Instructions to use lutfiia/dash-malone-7b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use lutfiia/dash-malone-7b with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-7B-Instruct")
model = PeftModel.from_pretrained(base_model, "lutfiia/dash-malone-7b")

Transformers

How to use lutfiia/dash-malone-7b with Transformers:

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

pipe = pipeline("text-generation", model="lutfiia/dash-malone-7b")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("lutfiia/dash-malone-7b", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use lutfiia/dash-malone-7b with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/lutfiia/dash-malone-7b

SGLang

How to use lutfiia/dash-malone-7b 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 "lutfiia/dash-malone-7b" \
    --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": "lutfiia/dash-malone-7b",
		"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 "lutfiia/dash-malone-7b" \
        --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": "lutfiia/dash-malone-7b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use lutfiia/dash-malone-7b with Docker Model Runner:
```
docker model run hf.co/lutfiia/dash-malone-7b
```

Dash Malone — film-noir LoRA for Qwen2.5-7B-Instruct

A QLoRA adapter that turns Qwen2.5-7B-Instruct into Dash Malone, a hard-boiled 1947 Los Angeles private detective who answers every question in terse, smoky noir prose — while keeping his facts straight.

Training details

Base: Qwen/Qwen2.5-7B-Instruct, 4-bit NF4 quantized
LoRA: r=16, alpha=32, dropout 0.05, targets: q/k/v/o + gate/up/down projections
Data: 728 chat examples (425 noir persona, 303 general instruction data to prevent forgetting)
3 epochs, lr 1e-4 cosine, batch 1 × grad-accum 8 — trained on a 6 GB RTX 3060 laptop in ~56 min
Trainable parameters: 40.4M (0.53% of the model)

Usage

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
import torch

bnb = BitsAndBytesConfig(load_in_4bit=True, bnb_4bit_quant_type="nf4",
                         bnb_4bit_use_double_quant=True,
                         bnb_4bit_compute_dtype=torch.bfloat16)
base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-7B-Instruct",
                                            quantization_config=bnb, device_map="auto")
model = PeftModel.from_pretrained(base, "YOUR-USERNAME/dash-malone-7b")
tok = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-7B-Instruct")

system = ("You are Dash Malone, a hard-boiled film-noir private detective in 1947 "
          "Los Angeles. Answer every question in terse, smoky noir prose: short "
          "sentences, rain and neon metaphors, world-weary but honorable.")
messages = [{"role": "system", "content": system},
            {"role": "user", "content": "Why is the sky blue?"}]
prompt = tok.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
ids = tok(prompt, return_tensors="pt").to(model.device)
out = model.generate(**ids, max_new_tokens=200)
print(tok.decode(out[0][ids["input_ids"].shape[1]:], skip_special_tokens=True))

Limitations

Trained for style, not factual accuracy: answers are in-character, not encyclopedic.
With a generic system prompt, the model behaves like a slightly terser Qwen2.5-7B-Instruct — the persona is switched on by the Dash system prompt above.

Framework versions

PEFT 0.20.0
transformers 5.17.0

Downloads last month: 21

Model tree for lutfiia/dash-malone-7b

Base model

Qwen/Qwen2.5-7B

Finetuned

Qwen/Qwen2.5-7B-Instruct

Adapter

(2823)

this model