Closed Chengqian-Zhang closed 1 month ago
doc/model/train-se-atten.md
91-91: Column: 1 Hard tabs --- 92-92: Column: 18 Hard tabs --- 93-93: Column: 23 Hard tabs --- 94-94: Column: 18 Hard tabs --- 95-95: Column: 17 Hard tabs --- 96-96: Column: 20 Hard tabs --- 97-97: Column: 25 Hard tabs --- 98-98: Column: 23 Hard tabs --- 99-99: Column: 18 Hard tabs --- 100-100: Column: 24 Hard tabs --- 101-101: Column: 23 Hard tabs --- 102-102: Column: 23 Hard tabs --- 103-103: Column: 18 Hard tabs --- 104-104: Column: 1 Hard tabs
doc/model/train-se-atten.md (1)
`129-138`: Clarify the explanation of `"tebd_input_mode"` parameter changes. The explanation of the `"tebd_input_mode"` parameter changes from `"concat"` to `"strip"` is provided, but it could benefit from a clearer distinction between the two modes. Consider adding a brief description of what each mode does at a high level to help readers understand the practical implications of these settings.
All modified and coverable lines are covered by tests :white_check_mark:
Project coverage is 75.18%. Comparing base (
3a7fbcf
) to head (9ad3a30
).
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
@iProzd I notice the documentation for se_atten is still old. Update: revised in #3832.
I think it is better to update the documentation in #3840.
Solve issue #3139
"se_atten_v2"
is inherited from"se_atten"
with the following parameter modifications:I add the equations for parameter
"tebd_input_mode"
.Summary by CodeRabbit
"tebd_input_mode"
parameter."se_atten_v2"
over"se_atten"
.se_atten_v2
.