2026-04-16 13:08:46.322031: 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)
I0416 13:08:46.437951 135676222270592 max_utils.py:238] Skipping jax distributed system due to skip_jax_distributed_system=True flag.
I0416 13:09:42.411810 135676222270592 max_utils.py:800] System Information: Jax Version: 0.9.2
I0416 13:09:42.411918 135676222270592 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0416 13:09:42.411952 135676222270592 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Apr 6 2026 20:48:10 (1775533690) cl/895581894
I0416 13:09:42.411978 135676222270592 train_utils.py:334] WARNING: 'dataset_path' might be pointing your local file system
I0416 13:09:42.412074 135676222270592 train.py:683] [DECOUPLED NO-OP] skipping cloud diagnostics wrapper.
I0416 13:09:42.916925 135676222270592 maxtext_utils.py:1517] Num_devices: 8, shape (1, 1, 1, 8, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0416 13:09:42.917213 135676222270592 checkpointing.py:677] Setting up checkpoint logger...
I0416 13:09:42.917259 135676222270592 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0416 13:09:42.917299 135676222270592 pytree_checkpoint_handler.py:577] save_device_host_concurrent_bytes=None
I0416 13:09:42.917796 135676222270592 base_pytree_checkpoint_handler.py:411] 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 0x7b64febf8a70>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0416 13:09:45.374950 135676222270592 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0416 13:09:45.375154 135676222270592 checkpoint_manager.py:702] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b5ea616fec0>}, handler_registry=None
I0416 13:09:45.375411 135676222270592 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b5ea616fec0>` 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`.
I0416 13:09:45.375452 135676222270592 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b5e8233a0c0>` 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`.
I0416 13:09:45.375482 135676222270592 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 0x7b5ea616fec0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b5ea616fec0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b5e8233a0c0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b5e8233a0c0>}).
I0416 13:09:45.375964 135676222270592 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.28
I0416 13:09:45.376013 135676222270592 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>.<lambda> at 0x7b5e82331300> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0416 13:09:45.510121 135676222270592 checkpoint_manager.py:1788] Found 0 checkpoint steps in gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints
I0416 13:09:45.510342 135676222270592 checkpoint_manager.py:921] [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_hns=False, 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), root_directory=gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b5e82338b30>
I0416 13:09:45.510420 135676222270592 checkpointing.py:301] Checkpoint manager created!
I0416 13:09:45.591743 135676222270592 dataset_info.py:707] Load dataset info from tests/assets/local_datasets/c4_en_dataset_minimal/c4/en/3.1.0
I0416 13:09:45.594745 135676222270592 reader.py:262] Creating a tf.data.Dataset reading 8 files located in folders: tests/assets/local_datasets/c4_en_dataset_minimal/c4/en/3.1.0.
I0416 13:09:45.647959 135676222270592 logging_logger.py:49] Constructing tf.data.Dataset __local_c4_builder for split train, from tests/assets/local_datasets/c4_en_dataset_minimal/c4/en/3.1.0
I0416 13:09:45.679090 135676222270592 tokenizer.py:245] Tokenizer path: src/maxtext/assets/tokenizers/tokenizer.llama2
I0416 13:09:45.679156 135676222270592 tokenizer.py:187] Loading sentencepiece tokenizer: src/maxtext/assets/tokenizers/tokenizer.llama2
I0416 13:09:46.652808 135676222270592 nnx_wrappers.py:437] Unknown Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0416 13:09:46.652916 135676222270592 nnx_wrappers.py:437] Unknown Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, None).
I0416 13:09:46.754303 135676222270592 attentions.py:1088] attentions/inputs_q Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0416 13:09:46.754381 135676222270592 attentions.py:1088] attentions/inputs_q Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, None).
I0416 13:09:46.769373 135676222270592 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0416 13:09:46.769423 135676222270592 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, None).
I0416 13:09:46.791411 135676222270592 attentions.py:1154] attentions/query Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0416 13:09:46.791466 135676222270592 attentions.py:1154] attentions/query Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None).
I0416 13:09:46.806432 135676222270592 attentions.py:1155] attentions/key Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0416 13:09:46.806482 135676222270592 attentions.py:1155] attentions/key Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None).
I0416 13:09:46.821439 135676222270592 attentions.py:1156] attentions/value Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0416 13:09:46.821489 135676222270592 attentions.py:1156] attentions/value Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None).
I0416 13:09:46.843579 135676222270592 attentions.py:1197] attentions/out Logical: bfloat16[8,2048,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0416 13:09:46.843635 135676222270592 attentions.py:1197] attentions/out Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None).
I0416 13:09:46.862661 135676222270592 linears.py:525] linears/x Logical: bfloat16[8,2048,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0416 13:09:46.862714 135676222270592 linears.py:525] linears/x Physical: bfloat16[8,2048,7168]....................................... ('fsdp', None, None).
I0416 13:09:47.049569 135676222270592 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0416 13:09:47.049659 135676222270592 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
[DECOUPLED NO-OP] gcs_storage: using stubs.
[DECOUPLED NO-OP] mldiagnostics: using stub.
[DECOUPLED NO-OP] mldiagnostics: using stub.
[DECOUPLED NO-OP] mldiagnostics: using stub.
[DECOUPLED NO-OP] workload_monitor: using stub.
[DECOUPLED NO-OP] vertex_tensorboard: using stub.
fsdp: 8
I0416 13:09:48.422908 135676222270592 maxtext_utils.py:1620] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0416 13:09:48.423014 135676222270592 maxtext_utils.py:1620] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0416 13:09:48.423078 135676222270592 maxtext_utils.py:1620] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0416 13:09:48.423128 135676222270592 maxtext_utils.py:1620] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0416 13:09:48.423169 135676222270592 maxtext_utils.py:1620] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0416 13:09:48.423198 135676222270592 maxtext_utils.py:1620] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0416 13:09:48.423238 135676222270592 maxtext_utils.py:1620] 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)
I0416 13:09:48.423277 135676222270592 maxtext_utils.py:1620] 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')
I0416 13:09:48.423304 135676222270592 maxtext_utils.py:1620] 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)
I0416 13:09:48.423330 135676222270592 maxtext_utils.py:1620] 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)
I0416 13:09:48.423365 135676222270592 maxtext_utils.py:1620] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0416 13:09:48.423398 135676222270592 maxtext_utils.py:1620] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0416 13:09:48.855905 135676222270592 train.py:155] train/xent Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length').
I0416 13:09:48.855985 135676222270592 train.py:155] train/xent Physical: float32[8,2048]............................................. ('fsdp', None).
I0416 13:09:48.870230 135676222270592 train.py:162] train/z_loss Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length').
I0416 13:09:48.870280 135676222270592 train.py:162] train/z_loss Physical: float32[8,2048]............................................. ('fsdp', None).
W0416 13:09:49.128278 3853394 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions.
I0416 13:09:49.240452 135676222270592 max_utils.py:791] Total memory size: 3.6 GB, Output size: 1.5 GB, Temp size: 2.0 GB, Argument size: 1.5 GB, Host temp size: 0.0 GB.
I0416 13:09:49.240932 135676222270592 max_utils.py:194] tensorboardX not available; using no-op SummaryWriter.
I0416 13:09:49.241088 135676222270592 metric_logger.py:289] number parameters: 1.104 billion
W0416 13:09:52.267439 3853394 pjrt_executable.cc:642] Assume version compatibility. PjRt-IFRT does not track XLA executable versions.
I0416 13:09:52.394989 135676222270592 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0416 13:09:52.517013 135676222270592 checkpointing.py:776] Waited 0.12201142311096191 seconds for step 0 to finish before starting checkpointing.
I0416 13:09:52.517447 135676222270592 checkpoint_manager.py:1983] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0416 13:09:52.517608 135676222270592 checkpoint_manager.py:1501] [process=0] Saving checkpoint at step 0
I0416 13:09:52.517954 135676222270592 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0.
I0416 13:09:52.622509 135676222270592 signaling_client.py:373] Using ThreadSafeKeyValueSignalingClient
I0416 13:09:52.711794 135561442494016 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0
I0416 13:09:52.729037 135676222270592 jax_array_handlers.py:347] Scheduling D2H of 39 prioritized jax.Array.
I0416 13:09:52.729137 135676222270592 replica_slices.py:410] 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
I0416 13:09:53.551824 135561432008256 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0/items
I0416 13:09:53.703468 135561379579456 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776344993442744872, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776344993442744872, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0/_CHECKPOINT_METADATA
I0416 13:09:53.832418 135676222270592 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 1.103781s
I0416 13:09:53.832937 135676222270592 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 10.204 GiB/s (total gbytes: 12.3 GiB) (time elapsed: a second) (per-host)
I0416 13:09:53.832999 135676222270592 base_pytree_checkpoint_handler.py:732] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 1.209545s (batch_requests_ready=0.101857s, total_serialization_initiated=1.107318s, others=0.000370s)
I0416 13:09:53.833081 135676222270592 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 1.210108s (all_items=0.000019s, per_item={'items': '0.00001931'}, temp_paths=1.210088)
I0416 13:09:53.833885 135561358607936 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0416 13:09:53.834007 135676222270592 async_checkpointer.py:561] Finished blocking save. Time taken: 1.316355s. Continuing background save to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0.
I0416 13:09:53.834269 135676222270592 checkpoint_manager.py:1549] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0416 13:09:53.834477 135561452979776 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0416 13:09:53.834575 135676222270592 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776344992.5174341, 'wait_for_prev_duration_secs': 4.0531158447265625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776344992.5176308, 'checkpointer_blocking_duration_secs': 1.3165335655212402, 'get_old_steps_start_time': 1776344993.8341937, 'get_old_steps_duration_secs': 2.9325485229492188e-05, 'checkpoint_manager_blocking_start_time': 1776344992.5173504, 'checkpoint_manager_blocking_duration_secs': 1.3171882629394531}
I0416 13:09:53.834738 135676222270592 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0416 13:09:53.834789 135676222270592 max_utils.py:750]
Memstats: After params initialized:
I0416 13:09:53.834880 135676222270592 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_0(process=0,(0,0,0,0))
I0416 13:09:53.834908 135676222270592 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_1(process=0,(1,0,0,0))
I0416 13:09:53.834928 135676222270592 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_2(process=0,(0,1,0,0))
I0416 13:09:53.834946 135676222270592 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_3(process=0,(1,1,0,0))
I0416 13:09:53.834964 135676222270592 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_4(process=0,(0,2,0,0))
I0416 13:09:53.834980 135676222270592 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_5(process=0,(1,2,0,0))
I0416 13:09:53.834996 135676222270592 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_6(process=0,(0,3,0,0))
I0416 13:09:53.835013 135676222270592 max_utils.py:756] Using (GB) 1.59 / 31.25 (5.088000%) on TPU_7(process=0,(1,3,0,0))
I0416 13:09:54.030006 3856408 google_auth_provider.cc:149] Using credentials at ~/.config/gcloud/application_default_credentials.json
I0416 13:09:54.030068 3856408 google_auth_provider.cc:156] Using OAuth2 AuthProvider
I0416 13:09:54.244637 135676222270592 metric_logger.py:185] completed step: 0, seconds: 3.153, TFLOP/s/device: 4.309, Tokens/s/device: 649.450, total_weights: 13328, loss: 10.828
I0416 13:09:54.245471 135676222270592 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/tensorboard/'
I0416 13:09:54.361016 135676222270592 metric_logger.py:185] completed step: 1, seconds: 1.848, TFLOP/s/device: 7.353, Tokens/s/device: 1108.359, total_weights: 12332, loss: 10.828
I0416 13:09:54.479108 135676222270592 metric_logger.py:185] completed step: 2, seconds: 0.009, TFLOP/s/device: 1485.419, Tokens/s/device: 223898.546, total_weights: 15161, loss: 9.823
I0416 13:09:54.596478 135676222270592 metric_logger.py:185] completed step: 3, seconds: 0.118, TFLOP/s/device: 115.007, Tokens/s/device: 17335.071, total_weights: 13327, loss: 9.405
I0416 13:09:55.011507 135561421522496 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0/items/array_metadatas/process_0
I0416 13:10:15.330835 135676222270592 metric_logger.py:185] completed step: 4, seconds: 0.118, TFLOP/s/device: 115.364, Tokens/s/device: 17388.942, total_weights: 11939, loss: 9.045
I0416 13:10:15.340193 135676222270592 metric_logger.py:185] completed step: 5, seconds: 0.117, TFLOP/s/device: 116.303, Tokens/s/device: 17530.494, total_weights: 15502, loss: 8.953
I0416 13:10:15.457128 135676222270592 metric_logger.py:185] completed step: 6, seconds: 20.734, TFLOP/s/device: 0.655, Tokens/s/device: 98.773, total_weights: 13864, loss: 8.775
I0416 13:10:15.574531 135676222270592 metric_logger.py:185] completed step: 7, seconds: 0.006, TFLOP/s/device: 2236.933, Tokens/s/device: 337174.844, total_weights: 12988, loss: 8.688
I0416 13:10:15.692163 135676222270592 metric_logger.py:185] completed step: 8, seconds: 0.118, TFLOP/s/device: 115.359, Tokens/s/device: 17388.204, total_weights: 13820, loss: 8.740
I0416 13:10:15.809691 135676222270592 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0416 13:10:15.810472 135676222270592 checkpointing.py:776] Waited 0.0008101463317871094 seconds for step 9 to finish before starting checkpointing.
I0416 13:10:15.810739 135676222270592 checkpoint_manager.py:1994] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0416 13:10:26.290278 135561369093696 base_pytree_checkpoint_handler.py:1217] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 32.262546s (commit=31.801486s, array_metadata_write=0.461060s)
I0416 13:10:26.291310 135561358607936 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/gbytes_per_sec: 375.337 MiB/s (total gbytes: 12.3 GiB) (time elapsed: 33 seconds) (per-host)
I0416 13:10:26.291365 135561358607936 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 32.457286s.
I0416 13:10:26.539389 135561358607936 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776344993442744872, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0/_CHECKPOINT_METADATA
I0416 13:10:26.730569 135561358607936 array_metadata_store.py:367] [process=0][thread=async_save] Skipped cross-host ArrayMetadata validation because only one process is found: process_index=0.
I0416 13:10:26.923391 135561379579456 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776344993442744872, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0/_CHECKPOINT_METADATA
I0416 13:10:27.182617 135561358607936 ocdbt_utils.py:56] Param validation support for Zarr3 will be added later (b/362328389).
I0416 13:10:27.182799 135561358607936 base_pytree_checkpoint_handler.py:1342] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 0.599321s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0/items
I0416 13:10:27.183461 135561358607936 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0/items
I0416 13:10:27.444991 135561358607936 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0
I0416 13:10:28.158375 135561358607936 atomicity.py:794] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0`.
I0416 13:10:28.158998 135561358607936 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 35.641356s. directory=gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/0
I0416 13:10:28.159087 135561358607936 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 34.325008s.
I0416 13:10:28.159209 135561452979776 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0416 13:10:28.159259 135561452979776 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0416 13:10:28.159311 135561452979776 checkpoint_manager.py:2103] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0416 13:10:28.159351 135561452979776 checkpoint_manager.py:2112] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0416 13:10:28.159466 135676222270592 checkpoint_manager.py:2006] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0416 13:10:28.159525 135676222270592 checkpoint_manager.py:1441] Waiting for previous save to complete took 12.348804 seconds. If this number is high, consider checkpointing less frequently.
I0416 13:10:28.160258 135676222270592 checkpoint_manager.py:1501] [process=0] Saving checkpoint at step 9
I0416 13:10:28.160491 135676222270592 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/9.
I0416 13:10:28.359014 135561452979776 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/9
I0416 13:10:28.363987 135676222270592 jax_array_handlers.py:347] Scheduling D2H of 39 prioritized jax.Array.
I0416 13:10:28.364095 135676222270592 replica_slices.py:410] 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
I0416 13:10:29.112252 135561369093696 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/9/items
I0416 13:10:37.964377 135676222270592 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 9.600792s
I0416 13:10:37.968466 135676222270592 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.272 GiB/s (total gbytes: 12.3 GiB) (time elapsed: 9 seconds) (per-host)
I0416 13:10:37.968531 135676222270592 base_pytree_checkpoint_handler.py:732] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 9.703175s (batch_requests_ready=0.095350s, total_serialization_initiated=9.603836s, others=0.003989s)
I0416 13:10:37.968606 135676222270592 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 9.703692s (all_items=0.000018s, per_item={'items': '0.00001764'}, temp_paths=9.703674)
I0416 13:10:37.970766 135561318762048 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0416 13:10:37.970846 135676222270592 async_checkpointer.py:561] Finished blocking save. Time taken: 9.810536s. Continuing background save to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/9.
I0416 13:10:37.971039 135676222270592 checkpoint_manager.py:1549] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0416 13:10:37.971283 135561308276288 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0416 13:10:37.971395 135676222270592 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776345015.810701, 'wait_for_prev_duration_secs': 12.348804473876953, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776345028.1602888, 'checkpointer_blocking_duration_secs': 9.81066370010376, 'get_old_steps_start_time': 1776345037.9709687, 'get_old_steps_duration_secs': 2.9802322387695312e-05, 'checkpoint_manager_blocking_start_time': 1776345015.8106637, 'checkpoint_manager_blocking_duration_secs': 22.160704851150513}
I0416 13:10:37.971521 135676222270592 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0416 13:10:37.971554 135676222270592 checkpoint_manager.py:1994] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0416 13:10:38.758295 135561421522496 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/9/items/array_metadatas/process_0
I0416 13:11:19.851295 135561369093696 base_pytree_checkpoint_handler.py:1217] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.882053s (commit=41.466585s, array_metadata_write=0.415468s)
I0416 13:11:19.852281 135561318762048 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/gbytes_per_sec: 244.961 MiB/s (total gbytes: 12.3 GiB) (time elapsed: 51 seconds) (per-host)
I0416 13:11:19.852334 135561318762048 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.881443s.
I0416 13:11:20.329868 135561318762048 array_metadata_store.py:367] [process=0][thread=async_save] Skipped cross-host ArrayMetadata validation because only one process is found: process_index=0.
I0416 13:11:20.750606 135561318762048 ocdbt_utils.py:56] Param validation support for Zarr3 will be added later (b/362328389).
I0416 13:11:20.750784 135561318762048 base_pytree_checkpoint_handler.py:1342] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 0.570109s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/9/items
I0416 13:11:20.751433 135561318762048 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/9/items
I0416 13:11:20.994559 135561318762048 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/9
I0416 13:11:21.703909 135561318762048 atomicity.py:794] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/9`.
I0416 13:11:21.704540 135561318762048 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 53.544237s. directory=gs://wanglance-maxtext/linen_ckpt_main_20260416_123638/linen_main_20260416_123638_10_shardy_false/checkpoints/9
I0416 13:11:21.704616 135561318762048 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 43.733726s.
I0416 13:11:21.704783 135561308276288 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0416 13:11:21.704886 135561308276288 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0416 13:11:21.704942 135561308276288 checkpoint_manager.py:2103] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0416 13:11:21.704982 135561308276288 checkpoint_manager.py:2112] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0416 13:11:21.705219 135676222270592 checkpoint_manager.py:2006] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0416 13:11:21.705354 135676222270592 checkpoint_manager.py:1983] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0416 13:11:21.705952 135676222270592 metric_logger.py:185] completed step: 9, seconds: 0.117, TFLOP/s/device: 115.774, Tokens/s/device: 17450.728, total_weights: 12300, loss: 8.710
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops