XPK Start: Mon Apr 20 21:01:14 UTC 2026 2026-04-20 21:01:18.241014: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered WARNING: All log messages before absl::InitializeLog() is called are written to STDERR E0000 00:00:1776718878.253369 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776718878.257067 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776718878.268564 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776718878.268592 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776718878.268594 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776718878.268596 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-20 21:01:37.472865: 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 21:01:37.995888 136993501005632 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-20 21:01:47,037:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0420 21:01:47.037906 136993501005632 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-20 21:01:47,040:jax._src.distributed:157: Connecting to JAX distributed service on mt-04-int8-qjomd-slice-job-0-0.mt-04-int8-qjomd:8482 I0420 21:01:47.040221 136993501005632 distributed.py:157] Connecting to JAX distributed service on mt-04-int8-qjomd-slice-job-0-0.mt-04-int8-qjomd:8482 I0420 21:01:48.026846 136993501005632 max_utils.py:284] Jax distributed system initialized! I0420 21:01:55.156393 136993501005632 max_utils.py:800] System Information: Jax Version: 0.8.1 I0420 21:01:55.156500 136993501005632 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0420 21:01:55.156541 136993501005632 max_utils.py:802] System Information: Jax Backend: PJRT C API TFRT TPU v6 lite Built on Nov 12 2025 14:16:36 (1762985796) cl/831091709 I0420 21:01:55.156579 136993501005632 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0420 21:01:55.766992 136993501005632 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0420 21:01:56.320697 136993501005632 checkpointing.py:688] Setting up checkpoint logger... I0420 21:01:56.320814 136993501005632 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0420 21:01:56.320864 136993501005632 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0420 21:01:56.321089 136993501005632 base_pytree_checkpoint_handler.py:415] 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 0x7c97c6974f50>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0420 21:01:59.186991 136993501005632 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0420 21:01:59.187355 136993501005632 checkpoint_manager.py:709] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c835850b770>}, handler_registry=None I0420 21:01:59.187592 136993501005632 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c835850b770>` 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 21:01:59.187640 136993501005632 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c8358475760>` 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 21:01:59.187676 136993501005632 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 0x7c835850b770>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c835850b770>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c8358475760>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c8358475760>}). I0420 21:01:59.188035 136993501005632 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0420 21:01:59.188122 136993501005632 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7c835820ccc0> timeout: 600 secs and primary_host=0 for async checkpoint writes I0420 21:02:00.315000 136993501005632 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints I0420 21:02:00.728539 136993501005632 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_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7c8358477260> I0420 21:02:00.728721 136993501005632 checkpointing.py:302] Checkpoint manager created! I0420 21:02:02.169028 136993501005632 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0420 21:02:02.169162 136993501005632 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0420 21:02:04.506390 136993501005632 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0420 21:02:04.506513 136993501005632 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0420 21:02:04.506568 136993501005632 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0420 21:02:04.506617 136993501005632 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0420 21:02:04.506656 136993501005632 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0420 21:02:04.506691 136993501005632 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0420 21:02:04.506730 136993501005632 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0420 21:02:04.506768 136993501005632 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0420 21:02:04.506804 136993501005632 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0420 21:02:04.506839 136993501005632 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0420 21:02:04.506873 136993501005632 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0420 21:02:04.506907 136993501005632 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0420 21:02:04.916302 136993501005632 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0420 21:02:04.916454 136993501005632 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 21:02:04.922024 136993501005632 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0420 21:02:04.922135 136993501005632 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 21:02:05.083270 136993501005632 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 21:02:05.083407 136993501005632 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 21:02:05.238824 136993501005632 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 21:02:05.238958 136993501005632 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 21:02:05.455071 136993501005632 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 21:02:05.455237 136993501005632 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 21:02:05.614418 136993501005632 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 21:02:05.614574 136993501005632 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 21:02:05.765987 136993501005632 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 21:02:05.766145 136993501005632 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 21:02:05.926834 136993501005632 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0420 21:02:05.926973 136993501005632 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 21:02:06.421328 136993501005632 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0420 21:02:06.421451 136993501005632 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0420 21:02:20.181508 136993501005632 max_utils.py:791] Total memory size: 1.8 GB, Output size: 0.4 GB, Temp size: 1.4 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB. I0420 21:02:20.190399 136993501005632 metric_logger.py:289] number parameters: 1.104 billion I0420 21:02:34.803157 136993501005632 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0420 21:02:35.037457 136993501005632 checkpointing.py:798] Waited 0.23428583145141602 seconds for step 0 to finish before starting checkpointing. I0420 21:02:35.040014 136993501005632 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 21:02:35.042064 136993501005632 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0 I0420 21:02:35.044911 136993501005632 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0. I0420 21:02:35.335251 136993501005632 signaling_client.py:364] Using JaxDistributedSignalingClient I0420 21:02:35.597314 136862656083712 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0 I0420 21:02:35.598430 136993501005632 jax_array_handlers.py:358] Scheduling D2H of 111 prioritized jax.Array. I0420 21:02:35.598494 136993501005632 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 21:02:35.844890 136993501005632 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.250075s I0420 21:02:35.845755 136993501005632 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.067 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5029323101043701 s) (per-host) I0420 21:02:35.845818 136993501005632 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.503012s (batch_requests_ready=0.239984s, total_serialization_initiated=0.262252s, others=0.000775s) I0420 21:02:35.845910 136993501005632 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.507066s (all_items=0.000019s, per_item={'items': '0.00001931'}, temp_paths=0.507047) I0420 21:02:35.846913 136864300267264 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0420 21:02:35.847074 136993501005632 async_checkpointer.py:561] Finished blocking save. Time taken: 0.804930s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0. I0420 21:02:35.879270 136993501005632 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0420 21:02:35.879601 136862647691008 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0420 21:02:35.879767 136993501005632 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776718955.0399961, 'wait_for_prev_duration_secs': 6.341934204101562e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776718955.0421162, 'checkpointer_blocking_duration_secs': 0.8050730228424072, 'get_old_steps_start_time': 1776718955.8472095, 'get_old_steps_duration_secs': 2.5987625122070312e-05, 'checkpoint_manager_blocking_start_time': 1776718955.0379214, 'checkpoint_manager_blocking_duration_secs': 0.8418056964874268} I0420 21:02:35.879891 136993501005632 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0420 21:02:35.879983 136993501005632 max_utils.py:750] Memstats: After params initialized: I0420 21:02:35.880036 136993501005632 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0420 21:02:35.880070 136993501005632 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0420 21:02:35.880114 136993501005632 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0420 21:02:35.880141 136993501005632 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0420 21:02:36.181348 136993501005632 metric_logger.py:185] completed step: 0, seconds: 14.280, TFLOP/s/device: 0.951, Tokens/s/device: 143.417, total_weights: 65536, loss: 10.872 I0420 21:02:36.184302 136993501005632 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/tensorboard/' I0420 21:02:36.634222 136993501005632 metric_logger.py:185] completed step: 1, seconds: 1.373, TFLOP/s/device: 9.899, Tokens/s/device: 1492.096, total_weights: 65536, loss: 10.872 I0420 21:02:36.854243 136993501005632 metric_logger.py:185] completed step: 2, seconds: 0.455, TFLOP/s/device: 29.877, Tokens/s/device: 4503.435, total_weights: 65536, loss: 10.258 I0420 21:02:37.083448 136993501005632 metric_logger.py:185] completed step: 3, seconds: 0.014, TFLOP/s/device: 941.458, Tokens/s/device: 141906.874, total_weights: 65536, loss: 9.727 I0420 21:02:37.406656 136865901438720 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/items I0420 21:02:59.127882 136862100780800 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776718957147974813, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776718957147974813, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/_CHECKPOINT_METADATA I0420 21:02:59.129434 136993501005632 metric_logger.py:185] completed step: 4, seconds: 0.221, TFLOP/s/device: 61.504, Tokens/s/device: 9270.534, total_weights: 65536, loss: 9.270 I0420 21:02:59.141135 136993501005632 metric_logger.py:185] completed step: 5, seconds: 0.229, TFLOP/s/device: 59.321, Tokens/s/device: 8941.552, total_weights: 65536, loss: 8.884 I0420 21:02:59.365653 136993501005632 metric_logger.py:185] completed step: 6, seconds: 22.045, TFLOP/s/device: 0.616, Tokens/s/device: 92.901, total_weights: 65536, loss: 8.584 I0420 21:02:59.594801 136993501005632 metric_logger.py:185] completed step: 7, seconds: 0.009, TFLOP/s/device: 1475.739, Tokens/s/device: 222439.448, total_weights: 65536, loss: 8.376 I0420 21:02:59.823614 136993501005632 metric_logger.py:185] completed step: 8, seconds: 0.226, TFLOP/s/device: 60.009, Tokens/s/device: 9045.177, total_weights: 65536, loss: 8.247 I0420 21:03:00.054677 136993501005632 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0420 21:03:00.055592 136993501005632 checkpointing.py:798] Waited 0.0009279251098632812 seconds for step 9 to finish before starting checkpointing. I0420 21:03:00.057503 136993501005632 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 21:03:00.272816 2789 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0420 21:03:02.024059 136864788190976 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 111 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/items/array_metadatas/process_0 I0420 21:03:34.500043 136864779798272 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 34.234573s (commit=32.927047s, array_metadata_write=1.307526s) I0420 21:03:34.501307 136864300267264 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 26.701 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 59.15845966339111 s) (per-host) I0420 21:03:34.501417 136864300267264 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 58.654389s. I0420 21:03:35.164890 136864300267264 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776718957147974813, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/_CHECKPOINT_METADATA I0420 21:03:35.976484 136862100780800 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776718957147974813, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/_CHECKPOINT_METADATA I0420 21:03:36.240118 136864300267264 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000731s. I0420 21:03:36.569459 136864300267264 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0420 21:03:37.520941 136864300267264 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.233176s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/items I0420 21:03:37.522846 136864300267264 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0/items I0420 21:03:37.936626 136864300267264 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0 I0420 21:03:39.735934 136864300267264 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0`. I0420 21:03:39.736768 136864300267264 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 64.694626s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/0 I0420 21:03:39.738272 136864300267264 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 63.891243s. I0420 21:03:39.738472 136862647691008 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 21:03:39.738617 136862647691008 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0420 21:03:39.738684 136862647691008 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0420 21:03:39.740614 136862647691008 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0420 21:03:39.740787 136993501005632 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0420 21:03:39.740922 136993501005632 checkpoint_manager.py:1458] Waiting for previous save to complete took 39.683422 seconds. If this number is high, consider checkpointing less frequently. I0420 21:03:39.742980 136993501005632 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9 I0420 21:03:39.746387 136993501005632 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9. I0420 21:03:40.317378 136864300267264 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9 I0420 21:03:40.322365 136993501005632 jax_array_handlers.py:358] Scheduling D2H of 111 prioritized jax.Array. I0420 21:03:40.322467 136993501005632 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 21:03:40.352373 136993501005632 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.033356s I0420 21:03:40.353232 136993501005632 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.095 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.3027822971343994 s) (per-host) I0420 21:03:40.353295 136993501005632 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.302863s (batch_requests_ready=0.258367s, total_serialization_initiated=0.043726s, others=0.000770s) I0420 21:03:40.353379 136993501005632 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.306828s (all_items=0.000016s, per_item={'items': '0.00001621'}, temp_paths=0.306812) I0420 21:03:40.354239 136865876260608 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0420 21:03:40.354401 136993501005632 async_checkpointer.py:561] Finished blocking save. Time taken: 0.611351s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9. I0420 21:03:40.801661 136993501005632 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0420 21:03:40.802026 136862647691008 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0420 21:03:40.802220 136993501005632 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776718980.0574706, 'wait_for_prev_duration_secs': 39.68342185020447, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776719019.7430198, 'checkpointer_blocking_duration_secs': 0.6114881038665771, 'get_old_steps_start_time': 1776719020.3545291, 'get_old_steps_duration_secs': 2.6702880859375e-05, 'checkpoint_manager_blocking_start_time': 1776718980.055875, 'checkpoint_manager_blocking_duration_secs': 40.74630665779114} I0420 21:03:40.802345 136993501005632 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0420 21:03:40.802396 136993501005632 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 21:03:42.100100 136860500207360 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9/items I0420 21:03:45.241350 136865893046016 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 111 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9/items/array_metadatas/process_0 I0420 21:04:25.151498 136865884653312 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.892719s (commit=40.680575s, array_metadata_write=1.212144s) I0420 21:04:25.153698 136865876260608 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 35.022 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.103214502334595 s) (per-host) I0420 21:04:25.153815 136865876260608 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 44.799458s. I0420 21:04:26.985906 136865876260608 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000443s. I0420 21:04:27.770401 136865876260608 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0420 21:04:28.737643 136865876260608 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.709735s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9/items I0420 21:04:28.740170 136865876260608 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9/items I0420 21:04:29.142514 136865876260608 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9 I0420 21:04:30.639455 136865876260608 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9`. I0420 21:04:30.640313 136865876260608 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 50.897267s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_04_int8/checkpoints/9 I0420 21:04:30.641779 136865876260608 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 50.287421s. I0420 21:04:30.641926 136862647691008 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 21:04:30.641988 136862647691008 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0420 21:04:30.642055 136862647691008 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0420 21:04:30.648676 136862647691008 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0420 21:04:30.648849 136993501005632 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0420 21:04:30.649033 136993501005632 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 21:04:30.649690 136993501005632 metric_logger.py:185] completed step: 9, seconds: 0.228, TFLOP/s/device: 59.550, Tokens/s/device: 8975.960, total_weights: 65536, loss: 8.171 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Mon Apr 20 21:04:44 UTC 2026 EXIT_CODE=0