Instructions to use TheBlokeAI/JackFram-Llama-68M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TheBlokeAI/JackFram-Llama-68M with Transformers:

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

pipe = pipeline("text-generation", model="TheBlokeAI/JackFram-Llama-68M")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("TheBlokeAI/JackFram-Llama-68M")
model = AutoModelForCausalLM.from_pretrained("TheBlokeAI/JackFram-Llama-68M", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use TheBlokeAI/JackFram-Llama-68M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "TheBlokeAI/JackFram-Llama-68M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TheBlokeAI/JackFram-Llama-68M",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/TheBlokeAI/JackFram-Llama-68M

SGLang

How to use TheBlokeAI/JackFram-Llama-68M 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 "TheBlokeAI/JackFram-Llama-68M" \
    --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": "TheBlokeAI/JackFram-Llama-68M",
		"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 "TheBlokeAI/JackFram-Llama-68M" \
        --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": "TheBlokeAI/JackFram-Llama-68M",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use TheBlokeAI/JackFram-Llama-68M with Docker Model Runner:
```
docker model run hf.co/TheBlokeAI/JackFram-Llama-68M
```

This is JackFram's Llama 68M, with the following changes:

Removed unnecessary files from training (optimizer.pt, scheduler.pt, rng_state.pth, train*)
Converted to safetensors.

This lowers the size of the repo to 260MiB.

All credit for the creation of this mini Llama goes to the original author(s).

This tiny Llama model can be used to test quantization and other processes, as its a fully compatible Llama 1 model, but with only two layers.

Downloads last month: 11

Safetensors

Model size

68M params

Tensor type

F32