元宝-多模态大模型算法研究员
Tencent
元宝-多模态大模型算法研究员
Job Description
元宝-多模态大模型算法研究员
📋 Job Overview
The role involves participating in the full-cycle R&D of audio large models, focusing on cross-modal alignment, multimodal understanding, and generation across text and speech. Responsibilities include data cleaning and preparation, model algorithm selection and optimization, and advancements in pre-training, supervised fine-tuning, and reinforcement learning. The position also entails end-to-end optimization of speech dialogue models for challenging scenarios and exploring new paradigms in multimodal model understanding and generation.
📍 Location: Beijing, China
🏢 Business Unit: CSIG
📄 Full Description
2. 负责语音对话大模型的端到端效果优化，提高在远场、低信噪比、多人、音乐等场景下的理解及生成效果，改善模型在方言、副语言信息等方面的理解能力，加强情感对话能力；
3. 探索多模态模型的理解和生成范式，跟进业界新的多模态大模型结构，从模型效果优化及降低全链路处理延时等多个方面开展前沿性研究及落地工作。
🎯 Key Responsibilities
- Participate in the full-cycle R&D of audio large models, including cross-modal alignment, multimodal understanding and generation, covering data cleaning and production for text and speech training, foundation model algorithm selection and optimization, focusing on technical iterations in pre-training, supervised fine-tuning, and reinforcement learning.
- Responsible for end-to-end effect optimization of speech dialogue large models, improving understanding and generation in far-field, low SNR, multi-person, music scenarios, enhancing model capabilities in dialects, paralinguistic information, and strengthening emotional dialogue abilities.
- Explore understanding and generation paradigms for multimodal models, follow industry advancements in new multimodal large model architectures, conduct cutting-edge research and implementation in model effect optimization, reducing full-chain processing latency, and other aspects.
Locations
- Beijing, China
Salary
400,000 - 800,000 CNY / yearly
Source: ai estimated
* This is an estimated range based on market data and may vary based on experience and qualifications.
Responsibilities
- Participate in the full-cycle R&D of audio large models, including cross-modal alignment, multimodal understanding and generation, covering data cleaning and production for text and speech training, foundation model algorithm selection and optimization, focusing on technical iterations in pre-training, supervised fine-tuning, and reinforcement learning.
- Responsible for end-to-end effect optimization of speech dialogue large models, improving understanding and generation in far-field, low SNR, multi-person, music scenarios, enhancing model capabilities in dialects, paralinguistic information, and strengthening emotional dialogue abilities.
- Explore understanding and generation paradigms for multimodal models, follow industry advancements in new multimodal large model architectures, conduct cutting-edge research and implementation in model effect optimization, reducing full-chain processing latency, and other aspects.
Tags & Categories
Answer 10 quick questions to check your fit for 元宝-多模态大模型算法研究员 @ Tencent.
Related Books and Jobs
元宝-多模态大模型算法研究员
Tencent
元宝-多模态大模型算法研究员
Job Description
元宝-多模态大模型算法研究员
📋 Job Overview
The role involves participating in the full-cycle R&D of audio large models, focusing on cross-modal alignment, multimodal understanding, and generation across text and speech. Responsibilities include data cleaning and preparation, model algorithm selection and optimization, and advancements in pre-training, supervised fine-tuning, and reinforcement learning. The position also entails end-to-end optimization of speech dialogue models for challenging scenarios and exploring new paradigms in multimodal model understanding and generation.
📍 Location: Beijing, China
🏢 Business Unit: CSIG
📄 Full Description
2. 负责语音对话大模型的端到端效果优化，提高在远场、低信噪比、多人、音乐等场景下的理解及生成效果，改善模型在方言、副语言信息等方面的理解能力，加强情感对话能力；
3. 探索多模态模型的理解和生成范式，跟进业界新的多模态大模型结构，从模型效果优化及降低全链路处理延时等多个方面开展前沿性研究及落地工作。
🎯 Key Responsibilities
- Participate in the full-cycle R&D of audio large models, including cross-modal alignment, multimodal understanding and generation, covering data cleaning and production for text and speech training, foundation model algorithm selection and optimization, focusing on technical iterations in pre-training, supervised fine-tuning, and reinforcement learning.
- Responsible for end-to-end effect optimization of speech dialogue large models, improving understanding and generation in far-field, low SNR, multi-person, music scenarios, enhancing model capabilities in dialects, paralinguistic information, and strengthening emotional dialogue abilities.
- Explore understanding and generation paradigms for multimodal models, follow industry advancements in new multimodal large model architectures, conduct cutting-edge research and implementation in model effect optimization, reducing full-chain processing latency, and other aspects.
Locations
- Beijing, China
Salary
400,000 - 800,000 CNY / yearly
Source: ai estimated
* This is an estimated range based on market data and may vary based on experience and qualifications.
Responsibilities
- Participate in the full-cycle R&D of audio large models, including cross-modal alignment, multimodal understanding and generation, covering data cleaning and production for text and speech training, foundation model algorithm selection and optimization, focusing on technical iterations in pre-training, supervised fine-tuning, and reinforcement learning.
- Responsible for end-to-end effect optimization of speech dialogue large models, improving understanding and generation in far-field, low SNR, multi-person, music scenarios, enhancing model capabilities in dialects, paralinguistic information, and strengthening emotional dialogue abilities.
- Explore understanding and generation paradigms for multimodal models, follow industry advancements in new multimodal large model architectures, conduct cutting-edge research and implementation in model effect optimization, reducing full-chain processing latency, and other aspects.
Tags & Categories
Answer 10 quick questions to check your fit for 元宝-多模态大模型算法研究员 @ Tencent.
Related Books and Jobs