legraphista/DeepSeek-V2-Lite-Chat-IMat-GGUF

Same weights, converted vs deepseek-ai/DeepSeek-V2-Lite-Chat. Same weights stored in another dtype (BF16 → F16/F32), plus 78 new tensors.

Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/82, MLP 0/107, Attention 0/108

On the Hub · Downloads on Model Pulse