MaxView

XPK Start: Mon Apr 20 18:09:08 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-20 18:09:32.401480: 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 18:09:32.579504 136079459305280 max_utils.py:273] Attempting to initialize the jax distributed system...
I0420 18:09:41.619747 136079459305280 distributed.py:149] Starting JAX distributed service on [::]:8482
I0420 18:09:41.622033 136079459305280 distributed.py:172] Connecting to JAX distributed service on mt-11-optimizer-offload-gcfj6-slice-job-0-0.mt-11-optimizer-offload-gcfj6:8482
I0420 18:09:42.603570 136079459305280 max_utils.py:284] Jax distributed system initialized!
I0420 18:09:47.828038 136079459305280 max_utils.py:800] System Information: Jax Version: 0.9.2
I0420 18:09:47.828138 136079459305280 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0420 18:09:47.828177 136079459305280 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365
I0420 18:09:47.828210 136079459305280 train_utils.py:378] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0420 18:09:48.544926 136079459305280 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0420 18:09:48.656328 136079459305280 checkpointing.py:688] Setting up checkpoint logger...
I0420 18:09:48.656442 136079459305280 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0420 18:09:48.656488 136079459305280 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0420 18:09:48.656692 136079459305280 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 0x7bc2d6d17650>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0420 18:09:51.623161 136079459305280 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0420 18:09:51.623393 136079459305280 checkpoint_manager.py:708] [process=1][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7baf085d7da0>}, handler_registry=None
I0420 18:09:51.623637 136079459305280 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7baf085d7da0>` 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 18:09:51.623701 136079459305280 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7baf085ebc80>` 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 18:09:51.623739 136079459305280 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 0x7baf085d7da0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7baf085d7da0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7baf085ebc80>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7baf085ebc80>}).
I0420 18:09:51.624045 136079459305280 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0420 18:09:51.624116 136079459305280 async_checkpointer.py:192] [process=1][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7baf2014cfe0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0420 18:09:52.981698 136079459305280 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_11_optimizer_offload_true/checkpoints
I0420 18:09:53.446450 136079459305280 checkpoint_manager.py:929] [process=1][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_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_11_optimizer_offload_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7baf401d1160>
I0420 18:09:53.446622 136079459305280 checkpointing.py:302] Checkpoint manager created!
I0420 18:09:54.404780 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to host
I0420 18:09:54.404895 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0420 18:09:54.404933 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.404966 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.404996 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405022 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0420 18:09:54.405050 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0420 18:09:54.405075 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405100 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405123 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405146 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405170 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405194 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to host
I0420 18:09:54.405218 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0420 18:09:54.405242 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405266 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405289 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405314 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0420 18:09:54.405337 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0420 18:09:54.405360 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405383 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405406 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405429 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405452 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0420 18:09:54.405475 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to host
I0420 18:09:54.405498 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to host
I0420 18:09:54.405521 136079459305280 maxtext_utils_nnx.py:115] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to host
I0420 18:09:54.406627 136079459305280 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0420 18:09:54.406697 136079459305280 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0420 18:09:55.861009 136079459305280 maxtext_utils.py:1836]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0420 18:09:55.861120 136079459305280 maxtext_utils.py:1836]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0420 18:09:55.861171 136079459305280 maxtext_utils.py:1836]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0420 18:09:55.861214 136079459305280 maxtext_utils.py:1836]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0420 18:09:55.861251 136079459305280 maxtext_utils.py:1836]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0420 18:09:55.861287 136079459305280 maxtext_utils.py:1836]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0420 18:09:55.861324 136079459305280 maxtext_utils.py:1836]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 18:09:55.861360 136079459305280 maxtext_utils.py:1836]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0420 18:09:55.861394 136079459305280 maxtext_utils.py:1836]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 18:09:55.861426 136079459305280 maxtext_utils.py:1836]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 18:09:55.861458 136079459305280 maxtext_utils.py:1836]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)
I0420 18:09:55.861489 136079459305280 maxtext_utils.py:1836]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')

I0420 18:09:56.266201 136079459305280 nnx_decoders.py:465] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 18:09:56.266294 136079459305280 nnx_decoders.py:465] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 18:09:56.271683 136079459305280 nnx_decoders.py:465] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 18:09:56.271736 136079459305280 nnx_decoders.py:465] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 18:09:56.287762 136079459305280 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 18:09:56.287818 136079459305280 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 18:09:56.303241 136079459305280 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 18:09:56.303299 136079459305280 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 18:09:56.327996 136079459305280 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 18:09:56.328063 136079459305280 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 18:09:56.343636 136079459305280 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 18:09:56.343704 136079459305280 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 18:09:56.359143 136079459305280 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 18:09:56.359204 136079459305280 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 18:09:56.393839 136079459305280 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0420 18:09:56.393947 136079459305280 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 18:09:56.413559 136079459305280 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0420 18:09:56.413620 136079459305280 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0420 18:09:56.575436 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to device
I0420 18:09:56.575557 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0420 18:09:56.575606 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.575644 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.575690 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.575721 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0420 18:09:56.575750 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0420 18:09:56.575777 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.575802 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.575828 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.575855 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.575879 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.575908 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0420 18:09:56.575935 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0420 18:09:56.575959 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.575983 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.576007 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.576031 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0420 18:09:56.576054 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0420 18:09:56.576078 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.576103 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.576129 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.576153 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.576176 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:09:56.576199 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0420 18:09:56.576223 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to device
I0420 18:09:56.576246 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to device
I0420 18:10:07.656347 136079459305280 max_utils.py:791] Total memory size: 1.3 GB, Output size: 0.1 GB, Temp size: 1.1 GB, Argument size: 0.1 GB, Host temp size: 0.0 GB.
I0420 18:10:07.658412 136079459305280 metric_logger.py:301] number parameters: 1.104 billion
I0420 18:10:08.194447 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to device
I0420 18:10:08.194560 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0420 18:10:08.194601 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.194642 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.194688 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.194720 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0420 18:10:08.194750 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0420 18:10:08.194776 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.194802 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.194830 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.194862 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.194889 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.194915 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0420 18:10:08.194944 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0420 18:10:08.194969 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.194993 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.195019 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.195044 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0420 18:10:08.195067 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0420 18:10:08.195091 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.195114 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.195137 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.195161 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.195188 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0420 18:10:08.195212 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0420 18:10:08.195235 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to device
I0420 18:10:08.195258 136079459305280 maxtext_utils_nnx.py:132] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to device
I0420 18:10:18.505835 136079459305280 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0420 18:10:18.673820 136079459305280 checkpointing.py:798] Waited 0.1679689884185791 seconds for step 0 to finish before starting checkpointing.
I0420 18:10:18.676271 136079459305280 checkpoint_manager.py:2009] [process=1][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 18:10:18.677833 136079459305280 checkpoint_manager.py:1512] [process=1] Saving checkpoint at step 0
I0420 18:10:18.679334 136079459305280 event_tracking.py:70] [process=1] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_11_optimizer_offload_true/checkpoints/0.
I0420 18:10:19.437903 136079459305280 signaling_client.py:364] Using JaxDistributedSignalingClient
I0420 18:10:19.438984 136079459305280 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0420 18:10:19.439040 136079459305280 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 18:10:19.705580 136079459305280 base_pytree_checkpoint_handler.py:154] [process=1][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.268447s
I0420 18:10:19.705767 136079459305280 base_pytree_checkpoint_handler.py:130] [process=1] /jax/orbax/write/blocking_gbytes_per_sec: 5.545 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.27817511558532715 s) (per-host)
I0420 18:10:19.705818 136079459305280 base_pytree_checkpoint_handler.py:768] [process=1][thread=MainThread] Initiated Pytree async_save. Time taken: 0.278235s (batch_requests_ready=0.002973s, total_serialization_initiated=0.275195s, others=0.000067s)
I0420 18:10:19.705914 136079459305280 composite_checkpoint_handler.py:715] [process=1][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.282383s (all_items=0.000017s, per_item={'items': '0.00001669'}, temp_paths=0.282367)
I0420 18:10:19.706682 136079459305280 event_tracking.py:125] [process=1] [async] Finished blocking save in 1.03 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_11_optimizer_offload_true/checkpoints/0.
I0420 18:10:19.707012 135947551221504 async_checkpointer.py:76] [process=1][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 18:30:19.706975
I0420 18:10:19.721232 136079459305280 checkpoint_manager.py:1560] [process=1][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 18:10:19.721538 135946787878656 async_checkpointer.py:280] [process=1][thread=save_finalize] Waiting for background save thread=async_save.
I0420 18:10:19.721705 136079459305280 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776708618.6762526, 'wait_for_prev_duration_secs': 6.008148193359375e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776708618.6778727, 'checkpointer_blocking_duration_secs': 1.029301404953003, 'get_old_steps_start_time': 1776708619.707202, 'get_old_steps_duration_secs': 2.86102294921875e-05, 'checkpoint_manager_blocking_start_time': 1776708618.6742182, 'checkpoint_manager_blocking_duration_secs': 1.0474483966827393}
I0420 18:10:19.721814 136079459305280 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0420 18:10:19.721884 136079459305280 max_utils.py:750] 
Memstats: After params initialized:
I0420 18:10:19.721934 136079459305280 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_2(process=1,(2,0,0,0))
I0420 18:10:19.721967 136079459305280 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_3(process=1,(3,0,0,0))
I0420 18:10:19.721992 136079459305280 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_6(process=1,(2,1,0,0))
I0420 18:10:19.722016 136079459305280 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_7(process=1,(3,1,0,0))
I0420 18:10:20.041825 136079459305280 metric_logger.py:196] completed step: 0, seconds: 10.846, TFLOP/s/device: 1.253, Tokens/s/device: 188.819, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52786.469
I0420 18:10:20.228632 136079459305280 metric_logger.py:196] completed step: 1, seconds: 1.534, TFLOP/s/device: 8.856, Tokens/s/device: 1334.903, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52786.469
I0420 18:10:20.684742 136079459305280 metric_logger.py:196] completed step: 2, seconds: 0.025, TFLOP/s/device: 541.277, Tokens/s/device: 81587.125, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28632.133
I0420 18:10:20.847434 136079459305280 metric_logger.py:196] completed step: 3, seconds: 0.446, TFLOP/s/device: 30.480, Tokens/s/device: 4594.267, total_weights: 65536, loss: 9.732, lm_loss: 9.732, perplexity: 16852.779
I0420 18:10:21.174245 136079459305280 metric_logger.py:196] completed step: 4, seconds: 0.180, TFLOP/s/device: 75.279, Tokens/s/device: 11346.889, total_weights: 65536, loss: 9.274, lm_loss: 9.274, perplexity: 10659.797
I0420 18:10:21.181935 136079459305280 metric_logger.py:196] completed step: 5, seconds: 0.162, TFLOP/s/device: 83.941, Tokens/s/device: 12652.519, total_weights: 65536, loss: 8.888, lm_loss: 8.888, perplexity: 7244.862
I0420 18:10:23.040746    2749 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0420 18:10:25.016459 135947056314112 array_metadata_store.py:203] [process=1][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_11_optimizer_offload_true/checkpoints/0/items/array_metadatas/process_1
I0420 18:10:43.905718 136079459305280 metric_logger.py:196] completed step: 6, seconds: 0.327, TFLOP/s/device: 41.490, Tokens/s/device: 6253.760, total_weights: 65536, loss: 8.589, lm_loss: 8.589, perplexity: 5371.227
I0420 18:10:44.068467 136079459305280 metric_logger.py:196] completed step: 7, seconds: 22.562, TFLOP/s/device: 0.602, Tokens/s/device: 90.772, total_weights: 65536, loss: 8.380, lm_loss: 8.380, perplexity: 4360.796
I0420 18:10:44.231403 136079459305280 metric_logger.py:196] completed step: 8, seconds: 0.169, TFLOP/s/device: 80.574, Tokens/s/device: 12145.005, total_weights: 65536, loss: 8.251, lm_loss: 8.251, perplexity: 3831.198
I0420 18:10:44.394977 136079459305280 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0420 18:10:44.395668 136079459305280 checkpointing.py:798] Waited 0.0006899833679199219 seconds for step 9 to finish before starting checkpointing.
I0420 18:10:44.397608 136079459305280 checkpoint_manager.py:2020] [process=1][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 18:10:57.841966 135947551221504 base_pytree_checkpoint_handler.py:130] [process=1] /jax/orbax/write/gbytes_per_sec: 41.118 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 38.414337158203125 s) (per-host)
I0420 18:10:57.842087 135947551221504 async_checkpointer.py:90] [process=1][thread=async_save] 3 Handler Commit operations completed. Time taken: 38.134962s.
I0420 18:11:06.062319 135947551221504 async_checkpointer.py:160] [process=1][thread=async_save] Background save thread done. Time taken: 46.355176s.
I0420 18:11:06.062597 135946787878656 async_checkpointer.py:288] [process=1][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 18:11:06.062746 135946787878656 async_checkpointer.py:298] [process=1][thread=save_finalize] No errors found in background save thread=async_save.
I0420 18:11:06.062800 135946787878656 checkpoint_manager.py:2137] [process=1][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0420 18:11:06.064541 135946787878656 checkpoint_manager.py:2146] [process=1][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0420 18:11:06.064716 136079459305280 checkpoint_manager.py:2032] [process=1][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0420 18:11:06.064856 136079459305280 checkpoint_manager.py:1452] Waiting for previous save to complete took 21.667242 seconds. If this number is high, consider checkpointing less frequently.
I0420 18:11:06.066613 136079459305280 checkpoint_manager.py:1512] [process=1] Saving checkpoint at step 9
I0420 18:11:06.068677 136079459305280 event_tracking.py:70] [process=1] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_11_optimizer_offload_true/checkpoints/9.
I0420 18:11:06.383599 136079459305280 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0420 18:11:06.383709 136079459305280 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 18:11:06.411560 136079459305280 base_pytree_checkpoint_handler.py:154] [process=1][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.029592s
I0420 18:11:06.411748 136079459305280 base_pytree_checkpoint_handler.py:130] [process=1] /jax/orbax/write/blocking_gbytes_per_sec: 42.106 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03663444519042969 s) (per-host)
I0420 18:11:06.411803 136079459305280 base_pytree_checkpoint_handler.py:768] [process=1][thread=MainThread] Initiated Pytree async_save. Time taken: 0.036699s (batch_requests_ready=0.002644s, total_serialization_initiated=0.033982s, others=0.000073s)
I0420 18:11:06.411899 136079459305280 composite_checkpoint_handler.py:715] [process=1][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.040982s (all_items=0.000014s, per_item={'items': '0.00001407'}, temp_paths=0.040968)
I0420 18:11:06.412666 136079459305280 event_tracking.py:125] [process=1] [async] Finished blocking save in 0.35 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_11_optimizer_offload_true/checkpoints/9.
I0420 18:11:06.412976 135946787878656 async_checkpointer.py:76] [process=1][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 18:31:06.412942
I0420 18:11:06.419481 136079459305280 checkpoint_manager.py:1560] [process=1][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 18:11:06.419795 135924990064384 async_checkpointer.py:280] [process=1][thread=save_finalize] Waiting for background save thread=async_save.
I0420 18:11:06.419957 136079459305280 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776708644.397578, 'wait_for_prev_duration_secs': 21.667242288589478, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776708666.0666528, 'checkpointer_blocking_duration_secs': 0.34647512435913086, 'get_old_steps_start_time': 1776708666.4131513, 'get_old_steps_duration_secs': 2.9325485229492188e-05, 'checkpoint_manager_blocking_start_time': 1776708644.395904, 'checkpoint_manager_blocking_duration_secs': 22.024019479751587}
I0420 18:11:06.420065 136079459305280 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0420 18:11:06.420109 136079459305280 checkpoint_manager.py:2020] [process=1][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 18:11:12.004790 135947056314112 array_metadata_store.py:203] [process=1][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_11_optimizer_offload_true/checkpoints/9/items/array_metadatas/process_1
I0420 18:11:48.992120 135946787878656 base_pytree_checkpoint_handler.py:130] [process=1] /jax/orbax/write/gbytes_per_sec: 37.063 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.61697006225586 s) (per-host)
I0420 18:11:48.992241 135946787878656 async_checkpointer.py:90] [process=1][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.579145s.
I0420 18:11:58.001932 135946787878656 async_checkpointer.py:160] [process=1][thread=async_save] Background save thread done. Time taken: 51.588820s.
I0420 18:11:58.002234 135924990064384 async_checkpointer.py:288] [process=1][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 18:11:58.002360 135924990064384 async_checkpointer.py:298] [process=1][thread=save_finalize] No errors found in background save thread=async_save.
I0420 18:11:58.002409 135924990064384 checkpoint_manager.py:2137] [process=1][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0420 18:11:58.004488 135924990064384 checkpoint_manager.py:2146] [process=1][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0420 18:11:58.004650 136079459305280 checkpoint_manager.py:2032] [process=1][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0420 18:11:58.004827 136079459305280 checkpoint_manager.py:2009] [process=1][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 18:11:58.005692 136079459305280 metric_logger.py:196] completed step: 9, seconds: 0.163, TFLOP/s/device: 83.271, Tokens/s/device: 12551.481, total_weights: 65536, loss: 8.175, lm_loss: 8.175, perplexity: 3550.804
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Mon Apr 20 18:12:09 UTC 2026
EXIT_CODE=0