Keural MoE 14B is an open-source language model by Hikaf143. Features: 15.4b LLM, VRAM: 30.7GB, Context: 4K, License: apache-2.0, MoE, Instruction-Based.
| LLM Name | Keural MoE 14B |
| Repository 🤗 | https://huggingface.co/Hikaf143/Keural-MoE-14B |
| Base Model(s) | |
| Model Size | 15.4b |
| Required VRAM | 30.7 GB |
| Updated | 2026-10-17 |
| Maintainer | Hikaf143 |
| Model Type | keural |
| Instruction-Based | Yes |
| Model Files | |
| Supported Languages | ko en |
| Model Architecture | KeuralMoECausalLM |
| License | apache-2.0 |
| Context Length | 4096 |
| Model Max Length | 4096 |
| Transformers Version | 4.40.0 |
| Tokenizer Class | LlamaTokenizer |
| Padding Token | <pad> |
| Vocabulary Size | 131074 |
| Torch Data Type | bfloat16 |