Feature Extraction

Encoder and embedding models that return vector representations.

Classification follows the Hugging Face task taxonomy. See the Hugging Face Feature Extraction task page for the ecosystem-level task definition.

Model families

Model family	Declared recipes	Exact checkpoint examples	Runtime CLI
`albert`	2	`albert/albert-base-v2`	`trtmc encode`
`bert`	5	`BAAI/bge-small-en-v1.5` `google-bert/bert-base-uncased` `sentence-transformers/all-MiniLM-L6-v2` `sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2`	`trtmc encode`
`convbert`	2	`YituTech/conv-bert-base`	`trtmc encode`
`deberta`	2	`microsoft/deberta-base`	`trtmc encode`
`distilbert`	2	`distilbert/distilbert-base-uncased`	`trtmc encode`
`dpr`	2	`facebook/dpr-ctx_encoder-single-nq-base`	`trtmc encode`
`eagle_vlm`	2	`nvidia/llama-nemotron-embed-vl-1b-v2`	`trtmc embed`
`electra`	2	`google/electra-base-discriminator`	`trtmc encode`
`fnet`	2	`google/fnet-base`	`trtmc encode`
`modernbert`	2	`answerdotai/ModernBERT-base`	`trtmc encode`
`mpnet`	2	`sentence-transformers/all-mpnet-base-v2`	`trtmc encode`
`roberta`	8	`FacebookAI/roberta-base` `FacebookAI/roberta-large` `FacebookAI/xlm-roberta-base` `almanach/camembert-base`	`trtmc encode`
`xlnet`	2	`xlnet/xlnet-base-cased`	`trtmc encode`