tinker-skills

Tinker Cookbook

We provide two libraries for the broader community to customize their language models: tinker and tinker-cookbook.

• tinker is a training SDK for researchers and developers to fine-tune language models. You send API requests to us and we handle the complexities of distributed training.
• tinker-cookbook includes realistic examples of fine-tuning language models. It builds on the Tinker API and provides common abstractions to fine-tune language models.

Installation

1. Sign up for Tinker here.
2. Once you have access, create an API key from the console and export it as environment variable TINKER_API_KEY.
3. Install tinker-cookbook (includes the tinker SDK as a dependency):

   # Latest stable release from PyPI
   uv pip install tinker-cookbook
   
   # Or install the nightly build
   uv pip install 'tinker-cookbook @ git+https://github.com/thinking-machines-lab/tinker-cookbook.git@nightly'
   

Tinker

Refer to the docs to start from basics. Here we introduce a few Tinker primitives - the basic components to fine-tune LLMs:

import tinker
service_client = tinker.ServiceClient()
training_client = service_client.create_lora_training_client(
  base_model="meta-llama/Llama-3.2-1B", rank=32,
)
training_client.forward_backward(...)
training_client.optim_step(...)
training_client.save_state(...)
training_client.load_state(...)

sampling_client = training_client.save_weights_and_get_sampling_client()
sampling_client.sample(...)

See tinker_cookbook/recipes/sl_loop.py and tinker_cookbook/recipes/rl_loop.py for minimal examples of using these primitives to fine-tune LLMs.

Tutorials

New to Tinker? The tutorials/ directory has progressive marimo notebooks that guide you from your first API call to building custom RL training pipelines:

#	Notebook	What you'll learn
101	Hello Tinker	Architecture overview, client hierarchy, sampling
102	First SFT	Renderers, datum construction, training loop, Kimi K2.5 scaling demo
103	Efficient Sampling	Concurrent futures, num_samples, batch evaluation throughput
104	First RL	GRPO on GSM8K: rewards, advantages, degenerate groups
301	Cookbook RL Abstractions	Env, EnvGroupBuilder, RLDataset, ProblemEnv
302	Custom RL Environment	Build your own ProblemEnv and RLDataset

Run any tutorial with marimo edit tutorials/101_hello_tinker.py. Rendered versions are available on the Tinker docs site.

To download the weights of any model:

rest_client = service_client.create_rest_client()
future = rest_client.get_checkpoint_archive_url_from_tinker_path(sampling_client.model_path)
with open(f"model-checkpoint.tar.gz", "wb") as f:
    f.write(future.result())

Tinker Cookbook

Besides these primitives, we also offer Tinker Cookbook (a.k.a. this repo), a library of a wide range of abstractions to help you customize training environments. tinker_cookbook/recipes/sl_basic.py and tinker_cookbook/recipes/rl_basic.py contain minimal examples to configure supervised learning and reinforcement learning.