MaxView

XPK Start: Tue Apr 21 06:42:24 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 06:42:49.114380: 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 06:42:49.327453 138731845056320 max_utils.py:273] Attempting to initialize the jax distributed system...
I0421 06:42:58.369599 138731845056320 distributed.py:149] Starting JAX distributed service on [::]:8482
I0421 06:42:58.372018 138731845056320 distributed.py:172] Connecting to JAX distributed service on mt-08-checkpoint-async--2i7ub-slice-job-0-0.mt-08-checkpoint-async--2i7ub:8482
I0421 06:42:59.554788 138731845056320 max_utils.py:284] Jax distributed system initialized!
I0421 06:43:05.974484 138731845056320 max_utils.py:800] System Information: Jax Version: 0.9.2
I0421 06:43:05.974590 138731845056320 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0421 06:43:05.974631 138731845056320 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 06:43:05.974686 138731845056320 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0421 06:43:06.665980 138731845056320 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 06:43:06.666268 138731845056320 checkpointing.py:677] Setting up checkpoint logger...
I0421 06:43:06.666328 138731845056320 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0421 06:43:06.666373 138731845056320 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0421 06:43:06.666728 138731845056320 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 0x7e2c6351c0b0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0421 06:43:07.964570 138731845056320 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0421 06:43:07.964828 138731845056320 checkpoint_manager.py:708] [process=4][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e18384219d0>}, handler_registry=None
I0421 06:43:07.965071 138731845056320 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e18384219d0>` 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 06:43:07.965120 138731845056320 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e1838425640>` 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 06:43:07.965157 138731845056320 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 0x7e18384219d0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e18384219d0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e1838425640>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e1838425640>}).
I0421 06:43:07.965496 138731845056320 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0421 06:43:07.965565 138731845056320 async_checkpointer.py:192] [process=4][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7e18181ed580> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0421 06:43:10.293549 138610967815936 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776753624032760373, 'commit_timestamp_nsecs': 1776753670884543251, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/9/_CHECKPOINT_METADATA
I0421 06:43:10.538767 138610976208640 checkpoint_manager.py:1871] Missing metrics for step 5
I0421 06:43:10.580069 138731845056320 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints
I0421 06:43:11.443575 138731845056320 checkpoint_manager.py:929] [process=4][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_08_checkpoint_async_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7e1838422270>
I0421 06:43:11.443767 138731845056320 checkpointing.py:301] Checkpoint manager created!
I0421 06:43:12.383187 138731845056320 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0421 06:43:12.383308 138731845056320 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 06:43:12.764474 138731845056320 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 06:43:12.764564 138731845056320 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 06:43:12.781079 138731845056320 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 06:43:12.781141 138731845056320 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 06:43:12.804872 138731845056320 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 06:43:12.804939 138731845056320 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 06:43:12.821434 138731845056320 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 06:43:12.821498 138731845056320 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 06:43:12.837929 138731845056320 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 06:43:12.837991 138731845056320 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 06:43:12.862771 138731845056320 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0421 06:43:12.862841 138731845056320 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 06:43:12.884856 138731845056320 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0421 06:43:12.884925 138731845056320 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0421 06:43:13.094302 138731845056320 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0421 06:43:13.094407 138731845056320 checkpointing.py:581] restoring from this run's directory step 9
I0421 06:43:13.339218 138731845056320 event_tracking.py:70] [process=4] [sync] Started load checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/9.
I0421 06:43:13.578599 138731845056320 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/9.
I0421 06:43:13.579040 138731845056320 event_tracking.py:125] [process=4] [sync] Finished blocking load in 0.24 seconds. Continuing load @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/9.
I0421 06:43:15.867144    1445 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0421 06:43:21.557423 138731845056320 jax_array_handlers.py:843] [process=4] /jax/orbax/read/worker/io/requested throughput: 277.433 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 5.696051359176636 s) (per-host)
I0421 06:43:21.562953 138731845056320 base_pytree_checkpoint_handler.py:130] [process=4] /jax/checkpoint/read/gbytes_per_sec: 241.857 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 6.543522119522095 s) (per-host)
I0421 06:43:23.267158 138731845056320 event_tracking.py:138] [process=4] [sync] Finished load in 9.93 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/9
I0421 06:43:23.267320 138731845056320 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints', 'checkpointer_start_time': 1776753793.3391204, 'checkpointer_duration_secs': 9.928168296813965, 'checkpoint_manager_start_time': 1776753793.094659, 'checkpoint_manager_duration_secs': 10.172631740570068}
fsdp: 32
I0421 06:43:23.683746 138731845056320 maxtext_utils.py:1668]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0421 06:43:23.683873 138731845056320 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 06:43:23.683928 138731845056320 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 06:43:23.683986 138731845056320 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 06:43:23.684042 138731845056320 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 06:43:23.684082 138731845056320 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 06:43:23.684133 138731845056320 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 06:43:23.684183 138731845056320 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 06:43:23.684221 138731845056320 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 06:43:23.684256 138731845056320 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 06:43:23.684301 138731845056320 maxtext_utils.py:1668]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0421 06:43:23.684346 138731845056320 maxtext_utils.py:1668]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')

I0421 06:43:24.175318 138731845056320 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 06:43:24.175412 138731845056320 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 06:43:24.191108 138731845056320 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 06:43:24.191175 138731845056320 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 06:43:35.223607 138731845056320 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 06:43:35.224447 138731845056320 metric_logger.py:301] number parameters: 1.104 billion
I0421 06:43:46.992964 138731845056320 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0421 06:43:47.149148 138731845056320 checkpointing.py:776] Waited 0.15616536140441895 seconds for step 10 to finish before starting checkpointing.
I0421 06:43:47.151486 138731845056320 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 06:43:47.153535 138731845056320 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 10
I0421 06:43:47.157074 138731845056320 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/10.
I0421 06:43:47.927513 138731845056320 signaling_client.py:364] Using JaxDistributedSignalingClient
I0421 06:43:47.928574 138731845056320 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 06:43:47.928630 138731845056320 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 06:43:48.206718 138731845056320 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.279280s
I0421 06:43:48.206887 138731845056320 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 5.446 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2832314968109131 s) (per-host)
I0421 06:43:48.206940 138731845056320 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.283296s (batch_requests_ready=0.002224s, total_serialization_initiated=0.281000s, others=0.000072s)
I0421 06:43:48.207032 138731845056320 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.287399s (all_items=0.000017s, per_item={'items': '0.00001717'}, temp_paths=0.287382)
I0421 06:43:48.207772 138731845056320 event_tracking.py:125] [process=4] [async] Finished blocking save in 1.05 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/10.
I0421 06:43:48.208106 138610389014272 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 07:03:48.208068
I0421 06:43:48.214096 138731845056320 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 06:43:48.214386 138608778401536 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 06:43:48.214545 138731845056320 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776753827.1514678, 'wait_for_prev_duration_secs': 6.699562072753906e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776753827.153621, 'checkpointer_blocking_duration_secs': 1.0546319484710693, 'get_old_steps_start_time': 1776753828.208276, 'get_old_steps_duration_secs': 3.647804260253906e-05, 'checkpoint_manager_blocking_start_time': 1776753827.149541, 'checkpoint_manager_blocking_duration_secs': 1.064967393875122}
I0421 06:43:48.214668 138731845056320 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0421 06:43:48.214722 138731845056320 max_utils.py:750] 
Memstats: After params initialized:
I0421 06:43:48.214775 138731845056320 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_16(process=4,(0,4,0,0))
I0421 06:43:48.214809 138731845056320 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_17(process=4,(1,4,0,0))
I0421 06:43:48.214838 138731845056320 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_20(process=4,(0,5,0,0))
I0421 06:43:48.214863 138731845056320 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_21(process=4,(1,5,0,0))
I0421 06:43:48.525339 138731845056320 metric_logger.py:196] completed step: 10, seconds: 11.768, TFLOP/s/device: 1.155, Tokens/s/device: 174.025, total_weights: 65536, loss: 6.833, lm_loss: 6.833, perplexity: 927.869
I0421 06:43:48.714414 138731845056320 metric_logger.py:196] completed step: 11, seconds: 1.531, TFLOP/s/device: 8.876, Tokens/s/device: 1337.835, total_weights: 65536, loss: 6.527, lm_loss: 6.527, perplexity: 683.659
I0421 06:43:48.862049 138731845056320 metric_logger.py:196] completed step: 12, seconds: 0.042, TFLOP/s/device: 324.864, Tokens/s/device: 48967.100, total_weights: 65536, loss: 6.265, lm_loss: 6.265, perplexity: 526.068
I0421 06:43:49.009440 138731845056320 metric_logger.py:196] completed step: 13, seconds: 0.153, TFLOP/s/device: 88.803, Tokens/s/device: 13385.358, total_weights: 65536, loss: 6.045, lm_loss: 6.045, perplexity: 422.001
I0421 06:43:49.013549 138731845056320 checkpointing.py:772] Waiting for step 15 to finish before checkpoint...
I0421 06:43:49.303534 138731845056320 checkpointing.py:776] Waited 0.2899491786956787 seconds for step 15 to finish before starting checkpointing.
I0421 06:43:49.306591 138731845056320 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 06:43:54.475250 138608795186944 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/10/items/array_metadatas/process_4
I0421 06:44:29.461776 138610389014272 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 38.028 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.53808355331421 s) (per-host)
I0421 06:44:29.461899 138610389014272 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.253686s.
I0421 06:44:39.093393 138610389014272 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 50.885163s.
I0421 06:44:39.093745 138608778401536 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 06:44:39.093865 138608778401536 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 06:44:39.093912 138608778401536 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0421 06:44:39.095442 138608778401536 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0421 06:44:39.095676 138731845056320 checkpoint_manager.py:2032] [process=4][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10.
W0421 06:44:39.095810 138731845056320 checkpoint_manager.py:1452] Waiting for previous save to complete took 49.789218 seconds. If this number is high, consider checkpointing less frequently.
I0421 06:44:39.097782 138731845056320 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 15
I0421 06:44:39.100354 138731845056320 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/15.
I0421 06:44:39.807687 138731845056320 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 06:44:39.807784 138731845056320 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 06:44:39.833429 138731845056320 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.027709s
I0421 06:44:39.833556 138731845056320 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 49.515 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03115367889404297 s) (per-host)
I0421 06:44:39.833603 138731845056320 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.031208s (batch_requests_ready=0.001812s, total_serialization_initiated=0.029335s, others=0.000061s)
I0421 06:44:39.833706 138731845056320 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.035982s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=0.035963)
I0421 06:44:39.834401 138731845056320 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.74 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/15.
I0421 06:44:39.834740 138608778401536 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 07:04:39.834702
I0421 06:44:39.844792 138731845056320 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 06:44:39.845057 138608795186944 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 06:44:39.845216 138731845056320 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776753829.306564, 'wait_for_prev_duration_secs': 49.78921842575073, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776753879.0978265, 'checkpointer_blocking_duration_secs': 0.737053632736206, 'get_old_steps_start_time': 1776753879.8349025, 'get_old_steps_duration_secs': 3.0517578125e-05, 'checkpoint_manager_blocking_start_time': 1776753829.3042166, 'checkpoint_manager_blocking_duration_secs': 50.54096484184265}
I0421 06:44:39.845328 138731845056320 checkpointing.py:408] Started an asynchronous checkpoint save for step 15
I0421 06:44:39.846166 138731845056320 metric_logger.py:196] completed step: 14, seconds: 0.148, TFLOP/s/device: 92.056, Tokens/s/device: 13875.715, total_weights: 65536, loss: 5.864, lm_loss: 5.864, perplexity: 352.193
I0421 06:44:39.852278 138731845056320 metric_logger.py:196] completed step: 15, seconds: 0.147, TFLOP/s/device: 92.277, Tokens/s/device: 13908.980, total_weights: 65536, loss: 5.719, lm_loss: 5.719, perplexity: 304.661
I0421 06:44:44.828882 138604995823360 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/15/items/array_metadatas/process_4
I0421 06:45:01.297561 138731845056320 metric_logger.py:196] completed step: 16, seconds: 50.837, TFLOP/s/device: 0.267, Tokens/s/device: 40.286, total_weights: 65536, loss: 5.605, lm_loss: 5.605, perplexity: 271.872
I0421 06:45:01.444958 138731845056320 metric_logger.py:196] completed step: 17, seconds: 21.299, TFLOP/s/device: 0.638, Tokens/s/device: 96.153, total_weights: 65536, loss: 5.517, lm_loss: 5.517, perplexity: 248.975
I0421 06:45:01.592278 138731845056320 metric_logger.py:196] completed step: 18, seconds: 0.152, TFLOP/s/device: 89.370, Tokens/s/device: 13470.760, total_weights: 65536, loss: 5.448, lm_loss: 5.448, perplexity: 232.333
I0421 06:45:01.739034 138731845056320 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0421 06:45:01.739729 138731845056320 checkpointing.py:776] Waited 0.0007169246673583984 seconds for step 19 to finish before starting checkpointing.
I0421 06:45:01.741798 138731845056320 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 06:45:21.668792 138608778401536 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 37.730 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.86634588241577 s) (per-host)
I0421 06:45:21.668922 138608778401536 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.834072s.
I0421 06:45:27.585738 138608778401536 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 47.750873s.
I0421 06:45:27.586045 138608795186944 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 06:45:27.586167 138608795186944 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 06:45:27.586232 138608795186944 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts...
I0421 06:45:27.587581 138608795186944 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts.
I0421 06:45:27.587780 138731845056320 checkpoint_manager.py:2032] [process=4][thread=MainThread][step=15][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=15.
W0421 06:45:27.587908 138731845056320 checkpoint_manager.py:1452] Waiting for previous save to complete took 25.846114 seconds. If this number is high, consider checkpointing less frequently.
I0421 06:45:27.589332 138731845056320 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 19
I0421 06:45:27.591752 138731845056320 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/19.
I0421 06:45:28.312376 138731845056320 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 06:45:28.312469 138731845056320 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 06:45:28.338062 138731845056320 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.026490s
I0421 06:45:28.338226 138731845056320 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 51.365 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.030031919479370117 s) (per-host)
I0421 06:45:28.338281 138731845056320 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.030097s (batch_requests_ready=0.001697s, total_serialization_initiated=0.028331s, others=0.000069s)
I0421 06:45:28.338374 138731845056320 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.034090s (all_items=0.000012s, per_item={'items': '0.00001192'}, temp_paths=0.034078)
I0421 06:45:28.339133 138731845056320 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/19.
I0421 06:45:28.339417 138604995823360 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 07:05:28.339385
I0421 06:45:28.342868 138731845056320 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 06:45:28.343108 138608795186944 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 06:45:28.343234 138731845056320 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776753901.7417705, 'wait_for_prev_duration_secs': 25.846113681793213, 'time_between_consecutive_saves_sec': 22.646276712417603, 'checkpointer_blocking_start_time': 1776753927.58937, 'checkpointer_blocking_duration_secs': 0.7501888275146484, 'get_old_steps_start_time': 1776753928.3395772, 'get_old_steps_duration_secs': 2.5272369384765625e-05, 'checkpoint_manager_blocking_start_time': 1776753901.7399654, 'checkpoint_manager_blocking_duration_secs': 26.603235483169556}
I0421 06:45:28.343358 138731845056320 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0421 06:45:28.343398 138731845056320 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 06:45:33.501526 138604987430656 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_08_checkpoint_async_true/checkpoints/19/items/array_metadatas/process_4
I0421 06:46:09.565254 138604995823360 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 38.287 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.25702977180481 s) (per-host)
I0421 06:46:09.565354 138604995823360 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.225826s.
I0421 06:46:16.385270 138604995823360 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 48.045723s.
I0421 06:46:16.385576 138608795186944 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 06:46:16.385711 138608795186944 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 06:46:16.385754 138608795186944 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0421 06:46:16.388252 138608795186944 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0421 06:46:16.388448 138731845056320 checkpoint_manager.py:2032] [process=4][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19.
I0421 06:46:16.388599 138731845056320 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 06:46:16.389681 138731845056320 metric_logger.py:196] completed step: 19, seconds: 0.147, TFLOP/s/device: 92.166, Tokens/s/device: 13892.281, 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 06:46:29 UTC 2026
EXIT_CODE=0