Jacobellis Dan (dgj335)
lightweight encoder
569067d
-
README_files
allow 1, 3, or 4 channels for entropy coding
-
3.06 kB
example usage
-
2.58 MB
allow 1, 3, or 4 channels for entropy coding
-
5.23 kB
allow 1, 3, or 4 channels for entropy coding
v0.0.1.pth
Detected Pickle imports (7)
- "_codecs.encode",
- "__main__.Args",
- "numpy.dtype",
- "numpy.core.multiarray.scalar",
- "collections.OrderedDict",
- "torch.FloatStorage",
- "torch._utils._rebuild_tensor_v2"
How to fix it?
339 MB
sd3 autoencoder without tft
v0.1.1.pth
Detected Pickle imports (7)
- "collections.OrderedDict",
- "__main__.Args",
- "torch.FloatStorage",
- "numpy.dtype",
- "torch._utils._rebuild_tensor_v2",
- "_codecs.encode",
- "numpy.core.multiarray.scalar"
How to fix it?
1.69 GB
true wavelet transform, 12.8x dr, 12 channels per scale, 8 bits, no hypercodec, 15 epochs
v0.1.2.pth
Detected Pickle imports (7)
- "collections.OrderedDict",
- "__main__.Args",
- "torch.FloatStorage",
- "numpy.dtype",
- "torch._utils._rebuild_tensor_v2",
- "_codecs.encode",
- "numpy.core.multiarray.scalar"
How to fix it?
1.69 GB
true wavelet transform, wavelet loss, 12.8x dr, 12 channels per scale, 8 bits, no hypercodec, 15 epochs
v0.1.3.pth
Detected Pickle imports (7)
- "numpy.core.multiarray.scalar",
- "_codecs.encode",
- "__main__.Args",
- "numpy.dtype",
- "torch._utils._rebuild_tensor_v2",
- "collections.OrderedDict",
- "torch.FloatStorage"
How to fix it?
1.93 GB
true wavelet, 12x dr, with hypercodec 15 epochs
v0.1.4.pth
Detected Pickle imports (7)
- "torch.FloatStorage",
- "numpy.core.multiarray.scalar",
- "__main__.Args",
- "numpy.dtype",
- "_codecs.encode",
- "torch._utils._rebuild_tensor_v2",
- "collections.OrderedDict"
How to fix it?
1.41 GB
true wavelet with hypercodec, 8x rr, 12x dr
v0.1.5.pth
Detected Pickle imports (7)
- "numpy.core.multiarray.scalar",
- "numpy.dtype",
- "_codecs.encode",
- "torch.FloatStorage",
- "__main__.Args",
- "collections.OrderedDict",
- "torch._utils._rebuild_tensor_v2"
How to fix it?
1.41 GB
true wavelet with hypercodec, modified quantization, 8x rr
v0.2.1.pth
Detected Pickle imports (7)
- "_codecs.encode",
- "__main__.Args",
- "numpy.core.multiarray.scalar",
- "torch.FloatStorage",
- "collections.OrderedDict",
- "torch._utils._rebuild_tensor_v2",
- "numpy.dtype"
How to fix it?
447 MB
symmetric, 12x dr
v0.2.2.pth
Detected Pickle imports (7)
- "numpy.core.multiarray.scalar",
- "collections.OrderedDict",
- "torch._utils._rebuild_tensor_v2",
- "_codecs.encode",
- "__main__.Args",
- "numpy.dtype",
- "torch.FloatStorage"
How to fix it?
445 MB
symmetric, 48x dr
v0.2.3.pth
Detected Pickle imports (7)
- "__main__.Args",
- "torch.FloatStorage",
- "numpy.core.multiarray.scalar",
- "collections.OrderedDict",
- "torch._utils._rebuild_tensor_v2",
- "_codecs.encode",
- "numpy.dtype"
How to fix it?
1.69 GB
symmetric fb, ternary quantization, 10d (16 bits)
v0.3.1.pth
Detected Pickle imports (7)
- "collections.OrderedDict",
- "__main__.Args",
- "torch._utils._rebuild_tensor_v2",
- "numpy.core.multiarray.scalar",
- "numpy.dtype",
- "torch.FloatStorage",
- "_codecs.encode"
How to fix it?
494 MB
32x rr 48x dr (baseline)
v0.3.2.pth
Detected Pickle imports (7)
- "collections.OrderedDict",
- "__main__.Args",
- "torch._utils._rebuild_tensor_v2",
- "numpy.core.multiarray.scalar",
- "numpy.dtype",
- "torch.FloatStorage",
- "_codecs.encode"
How to fix it?
494 MB
8 bits
v0.3.3.pth
Detected Pickle imports (7)
- "torch.FloatStorage",
- "__main__.Args",
- "numpy.core.multiarray.scalar",
- "torch._utils._rebuild_tensor_v2",
- "numpy.dtype",
- "collections.OrderedDict",
- "_codecs.encode"
How to fix it?
494 MB
deeper (6 instead of 4)
v0.3.4.pth
Detected Pickle imports (7)
- "torch._utils._rebuild_tensor_v2",
- "_codecs.encode",
- "numpy.dtype",
- "collections.OrderedDict",
- "torch.FloatStorage",
- "numpy.core.multiarray.scalar",
- "__main__.Args"
How to fix it?
635 MB
no rr, 768
v0.3.5.pth
Detected Pickle imports (7)
- "torch._utils._rebuild_tensor_v2",
- "_codecs.encode",
- "numpy.dtype",
- "collections.OrderedDict",
- "torch.FloatStorage",
- "numpy.core.multiarray.scalar",
- "__main__.Args"
How to fix it?
2.38 GB
no rr 1536
v0.3.6.pth
Detected Pickle imports (7)
- "torch._utils._rebuild_tensor_v2",
- "_codecs.encode",
- "numpy.dtype",
- "collections.OrderedDict",
- "torch.FloatStorage",
- "numpy.core.multiarray.scalar",
- "__main__.Args"
How to fix it?
494 MB
bior 4.4
v0.3.7.pth
Detected Pickle imports (7)
- "torch._utils._rebuild_tensor_v2",
- "__main__.Args",
- "_codecs.encode",
- "numpy.core.multiarray.scalar",
- "torch.FloatStorage",
- "collections.OrderedDict",
- "numpy.dtype"
How to fix it?
500 MB
LSDIR 540
v0.3.8.pth
Detected Pickle imports (7)
- "__main__.Args",
- "numpy.core.multiarray.scalar",
- "_codecs.encode",
- "collections.OrderedDict",
- "numpy.dtype",
- "torch._utils._rebuild_tensor_v2",
- "torch.FloatStorage"
How to fix it?
250 MB
16x rr
v0.4.1.pth
Detected Pickle imports (7)
- "__main__.Args",
- "numpy.core.multiarray.scalar",
- "_codecs.encode",
- "collections.OrderedDict",
- "numpy.dtype",
- "torch._utils._rebuild_tensor_v2",
- "torch.FloatStorage"
How to fix it?
195 MB
36 bits (6c 6b)
v0.4.2.pth
Detected Pickle imports (7)
- "__main__.Args",
- "numpy.core.multiarray.scalar",
- "_codecs.encode",
- "collections.OrderedDict",
- "numpy.dtype",
- "torch._utils._rebuild_tensor_v2",
- "torch.FloatStorage"
How to fix it?
195 MB
actually 36 bits (6c 6b)
v0.4.3.pth
Detected Pickle imports (7)
- "__main__.Args",
- "numpy.core.multiarray.scalar",
- "_codecs.encode",
- "collections.OrderedDict",
- "numpy.dtype",
- "torch._utils._rebuild_tensor_v2",
- "torch.FloatStorage"
How to fix it?
192 MB
36 bits (12c 3b)
v0.4.4.pth
Detected Pickle imports (7)
- "numpy.dtype",
- "numpy.core.multiarray.scalar",
- "__main__.Args",
- "torch.FloatStorage",
- "collections.OrderedDict",
- "torch._utils._rebuild_tensor_v2",
- "_codecs.encode"
How to fix it?
191 MB
70 bits (16c 5b)
v0.4.5.pth
Detected Pickle imports (7)
- "_codecs.encode",
- "torch.FloatStorage",
- "collections.OrderedDict",
- "torch._utils._rebuild_tensor_v2",
- "numpy.dtype",
- "__main__.Args",
- "numpy.core.multiarray.scalar"
How to fix it?
191 MB
(d)phi = 0.01, latent_dim=9, J=3, 4 bits
-
242 MB
(tfnorm) phi = 0, latent_dim=48, J=4, 6 bits
-
242 MB
(tfnorm) phi = 0.01, latent_dim=48, J=4, 6 bits
v0.5.3.pth
Detected Pickle imports (7)
- "torch.FloatStorage",
- "collections.OrderedDict",
- "__main__.Args",
- "numpy.core.multiarray.scalar",
- "numpy.dtype",
- "_codecs.encode",
- "torch._utils._rebuild_tensor_v2"
How to fix it?
191 MB
phi = 0.1, latent_dim=12, J=3, 5 bits
-
194 MB
(tfnorm) phi = 0.01, latent_dim=16, J=3, 5 bits
v0.6.1.pth
Detected Pickle imports (7)
- "torch._utils._rebuild_tensor_v2",
- "_codecs.encode",
- "__main__.Args",
- "collections.OrderedDict",
- "torch.FloatStorage",
- "numpy.core.multiarray.scalar",
- "numpy.dtype"
How to fix it?
195 MB
phi = 0, latent_dim=16, J=3, 5 bits
v0.6.1_ext.pth
Detected Pickle imports (7)
- "__main__.Args",
- "numpy.core.multiarray.scalar",
- "torch._utils._rebuild_tensor_v2",
- "_codecs.encode",
- "collections.OrderedDict",
- "numpy.dtype",
- "torch.FloatStorage"
How to fix it?
209 MB
extended training
v0.6.2.pth
Detected Pickle imports (7)
- "torch._utils._rebuild_tensor_v2",
- "_codecs.encode",
- "__main__.Args",
- "collections.OrderedDict",
- "torch.FloatStorage",
- "numpy.core.multiarray.scalar",
- "numpy.dtype"
How to fix it?
195 MB
(tf) phi = 1.0, latent_dim=16, J=3, 5 bits
v0.6.2_ext.pth
Detected Pickle imports (7)
- "torch.FloatStorage",
- "torch._utils._rebuild_tensor_v2",
- "numpy.dtype",
- "collections.OrderedDict",
- "numpy.core.multiarray.scalar",
- "__main__.Args",
- "_codecs.encode"
How to fix it?
206 MB
extended training
v0.6.3.pth
Detected Pickle imports (7)
- "torch.FloatStorage",
- "torch._utils._rebuild_tensor_v2",
- "numpy.dtype",
- "collections.OrderedDict",
- "numpy.core.multiarray.scalar",
- "__main__.Args",
- "_codecs.encode"
How to fix it?
195 MB
(tfabs) phi = 0.01, latent_dim=16, J=3, 5 bits
v0.6.3_ext.pth
Detected Pickle imports (7)
- "torch.FloatStorage",
- "torch._utils._rebuild_tensor_v2",
- "numpy.dtype",
- "collections.OrderedDict",
- "numpy.core.multiarray.scalar",
- "__main__.Args",
- "_codecs.encode"
How to fix it?
206 MB
extended training
v0.7.0.pth
Detected Pickle imports (7)
- "collections.OrderedDict",
- "torch.FloatStorage",
- "__main__.Config",
- "numpy.dtype",
- "_codecs.encode",
- "numpy.core.multiarray.scalar",
- "torch._utils._rebuild_tensor_v2"
How to fix it?
220 MB
256 encoder 8 layer decoder
v0.8.0.pth
Detected Pickle imports (7)
- "torch.FloatStorage",
- "collections.OrderedDict",
- "numpy.dtype",
- "_codecs.encode",
- "numpy.core.multiarray.scalar",
- "__main__.Config",
- "torch._utils._rebuild_tensor_v2"
How to fix it?
160 MB
16 channels 256 encoder
-
216 MB
retrain with smaller batch size and more extreme lr decay
v0.8.3.pth
Detected Pickle imports (7)
- "numpy.dtype",
- "torch.FloatStorage",
- "__main__.Config",
- "torch._utils._rebuild_tensor_v2",
- "collections.OrderedDict",
- "numpy.core.multiarray.scalar",
- "_codecs.encode"
How to fix it?
124 MB
lightweight encoder
