MaxView

XPK Start: Sat Apr 25 20:24:06 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
`rope_parameters`'s factor field must be a float >= 1, got 40
`rope_parameters`'s beta_fast field must be a float, got 32
`rope_parameters`'s beta_slow field must be a float, got 1
DeepseekV32Config got `key=rope_scaling` in kwargs but hasn't set it as attribute. For RoPE standardization you need to set `self.rope_parameters` in model's config. 
2026-04-25 20:24:31.151248: 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)
I0425 20:24:31.359945 132128585103168 max_utils.py:273] Attempting to initialize the jax distributed system...
I0425 20:24:40.400497 132128585103168 distributed.py:149] Starting JAX distributed service on [::]:8482
I0425 20:24:40.402863 132128585103168 distributed.py:172] Connecting to JAX distributed service on mt-06-grad-accum-zut4r-slice-job-0-0.mt-06-grad-accum-zut4r:8482
I0425 20:24:41.856747 132128585103168 max_utils.py:284] Jax distributed system initialized!
I0425 20:24:46.832417 132128585103168 max_utils.py:800] System Information: Jax Version: 0.9.2
I0425 20:24:46.832527 132128585103168 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0425 20:24:46.832566 132128585103168 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365
I0425 20:24:46.832602 132128585103168 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0425 20:24:47.531771 132128585103168 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0425 20:24:47.532048 132128585103168 checkpointing.py:677] Setting up checkpoint logger...
I0425 20:24:47.532119 132128585103168 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0425 20:24:47.532166 132128585103168 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0425 20:24:47.532506 132128585103168 base_pytree_checkpoint_handler.py:441] 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 0x782adde48530>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0425 20:24:50.440966 132128585103168 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0425 20:24:50.441226 132128585103168 checkpoint_manager.py:708] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78176c31b8c0>}, handler_registry=None
I0425 20:24:50.441464 132128585103168 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78176c31b8c0>` 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`.
I0425 20:24:50.441518 132128585103168 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78176c50bb00>` 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`.
I0425 20:24:50.441552 132128585103168 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 0x78176c31b8c0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78176c31b8c0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78176c50bb00>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78176c50bb00>}).
I0425 20:24:50.441867 132128585103168 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0425 20:24:50.441937 132128585103168 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7816547ef2e0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0425 20:24:51.944906 132128585103168 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_06_grad_accum/checkpoints
I0425 20:24:52.397874 132128585103168 checkpoint_manager.py:929] [process=6][thread=MainThread] CheckpointManager created,  primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_06_grad_accum/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x78176c31b3b0>
I0425 20:24:52.398053 132128585103168 checkpointing.py:301] Checkpoint manager created!
I0425 20:24:54.128715 132128585103168 nnx_wrappers.py:453] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0425 20:24:54.128833 132128585103168 nnx_wrappers.py:453] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0425 20:24:54.512712 132128585103168 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0425 20:24:54.512808 132128585103168 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0425 20:24:54.529304 132128585103168 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0425 20:24:54.529365 132128585103168 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0425 20:24:54.552987 132128585103168 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0425 20:24:54.553056 132128585103168 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 20:24:54.569546 132128585103168 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0425 20:24:54.569607 132128585103168 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 20:24:54.586060 132128585103168 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0425 20:24:54.586133 132128585103168 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 20:24:54.611032 132128585103168 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0425 20:24:54.611123 132128585103168 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 20:24:54.632322 132128585103168 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0425 20:24:54.632384 132128585103168 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0425 20:24:54.847301 132128585103168 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0425 20:24:54.847420 132128585103168 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0425 20:24:56.291841 132128585103168 maxtext_utils.py:1668]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0425 20:24:56.291969 132128585103168 maxtext_utils.py:1668]  params/params/decoder/layers/mlp/wi_0/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0425 20:24:56.292023 132128585103168 maxtext_utils.py:1668]  params/params/decoder/layers/mlp/wi_1/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0425 20:24:56.292080 132128585103168 maxtext_utils.py:1668]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   P('mlp', 'layers', 'embed')
    Physical:  (None, None, 'fsdp')
I0425 20:24:56.292155 132128585103168 maxtext_utils.py:1668]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0425 20:24:56.292195 132128585103168 maxtext_utils.py:1668]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0425 20:24:56.292247 132128585103168 maxtext_utils.py:1668]  params/params/decoder/layers/self_attention/key/kernel
    Shape:     float32[2048,16,16,128]
    Logical:   P('embed', 'layers', 'kv_heads', 'kv_head_dim')
    Physical:  ('fsdp', None, None, None)
I0425 20:24:56.292298 132128585103168 maxtext_utils.py:1668]  params/params/decoder/layers/self_attention/out/kernel
    Shape:     float32[16,16,128,2048]
    Logical:   P('heads', 'layers', 'kv', 'embed')
    Physical:  (None, None, None, 'fsdp')
I0425 20:24:56.292345 132128585103168 maxtext_utils.py:1668]  params/params/decoder/layers/self_attention/query/kernel
    Shape:     float32[2048,16,16,128]
    Logical:   P('embed', 'layers', 'q_heads', 'kv')
    Physical:  ('fsdp', None, None, None)
I0425 20:24:56.292382 132128585103168 maxtext_utils.py:1668]  params/params/decoder/layers/self_attention/value/kernel
    Shape:     float32[2048,16,16,128]
    Logical:   P('embed', 'layers', 'kv_heads', 'kv_head_dim')
    Physical:  ('fsdp', None, None, None)
I0425 20:24:56.292426 132128585103168 maxtext_utils.py:1668]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0425 20:24:56.292473 132128585103168 maxtext_utils.py:1668]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')

I0425 20:24:56.314898 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048]............................................... Unknown.
I0425 20:24:56.314963 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048]............................................... (None,).
I0425 20:24:56.330000 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048,16,7168]....................................... Unknown.
I0425 20:24:56.330053 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048,16,7168]....................................... ('fsdp', None, None).
I0425 20:24:56.360060 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[7168,16,2048]....................................... Unknown.
I0425 20:24:56.360140 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[7168,16,2048]....................................... (None, None, 'fsdp').
I0425 20:24:56.375223 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048,16]............................................ Unknown.
I0425 20:24:56.375279 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048,16]............................................ (None, None).
I0425 20:24:56.405349 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048,16,16,128]..................................... Unknown.
I0425 20:24:56.405409 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048,16,16,128]..................................... ('fsdp', None, None, None).
I0425 20:24:56.420384 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[16,16,128,2048]..................................... Unknown.
I0425 20:24:56.420438 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[16,16,128,2048]..................................... (None, None, None, 'fsdp').
I0425 20:24:56.471554 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048,32000]......................................... Unknown.
I0425 20:24:56.471650 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048,32000]......................................... ('fsdp', None).
I0425 20:24:56.486723 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[32000,2048]......................................... Unknown.
I0425 20:24:56.486778 132128585103168 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[32000,2048]......................................... (None, 'fsdp').
I0425 20:24:57.162975 132128585103168 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0425 20:24:57.163074 132128585103168 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0425 20:24:57.178818 132128585103168 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0425 20:24:57.178879 132128585103168 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0425 20:25:08.406000 132128585103168 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.
I0425 20:25:08.406846 132128585103168 metric_logger.py:301] number parameters: 1.104 billion
I0425 20:25:10.349760 132128585103168 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0425 20:25:21.701931 132128585103168 checkpointing.py:776] Waited 11.352148532867432 seconds for step 0 to finish before starting checkpointing.
I0425 20:25:21.704545 132128585103168 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0425 20:25:21.706416 132128585103168 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0
I0425 20:25:21.708031 132128585103168 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_06_grad_accum/checkpoints/0.
I0425 20:25:22.454822 132128585103168 signaling_client.py:364] Using JaxDistributedSignalingClient
I0425 20:25:22.455779 132128585103168 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0425 20:25:22.455834 132128585103168 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
I0425 20:25:22.725801 132128585103168 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.271054s
I0425 20:25:22.725969 132128585103168 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.581 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2763814926147461 s) (per-host)
I0425 20:25:22.726022 132128585103168 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.276443s (batch_requests_ready=0.002179s, total_serialization_initiated=0.274194s, others=0.000070s)
I0425 20:25:22.726133 132128585103168 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.280706s (all_items=0.000019s, per_item={'items': '0.00001860'}, temp_paths=0.280687)
I0425 20:25:22.726933 132128585103168 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.02 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_06_grad_accum/checkpoints/0.
I0425 20:25:22.727241 132003560228608 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-25 20:45:22.727205
I0425 20:25:22.742353 132128585103168 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0425 20:25:22.742665 132000354658048 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0425 20:25:22.742823 132128585103168 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_06_grad_accum/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777148721.7045271, 'wait_for_prev_duration_secs': 6.103515625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777148721.7064545, 'checkpointer_blocking_duration_secs': 1.020895004272461, 'get_old_steps_start_time': 1777148722.7273734, 'get_old_steps_duration_secs': 2.9087066650390625e-05, 'checkpoint_manager_blocking_start_time': 1777148721.702489, 'checkpoint_manager_blocking_duration_secs': 1.0402958393096924}
I0425 20:25:22.742933 132128585103168 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0425 20:25:22.742983 132128585103168 max_utils.py:750] 
Memstats: After params initialized:
I0425 20:25:22.743031 132128585103168 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0))
I0425 20:25:22.743062 132128585103168 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0))
I0425 20:25:22.743088 132128585103168 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0))
I0425 20:25:22.743128 132128585103168 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0))
I0425 20:25:23.053195 132128585103168 metric_logger.py:196] completed step: 0, seconds: 1.943, TFLOP/s/device: 27.974, Tokens/s/device: 4216.603, total_weights: 262144, loss: 10.877, lm_loss: 10.877, perplexity: 52959.059
I0425 20:25:23.666367 132128585103168 metric_logger.py:196] completed step: 1, seconds: 12.702, TFLOP/s/device: 4.279, Tokens/s/device: 644.944, total_weights: 262144, loss: 10.877, lm_loss: 10.877, perplexity: 52959.059
I0425 20:25:24.244688 132128585103168 metric_logger.py:196] completed step: 2, seconds: 0.035, TFLOP/s/device: 1540.665, Tokens/s/device: 232225.876, total_weights: 262144, loss: 10.563, lm_loss: 10.563, perplexity: 38662.707
I0425 20:25:24.822817 132128585103168 metric_logger.py:196] completed step: 3, seconds: 0.584, TFLOP/s/device: 93.060, Tokens/s/device: 14026.941, total_weights: 262144, loss: 10.272, lm_loss: 10.272, perplexity: 28909.668
I0425 20:25:25.979276 132128585103168 metric_logger.py:196] completed step: 4, seconds: 0.578, TFLOP/s/device: 94.004, Tokens/s/device: 14169.333, total_weights: 262144, loss: 10.022, lm_loss: 10.022, perplexity: 22524.992
I0425 20:25:25.985138 132128585103168 metric_logger.py:196] completed step: 5, seconds: 0.578, TFLOP/s/device: 93.982, Tokens/s/device: 14165.952, total_weights: 262144, loss: 9.820, lm_loss: 9.820, perplexity: 18401.865
I0425 20:25:26.761036    2555 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0425 20:25:28.750458 132001403459328 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_06_grad_accum/checkpoints/0/items/array_metadatas/process_6
I0425 20:25:59.396044 132003560228608 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 42.754 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 36.94641733169556 s) (per-host)
I0425 20:25:59.396197 132003560228608 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.668328s.
I0425 20:26:10.017383 132003560228608 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 47.289497s.
I0425 20:26:10.017704 132000354658048 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0425 20:26:10.017823 132000354658048 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0425 20:26:10.017874 132000354658048 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0425 20:26:10.019625 132000354658048 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0425 20:26:12.537189 132128585103168 metric_logger.py:196] completed step: 6, seconds: 1.157, TFLOP/s/device: 46.983, Tokens/s/device: 7081.764, total_weights: 262144, loss: 9.667, lm_loss: 9.667, perplexity: 15787.604
I0425 20:26:13.115464 132128585103168 metric_logger.py:196] completed step: 7, seconds: 45.974, TFLOP/s/device: 1.182, Tokens/s/device: 178.187, total_weights: 262144, loss: 9.561, lm_loss: 9.561, perplexity: 14203.827
I0425 20:26:13.693358 132128585103168 metric_logger.py:196] completed step: 8, seconds: 0.583, TFLOP/s/device: 93.161, Tokens/s/device: 14042.209, total_weights: 262144, loss: 9.496, lm_loss: 9.496, perplexity: 13302.920
I0425 20:26:14.270890 132128585103168 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0425 20:26:14.271594 132128585103168 checkpointing.py:776] Waited 0.0007233619689941406 seconds for step 9 to finish before starting checkpointing.
I0425 20:26:14.273540 132128585103168 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0425 20:26:14.275281 132128585103168 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9
I0425 20:26:14.276719 132128585103168 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_06_grad_accum/checkpoints/9.
I0425 20:26:14.991191 132128585103168 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0425 20:26:14.991280 132128585103168 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
I0425 20:26:15.030649 132128585103168 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.040353s
I0425 20:26:15.030831 132128585103168 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 34.743 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04439949989318848 s) (per-host)
I0425 20:26:15.030882 132128585103168 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.044461s (batch_requests_ready=0.001625s, total_serialization_initiated=0.042768s, others=0.000068s)
I0425 20:26:15.030969 132128585103168 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.048611s (all_items=0.000016s, per_item={'items': '0.00001621'}, temp_paths=0.048595)
I0425 20:26:15.031727 132128585103168 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.76 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_06_grad_accum/checkpoints/9.
I0425 20:26:15.032049 132000354658048 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-25 20:46:15.032012
I0425 20:26:15.033983 132128585103168 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0425 20:26:15.034216 132003059398400 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0425 20:26:15.034350 132128585103168 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_06_grad_accum/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777148774.2735069, 'wait_for_prev_duration_secs': 7.915496826171875e-05, 'time_between_consecutive_saves_sec': 4.2538416385650635, 'checkpointer_blocking_start_time': 1777148774.2753212, 'checkpointer_blocking_duration_secs': 0.7568788528442383, 'get_old_steps_start_time': 1777148775.0322247, 'get_old_steps_duration_secs': 2.8371810913085938e-05, 'checkpoint_manager_blocking_start_time': 1777148774.2718186, 'checkpoint_manager_blocking_duration_secs': 0.7624976634979248}
I0425 20:26:15.034504 132128585103168 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0425 20:26:15.034547 132128585103168 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0425 20:26:20.843640 131997137618688 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_06_grad_accum/checkpoints/9/items/array_metadatas/process_6
I0425 20:26:57.180955 132000354658048 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 37.436 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.19448399543762 s) (per-host)
I0425 20:26:57.181084 132000354658048 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.148913s.
I0425 20:27:04.132558 132000354658048 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 49.100371s.
I0425 20:27:04.132845 132003059398400 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0425 20:27:04.132966 132003059398400 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0425 20:27:04.133013 132003059398400 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0425 20:27:04.134499 132003059398400 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0425 20:27:04.134698 132128585103168 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0425 20:27:04.134848 132128585103168 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0425 20:27:04.135771 132128585103168 metric_logger.py:196] completed step: 9, seconds: 0.578, TFLOP/s/device: 93.996, Tokens/s/device: 14168.157, total_weights: 262144, loss: 9.457, lm_loss: 9.457, perplexity: 12802.546
Per train step:
 Total TFLOPs: 54.35 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Sat Apr 25 20:27:16 UTC 2026
EXIT_CODE=0