Fitting a 1D Gaussian using pymc

Question

I have a 1D array that has a a signal with a peak. The peak looks like a 1D Gaussian, hence I am trying to find the x value for the signal's peak (or the mean of the fit 1D Gaussian) using pymc. Below is the snippet of my pymc code and the plot showing the fit trace. So far, the model does not seem to be sensitive to the data. Not sure what I am doing wrong. Any help is appreciated.

signal ### 1D array with the signal peak
X = np.arange(1024)
with pm.Model() as model:
    ### gaussian fit
    #amplitude * np.exp(-((x - mean) / stddev)**2 / 2)
    amp = pm.Uniform('amp', 0,  max(fj_line) + 20*noise)
    mean = pm.Uniform('mean', 512 - 200 , 512 + 200)
    std = pm.Uniform('std', 0, 100)

    Y = pm.Normal('Y', mu=amp*np.exp(-((X - mean)/std)**2 / 2),
                  observed=signal)
    samp = pm.sample(tune=500,draws=2000,chains=2,cores=12)

pm.plot_trace(samp)

Below plot shows the actual signal (black line) and the fit trace

And the below figure shows the posterior of the fit parameters

The model fitting simply does not seem to be sensitive to the input data and I cant seem to figure out why. Any help/comments is appreciated. Thanks.