Jab1718/qwen3.8-flash-coder-44gb-selective-int8

Partial fine-tune vs Jab1718/qwen3.8-flash-coder-85gb-bf16. Partial change: 89% of the weights changed (attention, mlp), in 48 of 48 layers.

Tensors changed: Embeddings 0/1, Attention 16/492, Norms 0/181, MLP 96/480, lm_head 0/1, other 0/2

On the Hub · Downloads on Model Pulse