perplexity-ai

/

pplx-embed-v1-0.6b

like 205

Follow

Perplexity 2.58k

Runs improved batching and compile:

model_metadata:
  example_model_input:
    encoding_format: float
    input: text string
    model: model
model_name: BEI-Bert-pplx-embed
resources:
  accelerator: H100
  cpu: '1'
  memory: 10Gi
  use_gpu: true
trt_llm:
  build:
    base_model: encoder_bert
    checkpoint_repository:
      repo: perplexity-ai/pplx-embed-v1-0.6b
      revision: refs/pr/11
      source: HF
    max_num_tokens: 32768 
  runtime:
    webserver_default_route: /v1/embeddings