Shikhar Jaiswal
@maumueller This is because the current code is slightly hard-coded to run only on systems which have AVX-512 support (2017 or later CPUs). For running on non-AVX512 systems, a few...
Umm, actually I think the CI servers are making use of AVX-512 compatible processors. I don't see any of the builds failing. As for support for other platforms, it's just...
Ah, I didn't realise the repository had shifted from Travis. I'll look into this then.
@SachinG007 Can you please take a look?
Also can we have more expressive macro names than `O_T`, `I_T`, `O_F` and `I_F` in the weight files?
@affot The code pertaining to `SeeDot` is slightly unorganised. I've encountered this error before as well, and most of the times, it is because of the datasets and the model...
Hmm, in that case, could you look at `temp/Predictor/output/fixed/` and paste the outputs of `build.txt` and `exec.txt` please?
Okay, I must apologize for leading you astray. It seems that the nomenclature that is being followed internally is `usps-multiclass` instead of `usps10`. Could you try changing `datasets/bonsai/usps10` to `datasets/bonsai/usps-multiclass`...
Could you share the `input.sd` inside the `model/bonsai/usps-multiclass`?
@affot Sorry for the late reply. You're using the DSL code pertaining to `FastGRNN` for running the `Bonsai` code for `usps-multiclass`. The converter is currently broken, and we're trying to...