Skip to main content

Amr-amr's group workspace

Timestamps visible
2024-10-29 18:03:12
    throughput/device/tokens_per_second=355,493
2024-10-29 18:03:12
    throughput/device/batches_per_second=5.424
2024-10-29 18:03:12
2024-10-29 14:03:12.700 cn-l056.server.mila.quebec:0 olmo.train:967 INFO [step=242645/262144,epoch=0]
2024-10-29 18:03:12
    optim/total_grad_norm=0.7323
2024-10-29 18:03:12
    train/CrossEntropyLoss=3.866
2024-10-29 18:03:12
    train/Perplexity=47.77
2024-10-29 18:03:12
    throughput/total_tokens=127,215,861,760
2024-10-29 18:03:12
    throughput/total_training_Gflops=2,511,587,138
2024-10-29 18:03:12
    throughput/total_training_log_Gflops=21.64
2024-10-29 18:03:12
    throughput/device/tokens_per_second=355,988
2024-10-29 18:03:12
    throughput/device/batches_per_second=5.432
2024-10-29 18:03:12
2024-10-29 14:03:12.882 cn-l056.server.mila.quebec:0 olmo.train:967 INFO [step=242646/262144,epoch=0]
2024-10-29 18:03:12
    optim/total_grad_norm=0.6387
2024-10-29 18:03:12
    train/CrossEntropyLoss=3.814
2024-10-29 18:03:12
    train/Perplexity=45.34
2024-10-29 18:03:12
    throughput/total_tokens=127,216,386,048
2024-10-29 18:03:12
    throughput/total_training_Gflops=2,511,597,489
2024-10-29 18:03:12
    throughput/total_training_log_Gflops=21.64
2024-10-29 18:03:12
    throughput/device/tokens_per_second=355,983
2024-10-29 18:03:12
    throughput/device/batches_per_second=5.432