sanjuhs/nn-training-final-weights
By sanjuhs · ↓ 0 · ♥ 0 · Updated 2026-03-22
What's in this repo
This page lists the files in sanjuhs/nn-training-final-weights, including baseline_d192_l6.onnx, tcn_model.onnx, overnight_20260321_205258__conv_d320_l12_b16.onnx and more.
- 4 ONNX models (.onnx files)
- 11 files, including 4 .onnx, 4 .pth, 1 .md, 1 .json
- Last updated 2026-03-22
ONNX files
| File | Size | Download |
|---|---|---|
| baseline_d192_l6/baseline_d192_l6.onnx | — | Download |
| legacy_tcn_model/tcn_model.onnx | — | Download |
| overnight_conv_d320_l12_b16/overnight_20260321_205258__conv_d320_l12_b16.onnx | — | Download |
| overnight_conv_d320_l12_b20_nadam_huber/overnight_20260322_053753__conv_d320_l12_b20_nadam_huber.onnx | — | Download |
How to use
These ONNX files can be loaded with ONNX Runtime. To inspect a model graph, upload the file to the Netron viewer. Download links point to the original files on Hugging Face.
README
NN Training Final Weights
This model repository stores the retained final weights exported from the local NN_training project after cleanup.
Included families
overnight_conv_d320_l12_b20_nadam_huber: best overnight full transformer checkpoint by recorded sweep scoreovernight_conv_d320_l12_b16: strongest smile / expressiveness transformer checkpointbaseline_d192_l6: clean full-pipeline baseline transformer checkpointlegacy_tcn_model: legacy TCN reference checkpoint kept for comparison
Notes
- The transformer family outperformed the retained legacy TCN on the more meaningful mouth/smile/jaw metrics used in the local comparison.
manifest.jsoncontains the source paths, file sizes, and the comparison metrics used when selecting these artifacts.- This repository was created from a cleanup pass intended to keep GitHub pushes small while preserving the final model artifacts in Hugging Face.
Files (11)
.gitattributes
README.md
baseline_d192_l6/baseline_d192_l6.onnx
ONNX
baseline_d192_l6/baseline_d192_l6_best.pth
legacy_tcn_model/best_tcn_model.pth
legacy_tcn_model/tcn_model.onnx
ONNX
manifest.json
overnight_conv_d320_l12_b16/overnight_20260321_205258__conv_d320_l12_b16.onnx
ONNX
overnight_conv_d320_l12_b16/overnight_20260321_205258__conv_d320_l12_b16_best.pth
overnight_conv_d320_l12_b20_nadam_huber/overnight_20260322_053753__conv_d320_l12_b20_nadam_huber.onnx
ONNX
overnight_conv_d320_l12_b20_nadam_huber/overnight_20260322_053753__conv_d320_l12_b20_nadam_huber_best.pth
Related models
More from sanjuhs
No other models by this author.
More models
No task type declared.