OctoThinker/OctoThinker-8B-Short-Base
Different architecture vs meta-llama/Llama-3.2-3B. Different architecture: only 0 of 254/291 tensors match by name and shape.
4 derivatives checked
- mradermacher/OctoThinker-8B-Short-Base-i1-GGUF: Quantization
- mradermacher/OctoThinker-8B-Short-Base-GGUF: Same weights, converted
- mehuldamani/RLCR-hotpot-octo: Partial fine-tune
- mehuldamani/RLVR-hotpot-octo: Full fine-tune
1 other repos hold byte-for-byte identical weight files.