from chutes.chute import NodeSelector
from chutes.chute.template.sglang import build_sglang_chute

chute = build_sglang_chute(
    tee=True,
    username="chutes",
    readme="Qwen/Qwen3-32B, FP8 precision with DFLASH speculative decoding",
    model_name="Qwen/Qwen3-32B-FP8",
    name="Qwen/Qwen3-32B-TEE",
    revision="aa55da1ecc13d006e8b8e4f54579b1ea8c3db2df",
    image="chutes/sglang:0.5.13-post1c",
    official_package=True,
    concurrency=24,
    node_selector=NodeSelector(
        gpu_count=1,
        include=["pro_6000"],
    ),
    engine_args=(
        "--trust-remote-code "
        "--reasoning-parser qwen3 "
        "--speculative-algorithm DFLASH "
        "--speculative-draft-model-path chutesai/Qwen3-32B-FP8-DFLASH "
        "--speculative-num-draft-tokens 16 "
        "--speculative-draft-attention-backend triton "
        "--mem-fraction-static 0.85 "
        "--cuda-graph-max-bs 24 "
        "--tool-call-parser qwen25"
    ),
)