gemma-4-31B-it
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 2
- Elith and NVIDIA release synthetic Japanese financial personas dataset
Elith and NVIDIA released Nemotron-Personas-Finance-Japan, a commercial-usable synthetic dataset with 9 million personas across 1 million records for developing and evaluating financial AI without real customer data. Built from Japanese official statistics and household finance information, it was generated with NeMo SDG-PGMs, NeMo Data Designer, and gemma-4-31B-it, and is licensed under CC BY 4.0.
Hugging Face · 🔥 0 - NVIDIA and AI4Bharat unveil 2.8T-token Indic pretraining corpus
NVIDIA and AI4Bharat at IIT Madras have introduced Nemotron-pretraining-Indic-v1, a corpus of about 2.8 trillion tokens across the 22 languages listed in India's Constitution, saying it will be released soon with an open, GPU-accelerated data generation pipeline. The corpus includes native-script data for all 22 languages and romanized text for 12, spanning 34 language-script configurations; the pipeline captures prompts, LLM workflows and quality checks so developers can adapt the recipe. NVIDIA's technical report says Nemotron 3.5 Super VL scores rose on five Indic benchmarks after an Indic continued-pretraining stage, with gains of 4.98 to 9.82 percentage points, though the blog notes that the stage mixed synthetic and natural Indic data rather than isolating synthetic data; quality assessment is mainly automated with sampled human review, and the authors say retained text can still contain errors.
Hugging Face · 🔥 0
Experience and discussion from the community
Share my gemma-4-31B-it experienceAsk about gemma-4-31B-it
Nobody has shared their experience with gemma-4-31B-it yet.