BSC-LT/salamandra-7b-instruct
Full fine-tune vs BSC-LT/salamandra-7b. Fine-tune: 100% of the weights changed, across 32 of 32 layers.
Tensors changed: Embeddings 1/1, Norms 65/65, MLP 96/96, Attention 128/128, lm_head 1/1
14 derivatives checked
- RichardErkhov/BSC-LT_-_salamandra-7b-instruct-gguf: Full fine-tune
- hdnh2006/BSC-LT-salamandra-7b-instruct-gguf: Same weights, converted
- cstr/salamandra-7b-instruct-GGUF: Full fine-tune
- Fr4nkSt3in/salamandra-7b-instruct-2606-BF16-GGUF: Full fine-tune
- tensorblock/salamandra-7b-instruct-GGUF: Quantization
- Securelayer7/salamandra-7b-instruct-Uncensored-Abliterated: Abliterated
- BSC-LT/salamandraTA-7B-academic: Partial fine-tune
- BSC-LT/salamandra-7b-instruct-gptq: Quantization
- BSC-LT/speech-salamandra-es-en: Partial fine-tune (fewer layers)
- BSC-LT/salamandra-7b-instruct-guard: Identical copy
- projecte-aina/salamandra-7b-aligned-EADOP: Full fine-tune
- proxectonos/Carballo-Legal: Partial fine-tune
- proxectonos/Carvalho-Salamandra-Instruct: Partial fine-tune
- alberalm/salamandra-7b-instruct-description: LoRA-style
1 other repos hold byte-for-byte identical weight files.