MaxView

XPK Start: Wed Apr 22 19:17:53 UTC 2026
2026-04-22 19:17:57.523911: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
E0000 00:00:1776885477.537274      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776885477.541015      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776885477.552608      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776885477.552625      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776885477.552628      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776885477.552630      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-22 19:18:16.724557: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303)
I0422 19:18:17.251245 139074349336384 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-22 19:18:26,291:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0422 19:18:26.291758 139074349336384 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-22 19:18:26,297:jax._src.distributed:157: Connecting to JAX distributed service on mt-05-fp8-55ds1-slice-job-0-0.mt-05-fp8-55ds1:8482
I0422 19:18:26.297470 139074349336384 distributed.py:157] Connecting to JAX distributed service on mt-05-fp8-55ds1-slice-job-0-0.mt-05-fp8-55ds1:8482
I0422 19:18:27.607131 139074349336384 max_utils.py:284] Jax distributed system initialized!
I0422 19:18:34.396921 139074349336384 max_utils.py:800] System Information: Jax Version: 0.8.1
I0422 19:18:34.397028 139074349336384 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0422 19:18:34.397069 139074349336384 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Nov 12 2025 14:16:36 (1762985796) cl/831091709
I0422 19:18:34.397121 139074349336384 train_utils.py:347] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0422 19:18:35.013904 139074349336384 maxtext_utils.py:1517] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0422 19:18:35.014227 139074349336384 checkpointing.py:677] Setting up checkpoint logger...
I0422 19:18:35.014288 139074349336384 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0422 19:18:35.014334 139074349336384 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0422 19:18:35.014694 139074349336384 base_pytree_checkpoint_handler.py:415] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7e7c493ac890>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0422 19:18:37.897572 139074349336384 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0422 19:18:37.897817 139074349336384 checkpoint_manager.py:709] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e69d32d7920>}, handler_registry=None
I0422 19:18:37.898050 139074349336384 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e69d32d7920>` for item "items" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`.
I0422 19:18:37.898113 139074349336384 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e69d28677d0>` for item "metrics" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`.
I0422 19:18:37.898149 139074349336384 composite_checkpoint_handler.py:505] Initialized registry DefaultCheckpointHandlerRegistry({('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e69d32d7920>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e69d32d7920>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e69d28677d0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e69d28677d0>}).
I0422 19:18:37.898648 139074349336384 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0422 19:18:37.898727 139074349336384 async_checkpointer.py:177] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7e68300ce840> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0422 19:18:39.049420 139074349336384 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints
I0422 19:18:39.489746 139074349336384 checkpoint_manager.py:929] [process=6][thread=MainThread] CheckpointManager created,  primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7e69d2864fe0>
I0422 19:18:39.489919 139074349336384 checkpointing.py:301] Checkpoint manager created!
I0422 19:18:40.595680 139074349336384 nnx_wrappers.py:442] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0422 19:18:40.595788 139074349336384 nnx_wrappers.py:442] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 19:18:40.968004 139074349336384 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 19:18:40.968111 139074349336384 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 19:18:40.983792 139074349336384 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 19:18:40.983853 139074349336384 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 19:18:41.043221 139074349336384 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 19:18:41.043307 139074349336384 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 19:18:41.059239 139074349336384 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 19:18:41.059301 139074349336384 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 19:18:41.075078 139074349336384 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 19:18:41.075155 139074349336384 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 19:18:41.099895 139074349336384 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0422 19:18:41.099967 139074349336384 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 19:18:41.159776 139074349336384 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0422 19:18:41.159865 139074349336384 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0422 19:18:41.628984 139074349336384 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0422 19:18:41.629125 139074349336384 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0422 19:18:43.674982 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/input_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675126 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675182 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/kernel_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675222 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675255 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/output_grad_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675292 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/output_grad_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675323 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/input_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675353 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675381 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/kernel_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675408 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675436 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/output_grad_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675464 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/output_grad_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675491 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/input_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675522 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()

    Physical:  ()
I0422 19:18:43.675551 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/kernel_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675579 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675606 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/output_grad_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675633 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/output_grad_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675660 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/input_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675687 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675713 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/kernel_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675738 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675763 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/output_grad_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675788 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/output_grad_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675814 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/input_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675838 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675863 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/kernel_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675888 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675912 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/output_grad_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675937 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/output_grad_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675962 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/input_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.675987 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.676011 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/kernel_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.676035 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.676059 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/output_grad_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.676084 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/output_grad_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.676141 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/input_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.676174 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.676201 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/kernel_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.676226 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.676251 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/output_grad_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.676280 139074349336384 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/output_grad_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0422 19:18:43.676330 139074349336384 maxtext_utils.py:1620]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   PartitionSpec('norm',)
    Physical:  (None,)
I0422 19:18:43.676380 139074349336384 maxtext_utils.py:1620]  params/params/decoder/layers/mlp/wi_0/kernel
    Shape:     float32[2048,16,7168]
    Logical:   PartitionSpec('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0422 19:18:43.676414 139074349336384 maxtext_utils.py:1620]  params/params/decoder/layers/mlp/wi_1/kernel
    Shape:     float32[2048,16,7168]
    Logical:   PartitionSpec('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0422 19:18:43.676456 139074349336384 maxtext_utils.py:1620]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   PartitionSpec('mlp', 'layers', 'embed')
    Physical:  (None, None, 'fsdp')
I0422 19:18:43.676497 139074349336384 maxtext_utils.py:1620]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   PartitionSpec('norm', 'layers')
    Physical:  (None, None)
I0422 19:18:43.676529 139074349336384 maxtext_utils.py:1620]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   PartitionSpec('norm', 'layers')
    Physical:  (None, None)
I0422 19:18:43.676573 139074349336384 maxtext_utils.py:1620]  params/params/decoder/layers/self_attention/key/kernel
    Shape:     float32[2048,16,16,128]
    Logical:   PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim')
    Physical:  ('fsdp', None, None, None)
I0422 19:18:43.676618 139074349336384 maxtext_utils.py:1620]  params/params/decoder/layers/self_attention/out/kernel
    Shape:     float32[16,16,128,2048]
    Logical:   PartitionSpec('heads', 'layers', 'kv', 'embed')
    Physical:  (None, None, None, 'fsdp')
I0422 19:18:43.676652 139074349336384 maxtext_utils.py:1620]  params/params/decoder/layers/self_attention/query/kernel
    Shape:     float32[2048,16,16,128]
    Logical:   PartitionSpec('embed', 'layers', 'q_heads', 'kv')
    Physical:  ('fsdp', None, None, None)
I0422 19:18:43.676684 139074349336384 maxtext_utils.py:1620]  params/params/decoder/layers/self_attention/value/kernel
    Shape:     float32[2048,16,16,128]
    Logical:   PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim')
    Physical:  ('fsdp', None, None, None)
I0422 19:18:43.676732 139074349336384 maxtext_utils.py:1620]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   PartitionSpec('embed', 'vocab')
    Physical:  ('fsdp', None)
I0422 19:18:43.676777 139074349336384 maxtext_utils.py:1620]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   PartitionSpec('vocab', 'embed')
    Physical:  (None, 'fsdp')
I0422 19:18:45.447896 139074349336384 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0422 19:18:45.447989 139074349336384 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0422 19:18:45.463066 139074349336384 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0422 19:18:45.463139 139074349336384 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0422 19:18:58.654033 139074349336384 max_utils.py:791] Total memory size: 1.5 GB, Output size: 0.4 GB, Temp size: 1.1 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0422 19:18:58.655008 139074349336384 metric_logger.py:289] number parameters: 1.105 billion
I0422 19:19:13.608346 139074349336384 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0422 19:19:13.782468 139074349336384 checkpointing.py:776] Waited 0.1741018295288086 seconds for step 0 to finish before starting checkpointing.
I0422 19:19:13.785269 139074349336384 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 19:19:13.787064 139074349336384 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 0
I0422 19:19:13.790017 139074349336384 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints/0.
I0422 19:19:14.819473 139074349336384 signaling_client.py:364] Using JaxDistributedSignalingClient
I0422 19:19:14.820548 139074349336384 jax_array_handlers.py:358] Scheduling D2H of 81 prioritized jax.Array.
I0422 19:19:14.820613 139074349336384 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0422 19:19:15.185628 139074349336384 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.366280s
I0422 19:19:15.185801 139074349336384 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.447 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 1.0662360191345215 s) (per-host)
I0422 19:19:15.185855 139074349336384 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 1.066299s (batch_requests_ready=0.694691s, total_serialization_initiated=0.371536s, others=0.000072s)
I0422 19:19:15.185961 139074349336384 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 1.070343s (all_items=0.000020s, per_item={'items': '0.00001979'}, temp_paths=1.070323)
I0422 19:19:15.186901 138954263226112 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0422 19:19:15.187071 139074349336384 async_checkpointer.py:561] Finished blocking save. Time taken: 1.399916s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints/0.
I0422 19:19:15.189061 139074349336384 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 19:19:15.189356 138947787204352 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0422 19:19:15.189511 139074349336384 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776885553.7852376, 'wait_for_prev_duration_secs': 0.00012183189392089844, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776885553.7871232, 'checkpointer_blocking_duration_secs': 1.40006685256958, 'get_old_steps_start_time': 1776885555.1872108, 'get_old_steps_duration_secs': 2.6941299438476562e-05, 'checkpoint_manager_blocking_start_time': 1776885553.7832584, 'checkpoint_manager_blocking_duration_secs': 1.4062132835388184}
I0422 19:19:15.189626 139074349336384 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0422 19:19:15.189678 139074349336384 max_utils.py:750] 
Memstats: After params initialized:
I0422 19:19:15.189728 139074349336384 max_utils.py:756] 	Using (GB) 0.45 / 31.25 (1.440000%) on TPU_24(process=6,(0,6,0,0))
I0422 19:19:15.189763 139074349336384 max_utils.py:756] 	Using (GB) 0.45 / 31.25 (1.440000%) on TPU_25(process=6,(1,6,0,0))
I0422 19:19:15.189792 139074349336384 max_utils.py:756] 	Using (GB) 0.45 / 31.25 (1.440000%) on TPU_28(process=6,(0,7,0,0))
I0422 19:19:15.189818 139074349336384 max_utils.py:756] 	Using (GB) 0.45 / 31.25 (1.440000%) on TPU_29(process=6,(1,7,0,0))
I0422 19:19:15.465280 139074349336384 metric_logger.py:185] completed step: 0, seconds: 14.953, TFLOP/s/device: 0.909, Tokens/s/device: 136.960, total_weights: 65536, loss: 10.874
I0422 19:19:15.637415 139074349336384 metric_logger.py:185] completed step: 1, seconds: 1.855, TFLOP/s/device: 7.325, Tokens/s/device: 1104.111, total_weights: 65536, loss: 10.874
I0422 19:19:16.074402 139074349336384 metric_logger.py:185] completed step: 2, seconds: 0.015, TFLOP/s/device: 929.479, Tokens/s/device: 140101.245, total_weights: 65536, loss: 10.816
I0422 19:19:16.232827 139074349336384 metric_logger.py:185] completed step: 3, seconds: 0.438, TFLOP/s/device: 31.048, Tokens/s/device: 4679.945, total_weights: 65536, loss: 10.431
I0422 19:19:16.550007 139074349336384 metric_logger.py:185] completed step: 4, seconds: 0.164, TFLOP/s/device: 82.778, Tokens/s/device: 12477.154, total_weights: 65536, loss: 9.992
I0422 19:19:16.556432 139074349336384 metric_logger.py:185] completed step: 5, seconds: 0.158, TFLOP/s/device: 85.936, Tokens/s/device: 12953.171, total_weights: 65536, loss: 9.549
I0422 19:19:36.655489 139074349336384 metric_logger.py:185] completed step: 6, seconds: 0.318, TFLOP/s/device: 42.779, Tokens/s/device: 6448.058, total_weights: 65536, loss: 9.111
I0422 19:19:36.813849 139074349336384 metric_logger.py:185] completed step: 7, seconds: 19.942, TFLOP/s/device: 0.681, Tokens/s/device: 102.697, total_weights: 65536, loss: 8.685
I0422 19:19:36.971922 139074349336384 metric_logger.py:185] completed step: 8, seconds: 0.163, TFLOP/s/device: 83.353, Tokens/s/device: 12563.801, total_weights: 65536, loss: 8.281
I0422 19:19:36.976939 139074349336384 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0422 19:19:37.288321 139074349336384 checkpointing.py:776] Waited 0.31134581565856934 seconds for step 10 to finish before starting checkpointing.
I0422 19:19:37.291078 139074349336384 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 19:19:38.154925    2715 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0422 19:19:40.966204 138954271618816 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 81 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints/0/items/array_metadatas/process_6
I0422 19:19:54.666162 138954263226112 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 38.962 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 40.5465362071991 s) (per-host)
I0422 19:19:54.666282 138954263226112 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 39.479272s.
I0422 19:20:02.306082 138954263226112 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 47.119056s.
I0422 19:20:02.306390 138947787204352 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 19:20:02.306520 138947787204352 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0422 19:20:02.306571 138947787204352 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0422 19:20:02.308447 138947787204352 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0422 19:20:02.308644 139074349336384 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0422 19:20:02.308777 139074349336384 checkpoint_manager.py:1458] Waiting for previous save to complete took 25.017704 seconds. If this number is high, consider checkpointing less frequently.
I0422 19:20:02.310679 139074349336384 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 10
I0422 19:20:02.314284 139074349336384 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints/10.
I0422 19:20:03.598273 139074349336384 jax_array_handlers.py:358] Scheduling D2H of 81 prioritized jax.Array.
I0422 19:20:03.598370 139074349336384 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0422 19:20:03.654314 139074349336384 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.057253s
I0422 19:20:03.654478 139074349336384 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.492 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.6191508769989014 s) (per-host)
I0422 19:20:03.654535 139074349336384 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.619216s (batch_requests_ready=0.558415s, total_serialization_initiated=0.060726s, others=0.000075s)
I0422 19:20:03.654637 139074349336384 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.623409s (all_items=0.000015s, per_item={'items': '0.00001526'}, temp_paths=0.623393)
I0422 19:20:03.655555 138946699691776 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0422 19:20:03.655714 139074349336384 async_checkpointer.py:561] Finished blocking save. Time taken: 1.344964s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints/10.
I0422 19:20:03.786870 139074349336384 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 19:20:03.787259 138947787204352 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0422 19:20:03.787434 139074349336384 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776885577.2910511, 'wait_for_prev_duration_secs': 25.017703533172607, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776885602.310718, 'checkpointer_blocking_duration_secs': 1.3451042175292969, 'get_old_steps_start_time': 1776885603.655841, 'get_old_steps_duration_secs': 2.574920654296875e-05, 'checkpoint_manager_blocking_start_time': 1776885577.2892582, 'checkpoint_manager_blocking_duration_secs': 26.498137712478638}
I0422 19:20:03.787548 139074349336384 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0422 19:20:03.788021 139074349336384 metric_logger.py:185] completed step: 9, seconds: 0.159, TFLOP/s/device: 85.695, Tokens/s/device: 12916.898, total_weights: 65536, loss: 7.908
I0422 19:20:03.795117 139074349336384 metric_logger.py:185] completed step: 10, seconds: 0.158, TFLOP/s/device: 86.148, Tokens/s/device: 12985.119, total_weights: 65536, loss: 7.568
I0422 19:20:03.953045 139074349336384 metric_logger.py:185] completed step: 11, seconds: 26.816, TFLOP/s/device: 0.507, Tokens/s/device: 76.372, total_weights: 65536, loss: 7.288
I0422 19:20:04.111073 139074349336384 metric_logger.py:185] completed step: 12, seconds: 0.006, TFLOP/s/device: 2313.097, Tokens/s/device: 348655.090, total_weights: 65536, loss: 7.032
I0422 19:20:04.269458 139074349336384 metric_logger.py:185] completed step: 13, seconds: 0.159, TFLOP/s/device: 85.358, Tokens/s/device: 12866.100, total_weights: 65536, loss: 6.815
I0422 19:20:04.427425 139074349336384 metric_logger.py:185] completed step: 14, seconds: 0.158, TFLOP/s/device: 85.925, Tokens/s/device: 12951.615, total_weights: 65536, loss: 6.635
I0422 19:20:04.585612 139074349336384 metric_logger.py:185] completed step: 15, seconds: 0.158, TFLOP/s/device: 85.923, Tokens/s/device: 12951.287, total_weights: 65536, loss: 6.492
I0422 19:20:04.743882 139074349336384 metric_logger.py:185] completed step: 16, seconds: 0.158, TFLOP/s/device: 85.892, Tokens/s/device: 12946.620, total_weights: 65536, loss: 6.380
I0422 19:20:04.901820 139074349336384 metric_logger.py:185] completed step: 17, seconds: 0.158, TFLOP/s/device: 85.968, Tokens/s/device: 12958.007, total_weights: 65536, loss: 6.293
I0422 19:20:05.060044 139074349336384 metric_logger.py:185] completed step: 18, seconds: 0.158, TFLOP/s/device: 85.909, Tokens/s/device: 12949.158, total_weights: 65536, loss: 6.223
I0422 19:20:05.217931 139074349336384 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0422 19:20:05.218982 139074349336384 checkpointing.py:776] Waited 0.0010700225830078125 seconds for step 19 to finish before starting checkpointing.
I0422 19:20:05.220914 139074349336384 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 19:20:08.547747 138945681684224 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 81 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints/10/items/array_metadatas/process_6
I0422 19:20:45.743437 138946699691776 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 36.990 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.70806837081909 s) (per-host)
I0422 19:20:45.743556 138946699691776 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.087888s.
I0422 19:20:52.267305 138946699691776 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 48.611622s.
I0422 19:20:52.267582 138947787204352 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 19:20:52.267700 138947787204352 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0422 19:20:52.267747 138947787204352 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0422 19:20:52.269163 138947787204352 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0422 19:20:52.269346 139074349336384 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10.
W0422 19:20:52.269467 139074349336384 checkpoint_manager.py:1458] Waiting for previous save to complete took 47.048564 seconds. If this number is high, consider checkpointing less frequently.
I0422 19:20:52.270990 139074349336384 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 19
I0422 19:20:52.274363 139074349336384 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints/19.
I0422 19:20:52.819375 139074349336384 jax_array_handlers.py:358] Scheduling D2H of 81 prioritized jax.Array.
I0422 19:20:52.819472 139074349336384 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0422 19:20:52.875992 139074349336384 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.057751s
I0422 19:20:52.876174 139074349336384 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.021 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.30722951889038086 s) (per-host)
I0422 19:20:52.876225 139074349336384 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.307293s (batch_requests_ready=0.246022s, total_serialization_initiated=0.061202s, others=0.000069s)
I0422 19:20:52.876329 139074349336384 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.311394s (all_items=0.000012s, per_item={'items': '0.00001216'}, temp_paths=0.311382)
I0422 19:20:52.877205 138954271618816 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0422 19:20:52.877359 139074349336384 async_checkpointer.py:561] Finished blocking save. Time taken: 0.606301s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints/19.
I0422 19:20:53.334306 139074349336384 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 19:20:53.334676 138947787204352 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0422 19:20:53.334847 139074349336384 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776885605.220888, 'wait_for_prev_duration_secs': 47.048564195632935, 'time_between_consecutive_saves_sec': 2.912400960922241, 'checkpointer_blocking_start_time': 1776885652.271029, 'checkpointer_blocking_duration_secs': 0.6064348220825195, 'get_old_steps_start_time': 1776885652.8774776, 'get_old_steps_duration_secs': 2.1696090698242188e-05, 'checkpoint_manager_blocking_start_time': 1776885605.2192736, 'checkpoint_manager_blocking_duration_secs': 48.11553931236267}
I0422 19:20:53.334987 139074349336384 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0422 19:20:53.335031 139074349336384 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 19:20:58.817982 138945642735360 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 81 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260422_190741_05_fp8/checkpoints/19/items/array_metadatas/process_6
I0422 19:21:35.078788 138954271618816 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 37.162 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.509795904159546 s) (per-host)
I0422 19:21:35.078911 138954271618816 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.201595s.
I0422 19:21:43.091891 138954271618816 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 50.214562s.
I0422 19:21:43.092196 138947787204352 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 19:21:43.092316 138947787204352 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0422 19:21:43.092358 138947787204352 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0422 19:21:43.093923 138947787204352 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0422 19:21:43.094116 139074349336384 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19.
I0422 19:21:43.094256 139074349336384 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 19:21:43.095005 139074349336384 metric_logger.py:185] completed step: 19, seconds: 0.158, TFLOP/s/device: 85.963, Tokens/s/device: 12957.269, total_weights: 65536, loss: 6.167
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Wed Apr 22 19:21:53 UTC 2026
EXIT_CODE=0