Evaluation scripts of Chainer for deepmark
python evaluate/train_image.py
python evaluate/train_movie.py
python evaluate/train_audio.py
python evaluate/train_text.py
| Category | Arcitecture | Status | Comment |
|---|---|---|---|
| Image | InceptionV3-batchnorm | Y | |
| Alexnet-OWT | Y | ||
| VGG-D | Y | ||
| ResNet50 | Y | ||
| Movie | C3D | Need fix | Issue #5, waiting for MaxPoolingND (Chainer issue #1353) |
| Audio | DeepSpeech2 | Need fix | PR #20 |
| MSR 5FC layer | Teporary implementation | ||
| Text | BigLSTM | Working | Issue #2, PR #11 |
| SmallLSTM | Working | Issue #19 |
| Category | Single GPU | Multi GPU |
|---|---|---|
| Image | Y | N |
| Movie | Need fix | N |
| Audio | Need fix | N |
| Text | Need fix | N |
MIT (See LICENSE file)