LiquidAI/LFM2.5-350M
Full fine-tune vs LiquidAI/LFM2.5-350M-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
65 derivatives checked
- LiquidAI/LFM2.5-350M-GGUF: Same weights, converted
- LiquidAI/LFM2.5-VL-450M: Full fine-tune (fewer layers)
- itsZyn/ZynDwarf-1.1: LoRA-style
- OrdenWills/LFM2.5-350M-home-assistant-sft: LoRA-style
- mradermacher/LFM2.5-350M-GGUF: Same weights, converted
- LiquidAI/LFM2.5-350M-MLX-8bit: Quantization
- ZuzeTt/LFM2.5-VL-450M-GGUF: Full fine-tune
- mradermacher/LFM2.5-350M-i1-GGUF: Quantization
- LiquidAI/LFM2.5-350M-MLX-4bit: Quantization
- squ11z1/claude-oss-350m: Partial fine-tune
- SupraLabs/Supra-Title-350M-exp-GGUF: Full fine-tune
- Null-Guard/LFM2.5-350M-Uncensored-GGUF: Abliterated
- ipetrukha/LFM2.5-350M-MLX-4bit: Quantization
- YoussefElsafi/Aiko-350M-GGUF: Full fine-tune
- lmstudio-community/LFM2.5-350M-GGUF: Same weights, converted
- Piecrust/Spike-350M-GGUF: Partial fine-tune
- trymirai/LFM2.5-350M-L: Quantization
- DuoNeural/LFM2.5-350M-GTAP-v3-GGUF: Quantization
- trymirai/LFM2.5-350M-M: Quantization
- Z-L-Leo/my-sft-model-20260930-224915: Full fine-tune
- ipetrukha/LFM2.5-350M-MLX-8bit: Quantization
- Z-L-Leo/my-sft-model-20260930-004047: Full fine-tune
- LiquidAI/LFM2.5-350M-MLX-bf16: Same weights, repacked
- Z-L-Leo/my-sft-model-20260930-222018: Full fine-tune
- Z-L-Leo/my-sft-model-20260930-220416: Full fine-tune
- Z-L-Leo/my-sft-model-20260930-224405: Full fine-tune
- LiquidAI/LFM2.5-350M-MLX-6bit: Quantization
- BenjaminHelle/LFM2.5-350M-coder: Quantization
- lazos/lfm2.5-350m-frontend-agent: Full fine-tune
- coder3101/LFM2.5-VL-450M-heretic: Full fine-tune (fewer layers)
7 other repos hold byte-for-byte identical weight files.