TÜLU 3: Pushing Frontiers in Open Language Model Post-Training
Paper • 2411.15124 • Published • 68
llm.create_chat_completion(
messages = [
{
"role": "user",
"content": "What is the capital of France?"
}
]
)Overview
Tülu3 is a state-of-the-art instruction-following model family developed by Allen Institute for AI. It is designed to excel in a wide range of tasks beyond standard chat applications, including complex problem-solving in domains such as MATH, GSM8K, and IFEval. The Tülu3 series provides a fully open-source ecosystem, offering access to datasets, training code, and fine-tuning recipes to facilitate advanced model customization and experimentation.
Variants
| No | Variant | Cortex CLI command |
|---|---|---|
| 1 | Tulu3-8b | cortex run tulu3:8b |
Use it with Jan (UI)
- Install Jan using Quickstart
- Use in Jan model Hub:
cortexhub/tulu3
Use it with Cortex (CLI)
- Install Cortex using Quickstart
- Run the model with command:
cortex run tulu3
Credits
- Downloads last month
- 65
Hardware compatibility
Log In to add your hardware
2-bit
3-bit
4-bit
5-bit
6-bit
8-bit
# !pip install llama-cpp-python from llama_cpp import Llama llm = Llama.from_pretrained( repo_id="cortexso/tulu3", filename="", )