src/speech_decoding/models/v14_converged_v3/clip_sampler.py

Selected training excerpt · lines 105–109 · source SHA-256 aa70fc27ac09

105    counts = torch.tensor([c for _, c in aligned], dtype=torch.float64)106    which = int(torch.multinomial(counts / counts.sum(), 1, generator=generator).item())107    a_lo, count = aligned[which]108    step = int(torch.randint(0, count, (1,), generator=generator).item())109    return a_lo + step * start_align