HuggingFaceTB/SmolLM2-135M-Instruct
Full fine-tune vs HuggingFaceTB/SmolLM2-135M. Fine-tune: 100% of the weights changed, across 30 of 30 layers.
Tensors changed: Embeddings 1/1, Norms 61/61, MLP 90/90, Attention 120/120
73 derivatives checked
- HuggingFaceTB/SmolVLM-256M-Instruct: Full fine-tune (fewer layers)
- unsloth/SmolLM2-135M-Instruct-GGUF: Same weights, converted
- bartowski/SmolLM2-135M-Instruct-GGUF: Same weights, converted
- unsloth/SmolLM2-135M-Instruct: Same weights, repacked
- mradermacher/SmolLM2-135M-Instruct-GGUF: Quantization
- mlx-community/SmolLM2-135M-Instruct: Same weights, repacked
- jc-builds/SmolLM2-135M-Instruct-Q4_K_M-GGUF: Quantization
- lmstudio-community/SmolLM2-135M-Instruct-GGUF: Quantization
- mlx-community/SmolVLM-256M-Instruct-4bit: Quantization
- mradermacher/SmolLM2-135M-Instruct-i1-GGUF: Quantization
- QuantFactory/SmolLM2-Prompt-Enhance-GGUF: Full fine-tune
- mrbhazied/tuned-SmolLM2-135M-Instruct-test-s: Full fine-tune
- Miladsaeedi70/smollm2-135m-scientific-sft-lora: Adapter
- saidutta69/SmolLM2-135M-Instruct-heretic: Abliterated
- latishab/turnsense: LoRA-style
- SirSahOl/SmolLM2-135M-Instruct-chat-mlx-16bit: Same weights, repacked
- ykumards/smollm2-135m-bedtime-stories: Full fine-tune
- Mungert/SmolVLM-256M-Instruct-GGUF: Full fine-tune
- MaziyarPanahi/SmolLM2-135M-Instruct-GGUF: Same weights, converted
- 1000kavya/SmolLM2-135M-Alpaca-SFT: Full fine-tune
- jekunz/smollm-135m-lora-fineweb-faroese: Adapter
- styal/SmolLM2-135M-layertrop-topk-k3: Full fine-tune
- axana-labs/SmolLM2-135M-Instruct-GGUF: Same weights, converted
- prithivMLmods/SmolLM2-135M-F32-GGUF: Same weights, converted
- pavelklymenko/smollm2-135m-shakespeare: Full fine-tune
- Felladrin/gguf-Q8_0-SmolLM2-135M-Instruct: Quantization
- HackNetAyush/smollm2-135M-instruct-gguf-q8: Quantization
- unileon-robotics/SmolLM2-135M-Instruct-BehaviorTree-LoRA-GGUF: Not comparable yet
- leonidas123/smollm2-135m-math-lora-v1-merged: LoRA-style
- Wrooms/SMAXI-Community-135M: Full fine-tune
15 other repos hold byte-for-byte identical weight files.