Back to Latimer App— NEW MODEL LAUNCHED
LATIMER x INTEL
X
Latimer built a customized Small Language Model (“SLM”) with Intel using Microsoft’s Phi 3.5.
— BENEFIT
The model can operate locally and is downloaded to your browser.
— BENEFIT
It can operate with or without access to the Internet.
— BENEFIT
The model is very fast as it generally does not have to communicate with the cloud to generate a response.
— BENEFIT
It has access to the WEB or to Documents you upload if you find these features beneficial.
We will continue to develop "local" model functionality as it can have important consumer advantages for privacy, security, speed and resiliency.
— For Chrome USERS
CHROME USERS NEED TO ADJUST SETTINGS TO USE THIS LOCAL MODEL
Chrome default settings will not allow you to use this model. Once you download the model:
— Navigate to: chrome://flags/
— Search for: WebNN
— Change these settings to "ENABLE"
— Relaunch Chrome
— PERFORMANCE
SPECS RECOMMENDATION
For the best AI performance, we recommend:
Intel® Core™ Ultra Processors (Series 2) with at least 40 TOPS for the Neural Processing Unit (NPU)
Intel® Core™ Ultra 7 Processor 258V or higher
— Memory Size 32 GB
— Intel® Arc™ 140V GPU
— Intel® AI Boost 40 TOPS
— Memory Size 32 GB
— Intel® Arc™ 140V GPU
— Intel® AI Boost 40 TOPS