Instructions to use Blackroot/TensorProduct-Microllama with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Blackroot/TensorProduct-Microllama with Transformers:

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

pipe = pipeline("text-generation", model="Blackroot/TensorProduct-Microllama")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("Blackroot/TensorProduct-Microllama", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Blackroot/TensorProduct-Microllama with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Blackroot/TensorProduct-Microllama"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Blackroot/TensorProduct-Microllama",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Blackroot/TensorProduct-Microllama

SGLang

How to use Blackroot/TensorProduct-Microllama 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 "Blackroot/TensorProduct-Microllama" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Blackroot/TensorProduct-Microllama",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "Blackroot/TensorProduct-Microllama" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Blackroot/TensorProduct-Microllama",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Blackroot/TensorProduct-Microllama with Docker Model Runner:
```
docker model run hf.co/Blackroot/TensorProduct-Microllama
```

From scratch pretraining on english only no synthetic data, no code, 3 epochs of 1 gig of data for the ~125M param model.

Test network using Tensor Product Attention. Other than some alterations to the attention, such as 16 heads insted of 9 and using TPA, this is the same setup as https://huggingface.co/HuggingFaceTB/SmolLM2-135M-Instruct

Scripts:

inference.py to run the model with some test prompts
test_train.py runs with the exact configurations used to train this model and is the reproduction script. Data is assumed to be in JSONL format with "text":"example text", "text":"..."

Notes:

One of the primary reported benefits for TPA are for inference which are not really being leveraged at all, although you can probably fit a larger bsz than traditional MHA/GQA with this. This did save about 5% on params, that amount should scale much more as the network size increases. The run time is very similar to MHA/GQA at this scale.

Training Metrics

Dataset Information

Training data per epoch: 1 GB
Total tokens trained: 48,261,120
No sythetic data

Training Results

Final Train Loss: 3.0421
Final Train Perplexity: 20.95

Code

The code for tensor product attn is available at: https://github.com/tensorgi/T6.

Downloads last month: -; Downloads are not tracked for this model. How to track

Paper for Blackroot/TensorProduct-Microllama

Tensor Product Attention Is All You Need

Paper • 2501.06425 • Published Jan 11, 2025 • 91