-
Hi,
When I try to run step 1s -- dataset formatting -- I get the following error:
```
vocal_Saya.wav.reformatted.wav_10.flac_0022832960_0022952000.wav
vocal_Saya.wav.reformatted.wav_10.flac_0022…
-
### Feature
Switch distribution libraries...again!
The problem with `tfp` is that if we want to run non-shared parameters we need `vmap` the apply function (over params and observations), but this…
-
The disk-usage-based-eviction code logs almost everything it does at WARN level.
That made sense in the past but nowadays we run with `OnlyImitiate`, so, disk-usage-based eviction is the norm.
Demot…
-
Compiling current CMSIS for Host platfom (AMD64) while gcc's static code analysis feature is active results in serious looking errors.
GCC version:
```
$ gcc --version
gcc (Ubuntu 11.2.0-19ubunt…
-
**Gammapy version**
0.20.1
**Bug description**
When flux points are estimated by using FluxPointsEstimator, there are strange values on the flux points table like:
1. upper limit (e2dnde_ul) …
-
存储为npz文件的稀疏矩阵A和D张量的大小分别是(25718x25718)和(1x1),所以当在train.py里执行到D.mm(A).mm(D)这儿的时候,报错RuntimeError: mat1 and mat2 shapes cannot be multiplied (1x1 and 25718x25718)。
-
**Is your feature request related to a problem? Please describe.**
The `gempyor.utils.as_random_distribution/random_distribution_sampler` functions use a mix of `np.random` rngs and `scipy.stats` r…
-
We should only have the variant without `p`. Otherwise it would conflict with the new r-norm methods.
-
Hello, thanks for your excellent open-source code. May I ask when will you release your training code? I would like to tune the model but have a little problem regarding the batched training issue. Ho…
-
Hi guys,
I attempted to use VSSBlock based on mamba2 as the backbone in the model, with similar settings as in vmamba_base_m2
The parameters are
dims=128,ssm_d_state=64, ssm_ratio=1.0, ssm_dt_rank=…