Instructions to use 5ivatej/tinyllama-1.1b-early-exit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use 5ivatej/tinyllama-1.1b-early-exit with Transformers:

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

pipe = pipeline("text-generation", model="5ivatej/tinyllama-1.1b-early-exit")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("5ivatej/tinyllama-1.1b-early-exit", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use 5ivatej/tinyllama-1.1b-early-exit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "5ivatej/tinyllama-1.1b-early-exit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "5ivatej/tinyllama-1.1b-early-exit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/5ivatej/tinyllama-1.1b-early-exit

SGLang

How to use 5ivatej/tinyllama-1.1b-early-exit 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 "5ivatej/tinyllama-1.1b-early-exit" \
    --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": "5ivatej/tinyllama-1.1b-early-exit",
		"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 "5ivatej/tinyllama-1.1b-early-exit" \
        --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": "5ivatej/tinyllama-1.1b-early-exit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use 5ivatej/tinyllama-1.1b-early-exit with Docker Model Runner:
```
docker model run hf.co/5ivatej/tinyllama-1.1b-early-exit
```

TinyLlama Early-Exit Heads (Adapter)

Adapter-style early-exit classification heads for TinyLlama/TinyLlama-1.1B-Chat-v1.0.

Attach these heads to the base model to stop decoding early when a token’s next-token probability is already confident, reducing average compute per token.

💡 This repo contains only the early-exit heads + a small loader. You must also have the base model (it downloads automatically in the examples below).

TL;DR

One tiny linear head per transformer layer.
At each decoding step, compute layer-wise logits; if max_prob >= confidence_threshold, exit early.
Ships with a loader and a minimal generation helper.

Quickstart (no local files needed)

# pip install -U torch transformers safetensors huggingface_hub

from huggingface_hub import hf_hub_download
import importlib.util, sys

REPO_ID = "5ivatej/tinyllama-1.1b-early-exit"

# 1) Dynamically fetch the loader from the Hub
module_path = hf_hub_download(REPO_ID, "early_exit_wrapper.py")

# 2) Import it as a module
spec = importlib.util.spec_from_file_location("early_exit_wrapper", module_path)
early = importlib.util.module_from_spec(spec); sys.modules["early_exit_wrapper"] = early
spec.loader.exec_module(early)

# 3) Load wrapped model + tokenizer
wrapped, tok = early.load_early_exit_from_hub(REPO_ID)  # auto-picks CPU/MPS/CUDA & safe dtype

# 4) Generate with early exit
ids = early.generate_with_early_exit(
    "Explain early exit in one tweet.",
    wrapped, tok,
    max_new_tokens=64, temperature=0.7, top_p=0.9
)
print(tok.decode(ids[0], skip_special_tokens=True))

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

Model tree for 5ivatej/tinyllama-1.1b-early-exit

Base model

TinyLlama/TinyLlama-1.1B-Chat-v1.0

Finetuned

(586)

this model