mattshumer/Reflection-Llama-3.1-70B
Full fine-tune vs meta-llama/Llama-3.1-70B-Instruct. Fine-tune: 100% of the weights changed, across 80 of 80 layers.
Tensors changed: Norms 161/161, MLP 240/240, Attention 320/320
7 derivatives checked
- bartowski/Reflection-Llama-3.1-70B-GGUF: Quantization
- legraphista/Reflection-Llama-3.1-70B-IMat-GGUF: Same weights, converted
- bullerwins/Reflection-Llama-3.1-70B-GGUF: Quantization
- osanseviero/Reflection-Llama-3.1-70B-GGUF: Quantization
- tensorblock/Reflection-Llama-3.1-70B-GGUF: Quantization
- Lucyfer1718/spartan_nemotron: Full fine-tune (fewer layers)
- Lucyfer1718/Spartan_Alpha-Powered_v2.0: Full fine-tune (fewer layers)
1 other repos hold byte-for-byte identical weight files.