VLA-Adapter An Effective Paradigm for Tiny-Scale Vision-Language-Action Model
@_akhaliq
-
EchoX: Reducing Acoustic-Semantic Gap in Speech-to-Speech LLMs
By
–
EchoX
— AK (@_akhaliq) 12 septembre 2025
Towards Mitigating Acoustic-Semantic Gap via Echo Training for Speech-to-Speech LLMs pic.twitter.com/iEkoPnJn13EchoX Towards Mitigating Acoustic-Semantic Gap via Echo Training for Speech-to-Speech LLMs
-
Meta Releases MobileLLM-R1: Compact Edge Reasoning Model Under 1B Parameters
By
–
Meta just dropped MobileLLM-R1 on Hugging Face a edge reasoning model with fewer than 1B parameters 2×–5× Performance Boost over other fully open-source models: MobileLLM-R1 achieves ~5× higher MATH accuracy vs. Olmo-1.24B, and ~2× vs. SmolLM2-1.7B. Uses just 1/10 the
-
AnyCoder Tool Now Available on Hugging Face Spaces
By
–
Congrats both now available in anycoder: https://
huggingface.co/spaces/akhaliq
/anycoder
… -
Qwen3-Next-80B-A3B: Efficient MoE Model with Superior Performance
By
–
Qwen3-Next-80B-A3B is out 80B params, but only 3B activated per token → 10x cheaper training, 10x faster inference than Qwen3-32B.(esp. @ 32K+ context!) Qwen3-Next-80B-A3B-Instruct approaches our 235B flagship. Qwen3-Next-80B-A3B-Thinking outperforms
-
Majority Not Always Right: RL Training for Solution Aggregation
By
–
The Majority is not always right RL training for solution aggregation
-
Development of a Conversational Multilingual TTS Application
By
–
Vibe coding a chatterbox multilingual TTS app in anycoder
-
K2-Think: 32B Model Achieving Frontier Reasoning Performance
By
–
K2-Think a reasoning system that achieves frontier performance with just a 32B parameter model, surpassing or matching much larger models such as GPT-OSS 120B and DeepSeek v3.1 vibe coded a chat app for it in anycoder
-
Revolutionary Reinforcement Learning Framework for Diffusion Language Models
By
–
Revolutionizing Reinforcement Learning Framework for Diffusion Large Language Models
