XPK Start: Thu Apr 23 08:19:51 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.
2026-04-23 08:20:16.220850: 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)
I0423 08:20:16.435516 136781777966912 max_utils.py:273] Attempting to initialize the jax distributed system...
I0423 08:20:25.476342 136781777966912 distributed.py:149] Starting JAX distributed service on [::]:8482
I0423 08:20:25.478803 136781777966912 distributed.py:172] Connecting to JAX distributed service on mt-15-ocdbt-true-sv-owvtx-slice-job-0-0.mt-15-ocdbt-true-sv-owvtx:8482
I0423 08:20:26.625174 136781777966912 max_utils.py:284] Jax distributed system initialized!
I0423 08:20:32.912685 136781777966912 max_utils.py:800] System Information: Jax Version: 0.9.2
I0423 08:20:32.912790 136781777966912 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0423 08:20:32.912830 136781777966912 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
I0423 08:20:32.912863 136781777966912 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0423 08:20:33.607062 136781777966912 maxtext_utils.py:1604] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0423 08:20:33.607342 136781777966912 checkpointing.py:677] Setting up checkpoint logger...
I0423 08:20:33.607397 136781777966912 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0423 08:20:33.607442 136781777966912 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0423 08:20:33.607782 136781777966912 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 0x7c6645b3cd70>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0423 08:20:36.955439 136781777966912 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0423 08:20:36.955676 136781777966912 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 0x7c526c48ec60>}, handler_registry=None
I0423 08:20:36.955920 136781777966912 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c526c48ec60>` 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`.
I0423 08:20:36.955984 136781777966912 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c526c492630>` 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`.
I0423 08:20:36.956021 136781777966912 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 0x7c526c48ec60>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c526c48ec60>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c526c492630>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c526c492630>}).
I0423 08:20:36.956340 136781777966912 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0423 08:20:36.956411 136781777966912 async_checkpointer.py:192] [process=3][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7c526c2c1a80> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0423 08:20:38.119478 136781777966912 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints
I0423 08:20:38.122375 136781777966912 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_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7c526c48d190>
I0423 08:20:38.122575 136781777966912 checkpointing.py:301] Checkpoint manager created!
I0423 08:20:39.046202 136781777966912 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0423 08:20:39.046319 136781777966912 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 08:20:39.430169 136781777966912 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch_attn', 'activation_length_attn', 'activation_embed_attn').
I0423 08:20:39.430261 136781777966912 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 08:20:39.447231 136781777966912 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch_attn', 'activation_length_attn', 'activation_embed_attn').
I0423 08:20:39.447290 136781777966912 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 08:20:39.471311 136781777966912 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_length_attn', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 08:20:39.471382 136781777966912 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 08:20:39.492331 136781777966912 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_length_attn', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 08:20:39.492421 136781777966912 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 08:20:39.509558 136781777966912 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_length_attn', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 08:20:39.509623 136781777966912 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 08:20:39.534718 136781777966912 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch_attn', 'activation_length_attn', 'activation_heads', 'activation_kv').
I0423 08:20:39.534788 136781777966912 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 08:20:39.555945 136781777966912 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0423 08:20:39.556011 136781777966912 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0423 08:20:39.765699 136781777966912 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0423 08:20:39.765806 136781777966912 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0423 08:20:41.190520 136781777966912 maxtext_utils.py:1707] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0423 08:20:41.190651 136781777966912 maxtext_utils.py:1707] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0423 08:20:41.190706 136781777966912 maxtext_utils.py:1707] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0423 08:20:41.190764 136781777966912 maxtext_utils.py:1707] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0423 08:20:41.190817 136781777966912 maxtext_utils.py:1707] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0423 08:20:41.190856 136781777966912 maxtext_utils.py:1707] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0423 08:20:41.190907 136781777966912 maxtext_utils.py:1707] 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)
I0423 08:20:41.190973 136781777966912 maxtext_utils.py:1707] 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')
I0423 08:20:41.191017 136781777966912 maxtext_utils.py:1707] 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)
I0423 08:20:41.191055 136781777966912 maxtext_utils.py:1707] 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)
I0423 08:20:41.191104 136781777966912 maxtext_utils.py:1707] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0423 08:20:41.191154 136781777966912 maxtext_utils.py:1707] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0423 08:20:41.685897 136781777966912 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0423 08:20:41.686000 136781777966912 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0423 08:20:41.701715 136781777966912 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0423 08:20:41.701778 136781777966912 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0423 08:20:52.581555 136781777966912 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.
I0423 08:20:52.582370 136781777966912 metric_logger.py:301] number parameters: 1.104 billion
I0423 08:21:04.249623 136781777966912 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0423 08:21:04.402140 136781777966912 checkpointing.py:776] Waited 0.152496337890625 seconds for step 0 to finish before starting checkpointing.
I0423 08:21:04.404555 136781777966912 checkpoint_manager.py:2009] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 08:21:04.406333 136781777966912 checkpoint_manager.py:1512] [process=3] Saving checkpoint at step 0
I0423 08:21:04.407789 136781777966912 event_tracking.py:70] [process=3] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints/0.
I0423 08:21:05.174885 136781777966912 signaling_client.py:364] Using JaxDistributedSignalingClient
I0423 08:21:05.175855 136781777966912 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0423 08:21:05.175912 136781777966912 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
I0423 08:21:05.454153 136781777966912 base_pytree_checkpoint_handler.py:154] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.279339s
I0423 08:21:05.454326 136781777966912 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/blocking_gbytes_per_sec: 5.415 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2848954200744629 s) (per-host)
I0423 08:21:05.454379 136781777966912 base_pytree_checkpoint_handler.py:768] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.284959s (batch_requests_ready=0.002246s, total_serialization_initiated=0.282642s, others=0.000071s)
I0423 08:21:05.454479 136781777966912 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.289818s (all_items=0.000017s, per_item={'items': '0.00001740'}, temp_paths=0.289801)
I0423 08:21:05.455270 136781777966912 event_tracking.py:125] [process=3] [async] Finished blocking save in 1.05 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints/0.
I0423 08:21:05.455588 136653746190080 async_checkpointer.py:76] [process=3][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-23 08:41:05.455550
I0423 08:21:05.457610 136781777966912 checkpoint_manager.py:1560] [process=3][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 08:21:05.457894 136650520004352 async_checkpointer.py:280] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0423 08:21:05.458064 136781777966912 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776932464.4045377, 'wait_for_prev_duration_secs': 6.031990051269531e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776932464.4063725, 'checkpointer_blocking_duration_secs': 1.0493688583374023, 'get_old_steps_start_time': 1776932465.455769, 'get_old_steps_duration_secs': 3.075599670410156e-05, 'checkpoint_manager_blocking_start_time': 1776932464.4026632, 'checkpoint_manager_blocking_duration_secs': 1.0553624629974365}
I0423 08:21:05.458167 136781777966912 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0423 08:21:05.458217 136781777966912 max_utils.py:750]
Memstats: After params initialized:
I0423 08:21:05.458264 136781777966912 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_10(process=3,(2,2,0,0))
I0423 08:21:05.458297 136781777966912 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_11(process=3,(3,2,0,0))
I0423 08:21:05.458324 136781777966912 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_14(process=3,(2,3,0,0))
I0423 08:21:05.458352 136781777966912 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_15(process=3,(3,3,0,0))
I0423 08:21:05.771405 136781777966912 metric_logger.py:196] completed step: 0, seconds: 11.667, TFLOP/s/device: 1.165, Tokens/s/device: 175.536, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0423 08:21:05.949818 136781777966912 metric_logger.py:196] completed step: 1, seconds: 1.520, TFLOP/s/device: 8.937, Tokens/s/device: 1347.146, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0423 08:21:06.375367 136781777966912 metric_logger.py:196] completed step: 2, seconds: 0.031, TFLOP/s/device: 438.323, Tokens/s/device: 66068.779, total_weights: 65536, loss: 10.560, lm_loss: 10.560, perplexity: 38547.719
I0423 08:21:06.522801 136781777966912 metric_logger.py:196] completed step: 3, seconds: 0.426, TFLOP/s/device: 31.883, Tokens/s/device: 4805.797, total_weights: 65536, loss: 9.980, lm_loss: 9.980, perplexity: 21590.506
I0423 08:21:06.527358 136781777966912 checkpointing.py:772] Waiting for step 5 to finish before checkpoint...
I0423 08:21:06.816829 136781777966912 checkpointing.py:776] Waited 0.28943634033203125 seconds for step 5 to finish before starting checkpointing.
I0423 08:21:06.820001 136781777966912 checkpoint_manager.py:2020] [process=3][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 08:21:09.202199 2798 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0423 08:21:11.579179 136651066042112 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints/0/items/array_metadatas/process_3
I0423 08:21:41.912887 136653746190080 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/gbytes_per_sec: 42.990 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 36.74340510368347 s) (per-host)
I0423 08:21:41.913029 136653746190080 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.457330s.
I0423 08:21:52.886887 136653746190080 async_checkpointer.py:160] [process=3][thread=async_save] Background save thread done. Time taken: 47.431170s.
I0423 08:21:52.887224 136650520004352 async_checkpointer.py:288] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 08:21:52.887350 136650520004352 async_checkpointer.py:298] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0423 08:21:52.887402 136650520004352 checkpoint_manager.py:2137] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0423 08:21:52.889963 136650520004352 checkpoint_manager.py:2146] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0423 08:21:52.890158 136781777966912 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.
W0423 08:21:52.890311 136781777966912 checkpoint_manager.py:1452] Waiting for previous save to complete took 46.070307 seconds. If this number is high, consider checkpointing less frequently.
I0423 08:21:52.892121 136781777966912 checkpoint_manager.py:1512] [process=3] Saving checkpoint at step 5
I0423 08:21:52.894563 136781777966912 event_tracking.py:70] [process=3] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints/5.
I0423 08:21:53.636278 136781777966912 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0423 08:21:53.636371 136781777966912 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
I0423 08:21:53.676580 136781777966912 base_pytree_checkpoint_handler.py:154] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.041367s
I0423 08:21:53.676739 136781777966912 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/blocking_gbytes_per_sec: 34.395 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04484915733337402 s) (per-host)
I0423 08:21:53.676790 136781777966912 base_pytree_checkpoint_handler.py:768] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.044910s (batch_requests_ready=0.001841s, total_serialization_initiated=0.043000s, others=0.000069s)
I0423 08:21:53.676880 136781777966912 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.049911s (all_items=0.000015s, per_item={'items': '0.00001502'}, temp_paths=0.049896)
I0423 08:21:53.677525 136781777966912 event_tracking.py:125] [process=3] [async] Finished blocking save in 0.79 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints/5.
I0423 08:21:53.677867 136650520004352 async_checkpointer.py:76] [process=3][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-23 08:41:53.677828
I0423 08:21:53.694430 136781777966912 checkpoint_manager.py:1560] [process=3][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 08:21:53.694678 136653242889984 async_checkpointer.py:280] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0423 08:21:53.694804 136781777966912 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776932466.8199718, 'wait_for_prev_duration_secs': 46.07030701637268, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776932512.892158, 'checkpointer_blocking_duration_secs': 0.7858741283416748, 'get_old_steps_start_time': 1776932513.6780539, 'get_old_steps_duration_secs': 3.123283386230469e-05, 'checkpoint_manager_blocking_start_time': 1776932466.817512, 'checkpoint_manager_blocking_duration_secs': 46.877257347106934}
I0423 08:21:53.694921 136781777966912 checkpointing.py:408] Started an asynchronous checkpoint save for step 5
I0423 08:21:53.695725 136781777966912 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.865, Tokens/s/device: 13394.726, total_weights: 65536, loss: 9.460, lm_loss: 9.460, perplexity: 12840.355
I0423 08:21:53.702056 136781777966912 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.222, Tokens/s/device: 13900.673, total_weights: 65536, loss: 8.959, lm_loss: 8.959, perplexity: 7776.467
I0423 08:21:59.689967 136651066042112 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints/5/items/array_metadatas/process_3
I0423 08:22:16.623214 136781777966912 metric_logger.py:196] completed step: 6, seconds: 47.173, TFLOP/s/device: 0.288, Tokens/s/device: 43.415, total_weights: 65536, loss: 8.469, lm_loss: 8.469, perplexity: 4762.740
I0423 08:22:16.770509 136781777966912 metric_logger.py:196] completed step: 7, seconds: 22.775, TFLOP/s/device: 0.597, Tokens/s/device: 89.923, total_weights: 65536, loss: 8.003, lm_loss: 8.003, perplexity: 2989.151
I0423 08:22:16.917818 136781777966912 metric_logger.py:196] completed step: 8, seconds: 0.153, TFLOP/s/device: 88.905, Tokens/s/device: 13400.773, total_weights: 65536, loss: 7.572, lm_loss: 7.572, perplexity: 1943.098
I0423 08:22:17.064416 136781777966912 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0423 08:22:17.065111 136781777966912 checkpointing.py:776] Waited 0.0007154941558837891 seconds for step 9 to finish before starting checkpointing.
I0423 08:22:17.067213 136781777966912 checkpoint_manager.py:2020] [process=3][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 08:22:35.415174 136650520004352 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/gbytes_per_sec: 37.805 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.783249378204346 s) (per-host)
I0423 08:22:35.415293 136650520004352 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.737289s.
I0423 08:22:42.858949 136650520004352 async_checkpointer.py:160] [process=3][thread=async_save] Background save thread done. Time taken: 49.180917s.
I0423 08:22:42.859181 136653242889984 async_checkpointer.py:288] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 08:22:42.859239 136653242889984 async_checkpointer.py:298] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0423 08:22:42.859288 136653242889984 checkpoint_manager.py:2137] [process=3][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts...
I0423 08:22:42.861026 136653242889984 checkpoint_manager.py:2146] [process=3][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts.
I0423 08:22:42.861201 136781777966912 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.
W0423 08:22:42.861328 136781777966912 checkpoint_manager.py:1452] Waiting for previous save to complete took 25.794119 seconds. If this number is high, consider checkpointing less frequently.
I0423 08:22:42.862831 136781777966912 checkpoint_manager.py:1512] [process=3] Saving checkpoint at step 9
I0423 08:22:42.864955 136781777966912 event_tracking.py:70] [process=3] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints/9.
I0423 08:22:43.571614 136781777966912 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0423 08:22:43.571710 136781777966912 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
I0423 08:22:43.612025 136781777966912 base_pytree_checkpoint_handler.py:154] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.041422s
I0423 08:22:43.612189 136781777966912 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/blocking_gbytes_per_sec: 34.362 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.0448918342590332 s) (per-host)
I0423 08:22:43.612239 136781777966912 base_pytree_checkpoint_handler.py:768] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.044951s (batch_requests_ready=0.001666s, total_serialization_initiated=0.043221s, others=0.000065s)
I0423 08:22:43.612330 136781777966912 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.049141s (all_items=0.000010s, per_item={'items': '0.00001049'}, temp_paths=0.049130)
I0423 08:22:43.613072 136781777966912 event_tracking.py:125] [process=3] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints/9.
I0423 08:22:43.613414 136653242889984 async_checkpointer.py:76] [process=3][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-23 08:42:43.613376
I0423 08:22:43.615315 136781777966912 checkpoint_manager.py:1560] [process=3][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 08:22:43.615530 136652706019072 async_checkpointer.py:280] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0423 08:22:43.615683 136781777966912 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776932537.0671837, 'wait_for_prev_duration_secs': 25.794118642807007, 'time_between_consecutive_saves_sec': 24.177180290222168, 'checkpointer_blocking_start_time': 1776932562.8628702, 'checkpointer_blocking_duration_secs': 0.7506859302520752, 'get_old_steps_start_time': 1776932563.6135733, 'get_old_steps_duration_secs': 2.6464462280273438e-05, 'checkpoint_manager_blocking_start_time': 1776932537.065339, 'checkpoint_manager_blocking_duration_secs': 26.550313711166382}
I0423 08:22:43.615795 136781777966912 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0423 08:22:43.615838 136781777966912 checkpoint_manager.py:2020] [process=3][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 08:22:49.073188 136651066042112 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260423_071538/linen_xpk_main_20260423_071538_15_ocdbt_true/checkpoints/9/items/array_metadatas/process_3
I0423 08:23:25.504742 136653242889984 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/gbytes_per_sec: 37.666 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.9374041557312 s) (per-host)
I0423 08:23:25.504860 136653242889984 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.891331s.
I0423 08:23:32.813382 136653242889984 async_checkpointer.py:160] [process=3][thread=async_save] Background save thread done. Time taken: 49.199837s.
I0423 08:23:32.813672 136652706019072 async_checkpointer.py:288] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 08:23:32.813793 136652706019072 async_checkpointer.py:298] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0423 08:23:32.813832 136652706019072 checkpoint_manager.py:2137] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0423 08:23:32.815241 136652706019072 checkpoint_manager.py:2146] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0423 08:23:32.815409 136781777966912 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.
I0423 08:23:32.815557 136781777966912 checkpoint_manager.py:2009] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 08:23:32.816548 136781777966912 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.284, Tokens/s/device: 13910.114, 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: Thu Apr 23 08:23:41 UTC 2026
EXIT_CODE=0