MaxView

XPK Start: Thu Apr 23 19:18:26 UTC 2026
2026-04-23 19:18:30.303049: 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 00:00:1776971910.315815      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776971910.319617      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776971910.330876      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776971910.330897      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776971910.330899      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776971910.330901      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-23 19:18:49.494089: 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)
I0423 19:18:50.009409 135472160229184 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-23 19:18:59,050:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0423 19:18:59.050084 135472160229184 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-23 19:18:59,052:jax._src.distributed:157: Connecting to JAX distributed service on mt-05-fp8-stt5f-slice-job-0-0.mt-05-fp8-stt5f:8482
I0423 19:18:59.052536 135472160229184 distributed.py:157] Connecting to JAX distributed service on mt-05-fp8-stt5f-slice-job-0-0.mt-05-fp8-stt5f:8482
I0423 19:19:00.635245 135472160229184 max_utils.py:284] Jax distributed system initialized!
I0423 19:19:06.762602 135472160229184 max_utils.py:800] System Information: Jax Version: 0.8.1
I0423 19:19:06.762729 135472160229184 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0423 19:19:06.762772 135472160229184 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
I0423 19:19:06.762806 135472160229184 train_utils.py:347] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0423 19:19:07.374555 135472160229184 maxtext_utils.py:1517] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0423 19:19:07.374871 135472160229184 checkpointing.py:677] Setting up checkpoint logger...
I0423 19:19:07.374937 135472160229184 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0423 19:19:07.374982 135472160229184 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0423 19:19:07.375333 135472160229184 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 0x7b359f1750d0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0423 19:19:10.285498 135472160229184 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0423 19:19:10.285757 135472160229184 checkpoint_manager.py:709] [process=5][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b23786de1b0>}, handler_registry=None
I0423 19:19:10.285987 135472160229184 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b23786de1b0>` 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`.
I0423 19:19:10.286037 135472160229184 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b22f771b3e0>` 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`.
I0423 19:19:10.286073 135472160229184 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 0x7b23786de1b0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b23786de1b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b22f771b3e0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b22f771b3e0>}).
I0423 19:19:10.286562 135472160229184 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0423 19:19:10.286637 135472160229184 async_checkpointer.py:177] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b21c470eac0> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0423 19:19:11.727365 135472160229184 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints
I0423 19:19:12.149037 135472160229184 checkpoint_manager.py:929] [process=5][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_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b22f7718440>
I0423 19:19:12.149223 135472160229184 checkpointing.py:301] Checkpoint manager created!
I0423 19:19:13.252612 135472160229184 nnx_wrappers.py:442] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0423 19:19:13.252737 135472160229184 nnx_wrappers.py:442] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 19:19:13.628615 135472160229184 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0423 19:19:13.628718 135472160229184 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 19:19:13.644861 135472160229184 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0423 19:19:13.644926 135472160229184 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 19:19:13.704460 135472160229184 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 19:19:13.704546 135472160229184 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 19:19:13.720601 135472160229184 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 19:19:13.720684 135472160229184 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 19:19:13.736568 135472160229184 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 19:19:13.736626 135472160229184 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 19:19:13.760961 135472160229184 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0423 19:19:13.761033 135472160229184 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 19:19:13.820821 135472160229184 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0423 19:19:13.820905 135472160229184 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0423 19:19:14.286964 135472160229184 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0423 19:19:14.287091 135472160229184 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0423 19:19:16.307278 135472160229184 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:  ()
I0423 19:19:16.307404 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.307454 135472160229184 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:  ()
I0423 19:19:16.307494 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_0/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.307529 135472160229184 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:  ()
I0423 19:19:16.307562 135472160229184 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:  ()
I0423 19:19:16.307593 135472160229184 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:  ()
I0423 19:19:16.307624 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.307670 135472160229184 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:  ()
I0423 19:19:16.307707 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wi_1/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.307739 135472160229184 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:  ()
I0423 19:19:16.307771 135472160229184 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:  ()
I0423 19:19:16.307801 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/input_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.307833 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.307863 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/kernel_amax_history
    Shape:     float32[16,1024]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.307893 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.307922 135472160229184 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:  ()
I0423 19:19:16.307952 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/mlp/wo/Fp8DirectDotGeneralOp_0/output_grad_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.307984 135472160229184 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:  ()
I0423 19:19:16.308012 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.308040 135472160229184 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:  ()
I0423 19:19:16.308067 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/key/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.308094 135472160229184 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:  ()
I0423 19:19:16.308121 135472160229184 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:  ()
I0423 19:19:16.308148 135472160229184 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:  ()
I0423 19:19:16.308175 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.308202 135472160229184 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:  ()
I0423 19:19:16.308228 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/out/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.308254 135472160229184 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:  ()
I0423 19:19:16.308281 135472160229184 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:  ()
I0423 19:19:16.308307 135472160229184 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:  ()
I0423 19:19:16.308335 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.308361 135472160229184 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:  ()
I0423 19:19:16.308387 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/query/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.308413 135472160229184 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:  ()
I0423 19:19:16.308439 135472160229184 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:  ()
I0423 19:19:16.308465 135472160229184 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:  ()
I0423 19:19:16.308491 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/input_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.308517 135472160229184 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:  ()
I0423 19:19:16.308543 135472160229184 maxtext_utils.py:1620]  params/_overwrite_with_gradient/decoder/layers/self_attention/value/Fp8DirectDotGeneralOp_0/kernel_scale
    Shape:     float32[16,1]
    Logical:   PartitionSpec()
    Physical:  ()
I0423 19:19:16.308578 135472160229184 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:  ()
I0423 19:19:16.308629 135472160229184 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:  ()
I0423 19:19:16.308719 135472160229184 maxtext_utils.py:1620]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   PartitionSpec('norm',)
    Physical:  (None,)
I0423 19:19:16.308780 135472160229184 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)
I0423 19:19:16.308821 135472160229184 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)
I0423 19:19:16.308868 135472160229184 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')
I0423 19:19:16.308912 135472160229184 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)
I0423 19:19:16.308945 135472160229184 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)
I0423 19:19:16.308997 135472160229184 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)
I0423 19:19:16.309043 135472160229184 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')
I0423 19:19:16.309078 135472160229184 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)
I0423 19:19:16.309110 135472160229184 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)
I0423 19:19:16.309157 135472160229184 maxtext_utils.py:1620]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   PartitionSpec('embed', 'vocab')
    Physical:  ('fsdp', None)
I0423 19:19:16.309202 135472160229184 maxtext_utils.py:1620]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   PartitionSpec('vocab', 'embed')
    Physical:  (None, 'fsdp')

I0423 19:19:18.088542 135472160229184 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0423 19:19:18.088640 135472160229184 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0423 19:19:18.104782 135472160229184 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0423 19:19:18.104840 135472160229184 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0423 19:19:31.403897 135472160229184 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.
I0423 19:19:31.404923 135472160229184 metric_logger.py:289] number parameters: 1.105 billion
I0423 19:19:46.009320 135472160229184 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0423 19:19:46.267349 135472160229184 checkpointing.py:776] Waited 0.2580089569091797 seconds for step 0 to finish before starting checkpointing.
I0423 19:19:46.270040 135472160229184 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 19:19:46.272075 135472160229184 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 0
I0423 19:19:46.275080 135472160229184 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints/0.
I0423 19:19:46.861568 135472160229184 signaling_client.py:364] Using JaxDistributedSignalingClient
I0423 19:19:46.862550 135472160229184 jax_array_handlers.py:358] Scheduling D2H of 81 prioritized jax.Array.
I0423 19:19:46.862608 135472160229184 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
I0423 19:19:47.222301 135472160229184 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.360859s
I0423 19:19:47.222475 135472160229184 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.449 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.629910945892334 s) (per-host)
I0423 19:19:47.222527 135472160229184 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.629973s (batch_requests_ready=0.263816s, total_serialization_initiated=0.366087s, others=0.000069s)
I0423 19:19:47.222635 135472160229184 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.634050s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=0.634031)
I0423 19:19:47.223602 135348579464960 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started.
I0423 19:19:47.223737 135472160229184 async_checkpointer.py:561] Finished blocking save. Time taken: 0.951595s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints/0.
I0423 19:19:47.663132 135472160229184 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 19:19:47.663516 135346406803200 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0423 19:19:47.663695 135472160229184 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776971986.2700217, 'wait_for_prev_duration_secs': 6.222724914550781e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776971986.2721143, 'checkpointer_blocking_duration_secs': 0.9517989158630371, 'get_old_steps_start_time': 1776971987.2239318, 'get_old_steps_duration_secs': 2.5272369384765625e-05, 'checkpoint_manager_blocking_start_time': 1776971986.2682455, 'checkpoint_manager_blocking_duration_secs': 1.3953940868377686}
I0423 19:19:47.663810 135472160229184 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0423 19:19:47.663864 135472160229184 max_utils.py:750] 
Memstats: After params initialized:
I0423 19:19:47.663919 135472160229184 max_utils.py:756] 	Using (GB) 0.45 / 31.25 (1.440000%) on TPU_18(process=5,(2,4,0,0))
I0423 19:19:47.663955 135472160229184 max_utils.py:756] 	Using (GB) 0.45 / 31.25 (1.440000%) on TPU_19(process=5,(3,4,0,0))
I0423 19:19:47.663985 135472160229184 max_utils.py:756] 	Using (GB) 0.45 / 31.25 (1.440000%) on TPU_22(process=5,(2,5,0,0))
I0423 19:19:47.664011 135472160229184 max_utils.py:756] 	Using (GB) 0.45 / 31.25 (1.440000%) on TPU_23(process=5,(3,5,0,0))
I0423 19:19:47.931174 135472160229184 metric_logger.py:185] completed step: 0, seconds: 14.604, TFLOP/s/device: 0.930, Tokens/s/device: 140.233, total_weights: 65536, loss: 10.874
I0423 19:19:48.107251 135472160229184 metric_logger.py:185] completed step: 1, seconds: 1.920, TFLOP/s/device: 7.077, Tokens/s/device: 1066.737, total_weights: 65536, loss: 10.874
I0423 19:19:48.549098 135472160229184 metric_logger.py:185] completed step: 2, seconds: 0.019, TFLOP/s/device: 721.453, Tokens/s/device: 108745.288, total_weights: 65536, loss: 10.816
I0423 19:19:48.707304 135472160229184 metric_logger.py:185] completed step: 3, seconds: 0.442, TFLOP/s/device: 30.706, Tokens/s/device: 4628.343, total_weights: 65536, loss: 10.431
I0423 19:19:49.024808 135472160229184 metric_logger.py:185] completed step: 4, seconds: 0.163, TFLOP/s/device: 83.138, Tokens/s/device: 12531.436, total_weights: 65536, loss: 9.992
I0423 19:19:49.031105 135472160229184 metric_logger.py:185] completed step: 5, seconds: 0.159, TFLOP/s/device: 85.701, Tokens/s/device: 12917.713, total_weights: 65536, loss: 9.549
I0423 19:20:09.036336 135472160229184 metric_logger.py:185] completed step: 6, seconds: 0.317, TFLOP/s/device: 42.798, Tokens/s/device: 6450.963, total_weights: 65536, loss: 9.111
I0423 19:20:09.194728 135472160229184 metric_logger.py:185] completed step: 7, seconds: 19.848, TFLOP/s/device: 0.685, Tokens/s/device: 103.182, total_weights: 65536, loss: 8.685
I0423 19:20:09.352685 135472160229184 metric_logger.py:185] completed step: 8, seconds: 0.163, TFLOP/s/device: 83.414, Tokens/s/device: 12573.134, total_weights: 65536, loss: 8.281
I0423 19:20:09.357694 135472160229184 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0423 19:20:09.669177 135472160229184 checkpointing.py:776] Waited 0.3114495277404785 seconds for step 10 to finish before starting checkpointing.
I0423 19:20:09.672764 135472160229184 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 19:20:09.880641    2593 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0423 19:20:11.797894 135349073716992 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 81 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints/0/items/array_metadatas/process_5
I0423 19:20:26.975562 135348579464960 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 39.119 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 40.38295531272888 s) (per-host)
I0423 19:20:26.975695 135348579464960 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 39.751878s.
I0423 19:20:34.026056 135348579464960 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 46.802222s.
I0423 19:20:34.026328 135346406803200 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 19:20:34.026449 135346406803200 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0423 19:20:34.026499 135346406803200 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0423 19:20:34.028281 135346406803200 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0423 19:20:34.028456 135472160229184 checkpoint_manager.py:2036] [process=5][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0423 19:20:34.028592 135472160229184 checkpoint_manager.py:1458] Waiting for previous save to complete took 24.355834 seconds. If this number is high, consider checkpointing less frequently.
I0423 19:20:34.030350 135472160229184 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 10
I0423 19:20:34.033881 135472160229184 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints/10.
I0423 19:20:34.581968 135472160229184 jax_array_handlers.py:358] Scheduling D2H of 81 prioritized jax.Array.
I0423 19:20:34.582074 135472160229184 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
I0423 19:20:34.637360 135472160229184 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.056439s
I0423 19:20:34.637533 135472160229184 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 4.870 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.31681156158447266 s) (per-host)
I0423 19:20:34.637588 135472160229184 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.316874s (batch_requests_ready=0.256866s, total_serialization_initiated=0.059937s, others=0.000071s)
I0423 19:20:34.637703 135472160229184 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.320860s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.320844)
I0423 19:20:34.638557 135345886717696 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started.
I0423 19:20:34.638785 135472160229184 async_checkpointer.py:561] Finished blocking save. Time taken: 0.608364s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints/10.
I0423 19:20:34.650506 135472160229184 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 19:20:34.650797 135346406803200 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0423 19:20:34.650960 135472160229184 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776972009.6727343, 'wait_for_prev_duration_secs': 24.355833768844604, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776972034.030389, 'checkpointer_blocking_duration_secs': 0.6085069179534912, 'get_old_steps_start_time': 1776972034.638916, 'get_old_steps_duration_secs': 2.6702880859375e-05, 'checkpoint_manager_blocking_start_time': 1776972009.6699572, 'checkpoint_manager_blocking_duration_secs': 24.980968713760376}
I0423 19:20:34.651072 135472160229184 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0423 19:20:34.651504 135472160229184 metric_logger.py:185] completed step: 9, seconds: 0.158, TFLOP/s/device: 85.818, Tokens/s/device: 12935.418, total_weights: 65536, loss: 7.908
I0423 19:20:34.658833 135472160229184 metric_logger.py:185] completed step: 10, seconds: 0.158, TFLOP/s/device: 86.046, Tokens/s/device: 12969.824, total_weights: 65536, loss: 7.568
I0423 19:20:34.816259 135472160229184 metric_logger.py:185] completed step: 11, seconds: 25.299, TFLOP/s/device: 0.537, Tokens/s/device: 80.951, total_weights: 65536, loss: 7.288
I0423 19:20:34.974423 135472160229184 metric_logger.py:185] completed step: 12, seconds: 0.006, TFLOP/s/device: 2235.461, Tokens/s/device: 336952.945, total_weights: 65536, loss: 7.032
I0423 19:20:35.132683 135472160229184 metric_logger.py:185] completed step: 13, seconds: 0.159, TFLOP/s/device: 85.691, Tokens/s/device: 12916.246, total_weights: 65536, loss: 6.815
I0423 19:20:35.290863 135472160229184 metric_logger.py:185] completed step: 14, seconds: 0.158, TFLOP/s/device: 85.898, Tokens/s/device: 12947.521, total_weights: 65536, loss: 6.635
I0423 19:20:35.449117 135472160229184 metric_logger.py:185] completed step: 15, seconds: 0.158, TFLOP/s/device: 85.775, Tokens/s/device: 12928.966, total_weights: 65536, loss: 6.492
I0423 19:20:35.607203 135472160229184 metric_logger.py:185] completed step: 16, seconds: 0.158, TFLOP/s/device: 85.989, Tokens/s/device: 12961.205, total_weights: 65536, loss: 6.380
I0423 19:20:35.765342 135472160229184 metric_logger.py:185] completed step: 17, seconds: 0.158, TFLOP/s/device: 85.893, Tokens/s/device: 12946.784, total_weights: 65536, loss: 6.293
I0423 19:20:35.923553 135472160229184 metric_logger.py:185] completed step: 18, seconds: 0.158, TFLOP/s/device: 85.775, Tokens/s/device: 12928.885, total_weights: 65536, loss: 6.223
I0423 19:20:36.081406 135472160229184 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0423 19:20:36.082442 135472160229184 checkpointing.py:776] Waited 0.001055002212524414 seconds for step 19 to finish before starting checkpointing.
I0423 19:20:36.084421 135472160229184 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 19:20:39.833761 135346431981312 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 81 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints/10/items/array_metadatas/process_5
I0423 19:21:16.884024 135345886717696 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 37.116 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.56325340270996 s) (per-host)
I0423 19:21:16.884155 135345886717696 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.245423s.
I0423 19:21:26.999262 135345886717696 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 52.360516s.
I0423 19:21:26.999558 135346406803200 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 19:21:26.999694 135346406803200 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0423 19:21:26.999746 135346406803200 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0423 19:21:27.001299 135346406803200 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0423 19:21:27.001459 135472160229184 checkpoint_manager.py:2036] [process=5][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10.
W0423 19:21:27.001581 135472160229184 checkpoint_manager.py:1458] Waiting for previous save to complete took 50.917173 seconds. If this number is high, consider checkpointing less frequently.
I0423 19:21:27.003293 135472160229184 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 19
I0423 19:21:27.006799 135472160229184 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints/19.
I0423 19:21:27.551335 135472160229184 jax_array_handlers.py:358] Scheduling D2H of 81 prioritized jax.Array.
I0423 19:21:27.551436 135472160229184 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
I0423 19:21:27.608830 135472160229184 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.058749s
I0423 19:21:27.608998 135472160229184 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 4.865 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.31708431243896484 s) (per-host)
I0423 19:21:27.609056 135472160229184 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.317153s (batch_requests_ready=0.254823s, total_serialization_initiated=0.062256s, others=0.000074s)
I0423 19:21:27.609160 135472160229184 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.321376s (all_items=0.000012s, per_item={'items': '0.00001216'}, temp_paths=0.321364)
I0423 19:21:27.610116 135346431981312 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started.
I0423 19:21:27.610278 135472160229184 async_checkpointer.py:561] Finished blocking save. Time taken: 0.606909s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints/19.
I0423 19:21:28.038247 135472160229184 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 19:21:28.038613 135346406803200 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0423 19:21:28.038830 135472160229184 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776972036.0843918, 'wait_for_prev_duration_secs': 50.91717338562012, 'time_between_consecutive_saves_sec': 2.0560715198516846, 'checkpointer_blocking_start_time': 1776972087.0033362, 'checkpointer_blocking_duration_secs': 0.6070513725280762, 'get_old_steps_start_time': 1776972087.6104016, 'get_old_steps_duration_secs': 2.1219253540039062e-05, 'checkpoint_manager_blocking_start_time': 1776972036.082731, 'checkpoint_manager_blocking_duration_secs': 51.95606470108032}
I0423 19:21:28.038975 135472160229184 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0423 19:21:28.039020 135472160229184 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 19:21:33.171673 135346390017792 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 81 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840/linen_xpk_feat_Migrate_Decoder_And_Tests_to_NNX_20260423_190840_05_fp8/checkpoints/19/items/array_metadatas/process_5
I0423 19:22:09.292313 135346431981312 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 37.613 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.000357151031494 s) (per-host)
I0423 19:22:09.292441 135346431981312 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.682211s.
I0423 19:22:16.865880 135346431981312 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 49.255635s.
I0423 19:22:16.866105 135346406803200 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 19:22:16.866178 135346406803200 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0423 19:22:16.866229 135346406803200 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0423 19:22:16.867617 135346406803200 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0423 19:22:16.867801 135472160229184 checkpoint_manager.py:2036] [process=5][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19.
I0423 19:22:16.867956 135472160229184 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 19:22:16.868615 135472160229184 metric_logger.py:185] completed step: 19, seconds: 0.158, TFLOP/s/device: 86.150, Tokens/s/device: 12985.448, 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: Thu Apr 23 19:22:26 UTC 2026
EXIT_CODE=0