Relates #31 | Add SEACrowd Speech Classification

[sabilmakbar]

It's related to Issue #31 in Fleurs Submission, where the schema of Speech Classification isn't available on SEACrowd Schema and Configs. This PR is created to add a new schema of Speech Classification.

cc @holylovenia @SamuelCahyawijaya

[holylovenia]

Since we have text and text_multilabel schemas, isn't it better to be consistent and split the binary and the multilabel for speech classification into 2 separate schemas? What do you think? @sabilmakbar @SamuelCahyawijaya @fajri91

[sabilmakbar]

Apologies for mixing up things. That schema was supposed for Multiclass Classification, not Multilabel, will fix that soon. But I do agree that we shd separate it as 2. Would that be okay if we create another one for Multilabel regardless of schema adoption status (no one is using it rn)?

[sabilmakbar]

Hi, the codes have been updated & ready to be reviewed, implementing changes based on suggestion from Holy.

[holylovenia]

@sabilmakbar Thanks for the discussion! Let's change the schema names to SPEECH and SPEECH_MULTI so they're consistent with TEXT and TEXT_MULTI?

[sabilmakbar]

Apologies Holy, overlooked that one. Please review it again and see whether all of the checks have been completed. Thanks!

[holylovenia]

Thanks for the changes, @sabilmakbar! Looks good to me.

[SamuelCahyawijaya]

LGTM!