MaxView

XPK Start: Mon Apr 20 19:31:44 UTC 2026
2026-04-20 19:31:48.938297: 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 00:00:1776713508.950901      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776713508.954511      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776713508.965686      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776713508.965704      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776713508.965706      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776713508.965708      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-20 19:32:19.238787: 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)
I0420 19:32:19.765943 136588928440128 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-20 19:32:28,807:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0420 19:32:28.807990 136588928440128 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-20 19:32:28,810:jax._src.distributed:157: Connecting to JAX distributed service on mt-06-grad-accum-rzp58-slice-job-0-0.mt-06-grad-accum-rzp58:8482
I0420 19:32:28.810361 136588928440128 distributed.py:157] Connecting to JAX distributed service on mt-06-grad-accum-rzp58-slice-job-0-0.mt-06-grad-accum-rzp58:8482
I0420 19:32:47.511501 136588928440128 max_utils.py:284] Jax distributed system initialized!
I0420 19:32:54.436101 136588928440128 max_utils.py:800] System Information: Jax Version: 0.8.1
I0420 19:32:54.436211 136588928440128 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0420 19:32:54.436253 136588928440128 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
I0420 19:32:54.436288 136588928440128 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0420 19:32:55.067175 136588928440128 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0420 19:32:55.067785 136588928440128 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0420 19:32:55.067976 136588928440128 checkpointing.py:688] Setting up checkpoint logger...
I0420 19:32:55.068029 136588928440128 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0420 19:32:55.068072 136588928440128 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0420 19:32:55.068433 136588928440128 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 0x7c39a13793a0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0420 19:32:57.993778 136588928440128 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0420 19:32:57.994007 136588928440128 checkpoint_manager.py:709] [process=1][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c26d55d5880>}, handler_registry=None
I0420 19:32:57.994245 136588928440128 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c26d55d5880>` 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`.
I0420 19:32:57.994292 136588928440128 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c26d55de8a0>` 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`.
I0420 19:32:57.994327 136588928440128 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 0x7c26d55d5880>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c26d55d5880>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c26d55de8a0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c26d55de8a0>}).
I0420 19:32:57.994826 136588928440128 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0420 19:32:57.994903 136588928440128 async_checkpointer.py:177] [process=1][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7c26d4908180> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0420 19:32:59.786712 136588928440128 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_06_grad_accum/checkpoints
I0420 19:33:00.226784 136588928440128 checkpoint_manager.py:929] [process=1][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_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_06_grad_accum/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7c26d55d51f0>
I0420 19:33:00.226959 136588928440128 checkpointing.py:302] Checkpoint manager created!
I0420 19:33:01.872071 136588928440128 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 19:33:01.872179 136588928440128 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 19:33:02.247512 136588928440128 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 19:33:02.247606 136588928440128 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 19:33:02.264141 136588928440128 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 19:33:02.264202 136588928440128 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 19:33:02.287533 136588928440128 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 19:33:02.287599 136588928440128 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 19:33:02.303755 136588928440128 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 19:33:02.303819 136588928440128 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 19:33:02.320136 136588928440128 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 19:33:02.320201 136588928440128 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 19:33:02.345869 136588928440128 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0420 19:33:02.345935 136588928440128 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 19:33:02.365807 136588928440128 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0420 19:33:02.365882 136588928440128 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0420 19:33:02.557376 136588928440128 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0420 19:33:02.557492 136588928440128 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0420 19:33:03.969395 136588928440128 maxtext_utils.py:1740]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   PartitionSpec('norm',)
    Physical:  (None,)
I0420 19:33:03.969541 136588928440128 maxtext_utils.py:1740]  params/params/decoder/layers/mlp/wi_0/kernel
    Shape:     float32[2048,16,7168]
    Logical:   PartitionSpec('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0420 19:33:03.969599 136588928440128 maxtext_utils.py:1740]  params/params/decoder/layers/mlp/wi_1/kernel
    Shape:     float32[2048,16,7168]
    Logical:   PartitionSpec('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0420 19:33:03.969658 136588928440128 maxtext_utils.py:1740]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   PartitionSpec('mlp', 'layers', 'embed')
    Physical:  (None, None, 'fsdp')
I0420 19:33:03.969711 136588928440128 maxtext_utils.py:1740]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   PartitionSpec('norm', 'layers')
    Physical:  (None, None)
I0420 19:33:03.969749 136588928440128 maxtext_utils.py:1740]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   PartitionSpec('norm', 'layers')
    Physical:  (None, None)
I0420 19:33:03.969802 136588928440128 maxtext_utils.py:1740]  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)
I0420 19:33:03.969856 136588928440128 maxtext_utils.py:1740]  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')
I0420 19:33:03.969901 136588928440128 maxtext_utils.py:1740]  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)
I0420 19:33:03.969940 136588928440128 maxtext_utils.py:1740]  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)
I0420 19:33:03.969990 136588928440128 maxtext_utils.py:1740]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   PartitionSpec('embed', 'vocab')
    Physical:  ('fsdp', None)
I0420 19:33:03.970039 136588928440128 maxtext_utils.py:1740]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   PartitionSpec('vocab', 'embed')
    Physical:  (None, 'fsdp')

I0420 19:33:03.992932 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[2048]............................................... Unknown.
I0420 19:33:03.993002 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[2048]............................................... (None,).
I0420 19:33:04.008277 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[2048,16,7168]....................................... Unknown.
I0420 19:33:04.008336 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[2048,16,7168]....................................... ('fsdp', None, None).
I0420 19:33:04.038186 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[7168,16,2048]....................................... Unknown.
I0420 19:33:04.038252 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[7168,16,2048]....................................... (None, None, 'fsdp').
I0420 19:33:04.052896 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[2048,16]............................................ Unknown.
I0420 19:33:04.052952 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[2048,16]............................................ (None, None).
I0420 19:33:04.082431 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[2048,16,16,128]..................................... Unknown.
I0420 19:33:04.082499 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[2048,16,16,128]..................................... ('fsdp', None, None, None).
I0420 19:33:04.097565 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[16,16,128,2048]..................................... Unknown.
I0420 19:33:04.097620 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[16,16,128,2048]..................................... (None, None, None, 'fsdp').
I0420 19:33:04.141834 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[2048,32000]......................................... Unknown.
I0420 19:33:04.141904 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[2048,32000]......................................... ('fsdp', None).
I0420 19:33:04.156597 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Logical: float32[32000,2048]......................................... Unknown.
I0420 19:33:04.156658 136588928440128 gradient_accumulation.py:70] gradient_accumulation/inputs Physical: float32[32000,2048]......................................... (None, 'fsdp').
I0420 19:33:04.838996 136588928440128 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0420 19:33:04.839085 136588928440128 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0420 19:33:04.855389 136588928440128 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0420 19:33:04.855457 136588928440128 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0420 19:33:15.985301 136588928440128 max_utils.py:791] Total memory size: 1.7 GB, Output size: 0.4 GB, Temp size: 1.3 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0420 19:33:15.986203 136588928440128 metric_logger.py:289] number parameters: 1.104 billion
I0420 19:33:18.110027 136588928440128 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0420 19:33:28.309561 136588928440128 checkpointing.py:798] Waited 10.199514389038086 seconds for step 0 to finish before starting checkpointing.
I0420 19:33:28.312143 136588928440128 checkpoint_manager.py:2013] [process=1][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 19:33:28.313875 136588928440128 checkpoint_manager.py:1518] [process=1] Saving checkpoint at step 0
I0420 19:33:28.317746 136588928440128 async_checkpointer.py:452] [process=1] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_06_grad_accum/checkpoints/0.
I0420 19:33:28.900827 136588928440128 signaling_client.py:364] Using JaxDistributedSignalingClient
I0420 19:33:28.901947 136588928440128 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array.
I0420 19:33:28.902007 136588928440128 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
I0420 19:33:29.144021 136588928440128 base_pytree_checkpoint_handler.py:153] [process=1][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.243274s
I0420 19:33:29.144194 136588928440128 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.097 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.49809765815734863 s) (per-host)
I0420 19:33:29.144246 136588928440128 base_pytree_checkpoint_handler.py:737] [process=1][thread=MainThread] Initiated Pytree async_save. Time taken: 0.498159s (batch_requests_ready=0.251341s, total_serialization_initiated=0.246750s, others=0.000068s)
I0420 19:33:29.144335 136588928440128 composite_checkpoint_handler.py:715] [process=1][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.502238s (all_items=0.000019s, per_item={'items': '0.00001931'}, temp_paths=0.502218)
I0420 19:33:29.145271 136466034304768 async_checkpointer.py:79] [process=1][thread=async_save] Background save thread started.
I0420 19:33:29.145448 136588928440128 async_checkpointer.py:561] Finished blocking save. Time taken: 0.831501s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_06_grad_accum/checkpoints/0.
I0420 19:33:29.185034 136588928440128 checkpoint_manager.py:1566] [process=1][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 19:33:29.185337 136462813079296 async_checkpointer.py:265] [process=1][thread=save_finalize] Waiting for background save thread=async_save.
I0420 19:33:29.185513 136588928440128 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_06_grad_accum/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776713608.3121247, 'wait_for_prev_duration_secs': 6.198883056640625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776713608.3139145, 'checkpointer_blocking_duration_secs': 0.8316383361816406, 'get_old_steps_start_time': 1776713609.1455722, 'get_old_steps_duration_secs': 2.47955322265625e-05, 'checkpoint_manager_blocking_start_time': 1776713608.3100672, 'checkpoint_manager_blocking_duration_secs': 0.8754069805145264}
I0420 19:33:29.185620 136588928440128 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0420 19:33:29.185672 136588928440128 max_utils.py:750] 
Memstats: After params initialized:
I0420 19:33:29.185722 136588928440128 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_2(process=1,(2,0,0,0))
I0420 19:33:29.185757 136588928440128 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_3(process=1,(3,0,0,0))
I0420 19:33:29.185786 136588928440128 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_6(process=1,(2,1,0,0))
I0420 19:33:29.185811 136588928440128 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_7(process=1,(3,1,0,0))
I0420 19:33:29.446970 136588928440128 metric_logger.py:185] completed step: 0, seconds: 2.124, TFLOP/s/device: 25.591, Tokens/s/device: 3857.390, total_weights: 262144, loss: 10.877
I0420 19:33:30.051732 136588928440128 metric_logger.py:185] completed step: 1, seconds: 11.336, TFLOP/s/device: 4.794, Tokens/s/device: 722.670, total_weights: 262144, loss: 10.877
I0420 19:33:30.630294 136588928440128 metric_logger.py:185] completed step: 2, seconds: 0.026, TFLOP/s/device: 2059.358, Tokens/s/device: 310408.852, total_weights: 262144, loss: 10.563
I0420 19:33:31.209026 136588928440128 metric_logger.py:185] completed step: 3, seconds: 0.584, TFLOP/s/device: 93.076, Tokens/s/device: 14029.415, total_weights: 262144, loss: 10.272
I0420 19:33:31.819901    2370 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0420 19:33:32.366281 136588928440128 metric_logger.py:185] completed step: 4, seconds: 0.579, TFLOP/s/device: 93.917, Tokens/s/device: 14156.160, total_weights: 262144, loss: 10.022
I0420 19:33:32.371821 136588928440128 metric_logger.py:185] completed step: 5, seconds: 0.579, TFLOP/s/device: 93.900, Tokens/s/device: 14153.616, total_weights: 262144, loss: 9.820
I0420 19:33:33.759203 136466042697472 array_metadata_store.py:203] [process=1][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_06_grad_accum/checkpoints/0/items/array_metadatas/process_1
I0420 19:34:06.288543 136466034304768 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/gbytes_per_sec: 41.963 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.64242196083069 s) (per-host)
I0420 19:34:06.288642 136466034304768 async_checkpointer.py:90] [process=1][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.143263s.
I0420 19:34:10.637470 136588928440128 metric_logger.py:185] completed step: 6, seconds: 1.158, TFLOP/s/device: 46.953, Tokens/s/device: 7077.285, total_weights: 262144, loss: 9.667
I0420 19:34:11.215724 136588928440128 metric_logger.py:185] completed step: 7, seconds: 37.687, TFLOP/s/device: 1.442, Tokens/s/device: 217.370, total_weights: 262144, loss: 9.561
I0420 19:34:11.794228 136588928440128 metric_logger.py:185] completed step: 8, seconds: 0.584, TFLOP/s/device: 93.062, Tokens/s/device: 14027.349, total_weights: 262144, loss: 9.496
I0420 19:34:12.372312 136588928440128 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0420 19:34:12.372950 136588928440128 checkpointing.py:798] Waited 0.0006580352783203125 seconds for step 9 to finish before starting checkpointing.
I0420 19:34:12.375130 136588928440128 checkpoint_manager.py:2024] [process=1][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 19:34:49.965987 136466034304768 async_checkpointer.py:144] [process=1][thread=async_save] Background save thread done. Time taken: 80.820592s.
I0420 19:34:49.966343 136462813079296 async_checkpointer.py:273] [process=1][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 19:34:49.966493 136462813079296 async_checkpointer.py:283] [process=1][thread=save_finalize] No errors found in background save thread=async_save.
I0420 19:34:49.966547 136462813079296 checkpoint_manager.py:2133] [process=1][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0420 19:34:49.967885 136462813079296 checkpoint_manager.py:2142] [process=1][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0420 19:34:49.968049 136588928440128 checkpoint_manager.py:2036] [process=1][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0420 19:34:49.968198 136588928440128 checkpoint_manager.py:1458] Waiting for previous save to complete took 37.593078 seconds. If this number is high, consider checkpointing less frequently.
I0420 19:34:49.970166 136588928440128 checkpoint_manager.py:1518] [process=1] Saving checkpoint at step 9
I0420 19:34:49.973850 136588928440128 async_checkpointer.py:452] [process=1] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_06_grad_accum/checkpoints/9.
I0420 19:34:50.528541 136588928440128 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array.
I0420 19:34:50.528640 136588928440128 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
I0420 19:34:50.571053 136588928440128 base_pytree_checkpoint_handler.py:153] [process=1][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.043433s
I0420 19:34:50.571219 136588928440128 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.139 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.30019116401672363 s) (per-host)
I0420 19:34:50.571271 136588928440128 base_pytree_checkpoint_handler.py:737] [process=1][thread=MainThread] Initiated Pytree async_save. Time taken: 0.300251s (batch_requests_ready=0.254882s, total_serialization_initiated=0.045302s, others=0.000067s)
I0420 19:34:50.571359 136588928440128 composite_checkpoint_handler.py:715] [process=1][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.304205s (all_items=0.000015s, per_item={'items': '0.00001502'}, temp_paths=0.304190)
I0420 19:34:50.572229 136462804686592 async_checkpointer.py:79] [process=1][thread=async_save] Background save thread started.
I0420 19:34:50.572384 136588928440128 async_checkpointer.py:561] Finished blocking save. Time taken: 0.602147s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_06_grad_accum/checkpoints/9.
I0420 19:34:50.984999 136588928440128 checkpoint_manager.py:1566] [process=1][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 19:34:50.985364 136462813079296 async_checkpointer.py:265] [process=1][thread=save_finalize] Waiting for background save thread=async_save.
I0420 19:34:50.985558 136588928440128 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_06_grad_accum/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776713652.3750954, 'wait_for_prev_duration_secs': 37.593077659606934, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776713689.9702046, 'checkpointer_blocking_duration_secs': 0.6023013591766357, 'get_old_steps_start_time': 1776713690.5725234, 'get_old_steps_duration_secs': 2.574920654296875e-05, 'checkpoint_manager_blocking_start_time': 1776713652.3731701, 'checkpoint_manager_blocking_duration_secs': 38.61235070228577}
I0420 19:34:50.985663 136588928440128 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0420 19:34:50.985705 136588928440128 checkpoint_manager.py:2024] [process=1][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 19:34:55.754142 136466042697472 array_metadata_store.py:203] [process=1][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_06_grad_accum/checkpoints/9/items/array_metadatas/process_1
I0420 19:35:32.385417 136462804686592 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/gbytes_per_sec: 37.507 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.114333629608154 s) (per-host)
I0420 19:35:32.385535 136462804686592 async_checkpointer.py:90] [process=1][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.813197s.
I0420 19:35:39.573629 136462804686592 async_checkpointer.py:144] [process=1][thread=async_save] Background save thread done. Time taken: 49.001277s.
I0420 19:35:39.573931 136462813079296 async_checkpointer.py:273] [process=1][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 19:35:39.574047 136462813079296 async_checkpointer.py:283] [process=1][thread=save_finalize] No errors found in background save thread=async_save.
I0420 19:35:39.574093 136462813079296 checkpoint_manager.py:2133] [process=1][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0420 19:35:39.575631 136462813079296 checkpoint_manager.py:2142] [process=1][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0420 19:35:39.575824 136588928440128 checkpoint_manager.py:2036] [process=1][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0420 19:35:39.576059 136588928440128 checkpoint_manager.py:2013] [process=1][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 19:35:39.576659 136588928440128 metric_logger.py:185] completed step: 9, seconds: 0.578, TFLOP/s/device: 93.978, Tokens/s/device: 14165.315, total_weights: 262144, loss: 9.457
Per train step:
 Total TFLOPs: 54.35 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Mon Apr 20 19:35:50 UTC 2026
EXIT_CODE=0