MaxView

← Back to run

Log Summary

XPK Start: Mon Apr 20 13:49:33 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 13:49:57.607286: 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 13:49:57.784955 136767582992192 max_utils.py:273] Attempting to initialize the jax distributed system...
I0420 13:50:06.827689 136767582992192 distributed.py:149] Starting JAX distributed service on [::]:8482
I0420 13:50:06.830151 136767582992192 distributed.py:172] Connecting to JAX distributed service on mt-01-base-3orra-slice-job-0-0.mt-01-base-3orra:8482
I0420 13:50:08.058714 136767582992192 max_utils.py:284] Jax distributed system initialized!
I0420 13:50:14.389727 136767582992192 max_utils.py:800] System Information: Jax Version: 0.9.2
I0420 13:50:14.389826 136767582992192 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0420 13:50:14.389866 136767582992192 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 13:50:14.389902 136767582992192 train_utils.py:378] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0420 13:50:15.431622 136767582992192 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0420 13:50:15.545255 136767582992192 checkpointing.py:688] Setting up checkpoint logger...
I0420 13:50:15.545364 136767582992192 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0420 13:50:15.545409 136767582992192 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0420 13:50:15.545602 136767582992192 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 0x7c62f7848890>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0420 13:50:19.161854 136767582992192 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0420 13:50:19.162042 136767582992192 checkpoint_manager.py:708] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c61ebc38ad0>}, handler_registry=None
I0420 13:50:19.162329 136767582992192 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c61ebc38ad0>` 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 13:50:19.162384 136767582992192 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c4f18256660>` 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 13:50:19.162431 136767582992192 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 0x7c61ebc38ad0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c61ebc38ad0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c4f18256660>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c4f18256660>}).
I0420 13:50:19.162776 136767582992192 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0420 13:50:19.162858 136767582992192 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7c4f186d85e0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0420 13:50:20.774737 136767582992192 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints
I0420 13:50:21.201319 136767582992192 checkpoint_manager.py:929] [process=0][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/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7c61ebe197c0>
I0420 13:50:21.201490 136767582992192 checkpointing.py:302] Checkpoint manager created!
I0420 13:50:22.161035 136767582992192 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0420 13:50:22.161163 136767582992192 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0420 13:50:23.554853 136767582992192 maxtext_utils.py:1836]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0420 13:50:23.554966 136767582992192 maxtext_utils.py:1836]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0420 13:50:23.555018 136767582992192 maxtext_utils.py:1836]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0420 13:50:23.555060 136767582992192 maxtext_utils.py:1836]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0420 13:50:23.555110 136767582992192 maxtext_utils.py:1836]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0420 13:50:23.555146 136767582992192 maxtext_utils.py:1836]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0420 13:50:23.555181 136767582992192 maxtext_utils.py:1836]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 13:50:23.555216 136767582992192 maxtext_utils.py:1836]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0420 13:50:23.555248 136767582992192 maxtext_utils.py:1836]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 13:50:23.555280 136767582992192 maxtext_utils.py:1836]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 13:50:23.555311 136767582992192 maxtext_utils.py:1836]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)
I0420 13:50:23.555343 136767582992192 maxtext_utils.py:1836]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')

I0420 13:50:23.955815 136767582992192 nnx_decoders.py:465] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 13:50:23.955904 136767582992192 nnx_decoders.py:465] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 13:50:23.962273 136767582992192 nnx_decoders.py:465] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 13:50:23.962333 136767582992192 nnx_decoders.py:465] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 13:50:23.978678 136767582992192 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 13:50:23.978737 136767582992192 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 13:50:23.994531 136767582992192 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 13:50:23.994587 136767582992192 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 13:50:24.018362 136767582992192 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 13:50:24.018429 136767582992192 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 13:50:24.039536 136767582992192 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 13:50:24.039653 136767582992192 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 13:50:24.055371 136767582992192 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 13:50:24.055432 136767582992192 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 13:50:24.085371 136767582992192 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0420 13:50:24.085441 136767582992192 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 13:50:24.104987 136767582992192 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0420 13:50:24.105051 136767582992192 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0420 13:50:35.568269 136767582992192 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 13:50:35.577795 136767582992192 metric_logger.py:301] number parameters: 1.104 billion
I0420 13:50:47.168127 136767582992192 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0420 13:50:47.319589 136767582992192 checkpointing.py:798] Waited 0.15143489837646484 seconds for step 0 to finish before starting checkpointing.
I0420 13:50:47.321943 136767582992192 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 13:50:47.323646 136767582992192 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 0
I0420 13:50:47.325046 136767582992192 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0.
I0420 13:50:47.652037 136767582992192 signaling_client.py:364] Using JaxDistributedSignalingClient
I0420 13:50:47.671710 136767582992192 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0420 13:50:47.671781 136767582992192 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 13:50:47.949806 136638671873792 atomicity.py:140] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0
I0420 13:50:47.951507 136767582992192 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.281219s
I0420 13:50:47.952336 136767582992192 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.280 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.29213929176330566 s) (per-host)
I0420 13:50:47.952422 136767582992192 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.292249s (batch_requests_ready=0.003042s, total_serialization_initiated=0.288450s, others=0.000758s)
I0420 13:50:47.952507 136767582992192 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.296695s (all_items=0.000032s, per_item={'items': '0.00003242'}, temp_paths=0.296662)
I0420 13:50:47.953195 136767582992192 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.63 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0.
I0420 13:50:47.953499 136637549242112 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 14:10:47.953459
I0420 13:50:47.957437 136767582992192 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 13:50:47.957662 136638680266496 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0420 13:50:47.957784 136767582992192 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776693047.3219156, 'wait_for_prev_duration_secs': 7.224082946777344e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776693047.3236895, 'checkpointer_blocking_duration_secs': 0.6299676895141602, 'get_old_steps_start_time': 1776693047.9536817, 'get_old_steps_duration_secs': 2.8371810913085938e-05, 'checkpoint_manager_blocking_start_time': 1776693047.319987, 'checkpoint_manager_blocking_duration_secs': 0.6377613544464111}
I0420 13:50:47.957896 136767582992192 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0420 13:50:47.957966 136767582992192 max_utils.py:750] 
Memstats: After params initialized:
I0420 13:50:47.958013 136767582992192 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0))
I0420 13:50:47.958045 136767582992192 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0))
I0420 13:50:47.958071 136767582992192 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0))
I0420 13:50:47.958107 136767582992192 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0))
I0420 13:50:48.308409 136767582992192 metric_logger.py:196] completed step: 0, seconds: 11.252, TFLOP/s/device: 1.208, Tokens/s/device: 182.020, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52786.469
I0420 13:50:48.311596 136767582992192 metric_logger.py:281] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/tensorboard/'
I0420 13:50:48.758564 136767582992192 metric_logger.py:196] completed step: 1, seconds: 1.137, TFLOP/s/device: 11.955, Tokens/s/device: 1801.932, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52786.469
I0420 13:50:48.900364 136767582992192 metric_logger.py:196] completed step: 2, seconds: 0.451, TFLOP/s/device: 30.126, Tokens/s/device: 4540.869, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28632.133
I0420 13:50:49.048122 136767582992192 metric_logger.py:196] completed step: 3, seconds: 0.013, TFLOP/s/device: 1045.485, Tokens/s/device: 157586.950, total_weights: 65536, loss: 9.732, lm_loss: 9.732, perplexity: 16855.789
I0420 13:50:49.817995 136638655088384 atomicity.py:140] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0/items
I0420 13:50:49.957582 136633541256960 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776693049526717107, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776693049526717107, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0/_CHECKPOINT_METADATA
I0420 13:50:50.954436    2808 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0420 13:50:52.761049 136637603362560 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0/items/array_metadatas/process_0
I0420 13:51:11.344677 136767582992192 metric_logger.py:196] completed step: 4, seconds: 0.143, TFLOP/s/device: 94.759, Tokens/s/device: 14283.124, total_weights: 65536, loss: 9.274, lm_loss: 9.274, perplexity: 10661.594
I0420 13:51:11.355245 136767582992192 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.407, Tokens/s/device: 13928.562, total_weights: 65536, loss: 8.888, lm_loss: 8.888, perplexity: 7245.978
I0420 13:51:11.499780 136767582992192 metric_logger.py:196] completed step: 6, seconds: 22.295, TFLOP/s/device: 0.609, Tokens/s/device: 91.861, total_weights: 65536, loss: 8.589, lm_loss: 8.589, perplexity: 5371.156
I0420 13:51:11.647284 136767582992192 metric_logger.py:196] completed step: 7, seconds: 0.008, TFLOP/s/device: 1629.543, Tokens/s/device: 245622.451, total_weights: 65536, loss: 8.380, lm_loss: 8.380, perplexity: 4360.465
I0420 13:51:11.794725 136767582992192 metric_logger.py:196] completed step: 8, seconds: 0.146, TFLOP/s/device: 93.103, Tokens/s/device: 14033.453, total_weights: 65536, loss: 8.251, lm_loss: 8.251, perplexity: 3831.256
I0420 13:51:11.942764 136767582992192 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0420 13:51:11.943516 136767582992192 checkpointing.py:798] Waited 0.0007648468017578125 seconds for step 9 to finish before starting checkpointing.
I0420 13:51:11.945461 136767582992192 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 13:51:28.236940 136637557634816 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.280446s (commit=35.968823s, array_metadata_write=1.311623s)
I0420 13:51:28.238187 136637549242112 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 38.928 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 40.577948808670044 s) (per-host)
I0420 13:51:28.238322 136637549242112 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 40.284707s.
I0420 13:51:29.017252 136637549242112 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776693049526717107, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0/_CHECKPOINT_METADATA
I0420 13:51:29.741348 136637549242112 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000376s.
I0420 13:51:29.782929 136633541256960 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776693049526717107, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0/_CHECKPOINT_METADATA
I0420 13:51:30.560498 136637549242112 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0420 13:51:31.517873 136637549242112 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.373265s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0/items
I0420 13:51:31.519618 136637549242112 atomicity.py:666] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0/items
I0420 13:51:32.073238 136637549242112 atomicity.py:666] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0
I0420 13:51:33.647851 136637549242112 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0`.
I0420 13:51:33.648725 136637549242112 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 46.32 seconds @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/0
I0420 13:51:33.650239 136637549242112 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 45.696622s.
I0420 13:51:33.650432 136638680266496 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 13:51:33.650558 136638680266496 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0420 13:51:33.650623 136638680266496 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0420 13:51:33.653279 136638680266496 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0420 13:51:33.653471 136767582992192 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0420 13:51:33.653606 136767582992192 checkpoint_manager.py:1452] Waiting for previous save to complete took 21.708146 seconds. If this number is high, consider checkpointing less frequently.
I0420 13:51:33.655514 136767582992192 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 9
I0420 13:51:33.657616 136767582992192 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/9.
I0420 13:51:34.389385 136767582992192 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0420 13:51:34.389482 136767582992192 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 13:51:34.417650 136767582992192 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.029757s
I0420 13:51:34.418464 136767582992192 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 40.121 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03844809532165527 s) (per-host)
I0420 13:51:34.418525 136767582992192 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.038525s (batch_requests_ready=0.002601s, total_serialization_initiated=0.035202s, others=0.000722s)
I0420 13:51:34.418601 136767582992192 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.042568s (all_items=0.000018s, per_item={'items': '0.00001788'}, temp_paths=0.042550)
I0420 13:51:34.419269 136767582992192 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.76 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/9.
I0420 13:51:34.419546 136638680266496 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 14:11:34.419523
I0420 13:51:34.428997 136767582992192 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 13:51:34.429239 136636490819328 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0420 13:51:34.429366 136767582992192 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776693071.9454277, 'wait_for_prev_duration_secs': 21.7081458568573, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776693093.6555533, 'checkpointer_blocking_duration_secs': 0.7641003131866455, 'get_old_steps_start_time': 1776693094.4196777, 'get_old_steps_duration_secs': 3.075599670410156e-05, 'checkpoint_manager_blocking_start_time': 1776693071.9437392, 'checkpoint_manager_blocking_duration_secs': 22.485591650009155}
I0420 13:51:34.429494 136767582992192 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0420 13:51:34.429540 136767582992192 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 13:51:34.634294 136637549242112 atomicity.py:140] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/9
I0420 13:51:36.515655 136613942208256 atomicity.py:140] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/9/items
I0420 13:51:39.440390 136636507604736 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/9/items/array_metadatas/process_0
I0420 13:52:19.608143 136636499212032 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.971172s (commit=40.722971s, array_metadata_write=1.248201s)
I0420 13:52:19.610315 136638680266496 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 34.924 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.23027181625366 s) (per-host)
I0420 13:52:19.610426 136638680266496 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.190794s.
I0420 13:52:21.326666 136638680266496 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0420 13:52:21.494881 136638680266496 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000406s.
I0420 13:52:21.971719 136638680266496 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.479841s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/9/items
I0420 13:52:21.973568 136638680266496 atomicity.py:666] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/9/items
I0420 13:52:22.521052 136638680266496 atomicity.py:666] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/9
I0420 13:52:23.961120 136638680266496 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/9`.
I0420 13:52:23.962005 136638680266496 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 50.31 seconds @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158/nnx_xpk_feat_nnx_linen_converter_and_sharding_tools_20260420_112158_01_base/checkpoints/9
I0420 13:52:23.963725 136638680266496 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 49.544091s.
I0420 13:52:23.963916 136636490819328 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 13:52:23.964032 136636490819328 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0420 13:52:23.964108 136636490819328 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0420 13:52:23.966972 136636490819328 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0420 13:52:23.967121 136767582992192 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0420 13:52:23.967245 136767582992192 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 13:52:23.968315 136767582992192 metric_logger.py:196] completed step: 9, seconds: 0.148, TFLOP/s/device: 92.111, Tokens/s/device: 13883.993, total_weights: 65536, loss: 8.175, lm_loss: 8.175, perplexity: 3551.438
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Mon Apr 20 13:52:34 UTC 2026
EXIT_CODE=0