Nllb 3.3B Salt Lr2e 4 Fp16 is an open-source language model by jq. Features: 3.3b LLM, VRAM: 13.4GB, Context: 1K, Quantized, LLM Explorer Score: 0.17.
| LLM Name | Nllb 3.3B Salt Lr2e 4 Fp16 |
| Repository 🤗 | https://huggingface.co/jq/nllb-3.3b-salt-lr2e-4-fp16 |
| Model Size | 3.3b |
| Required VRAM | 13.4 GB |
| Updated | 2026-08-04 |
| Maintainer | jq |
| Model Type | m2m_100 |
| Model Files | |
| Quantization Type | fp16 |
| Model Architecture | TrainableM2MForConditionalGeneration |
| Context Length | 1024 |
| Model Max Length | 1024 |
| Transformers Version | 4.50.3 |
| Tokenizer Class | NllbTokenizer |
| Padding Token | <pad> |
| Vocabulary Size | 256206 |
| Torch Data Type | float32 |
| Activation Function | relu |