ibm-granite/granite-3.3-2b-instruct
Full fine-tune vs ibm-granite/granite-3.3-2b-base. Fine-tune: 100% of the weights changed, across 40 of 40 layers.
Tensors changed: Norms 81/81, MLP 120/120, Attention 160/160
24 derivatives checked
- ibm-granite/granite-speech-3.3-2b: Same weights, repacked (fewer layers)
- ibm-granite/granite-3.3-2b-instruct-GGUF: Same weights, converted
- bartowski/ibm-granite_granite-3.3-2b-instruct-GGUF: Same weights, converted
- mradermacher/granite-3.3-2b-instruct-i1-GGUF: Quantization
- SamsungSDS-Research/SGuard-ContentFilter-2B-v1: Full fine-tune
- DevQuasar-12/ibm-granite.granite-3.3-2b-instruct-GGUF: Same weights, converted
- arzaan789/granite-3.3-2b-uncensored: Full fine-tune
- mlx-community/granite-3.3-2b-instruct-4bit: Quantization
- SamsungSDS-Research/SGuard-JailbreakFilter-2B-v1: Full fine-tune
- mradermacher/granite-3.3-2b-instruct-GGUF: Same weights, converted
- drstupidity/granite-3.3-2b-customer-support: LoRA-style
- webAI-Official/TwIL-LM2: LoRA-style
- unsloth/granite-3.3-2b-instruct: Identical copy
- lmstudio-community/granite-3.3-2b-instruct-GGUF: Quantization
- logan7000/llm-math345-gt-granite2b-endpoint: Full fine-tune
- tensorblock/ibm-granite_granite-3.3-2b-instruct-GGUF: Quantization
- logan7000/cogrpo-n3-ring-qwen25-3b-x-llama32-3b-x-granite2b-math345-groupC-granite2b-end: Full fine-tune
- MihaiPopa-1/granite-3.3-2b-instruct-heretic-safety-defiltered: Abliterated
- agentlans/granite-3.3-2b-instruct-critical-thinking: LoRA-style
- agentlans/granite-3.3-2b-instruct-ethics: LoRA-style
- agentlans/granite-3.3-2b-instruct-literary-writer-v0.1: LoRA-style
- powermove72/Granite-3.3-2B-Avg-SliceWeighted: Full fine-tune
- powermove72/Granite-3.3-2B-Slerp: Full fine-tune
- powermove72/Granite-3.3-2B-Avg: Full fine-tune
1 other repos hold byte-for-byte identical weight files.