Dynesty likelihood error while fitting spectroscopy and photometry together

[GSBhargaviR]

Hello,
I'm attempting to jointly fit the spectroscopy and photometry of a quiescent galaxy at z = 1.948 using Prospector v1.4.0 with the Dynesty sampler. While I previously used the C3K stellar library without any issues, I recently switched to MILES and have been encountering the following error.

I am only fitting for basic parameters:
'zred', 'logzsol', 'dust2', 'logmass', 'logsfr_ratios_1', 'logsfr_ratios_2', 'logsfr_ratios_3', 'logsfr_ratios_4', 'logsfr_ratios_5', 'logsfr_ratios_6', 'logsfr_ratios_7', 'logsfr_ratios_8', 'logsfr_ratios_9', 'sigma_smooth', 'spec_jitter'
I'm using polyorder=6 for the fit.
My data extends beyond the high-resolution coverage of the MILES library, so I masked the spectroscopic range to 3525–7500 Å. However, the issue still persists.

I also tried masking the spectroscopic data more aggressively on the red end, and at one point, limiting the range to just 3525–3800 Å seemed to work. However, this is a very narrow wavelength range, which seems odd.
That said, fitting the photometry alone runs without any issues.
Would you be able to help me troubleshoot this?