Hi,
After following the step-by-step guide, I was able to train a model on
44K sentences for the language pair fr-en. I am trying to repeat
training but this time on approx. 1.4 million sentences from europarl
corpus. The experiment runs fine till the step of building n-gram
models. However, it seems to fail while training a phrase-based model.
I am using cygwin on windows XP with RAM 3.25GB
========
Exception: STATUS_ACCESS_VIOLATION at eip=610D3C8E
eax=00000000 ebx=00000000 ecx=61150140 edx=FFFFFFFF esi=00000000
edi=7FDB2CD8
ebp=0022C258 esp=0022C240
program=C:\cygwin\home\moses\tools\bin\snt2cooc.out, pid 4348, thread main
cs=001B ds=0023 es=0023 fs=003B gs=0000 ss=0023
Stack trace:
Frame Function Args
0022C258 610D3C8E (00000000, 00000000, 00000000, FFFFFFFF)
0022C368 610D75D0 (0044B000, 00000000, 00000014, 00010000)
0022C468 610DB0F3 (0044B000, 00010000, 00000000, FFFFFFFF)
0022C4D8 610B5178 (00000004, 0022C480, 0022C67C, 0041623F)
0022C538 004055BA (00000018, 5F9D58D0, 745B46A8, 7FDB2CDC)
0022C568 00443DA4 (7FDB2CD8, 00000000, 7816CCB0, 0022C658)
0022C598 00443B40 (7FDB2CD8, 7E5CE968, 0022C658, 00402000)
0022CD78 00402DB8 (00000004, 61210304, 007100F8, 61004A1D)
0022CDA8 61006DDA (00000000, 0022CDE0, 610066E0, 7FFDF000)
End of stack trace
1 [main] snt2cooc.out 4348
C:\cygwin\home\moses\tools\bin\snt2cooc.out: *** fatal error - cmalloc
would have returned NULL
=========
Am I running short of RAM? Could someone help?
Thanks,
Niraj
_______________________________________________
Moses-support mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/moses-support
