Loading…
Loading…
Models / mmbert-32k-yarn
GPU memory needed (smallest version of each): mmbert-32k-yarn ~307M about 2.2 GB. The cheapest runs on RTX 3060 from $0.10/h right now.
1 model (~307M) · ↓ 46.9K downloads a month · from $0.10/h on RTX 3060 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 |
|---|---|---|---|---|
| mmbert embed 32k 2d matryoshkavllm-sr/mmbert-embed-32k-2d-matryoshka | vLLM | 610 MB | 2.2 GB | $0.10/h RTX 3060 |
| mmbert32k factcheck classifier mergedvllm-sr/mmbert32k-factcheck-classifier-merged | vLLM | 620 MB | 2.2 GB | $0.10/h RTX 3060 |
| mmbert32k jailbreak detector mergedvllm-sr/mmbert32k-jailbreak-detector-merged | vLLM | 620 MB | 2.2 GB | $0.10/h RTX 3060 |
| mmbert32k intent classifier mergedvllm-sr/mmbert32k-intent-classifier-merged | vLLM | 620 MB | 2.2 GB | $0.10/h RTX 3060 |
| mmbert32k feedback detector mergedvllm-sr/mmbert32k-feedback-detector-merged | vLLM | 620 MB | 2.2 GB | $0.10/h RTX 3060 |
| mmbert32k modality router mergedvllm-sr/mmbert32k-modality-router-merged | vLLM | 620 MB | 2.2 GB | $0.10/h RTX 3060 |
| elephant embeddings v1 text smallvllm-sr/elephant-embeddings-v1-text-small | vLLM | 610 MB | 2.2 GB | $0.10/h RTX 3060 |
| mmbert embed financevllm-sr/mmbert-embed-finance | vLLM | 610 MB | 2.2 GB | $0.10/h RTX 3060 |
| mmbert embed medicalvllm-sr/mmbert-embed-medical | vLLM | 610 MB | 2.2 GB | $0.10/h RTX 3060 |
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").