# HELP sglang:num_requests_total n
# TYPE sglang:num_requests_total counter
sglang:num_requests_total{is_streaming="true",model_name="Qwen/Qwen3-8B"} 500.0
# HELP sglang:prompt_tokens_histogram p
# TYPE sglang:prompt_tokens_histogram histogram
sglang:prompt_tokens_histogram_bucket{le="256.0",model_name="Qwen/Qwen3-8B"} 0.0
sglang:prompt_tokens_histogram_bucket{le="1024.0",model_name="Qwen/Qwen3-8B"} 500.0
sglang:prompt_tokens_histogram_bucket{le="4096.0",model_name="Qwen/Qwen3-8B"} 500.0
sglang:prompt_tokens_histogram_bucket{le="+Inf",model_name="Qwen/Qwen3-8B"} 500.0
sglang:prompt_tokens_histogram_count{model_name="Qwen/Qwen3-8B"} 500.0
sglang:prompt_tokens_histogram_sum{model_name="Qwen/Qwen3-8B"} 150000.0
# HELP sglang:generation_tokens_histogram g
# TYPE sglang:generation_tokens_histogram histogram
sglang:generation_tokens_histogram_bucket{le="128.0",model_name="Qwen/Qwen3-8B"} 500.0
sglang:generation_tokens_histogram_bucket{le="256.0",model_name="Qwen/Qwen3-8B"} 500.0
sglang:generation_tokens_histogram_bucket{le="1024.0",model_name="Qwen/Qwen3-8B"} 500.0
sglang:generation_tokens_histogram_bucket{le="+Inf",model_name="Qwen/Qwen3-8B"} 500.0
sglang:generation_tokens_histogram_count{model_name="Qwen/Qwen3-8B"} 500.0
sglang:generation_tokens_histogram_sum{model_name="Qwen/Qwen3-8B"} 40000.0
# HELP sglang:e2e_request_latency_seconds e
# TYPE sglang:e2e_request_latency_seconds histogram
sglang:e2e_request_latency_seconds_bucket{is_streaming="true",le="1.0",model_name="Qwen/Qwen3-8B"} 0.0
sglang:e2e_request_latency_seconds_bucket{is_streaming="true",le="2.0",model_name="Qwen/Qwen3-8B"} 500.0
sglang:e2e_request_latency_seconds_bucket{is_streaming="true",le="4.0",model_name="Qwen/Qwen3-8B"} 500.0
sglang:e2e_request_latency_seconds_bucket{is_streaming="true",le="8.0",model_name="Qwen/Qwen3-8B"} 500.0
sglang:e2e_request_latency_seconds_bucket{is_streaming="true",le="+Inf",model_name="Qwen/Qwen3-8B"} 500.0
sglang:e2e_request_latency_seconds_count{is_streaming="true",model_name="Qwen/Qwen3-8B"} 500.0
sglang:e2e_request_latency_seconds_sum{is_streaming="true",model_name="Qwen/Qwen3-8B"} 750.0
# HELP sglang:token_usage t
# TYPE sglang:token_usage gauge
sglang:token_usage{model_name="Qwen/Qwen3-8B"} 0.2
# HELP sglang:full_token_usage t
# TYPE sglang:full_token_usage gauge
sglang:full_token_usage{model_name="Qwen/Qwen3-8B"} 0.2
# HELP sglang:swa_token_usage t
# TYPE sglang:swa_token_usage gauge
sglang:swa_token_usage{model_name="Qwen/Qwen3-8B"} 0.0
# HELP sglang:mamba_usage t
# TYPE sglang:mamba_usage gauge
sglang:mamba_usage{model_name="Qwen/Qwen3-8B"} 0.0
# HELP sglang:num_running_reqs r
# TYPE sglang:num_running_reqs gauge
sglang:num_running_reqs{model_name="Qwen/Qwen3-8B"} 5.0
# HELP sglang:num_queue_reqs q
# TYPE sglang:num_queue_reqs gauge
sglang:num_queue_reqs{model_name="Qwen/Qwen3-8B"} 0.0
# HELP sglang:cache_hit_rate c
# TYPE sglang:cache_hit_rate gauge
sglang:cache_hit_rate{model_name="Qwen/Qwen3-8B"} 0.3
# HELP sglang:estimated_flops_per_gpu_total f
# TYPE sglang:estimated_flops_per_gpu_total counter
sglang:estimated_flops_per_gpu_total{model_name="Qwen/Qwen3-8B"} 5e+14
# HELP sglang:estimated_read_bytes_per_gpu_total rb
# TYPE sglang:estimated_read_bytes_per_gpu_total counter
sglang:estimated_read_bytes_per_gpu_total{model_name="Qwen/Qwen3-8B"} 5e+13
# HELP sglang:estimated_write_bytes_per_gpu_total wb
# TYPE sglang:estimated_write_bytes_per_gpu_total counter
sglang:estimated_write_bytes_per_gpu_total{model_name="Qwen/Qwen3-8B"} 1e+12
