Instructions to use devon7y/Toronto-Mans-4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use devon7y/Toronto-Mans-4B with Transformers:

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

pipe = pipeline("text-generation", model="devon7y/Toronto-Mans-4B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("devon7y/Toronto-Mans-4B")
model = AutoModelForCausalLM.from_pretrained("devon7y/Toronto-Mans-4B", 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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use devon7y/Toronto-Mans-4B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "devon7y/Toronto-Mans-4B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "devon7y/Toronto-Mans-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/devon7y/Toronto-Mans-4B

SGLang

How to use devon7y/Toronto-Mans-4B 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 "devon7y/Toronto-Mans-4B" \
    --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": "devon7y/Toronto-Mans-4B",
		"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 "devon7y/Toronto-Mans-4B" \
        --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": "devon7y/Toronto-Mans-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use devon7y/Toronto-Mans-4B with Docker Model Runner:
```
docker model run hf.co/devon7y/Toronto-Mans-4B
```

devon7y commited on Sep 17

Commit

ae5deeb

verified ·

1 Parent(s): 3d7b1e5

Card: try-it-now banner linking the ZeroGPU Spaces

Browse files

Files changed (1) hide show

README.md +7 -0

README.md CHANGED Viewed

@@ -20,6 +20,13 @@ datasets:
 # Toronto-Mans-4B
 > [!Note]
 > Wagwan. Dis is the likkle one: Qwen3.5-4B with the same Toronto fine-tune folded in (LoRA merged, no adapter, allow it), 8 GB in bf16, runs on a laptop GPU or a free ZeroGPU slice. Same voice as the big man [Toronto-Mans-35B-A3B](https://huggingface.co/devon7y/Toronto-Mans-35B-A3B), a likkle less polish, bare faster.
 >

 # Toronto-Mans-4B
+> ## 🚋 TRY MANS RIGHT NOW, NO DOWNLOAD, NO SETUP
+> **[▶ Chat with Toronto Mans 35B](https://huggingface.co/spaces/devon7y/toronto-mans-35b)** (the real ting, takes a likkle minute to wake up) · **[▶ Chat with Toronto Mans 4B](https://huggingface.co/spaces/devon7y/toronto-mans-4b)** (replies in seconds) · 9B coming, styll
+>
+> [![Try 35B](https://img.shields.io/badge/%F0%9F%9A%8B%20Try%20it-35B%20on%20ZeroGPU-c8102e?style=for-the-badge)](https://huggingface.co/spaces/devon7y/toronto-mans-35b) [![Try 4B](https://img.shields.io/badge/%F0%9F%9A%8B%20Try%20it-4B%20on%20ZeroGPU-c8102e?style=for-the-badge)](https://huggingface.co/spaces/devon7y/toronto-mans-4b)
+>
+> Free to use, runs on Hugging Face ZeroGPU, say anything and pree the reply. Wallahi it's live.
 > [!Note]
 > Wagwan. Dis is the likkle one: Qwen3.5-4B with the same Toronto fine-tune folded in (LoRA merged, no adapter, allow it), 8 GB in bf16, runs on a laptop GPU or a free ZeroGPU slice. Same voice as the big man [Toronto-Mans-35B-A3B](https://huggingface.co/devon7y/Toronto-Mans-35B-A3B), a likkle less polish, bare faster.
 >