MaxView

XPK Start: Wed Apr 22 21:39:43 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
`rope_parameters`'s factor field must be a float >= 1, got 40
`rope_parameters`'s beta_fast field must be a float, got 32
`rope_parameters`'s beta_slow field must be a float, got 1
DeepseekV32Config got `key=rope_scaling` in kwargs but hasn't set it as attribute. For RoPE standardization you need to set `self.rope_parameters` in model's config. 
2026-04-22 21:40:08.636460: 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)
I0422 21:40:08.848138 139448336062272 max_utils.py:273] Attempting to initialize the jax distributed system...
I0422 21:40:17.889979 139448336062272 distributed.py:149] Starting JAX distributed service on [::]:8482
I0422 21:40:17.892456 139448336062272 distributed.py:172] Connecting to JAX distributed service on mt-06-grad-accum-n9k6p-slice-job-0-0.mt-06-grad-accum-n9k6p:8482
I0422 21:40:18.875946 139448336062272 max_utils.py:284] Jax distributed system initialized!
I0422 21:40:25.068447 139448336062272 max_utils.py:800] System Information: Jax Version: 0.9.2
I0422 21:40:25.068551 139448336062272 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0422 21:40:25.068589 139448336062272 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365
I0422 21:40:25.068625 139448336062272 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0422 21:40:25.772820 139448336062272 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0422 21:40:25.773116 139448336062272 checkpointing.py:677] Setting up checkpoint logger...
I0422 21:40:25.773173 139448336062272 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0422 21:40:25.773218 139448336062272 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0422 21:40:25.773557 139448336062272 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7ed34eb1c140>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0422 21:40:28.673063 139448336062272 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0422 21:40:28.673328 139448336062272 checkpoint_manager.py:708] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ebf2c387ad0>}, handler_registry=None
I0422 21:40:28.673565 139448336062272 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ebf2c387ad0>` 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`.
I0422 21:40:28.673613 139448336062272 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ebf2c3894f0>` 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`.
I0422 21:40:28.673648 139448336062272 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 0x7ebf2c387ad0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ebf2c387ad0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ebf2c3894f0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ebf2c3894f0>}).
I0422 21:40:28.673964 139448336062272 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0422 21:40:28.674031 139448336062272 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7ebf2c12b2e0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0422 21:40:30.523628 139448336062272 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260422_212603/linen_xpk_test_pipeline_scan_nnx_20260422_212603_06_grad_accum/checkpoints
I0422 21:40:30.850876 139448336062272 checkpoint_manager.py:929] [process=6][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/linen_ckpt_xpk_test_pipeline_scan_nnx_20260422_212603/linen_xpk_test_pipeline_scan_nnx_20260422_212603_06_grad_accum/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7ebf2c385400>
I0422 21:40:30.851049 139448336062272 checkpointing.py:301] Checkpoint manager created!
I0422 21:40:32.581786 139448336062272 nnx_wrappers.py:453] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0422 21:40:32.581908 139448336062272 nnx_wrappers.py:453] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 21:40:32.975747 139448336062272 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 21:40:32.975839 139448336062272 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 21:40:32.992215 139448336062272 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 21:40:32.992273 139448336062272 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 21:40:33.015909 139448336062272 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 21:40:33.015980 139448336062272 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 21:40:33.032632 139448336062272 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 21:40:33.032695 139448336062272 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 21:40:33.049163 139448336062272 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 21:40:33.049224 139448336062272 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 21:40:33.074655 139448336062272 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0422 21:40:33.074727 139448336062272 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 21:40:33.095931 139448336062272 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0422 21:40:33.095997 139448336062272 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0422 21:40:33.316598 139448336062272 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0422 21:40:33.316711 139448336062272 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0422 21:40:34.743480 139448336062272 maxtext_utils.py:1668]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0422 21:40:34.743615 139448336062272 maxtext_utils.py:1668]  params/params/decoder/layers/mlp/wi_0/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0422 21:40:34.743669 139448336062272 maxtext_utils.py:1668]  params/params/decoder/layers/mlp/wi_1/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0422 21:40:34.743727 139448336062272 maxtext_utils.py:1668]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   P('mlp', 'layers', 'embed')
    Physical:  (None, None, 'fsdp')
I0422 21:40:34.743779 139448336062272 maxtext_utils.py:1668]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0422 21:40:34.743818 139448336062272 maxtext_utils.py:1668]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0422 21:40:34.743871 139448336062272 maxtext_utils.py:1668]  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)
I0422 21:40:34.743925 139448336062272 maxtext_utils.py:1668]  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')
I0422 21:40:34.743966 139448336062272 maxtext_utils.py:1668]  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)
I0422 21:40:34.744004 139448336062272 maxtext_utils.py:1668]  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)
I0422 21:40:34.744053 139448336062272 maxtext_utils.py:1668]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0422 21:40:34.744115 139448336062272 maxtext_utils.py:1668]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')

I0422 21:40:34.770821 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048]............................................... Unknown.
I0422 21:40:34.770914 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048]............................................... (None,).
I0422 21:40:34.785834 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048,16,7168]....................................... Unknown.
I0422 21:40:34.785892 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048,16,7168]....................................... ('fsdp', None, None).
I0422 21:40:34.820302 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[7168,16,2048]....................................... Unknown.
I0422 21:40:34.820376 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[7168,16,2048]....................................... (None, None, 'fsdp').
I0422 21:40:34.835370 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048,16]............................................ Unknown.
I0422 21:40:34.835428 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048,16]............................................ (None, None).
I0422 21:40:34.865053 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048,16,16,128]..................................... Unknown.
I0422 21:40:34.865129 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048,16,16,128]..................................... ('fsdp', None, None, None).
I0422 21:40:34.880185 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[16,16,128,2048]..................................... Unknown.
I0422 21:40:34.880242 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[16,16,128,2048]..................................... (None, None, None, 'fsdp').
I0422 21:40:34.924326 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048,32000]......................................... Unknown.
I0422 21:40:34.924399 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048,32000]......................................... ('fsdp', None).
I0422 21:40:34.939352 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[32000,2048]......................................... Unknown.
I0422 21:40:34.939409 139448336062272 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[32000,2048]......................................... (None, 'fsdp').
I0422 21:40:35.611174 139448336062272 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0422 21:40:35.611271 139448336062272 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0422 21:40:35.626860 139448336062272 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0422 21:40:35.626921 139448336062272 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0422 21:40:46.899900 139448336062272 max_utils.py:791] Total memory size: 1.7 GB, Output size: 0.4 GB, Temp size: 1.3 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0422 21:40:46.900782 139448336062272 metric_logger.py:301] number parameters: 1.104 billion
I0422 21:40:48.769255 139448336062272 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0422 21:41:00.072702 139448336062272 checkpointing.py:776] Waited 11.303431987762451 seconds for step 0 to finish before starting checkpointing.
I0422 21:41:00.075047 139448336062272 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 21:41:00.077039 139448336062272 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0
I0422 21:41:00.078528 139448336062272 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260422_212603/linen_xpk_test_pipeline_scan_nnx_20260422_212603_06_grad_accum/checkpoints/0.
I0422 21:41:00.800614 139448336062272 signaling_client.py:364] Using JaxDistributedSignalingClient
I0422 21:41:00.801488 139448336062272 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0422 21:41:00.801546 139448336062272 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
I0422 21:41:01.071124 139448336062272 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.270571s
I0422 21:41:01.071300 139448336062272 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.587 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2760813236236572 s) (per-host)
I0422 21:41:01.071354 139448336062272 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.276146s (batch_requests_ready=0.002304s, total_serialization_initiated=0.273769s, others=0.000073s)
I0422 21:41:01.071448 139448336062272 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.280288s (all_items=0.000034s, per_item={'items': '0.00003362'}, temp_paths=0.280254)
I0422 21:41:01.072229 139448336062272 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.00 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260422_212603/linen_xpk_test_pipeline_scan_nnx_20260422_212603_06_grad_accum/checkpoints/0.
I0422 21:41:01.072596 139320775214848 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-22 22:01:01.072558
I0422 21:41:01.085941 139448336062272 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 21:41:01.086189 139319152338688 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0422 21:41:01.086322 139448336062272 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260422_212603/linen_xpk_test_pipeline_scan_nnx_20260422_212603_06_grad_accum/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776894060.0750277, 'wait_for_prev_duration_secs': 6.0558319091796875e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776894060.0770772, 'checkpointer_blocking_duration_secs': 0.9956274032592773, 'get_old_steps_start_time': 1776894061.0727324, 'get_old_steps_duration_secs': 2.9802322387695312e-05, 'checkpoint_manager_blocking_start_time': 1776894060.073236, 'checkpoint_manager_blocking_duration_secs': 1.0130500793457031}
I0422 21:41:01.086429 139448336062272 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0422 21:41:01.086478 139448336062272 max_utils.py:750] 
Memstats: After params initialized:
I0422 21:41:01.086525 139448336062272 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0))
I0422 21:41:01.086558 139448336062272 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0))
I0422 21:41:01.086587 139448336062272 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0))
I0422 21:41:01.086612 139448336062272 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0))
I0422 21:41:01.400391 139448336062272 metric_logger.py:196] completed step: 0, seconds: 1.868, TFLOP/s/device: 29.089, Tokens/s/device: 4384.617, total_weights: 262144, loss: 10.877, lm_loss: 10.877, perplexity: 52959.059
I0422 21:41:02.007756 139448336062272 metric_logger.py:196] completed step: 1, seconds: 12.630, TFLOP/s/device: 4.303, Tokens/s/device: 648.639, total_weights: 262144, loss: 10.877, lm_loss: 10.877, perplexity: 52959.059
I0422 21:41:02.586243 139448336062272 metric_logger.py:196] completed step: 2, seconds: 0.029, TFLOP/s/device: 1849.910, Tokens/s/device: 278838.626, total_weights: 262144, loss: 10.563, lm_loss: 10.563, perplexity: 38662.707
I0422 21:41:03.164280 139448336062272 metric_logger.py:196] completed step: 3, seconds: 0.584, TFLOP/s/device: 93.030, Tokens/s/device: 14022.475, total_weights: 262144, loss: 10.272, lm_loss: 10.272, perplexity: 28909.668
I0422 21:41:04.083985    2610 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0422 21:41:04.320825 139448336062272 metric_logger.py:196] completed step: 4, seconds: 0.578, TFLOP/s/device: 93.977, Tokens/s/device: 14165.291, total_weights: 262144, loss: 10.022, lm_loss: 10.022, perplexity: 22524.992
I0422 21:41:04.328322 139448336062272 metric_logger.py:196] completed step: 5, seconds: 0.578, TFLOP/s/device: 94.004, Tokens/s/device: 14169.309, total_weights: 262144, loss: 9.820, lm_loss: 9.820, perplexity: 18401.865
I0422 21:41:06.498032 139319160731392 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260422_212603/linen_xpk_test_pipeline_scan_nnx_20260422_212603_06_grad_accum/checkpoints/0/items/array_metadatas/process_6
I0422 21:41:38.764430 139320775214848 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 41.602 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.96917462348938 s) (per-host)
I0422 21:41:38.764568 139320775214848 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.691331s.
I0422 21:41:45.702156 139320775214848 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 44.628902s.
I0422 21:41:45.702461 139319152338688 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 21:41:45.702582 139319152338688 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0422 21:41:45.702633 139319152338688 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0422 21:41:45.704944 139319152338688 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0422 21:41:50.971765 139448336062272 metric_logger.py:196] completed step: 6, seconds: 1.158, TFLOP/s/device: 46.926, Tokens/s/device: 7073.234, total_weights: 262144, loss: 9.667, lm_loss: 9.667, perplexity: 15787.604
I0422 21:41:51.549712 139448336062272 metric_logger.py:196] completed step: 7, seconds: 46.066, TFLOP/s/device: 1.180, Tokens/s/device: 177.832, total_weights: 262144, loss: 9.561, lm_loss: 9.561, perplexity: 14203.827
I0422 21:41:52.127915 139448336062272 metric_logger.py:196] completed step: 8, seconds: 0.583, TFLOP/s/device: 93.153, Tokens/s/device: 14041.078, total_weights: 262144, loss: 9.496, lm_loss: 9.496, perplexity: 13302.920
I0422 21:41:52.704906 139448336062272 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0422 21:41:52.705589 139448336062272 checkpointing.py:776] Waited 0.0007045269012451172 seconds for step 9 to finish before starting checkpointing.
I0422 21:41:52.707720 139448336062272 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 21:41:52.709536 139448336062272 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9
I0422 21:41:52.710954 139448336062272 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260422_212603/linen_xpk_test_pipeline_scan_nnx_20260422_212603_06_grad_accum/checkpoints/9.
I0422 21:41:53.442853 139448336062272 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0422 21:41:53.442948 139448336062272 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
I0422 21:41:53.478981 139448336062272 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.037022s
I0422 21:41:53.479174 139448336062272 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 37.426 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.041216135025024414 s) (per-host)
I0422 21:41:53.479224 139448336062272 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.041277s (batch_requests_ready=0.001755s, total_serialization_initiated=0.039453s, others=0.000068s)
I0422 21:41:53.479315 139448336062272 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.045444s (all_items=0.000015s, per_item={'items': '0.00001454'}, temp_paths=0.045430)
I0422 21:41:53.479969 139448336062272 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.77 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260422_212603/linen_xpk_test_pipeline_scan_nnx_20260422_212603_06_grad_accum/checkpoints/9.
I0422 21:41:53.480313 139320775214848 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-22 22:01:53.480272
I0422 21:41:53.482619 139448336062272 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 21:41:53.482842 139324569937664 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0422 21:41:53.482968 139448336062272 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260422_212603/linen_xpk_test_pipeline_scan_nnx_20260422_212603_06_grad_accum/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776894112.7076855, 'wait_for_prev_duration_secs': 7.486343383789062e-05, 'time_between_consecutive_saves_sec': 7.002701759338379, 'checkpointer_blocking_start_time': 1776894112.7095761, 'checkpointer_blocking_duration_secs': 0.7708845138549805, 'get_old_steps_start_time': 1776894113.4804869, 'get_old_steps_duration_secs': 2.9325485229492188e-05, 'checkpoint_manager_blocking_start_time': 1776894112.7058136, 'checkpoint_manager_blocking_duration_secs': 0.7771186828613281}
I0422 21:41:53.483077 139448336062272 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0422 21:41:53.483142 139448336062272 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 21:41:59.968355 139314352604928 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260422_212603/linen_xpk_test_pipeline_scan_nnx_20260422_212603_06_grad_accum/checkpoints/9/items/array_metadatas/process_6
I0422 21:42:35.795402 139320775214848 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 37.292 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.35740518569946 s) (per-host)
I0422 21:42:35.795528 139320775214848 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.315102s.
I0422 21:42:41.532993 139320775214848 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 48.052552s.
I0422 21:42:41.533302 139324569937664 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 21:42:41.533435 139324569937664 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0422 21:42:41.533483 139324569937664 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0422 21:42:41.535538 139324569937664 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0422 21:42:41.535718 139448336062272 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0422 21:42:41.535867 139448336062272 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 21:42:41.536868 139448336062272 metric_logger.py:196] completed step: 9, seconds: 0.578, TFLOP/s/device: 94.051, Tokens/s/device: 14176.444, total_weights: 262144, loss: 9.457, lm_loss: 9.457, perplexity: 12802.546
Per train step:
 Total TFLOPs: 54.35 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Wed Apr 22 21:42:53 UTC 2026
EXIT_CODE=0