text stringlengths 54 2.3k |
|---|
[2026-03-20 19:55:23] [Rank 0] step:14661/15000 train_loss_per_token:0.7776 train_time:8238795ms step_avg:561.95ms |
[2026-03-20 19:55:23] [Rank 0] step:14661/15000 train_loss_per_token:0.7776 train_time:8238795ms step_avg:561.95ms |
[2026-03-20 19:55:35] [Rank 0] step:14681/15000 train_loss_per_token:0.7783 train_time:8250991ms step_avg:562.02ms |
[2026-03-20 19:55:35] [Rank 0] step:14681/15000 train_loss_per_token:0.7783 train_time:8250991ms step_avg:562.02ms |
[2026-03-20 19:55:48] [Rank 0] step:14701/15000 train_loss_per_token:0.7888 train_time:8263162ms step_avg:562.08ms |
[2026-03-20 19:55:48] [Rank 0] step:14701/15000 train_loss_per_token:0.7888 train_time:8263162ms step_avg:562.08ms |
[2026-03-20 19:56:00] [Rank 0] step:14721/15000 train_loss_per_token:0.7717 train_time:8275302ms step_avg:562.14ms |
[2026-03-20 19:56:00] [Rank 0] step:14721/15000 train_loss_per_token:0.7717 train_time:8275302ms step_avg:562.14ms |
[2026-03-20 19:56:12] [Rank 0] step:14741/15000 train_loss_per_token:0.8206 train_time:8287519ms step_avg:562.21ms |
[2026-03-20 19:56:12] [Rank 0] step:14741/15000 train_loss_per_token:0.8206 train_time:8287519ms step_avg:562.21ms |
[2026-03-20 19:56:24] [Rank 0] step:14761/15000 train_loss_per_token:0.8025 train_time:8299918ms step_avg:562.29ms |
[2026-03-20 19:56:24] [Rank 0] step:14761/15000 train_loss_per_token:0.8025 train_time:8299918ms step_avg:562.29ms |
[2026-03-20 19:56:37] [Rank 0] step:14781/15000 train_loss_per_token:0.7786 train_time:8312129ms step_avg:562.35ms |
[2026-03-20 19:56:37] [Rank 0] step:14781/15000 train_loss_per_token:0.7786 train_time:8312129ms step_avg:562.35ms |
[2026-03-20 19:56:49] [Rank 0] step:14801/15000 train_loss_per_token:0.7814 train_time:8324355ms step_avg:562.42ms |
[2026-03-20 19:56:49] [Rank 0] step:14801/15000 train_loss_per_token:0.7814 train_time:8324355ms step_avg:562.42ms |
[2026-03-20 19:57:01] [Rank 0] step:14821/15000 train_loss_per_token:0.7869 train_time:8336533ms step_avg:562.48ms |
[2026-03-20 19:57:01] [Rank 0] step:14821/15000 train_loss_per_token:0.7869 train_time:8336533ms step_avg:562.48ms |
[2026-03-20 19:57:13] [Rank 0] step:14841/15000 train_loss_per_token:0.7997 train_time:8348728ms step_avg:562.54ms |
[2026-03-20 19:57:13] [Rank 0] step:14841/15000 train_loss_per_token:0.7997 train_time:8348728ms step_avg:562.54ms |
[2026-03-20 19:57:25] [Rank 0] step:14861/15000 train_loss_per_token:0.7802 train_time:8360980ms step_avg:562.61ms |
[2026-03-20 19:57:25] [Rank 0] step:14861/15000 train_loss_per_token:0.7802 train_time:8360980ms step_avg:562.61ms |
[2026-03-20 19:57:38] [Rank 0] step:14881/15000 train_loss_per_token:0.7913 train_time:8373204ms step_avg:562.68ms |
[2026-03-20 19:57:38] [Rank 0] step:14881/15000 train_loss_per_token:0.7913 train_time:8373204ms step_avg:562.68ms |
[2026-03-20 19:57:50] [Rank 0] step:14901/15000 train_loss_per_token:0.7985 train_time:8385433ms step_avg:562.74ms |
[2026-03-20 19:57:50] [Rank 0] step:14901/15000 train_loss_per_token:0.7985 train_time:8385433ms step_avg:562.74ms |
[2026-03-20 19:58:02] [Rank 0] step:14921/15000 train_loss_per_token:0.7817 train_time:8397721ms step_avg:562.81ms |
[2026-03-20 19:58:02] [Rank 0] step:14921/15000 train_loss_per_token:0.7817 train_time:8397721ms step_avg:562.81ms |
[2026-03-20 19:58:14] [Rank 0] step:14941/15000 train_loss_per_token:0.7902 train_time:8409921ms step_avg:562.88ms |
[2026-03-20 19:58:14] [Rank 0] step:14941/15000 train_loss_per_token:0.7902 train_time:8409921ms step_avg:562.88ms |
[2026-03-20 19:58:27] [Rank 0] step:14961/15000 train_loss_per_token:0.7878 train_time:8422118ms step_avg:562.94ms |
[2026-03-20 19:58:27] [Rank 0] step:14961/15000 train_loss_per_token:0.7878 train_time:8422118ms step_avg:562.94ms |
[2026-03-20 19:58:39] [Rank 0] step:14981/15000 train_loss_per_token:0.7670 train_time:8434255ms step_avg:563.00ms |
[2026-03-20 19:58:39] [Rank 0] step:14981/15000 train_loss_per_token:0.7670 train_time:8434255ms step_avg:563.00ms |
[2026-03-20 19:58:53] [Rank 0] PRINT: Saved checkpoint at step 14999 to /home/aiops/zhangfz/MUON_sharpness/logs_sharpness_clean_accu/muon_lr_search/mode_0_param_qkvo_adam_lr_0.01_muon_lr_0.1_seed_42/checkpoints/ckpt_latest.pt |
[2026-03-20 19:58:53] [Rank 0] PRINT: Saved checkpoint at step 14999 to /home/aiops/zhangfz/MUON_sharpness/logs_sharpness_clean_accu/muon_lr_search/mode_0_param_qkvo_adam_lr_0.01_muon_lr_0.1_seed_42/checkpoints/ckpt_latest.pt |
[2026-03-20 19:58:53] [Rank 0] step:15000/15000 train_loss_per_token:0.7939 train_time:8448435ms step_avg:563.23ms |
[2026-03-20 19:58:53] [Rank 0] step:15000/15000 train_loss_per_token:0.7939 train_time:8448435ms step_avg:563.23ms |
[2026-03-20 19:58:53] [Rank 0] PRINT: Warning: val_tokens (1966080) not perfectly divisible by val_batch_size (262144). Some tokens might be missed. |
[2026-03-20 19:58:53] [Rank 0] PRINT: Warning: val_tokens (1966080) not perfectly divisible by val_batch_size (262144). Some tokens might be missed. |
[2026-03-20 19:58:54] [Rank 0] PRINT: step:15000/15000 train_loss_per_token:0.7939 val_loss:3.1442 train_time:8448440ms step_avg:563.23ms |
[2026-03-20 19:58:54] [Rank 0] PRINT: step:15000/15000 train_loss_per_token:0.7939 val_loss:3.1442 train_time:8448440ms step_avg:563.23ms |
[2026-03-20 19:58:54] [Rank 0] PRINT: Training completed. Checkpoint deleted. |
[2026-03-20 19:58:54] [Rank 0] PRINT: Training completed. Checkpoint deleted. |
[2026-03-20 19:58:54] [Rank 0] PRINT: --- Training Finished: Fri Mar 20 19:58:54 2026 --- |
[2026-03-20 19:58:54] [Rank 0] PRINT: --- Training Finished: Fri Mar 20 19:58:54 2026 --- |
[2026-03-20 19:58:54] [Rank 0] PRINT: Peak memory allocated: 11838 MiB reserved: 15656 MiB |
[2026-03-20 19:58:54] [Rank 0] PRINT: Peak memory allocated: 11838 MiB reserved: 15656 MiB |
[2026-03-21 00:47:42] [Rank 0] PRINT: --- Script Start: Sat Mar 21 00:47:42 2026 --- |
[2026-03-21 00:47:42] [Rank 0] PRINT: --- Script Start: Sat Mar 21 00:47:42 2026 --- |
[2026-03-21 00:47:42] [Rank 0] PRINT: Parsed CLI args: Namespace(unet=False, seed=43, optimizer_mode=0, model_parameterization='qkvo', adam_lr=0.01, muon_lr=0.1, base_dir='/home/aiops/zhangfz/MUON_sharpness/logs_sharpness_clean_accu/muon_lr_search', analyze_sharpness=False, sharpness_analysis_interval=500, hvp_chunk_si... |
[2026-03-21 00:47:42] [Rank 0] PRINT: Parsed CLI args: Namespace(unet=False, seed=43, optimizer_mode=0, model_parameterization='qkvo', adam_lr=0.01, muon_lr=0.1, base_dir='/home/aiops/zhangfz/MUON_sharpness/logs_sharpness_clean_accu/muon_lr_search', analyze_sharpness=False, sharpness_analysis_interval=500, hvp_chunk_si... |
[2026-03-21 00:47:42] [Rank 0] PRINT: Hyperparameters: Hyperparameters() |
[2026-03-21 00:47:42] [Rank 0] PRINT: Hyperparameters: Hyperparameters() |
[2026-03-21 00:47:42] [Rank 0] PRINT: Using fixed seed: 43 |
[2026-03-21 00:47:42] [Rank 0] PRINT: Using fixed seed: 43 |
[2026-03-21 00:47:42] [Rank 0] PRINT: Run directory: /home/aiops/zhangfz/MUON_sharpness/logs_sharpness_clean_accu/muon_lr_search/mode_0_param_qkvo_adam_lr_0.01_muon_lr_0.1_seed_43 |
[2026-03-21 00:47:42] [Rank 0] PRINT: Run directory: /home/aiops/zhangfz/MUON_sharpness/logs_sharpness_clean_accu/muon_lr_search/mode_0_param_qkvo_adam_lr_0.01_muon_lr_0.1_seed_43 |
[2026-03-21 00:47:42] [Rank 0] PRINT: code_sha256=2c01a5e64ccec55852d52b5c8539f402caebedf5454fe87ff6bbc52b316b754e |
[2026-03-21 00:47:42] [Rank 0] PRINT: code_sha256=2c01a5e64ccec55852d52b5c8539f402caebedf5454fe87ff6bbc52b316b754e |
[2026-03-21 00:47:42] [Rank 0] PRINT: Constructing model... |
[2026-03-21 00:47:42] [Rank 0] PRINT: Constructing model... |
[2026-03-21 00:47:44] [Rank 0] PRINT: Broadcasting model parameters... |
[2026-03-21 00:47:44] [Rank 0] PRINT: Broadcasting model parameters... |
[2026-03-21 00:47:44] [Rank 0] PRINT: Model constructed and broadcasted. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Model constructed and broadcasted. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Collecting parameters for optimizers... |
[2026-03-21 00:47:44] [Rank 0] PRINT: Collecting parameters for optimizers... |
[2026-03-21 00:47:44] [Rank 0] PRINT: Warning - Parameter torch.Size([50257, 768]) ended up in scalar_params but has ndim >= 2. Check grouping. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Warning - Parameter torch.Size([50257, 768]) ended up in scalar_params but has ndim >= 2. Check grouping. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Warning - Parameter torch.Size([50257, 768]) ended up in scalar_params but has ndim >= 2. Check grouping. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Warning - Parameter torch.Size([50257, 768]) ended up in scalar_params but has ndim >= 2. Check grouping. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Warning - Parameter torch.Size([50257, 768]) ended up in scalar_params but has ndim >= 2. Check grouping. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Warning - Parameter torch.Size([50257, 768]) ended up in scalar_params but has ndim >= 2. Check grouping. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Configuring optimizers for EXPERIMENT_MODE = 0 |
[2026-03-21 00:47:44] [Rank 0] PRINT: Configuring optimizers for EXPERIMENT_MODE = 0 |
[2026-03-21 00:47:44] [Rank 0] PRINT: Mode 0: Muon on ALL Attention (QKVO) and ALL MLP matrices. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Mode 0: Muon on ALL Attention (QKVO) and ALL MLP matrices. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Optimizers configured. Total optimizers: 2 |
[2026-03-21 00:47:44] [Rank 0] PRINT: Optimizers configured. Total optimizers: 2 |
[2026-03-21 00:47:44] [Rank 0] PRINT: Muon optimizer is active with 72 parameters. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Muon optimizer is active with 72 parameters. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Compiling model with TorchInductor... |
[2026-03-21 00:47:44] [Rank 0] PRINT: Compiling model with TorchInductor... |
[2026-03-21 00:47:44] [Rank 0] PRINT: Model compilation complete. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Model compilation complete. |
[2026-03-21 00:47:44] [Rank 0] PRINT: No checkpoint found. Starting training from scratch. |
[2026-03-21 00:47:44] [Rank 0] PRINT: No checkpoint found. Starting training from scratch. |
[2026-03-21 00:47:44] [Rank 0] PRINT: Starting warmup... |
[2026-03-21 00:47:44] [Rank 0] PRINT: Starting warmup... |
[2026-03-21 00:48:27] [Rank 0] PRINT: Warmup complete. |
[2026-03-21 00:48:27] [Rank 0] PRINT: Warmup complete. |
[2026-03-21 00:48:27] [Rank 0] PRINT: Starting training... |
[2026-03-21 00:48:27] [Rank 0] PRINT: Starting training... |
[2026-03-21 00:48:27] [Rank 0] PRINT: Warning: val_tokens (1966080) not perfectly divisible by val_batch_size (262144). Some tokens might be missed. |
[2026-03-21 00:48:27] [Rank 0] PRINT: Warning: val_tokens (1966080) not perfectly divisible by val_batch_size (262144). Some tokens might be missed. |
[2026-03-21 00:49:41] [Rank 0] PRINT: step:0/15000 val_loss:10.8258 train_time:0ms step_avg:0.00ms |
[2026-03-21 00:49:41] [Rank 0] PRINT: step:0/15000 val_loss:10.8258 train_time:0ms step_avg:0.00ms |
[2026-03-21 00:49:44] [Rank 0] PRINT: Saved checkpoint at step 0 to /home/aiops/zhangfz/MUON_sharpness/logs_sharpness_clean_accu/muon_lr_search/mode_0_param_qkvo_adam_lr_0.01_muon_lr_0.1_seed_43/checkpoints/ckpt_latest.pt |
[2026-03-21 00:49:44] [Rank 0] PRINT: Saved checkpoint at step 0 to /home/aiops/zhangfz/MUON_sharpness/logs_sharpness_clean_accu/muon_lr_search/mode_0_param_qkvo_adam_lr_0.01_muon_lr_0.1_seed_43/checkpoints/ckpt_latest.pt |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.