NLP Wiki

User Tools

Site Tools

Trace: • experimental_method • alternative_training_methods • zero-shot_learning • distributed_training • learning_rate • state-space_models • patent_domain_nlp
nlp:seq2seq

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
nlp:seq2seq [2025/04/01 06:15] – [Decoding Strategies] jmflanignlp:seq2seq [2025/05/29 07:15] (current) – [Decoding Strategies] jmflanig
Line 17: Line 17:
     * [[https://arxiv.org/pdf/2109.05093.pdf|Scholak et al 2021 - PICARD: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models]]     * [[https://arxiv.org/pdf/2109.05093.pdf|Scholak et al 2021 - PICARD: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models]]
     * [[https://arxiv.org/pdf/2201.11227.pdf|Poesia et al 2022 - Synchromesh: Reliable Code Generation from Pre-trained Language Models]] They created a tool that will take a ANTLR parser and a string, and give you the set of valid next token completions (see sect 3.1).     * [[https://arxiv.org/pdf/2201.11227.pdf|Poesia et al 2022 - Synchromesh: Reliable Code Generation from Pre-trained Language Models]] They created a tool that will take a ANTLR parser and a string, and give you the set of valid next token completions (see sect 3.1).
 +    * **[[https://arxiv.org/pdf/2305.13971|Geng et al 2023 - Grammar-Constrained Decoding for Structured NLP Tasks without Finetuning]]** Elegant solution. Uses Grammatical Framework to constrain the outputs.
     * **[[https://arxiv.org/pdf/2403.06988|Beurer-Kellner et al 2024 - Guiding LLMs The Right Way: Fast, Non-Invasive Constrained Generation]]**     * **[[https://arxiv.org/pdf/2403.06988|Beurer-Kellner et al 2024 - Guiding LLMs The Right Way: Fast, Non-Invasive Constrained Generation]]**
   * **Parallel Decoding**   * **Parallel Decoding**
Line 26: Line 27:
     * [[https://arxiv.org/pdf/2211.17192|Leviathan et al 2024 - Fast Inference from Transformers via Speculative Decoding]]     * [[https://arxiv.org/pdf/2211.17192|Leviathan et al 2024 - Fast Inference from Transformers via Speculative Decoding]]
     * [[https://arxiv.org/pdf/2404.11912|Sun et al 2024 - TriForce: Lossless Acceleration of Long Sequence Generation with Hierarchical Speculative Decoding]]     * [[https://arxiv.org/pdf/2404.11912|Sun et al 2024 - TriForce: Lossless Acceleration of Long Sequence Generation with Hierarchical Speculative Decoding]]
 +    * [[https://arxiv.org/pdf/2502.17421|Yang et al 2025 - LongSpec: Long-Context Speculative Decoding with Efficient Drafting and Verification]]
 +    * [[https://arxiv.org/pdf/2505.20776|Cha et al 2025 - SpecExtend: A Drop-in Enhancement for Speculative Decoding of Long Sequences]]
   * **Miscellaneous Decoding Techniques**   * **Miscellaneous Decoding Techniques**
     * Contrastive Decoding     * Contrastive Decoding
nlp/seq2seq.1743488111.txt.gz · Last modified: 2025/04/01 06:15 by jmflanig
Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki