LiquidAI/LFM2.5-1.2B-Thinking
Full fine-tune vs LiquidAI/LFM2.5-1.2B-Base. Fine-tune: 100% of the weights changed, across 16 of 16 layers.
Tensors changed: Embeddings 2/2, other 30/30, MLP 48/48, Norms 44/44, Attention 24/24
44 derivatives checked
- LiquidAI/LFM2.5-1.2B-Thinking-GGUF: Same weights, converted
- unsloth/LFM2.5-1.2B-Thinking-GGUF: Same weights, converted
- LiquidAI/LFM2.5-1.2B-Thinking-MLX-4bit: Quantization
- mlx-community/LFM2.5-1.2B-Thinking-6bit: Quantization
- mradermacher/LFM2.5-1.2B-Thinking-i1-GGUF: Quantization
- Compactbot/liquidagent-1.2b: Full fine-tune
- lmstudio-community/LFM2.5-1.2B-Thinking-GGUF: Partial fine-tune
- lmstudio-community/LFM2.5-1.2B-Thinking-MLX-bf16: Same weights, repacked
- Atomic-Germ/Qwen3.8-Distilled-1.2B-NPU2: LoRA-style
- OrionLLM/GRM-3.2-Turf: LoRA-style
- mradermacher/LFM2.5-1.2B-Thinking-GGUF: Same weights, converted
- traderpedroso/LFM2.5-1.2B-CLAUDE-THINKING: LoRA-style
- LiquidAI/LFM2.5-1.2B-Thinking-MLX-8bit: Quantization
- niranjanh123/lfm-sft-hotter-colder-gemini-simplified-reasoning: Partial fine-tune
- lmstudio-community/LFM2.5-1.2B-Thinking-MLX-8bit: Quantization
- trymirai/LFM2.5-1.2B-Thinking-L: Quantization
- trymirai/LFM2.5-1.2B-Thinking-M: Quantization
- LiquidAI/LFM2.5-1.2B-Thinking-MLX-bf16: Same weights, repacked
- mlx-community/LFM2.5-1.2B-Thinking-4bit: Quantization
- enseven/lfm-2.5-think-code-GGUF: Partial fine-tune
- Nichonauta/LFM2.5-1.2B-Thinking-ToMoE: Different architecture
- LiquidAI/LFM2.5-1.2B-Thinking-MLX-6bit: Quantization
- mlx-community/LFM2.5-1.2B-Thinking-8bit: Quantization
- MuXodious/LFM2.5-1.2B-Thinking-absolute-heresy: Abliterated
- immortaltatsu/ghostai-lfm2.5-1.2b-app-v2: Full fine-tune
- Mincofficial/Sonoma-1.2B-Preview: Partial fine-tune
- z4iXsECHS6ND-heighTSAI-INSEGCSyX6T-SYS-ROD/LFM2.5-1.2B-Thinking-GGUF: Same weights, converted
- Rumiii/LiquiMedThink1.2B: LoRA-style
- 0xzknw/LFM2.5-1.2B-Thinking-Heretic-NX-Residual-Stream: Abliterated
- 0xzknw/LFM2.5-1.2B-Thinking-Heretic: Abliterated
2 other repos hold byte-for-byte identical weight files.