Comments (1)
It's feasible for the right model size. There are two places you could look as starting points:
- The Transformer LM example. That does training for a Transformer model
- The LoRA example. In this example you could remove the call to
model.freeze
and the lora layers and that should give you full model training.
Watch out for lower precision though. Without other changes you'd likely want to use float32 for full model training.
from mlx-examples.
Related Issues (20)
- Please share the sample of chat format for LoRA Fine-tuning HOT 4
- [Feature Request] A better Finetune method
- Errors while generating HOT 2
- DBRX Instruct prompt templating issue HOT 5
- Add support for llava-hf/llava-v1.6-mistral-7b-hf HOT 2
- Regarding with #625 HOT 5
- Segmentation fault during training HOT 22
- Unable to load mlx-community/dbrx-instruct-4bit locally for use in QLoRA HOT 6
- Server produce gibberish output for some model HOT 9
- Jamba Model Conversion to MLX
- Failed to run mlx-community/c4ai-command-r-plus-4bit HOT 15
- Enso error when running llama-2- w/70B parameters HOT 9
- Will it support chatglm series?
- CVAE Example is 3x the speed of the M1 Max on an M3 Max (with MLX 0.9.1) HOT 1
- interesting new finetuning approach from stanford - ReFT
- Trouble with Triplet Loss Implementation in MLX HOT 2
- Command-R-Plus, Context Window Limitations HOT 42
- Can we integrate schedule free learning rate over epochs from Meta research into MLX fine tuning? HOT 1
- Failed to convert gemma model. HOT 1
- libc++abi: terminating due to uncaught exception of type std::runtime_error HOT 20
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from mlx-examples.