🦞🌯 Lobster Roll

Stories by jinqueeny

Llama-Factory: Unified, Efficient Fine-Tuning for 100 Open LLMs (github.com)
Paper2video: Automatic video generation from scientific papers (arxiv.org)
Sheet0, a data agent transform webpages to structured spreadsheets (sheet0.com)
OCaml,an industrial-strength functional programming language (ocaml.org)
The First Large Visual Memory Model (memories.ai)
Ling-1T: 1T-parameter model with 50B active parameters per token (huggingface.co)
PowerMem – Persistent memory layer for AI agents (github.com)
MemMachine, an open-source memory layer for advanced AI agents (github.com)
Ursa: A Lakehouse-Native Data Streaming Engine for Kafka (vldb.org)
LongCat-Flash, a language model with 560B total parameters, MoE architecture (github.com)
Eigent, a Multi-agent Workforce desktop application (github.com)
Alma – AI desktop app with persistent memory and tool use across AI providers (alma.now)
Show HN: CATArena – Evaluating LLM agents via dynamic enviroment interactions (github.com)
Act-1: A Robot Foundation Model Trained on Zero Robot Data (sunday.ai)
Ten Years Building TiDB (medium.com)
TiDB X: Context-aware scaling for distributed SQL with object storage backbone (pingcap.com)
BANG: Dividing 3D Assets via Generative Exploded Dynamics (arxiv.org)
Fello, an agentic browser that acts (fellou.ai)
The open source AI coding agent (opencode.ai)
Boxlite, SQLite-like agent runtime (github.com)
Stop benchmarking LLMs. Make them fight (github.com)
Alma, Elegant AI Provider Orchestration (alma.now)
WeKnora – LLM-Powered Document Understanding and Retrieval Framework (github.com)
LLaDA2.0: Scaling Up Diffusion Language Models to 100B [pdf] (github.com)
Acontext, Turn Your Agent's Task History into Reusable Skills (SOPs) (github.com)
Optimizing RocksDB in TiKV (Part 1) – The Battle Against the DB Mutex (medium.com)
Datus-agent, Claude-Code-like CLI for data engineers (github.com)
TLDW turns long-form YouTube videos into a structured learning workspace (github.com)
Example skills that demonstrate what's possible with Claude's skills system (github.com)
Ring-1T: Trillion-Parameter Model Trained with RLVR and RLHF (ant-ling.medium.com)