zhangshuai/mlx-swift-examples
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add Llama 3.1 (#98)

Browse Source 
* Update Mistral 7B config

* Add Mistral NeMo

* Update for Llama 3.1

* Align LlamaConfiguration with Python implementation

* Fix model configuration names

* Refine DynamicNTKScalingRoPE

* compute base only once

---------

Co-authored-by: Awni Hannun <awni@apple.com>
This commit is contained in:
Anthony
2024-07-26 22:05:42 +02:00
committed by GitHub
parent c4fda0e036
commit ac6bdfccec
3 changed files with 200 additions and 84 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
Applications/LLMEval/ContentView.swift
View File

@@ -159,7 +159,7 @@ class LLMEvaluator {
/// this controls which model loads -- phi4bit is one of the smaller ones so this will fit on
/// more devices
let modelConfiguration = ModelConfiguration.phi34bit
let modelConfiguration = ModelConfiguration.phi3_4bit
/// parameters controlling the output
let generateParameters = GenerateParameters(temperature: 0.6)