Fix IMRPhenomPv2 batch inputs bug

[ravioli1369]

Addresses #179 and #176 (comment)

[ravioli1369]

@deepchatterjeeligo, these are the parameters that pass the O(1e-2) tests but not the O(1e-3) ones

mass_1	mass_2	chi1x	chi1y	chi1z	chi2x	chi2y	chi2z	distance	inclination	phase	tc	f_ref	theta_jn	phi_jl	tilt_1	tilt_2	phi_12	a_1	a_2
274.208	36.652	-0.267	0.173	-0.007	0.004	-0.003	0.134	138.137	1.701	3.395	0.0	20	1.877	2.976	1.594	0.035	3.033	0.318	0.135
118.699	77.099	0.411	-0.308	-0.717	0.057	0.031	-0.532	201.967	0.301	2.815	0.0	20	0.269	0.834	2.520	3.020	1.134	0.881	0.536
133.840	59.373	0.369	-0.092	0.721	-0.012	-0.181	-0.132	146.473	0.753	3.983	0.0	20	0.854	0.654	0.486	2.199	4.890	0.815	0.225
201.940	28.939	0.006	-0.087	-0.026	-0.192	0.023	-0.142	111.424	1.060	2.409	0.0	20	0.908	3.906	1.863	2.204	4.524	0.091	0.240
194.882	42.329	-0.265	0.086	0.751	-0.206	-0.297	0.490	225.858	2.961	2.123	0.0	20	3.084	3.557	0.355	0.636	1.278	0.801	0.609
145.800	44.009	0.224	-0.468	0.650	0.053	0.043	-0.053	144.681	1.044	2.184	0.0	20	0.779	4.082	0.674	2.237	1.807	0.832	0.087

[deepchatterjeeligo]

@ravioli1369 thanks for tracking that down. So it does not seem like the large spin values are an issue. The failures are caused due to large chirp mass values and small distances which means they are just louder waveforms.