TheDrummer/MS-Interleaved-Upscale-39B
Partial fine-tune (fewer layers) vs unsloth/Mistral-Small-Instruct-2409. unsloth/Mistral-Small-Instruct-2409 matches part of TheDrummer/MS-Interleaved-Upscale-39B (507 of its 894 tensors; the rest of TheDrummer/MS-Interleaved-Upscale-39B has no counterpart). On the shared part: Partial change: 63% of the weights changed (attention, mlp, norms), in 36 of 56 layers.
Tensors changed: Embeddings 0/1, Norms 72/113, MLP 108/168, Attention 144/224, lm_head 0/1
4 derivatives checked
- mradermacher/MS-Interleaved-Upscale-39B-i1-GGUF: Quantization
- TheDrummer/Skyfall-39B-v1-GGUF: Full fine-tune
- mradermacher/MS-Interleaved-Upscale-39B-GGUF: Quantization
- TheDrummer/Skyfall-39B-v1: Full fine-tune