Results 281 to 290 of about 14,757,385 (333)
Some of the next articles are maybe not open access.
2014
Technical report (informe tecnico) of 2023 field season approved by the Consejo de Arqueologia, INAH, Mexico.
openaire +1 more source
Technical report (informe tecnico) of 2023 field season approved by the Consejo de Arqueologia, INAH, Mexico.
openaire +1 more source
arXiv.org
We introduce Qwen2.5-1M, a series of models that extend the context length to 1 million tokens. Compared to the previous 128K version, the Qwen2.5-1M series have significantly enhanced long-context capabilities through long-context pre-training and post ...
An Yang +27 more
semanticscholar +1 more source
We introduce Qwen2.5-1M, a series of models that extend the context length to 1 million tokens. Compared to the previous 128K version, the Qwen2.5-1M series have significantly enhanced long-context capabilities through long-context pre-training and post ...
An Yang +27 more
semanticscholar +1 more source
Qwen2.5-Math Technical Report: Toward Mathematical Expert Model via Self-Improvement
arXiv.orgIn this report, we present a series of math-specific large language models: Qwen2.5-Math and Qwen2.5-Math-Instruct-1.5B/7B/72B. The core innovation of the Qwen2.5 series lies in integrating the philosophy of self-improvement throughout the entire ...
An Yang +15 more
semanticscholar +1 more source
arXiv.org
We introduce the latest progress of Qwen-Audio, a large-scale audio-language model called Qwen2-Audio, which is capable of accepting various audio signal inputs and performing audio analysis or direct textual responses with regard to speech instructions.
Yunfei Chu +11 more
semanticscholar +1 more source
We introduce the latest progress of Qwen-Audio, a large-scale audio-language model called Qwen2-Audio, which is capable of accepting various audio signal inputs and performing audio analysis or direct textual responses with regard to speech instructions.
Yunfei Chu +11 more
semanticscholar +1 more source
PaddleOCR 3.0 Technical Report
arXiv.orgThis technical report introduces PaddleOCR 3.0, an Apache-licensed open-source toolkit for OCR and document parsing. To address the growing demand for document understanding in the era of large language models, PaddleOCR 3.0 presents three major ...
Cheng Cui +18 more
semanticscholar +1 more source
arXiv.org
We present Seedream 3.0, a high-performance Chinese-English bilingual image generation foundation model. We develop several technical improvements to address existing challenges in Seedream 2.0, including alignment with complicated prompts, fine-grained ...
Yu Gao +30 more
semanticscholar +1 more source
We present Seedream 3.0, a high-performance Chinese-English bilingual image generation foundation model. We develop several technical improvements to address existing challenges in Seedream 2.0, including alignment with complicated prompts, fine-grained ...
Yu Gao +30 more
semanticscholar +1 more source
arXiv.org
We present Qwen3-Omni, a single multimodal model that, for the first time, maintains state-of-the-art performance across text, image, audio, and video without any degradation relative to single-modal counterparts.
Jin Xu +37 more
semanticscholar +1 more source
We present Qwen3-Omni, a single multimodal model that, for the first time, maintains state-of-the-art performance across text, image, audio, and video without any degradation relative to single-modal counterparts.
Jin Xu +37 more
semanticscholar +1 more source
Skywork Open Reasoner 1 Technical Report
arXiv.orgThe success of DeepSeek-R1 underscores the significant role of reinforcement learning (RL) in enhancing the reasoning capabilities of large language models (LLMs). In this work, we present Skywork-OR1, an effective and scalable RL implementation for long
Jujie He +16 more
semanticscholar +1 more source
Phi-4-reasoning Technical Report
arXiv.orgWe introduce Phi-4-reasoning, a 14-billion parameter reasoning model that achieves strong performance on complex reasoning tasks. Trained via supervised fine-tuning of Phi-4 on carefully curated set of"teachable"prompts-selected for the right level of ...
Marah Abdin +22 more
semanticscholar +1 more source
Baichuan-Omni-1.5 Technical Report
arXiv.orgWe introduce Baichuan-Omni-1.5, an omni-modal model that not only has omni-modal understanding capabilities but also provides end-to-end audio generation capabilities.
Yadong Li +91 more
semanticscholar +1 more source