Comments (3)
Hi,
This is not implemented by the framework. The usual process is straight forward: Load the pretrained weights for all but the last DNN layers. Then, do your normal training on your fine-tune data.
from deepfilternet.
This issue is stale because it has been open 90 days with no activity. Remove stale label or comment or this will be closed in 7 days.
from deepfilternet.
Hi,
This is not implemented by the framework. The usual process is straight forward: Load the pretrained weights for all but the last DNN layers. Then, do your normal training on your fine-tune data.
Hey @Rikorose I'd love to finetune on the existing pretrained weights as well. Could you provide a more detailed description or examples on how to go about doing the "load pretrained weights all but for last DNN layers" step?
The original model is excellent so would love to be able to use that, I just want to experiment with my own noise types to see if it can recover some lost aspects.
No worries if too lengthy/much of a hassle though.
from deepfilternet.
Related Issues (20)
- How to use the pretrain-model?
- audio super resolution
- Clarification on Training Configuration for DeepFilterNet3 (dataset.cfg) HOT 1
- Cannot use DeepFilterNet2 model (CLI)
- Pre-built binaries of `libdeepfilter` (C API)
- Can't reproduce DeepFilterNet2 and DeepFilterNet3 results on Voicebank Demand test set HOT 6
- Unable to use libDF due to debug checks failing: duplicate name /convt3/Conv.bias
- Error(s) in loading state_dict for DfNet: Unexpected key(s) in state_dict: "erb_comp.c", "erb_comp.mn".
- Out-of-memory when running deepFilter on long audio files
- Using ladspa with mono input produces an error
- Pipewire filter-chain source support in Windows
- Fine Tuning DeepFilterNet
- Installation problem
- requirement files installation problem
- GPU Memory torch.cuda.OutOfMemoryError
- [BUG] Only works with sample rate == 48000 HOT 3
- Multiple GPU training via DistributedDataParallel
- How to quantify the model
- 16k model training configuration and 16k pre-trained model
- RuntimeError: Couldn't find appropriate backend to handle uri <noisy_audio_file> and format None.
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 deepfilternet.