Model comparison
RoBERTa Large vs StarCoder2 15B
Capability radar, value bars, and side-by-side posture.
RoBERTa Large
Not scored
value score
StarCoder2 15B
70
value score
RoBERTa Large ctx
—
StarCoder2 15B ctx
16K
Capability radar
Value · context · multimodal · openness · speed posture
Value head-to-head
- StarCoder2 15B70
Attribute tape
| Field | RoBERTa Large | StarCoder2 15B |
|---|---|---|
| Developer | Hugging Face | Hugging Face |
| Context | — | 16K |
| Modalities | Fill-mask | Text, Code |
| Openness | — | Open Weights |
| Speed | — | — |
| Price | — | — |
| Value score | — | 70 |
| Summary | RoBERTa Large is a 355M-parameter English masked-language model pretrained by Meta's FacebookAI lab, designed as a robust text encoder for downstream NLP tasks such as classification, named entity recognition, and question answering. It improves on BERT-style pretraining through dynamic masking, removal of the next-sentence-prediction objective, larger training data, and larger batch sizes. | BigCode open coding model on Hugging Face. |