Instructions to use MiniLLM/MiniPLM-Qwen-200M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MiniLLM/MiniPLM-Qwen-200M with Transformers:

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

pipe = pipeline("text-generation", model="MiniLLM/MiniPLM-Qwen-200M")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MiniLLM/MiniPLM-Qwen-200M")
model = AutoModelForCausalLM.from_pretrained("MiniLLM/MiniPLM-Qwen-200M")
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

vLLM

How to use MiniLLM/MiniPLM-Qwen-200M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MiniLLM/MiniPLM-Qwen-200M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MiniLLM/MiniPLM-Qwen-200M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/MiniLLM/MiniPLM-Qwen-200M

SGLang

How to use MiniLLM/MiniPLM-Qwen-200M 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 "MiniLLM/MiniPLM-Qwen-200M" \
    --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": "MiniLLM/MiniPLM-Qwen-200M",
		"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 "MiniLLM/MiniPLM-Qwen-200M" \
        --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": "MiniLLM/MiniPLM-Qwen-200M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use MiniLLM/MiniPLM-Qwen-200M with Docker Model Runner:
```
docker model run hf.co/MiniLLM/MiniPLM-Qwen-200M
```

t1101675 commited on Oct 20, 2024

Commit

543ee2a

verified ·

1 Parent(s): bffed11

Update README.md

Browse files

Files changed (1) hide show

README.md +7 -2

README.md CHANGED Viewed

@@ -3,6 +3,7 @@ library_name: transformers
 license: apache-2.0
 datasets:
 - monology/pile-uncopyrighted
 language:
 - en
 metrics:
@@ -16,6 +17,8 @@ pipeline_tag: text-generation
 **MiniPLM-QWen-200M** is a 200M model with QWen achitecture pre-trained from scratch on [the Pile](https://huggingface.co/datasets/monology/pile-uncopyrighted) using the MiniPLM knowledge distillation framework with the [offcial QWen1.5-1.8B](https://huggingface.co/Qwen/Qwen1.5-1.8B) as the teacher model.
 <p align='left'>
     <img src="https://cdn-uploads.huggingface.co/production/uploads/624ac662102fcdff87be51b9/2BqT0NgkmIXYlktovw9kG.png" width="1000">
 </p>
@@ -29,7 +32,9 @@ MiniPLM models achieves better performance given the same computation and scales
 </p>
 ## Baseline Models
-+ [Conventional Pre-Training](https://huggingface.co/MiniLLM/Pretrain-Qwen-1.2B)
-+ [VanillaKD](https://huggingface.co/MiniLLM/VanillaKD-Pretrain-Qwen-1.2B)
 ## Citation

 license: apache-2.0
 datasets:
 - monology/pile-uncopyrighted
+- MiniLLM/pile-diff_samp-qwen_1.8B-qwen_104M-r0.5
 language:
 - en
 metrics:
 **MiniPLM-QWen-200M** is a 200M model with QWen achitecture pre-trained from scratch on [the Pile](https://huggingface.co/datasets/monology/pile-uncopyrighted) using the MiniPLM knowledge distillation framework with the [offcial QWen1.5-1.8B](https://huggingface.co/Qwen/Qwen1.5-1.8B) as the teacher model.
+We also open-source the [pre-training corpus](https://huggingface.co/datasets/MiniLLM/pile-diff_samp-qwen_1.8B-qwen_104M-r0.5) refined by Difference Sampling in MiniPLM for reproducibility.
 <p align='left'>
     <img src="https://cdn-uploads.huggingface.co/production/uploads/624ac662102fcdff87be51b9/2BqT0NgkmIXYlktovw9kG.png" width="1000">
 </p>
 </p>
 ## Baseline Models
++ [Conventional Pre-Training](https://huggingface.co/MiniLLM/Pretrain-Qwen-200M)
++ [VanillaKD](https://huggingface.co/MiniLLM/VanillaKD-Pretrain-Qwen-200M)
 ## Citation
+TODO