Add strain channel name field to `MLy` GraceDb model

MLy record (see https://gracedb-playground.ligo.org/events/G2273277/view/) do not report the chanal used for the analysis even if is correctly uploaded to gracedb: https://gracedb-playground.ligo.org/events/G2273277/files/T_1411734936.141145_HL.json,0

{
    "gpstime": 1411734936.141145,
    "far": 1.0437078851319122e-06,
    "ifos": "H1,L1",
    "channels": [
        "H1:GDS-CALIB_STRAIN_INJ1_O3Replay",
        "L1:GDS-CALIB_STRAIN_INJ1_O3Replay"
    ],
    "scores": {
        "coincidence": 0.9636721014976501,
        "coherency": 0.0931749939918518,
        "combined": 0.08979014226715876
    },
    "instance_start_time": 1411734936.0,
    "SNR": 10.43516549461369,
    "central_time": 1411734936.141145,
    "duration": 0.125,
    "central_freq": 63.502407310308236,
    "bandwidth": 80.0,
    "start_time": 0.0625,
    "end_time": 0.1875,
    "start_frequency": 24.0,
    "end_frequency": 104.0
}

Machine Learning Info" contains fields irrelevant to MLy, and the channel (channels?) information is not ingested from MLy. Concerning the playground deployment, there is no connection between the MLy pipeline and the gawk one (not to be deployed during O4?), and there is no reason (and it is confusing) to mix incompatible information.

aframe: https://gracedb-playground.ligo.org/events/G2275944/view/ does just import the channel in an incomplete form unsuited to have VIRGO and KAGRA frame channels name.

gawk: find out to example in playground-test

The "Requirements for search pipelines to run online" require that The BURST pipeline should upload information about the event duration or some metric from which it can be calculated.

Related issue: #354 (closed) Related merge request: !234 (merged)