Instructions to use superwhisper/s1-mini with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use superwhisper/s1-mini with Transformers:

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

pipe = pipeline("text-generation", model="superwhisper/s1-mini")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("superwhisper/s1-mini")
model = AutoModelForCausalLM.from_pretrained("superwhisper/s1-mini", 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 superwhisper/s1-mini with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "superwhisper/s1-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": "superwhisper/s1-mini",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/superwhisper/s1-mini

SGLang

How to use superwhisper/s1-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 "superwhisper/s1-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": "superwhisper/s1-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 "superwhisper/s1-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": "superwhisper/s1-mini",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use superwhisper/s1-mini with Docker Model Runner:
```
docker model run hf.co/superwhisper/s1-mini
```

Super cool!

by owao - opened 6 days ago

Discussion

owao

6 days ago

•

edited 6 days ago

Thanks for sharing it! I'm hyped to try it! It will be really useful for my little TTS project, especially for granite-nar! So far I was eventually using LLMs to do the job but way overkill for the task! Was great idea to fine tune a tiny one, especially this old Qwen3-0.6B!
I'm curious, as I imagine the training didn't take so long: did you try other tiny ones? Also, I guess you didn't go above 0.6B for latency, ~~but how well does it infer from context? Does longer audio result in better corrections?~~ Oh I just saw you had a space! I'm gonna try it ;)

softmaximalist

Superwhisper org 3 days ago

We haven't explored many base models as the all the other components are much more important except certain cases. However, we are already using our knowledge and learnings from S1-mini to keep improving and also work on future models.

owao

2 days ago

Wow lol I just realized you were the guys behind superwhisper.com! I don't how I missed that when it was literally the name of your org :D
I didn't have time to try it yet, but I'm working the hype up.

softmaximalist

Superwhisper org 2 days ago

lol, nice to see some users here. Note that we will soon make some updates which should improve the model (especially if you are using the gguf file format) quite noticeably. If you are using the app, there are some in-app improvements as well!

owao

2 days ago

Nah I'm on linux I actually don't use Superwhisper! 😅 I use my own 😎

softmaximalist

Superwhisper org 2 days ago

Oh I see. It's still nice to see the model getting is used in the linux and open source app space!

nub235

1 day ago

Have you guys considered having it to be able to apply custom vocabulary? Like you give it a list of vocabulary terms and maybe like aliases and it can recognize those in the transcript and apply them?

softmaximalist

Superwhisper org about 22 hours ago

We do have that feature in our app but it's not done by S1-mini.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment