Comments (6)
Hi,
this is most probably a problem with the number of labels. Does your dataset use a different number of labels/characters? Please check if this is correct in the config.
from returnn.
Thanks..That worked. Also I want to ask if we can train only a few layers from the original model(transfer learning)?
from returnn.
There is a parameter which you can add in the layer definition in json, I think you have to set "trainable": false for the layers which should not be updated.
Please have a try, and maybe check the code for some details.
If it doesn't work, feel free to post again here.
from returnn.
Hi, Thank you for the reply. I dont understand following parameters that are used in config_real:
- "max_seqs": 10,
- "nadam": 1,
- "reinit": 0,
- "log_verbosity": 5,
Further, If training is interrupted at some step, how can i continue from the previous saved model? can i do it with task = forward or else?
from returnn.
Hi,
max_seqs is an upper bound on the number of sequences (i.e. images in your case), which are processed together in one mini-batch
nadam: 1 means, that Adam with Nesterov momentum is used as the optimizer
reinit: 0 there I'm not sure what exactly it does, better just leave it like this
log_verbosity: 5 controls how many messages are printed to the log. 5 is the highest verbosity, so everything will be printed. You can reduce it if you want less output
I think the default behaviour should be to continue the training if a previous model is available
from returnn.
Thank you very much for the prompt reply.
from returnn.
Related Issues (20)
- PyTorch collect model statistics
- PyTorch recover after CUDA OOM with restart does not work with CUDA HOT 3
- PyTorch distributed training CPU OOM with sync_on_cpu HOT 1
- Support `torch.compile` for RF
- RF backend: PyTorch code
- Different effective learning rate reported over gpus HOT 11
- CUDA error: initialization error HOT 3
- MultiProcDataset inside PyTorch DataLoader with num_workers>0, multiple issues HOT 4
- RuntimeError: CUDA error: unspecified launch failure HOT 2
- NonDaemonicSpawnProcess hangs at exit HOT 2
- High memory usage with datasets (specifically when multi procs are used)
- Hang at exit in TDL worker in multiprocessing `_run_finalizers`, deadlock in `_wait_for_tstate_lock`? HOT 6
- Hang HOT 2
- Returnn Native after using different apptainer uses old compilation HOT 6
- MetaDataset with sequence list filter file
- HDFDataset (or generic dataset) post processing HOT 15
- Dataset batching like ESPnet support
- torch.nn.functional.conv2d: RuntimeError: GET was unable to find an engine to execute this computation HOT 1
- TensorFlow 2.14 degradation in WER HOT 2
- Updates for recent TensorFlow version
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 returnn.