China DOS Union

In one directory there are the following files:
abc.ncmm_0
abc.ncmm_1
abc.ncmm_2
abc.ncmm_3
abc.ncmm_4
abc.ncmm_5
I wrote a piece of code to display the specified content in each file:
@echo off
for /f "delims=" %%i in ('dir /b /on *ncmm*') do (
for /f "delims=!" %%a in ('findstr ! %%i') do (
echo %%a
)
)
pause
The displayed content is as follows:
<td width=83 align=center rowspan=2>IN102</td> <
<td width=83 align=center rowspan=2>IN103</td> <
<td width=83 align=center rowspan=2>IN104</td> <
<td width=83 align=center rowspan=2>IN105</td> <
<td width=83 align=center rowspan=2>IN105</td> <
<td width=83 align=center rowspan=2>IN106</td> <
There are two IN105 entries. Now I want to delete the file abc.ncmm_3 corresponding to the first IN105, and then merge the remaining files. What exactly should I do? Thanks.

I didn't explain it clearly. The situation is like this: these ncmm files are output by another program. The number of files is random, maybe three or five, maybe dozens. The numeric suffixes of files whose specified content is duplicated are consecutive. It could be 1 2 3, or it could be 8 9, but there will never be a case like 2 and 4 having the same content (if that happens, then 2 3 4 would all have the same content).

Oh, my understanding was a bit off too.

One more question: do you want to merge the filtered files, or merge the original files?

Merge the filtered files