Davidsv/SUONG-1

Full fine-tune (fewer layers) vs OpenPipe/mistral-ft-optimized-1218. Davidsv/SUONG-1 matches part of OpenPipe/mistral-ft-optimized-1218 (111 of its 291 tensors; the rest of OpenPipe/mistral-ft-optimized-1218 has no counterpart). On the shared part: Fine-tune: 99% of the weights changed, across 12 of 12 layers.

Tensors changed: Embeddings 1/1, Norms 13/25, MLP 36/36, Attention 44/48, lm_head 1/1

1 derivatives checked

• mradermacher/SUONG-1-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse