This dataset presents a catalog of stellar atmospheric parameters and abundances of 16 chemical elements derived from low-resolution spectra of LAMOST (Large Sky Area Multi-Object Fiber Spectroscopic Telescope) DR11, estimated using the Spectral Transformer Encoder (SpecTE) model. SpecTE learns the mapping relationship between low-quality and high-quality spectra through pre-training, enabling the model to extract features from low-quality spectra and significantly improve the estimation accuracy of low-quality spectra. We pre-trained SpecTE using repeated observations of high/low signal-to-noise ratio data from LAMOST DR11, then conducted further training using cross-matched common stars from LAMOST DR11 and APOGEE DR17 catalogs. The catalog contains estimated parameters and their uncertainties for 9.8 million low-resolution spectra from LAMOST DR11, including: observation spectrum identifier (obsid), coordinate information (ra, dec), spectral signal-to-noise ratio (snrg), effective temperature (Teff), surface gravity (logg), radial velocity (RV), abundance information of 16 elements (X/H), and 1σ parameter uncertainties (X_uncertainty). Under S/Ng ≥ 5 conditions, the MAEs of SpecTE estimates are as follows: Teff and log g achieve precisions of 45K and 0.08dex respectively; elements Fe, Mg, Si, Ni, Ca, C show 0.037dex~0.05dex; Al, Mn, O, S, K show 0.05dex~0.077dex; Ti, N, Cr show 0.09dex~0.12dex; V and Na show 0.15dex and 0.21dex respectively. Compared with DD-Payne and StarGRUNet models, SpecTE demonstrates higher accuracy and robustness. Compared with APOGEE (Apache Point Galactic Evolution Experiment) and GALAH (The Galactic Archaeology with HERMES) surveys, SpecTE exhibits better consistency within reasonable deviation ranges. This catalog provides 9.8 million low-resolution spectra from LAMOST DR11 for reference in astronomical science exploration and data processing algorithm research.