XPK Start: Tue Apr 21 07:20:33 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.
2026-04-21 07:20:58.704613: 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)
I0421 07:20:58.918594 135117366056768 max_utils.py:273] Attempting to initialize the jax distributed system...
I0421 07:21:07.970000 135117366056768 distributed.py:149] Starting JAX distributed service on [::]:8482
I0421 07:21:07.972845 135117366056768 distributed.py:172] Connecting to JAX distributed service on mt-15-ocdbt-true-rs-b37aa-slice-job-0-0.mt-15-ocdbt-true-rs-b37aa:8482
I0421 07:21:09.466885 135117366056768 max_utils.py:284] Jax distributed system initialized!
I0421 07:21:15.768846 135117366056768 max_utils.py:800] System Information: Jax Version: 0.9.2
I0421 07:21:15.768950 135117366056768 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0421 07:21:15.768991 135117366056768 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
I0421 07:21:15.769026 135117366056768 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0421 07:21:16.467812 135117366056768 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 07:21:16.468089 135117366056768 checkpointing.py:677] Setting up checkpoint logger...
I0421 07:21:16.468144 135117366056768 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0421 07:21:16.468188 135117366056768 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0421 07:21:16.468533 135117366056768 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 0x7ae326be9b80>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0421 07:21:17.750088 135117366056768 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0421 07:21:17.750327 135117366056768 checkpoint_manager.py:708] [process=7][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ace6c2b7a70>}, handler_registry=None
I0421 07:21:17.750569 135117366056768 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ace6c2b7a70>` 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`.
I0421 07:21:17.750620 135117366056768 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ace6c2bc2f0>` 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`.
I0421 07:21:17.750660 135117366056768 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 0x7ace6c2b7a70>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ace6c2b7a70>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ace6c2bc2f0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ace6c2bc2f0>}).
I0421 07:21:17.751010 135117366056768 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0421 07:21:17.751086 135117366056768 async_checkpointer.py:192] [process=7][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7acd787e1580> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0421 07:21:19.588638 134995746203392 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776755815390245600, 'commit_timestamp_nsecs': 1776755858135992787, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/0/_CHECKPOINT_METADATA
I0421 07:21:19.839637 134995746203392 checkpoint_manager.py:1871] Missing metrics for step 0
I0421 07:21:20.279444 135117366056768 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints
I0421 07:21:20.301360 135117366056768 checkpoint_manager.py:929] [process=7][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_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7ace6c2ba0f0>
I0421 07:21:20.301497 135117366056768 checkpointing.py:301] Checkpoint manager created!
I0421 07:21:21.251026 135117366056768 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0421 07:21:21.251148 135117366056768 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 07:21:21.633355 135117366056768 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 07:21:21.633452 135117366056768 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 07:21:21.649961 135117366056768 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 07:21:21.650026 135117366056768 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 07:21:21.673686 135117366056768 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0421 07:21:21.673765 135117366056768 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 07:21:21.690264 135117366056768 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0421 07:21:21.690326 135117366056768 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 07:21:21.706700 135117366056768 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0421 07:21:21.706776 135117366056768 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 07:21:21.731767 135117366056768 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0421 07:21:21.731839 135117366056768 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 07:21:21.752941 135117366056768 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0421 07:21:21.753016 135117366056768 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0421 07:21:21.963941 135117366056768 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0421 07:21:21.964056 135117366056768 checkpointing.py:581] restoring from this run's directory step 9
I0421 07:21:22.216640 135117366056768 event_tracking.py:70] [process=7] [sync] Started load checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/9.
I0421 07:21:22.469499 135117366056768 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/9.
I0421 07:21:22.469952 135117366056768 event_tracking.py:125] [process=7] [sync] Finished blocking load in 0.25 seconds. Continuing load @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/9.
I0421 07:21:23.903826 1461 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0421 07:21:30.954233 135117366056768 jax_array_handlers.py:843] [process=7] /jax/orbax/read/worker/io/requested throughput: 223.953 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 7.056260347366333 s) (per-host)
I0421 07:21:30.960073 135117366056768 base_pytree_checkpoint_handler.py:130] [process=7] /jax/checkpoint/read/gbytes_per_sec: 211.257 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 7.491331338882446 s) (per-host)
I0421 07:21:31.458312 135117366056768 event_tracking.py:138] [process=7] [sync] Finished load in 9.24 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/9
I0421 07:21:31.458471 135117366056768 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints', 'checkpointer_start_time': 1776756082.2165315, 'checkpointer_duration_secs': 9.241907358169556, 'checkpoint_manager_start_time': 1776756081.9642992, 'checkpoint_manager_duration_secs': 9.494142055511475}
fsdp: 32
I0421 07:21:31.880061 135117366056768 maxtext_utils.py:1668] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0421 07:21:31.880228 135117366056768 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)
I0421 07:21:31.880307 135117366056768 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)
I0421 07:21:31.880398 135117366056768 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')
I0421 07:21:31.880481 135117366056768 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)
I0421 07:21:31.880540 135117366056768 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)
I0421 07:21:31.880617 135117366056768 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)
I0421 07:21:31.880693 135117366056768 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')
I0421 07:21:31.880776 135117366056768 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)
I0421 07:21:31.880846 135117366056768 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)
I0421 07:21:31.880920 135117366056768 maxtext_utils.py:1668] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0421 07:21:31.880993 135117366056768 maxtext_utils.py:1668] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0421 07:21:32.369723 135117366056768 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 07:21:32.369845 135117366056768 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 07:21:32.385478 135117366056768 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 07:21:32.385541 135117366056768 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 07:21:43.541895 135117366056768 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.
I0421 07:21:43.542811 135117366056768 metric_logger.py:301] number parameters: 1.104 billion
I0421 07:21:55.009855 135117366056768 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0421 07:21:55.164096 135117366056768 checkpointing.py:776] Waited 0.15421247482299805 seconds for step 10 to finish before starting checkpointing.
I0421 07:21:55.166661 135117366056768 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 07:21:55.168601 135117366056768 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 10
I0421 07:21:55.170169 135117366056768 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/10.
I0421 07:21:55.907931 135117366056768 signaling_client.py:364] Using JaxDistributedSignalingClient
I0421 07:21:55.908949 135117366056768 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 07:21:55.909008 135117366056768 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
I0421 07:21:56.199403 135117366056768 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.291547s
I0421 07:21:56.199578 135117366056768 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 5.221 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2954590320587158 s) (per-host)
I0421 07:21:56.199633 135117366056768 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.295524s (batch_requests_ready=0.002174s, total_serialization_initiated=0.293277s, others=0.000073s)
I0421 07:21:56.199720 135117366056768 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.300202s (all_items=0.000017s, per_item={'items': '0.00001717'}, temp_paths=0.300185)
I0421 07:21:56.200498 135117366056768 event_tracking.py:125] [process=7] [async] Finished blocking save in 1.03 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/10.
I0421 07:21:56.200807 134995142223616 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 07:41:56.200775
I0421 07:21:56.202702 135117366056768 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 07:21:56.203012 134993028310784 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save.
I0421 07:21:56.203146 135117366056768 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776756115.166643, 'wait_for_prev_duration_secs': 6.198883056640625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776756115.1686423, 'checkpointer_blocking_duration_secs': 1.032320261001587, 'get_old_steps_start_time': 1776756116.2009852, 'get_old_steps_duration_secs': 3.24249267578125e-05, 'checkpoint_manager_blocking_start_time': 1776756115.1646411, 'checkpoint_manager_blocking_duration_secs': 1.0384712219238281}
I0421 07:21:56.203301 135117366056768 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0421 07:21:56.203351 135117366056768 max_utils.py:750]
Memstats: After params initialized:
I0421 07:21:56.203400 135117366056768 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_26(process=7,(2,6,0,0))
I0421 07:21:56.203432 135117366056768 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_27(process=7,(3,6,0,0))
I0421 07:21:56.203459 135117366056768 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_30(process=7,(2,7,0,0))
I0421 07:21:56.203485 135117366056768 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_31(process=7,(3,7,0,0))
I0421 07:21:56.516851 135117366056768 metric_logger.py:196] completed step: 10, seconds: 11.467, TFLOP/s/device: 1.185, Tokens/s/device: 178.601, total_weights: 65536, loss: 6.833, lm_loss: 6.833, perplexity: 927.869
I0421 07:21:56.685648 135117366056768 metric_logger.py:196] completed step: 11, seconds: 1.505, TFLOP/s/device: 9.025, Tokens/s/device: 1360.404, total_weights: 65536, loss: 6.527, lm_loss: 6.527, perplexity: 683.659
I0421 07:21:56.832964 135117366056768 metric_logger.py:196] completed step: 12, seconds: 0.021, TFLOP/s/device: 635.358, Tokens/s/device: 95768.062, total_weights: 65536, loss: 6.265, lm_loss: 6.265, perplexity: 526.068
I0421 07:21:56.980589 135117366056768 metric_logger.py:196] completed step: 13, seconds: 0.153, TFLOP/s/device: 88.649, Tokens/s/device: 13362.128, total_weights: 65536, loss: 6.045, lm_loss: 6.045, perplexity: 422.001
I0421 07:21:56.984864 135117366056768 checkpointing.py:772] Waiting for step 15 to finish before checkpoint...
I0421 07:21:57.274556 135117366056768 checkpointing.py:776] Waited 0.2896571159362793 seconds for step 15 to finish before starting checkpointing.
I0421 07:21:57.277688 135117366056768 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 07:22:00.932679 134993036703488 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/10/items/array_metadatas/process_7
I0421 07:22:38.129878 134995142223616 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 37.408 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.22572135925293 s) (per-host)
I0421 07:22:38.130013 134995142223616 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.929092s.
I0421 07:22:45.953472 134995142223616 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 49.752532s.
I0421 07:22:45.953765 134993028310784 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 07:22:45.953883 134993028310784 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save.
I0421 07:22:45.953934 134993028310784 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0421 07:22:45.955294 134993028310784 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0421 07:22:45.955458 135117366056768 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10.
W0421 07:22:45.955593 135117366056768 checkpoint_manager.py:1452] Waiting for previous save to complete took 48.677903 seconds. If this number is high, consider checkpointing less frequently.
I0421 07:22:45.957600 135117366056768 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 15
I0421 07:22:45.960109 135117366056768 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/15.
I0421 07:22:46.269265 135117366056768 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 07:22:46.269362 135117366056768 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
I0421 07:22:46.293865 135117366056768 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.025987s
I0421 07:22:46.293994 135117366056768 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 52.471 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.029398679733276367 s) (per-host)
I0421 07:22:46.294048 135117366056768 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.029455s (batch_requests_ready=0.001764s, total_serialization_initiated=0.027630s, others=0.000061s)
I0421 07:22:46.294128 135117366056768 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.034178s (all_items=0.000012s, per_item={'items': '0.00001168'}, temp_paths=0.034166)
I0421 07:22:46.294770 135117366056768 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.34 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/15.
I0421 07:22:46.295124 134993028310784 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 07:42:46.295085
I0421 07:22:46.306916 135117366056768 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 07:22:46.307210 134988716558080 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save.
I0421 07:22:46.307370 135117366056768 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776756117.2776597, 'wait_for_prev_duration_secs': 48.677903175354004, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776756165.9576404, 'checkpointer_blocking_duration_secs': 0.33763551712036133, 'get_old_steps_start_time': 1776756166.2952998, 'get_old_steps_duration_secs': 2.9802322387695312e-05, 'checkpoint_manager_blocking_start_time': 1776756117.275173, 'checkpoint_manager_blocking_duration_secs': 49.032163858413696}
I0421 07:22:46.307479 135117366056768 checkpointing.py:408] Started an asynchronous checkpoint save for step 15
I0421 07:22:46.308264 135117366056768 metric_logger.py:196] completed step: 14, seconds: 0.147, TFLOP/s/device: 92.189, Tokens/s/device: 13895.768, total_weights: 65536, loss: 5.864, lm_loss: 5.864, perplexity: 352.193
I0421 07:22:46.314266 135117366056768 metric_logger.py:196] completed step: 15, seconds: 0.147, TFLOP/s/device: 92.164, Tokens/s/device: 13891.904, total_weights: 65536, loss: 5.719, lm_loss: 5.719, perplexity: 304.661
I0421 07:22:51.823706 134993036703488 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/15/items/array_metadatas/process_7
I0421 07:23:07.303431 135117366056768 metric_logger.py:196] completed step: 16, seconds: 49.328, TFLOP/s/device: 0.275, Tokens/s/device: 41.518, total_weights: 65536, loss: 5.605, lm_loss: 5.605, perplexity: 271.872
I0421 07:23:07.450787 135117366056768 metric_logger.py:196] completed step: 17, seconds: 20.843, TFLOP/s/device: 0.652, Tokens/s/device: 98.259, total_weights: 65536, loss: 5.517, lm_loss: 5.517, perplexity: 248.975
I0421 07:23:07.598040 135117366056768 metric_logger.py:196] completed step: 18, seconds: 0.152, TFLOP/s/device: 89.143, Tokens/s/device: 13436.557, total_weights: 65536, loss: 5.448, lm_loss: 5.448, perplexity: 232.333
I0421 07:23:07.744681 135117366056768 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0421 07:23:07.745430 135117366056768 checkpointing.py:776] Waited 0.00077056884765625 seconds for step 19 to finish before starting checkpointing.
I0421 07:23:07.747457 135117366056768 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 07:23:28.154448 134993028310784 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 37.708 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.88983178138733 s) (per-host)
I0421 07:23:28.154554 134993028310784 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.859314s.
I0421 07:23:36.165151 134993028310784 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 49.869895s.
I0421 07:23:36.165484 134988716558080 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 07:23:36.165604 134988716558080 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save.
I0421 07:23:36.165652 134988716558080 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts...
I0421 07:23:36.167055 134988716558080 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts.
I0421 07:23:36.167281 135117366056768 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=15][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=15.
W0421 07:23:36.167411 135117366056768 checkpoint_manager.py:1452] Waiting for previous save to complete took 28.419960 seconds. If this number is high, consider checkpointing less frequently.
I0421 07:23:36.168901 135117366056768 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 19
I0421 07:23:36.171070 135117366056768 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/19.
I0421 07:23:36.467515 135117366056768 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 07:23:36.467609 135117366056768 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
I0421 07:23:36.498034 135117366056768 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.031463s
I0421 07:23:36.498169 135117366056768 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 44.186 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.034911155700683594 s) (per-host)
I0421 07:23:36.498218 135117366056768 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.034968s (batch_requests_ready=0.001665s, total_serialization_initiated=0.033241s, others=0.000061s)
I0421 07:23:36.498290 135117366056768 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.039177s (all_items=0.000011s, per_item={'items': '0.00001097'}, temp_paths=0.039166)
I0421 07:23:36.498999 135117366056768 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.33 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/19.
I0421 07:23:36.499333 134988716558080 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 07:43:36.499293
I0421 07:23:36.501301 135117366056768 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 07:23:36.501584 134989278607104 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save.
I0421 07:23:36.501759 135117366056768 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776756187.7474282, 'wait_for_prev_duration_secs': 28.419960260391235, 'time_between_consecutive_saves_sec': 21.79209280014038, 'checkpointer_blocking_start_time': 1776756216.1689415, 'checkpointer_blocking_duration_secs': 0.33054161071777344, 'get_old_steps_start_time': 1776756216.4995015, 'get_old_steps_duration_secs': 2.5987625122070312e-05, 'checkpoint_manager_blocking_start_time': 1776756187.7456615, 'checkpoint_manager_blocking_duration_secs': 28.756052255630493}
I0421 07:23:36.501867 135117366056768 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0421 07:23:36.501909 135117366056768 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 07:23:43.342241 134993036703488 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/19/items/array_metadatas/process_7
I0421 07:24:19.436881 134988716558080 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 36.757 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.97358536720276 s) (per-host)
I0421 07:24:19.437007 134988716558080 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.937555s.
I0421 07:24:26.662820 134988716558080 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 50.163353s.
I0421 07:24:26.663131 134989278607104 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 07:24:26.663260 134989278607104 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save.
I0421 07:24:26.663311 134989278607104 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0421 07:24:26.664764 134989278607104 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0421 07:24:26.664946 135117366056768 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19.
I0421 07:24:26.665100 135117366056768 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 07:24:26.666086 135117366056768 metric_logger.py:196] completed step: 19, seconds: 0.147, TFLOP/s/device: 92.300, Tokens/s/device: 13912.476, total_weights: 65536, loss: 5.391, lm_loss: 5.391, perplexity: 219.349
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Tue Apr 21 07:24:38 UTC 2026
EXIT_CODE=0