Loading…
Loading…
Models / mMiniLMv2-L12-H384-distilled-from-XLMR-Large
1 model (~118M) · ↓ 1.9M downloads a month · from $0.08/h on GTX 1660 SUPER right now. Each model's versions do the same job; quantized ones (AWQ, FP8, 4-bit) need less GPU memory and run on cheaper machines.
| Version | Runtime | Weights | GPU memory | Cheapest machine now |
|---|---|---|---|---|
| mmarco mMiniLMv2 L12 H384 v1 (cross-encoder)cross-encoder/mmarco-mMiniLMv2-L12-H384-v1 | Text Embeddings Inference | 240 MB | 1.8 GB | $0.08/h GTX 1660 SUPER |
| rerank indonesia (madebyaris)madebyaris/rerank-indonesia | Text Embeddings Inference | 240 MB | 1.8 GB | $0.08/h GTX 1660 SUPER |
| mmarco mMiniLMv2 L12 H384 v1 onnx cpu qint8 (temsa)temsa/mmarco-mMiniLMv2-L12-H384-v1-onnx-cpu-qint8 | ONNX Runtime | 410 MB | 1.6 GB | $0.08/h GTX 1660 SUPER |
| mmarco mMiniLMv2 L12 H384 v1 (SugoLabs)SugoLabs/mmarco-mMiniLMv2-L12-H384-v1 | ONNX Runtime | 120 MB | 1.2 GB | $0.08/h GTX 1660 SUPER |
Model details come from Hugging Face and are refreshed daily; prices follow the live offers. GPU memory is GridPup's estimate: the weights, room for an 8,192-token conversation (chat models) and the runtime's own memory. Large models are split across the GPUs of one machine ("4x RTX 4090").