{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2018:OUS2V7WTSEBXE5QB5DD5D5UP5V","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"7f6828ce44c25325c6b03c2ddae05a958c176bdef41406429f96ce8b335e89c0","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2018-05-14T15:23:40Z","title_canon_sha256":"581b10b0378e5410103e30d29e2f0dc3343b24886b8d829b2d239c3031e02eea"},"schema_version":"1.0","source":{"id":"1805.05225","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1805.05225","created_at":"2026-07-04T23:51:18Z"},{"alias_kind":"arxiv_version","alias_value":"1805.05225v2","created_at":"2026-07-04T23:51:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1805.05225","created_at":"2026-07-04T23:51:18Z"},{"alias_kind":"pith_short_12","alias_value":"OUS2V7WTSEBX","created_at":"2026-07-04T23:51:18Z"},{"alias_kind":"pith_short_16","alias_value":"OUS2V7WTSEBXE5QB","created_at":"2026-07-04T23:51:18Z"},{"alias_kind":"pith_short_8","alias_value":"OUS2V7WT","created_at":"2026-07-04T23:51:18Z"}],"graph_snapshots":[{"event_id":"sha256:7fff7ccc0ed652f3a9f60662590bf4dab9da6ca4b5eaf728108da0467923be1b","target":"graph","created_at":"2026-07-04T23:51:18Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/1805.05225/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We compare the fast training and decoding speed of RETURNN of attention models for translation, due to fast CUDA LSTM kernels, and a fast pure TensorFlow beam search decoder. We show that a layer-wise pretraining scheme for recurrent attention models gives over 1% BLEU improvement absolute and it allows to train deeper recurrent encoder networks. Promising preliminary results on max. expected BLEU training are presented. We are able to train state-of-the-art models for translation and end-to-end models for speech recognition and show results on WMT 2017 and Switchboard. The flexibility of RETU","authors_text":"Albert Zeyer, Hermann Ney, Tamer Alkhouli","cross_cats":["cs.AI","cs.CL"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2018-05-14T15:23:40Z","title":"RETURNN as a Generic Flexible Neural Toolkit with Application to Translation and Speech Recognition"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1805.05225","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:dd3cae971c6dc8220eca81347e389e855eb619b63615d3cf5c688674bb83b2f0","target":"record","created_at":"2026-07-04T23:51:18Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"7f6828ce44c25325c6b03c2ddae05a958c176bdef41406429f96ce8b335e89c0","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2018-05-14T15:23:40Z","title_canon_sha256":"581b10b0378e5410103e30d29e2f0dc3343b24886b8d829b2d239c3031e02eea"},"schema_version":"1.0","source":{"id":"1805.05225","kind":"arxiv","version":2}},"canonical_sha256":"7525aafed39103727601e8c7d1f68fed7cabb0f3108c2ff24784322611ab78ab","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7525aafed39103727601e8c7d1f68fed7cabb0f3108c2ff24784322611ab78ab","first_computed_at":"2026-07-04T23:51:18.342102Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-04T23:51:18.342102Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"5wInWEodZPFf8fAaWAOrLc9SJ8cI9ChNACM5t483S9UoPiJZvhxAmBHByQG5F5gnZJFOAEw+r+iIdrERDyf3Aw==","signature_status":"signed_v1","signed_at":"2026-07-04T23:51:18.342580Z","signed_message":"canonical_sha256_bytes"},"source_id":"1805.05225","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:dd3cae971c6dc8220eca81347e389e855eb619b63615d3cf5c688674bb83b2f0","sha256:7fff7ccc0ed652f3a9f60662590bf4dab9da6ca4b5eaf728108da0467923be1b"],"state_sha256":"ab2cc91c277c206f8bef6a0a2ea8e23ccaba5e558bb72739fc68a08678633276"}