MaxView

XPK Start: Tue Apr 21 09:10:04 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-21 09:10:28.661256: 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 09:10:29.200793 134292041258816 max_utils.py:273] Attempting to initialize the jax distributed system...
I0421 09:10:38.243062 134292041258816 distributed.py:149] Starting JAX distributed service on [::]:8482
I0421 09:10:38.245286 134292041258816 distributed.py:172] Connecting to JAX distributed service on mt-10-shardy-false-qlt2c-slice-job-0-0.mt-10-shardy-false-qlt2c:8482
I0421 09:10:39.628569 134292041258816 max_utils.py:284] Jax distributed system initialized!
I0421 09:10:45.545821 134292041258816 max_utils.py:800] System Information: Jax Version: 0.9.2
I0421 09:10:45.545922 134292041258816 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0421 09:10:45.545963 134292041258816 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 09:10:45.545997 134292041258816 train_utils.py:378] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0421 09:10:46.234372 134292041258816 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 09:10:46.234972 134292041258816 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 09:10:46.235149 134292041258816 checkpointing.py:688] Setting up checkpoint logger...
I0421 09:10:46.235204 134292041258816 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0421 09:10:46.235248 134292041258816 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0421 09:10:46.235563 134292041258816 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 0x7a22b1718770>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0421 09:10:49.014318 134292041258816 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0421 09:10:49.014595 134292041258816 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 0x7a0e2814fb90>}, handler_registry=None
I0421 09:10:49.014846 134292041258816 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7a0e2814fb90>` 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 09:10:49.014895 134292041258816 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a0e2814e450>` 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 09:10:49.014930 134292041258816 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 0x7a0e2814fb90>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7a0e2814fb90>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a0e2814e450>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a0e2814e450>}).
I0421 09:10:49.015230 134292041258816 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0421 09:10:49.015299 134292041258816 async_checkpointer.py:192] [process=4][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7a0df84cfec0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0421 09:10:50.138187 134292041258816 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_10_shardy_false/checkpoints
I0421 09:10:50.597477 134292041258816 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_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_10_shardy_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7a0e282d5640>
I0421 09:10:50.597647 134292041258816 checkpointing.py:302] Checkpoint manager created!
I0421 09:10:51.543457 134292041258816 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0421 09:10:51.543559 134292041258816 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 09:10:51.923092 134292041258816 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 09:10:51.923185 134292041258816 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 09:10:51.939478 134292041258816 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 09:10:51.939535 134292041258816 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 09:10:51.963173 134292041258816 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 09:10:51.963245 134292041258816 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 09:10:51.979524 134292041258816 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 09:10:51.979588 134292041258816 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 09:10:51.995770 134292041258816 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 09:10:51.995828 134292041258816 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 09:10:52.021076 134292041258816 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0421 09:10:52.021142 134292041258816 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 09:10:52.042071 134292041258816 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0421 09:10:52.042134 134292041258816 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0421 09:10:52.257053 134292041258816 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0421 09:10:52.257157 134292041258816 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0421 09:10:53.653857 134292041258816 maxtext_utils.py:1821]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0421 09:10:53.653981 134292041258816 maxtext_utils.py:1821]  params/params/decoder/layers/mlp/wi_0/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0421 09:10:53.654037 134292041258816 maxtext_utils.py:1821]  params/params/decoder/layers/mlp/wi_1/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0421 09:10:53.654104 134292041258816 maxtext_utils.py:1821]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   P('mlp', 'layers', 'embed')
    Physical:  (None, None, 'fsdp')
I0421 09:10:53.654160 134292041258816 maxtext_utils.py:1821]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0421 09:10:53.654203 134292041258816 maxtext_utils.py:1821]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0421 09:10:53.654257 134292041258816 maxtext_utils.py:1821]  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 09:10:53.654312 134292041258816 maxtext_utils.py:1821]  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 09:10:53.654355 134292041258816 maxtext_utils.py:1821]  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 09:10:53.654392 134292041258816 maxtext_utils.py:1821]  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 09:10:53.654441 134292041258816 maxtext_utils.py:1821]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)

I0421 09:10:53.654492 134292041258816 maxtext_utils.py:1821]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')
I0421 09:10:54.142816 134292041258816 train.py:157] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 09:10:54.142907 134292041258816 train.py:157] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 09:10:54.158197 134292041258816 train.py:164] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 09:10:54.158258 134292041258816 train.py:164] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 09:11:04.837998 134292041258816 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 09:11:04.838753 134292041258816 metric_logger.py:301] number parameters: 1.104 billion
I0421 09:11:15.820400 134292041258816 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0421 09:11:16.418406 134292041258816 checkpointing.py:798] Waited 0.5979838371276855 seconds for step 0 to finish before starting checkpointing.
I0421 09:11:16.422150 134292041258816 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 09:11:16.423562 134292041258816 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 0
I0421 09:11:16.424928 134292041258816 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_10_shardy_false/checkpoints/0.
I0421 09:11:17.196480 134292041258816 signaling_client.py:364] Using JaxDistributedSignalingClient
I0421 09:11:17.197668 134292041258816 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 09:11:17.197808 134292041258816 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 09:11:17.474518 134292041258816 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.278112s
I0421 09:11:17.474686 134292041258816 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 5.439 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28363537788391113 s) (per-host)
I0421 09:11:17.474766 134292041258816 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.283724s (batch_requests_ready=0.002330s, total_serialization_initiated=0.281298s, others=0.000097s)
I0421 09:11:17.474863 134292041258816 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.287843s (all_items=0.000017s, per_item={'items': '0.00001740'}, temp_paths=0.287826)
I0421 09:11:17.475666 134292041258816 event_tracking.py:125] [process=4] [async] Finished blocking save in 1.05 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_10_shardy_false/checkpoints/0.
I0421 09:11:17.476015 134162698729216 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 09:31:17.475976
I0421 09:11:17.481328 134292041258816 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 09:11:17.481608 134161587230464 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 09:11:17.481748 134292041258816 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776762676.4221313, 'wait_for_prev_duration_secs': 6.127357482910156e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776762676.4236002, 'checkpointer_blocking_duration_secs': 1.052565574645996, 'get_old_steps_start_time': 1776762677.4761884, 'get_old_steps_duration_secs': 3.0517578125e-05, 'checkpoint_manager_blocking_start_time': 1776762676.4189324, 'checkpoint_manager_blocking_duration_secs': 1.0627782344818115}
I0421 09:11:17.481864 134292041258816 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0421 09:11:17.481914 134292041258816 max_utils.py:750] 
Memstats: After params initialized:
I0421 09:11:17.481966 134292041258816 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_16(process=4,(0,4,0,0))
I0421 09:11:17.481995 134292041258816 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_17(process=4,(1,4,0,0))
I0421 09:11:17.482023 134292041258816 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_20(process=4,(0,5,0,0))
I0421 09:11:17.482048 134292041258816 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_21(process=4,(1,5,0,0))
I0421 09:11:17.796446 134292041258816 metric_logger.py:196] completed step: 0, seconds: 10.982, TFLOP/s/device: 1.237, Tokens/s/device: 186.495, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0421 09:11:17.970551 134292041258816 metric_logger.py:196] completed step: 1, seconds: 1.975, TFLOP/s/device: 6.881, Tokens/s/device: 1037.188, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0421 09:11:18.370229 134292041258816 metric_logger.py:196] completed step: 2, seconds: 0.027, TFLOP/s/device: 507.228, Tokens/s/device: 76454.997, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129
I0421 09:11:18.517707 134292041258816 metric_logger.py:196] completed step: 3, seconds: 0.400, TFLOP/s/device: 33.968, Tokens/s/device: 5119.962, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203
I0421 09:11:18.812775 134292041258816 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.897, Tokens/s/device: 13399.458, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614
I0421 09:11:18.818591 134292041258816 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.184, Tokens/s/device: 13895.014, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717
I0421 09:11:20.403543    2810 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0421 09:11:22.410659 134161595623168 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_10_shardy_false/checkpoints/0/items/array_metadatas/process_4
I0421 09:11:40.712626 134292041258816 metric_logger.py:196] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.976, Tokens/s/device: 6930.016, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759
I0421 09:11:40.859936 134292041258816 metric_logger.py:196] completed step: 7, seconds: 21.747, TFLOP/s/device: 0.625, Tokens/s/device: 94.173, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905
I0421 09:11:41.007246 134292041258816 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.321, Tokens/s/device: 13463.410, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406
I0421 09:11:41.154008 134292041258816 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0421 09:11:41.154581 134292041258816 checkpointing.py:798] Waited 0.0005936622619628906 seconds for step 9 to finish before starting checkpointing.
I0421 09:11:41.156810 134292041258816 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 09:11:54.945503 134162698729216 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 41.839 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.75441384315491 s) (per-host)
I0421 09:11:54.945625 134162698729216 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.469498s.
I0421 09:12:02.354093 134162698729216 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 44.877947s.
I0421 09:12:02.354321 134161587230464 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 09:12:02.354392 134161587230464 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 09:12:02.354447 134161587230464 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0421 09:12:02.356012 134161587230464 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0421 09:12:02.356173 134292041258816 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.
W0421 09:12:02.356308 134292041258816 checkpoint_manager.py:1452] Waiting for previous save to complete took 21.199498 seconds. If this number is high, consider checkpointing less frequently.
I0421 09:12:02.358223 134292041258816 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 9
I0421 09:12:02.360290 134292041258816 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_10_shardy_false/checkpoints/9.
I0421 09:12:03.074122 134292041258816 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 09:12:03.074219 134292041258816 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 09:12:03.110717 134292041258816 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.037629s
I0421 09:12:03.110888 134292041258816 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 37.423 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04122042655944824 s) (per-host)
I0421 09:12:03.110938 134292041258816 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.041280s (batch_requests_ready=0.001909s, total_serialization_initiated=0.039303s, others=0.000068s)
I0421 09:12:03.111030 134292041258816 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.045476s (all_items=0.000015s, per_item={'items': '0.00001526'}, temp_paths=0.045461)
I0421 09:12:03.111759 134292041258816 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_10_shardy_false/checkpoints/9.
I0421 09:12:03.112069 134161587230464 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 09:32:03.112036
I0421 09:12:03.113990 134292041258816 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 09:12:03.114267 134140984809216 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 09:12:03.114401 134292041258816 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776762701.1567805, 'wait_for_prev_duration_secs': 21.199497938156128, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776762722.358264, 'checkpointer_blocking_duration_secs': 0.7539465427398682, 'get_old_steps_start_time': 1776762723.1122372, 'get_old_steps_duration_secs': 2.9802322387695312e-05, 'checkpoint_manager_blocking_start_time': 1776762701.154824, 'checkpoint_manager_blocking_duration_secs': 21.959545850753784}
I0421 09:12:03.114572 134292041258816 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0421 09:12:03.114617 134292041258816 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 09:12:08.663576 134161595623168 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_10_shardy_false/checkpoints/9/items/array_metadatas/process_4
I0421 09:12:44.843667 134161587230464 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 37.813 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.773966550827026 s) (per-host)
I0421 09:12:44.843786 134161587230464 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.731603s.
I0421 09:12:53.601436 134161587230464 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 50.489233s.
I0421 09:12:53.601823 134140984809216 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 09:12:53.601940 134140984809216 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 09:12:53.601990 134140984809216 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0421 09:12:53.604284 134140984809216 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0421 09:12:53.604467 134292041258816 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.
I0421 09:12:53.604639 134292041258816 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 09:12:53.605456 134292041258816 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.246, Tokens/s/device: 13904.259, 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: Tue Apr 21 09:13:05 UTC 2026
EXIT_CODE=0