🚀 Introducing Spatial-SSRL, the first study which proposes a Self-Supervised Reinforcement Learning paradigm for spatial understanding. 💡 Spatial-SSRL a lightweight tool-free framework that aims at enhancing spatial intelligence and is natually compatible with the RLVR training

🚀Introducing #CapRL, the first study of applying GRPO for the open-ended and subjective image captioning task. 🤯 🤖The trained CapRL-3B model achieves image captioning performance comparable to Qwen2.5-VL-72B. ✨CapRL introduces a novel training framework that redefines caption

🚀 Big news for #lmdeploy v0.10.1! 🥳Our #FP8 high-performance inference is no longer limited to the latest #GPUs. It now supports all #NVIDIA architectures from V100 onwards, bringing major speedups to more users. 🤗 https://t.co/bPJfr9rz5p

Introducing #DLCompiler and #DLBlas. Developers can achieve performance close to the hardware peak without manual tuning. And for the first time, Triton OP achieves extreme performance optimization on DSA chips. DLCompiler: https://t.co/xZKZh98NOf DLBlas： https://t.co/FSRx2EHghL

🔥LMDeploy v0.10.0 released! 😊Supercharges OpenAI’s GPT-OSS MXFP4 models. 😊Delivers exceptional performance for GPT-OSS models on V100 and higher GPUs. 😊On H800 & A100, LMDeploy outperforms vLLM across all scenarios—faster, more efficient inference! 🤗 https://t.co/bPJfr9rz5p

🔥Introducing Intern-S1-mini, a lightweight open-source multimodal reasoning model based on the same techniques as Intern-S1. 🥳With just 8B parameters, it’s optimized for fast deployment and easy customization. - Strong general capabilities while excelling in specialized

🔥🔥🔥 😊Try our #InternS1😋 😉Model: https://t.co/qlUYXkrNCn 😉GitHub: https://t.co/yEUApYSGTH 😉Chat here:

Our paper won an outstanding paper on ACL 2025. Try our best open-source multimodal reasoning model Intern-S1 at https://t.co/qlUYXkrNCn. This 241B MoE model combines strong general-task capabilities with state-of-the-art performance on a wide range of scientific tasks,

Try our new multimodal reasoning model #InternS1😋 Model: https://t.co/qlUYXkrNCn GitHub: https://t.co/yEUApYSGTH Chat here: https://t.co/FBGtILjuiu

🚀Introducing Intern-S1, our most advanced open-source multimodal reasoning model yet! 🥳Strong general-task capabilities + SOTA performance on scientific tasks, rivaling leading closed-source commercial models. 🥰Built upon a 235B MoE language model and a 6B Vision encoder.

🚀 Introducing #POLAR: Bring Reward Model into a New Pre-training Era! ✨ Say goodbye to reward models with poor generalization! POLAR (Policy Discriminative Learning) is a groundbreaking pre-training paradigm that trains reward models to distinguish policy distributions,

🥳Trained through #InternBootcamp, #InternThinker now combines pro-level Go skills with transparent reasoning. 😉In each game, it acts as a patient, insightful coach—analyzing the board, comparing moves, and clearly explaining each decision. 🤗Try it now: https://t.co/dMJEbOui5q

🥳Introducing #InternBootcamp, an easy-to-use and extensible library for training large reasoning models. Unlimited automatic question generation and result verification. Over 1,000 verifiable tasks covering logic, puzzles, algorithms, games, and more. 🤗 https://t.co/uLtrAghjUz

🥳Thrill to release the full RL training code of #OREAL! 😊Now you can fully reproduce the results of OREAL-7B/32B. Using #DeepSeek-R1-Distill-Qwen-32B, you can further obtain a model has 95.6 on MATH-500! 🤗Code: https://t.co/64qos7qln3 🤗Based on:

🥳Introducing #OREAL, a new RL method for math reasoning. 😊With OREAL, a 7B model achieves 94.0 pass@1 on MATH-500, matching many 32B models, while OREAL-32B achieves 95.0 pass@1, surpassing #DeepSeek-R1 Distilled models. 🤗Paper/Model/Data: https://t.co/IVD72J8BHN

🥳With PowerServe, InternLM3-8B-Instruct runs on Android devices equipped with Qualcomm NPUs. 😉Feel free to try out our new model at https://t.co/vuH5Y1mDD2

📢 We've just integrated InternLM's models into the 🐫 CAMEL-AI framework! InternLM is a multilingual, multi-billion-parameter base model trained on trillions of tokens, and we now support their whole family of models. 🚀 Newly Supported Models: ✅ internlm3-latest ✅

Now you can run InternLM3-8B in Qualcomm NPU with PowerServe!

🥳InternLM3-8B-Instruct takes on the easy problem from LeetCode Weekly Contest 431. 😉Leetcode Weekly Contest 431： https://t.co/1OhZdOW5j2 😉Feel free to try out our new model at https://t.co/m7aLtxcxMX

🥳Here are some popular examples from the InternLM3-8B-Instruct model. 😉Feel free to try out our new model at https://t.co/m7aLtxcxMX.