After upgrade to plotly 4.14.3 all my charts are broken

After upgrading plotly.py to 4.14.3 from 4.13.0 my charts now display incorrectly, see sample below.
Charts are created from pandas dataframe data.
Do I need to do something different now when working with pandas and plotly.
If use cufflinks to make some charts and they are still OK.

Sample of fig creation is:
fig=make_subplots(rows=17, cols=1,shared_xaxes=True,
subplot_titles=(ā€œHITSā€, ā€œACCESSESā€, ā€œLockBlkā€,ā€œHIT %ā€, ā€œDelAck NLā€, ā€œDelAck FCā€, ā€œDelAck SSDā€, ā€œPages Freeā€,ā€œPages Cleanā€,ā€œPages Write1ā€,ā€œPages WriteNā€,ā€œPages WriteSchedā€,ā€œPages Writingā€,ā€œPages DcowPendingā€,ā€œCMP Dirty/Max NLā€,ā€œCMP Dirty/Max FCā€,ā€œCMP Dirty/Max SSDā€),vertical_spacing=0.03)
for c2print in col_2_print:
chart_name,chart_row_no,chart_col_no,axis_color = xlate_4_col_names[c2print]
fig.add_trace(go.Scatter(x=fig_df.index, y=fig_df.loc[:,(c2print)],
name = chart_name,
line=dict(color=(axis_color)),
hoverlabel = dict(namelength = -1)),row=chart_row_no, col=chart_col_no)

I think I worked it out.
My df has dtype = object, if I change it to float all is back to normal.

eg:
y=cpu_df.loc[:,(c2print)].astype(ā€˜floatā€™)