Deep　learning　can　directly　extract　useful　features　from　raw　signals　without　laborious　manual　feature　extraction,　which　has　become　popular　for　the　automatic　sleep　stage　scoring　task.　However,　most　of　the　existing　models　focused　only　on　extracting　information　from　the　input　sequence,　without　paying　attention　to　the　explicit　transition　rules　of　the　output　sequence.　We　proposed　an　end-to-end　deep　learning　model　for　automatic　sleep　stage　scoring　based　on　raw　single-channel　EEG　signals.　In　this　model,　we　used　Convolutional　Neural　Network　(CNN)　to　extract　features　from　raw　signals,　and　used　Long　Short-Term　Memory　(LSTM)　to　learn　implicit　information　from　the　input　sequence.　Then,　we　utilized　Conditional　Random　Field　(CRF)　to　learn　explicit　transition　rules　of　labels　from　the　output　sequence.　We　evaluated　our　model　on　the　Fpz-Cz　and　Pz-Oz　channels　of　the　SleepEDF-20　and　the　SleepEDF-78　datasets.　The　results　showed　that　our　method　achieved　competitive　or　better　performances　compared　to　the　state-of-the-art　methods.　©　2021　ACM.