MaxView

XPK Start: Sat Apr 25 21:04:51 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 21:05:16.480865: 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 21:05:16.692736 132448011401024 max_utils.py:273] Attempting to initialize the jax distributed system...
I0425 21:05:25.732800 132448011401024 distributed.py:149] Starting JAX distributed service on [::]:8482
I0425 21:05:25.735193 132448011401024 distributed.py:172] Connecting to JAX distributed service on mt-15-ocdbt-true-sv-lxqrg-slice-job-0-0.mt-15-ocdbt-true-sv-lxqrg:8482
I0425 21:05:26.711502 132448011401024 max_utils.py:284] Jax distributed system initialized!
I0425 21:05:32.838161 132448011401024 max_utils.py:800] System Information: Jax Version: 0.9.2
I0425 21:05:32.838266 132448011401024 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0425 21:05:32.838307 132448011401024 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 21:05:32.838343 132448011401024 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0425 21:05:33.536073 132448011401024 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0425 21:05:33.536345 132448011401024 checkpointing.py:677] Setting up checkpoint logger...
I0425 21:05:33.536401 132448011401024 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0425 21:05:33.536445 132448011401024 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0425 21:05:33.536790 132448011401024 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 0x787553b18bf0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0425 21:05:36.450228 132448011401024 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0425 21:05:36.450457 132448011401024 checkpoint_manager.py:708] [process=3][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7861ac6c75f0>}, handler_registry=None
I0425 21:05:36.450696 132448011401024 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7861ac6c75f0>` 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 21:05:36.450744 132448011401024 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7861ac6c8740>` 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 21:05:36.450781 132448011401024 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 0x7861ac6c75f0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7861ac6c75f0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7861ac6c8740>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7861ac6c8740>}).
I0425 21:05:36.451124 132448011401024 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0425 21:05:36.451194 132448011401024 async_checkpointer.py:192] [process=3][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7861ac4df2e0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0425 21:05:37.609258 132448011401024 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_15_ocdbt_true/checkpoints
I0425 21:05:38.434069 132448011401024 checkpoint_manager.py:929] [process=3][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=5), 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_15_ocdbt_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7861ac6c7920>
I0425 21:05:38.434245 132448011401024 checkpointing.py:301] Checkpoint manager created!
I0425 21:05:39.366145 132448011401024 nnx_wrappers.py:453] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0425 21:05:39.366261 132448011401024 nnx_wrappers.py:453] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0425 21:05:39.753498 132448011401024 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0425 21:05:39.753590 132448011401024 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0425 21:05:39.770315 132448011401024 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0425 21:05:39.770375 132448011401024 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0425 21:05:39.794101 132448011401024 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 21:05:39.794172 132448011401024 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 21:05:39.810777 132448011401024 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 21:05:39.810836 132448011401024 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 21:05:39.827396 132448011401024 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 21:05:39.827456 132448011401024 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 21:05:39.852310 132448011401024 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0425 21:05:39.852380 132448011401024 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 21:05:39.873517 132448011401024 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0425 21:05:39.873589 132448011401024 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0425 21:05:40.083746 132448011401024 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0425 21:05:40.083852 132448011401024 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0425 21:05:41.509003 132448011401024 maxtext_utils.py:1668]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0425 21:05:41.509127 132448011401024 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 21:05:41.509179 132448011401024 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 21:05:41.509244 132448011401024 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 21:05:41.509295 132448011401024 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 21:05:41.509334 132448011401024 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 21:05:41.509386 132448011401024 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 21:05:41.509443 132448011401024 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 21:05:41.509483 132448011401024 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 21:05:41.509518 132448011401024 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 21:05:41.509565 132448011401024 maxtext_utils.py:1668]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0425 21:05:41.509611 132448011401024 maxtext_utils.py:1668]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')

I0425 21:05:42.002388 132448011401024 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0425 21:05:42.002486 132448011401024 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0425 21:05:42.018074 132448011401024 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0425 21:05:42.018133 132448011401024 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0425 21:05:52.954830 132448011401024 max_utils.py:791] Total memory size: 1.5 GB, Output size: 0.4 GB, Temp size: 1.1 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0425 21:05:52.955558 132448011401024 metric_logger.py:301] number parameters: 1.104 billion
I0425 21:06:04.230443 132448011401024 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0425 21:06:04.682018 132448011401024 checkpointing.py:776] Waited 0.4515526294708252 seconds for step 0 to finish before starting checkpointing.
I0425 21:06:04.684407 132448011401024 checkpoint_manager.py:2009] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0425 21:06:04.686015 132448011401024 checkpoint_manager.py:1512] [process=3] Saving checkpoint at step 0
I0425 21:06:04.687450 132448011401024 event_tracking.py:70] [process=3] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_15_ocdbt_true/checkpoints/0.
I0425 21:06:05.427533 132448011401024 signaling_client.py:364] Using JaxDistributedSignalingClient
I0425 21:06:05.428487 132448011401024 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0425 21:06:05.428543 132448011401024 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 21:06:05.709507 132448011401024 base_pytree_checkpoint_handler.py:154] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.282053s
I0425 21:06:05.709675 132448011401024 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/blocking_gbytes_per_sec: 5.358 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28789591789245605 s) (per-host)
I0425 21:06:05.709728 132448011401024 base_pytree_checkpoint_handler.py:768] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.287958s (batch_requests_ready=0.002637s, total_serialization_initiated=0.285250s, others=0.000070s)
I0425 21:06:05.709822 132448011401024 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.292313s (all_items=0.000019s, per_item={'items': '0.00001860'}, temp_paths=0.292295)
I0425 21:06:05.710627 132448011401024 event_tracking.py:125] [process=3] [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_15_ocdbt_true/checkpoints/0.
I0425 21:06:05.710977 132323127224064 async_checkpointer.py:76] [process=3][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-25 21:26:05.710941
I0425 21:06:05.713447 132448011401024 checkpoint_manager.py:1560] [process=3][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0425 21:06:05.713689 132322048915200 async_checkpointer.py:280] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0425 21:06:05.713838 132448011401024 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_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777151164.6843894, 'wait_for_prev_duration_secs': 6.151199340820312e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777151164.686052, 'checkpointer_blocking_duration_secs': 1.0250773429870605, 'get_old_steps_start_time': 1777151165.711154, 'get_old_steps_duration_secs': 2.7179718017578125e-05, 'checkpoint_manager_blocking_start_time': 1777151164.6824195, 'checkpoint_manager_blocking_duration_secs': 1.0313830375671387}
I0425 21:06:05.713959 132448011401024 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0425 21:06:05.714010 132448011401024 max_utils.py:750] 
Memstats: After params initialized:
I0425 21:06:05.714059 132448011401024 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_10(process=3,(2,2,0,0))
I0425 21:06:05.714092 132448011401024 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_11(process=3,(3,2,0,0))
I0425 21:06:05.714121 132448011401024 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_14(process=3,(2,3,0,0))
I0425 21:06:05.714149 132448011401024 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_15(process=3,(3,3,0,0))
I0425 21:06:06.024198 132448011401024 metric_logger.py:196] completed step: 0, seconds: 11.275, TFLOP/s/device: 1.205, Tokens/s/device: 181.644, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0425 21:06:06.213579 132448011401024 metric_logger.py:196] completed step: 1, seconds: 1.792, TFLOP/s/device: 7.581, Tokens/s/device: 1142.691, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0425 21:06:06.640588 132448011401024 metric_logger.py:196] completed step: 2, seconds: 0.042, TFLOP/s/device: 324.864, Tokens/s/device: 48967.100, total_weights: 65536, loss: 10.560, lm_loss: 10.560, perplexity: 38547.719
I0425 21:06:06.788023 132448011401024 metric_logger.py:196] completed step: 3, seconds: 0.428, TFLOP/s/device: 31.765, Tokens/s/device: 4787.944, total_weights: 65536, loss: 9.980, lm_loss: 9.980, perplexity: 21590.506
I0425 21:06:06.792644 132448011401024 checkpointing.py:772] Waiting for step 5 to finish before checkpoint...
I0425 21:06:07.082080 132448011401024 checkpointing.py:776] Waited 0.28940439224243164 seconds for step 5 to finish before starting checkpointing.
I0425 21:06:07.084988 132448011401024 checkpoint_manager.py:2020] [process=3][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0425 21:06:09.119374    2798 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0425 21:06:11.176373 132322091570944 array_metadata_store.py:203] [process=3][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_15_ocdbt_true/checkpoints/0/items/array_metadatas/process_3
I0425 21:06:42.693625 132323127224064 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/gbytes_per_sec: 42.380 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.271808385849 s) (per-host)
I0425 21:06:42.693747 132323127224064 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.982656s.
I0425 21:06:56.580450 132323127224064 async_checkpointer.py:160] [process=3][thread=async_save] Background save thread done. Time taken: 50.869333s.
I0425 21:06:56.580702 132322048915200 async_checkpointer.py:288] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0425 21:06:56.580828 132322048915200 async_checkpointer.py:298] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0425 21:06:56.580880 132322048915200 checkpoint_manager.py:2137] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0425 21:06:56.583408 132322048915200 checkpoint_manager.py:2146] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0425 21:06:56.583586 132448011401024 checkpoint_manager.py:2032] [process=3][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0425 21:06:56.583719 132448011401024 checkpoint_manager.py:1452] Waiting for previous save to complete took 49.498733 seconds. If this number is high, consider checkpointing less frequently.
I0425 21:06:56.585756 132448011401024 checkpoint_manager.py:1512] [process=3] Saving checkpoint at step 5
I0425 21:06:56.588174 132448011401024 event_tracking.py:70] [process=3] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_15_ocdbt_true/checkpoints/5.
I0425 21:06:57.731433 132448011401024 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0425 21:06:57.731528 132448011401024 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 21:06:57.766389 132448011401024 base_pytree_checkpoint_handler.py:154] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.035993s
I0425 21:06:57.766560 132448011401024 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/blocking_gbytes_per_sec: 39.033 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.039520263671875 s) (per-host)
I0425 21:06:57.766611 132448011401024 base_pytree_checkpoint_handler.py:768] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.039581s (batch_requests_ready=0.001874s, total_serialization_initiated=0.037637s, others=0.000069s)
I0425 21:06:57.766700 132448011401024 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.044554s (all_items=0.000015s, per_item={'items': '0.00001478'}, temp_paths=0.044540)
I0425 21:06:57.767363 132448011401024 event_tracking.py:125] [process=3] [async] Finished blocking save in 1.18 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_15_ocdbt_true/checkpoints/5.
I0425 21:06:57.767644 132317255292672 async_checkpointer.py:76] [process=3][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-25 21:26:57.767612
I0425 21:06:57.777238 132448011401024 checkpoint_manager.py:1560] [process=3][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize
I0425 21:06:57.777475 132322048915200 async_checkpointer.py:280] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0425 21:06:57.777601 132448011401024 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777151167.084959, 'wait_for_prev_duration_secs': 49.498732566833496, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777151216.585795, 'checkpointer_blocking_duration_secs': 1.1819970607757568, 'get_old_steps_start_time': 1777151217.7678118, 'get_old_steps_duration_secs': 2.6941299438476562e-05, 'checkpoint_manager_blocking_start_time': 1777151167.0827081, 'checkpoint_manager_blocking_duration_secs': 50.69486117362976}
I0425 21:06:57.777708 132448011401024 checkpointing.py:408] Started an asynchronous checkpoint save for step 5
I0425 21:06:57.778523 132448011401024 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.808, Tokens/s/device: 13386.146, total_weights: 65536, loss: 9.460, lm_loss: 9.460, perplexity: 12840.355
I0425 21:06:57.784333 132448011401024 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.258, Tokens/s/device: 13906.147, total_weights: 65536, loss: 8.959, lm_loss: 8.959, perplexity: 7776.467
I0425 21:07:10.171400 132322091570944 array_metadata_store.py:203] [process=3][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_15_ocdbt_true/checkpoints/5/items/array_metadatas/process_3
I0425 21:07:25.912241 132448011401024 metric_logger.py:196] completed step: 6, seconds: 50.990, TFLOP/s/device: 0.266, Tokens/s/device: 40.165, total_weights: 65536, loss: 8.469, lm_loss: 8.469, perplexity: 4762.740
I0425 21:07:26.059583 132448011401024 metric_logger.py:196] completed step: 7, seconds: 27.982, TFLOP/s/device: 0.486, Tokens/s/device: 73.190, total_weights: 65536, loss: 8.003, lm_loss: 8.003, perplexity: 2989.151
I0425 21:07:26.206971 132448011401024 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.339, Tokens/s/device: 13466.154, total_weights: 65536, loss: 7.572, lm_loss: 7.572, perplexity: 1943.098
I0425 21:07:26.353593 132448011401024 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0425 21:07:26.354305 132448011401024 checkpointing.py:776] Waited 0.0007317066192626953 seconds for step 9 to finish before starting checkpointing.
I0425 21:07:26.356277 132448011401024 checkpoint_manager.py:2020] [process=3][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0425 21:07:44.085257 132317255292672 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/gbytes_per_sec: 34.074 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 46.35817742347717 s) (per-host)
I0425 21:07:44.085364 132317255292672 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 46.317605s.
I0425 21:07:54.588842 132317255292672 async_checkpointer.py:160] [process=3][thread=async_save] Background save thread done. Time taken: 56.821068s.
I0425 21:07:54.589080 132322048915200 async_checkpointer.py:288] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0425 21:07:54.589137 132322048915200 async_checkpointer.py:298] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0425 21:07:54.589183 132322048915200 checkpoint_manager.py:2137] [process=3][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts...
I0425 21:07:54.591813 132322048915200 checkpoint_manager.py:2146] [process=3][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts.
I0425 21:07:54.591998 132448011401024 checkpoint_manager.py:2032] [process=3][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5.
W0425 21:07:54.592124 132448011401024 checkpoint_manager.py:1452] Waiting for previous save to complete took 28.235854 seconds. If this number is high, consider checkpointing less frequently.
I0425 21:07:54.593612 132448011401024 checkpoint_manager.py:1512] [process=3] Saving checkpoint at step 9
I0425 21:07:54.595895 132448011401024 event_tracking.py:70] [process=3] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_15_ocdbt_true/checkpoints/9.
I0425 21:07:54.880412 132448011401024 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0425 21:07:54.880507 132448011401024 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 21:07:54.915896 132448011401024 base_pytree_checkpoint_handler.py:154] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.036322s
I0425 21:07:54.916067 132448011401024 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/blocking_gbytes_per_sec: 38.733 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.039825439453125 s) (per-host)
I0425 21:07:54.916115 132448011401024 base_pytree_checkpoint_handler.py:768] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.039884s (batch_requests_ready=0.001769s, total_serialization_initiated=0.038052s, others=0.000063s)
I0425 21:07:54.916201 132448011401024 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.043910s (all_items=0.000011s, per_item={'items': '0.00001097'}, temp_paths=0.043899)
I0425 21:07:54.916912 132448011401024 event_tracking.py:125] [process=3] [async] Finished blocking save in 0.32 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260425_201227/linen_xpk_test_pipeline_scan_nnx_20260425_201227_15_ocdbt_true/checkpoints/9.
I0425 21:07:54.917240 132322048915200 async_checkpointer.py:76] [process=3][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-25 21:27:54.917204
I0425 21:07:54.922188 132448011401024 checkpoint_manager.py:1560] [process=3][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0425 21:07:54.922452 132322091570944 async_checkpointer.py:280] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0425 21:07:54.922611 132448011401024 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_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777151246.3562489, 'wait_for_prev_duration_secs': 28.235853672027588, 'time_between_consecutive_saves_sec': 29.772801637649536, 'checkpointer_blocking_start_time': 1777151274.593649, 'checkpointer_blocking_duration_secs': 0.32373499870300293, 'get_old_steps_start_time': 1777151274.9174018, 'get_old_steps_duration_secs': 2.2411346435546875e-05, 'checkpoint_manager_blocking_start_time': 1777151246.3545353, 'checkpoint_manager_blocking_duration_secs': 28.568044900894165}
I0425 21:07:54.922717 132448011401024 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0425 21:07:54.922762 132448011401024 checkpoint_manager.py:2020] [process=3][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0425 21:08:02.279642 132321516689152 array_metadata_store.py:203] [process=3][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_15_ocdbt_true/checkpoints/9/items/array_metadatas/process_3
I0425 21:08:37.878661 132322048915200 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/gbytes_per_sec: 36.733 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.00237965583801 s) (per-host)
I0425 21:08:37.878798 132322048915200 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.961447s.
I0425 21:08:45.236351 132322048915200 async_checkpointer.py:160] [process=3][thread=async_save] Background save thread done. Time taken: 50.318986s.
I0425 21:08:45.236651 132322091570944 async_checkpointer.py:288] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0425 21:08:45.236773 132322091570944 async_checkpointer.py:298] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0425 21:08:45.236815 132322091570944 checkpoint_manager.py:2137] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0425 21:08:45.239098 132322091570944 checkpoint_manager.py:2146] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0425 21:08:45.239280 132448011401024 checkpoint_manager.py:2032] [process=3][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0425 21:08:45.239442 132448011401024 checkpoint_manager.py:2009] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0425 21:08:45.240410 132448011401024 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.310, Tokens/s/device: 13913.989, total_weights: 65536, loss: 7.181, lm_loss: 7.181, perplexity: 1314.495
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Sat Apr 25 21:08:57 UTC 2026
EXIT_CODE=0