[2025-10-02 23:42:16] Finish: obj=GenerateReqInput(video_data=None, rid='5799660f79fd4dc481ac2226d5b97102', return_logprob=False, logprob_start_len=-1, top_logprobs_num=0, token_ids_logprob=None, return_text_in_logprobs=True, stream=True, log_metrics=True, return_hidden_states=False, modalities=[], session_params=None, lora_id=None, custom_logit_processor=None, bootstrap_host=None, bootstrap_port=None, bootstrap_room=None, bootstrap_pair_key=None, data_parallel_rank=None, background=False, conversation_id=None, priority=None, extra_key=None, no_logs=False, custom_labels=None, label=None, return_bytes=False), out={'meta_info': {'id': '5799660f79fd4dc481ac2226d5b97102', 'finish_reason': {'type': 'stop', 'matched': 2}, 'prompt_tokens': 23, 'weight_version': 'default', 'completion_tokens': 32, 'cached_tokens': 0, 'spec_verify_ct': 14, 'spec_accept_rate': 0.40476190476190477, 'spec_accept_length': 2.2857142857142856, 'e2e_latency': 0.15333986282348633}}

curl -s http://localhost:8001/metrics | grep -E "spec_accept"
# HELP sglang:spec_accept_length The average acceptance length of speculative decoding.
# TYPE sglang:spec_accept_length gauge
sglang:spec_accept_length{engine_type="unified",model_name="meta-llama/Llama-2-7b-chat-hf",pp_rank="0",tp_rank="0"} 2.4782608695652173
# HELP sglang:spec_accept_rate The average acceptance rate of speculative decoding (`accepted tokens / total draft tokens` in batch).
# TYPE sglang:spec_accept_rate gauge
sglang:spec_accept_rate{engine_type="unified",model_name="meta-llama/Llama-2-7b-chat-hf",pp_rank="0",tp_rank="0"} 0.8260869565217391