Data Prep

[stefi-says]

COmpile the training data form diferent sources in to flagged dataset containing the features necessarie to do the prediction

[stefi-says]

schema specified on the paper of BERT4ETH:

'hash, nonce ,block_hash ,block_number ,transaction_index ,from_address ,to_address ,value, gas, gas_price, input, block_timestamp, max_fee_per_gas ,max_priority_fee_per_gas ,transaction_type

Not all the columns are used during the code. MOre information on the coluns used can be found o the paper at : 4.2.1

[stefi-says]

this issue depends on the completion of issue #5 and #8

[epowell101]

Here is the specification from the docs for Ethereum-etl as mentioned by the authors: https://ethereum-etl.readthedocs.io/en/latest/schema/

[epowell101]

There are two steps really, one is to preprocesses the data for our Sybils and the second is to then create an embedding from that preprocessing.