MACHINE LEARNING

NL2SQL Techniques Survey: LLM Models and Evaluation Methods

By

–

18 August 2024 18h24

10/ Survey of NL2QL – a comprehensive overview of NL2SQL techniques powered by LLMs; covers models, data collection, evaluation methods, and error analysis.

rStar Enhances Small Language Models Reasoning Without Fine-tuning

By

–

18 August 2024 18h24

7/ rStar – introduces self-play mutual reasoning to improve the reasoning capabilities of small language models without fine-tuning or superior models; MCTS is augmented with human-like reasoning actions, obtained from SLMs, to build richer reasoning trajectories…

Optimal Scaling of LLM Inference-Time Compute

By

–

18 August 2024 18h24

8/ Scaling LLM Test-Time Compute Optimally – investigates the scaling behaviors of inference-time computation in LLMs; in particular, it analyses how much an LLM can be improved provided a fixed amount of inference-time compute; finds that the effectiveness of different scaling

RAGChecker: Fine-Grained RAG Evaluation Framework

By

–

18 August 2024 18h24

5/ RAGChecker – a fine-grained evaluation framework for diagnosing retrieval and generation modules in RAG; shows that RAGChecker has better correlations with human judgment; reports several revealing insightful patterns and trade-offs in design choices of RAG architectures.

LongWriter AgentWrite Enables LLMs Generate Beyond 20K Words

By

–

18 August 2024 18h24

3/ LongWriter – proposes AgentWrite to enable off-the-shelf LLMs to generate coherent outputs beyond 20K words; AgentWrite breaks down the long generation task into subtasks and in a divide-and-conquer approach generates…

EfficientRAG: Auto-Encoder LM for Intelligent Chunk Retrieval

By

–

18 August 2024 18h24

4/ EfficientRAG – trains an auto-encoder LM to label and tag chunks; it retrieves relevant chunks, tags them as either or , and annotates chunks for continuous processing; then a filter model is trained to formulate the next-hop query based on the

→ View original post on X — @nandodf

JAX vs PyTorch: Choosing the right ML framework

By

@nandodf

–

18 August 2024 17h55

It depends. If you have TPUs, Jax. If you have GPUs, PyTorch. It also depends on who your community is, and whether you need to distribute models or not. For simple ML, either is fine and then it comes down to popularity within your circle.