MaxView

← Back to run

Log Summary

XPK Start: Mon Apr 20 13:55:43 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 13:56:07.348239: 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 13:56:07.526261 133423096973120 max_utils.py:273] Attempting to initialize the jax distributed system...
I0420 13:56:16.567550 133423096973120 distributed.py:149] Starting JAX distributed service on [::]:8482
I0420 13:56:16.569875 133423096973120 distributed.py:172] Connecting to JAX distributed service on mt-03-dropout-4q4bk-slice-job-0-0.mt-03-dropout-4q4bk:8482
I0420 13:56:17.277732 133423096973120 max_utils.py:284] Jax distributed system initialized!
I0420 13:56:23.376187 133423096973120 max_utils.py:800] System Information: Jax Version: 0.9.2
I0420 13:56:23.376287 133423096973120 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0420 13:56:23.376327 133423096973120 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 13:56:23.376360 133423096973120 train_utils.py:378] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0420 13:56:24.067476 133423096973120 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0420 13:56:24.179882 133423096973120 checkpointing.py:688] Setting up checkpoint logger...
I0420 13:56:24.179997 133423096973120 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0420 13:56:24.180042 133423096973120 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0420 13:56:24.180253 133423096973120 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 0x795865717b90>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0420 13:56:27.193420 133423096973120 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0420 13:56:27.193696 133423096973120 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 0x7943ec40fef0>}, handler_registry=None
I0420 13:56:27.193942 133423096973120 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7943ec40fef0>` 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 13:56:27.193992 133423096973120 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7943ec489880>` 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 13:56:27.194028 133423096973120 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 0x7943ec40fef0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7943ec40fef0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7943ec489880>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7943ec489880>}).
I0420 13:56:27.194351 133423096973120 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0420 13:56:27.194421 133423096973120 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7943ec4e5120> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0420 13:56:28.414373 133423096973120 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_03_dropout/checkpoints
I0420 13:56:28.416620 133423096973120 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_03_dropout/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7942f4607b60>
I0420 13:56:28.416730 133423096973120 checkpointing.py:302] Checkpoint manager created!
I0420 13:56:28.775812 133423096973120 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0420 13:56:28.775933 133423096973120 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0420 13:56:30.245004 133423096973120 maxtext_utils.py:1836]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0420 13:56:30.245130 133423096973120 maxtext_utils.py:1836]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0420 13:56:30.245187 133423096973120 maxtext_utils.py:1836]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0420 13:56:30.245231 133423096973120 maxtext_utils.py:1836]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0420 13:56:30.245266 133423096973120 maxtext_utils.py:1836]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0420 13:56:30.245300 133423096973120 maxtext_utils.py:1836]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0420 13:56:30.245335 133423096973120 maxtext_utils.py:1836]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 13:56:30.245369 133423096973120 maxtext_utils.py:1836]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0420 13:56:30.245402 133423096973120 maxtext_utils.py:1836]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 13:56:30.245433 133423096973120 maxtext_utils.py:1836]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 13:56:30.245465 133423096973120 maxtext_utils.py:1836]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)
I0420 13:56:30.245496 133423096973120 maxtext_utils.py:1836]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')

I0420 13:56:30.653808 133423096973120 nnx_decoders.py:465] nnx_decoders/carry Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 13:56:30.653906 133423096973120 nnx_decoders.py:465] nnx_decoders/carry Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0420 13:56:30.660071 133423096973120 nnx_decoders.py:465] Unknown Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 13:56:30.660142 133423096973120 nnx_decoders.py:465] Unknown Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0420 13:56:30.676516 133423096973120 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 13:56:30.676574 133423096973120 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0420 13:56:30.692137 133423096973120 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 13:56:30.692194 133423096973120 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0420 13:56:30.716239 133423096973120 attentions.py:1154] attentions/query Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 13:56:30.716310 133423096973120 attentions.py:1154] attentions/query Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0420 13:56:30.732068 133423096973120 attentions.py:1155] attentions/key Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 13:56:30.732143 133423096973120 attentions.py:1155] attentions/key Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0420 13:56:30.747813 133423096973120 attentions.py:1156] attentions/value Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 13:56:30.747874 133423096973120 attentions.py:1156] attentions/value Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0420 13:56:30.777841 133423096973120 attentions.py:1197] attentions/out Logical: bfloat16[32,128,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0420 13:56:30.777913 133423096973120 attentions.py:1197] attentions/out Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0420 13:56:30.803671 133423096973120 linears.py:525] linears/x Logical: bfloat16[32,128,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0420 13:56:30.803737 133423096973120 linears.py:525] linears/x Physical: bfloat16[32,128,7168]....................................... ('fsdp', None, None).
I0420 13:56:34.598072 133423096973120 max_utils.py:791] Total memory size: 0.8 GB, Output size: 0.4 GB, Temp size: 0.4 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0420 13:56:34.607529 133423096973120 metric_logger.py:301] number parameters: 1.104 billion
I0420 13:56:38.876276 133423096973120 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0420 13:56:38.948857 133423096973120 checkpointing.py:798] Waited 0.07256674766540527 seconds for step 0 to finish before starting checkpointing.
I0420 13:56:38.951274 133423096973120 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 13:56:38.953268 133423096973120 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 0
I0420 13:56:38.954778 133423096973120 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_03_dropout/checkpoints/0.
I0420 13:56:39.237828 133423096973120 signaling_client.py:364] Using JaxDistributedSignalingClient
I0420 13:56:39.257177 133423096973120 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0420 13:56:39.257246 133423096973120 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 13:56:39.516978 133294813329152 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_03_dropout/checkpoints/0
I0420 13:56:39.533641 133423096973120 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.277886s
I0420 13:56:39.534487 133423096973120 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.338 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2889838218688965 s) (per-host)
I0420 13:56:39.534563 133423096973120 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.289081s (batch_requests_ready=0.003213s, total_serialization_initiated=0.285093s, others=0.000775s)
I0420 13:56:39.534646 133423096973120 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.293155s (all_items=0.000018s, per_item={'items': '0.00001764'}, temp_paths=0.293138)
I0420 13:56:39.535452 133423096973120 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.58 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_03_dropout/checkpoints/0.
I0420 13:56:39.535745 133294821721856 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 14:16:39.535713
I0420 13:56:39.548745 133423096973120 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 13:56:39.549002 133294255486720 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0420 13:56:39.549147 133423096973120 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_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776693398.951255, 'wait_for_prev_duration_secs': 6.103515625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776693398.9533088, 'checkpointer_blocking_duration_secs': 0.5825815200805664, 'get_old_steps_start_time': 1776693399.5359156, 'get_old_steps_duration_secs': 2.6464462280273438e-05, 'checkpoint_manager_blocking_start_time': 1776693398.9492378, 'checkpoint_manager_blocking_duration_secs': 0.5998709201812744}
I0420 13:56:39.549313 133423096973120 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0420 13:56:39.549387 133423096973120 max_utils.py:750] 
Memstats: After params initialized:
I0420 13:56:39.549439 133423096973120 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_0(process=0,(0,0,0,0))
I0420 13:56:39.549475 133423096973120 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_1(process=0,(1,0,0,0))
I0420 13:56:39.549505 133423096973120 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_4(process=0,(0,1,0,0))
I0420 13:56:39.549528 133423096973120 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_5(process=0,(1,1,0,0))
I0420 13:56:39.907934 133423096973120 metric_logger.py:196] completed step: 0, seconds: 4.068, TFLOP/s/device: 0.197, Tokens/s/device: 31.467, total_weights: 4096, loss: 10.874, lm_loss: 10.874, perplexity: 52805.141
I0420 13:56:39.911362 133423096973120 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_03_dropout/tensorboard/'
I0420 13:56:40.344421 133423096973120 metric_logger.py:196] completed step: 1, seconds: 1.028, TFLOP/s/device: 0.779, Tokens/s/device: 124.524, total_weights: 4096, loss: 10.892, lm_loss: 10.892, perplexity: 53764.488
I0420 13:56:40.409189 133423096973120 metric_logger.py:196] completed step: 2, seconds: 0.438, TFLOP/s/device: 1.831, Tokens/s/device: 292.570, total_weights: 4096, loss: 9.969, lm_loss: 9.969, perplexity: 21350.100
I0420 13:56:40.479913 133423096973120 metric_logger.py:196] completed step: 3, seconds: 0.013, TFLOP/s/device: 63.587, Tokens/s/device: 10162.763, total_weights: 4096, loss: 9.125, lm_loss: 9.125, perplexity: 9186.262
I0420 13:56:41.911747 133293236287232 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776693401492117358, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776693401492117358, "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_03_dropout/checkpoints/0/_CHECKPOINT_METADATA
I0420 13:56:42.189842 133294796543744 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_03_dropout/checkpoints/0/items
I0420 13:56:43.348491    2784 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0420 13:56:45.181933 133294272272128 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_03_dropout/checkpoints/0/items/array_metadatas/process_0
I0420 13:56:50.671185 133423096973120 metric_logger.py:196] completed step: 4, seconds: 0.067, TFLOP/s/device: 11.989, Tokens/s/device: 1916.139, total_weights: 4096, loss: 8.469, lm_loss: 8.469, perplexity: 4765.735
I0420 13:56:50.683145 133423096973120 metric_logger.py:196] completed step: 5, seconds: 0.070, TFLOP/s/device: 11.521, Tokens/s/device: 1841.276, total_weights: 4096, loss: 7.908, lm_loss: 7.908, perplexity: 2719.938
I0420 13:56:50.749663 133423096973120 metric_logger.py:196] completed step: 6, seconds: 10.191, TFLOP/s/device: 0.079, Tokens/s/device: 12.560, total_weights: 4096, loss: 7.554, lm_loss: 7.554, perplexity: 1909.063
I0420 13:56:51.182925 133423096973120 metric_logger.py:196] completed step: 7, seconds: 0.008, TFLOP/s/device: 96.596, Tokens/s/device: 15438.427, total_weights: 4096, loss: 7.313, lm_loss: 7.313, perplexity: 1500.206
I0420 13:56:51.192458 133423096973120 metric_logger.py:196] completed step: 8, seconds: 0.491, TFLOP/s/device: 1.630, Tokens/s/device: 260.539, total_weights: 4096, loss: 7.091, lm_loss: 7.091, perplexity: 1200.912
I0420 13:56:51.261186 133423096973120 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0420 13:56:51.261889 133423096973120 checkpointing.py:798] Waited 0.0007131099700927734 seconds for step 9 to finish before starting checkpointing.
I0420 13:56:51.264186 133423096973120 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 13:57:15.844723 133294263879424 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 32.493562s (commit=31.211291s, array_metadata_write=1.282271s)
I0420 13:57:15.845785 133294821721856 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 43.158 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 36.60025501251221 s) (per-host)
I0420 13:57:15.845923 133294821721856 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.310056s.
I0420 13:57:16.642491 133294821721856 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776693401492117358, '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_03_dropout/checkpoints/0/_CHECKPOINT_METADATA
I0420 13:57:17.372014 133293236287232 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776693401492117358, '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_03_dropout/checkpoints/0/_CHECKPOINT_METADATA
I0420 13:57:17.382929 133294821721856 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000473s.
I0420 13:57:18.160614 133294821721856 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0420 13:57:19.133688 133294821721856 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.364480s. 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_03_dropout/checkpoints/0/items
I0420 13:57:19.135591 133294821721856 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_03_dropout/checkpoints/0/items
I0420 13:57:19.692914 133294821721856 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_03_dropout/checkpoints/0
I0420 13:57:21.095814 133294821721856 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_03_dropout/checkpoints/0`.
I0420 13:57:21.096517 133294821721856 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 42.14 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_03_dropout/checkpoints/0
I0420 13:57:21.097773 133294821721856 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 41.561914s.
I0420 13:57:21.097982 133294255486720 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 13:57:21.098127 133294255486720 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0420 13:57:21.098194 133294255486720 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0420 13:57:21.100034 133294255486720 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0420 13:57:21.100176 133423096973120 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 13:57:21.100261 133423096973120 checkpoint_manager.py:1452] Waiting for previous save to complete took 29.836078 seconds. If this number is high, consider checkpointing less frequently.
I0420 13:57:21.102196 133423096973120 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 9
I0420 13:57:21.104299 133423096973120 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_03_dropout/checkpoints/9.
I0420 13:57:22.831666 133423096973120 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0420 13:57:22.831762 133423096973120 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 13:57:22.865370 133423096973120 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.035120s
I0420 13:57:22.866250 133423096973120 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 34.887 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04421663284301758 s) (per-host)
I0420 13:57:22.866326 133423096973120 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.044310s (batch_requests_ready=0.002876s, total_serialization_initiated=0.040627s, others=0.000807s)
I0420 13:57:22.866411 133423096973120 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.048446s (all_items=0.000017s, per_item={'items': '0.00001740'}, temp_paths=0.048429)
I0420 13:57:22.867171 133423096973120 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.76 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_03_dropout/checkpoints/9.
I0420 13:57:22.867519 133294255486720 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 14:17:22.867482
I0420 13:57:22.869521 133423096973120 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 13:57:22.869797 133292643251968 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0420 13:57:22.869961 133423096973120 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_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776693411.2641551, 'wait_for_prev_duration_secs': 29.83607840538025, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776693441.1022336, 'checkpointer_blocking_duration_secs': 1.7654194831848145, 'get_old_steps_start_time': 1776693442.8676765, 'get_old_steps_duration_secs': 2.9325485229492188e-05, 'checkpoint_manager_blocking_start_time': 1776693411.2621577, 'checkpoint_manager_blocking_duration_secs': 31.60776996612549}
I0420 13:57:22.870092 133423096973120 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0420 13:57:22.870140 133423096973120 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 13:57:23.078173 133294821721856 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_03_dropout/checkpoints/9
I0420 13:57:25.758610 133293771052800 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_03_dropout/checkpoints/9/items
I0420 13:57:29.236642 133292660037376 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_03_dropout/checkpoints/9/items/array_metadatas/process_0
I0420 13:58:08.750507 133292651644672 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.349327s (commit=40.123628s, array_metadata_write=1.225698s)
I0420 13:58:08.751717 133294255486720 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 34.392 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.929654598236084 s) (per-host)
I0420 13:58:08.751849 133294255486720 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.884219s.
I0420 13:58:10.527858 133294255486720 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000423s.
I0420 13:58:10.530459 133294255486720 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0420 13:58:11.589123 133294255486720 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.048446s. 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_03_dropout/checkpoints/9/items
I0420 13:58:11.590887 133294255486720 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_03_dropout/checkpoints/9/items
I0420 13:58:12.176865 133294255486720 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_03_dropout/checkpoints/9
I0420 13:58:13.557506 133294255486720 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_03_dropout/checkpoints/9`.
I0420 13:58:13.558336 133294255486720 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 52.46 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_03_dropout/checkpoints/9
I0420 13:58:13.559729 133294255486720 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 50.692097s.
I0420 13:58:13.559916 133292643251968 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 13:58:13.560034 133292643251968 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0420 13:58:13.560110 133292643251968 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0420 13:58:13.562075 133292643251968 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0420 13:58:13.562267 133423096973120 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 13:58:13.562466 133423096973120 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 13:58:13.563642 133423096973120 metric_logger.py:196] completed step: 9, seconds: 0.010, TFLOP/s/device: 79.444, Tokens/s/device: 12697.153, total_weights: 4096, loss: 7.016, lm_loss: 7.016, perplexity: 1113.843
Per train step:
 Total TFLOPs: 0.80 
 split as 99.60% learnable weight flops and 0.40% attention flops
XPK End: Mon Apr 20 13:58:25 UTC 2026
EXIT_CODE=0