Instructions to use Azrail/smallm_70_instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Azrail/smallm_70_instruct with Transformers:

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

pipe = pipeline("text-generation", model="Azrail/smallm_70_instruct", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("Azrail/smallm_70_instruct", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Azrail/smallm_70_instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Azrail/smallm_70_instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Azrail/smallm_70_instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Azrail/smallm_70_instruct

SGLang

How to use Azrail/smallm_70_instruct 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 "Azrail/smallm_70_instruct" \
    --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": "Azrail/smallm_70_instruct",
		"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 "Azrail/smallm_70_instruct" \
        --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": "Azrail/smallm_70_instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Azrail/smallm_70_instruct with Docker Model Runner:
```
docker model run hf.co/Azrail/smallm_70_instruct
```

Azrail commited on Apr 16, 2025

Commit

4125739

verified ·

1 Parent(s): d9b459c

Update README.md

Browse files

Files changed (1) hide show

README.md +3 -19

README.md CHANGED Viewed

@@ -18,10 +18,11 @@ It has been trained using [TRL](https://github.com/huggingface/trl).
 ## Quick start
 ```python
-from transformers import pipeline
 question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
-generator = pipeline("text-generation", model="Azrail/smallm_70_instruct", device="cuda")
 output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
 print(output["generated_text"])
 ```
@@ -40,20 +41,3 @@ This model was trained with SFT.
 - Pytorch: 2.6.0+cu126
 - Datasets: 3.5.0
 - Tokenizers: 0.21.1
-## Citations
-Cite TRL as:
-```bibtex
-@misc{vonwerra2022trl,
-	title        = {{TRL: Transformer Reinforcement Learning}},
-	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
-	year         = 2020,
-	journal      = {GitHub repository},
-	publisher    = {GitHub},
-	howpublished = {\url{https://github.com/huggingface/trl}}
-}
-```

 ## Quick start
 ```python
+from transformers import pipeline, AutoTokenizer
 question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
+tokenizer = AutoTokenizer.from_pretrained("Azrail/smallm_70_instruct")
+generator = pipeline("text-generation", model="Azrail/smallm_70_instruct", device="cuda", trust_remote_code=True, tokenizer=tokenizer)
 output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
 print(output["generated_text"])
 ```
 - Pytorch: 2.6.0+cu126
 - Datasets: 3.5.0
 - Tokenizers: 0.21.1