XPK Start: Thu Apr 23 18:05:27 UTC 2026 2026-04-23 18:05:31.590909: 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 00:00:1776967531.603909 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776967531.607617 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776967531.619148 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776967531.619168 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776967531.619171 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776967531.619173 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-23 18:05:50.756877: 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) I0423 18:05:51.283119 138309289326400 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-23 18:06:00,322:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0423 18:06:00.322698 138309289326400 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-23 18:06:00,325:jax._src.distributed:157: Connecting to JAX distributed service on mt-10-shardy-false-7gheh-slice-job-0-0.mt-10-shardy-false-7gheh:8482 I0423 18:06:00.325017 138309289326400 distributed.py:157] Connecting to JAX distributed service on mt-10-shardy-false-7gheh-slice-job-0-0.mt-10-shardy-false-7gheh:8482 I0423 18:06:01.488793 138309289326400 max_utils.py:284] Jax distributed system initialized! I0423 18:06:08.330958 138309289326400 max_utils.py:800] System Information: Jax Version: 0.8.1 I0423 18:06:08.331063 138309289326400 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0423 18:06:08.331103 138309289326400 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 I0423 18:06:08.331137 138309289326400 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0423 18:06:08.952687 138309289326400 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0423 18:06:09.066960 138309289326400 checkpointing.py:688] Setting up checkpoint logger... I0423 18:06:09.067079 138309289326400 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0423 18:06:09.067125 138309289326400 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0423 18:06:09.067334 138309289326400 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 0x7dca2cb75010>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0423 18:06:11.996079 138309289326400 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0423 18:06:11.996424 138309289326400 checkpoint_manager.py:709] [process=3][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7db7106b02f0>}, handler_registry=None I0423 18:06:11.996672 138309289326400 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7db7106b02f0>` 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`. I0423 18:06:11.996722 138309289326400 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7db5b42f0170>` 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`. I0423 18:06:11.996757 138309289326400 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 0x7db7106b02f0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7db7106b02f0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7db5b42f0170>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7db5b42f0170>}). I0423 18:06:11.997099 138309289326400 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0423 18:06:11.997179 138309289326400 async_checkpointer.py:177] [process=3][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7db6d3eb0f40> timeout: 600 secs and primary_host=0 for async checkpoint writes I0423 18:06:13.150840 138309289326400 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_10_shardy_false/checkpoints I0423 18:06:13.153033 138309289326400 checkpoint_manager.py:929] [process=3][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_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_10_shardy_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7db5b42f07d0> I0423 18:06:13.153143 138309289326400 checkpointing.py:302] Checkpoint manager created! I0423 18:06:14.099392 138309289326400 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0423 18:06:14.099510 138309289326400 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0423 18:06:15.460529 138309289326400 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0423 18:06:15.460677 138309289326400 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0423 18:06:15.460737 138309289326400 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0423 18:06:15.460782 138309289326400 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0423 18:06:15.460824 138309289326400 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0423 18:06:15.460882 138309289326400 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0423 18:06:15.461007 138309289326400 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0423 18:06:15.461058 138309289326400 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0423 18:06:15.461098 138309289326400 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0423 18:06:15.461135 138309289326400 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0423 18:06:15.461169 138309289326400 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0423 18:06:15.461202 138309289326400 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0423 18:06:15.863328 138309289326400 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 18:06:15.863471 138309289326400 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 18:06:15.868816 138309289326400 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 18:06:15.868909 138309289326400 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 18:06:16.014578 138309289326400 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 18:06:16.014713 138309289326400 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 18:06:16.156953 138309289326400 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 18:06:16.157085 138309289326400 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 18:06:16.313524 138309289326400 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 18:06:16.313660 138309289326400 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 18:06:16.461292 138309289326400 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 18:06:16.461432 138309289326400 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 18:06:16.605442 138309289326400 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 18:06:16.605580 138309289326400 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 18:06:16.760580 138309289326400 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0423 18:06:16.760720 138309289326400 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 18:06:16.783066 138309289326400 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0423 18:06:16.783160 138309289326400 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0423 18:06:27.864256 138309289326400 max_utils.py:791] Total memory size: 1.5 GB, Output size: 0.4 GB, Temp size: 1.1 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB. I0423 18:06:27.866695 138309289326400 metric_logger.py:289] number parameters: 1.104 billion I0423 18:06:38.890136 138309289326400 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0423 18:06:39.042608 138309289326400 checkpointing.py:798] Waited 0.15245771408081055 seconds for step 0 to finish before starting checkpointing. I0423 18:06:39.045085 138309289326400 checkpoint_manager.py:2013] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 18:06:39.046725 138309289326400 checkpoint_manager.py:1518] [process=3] Saving checkpoint at step 0 I0423 18:06:39.049456 138309289326400 async_checkpointer.py:452] [process=3] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_10_shardy_false/checkpoints/0. I0423 18:06:40.120486 138309289326400 signaling_client.py:364] Using JaxDistributedSignalingClient I0423 18:06:40.121411 138309289326400 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0423 18:06:40.121471 138309289326400 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 I0423 18:06:40.363473 138309289326400 base_pytree_checkpoint_handler.py:153] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.244121s I0423 18:06:40.363651 138309289326400 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.747 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5615110397338867 s) (per-host) I0423 18:06:40.363716 138309289326400 base_pytree_checkpoint_handler.py:737] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.561586s (batch_requests_ready=0.310097s, total_serialization_initiated=0.251405s, others=0.000084s) I0423 18:06:40.363825 138309289326400 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.565585s (all_items=0.000020s, per_item={'items': '0.00001955'}, temp_paths=0.565566) I0423 18:06:40.364732 138186537821952 async_checkpointer.py:79] [process=3][thread=async_save] Background save thread started. I0423 18:06:40.364908 138309289326400 async_checkpointer.py:561] Finished blocking save. Time taken: 1.318114s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_10_shardy_false/checkpoints/0. I0423 18:06:40.765234 138309289326400 checkpoint_manager.py:1566] [process=3][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0423 18:06:40.765622 138181152327424 async_checkpointer.py:265] [process=3][thread=save_finalize] Waiting for background save thread=async_save. I0423 18:06:40.765792 138309289326400 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776967599.0450678, 'wait_for_prev_duration_secs': 6.29425048828125e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776967599.0467622, 'checkpointer_blocking_duration_secs': 1.3182706832885742, 'get_old_steps_start_time': 1776967600.365053, 'get_old_steps_duration_secs': 2.5033950805664062e-05, 'checkpoint_manager_blocking_start_time': 1776967599.0430872, 'checkpoint_manager_blocking_duration_secs': 1.7226622104644775} I0423 18:06:40.765904 138309289326400 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0423 18:06:40.765990 138309289326400 max_utils.py:750] Memstats: After params initialized: I0423 18:06:40.766049 138309289326400 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_10(process=3,(2,2,0,0)) I0423 18:06:40.766083 138309289326400 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_11(process=3,(3,2,0,0)) I0423 18:06:40.766110 138309289326400 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_14(process=3,(2,3,0,0)) I0423 18:06:40.766137 138309289326400 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_15(process=3,(3,3,0,0)) I0423 18:06:41.026542 138309289326400 metric_logger.py:185] completed step: 0, seconds: 11.022, TFLOP/s/device: 1.233, Tokens/s/device: 185.803, total_weights: 65536, loss: 10.874 I0423 18:06:41.232086 138309289326400 metric_logger.py:185] completed step: 1, seconds: 2.135, TFLOP/s/device: 6.365, Tokens/s/device: 959.398, total_weights: 65536, loss: 10.874 I0423 18:06:41.643375 138309289326400 metric_logger.py:185] completed step: 2, seconds: 0.060, TFLOP/s/device: 227.172, Tokens/s/device: 34241.766, total_weights: 65536, loss: 10.262 I0423 18:06:41.790665 138309289326400 metric_logger.py:185] completed step: 3, seconds: 0.402, TFLOP/s/device: 33.809, Tokens/s/device: 5096.125, total_weights: 65536, loss: 9.732 I0423 18:06:42.085972 138309289326400 metric_logger.py:185] completed step: 4, seconds: 0.163, TFLOP/s/device: 83.283, Tokens/s/device: 12553.250, total_weights: 65536, loss: 9.274 I0423 18:06:42.093269 138309289326400 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.533, Tokens/s/device: 13947.533, total_weights: 65536, loss: 8.888 I0423 18:07:01.593323 138309289326400 metric_logger.py:185] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.930, Tokens/s/device: 6923.082, total_weights: 65536, loss: 8.589 I0423 18:07:01.740699 138309289326400 metric_logger.py:185] completed step: 7, seconds: 19.354, TFLOP/s/device: 0.702, Tokens/s/device: 105.818, total_weights: 65536, loss: 8.380 I0423 18:07:01.888443 138309289326400 metric_logger.py:185] completed step: 8, seconds: 0.154, TFLOP/s/device: 88.181, Tokens/s/device: 13291.624, total_weights: 65536, loss: 8.251 I0423 18:07:02.036896 138309289326400 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0423 18:07:02.037504 138309289326400 checkpointing.py:798] Waited 0.0006189346313476562 seconds for step 9 to finish before starting checkpointing. I0423 18:07:02.039414 138309289326400 checkpoint_manager.py:2024] [process=3][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 18:07:02.761954 2645 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0423 18:07:04.738787 138186546214656 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_10_shardy_false/checkpoints/0/items/array_metadatas/process_3 I0423 18:07:36.788278 138186537821952 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/gbytes_per_sec: 27.719 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 56.98609900474548 s) (per-host) I0423 18:07:36.788401 138186537821952 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 56.423545s. I0423 18:07:44.375200 138186537821952 async_checkpointer.py:144] [process=3][thread=async_save] Background save thread done. Time taken: 64.010329s. I0423 18:07:44.375508 138181152327424 async_checkpointer.py:273] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 18:07:44.375630 138181152327424 async_checkpointer.py:283] [process=3][thread=save_finalize] No errors found in background save thread=async_save. I0423 18:07:44.375683 138181152327424 checkpoint_manager.py:2133] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0423 18:07:44.377425 138181152327424 checkpoint_manager.py:2142] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0423 18:07:44.377608 138309289326400 checkpoint_manager.py:2036] [process=3][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0423 18:07:44.377737 138309289326400 checkpoint_manager.py:1458] Waiting for previous save to complete took 42.338326 seconds. If this number is high, consider checkpointing less frequently. I0423 18:07:44.379511 138309289326400 checkpoint_manager.py:1518] [process=3] Saving checkpoint at step 9 I0423 18:07:44.382843 138309289326400 async_checkpointer.py:452] [process=3] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_10_shardy_false/checkpoints/9. I0423 18:07:44.923433 138309289326400 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0423 18:07:44.923584 138309289326400 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 I0423 18:07:44.969435 138309289326400 base_pytree_checkpoint_handler.py:153] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.047980s I0423 18:07:44.969602 138309289326400 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.036 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.30631017684936523 s) (per-host) I0423 18:07:44.969656 138309289326400 base_pytree_checkpoint_handler.py:737] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.306372s (batch_requests_ready=0.253694s, total_serialization_initiated=0.052609s, others=0.000069s) I0423 18:07:44.969754 138309289326400 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.310208s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=0.310193) I0423 18:07:44.970634 138182226069248 async_checkpointer.py:79] [process=3][thread=async_save] Background save thread started. I0423 18:07:44.970801 138309289326400 async_checkpointer.py:561] Finished blocking save. Time taken: 0.591223s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_10_shardy_false/checkpoints/9. I0423 18:07:45.421296 138309289326400 checkpoint_manager.py:1566] [process=3][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0423 18:07:45.421698 138181152327424 async_checkpointer.py:265] [process=3][thread=save_finalize] Waiting for background save thread=async_save. I0423 18:07:45.421878 138309289326400 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776967622.0393865, 'wait_for_prev_duration_secs': 42.33832621574402, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776967664.3795488, 'checkpointer_blocking_duration_secs': 0.5913581848144531, 'get_old_steps_start_time': 1776967664.9709408, 'get_old_steps_duration_secs': 2.8371810913085938e-05, 'checkpoint_manager_blocking_start_time': 1776967622.037738, 'checkpoint_manager_blocking_duration_secs': 43.38410139083862} I0423 18:07:45.422004 138309289326400 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0423 18:07:45.422049 138309289326400 checkpoint_manager.py:2024] [process=3][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 18:07:50.451945 138186546214656 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_10_shardy_false/checkpoints/9/items/array_metadatas/process_3 I0423 18:08:27.500411 138182226069248 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/gbytes_per_sec: 36.875 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.83708119392395 s) (per-host) I0423 18:08:27.500522 138182226069248 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.529772s. I0423 18:08:35.740137 138182226069248 async_checkpointer.py:144] [process=3][thread=async_save] Background save thread done. Time taken: 50.769370s. I0423 18:08:35.740391 138181152327424 async_checkpointer.py:273] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 18:08:35.740457 138181152327424 async_checkpointer.py:283] [process=3][thread=save_finalize] No errors found in background save thread=async_save. I0423 18:08:35.740517 138181152327424 checkpoint_manager.py:2133] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0423 18:08:35.742131 138181152327424 checkpoint_manager.py:2142] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0423 18:08:35.742303 138309289326400 checkpoint_manager.py:2036] [process=3][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0423 18:08:35.742489 138309289326400 checkpoint_manager.py:2013] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 18:08:35.743162 138309289326400 metric_logger.py:185] completed step: 9, seconds: 0.146, TFLOP/s/device: 93.075, Tokens/s/device: 14029.319, total_weights: 65536, loss: 8.175 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Thu Apr 23 18:08:48 UTC 2026 EXIT_CODE=0