yuniaXian
AI researcher (AI, NLP) @americanexpress PhD @ubsuny, #NLP, #LLM, #Gen-AI, #Recommender, #Risk @facebook @amzn.
American Express @americanexpressNew York City
Pinned Repositories
brain_activities_classification
(Kraggle) HMS - Harmful Brain Activity Classification
fairseq_kg
Implement of Knowledge graph to text model. Integrated with Fairseq (Meta Fair research library))
llm_langchain_projects
Collection of llm_langchain_projects: Autolabelling, Search and Indexing
llm_peft
LLM PEFT finetuning scripts for llama2 and Chatglm
multimodal_model_training
Training codes of multimodal components: clip, ViT, detr
ppo_llm_DeepSpeed
Customized llm PPO (reinforcement learning) pipeline with deepSpeed. For Amex external usage. Training reward model, actor-critic models with referenced supervised fine-tuned model
rasa_calm_chatbot
💬 Customized rasa chatbot framework based on llm to automate text- and voice-based conversations
yuniaXian.github.io
personal page
yuniaXian's Repositories
yuniaXian/llm_langchain_projects
Collection of llm_langchain_projects: Autolabelling, Search and Indexing
yuniaXian/fairseq_kg
Implement of Knowledge graph to text model. Integrated with Fairseq (Meta Fair research library))
yuniaXian/multimodal_model_training
Training codes of multimodal components: clip, ViT, detr
yuniaXian/llm_peft
LLM PEFT finetuning scripts for llama2 and Chatglm
yuniaXian/ppo_llm_DeepSpeed
Customized llm PPO (reinforcement learning) pipeline with deepSpeed. For Amex external usage. Training reward model, actor-critic models with referenced supervised fine-tuned model
yuniaXian/rasa_calm_chatbot
💬 Customized rasa chatbot framework based on llm to automate text- and voice-based conversations
yuniaXian/brain_activities_classification
(Kraggle) HMS - Harmful Brain Activity Classification
yuniaXian/yuniaXian.github.io
personal page