Instructions to use Pageshift-Entertainment/pagestorm-research-preview-14b-full-book with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Pageshift-Entertainment/pagestorm-research-preview-14b-full-book with Transformers:

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

pipe = pipeline("text-generation", model="Pageshift-Entertainment/pagestorm-research-preview-14b-full-book")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Pageshift-Entertainment/pagestorm-research-preview-14b-full-book")
model = AutoModelForCausalLM.from_pretrained("Pageshift-Entertainment/pagestorm-research-preview-14b-full-book")
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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Pageshift-Entertainment/pagestorm-research-preview-14b-full-book with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Pageshift-Entertainment/pagestorm-research-preview-14b-full-book"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Pageshift-Entertainment/pagestorm-research-preview-14b-full-book",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Pageshift-Entertainment/pagestorm-research-preview-14b-full-book

SGLang

How to use Pageshift-Entertainment/pagestorm-research-preview-14b-full-book 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 "Pageshift-Entertainment/pagestorm-research-preview-14b-full-book" \
    --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": "Pageshift-Entertainment/pagestorm-research-preview-14b-full-book",
		"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 "Pageshift-Entertainment/pagestorm-research-preview-14b-full-book" \
        --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": "Pageshift-Entertainment/pagestorm-research-preview-14b-full-book",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Pageshift-Entertainment/pagestorm-research-preview-14b-full-book with Docker Model Runner:
```
docker model run hf.co/Pageshift-Entertainment/pagestorm-research-preview-14b-full-book
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

PageStorm Research Preview 14B Full Book

Paper: `2605.17064`

Collection: `PageStorm Research Preview`

Twitter / X: `https://x.com/pageshiftAI`

Job Openings: `https://pageshift.ai/hiring`

This is our main model described in the paper, and it was trained to produce a full novel from a single user prompt. It was initialized from the Mistral3 14B base and was trained for 150B tokens at a sequence length of 256K Tokens on our LongPage dataset. The Training happened on a V6e-256 TPU Pod provided by Google’s TPU Research Cloud. The model is single-turn only and cannot receive additional user prompts after the initial one.

Important Safety Note: This model was trained without safety guidelines in place. It cannot refuse any requests on its own and should be deployed behind a separate safety classifier.

Usage Example

pip install "git+https://github.com/Pageshift-ai/pagestorm.git"

pagestorm generate \
  --repo-id Pageshift-Entertainment/pagestorm-research-preview-14b-full-book \
  --prompt "Thriller in Zurich"

Downloads last month: 7

Safetensors

Model size

14B params

Tensor type

BF16

Model tree for Pageshift-Entertainment/pagestorm-research-preview-14b-full-book

Quantizations

2 models

Dataset used to train Pageshift-Entertainment/pagestorm-research-preview-14b-full-book

Collection including Pageshift-Entertainment/pagestorm-research-preview-14b-full-book

PageStorm Research Preview

Collection

2 items • Updated about 9 hours ago

Paper for Pageshift-Entertainment/pagestorm-research-preview-14b-full-book

Towards Human-Level Book-Writing Capability

Paper • 2605.17064 • Published May 16 • 2