Instructions to use moelanoby/Sensitive-Qwen-0.5B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use moelanoby/Sensitive-Qwen-0.5B with Transformers:

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

pipe = pipeline("text-generation", model="moelanoby/Sensitive-Qwen-0.5B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("moelanoby/Sensitive-Qwen-0.5B")
model = AutoModelForCausalLM.from_pretrained("moelanoby/Sensitive-Qwen-0.5B", device_map="auto")
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 Settings

vLLM

How to use moelanoby/Sensitive-Qwen-0.5B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "moelanoby/Sensitive-Qwen-0.5B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moelanoby/Sensitive-Qwen-0.5B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/moelanoby/Sensitive-Qwen-0.5B

SGLang

How to use moelanoby/Sensitive-Qwen-0.5B 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 "moelanoby/Sensitive-Qwen-0.5B" \
    --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": "moelanoby/Sensitive-Qwen-0.5B",
		"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 "moelanoby/Sensitive-Qwen-0.5B" \
        --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": "moelanoby/Sensitive-Qwen-0.5B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use moelanoby/Sensitive-Qwen-0.5B with Docker Model Runner:
```
docker model run hf.co/moelanoby/Sensitive-Qwen-0.5B
```

To load this model from the Hugging Face Hub in Python:

# 0. If in Colab and it's a new session, or if model is private, authenticate:
# from huggingface_hub import notebook_login; notebook_login()

# 1. Import necessary libraries:
from transformers import AutoModelForCausalLM, AutoTokenizer
# The following torch imports might be needed if you were to define the classes manually,
# but trust_remote_code=True should handle it by loading them from the Hub.
# import torch
# import torch.nn as nn

# 2. Define your model ID:
MODEL_ID = "moelanoby/Sensitive-Qwen-0.5B"

# 3. Load tokenizer and model (trust_remote_code=True is CRUCIAL):
# This allows Transformers to download and use the Python file ('LLMadd.py')
# from your Hub repository, which contains the definitions for
# `SensitivityModule` and `SensitiveBottleneckLayer`.

try:
    tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, trust_remote_code=True)
    model = AutoModelForCausalLM.from_pretrained(MODEL_ID, trust_remote_code=True, device_map='auto') # Add other params as needed
    print(f'Model {MODEL_ID} loaded successfully!')
except Exception as e:
    print(f'Error loading model: {e}')
    print('Ensure the custom code file (LLMadd.py) in the Hub repo is correct and classes are defined.')

# 4. Example generation (adjust based on your model's chat template, e.g., Qwen2-Instruct):
# prompt = "What is the capital of France?"
# messages = [{"role": "user", "content": prompt}]
# text_input = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
# model_inputs = tokenizer([text_input], return_tensors="pt").to(model.device)
# generated_ids = model.generate(model_inputs.input_ids, max_new_tokens=50)
# result = tokenizer.batch_decode(generated_ids[:, model_inputs.input_ids.shape[-1]:], skip_special_tokens=True)[0]
# print(f'Generated: {result}')

IMPORTANT: trust_remote_code=True allows the execution of Python code from the 'moelanoby/Sensitive-Qwen-0.5B' repository on Hugging Face Hub.

Downloads last month: 4

Safetensors

Model size

0.5B params

Tensor type

BF16