reaperdoesntknow/Symbiotic-8B

Same weights, converted vs Qwen/Qwen3-8B. Same weights stored in another dtype (BF16 → F32).

Tensors changed: Embeddings 0/1, Norms 0/145, MLP 0/108, Attention 0/144, lm_head 0/1

2 derivatives checked

• mradermacher/Symbiotic-8B-i1-GGUF: Quantization
• mradermacher/Symbiotic-8B-GGUF: Same weights, converted

10 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse