Hugging Face · Schema
Info
Properties
| Name | Type | Description |
|---|---|---|
| model_id | string | Loaded model ID |
| model_sha | string | Model revision SHA |
| model_dtype | string | Model data type |
| model_device_type | string | Device type (cuda, cpu) |
| model_pipeline_tag | string | Pipeline task tag |
| max_concurrent_requests | integer | |
| max_best_of | integer | |
| max_stop_sequences | integer | |
| max_input_length | integer | Maximum input token length |
| max_total_tokens | integer | Maximum total tokens (input + output) |
| waiting_served_ratio | number | |
| max_batch_total_tokens | integer | |
| max_waiting_tokens | integer | |
| validation_workers | integer | |
| max_client_batch_size | integer | |
| version | string | TGI version |
| sha | string | TGI build SHA |
| docker_label | string | Docker image label |