MaxView

XPK Start: Mon Apr 20 06:36:28 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-20 06:36:52.817457: 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)
I0420 06:36:52.996236 140303520335680 max_utils.py:273] Attempting to initialize the jax distributed system...
I0420 06:37:02.036441 140303520335680 distributed.py:149] Starting JAX distributed service on [::]:8482
I0420 06:37:02.038743 140303520335680 distributed.py:172] Connecting to JAX distributed service on mt-10-shardy-false-a4q8f-slice-job-0-0.mt-10-shardy-false-a4q8f:8482
I0420 06:37:03.392441 140303520335680 max_utils.py:284] Jax distributed system initialized!
I0420 06:37:09.457650 140303520335680 max_utils.py:800] System Information: Jax Version: 0.9.2
I0420 06:37:09.457749 140303520335680 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0420 06:37:09.457787 140303520335680 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
I0420 06:37:09.457822 140303520335680 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0420 06:37:10.146970 140303520335680 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0420 06:37:10.147238 140303520335680 checkpointing.py:677] Setting up checkpoint logger...
I0420 06:37:10.147292 140303520335680 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0420 06:37:10.147351 140303520335680 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0420 06:37:10.147685 140303520335680 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 0x7f9a3de383e0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0420 06:37:13.170461 140303520335680 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0420 06:37:13.170694 140303520335680 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 0x7f862817e0c0>}, handler_registry=None
I0420 06:37:13.170926 140303520335680 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f862817e0c0>` 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`.
I0420 06:37:13.170974 140303520335680 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f8608775fd0>` 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`.
I0420 06:37:13.171009 140303520335680 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 0x7f862817e0c0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f862817e0c0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f8608775fd0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f8608775fd0>}).
I0420 06:37:13.171317 140303520335680 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0420 06:37:13.171399 140303520335680 async_checkpointer.py:192] [process=4][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7f852030fd80> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0420 06:37:14.209804 140303520335680 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_10_shardy_false/checkpoints
I0420 06:37:14.235276 140303520335680 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=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_10_shardy_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7f86087724e0>
I0420 06:37:14.235427 140303520335680 checkpointing.py:301] Checkpoint manager created!
I0420 06:37:15.166032 140303520335680 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 06:37:15.166145 140303520335680 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 06:37:15.541881 140303520335680 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 06:37:15.541969 140303520335680 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 06:37:15.558207 140303520335680 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 06:37:15.558268 140303520335680 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 06:37:15.582289 140303520335680 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 06:37:15.582365 140303520335680 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 06:37:15.598770 140303520335680 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 06:37:15.598829 140303520335680 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 06:37:15.615001 140303520335680 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 06:37:15.615056 140303520335680 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 06:37:15.639323 140303520335680 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0420 06:37:15.639401 140303520335680 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 06:37:15.660988 140303520335680 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0420 06:37:15.661051 140303520335680 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0420 06:37:15.874155 140303520335680 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0420 06:37:15.874258 140303520335680 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0420 06:37:17.260586 140303520335680 maxtext_utils.py:1654]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0420 06:37:17.260709 140303520335680 maxtext_utils.py:1654]  params/params/decoder/layers/mlp/wi_0/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0420 06:37:17.260760 140303520335680 maxtext_utils.py:1654]  params/params/decoder/layers/mlp/wi_1/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0420 06:37:17.260817 140303520335680 maxtext_utils.py:1654]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   P('mlp', 'layers', 'embed')
    Physical:  (None, None, 'fsdp')
I0420 06:37:17.260870 140303520335680 maxtext_utils.py:1654]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0420 06:37:17.260908 140303520335680 maxtext_utils.py:1654]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0420 06:37:17.260960 140303520335680 maxtext_utils.py:1654]  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)
I0420 06:37:17.261013 140303520335680 maxtext_utils.py:1654]  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')
I0420 06:37:17.261053 140303520335680 maxtext_utils.py:1654]  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)
I0420 06:37:17.261090 140303520335680 maxtext_utils.py:1654]  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)
I0420 06:37:17.261137 140303520335680 maxtext_utils.py:1654]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)

I0420 06:37:17.261186 140303520335680 maxtext_utils.py:1654]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')
I0420 06:37:17.752678 140303520335680 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0420 06:37:17.752771 140303520335680 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0420 06:37:17.768085 140303520335680 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0420 06:37:17.768143 140303520335680 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0420 06:37:28.637199 140303520335680 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.
I0420 06:37:28.637992 140303520335680 metric_logger.py:301] number parameters: 1.104 billion
I0420 06:37:39.739447 140303520335680 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0420 06:37:39.894698 140303520335680 checkpointing.py:776] Waited 0.15523552894592285 seconds for step 0 to finish before starting checkpointing.
I0420 06:37:39.897193 140303520335680 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 06:37:39.899005 140303520335680 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 0
I0420 06:37:39.900465 140303520335680 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_10_shardy_false/checkpoints/0.
I0420 06:37:40.197223 140303520335680 signaling_client.py:364] Using JaxDistributedSignalingClient
I0420 06:37:40.198130 140303520335680 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0420 06:37:40.198186 140303520335680 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
I0420 06:37:40.468876 140303520335680 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.271724s
I0420 06:37:40.469051 140303520335680 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 5.555 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.27767014503479004 s) (per-host)
I0420 06:37:40.469105 140303520335680 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.277735s (batch_requests_ready=0.002749s, total_serialization_initiated=0.274914s, others=0.000072s)
I0420 06:37:40.469196 140303520335680 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.281818s (all_items=0.000018s, per_item={'items': '0.00001836'}, temp_paths=0.281800)
I0420 06:37:40.470049 140303520335680 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.57 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_10_shardy_false/checkpoints/0.
I0420 06:37:40.470386 140173947721472 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 06:57:40.470349
I0420 06:37:40.489117 140303520335680 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 06:37:40.489449 140169680254720 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0420 06:37:40.489610 140303520335680 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776667059.8971744, 'wait_for_prev_duration_secs': 6.365776062011719e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776667059.8990428, 'checkpointer_blocking_duration_secs': 0.5714950561523438, 'get_old_steps_start_time': 1776667060.470562, 'get_old_steps_duration_secs': 3.0994415283203125e-05, 'checkpoint_manager_blocking_start_time': 1776667059.8950903, 'checkpoint_manager_blocking_duration_secs': 0.5944805145263672}
I0420 06:37:40.489720 140303520335680 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0420 06:37:40.489769 140303520335680 max_utils.py:750] 
Memstats: After params initialized:
I0420 06:37:40.489825 140303520335680 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_16(process=4,(0,4,0,0))
I0420 06:37:40.489857 140303520335680 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_17(process=4,(1,4,0,0))
I0420 06:37:40.489884 140303520335680 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_20(process=4,(0,5,0,0))
I0420 06:37:40.489908 140303520335680 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_21(process=4,(1,5,0,0))
I0420 06:37:40.797905 140303520335680 metric_logger.py:196] completed step: 0, seconds: 11.101, TFLOP/s/device: 1.224, Tokens/s/device: 184.482, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0420 06:37:40.988236 140303520335680 metric_logger.py:196] completed step: 1, seconds: 1.057, TFLOP/s/device: 12.855, Tokens/s/device: 1937.609, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0420 06:37:41.404200 140303520335680 metric_logger.py:196] completed step: 2, seconds: 0.043, TFLOP/s/device: 316.141, Tokens/s/device: 47652.287, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129
I0420 06:37:41.551576 140303520335680 metric_logger.py:196] completed step: 3, seconds: 0.416, TFLOP/s/device: 32.661, Tokens/s/device: 4922.994, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203
I0420 06:37:41.847010 140303520335680 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.629, Tokens/s/device: 13359.077, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614
I0420 06:37:41.852910 140303520335680 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.178, Tokens/s/device: 13894.071, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717
I0420 06:37:43.627639    2829 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0420 06:37:45.783555 140171261503232 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_10_shardy_false/checkpoints/0/items/array_metadatas/process_4
I0420 06:38:03.682711 140303520335680 metric_logger.py:196] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.934, Tokens/s/device: 6923.714, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759
I0420 06:38:03.830227 140303520335680 metric_logger.py:196] completed step: 7, seconds: 21.683, TFLOP/s/device: 0.627, Tokens/s/device: 94.452, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905
I0420 06:38:03.977715 140303520335680 metric_logger.py:196] completed step: 8, seconds: 0.153, TFLOP/s/device: 89.074, Tokens/s/device: 13426.163, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406
I0420 06:38:04.124445 140303520335680 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0420 06:38:04.125126 140303520335680 checkpointing.py:776] Waited 0.0006983280181884766 seconds for step 9 to finish before starting checkpointing.
I0420 06:38:04.127187 140303520335680 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 06:38:17.693174 140173947721472 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 42.121 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.50175499916077 s) (per-host)
I0420 06:38:17.693296 140173947721472 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.222796s.
I0420 06:38:26.885671 140173947721472 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 46.415154s.
I0420 06:38:26.885957 140169680254720 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 06:38:26.886072 140169680254720 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0420 06:38:26.886123 140169680254720 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0420 06:38:26.889994 140169680254720 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0420 06:38:26.890183 140303520335680 checkpoint_manager.py:2032] [process=4][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0420 06:38:26.890310 140303520335680 checkpoint_manager.py:1452] Waiting for previous save to complete took 22.763124 seconds. If this number is high, consider checkpointing less frequently.
I0420 06:38:26.892257 140303520335680 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 9
I0420 06:38:26.894467 140303520335680 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_10_shardy_false/checkpoints/9.
I0420 06:38:27.181206 140303520335680 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0420 06:38:27.181298 140303520335680 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
I0420 06:38:27.215689 140303520335680 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.035551s
I0420 06:38:27.215855 140303520335680 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 39.448 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03910374641418457 s) (per-host)
I0420 06:38:27.215904 140303520335680 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.039163s (batch_requests_ready=0.001900s, total_serialization_initiated=0.037196s, others=0.000067s)
I0420 06:38:27.215992 140303520335680 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.043118s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.043103)
I0420 06:38:27.216650 140303520335680 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.32 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_10_shardy_false/checkpoints/9.
I0420 06:38:27.216997 140152471021312 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 06:58:27.216959
I0420 06:38:27.219008 140303520335680 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 06:38:27.219241 140169680254720 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0420 06:38:27.219374 140303520335680 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776667084.1271555, 'wait_for_prev_duration_secs': 22.763124227523804, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776667106.8922973, 'checkpointer_blocking_duration_secs': 0.32485365867614746, 'get_old_steps_start_time': 1776667107.217172, 'get_old_steps_duration_secs': 3.0517578125e-05, 'checkpoint_manager_blocking_start_time': 1776667084.1254513, 'checkpoint_manager_blocking_duration_secs': 23.093878030776978}
I0420 06:38:27.219477 140303520335680 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0420 06:38:27.219521 140303520335680 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 06:38:32.797828 140171261503232 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_10_shardy_false/checkpoints/9/items/array_metadatas/process_4
I0420 06:39:09.518104 140152471021312 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 37.306 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.34131145477295 s) (per-host)
I0420 06:39:09.518229 140152471021312 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.301107s.
I0420 06:39:17.727494 140152471021312 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 50.510357s.
I0420 06:39:17.727697 140169680254720 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 06:39:17.727765 140169680254720 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0420 06:39:17.727810 140169680254720 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0420 06:39:17.729147 140169680254720 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0420 06:39:17.729308 140303520335680 checkpoint_manager.py:2032] [process=4][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0420 06:39:17.729467 140303520335680 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 06:39:17.730411 140303520335680 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.175, Tokens/s/device: 13893.600, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.635
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Mon Apr 20 06:39:29 UTC 2026
EXIT_CODE=0