Causal language models fine-tuned on DailyDialog for emotion-conditioned dialogue generation using explicit emotion tags.
Mario Rodríguez-Cantelar
mario-rc
AI & ML interests
NLP and Dialogue Generation & Management and Natural Language Understanding
Recent Activity
updated a model 9 days ago
mario-rc/emotional-rlaif-dpo-phi-3-small-8k-instruct updated a model 9 days ago
mario-rc/emotional-rlaif-ppo-phi-3-small-8k-instruct updated a model 9 days ago
mario-rc/emotional-rlaif-dpo-mistral-7b-instruct-v0.3Organizations
None yet
Multi-domain reward models for chat responses. They use 23 fine-grained regression objectives and a prompt-conditioned gating network.
-
mario-rc/multi-domain-rm-fsfairx-gemma-2-9b-it
Text Classification • 9B • Updated • 43 -
mario-rc/multi-domain-rm-fsfairx-llama-3-8b-it
Text Classification • 8B • Updated • 40 -
mario-rc/multi-domain-rm-mistral-7b-it
Text Classification • 7B • Updated • 38 -
mario-rc/multi-domain-rm-qwen-3-nemotron-8b-it
Text Classification • 8B • Updated • 38 • 1
Multilingual sequence classifiers fine-tuned for emotion recognition across seven categories in English and Spanish dialogues.
-
mario-rc/emotional-classifier-bert-base-multilingual-cased
Text Classification • 0.2B • Updated • 13 -
mario-rc/emotional-classifier-bert-base-multilingual-uncased
Text Classification • 0.2B • Updated • 7 -
mario-rc/multilingual-emotional-classifier-xlm-roberta-base
Text Classification • 0.3B • Updated • 45 -
mario-rc/multilingual-emotional-classifier-xlm-roberta-large
Text Classification • 0.6B • Updated • 27
Causal language models fine-tuned on DailyDialog for emotion-conditioned dialogue generation using explicit emotion tags.
Multi-domain reward models for chat responses. They use 23 fine-grained regression objectives and a prompt-conditioned gating network.
-
mario-rc/multi-domain-rm-fsfairx-gemma-2-9b-it
Text Classification • 9B • Updated • 43 -
mario-rc/multi-domain-rm-fsfairx-llama-3-8b-it
Text Classification • 8B • Updated • 40 -
mario-rc/multi-domain-rm-mistral-7b-it
Text Classification • 7B • Updated • 38 -
mario-rc/multi-domain-rm-qwen-3-nemotron-8b-it
Text Classification • 8B • Updated • 38 • 1
Instruction-tuned language models aligned via an emotional RLAIF workflow (PPO/DPO LoRA) for emotion-conditioned and empathetic dialogue generation.
Multilingual sequence classifiers fine-tuned for emotion recognition across seven categories in English and Spanish dialogues.
-
mario-rc/emotional-classifier-bert-base-multilingual-cased
Text Classification • 0.2B • Updated • 13 -
mario-rc/emotional-classifier-bert-base-multilingual-uncased
Text Classification • 0.2B • Updated • 7 -
mario-rc/multilingual-emotional-classifier-xlm-roberta-base
Text Classification • 0.3B • Updated • 45 -
mario-rc/multilingual-emotional-classifier-xlm-roberta-large
Text Classification • 0.6B • Updated • 27