Llama3.1 8B TI Instruct
Llama TI is an improved Llama (from Meta AI), some aspects of the model have been revised and some features have been added.
Info
Main
The model is based on Meta-Llama-3.1-8B-Instruct, and has the same 8.03B parameters. The Llama3 architecture (LlamaForCausalLM) has been preserved and the model launch methods are the same.
Differences
Thanks to additional training and advanced merging, it was possible to improve mathematical, biological, reasoning and writing skills.
Now the model can:
- Count well and solve mathematical/physical problems
- Reason/think logically
- Write creatively (in many languages)
- Code well
- Process/analyze large texts
Where is the base version?
It is available here!
Special thanks to:
Meta AI, NVIDIA, Arcee AI, SkyWork, NousReaserch, Unsloth and Project Fluently.
Developed and uploaded by ehristoforu.
- Downloads last month
- 100
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.