Instructions to use HuggingFaceFW/ablation-model-the-pile with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use HuggingFaceFW/ablation-model-the-pile with Transformers:

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

pipe = pipeline("text-generation", model="HuggingFaceFW/ablation-model-the-pile")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("HuggingFaceFW/ablation-model-the-pile")
model = AutoModelForCausalLM.from_pretrained("HuggingFaceFW/ablation-model-the-pile")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use HuggingFaceFW/ablation-model-the-pile with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/HuggingFaceFW/ablation-model-the-pile

SGLang

How to use HuggingFaceFW/ablation-model-the-pile 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 "HuggingFaceFW/ablation-model-the-pile" \
    --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": "HuggingFaceFW/ablation-model-the-pile",
		"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 "HuggingFaceFW/ablation-model-the-pile" \
        --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": "HuggingFaceFW/ablation-model-the-pile",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use HuggingFaceFW/ablation-model-the-pile with Docker Model Runner:
```
docker model run hf.co/HuggingFaceFW/ablation-model-the-pile
```

Intermediate checkpoints

by przvl - opened Apr 26, 2024

Discussion

przvl

Apr 26, 2024

Thanks for your work 🤗
Could you also provide access to intermediate checkpoints of all ablation models to be able to compare their perfomance at different training steps?

guipenedo

Apr 26, 2024

Hi! these are already available, you can check the different branches of this repo, each has a different checkpoint. You can load a given checkpoint by providing revision=(branch name)

przvl

Apr 26, 2024

Perfect, thanks!

przvl changed discussion status to closed Apr 26, 2024

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