Instructions to use tencent/Hy-Memory-30B-A3B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tencent/Hy-Memory-30B-A3B with Transformers:

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

pipe = pipeline("text-generation", model="tencent/Hy-Memory-30B-A3B")
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("tencent/Hy-Memory-30B-A3B")
model = AutoModelForCausalLM.from_pretrained("tencent/Hy-Memory-30B-A3B", 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 tencent/Hy-Memory-30B-A3B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tencent/Hy-Memory-30B-A3B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tencent/Hy-Memory-30B-A3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/tencent/Hy-Memory-30B-A3B

SGLang

How to use tencent/Hy-Memory-30B-A3B 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 "tencent/Hy-Memory-30B-A3B" \
    --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": "tencent/Hy-Memory-30B-A3B",
		"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 "tencent/Hy-Memory-30B-A3B" \
        --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": "tencent/Hy-Memory-30B-A3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use tencent/Hy-Memory-30B-A3B with Docker Model Runner:
```
docker model run hf.co/tencent/Hy-Memory-30B-A3B
```

Actual Context Length and RoPE Scaling to 1M Tokens

by rikunarita-3 - opened 9 days ago

Discussion

rikunarita-3

9 days ago

Is the 256K context length specified in config.json correct in practice?

"max_position_embeddings": 262144,

Also, is it possible to extend the context length to 1M tokens using RoPE scaling?

seventhsummer

Tencent org 7 days ago

256k is the base model context length.
so, the config is correct.

we trained in 8k token chunk, we advice not too long tokens in, will result better performace.

rikunarita-3

6 days ago

Thanks for the clarification! One last question: could Resonance RoPE / Resonance YaRN be used to extend the context length to 1M tokens without fine-tuning, even if performance may degrade?

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