Instructions to use BAAI/Infinity-Instruct-3M-0625-Qwen2-7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use BAAI/Infinity-Instruct-3M-0625-Qwen2-7B with Transformers:

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

pipe = pipeline("text-generation", model="BAAI/Infinity-Instruct-3M-0625-Qwen2-7B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("BAAI/Infinity-Instruct-3M-0625-Qwen2-7B")
model = AutoModelForCausalLM.from_pretrained("BAAI/Infinity-Instruct-3M-0625-Qwen2-7B", 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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use BAAI/Infinity-Instruct-3M-0625-Qwen2-7B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "BAAI/Infinity-Instruct-3M-0625-Qwen2-7B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "BAAI/Infinity-Instruct-3M-0625-Qwen2-7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/BAAI/Infinity-Instruct-3M-0625-Qwen2-7B

SGLang

How to use BAAI/Infinity-Instruct-3M-0625-Qwen2-7B 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 "BAAI/Infinity-Instruct-3M-0625-Qwen2-7B" \
    --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": "BAAI/Infinity-Instruct-3M-0625-Qwen2-7B",
		"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 "BAAI/Infinity-Instruct-3M-0625-Qwen2-7B" \
        --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": "BAAI/Infinity-Instruct-3M-0625-Qwen2-7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use BAAI/Infinity-Instruct-3M-0625-Qwen2-7B with Docker Model Runner:
```
docker model run hf.co/BAAI/Infinity-Instruct-3M-0625-Qwen2-7B
```

Does two stage training use same hyperparamers?

by bbruceyuan - opened Aug 25, 2024

Discussion

bbruceyuan

Aug 25, 2024

In model card. There is a description:

First, we apply the foundational dataset Infinity-Instruct-3M to improve the foundational ability (math & code) of Qwen2-7B, and get the foundational instruct model Infinity-Instruct-3M-Qwen2-7B. Then we finetune the Infinity-Instruct-3M-Qwen2-7B to get the stronger chat model Infinity-Instruct-3M-0625-Qwen2-7B. Here is the training hyperparamers.

Question: there are two stages and only one group training hyperparamers. so does both two stage SFT training use same hyperparamers?

hyxmmm

Beijing Academy of Artificial Intelligence org Sep 3, 2024

Yes, you can use the same set of hyperparameters for the two stages training.

liweipe

Sep 10, 2024

Hello, which template do you use to fine tune from pretrained model to Foundational Instruct model? I assume you used the chat template to finetune the final chat model but what about the intermediate stage. Also use chat template with system prompt?

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