Emperorizzis/ASTRA-32B-Thinking-v1
Full fine-tune vs Qwen/Qwen3-32B. Fine-tune: 100% of the weights changed, across 64 of 64 layers.
Tensors changed: Embeddings 1/1, Norms 206/257, MLP 192/192, Attention 256/256, lm_head 1/1
3 derivatives checked
- mradermacher/ASTRA-32B-Thinking-v1-i1-GGUF: Quantization
- m-i/ASTRA-32B-Thinking-v1-mlx-5Bit: Quantization
- mradermacher/ASTRA-32B-Thinking-v1-GGUF: Quantization