MaxView

XPK Start: Thu Apr 23 18:09:50 UTC 2026
2026-04-23 18:09:54.757464: 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 00:00:1776967794.770482      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776967794.774254      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776967794.786009      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776967794.786028      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776967794.786030      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776967794.786033      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-23 18:10:14.115360: 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 18:10:14.644080 132339182122816 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-23 18:10:23,686:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0423 18:10:23.686400 132339182122816 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-23 18:10:23,688:jax._src.distributed:157: Connecting to JAX distributed service on mt-11-optimizer-offload-6xy6m-slice-job-0-0.mt-11-optimizer-offload-6xy6m:8482
I0423 18:10:23.688758 132339182122816 distributed.py:157] Connecting to JAX distributed service on mt-11-optimizer-offload-6xy6m-slice-job-0-0.mt-11-optimizer-offload-6xy6m:8482
I0423 18:10:24.485420 132339182122816 max_utils.py:284] Jax distributed system initialized!
I0423 18:10:32.138070 132339182122816 max_utils.py:800] System Information: Jax Version: 0.8.1
I0423 18:10:32.138195 132339182122816 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0423 18:10:32.138237 132339182122816 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 18:10:32.138273 132339182122816 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0423 18:10:32.779440 132339182122816 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0423 18:10:32.894876 132339182122816 checkpointing.py:688] Setting up checkpoint logger...
I0423 18:10:32.894994 132339182122816 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0423 18:10:32.895039 132339182122816 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0423 18:10:32.895293 132339182122816 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 0x785c23978ef0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0423 18:10:36.642029 132339182122816 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0423 18:10:36.642423 132339182122816 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 0x7848b7b11220>}, handler_registry=None
I0423 18:10:36.642660 132339182122816 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7848b7b11220>` 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 18:10:36.642709 132339182122816 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7848b6c565d0>` 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 18:10:36.642743 132339182122816 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 0x7848b7b11220>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7848b7b11220>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7848b6c565d0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7848b6c565d0>}).
I0423 18:10:36.643066 132339182122816 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0423 18:10:36.643169 132339182122816 async_checkpointer.py:177] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7848b6bd8f40> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0423 18:10:37.347442 132339182122816 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints
I0423 18:10:37.357717 132339182122816 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/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7848b7eb1850>
I0423 18:10:37.357838 132339182122816 checkpointing.py:302] Checkpoint manager created!
I0423 18:10:38.288253 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to host
I0423 18:10:38.288367 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0423 18:10:38.288407 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288442 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288481 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288513 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0423 18:10:38.288544 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0423 18:10:38.288570 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288596 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288620 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288645 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288668 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288693 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to host
I0423 18:10:38.288717 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0423 18:10:38.288742 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288766 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288789 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288812 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0423 18:10:38.288836 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0423 18:10:38.288860 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288882 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288906 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288929 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288952 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0423 18:10:38.288976 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to host
I0423 18:10:38.288999 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to host
I0423 18:10:38.289022 132339182122816 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to host
I0423 18:10:38.290156 132339182122816 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0423 18:10:38.290213 132339182122816 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0423 18:10:40.168366 132339182122816 maxtext_utils.py:1749]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0423 18:10:40.168496 132339182122816 maxtext_utils.py:1749]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0423 18:10:40.168554 132339182122816 maxtext_utils.py:1749]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0423 18:10:40.168599 132339182122816 maxtext_utils.py:1749]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0423 18:10:40.168639 132339182122816 maxtext_utils.py:1749]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0423 18:10:40.168676 132339182122816 maxtext_utils.py:1749]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0423 18:10:40.168716 132339182122816 maxtext_utils.py:1749]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0423 18:10:40.168756 132339182122816 maxtext_utils.py:1749]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0423 18:10:40.168793 132339182122816 maxtext_utils.py:1749]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0423 18:10:40.168828 132339182122816 maxtext_utils.py:1749]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0423 18:10:40.168861 132339182122816 maxtext_utils.py:1749]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)
I0423 18:10:40.168894 132339182122816 maxtext_utils.py:1749]  token_embedder/embedding/value
    Shape:     float32[32000,2048]

    Physical:  (None, 'fsdp')
I0423 18:10:40.576501 132339182122816 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0423 18:10:40.576658 132339182122816 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 18:10:40.582042 132339182122816 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0423 18:10:40.582148 132339182122816 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 18:10:40.729324 132339182122816 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0423 18:10:40.729451 132339182122816 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 18:10:40.876702 132339182122816 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0423 18:10:40.876829 132339182122816 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 18:10:41.034162 132339182122816 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 18:10:41.034296 132339182122816 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 18:10:41.176085 132339182122816 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 18:10:41.176235 132339182122816 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 18:10:41.315946 132339182122816 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 18:10:41.316072 132339182122816 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 18:10:41.466475 132339182122816 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0423 18:10:41.466610 132339182122816 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 18:10:41.489390 132339182122816 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0423 18:10:41.489482 132339182122816 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0423 18:10:41.630543 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to device
I0423 18:10:41.630653 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 18:10:41.630693 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.630728 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.630756 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.630784 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 18:10:41.630810 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 18:10:41.630835 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.630861 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.630886 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.630910 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.630934 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.630959 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0423 18:10:41.630984 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 18:10:41.631008 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.631031 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.631055 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.631079 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 18:10:41.631114 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 18:10:41.631139 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.631164 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.631187 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.631211 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.631235 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:41.631260 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0423 18:10:41.631283 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to device
I0423 18:10:41.631305 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to device
I0423 18:10:52.313270 132339182122816 max_utils.py:791] Total memory size: 1.3 GB, Output size: 0.1 GB, Temp size: 1.1 GB, Argument size: 0.1 GB, Host temp size: 0.0 GB.
I0423 18:10:52.315526 132339182122816 metric_logger.py:289] number parameters: 1.104 billion
I0423 18:10:52.862301 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to device
I0423 18:10:52.862411 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 18:10:52.862452 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.862486 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.862514 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.862542 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 18:10:52.862570 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 18:10:52.862596 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.862622 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.862659 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.862708 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.862755 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.862800 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0423 18:10:52.862842 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 18:10:52.862884 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.862915 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.862950 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.862986 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 18:10:52.863022 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0423 18:10:52.863053 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.863089 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.863161 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.863202 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.863236 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0423 18:10:52.863274 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0423 18:10:52.863313 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to device
I0423 18:10:52.863347 132339182122816 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to device
I0423 18:11:03.097441 132339182122816 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0423 18:11:03.747688 132339182122816 checkpointing.py:798] Waited 0.6502318382263184 seconds for step 0 to finish before starting checkpointing.
I0423 18:11:03.750142 132339182122816 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 18:11:03.751877 132339182122816 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 0
I0423 18:11:03.754541 132339182122816 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints/0.
I0423 18:11:04.779682 132339182122816 signaling_client.py:364] Using JaxDistributedSignalingClient
I0423 18:11:04.780943 132339182122816 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0423 18:11:04.781003 132339182122816 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 18:11:05.033723 132339182122816 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.255224s
I0423 18:11:05.033897 132339182122816 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.619 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.9526996612548828 s) (per-host)
I0423 18:11:05.033949 132339182122816 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.952761s (batch_requests_ready=0.689914s, total_serialization_initiated=0.262778s, others=0.000069s)
I0423 18:11:05.034048 132339182122816 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.956798s (all_items=0.000035s, per_item={'items': '0.00003505'}, temp_paths=0.956763)
I0423 18:11:05.035023 132217112082176 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0423 18:11:05.035160 132339182122816 async_checkpointer.py:561] Finished blocking save. Time taken: 1.283210s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints/0.
I0423 18:11:05.037266 132339182122816 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 18:11:05.037495 132205837793024 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0423 18:11:05.037654 132339182122816 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776967863.7501235, 'wait_for_prev_duration_secs': 6.365776062011719e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776967863.7519166, 'checkpointer_blocking_duration_secs': 1.283442497253418, 'get_old_steps_start_time': 1776967865.0353808, 'get_old_steps_duration_secs': 2.956390380859375e-05, 'checkpoint_manager_blocking_start_time': 1776967863.7481713, 'checkpoint_manager_blocking_duration_secs': 1.2894368171691895}
I0423 18:11:05.037784 132339182122816 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0423 18:11:05.037883 132339182122816 max_utils.py:750] 
Memstats: After params initialized:
I0423 18:11:05.037959 132339182122816 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_24(process=6,(0,6,0,0))
I0423 18:11:05.038014 132339182122816 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_25(process=6,(1,6,0,0))
I0423 18:11:05.038066 132339182122816 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_28(process=6,(0,7,0,0))
I0423 18:11:05.038128 132339182122816 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_29(process=6,(1,7,0,0))
I0423 18:11:05.304106 132339182122816 metric_logger.py:185] completed step: 0, seconds: 10.781, TFLOP/s/device: 1.260, Tokens/s/device: 189.967, total_weights: 65536, loss: 10.874
I0423 18:11:05.503790 132339182122816 metric_logger.py:185] completed step: 1, seconds: 2.205, TFLOP/s/device: 6.162, Tokens/s/device: 928.844, total_weights: 65536, loss: 10.874
I0423 18:11:05.906886 132339182122816 metric_logger.py:185] completed step: 2, seconds: 0.039, TFLOP/s/device: 347.657, Tokens/s/device: 52402.641, total_weights: 65536, loss: 10.262
I0423 18:11:06.070560 132339182122816 metric_logger.py:185] completed step: 3, seconds: 0.392, TFLOP/s/device: 34.677, Tokens/s/device: 5226.890, total_weights: 65536, loss: 9.732
I0423 18:11:06.398299 132339182122816 metric_logger.py:185] completed step: 4, seconds: 0.180, TFLOP/s/device: 75.440, Tokens/s/device: 11371.081, total_weights: 65536, loss: 9.274
I0423 18:11:06.405663 132339182122816 metric_logger.py:185] completed step: 5, seconds: 0.163, TFLOP/s/device: 83.313, Tokens/s/device: 12557.792, total_weights: 65536, loss: 8.888
I0423 18:11:25.313803 132339182122816 metric_logger.py:185] completed step: 6, seconds: 0.328, TFLOP/s/device: 41.387, Tokens/s/device: 6238.311, total_weights: 65536, loss: 8.589
I0423 18:11:25.477416 132339182122816 metric_logger.py:185] completed step: 7, seconds: 18.745, TFLOP/s/device: 0.725, Tokens/s/device: 109.253, total_weights: 65536, loss: 8.380
I0423 18:11:25.641030 132339182122816 metric_logger.py:185] completed step: 8, seconds: 0.170, TFLOP/s/device: 79.902, Tokens/s/device: 12043.729, total_weights: 65536, loss: 8.251
I0423 18:11:25.805602 132339182122816 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0423 18:11:25.806351 132339182122816 checkpointing.py:798] Waited 0.0007581710815429688 seconds for step 9 to finish before starting checkpointing.
I0423 18:11:25.808442 132339182122816 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 18:11:26.635366    2593 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0423 18:11:28.591324 132217120474880 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints/0/items/array_metadatas/process_6
I0423 18:11:59.014850 132217112082176 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 28.753 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 54.93360686302185 s) (per-host)
I0423 18:11:59.014968 132217112082176 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 53.979728s.
I0423 18:12:09.089758 132217112082176 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 64.054503s.
I0423 18:12:09.090076 132205837793024 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 18:12:09.090207 132205837793024 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0423 18:12:09.090265 132205837793024 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0423 18:12:09.093001 132205837793024 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0423 18:12:09.093187 132339182122816 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.
W0423 18:12:09.093356 132339182122816 checkpoint_manager.py:1458] Waiting for previous save to complete took 43.284916 seconds. If this number is high, consider checkpointing less frequently.
I0423 18:12:09.095469 132339182122816 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 9
I0423 18:12:09.098844 132339182122816 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints/9.
I0423 18:12:10.078481 132339182122816 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0423 18:12:10.078638 132339182122816 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 18:12:10.137073 132339182122816 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.060671s
I0423 18:12:10.137259 132339182122816 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 4.767 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.32358646392822266 s) (per-host)
I0423 18:12:10.137317 132339182122816 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.323648s (batch_requests_ready=0.257906s, total_serialization_initiated=0.065671s, others=0.000071s)
I0423 18:12:10.137417 132339182122816 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.327660s (all_items=0.000015s, per_item={'items': '0.00001454'}, temp_paths=0.327646)
I0423 18:12:10.138297 132207440013056 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0423 18:12:10.138456 132339182122816 async_checkpointer.py:561] Finished blocking save. Time taken: 1.042914s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints/9.
I0423 18:12:10.140317 132339182122816 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 18:12:10.140558 132205837793024 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0423 18:12:10.140679 132339182122816 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776967885.8084133, 'wait_for_prev_duration_secs': 43.284915924072266, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776967929.0955105, 'checkpointer_blocking_duration_secs': 1.0430548191070557, 'get_old_steps_start_time': 1776967930.1385846, 'get_old_steps_duration_secs': 2.4318695068359375e-05, 'checkpoint_manager_blocking_start_time': 1776967885.8066325, 'checkpoint_manager_blocking_duration_secs': 44.33401083946228}
I0423 18:12:10.140784 132339182122816 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0423 18:12:10.140827 132339182122816 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 18:12:15.575561 132217120474880 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints/9/items/array_metadatas/process_6
I0423 18:12:51.592752 132207440013056 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 37.807 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.77903723716736 s) (per-host)
I0423 18:12:51.592873 132207440013056 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.454464s.
I0423 18:12:59.309368 132207440013056 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 49.170943s.
I0423 18:12:59.309674 132205837793024 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 18:12:59.309808 132205837793024 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0423 18:12:59.309857 132205837793024 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0423 18:12:59.312010 132205837793024 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0423 18:12:59.312176 132339182122816 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0423 18:12:59.312341 132339182122816 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 18:12:59.312949 132339182122816 metric_logger.py:185] completed step: 9, seconds: 0.163, TFLOP/s/device: 83.275, Tokens/s/device: 12552.173, total_weights: 65536, loss: 8.175
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Thu Apr 23 18:13:11 UTC 2026
EXIT_CODE=0