks are provided for both base model and checkpoint-based fine-tuning.

⚠️ Notes

The model may sometimes hallucinate, as is common with LLMs.
For best results, use a large, high-quality dataset for further fine-tuning to avoid overfitting.

📚 References

Hugging Face Models

Datasets

GitHub Repositories

Papers

DeepSeek R1 Paper

For all usage instructions, fine-tuning guides, and code, please see the GitHub repository.

Downloads last month: 2

Safetensors

Model size

8B params

Tensor type

F16

F32

Model tree for Aeshp/deepseekR1_tunedchat

Base model

deepseek-ai/DeepSeek-R1-0528

Quantized

(48)

this model

Aeshp
/

deepseekR1_tunedchat