MCPcopy Create free account
hub / github.com/RL-Align/RL-Kernel / BenchmarkConfig

Class BenchmarkConfig

benchmarks/benchmark_rl_kernels.py:56–69  ·  view source on GitHub ↗

Source from the content-addressed store, hash-verified

54
55@dataclass(frozen=True)
56class BenchmarkConfig:
57 case: str
58 candidate: str
59 device: torch.device
60 dtype: torch.dtype
61 num_prompts: int
62 samples_per_prompt: int
63 prompt_len: int
64 completion_len: int
65 vocab_size: int
66 mask_density: float
67 seed: int
68 warmup: int
69 repeat: int
70
71
72def _parse_int_list(value: str) -> list[int]:

Callers 1

mainFunction · 0.70

Calls

no outgoing calls

Tested by

no test coverage detected