https://github.com/OpenMOSS/FRoM-W1
AI & ML interests
LLM
Recent Activity
View all activity
Papers
Beyond Real: Imaginary Extension of Rotary Position Embeddings for Long-Context LLMs
SRPO: Self-Referential Policy Optimization for Vision-Language-Action Models
Proactive Robot Manipulation in Omni-modal Context
Open source weights of Lorsa modules introduced in "Towards Understanding the Nature of Attention with Low-Rank Sparse Decomposition".
The MHA2MLA model published in the paper "Towards Economical Inference: Enabling DeepSeek's Multi-Head Latent Attention in Any Transformer-Based LLMs"
-
Towards Economical Inference: Enabling DeepSeek's Multi-Head Latent Attention in Any Transformer-based LLMs
Paper β’ 2502.14837 β’ Published β’ 3 -
OpenMOSS-Team/Llama-2-7B-MLA-d_kv_16
Text Generation β’ 6B β’ Updated β’ 14 -
OpenMOSS-Team/Llama-2-7B-MLA-d_kv_32
Text Generation β’ 6B β’ Updated β’ 13 -
OpenMOSS-Team/Llama-2-7B-MLA-d_kv_64
Text Generation β’ 7B β’ Updated β’ 11
An Efficient Training Framework for Diffusion Language Models
True Speech-to-Speech Langugage Model
-
OpenMOSS-Team/Embodied_R1-ScienceWorld
8B β’ Updated β’ 6 -
OpenMOSS-Team/Embodied_Planner-R1-Alfworld
8B β’ Updated β’ 7 -
Unleashing Embodied Task Planning Ability in LLMs via Reinforcement Learning
Paper β’ 2506.23127 β’ Published β’ 1 -
World Modeling Makes a Better Planner: Dual Preference Optimization for Embodied Task Planning
Paper β’ 2503.10480 β’ Published β’ 55
The MHA2MLA model published in the paper "Towards Economical Inference: Enabling DeepSeek's Multi-Head Latent Attention in Any Transformer-Based LLMs"
-
OpenMOSS-Team/SmolLM-135M-MLA-d_kv_8-refactor
Text Generation β’ 0.1B β’ Updated β’ 20 -
OpenMOSS-Team/SmolLM-135M-MLA-d_kv_32-refactor
Text Generation β’ 0.1B β’ Updated β’ 14 -
OpenMOSS-Team/SmolLM-135M-MLA-d_kv_16-refactor
Text Generation β’ 0.1B β’ Updated β’ 9 -
OpenMOSS-Team/SmolLM-360M-MLA-d_kv_8-refactor
Text Generation β’ 0.3B β’ Updated β’ 7
-
OpenMOSS-Team/moss-moon-003-sft-plugin
Text Generation β’ Updated β’ 35 β’ 69 -
OpenMOSS-Team/moss-moon-003-sft
Text Generation β’ Updated β’ 99 β’ 127 -
OpenMOSS-Team/moss-moon-003-base
Text Generation β’ Updated β’ 226 β’ 131 -
OpenMOSS-Team/moss-moon-003-sft-int4
Text Generation β’ Updated β’ 55 β’ 40
https://github.com/OpenMOSS/FRoM-W1
An Efficient Training Framework for Diffusion Language Models
Proactive Robot Manipulation in Omni-modal Context
True Speech-to-Speech Langugage Model
-
OpenMOSS-Team/Embodied_R1-ScienceWorld
8B β’ Updated β’ 6 -
OpenMOSS-Team/Embodied_Planner-R1-Alfworld
8B β’ Updated β’ 7 -
Unleashing Embodied Task Planning Ability in LLMs via Reinforcement Learning
Paper β’ 2506.23127 β’ Published β’ 1 -
World Modeling Makes a Better Planner: Dual Preference Optimization for Embodied Task Planning
Paper β’ 2503.10480 β’ Published β’ 55
Open source weights of Lorsa modules introduced in "Towards Understanding the Nature of Attention with Low-Rank Sparse Decomposition".
The MHA2MLA model published in the paper "Towards Economical Inference: Enabling DeepSeek's Multi-Head Latent Attention in Any Transformer-Based LLMs"
-
OpenMOSS-Team/SmolLM-135M-MLA-d_kv_8-refactor
Text Generation β’ 0.1B β’ Updated β’ 20 -
OpenMOSS-Team/SmolLM-135M-MLA-d_kv_32-refactor
Text Generation β’ 0.1B β’ Updated β’ 14 -
OpenMOSS-Team/SmolLM-135M-MLA-d_kv_16-refactor
Text Generation β’ 0.1B β’ Updated β’ 9 -
OpenMOSS-Team/SmolLM-360M-MLA-d_kv_8-refactor
Text Generation β’ 0.3B β’ Updated β’ 7
The MHA2MLA model published in the paper "Towards Economical Inference: Enabling DeepSeek's Multi-Head Latent Attention in Any Transformer-Based LLMs"
-
Towards Economical Inference: Enabling DeepSeek's Multi-Head Latent Attention in Any Transformer-based LLMs
Paper β’ 2502.14837 β’ Published β’ 3 -
OpenMOSS-Team/Llama-2-7B-MLA-d_kv_16
Text Generation β’ 6B β’ Updated β’ 14 -
OpenMOSS-Team/Llama-2-7B-MLA-d_kv_32
Text Generation β’ 6B β’ Updated β’ 13 -
OpenMOSS-Team/Llama-2-7B-MLA-d_kv_64
Text Generation β’ 7B β’ Updated β’ 11
-
OpenMOSS-Team/moss-moon-003-sft-plugin
Text Generation β’ Updated β’ 35 β’ 69 -
OpenMOSS-Team/moss-moon-003-sft
Text Generation β’ Updated β’ 99 β’ 127 -
OpenMOSS-Team/moss-moon-003-base
Text Generation β’ Updated β’ 226 β’ 131 -
OpenMOSS-Team/moss-moon-003-sft-int4
Text Generation β’ Updated β’ 55 β’ 40