LICENSE
MANIFEST.in
README.md
THIRD_PARTY_NOTICES.md
pyproject.toml
src/ai2_olmo_core.egg-info/PKG-INFO
src/ai2_olmo_core.egg-info/SOURCES.txt
src/ai2_olmo_core.egg-info/dependency_links.txt
src/ai2_olmo_core.egg-info/requires.txt
src/ai2_olmo_core.egg-info/top_level.txt
src/olmo_core/__init__.py
src/olmo_core/_nvtx.py
src/olmo_core/aliases.py
src/olmo_core/config.py
src/olmo_core/doc_utils.py
src/olmo_core/exceptions.py
src/olmo_core/fs_cache.py
src/olmo_core/io.py
src/olmo_core/mxfp8_config.py
src/olmo_core/py.typed
src/olmo_core/script_utils.py
src/olmo_core/utils.py
src/olmo_core/version.py
src/olmo_core/data/__init__.py
src/olmo_core/data/collator.py
src/olmo_core/data/data_loader.py
src/olmo_core/data/numpy_dataset.py
src/olmo_core/data/source_mixture.py
src/olmo_core/data/tokenizer.py
src/olmo_core/data/types.py
src/olmo_core/data/utils.py
src/olmo_core/data/composable/__init__.py
src/olmo_core/data/composable/concat_and_chunk_instance_source.py
src/olmo_core/data/composable/data_loader.py
src/olmo_core/data/composable/instance_source.py
src/olmo_core/data/composable/mixing_document_source.py
src/olmo_core/data/composable/mixing_instance_source.py
src/olmo_core/data/composable/mixing_token_source.py
src/olmo_core/data/composable/numpy_document_source.py
src/olmo_core/data/composable/packing_instance_source.py
src/olmo_core/data/composable/random_instance_source.py
src/olmo_core/data/composable/sampling_document_source.py
src/olmo_core/data/composable/sampling_instance_source.py
src/olmo_core/data/composable/sampling_token_source.py
src/olmo_core/data/composable/sliced_instance_source.py
src/olmo_core/data/composable/sliced_token_source.py
src/olmo_core/data/composable/source_abc.py
src/olmo_core/data/composable/token_source.py
src/olmo_core/data/composable/utils.py
src/olmo_core/data/composable/visualize.py
src/olmo_core/data/misc/__init__.py
src/olmo_core/data/misc/code_fresh_export.py
src/olmo_core/data/mixes/OLMo-longmino-mix-0625.txt
src/olmo_core/data/mixes/OLMo-longmino-mix-0925.txt
src/olmo_core/data/mixes/OLMo-midtraining-mix-0625-100B.txt
src/olmo_core/data/mixes/OLMo-midtraining-mix-0925-ingredient1-100B.txt
src/olmo_core/data/mixes/OLMo-midtraining-mix-0925-ingredient2-100B.txt
src/olmo_core/data/mixes/OLMo-mix-0625-150Bsample.txt
src/olmo_core/data/mixes/OLMo-mix-0625-700Bsample.txt
src/olmo_core/data/mixes/OLMo-mix-0625-official.txt
src/olmo_core/data/mixes/OLMo-mix-0625.txt
src/olmo_core/data/mixes/OLMo-mix-0925-official.txt
src/olmo_core/data/mixes/OLMo-mix-0925.txt
src/olmo_core/data/mixes/OLMoE-mix-0824-dev.txt
src/olmo_core/data/mixes/OLMoE-mix-0824.txt
src/olmo_core/data/mixes/__init__.py
src/olmo_core/data/mixes/c4-ppl-validation.txt
src/olmo_core/data/mixes/code-fresh-ppl-validation.txt
src/olmo_core/data/mixes/dolma17.txt
src/olmo_core/data/mixes/v3-small-ppl-validation.txt
src/olmo_core/distributed/__init__.py
src/olmo_core/distributed/nn.py
src/olmo_core/distributed/utils.py
src/olmo_core/distributed/checkpoint/__init__.py
src/olmo_core/distributed/checkpoint/filesystem.py
src/olmo_core/distributed/checkpoint/utils.py
src/olmo_core/distributed/parallel/__init__.py
src/olmo_core/distributed/parallel/context_parallel.py
src/olmo_core/distributed/parallel/data_parallel.py
src/olmo_core/distributed/parallel/expert_parallel.py
src/olmo_core/distributed/parallel/pipeline_parallel.py
src/olmo_core/distributed/parallel/tensor_parallel.py
src/olmo_core/eval/__init__.py
src/olmo_core/eval/evaluator.py
src/olmo_core/eval/lm_evaluator.py
src/olmo_core/eval/metrics.py
src/olmo_core/eval/task_groups.py
src/olmo_core/float8/__init__.py
src/olmo_core/float8/ao.py
src/olmo_core/generate/__init__.py
src/olmo_core/generate/chat.py
src/olmo_core/generate/sampling.py
src/olmo_core/generate/utils.py
src/olmo_core/generate/generation_module/__init__.py
src/olmo_core/generate/generation_module/config.py
src/olmo_core/generate/generation_module/generation_module.py
src/olmo_core/generate/generation_module/transformer/__init__.py
src/olmo_core/generate/generation_module/transformer/config.py
src/olmo_core/generate/generation_module/transformer/generation_module.py
src/olmo_core/internal/__init__.py
src/olmo_core/internal/common.py
src/olmo_core/internal/cookbook.py
src/olmo_core/internal/experiment.py
src/olmo_core/kernels/__init__.py
src/olmo_core/kernels/build_symm_mem_vdev2d_ext.py
src/olmo_core/kernels/cuda_build_utils.py
src/olmo_core/kernels/cuda_extension_utils.py
src/olmo_core/kernels/grouped_mm.py
src/olmo_core/kernels/grouped_mm_row_offset.py
src/olmo_core/kernels/moe.py
src/olmo_core/kernels/moe_chunk_reorder.py
src/olmo_core/kernels/moe_permute_drop.py
src/olmo_core/kernels/moe_unpermute_bwd.py
src/olmo_core/kernels/mxfp8_linear.py
src/olmo_core/kernels/mxfp8_tensor.py
src/olmo_core/kernels/mxfp8_utils.py
src/olmo_core/kernels/nccl_rma_p2p.py
src/olmo_core/kernels/olmo_symm_mem.py
src/olmo_core/kernels/scaled_grouped_mm.py
src/olmo_core/kernels/swiglu.py
src/olmo_core/kernels/symm_mem_vdev2d.py
src/olmo_core/kernels/cuda/CMakeLists.txt
src/olmo_core/kernels/cuda/grouped_mm_out.cpp
src/olmo_core/kernels/cuda/grouped_mm_row_offset.cu
src/olmo_core/kernels/cuda/moe_chunk_reorder.cpp
src/olmo_core/kernels/cuda/moe_chunk_reorder_kernel.cu
src/olmo_core/kernels/cuda/moe_permute_drop.cpp
src/olmo_core/kernels/cuda/moe_permute_drop_kernel.cu
src/olmo_core/kernels/cuda/moe_unpermute_bwd.cpp
src/olmo_core/kernels/cuda/moe_unpermute_bwd_kernel.cu
src/olmo_core/kernels/cuda/nccl_rma_p2p.cpp
src/olmo_core/kernels/cuda/olmo_symm_mem_all_to_all.cuh
src/olmo_core/kernels/cuda/olmo_symm_mem_all_to_all_kernels.cuh
src/olmo_core/kernels/cuda/olmo_symm_mem_bindings.cpp
src/olmo_core/kernels/cuda/olmo_symm_mem_common.cuh
src/olmo_core/kernels/cuda/olmo_symm_mem_kernels.cu
src/olmo_core/kernels/cuda/olmo_symm_mem_launch_helpers.cuh
src/olmo_core/kernels/cuda/olmo_symm_mem_rowwise.cuh
src/olmo_core/kernels/cuda/olmo_symm_mem_rowwise_kernels.cuh
src/olmo_core/kernels/cuda/olmo_symm_mem_runtime.cuh
src/olmo_core/kernels/cuda/olmo_symm_mem_stream_sync.cuh
src/olmo_core/kernels/cuda/scaled_grouped_mm_out.cpp
src/olmo_core/launch/__init__.py
src/olmo_core/launch/beaker.py
src/olmo_core/launch/beaker_presets.py
src/olmo_core/launch/reorder_ranks_in_gcp.py
src/olmo_core/launch/select_beaker_hosts.py
src/olmo_core/launch/utils.py
src/olmo_core/nn/__init__.py
src/olmo_core/nn/buffer_cache.py
src/olmo_core/nn/config.py
src/olmo_core/nn/convolution.py
src/olmo_core/nn/cross_entropy_loss.py
src/olmo_core/nn/feed_forward.py
src/olmo_core/nn/fp8_weight.py
src/olmo_core/nn/layer_norm.py
src/olmo_core/nn/lm_head.py
src/olmo_core/nn/mxfp8_linear.py
src/olmo_core/nn/output_discard_checkpoint.py
src/olmo_core/nn/residual_stream.py
src/olmo_core/nn/rope.py
src/olmo_core/nn/utils.py
src/olmo_core/nn/_csrc/share_storage.cpp
src/olmo_core/nn/attention/__init__.py
src/olmo_core/nn/attention/backend.py
src/olmo_core/nn/attention/base.py
src/olmo_core/nn/attention/flash_attn_api.py
src/olmo_core/nn/attention/flash_linear_attn_api.py
src/olmo_core/nn/attention/kda.py
src/olmo_core/nn/attention/kv_cache.py
src/olmo_core/nn/attention/recurrent.py
src/olmo_core/nn/attention/ring.py
src/olmo_core/nn/attention/te_attn_api.py
src/olmo_core/nn/conversion/__init__.py
src/olmo_core/nn/conversion/state_converter.py
src/olmo_core/nn/conversion/state_mapping.py
src/olmo_core/nn/ddp/__init__.py
src/olmo_core/nn/ddp/block.py
src/olmo_core/nn/ddp/model.py
src/olmo_core/nn/functional/__init__.py
src/olmo_core/nn/functional/cross_entropy_loss.py
src/olmo_core/nn/hf/__init__.py
src/olmo_core/nn/hf/checkpoint.py
src/olmo_core/nn/hf/config.py
src/olmo_core/nn/hf/convert.py
src/olmo_core/nn/hf/convert_checkpoint.py
src/olmo_core/nn/hf/tokenizer.py
src/olmo_core/nn/moe/__init__.py
src/olmo_core/nn/moe/emo.py
src/olmo_core/nn/moe/loss.py
src/olmo_core/nn/moe/mlp.py
src/olmo_core/nn/moe/moe.py
src/olmo_core/nn/moe/parallel_mlp.py
src/olmo_core/nn/moe/router.py
src/olmo_core/nn/moe/utils.py
src/olmo_core/nn/moe/v2/__init__.py
src/olmo_core/nn/moe/v2/_nvtx_colors.py
src/olmo_core/nn/moe/v2/activation_debug.py
src/olmo_core/nn/moe/v2/checkpointing.py
src/olmo_core/nn/moe/v2/comm.py
src/olmo_core/nn/moe/v2/emo_router.py
src/olmo_core/nn/moe/v2/ep_config.py
src/olmo_core/nn/moe/v2/ep_deepep_v2.py
src/olmo_core/nn/moe/v2/ep_no_sync_1d.py
src/olmo_core/nn/moe/v2/ep_no_sync_buffers.py
src/olmo_core/nn/moe/v2/ep_no_sync_common.py
src/olmo_core/nn/moe/v2/ep_no_sync_rowwise.py
src/olmo_core/nn/moe/v2/ep_no_sync_rowwise_helpers.py
src/olmo_core/nn/moe/v2/ep_no_sync_rowwise_wave.py
src/olmo_core/nn/moe/v2/ep_no_sync_tbo_rowwise.py
src/olmo_core/nn/moe/v2/ep_sync_1d.py
src/olmo_core/nn/moe/v2/fp8.py
src/olmo_core/nn/moe/v2/gpt_oss.py
src/olmo_core/nn/moe/v2/nemotron.py
src/olmo_core/nn/moe/v2/no_ep.py
src/olmo_core/nn/moe/v2/output_discard_checkpoint.py
src/olmo_core/nn/moe/v2/qwen.py
src/olmo_core/nn/moe/v2/routed_experts.py
src/olmo_core/nn/moe/v2/router.py
src/olmo_core/nn/moe/v2/shared_experts.py
src/olmo_core/nn/moe/v2/hf/__init__.py
src/olmo_core/nn/moe/v2/hf/configuration_olmo3moe.py
src/olmo_core/nn/moe/v2/hf/modeling_olmo3moe.py
src/olmo_core/nn/moe/v2/te/__init__.py
src/olmo_core/nn/moe/v2/te/cpu_offload.py
src/olmo_core/nn/parallel/__init__.py
src/olmo_core/nn/parallel/distributed.py
src/olmo_core/nn/transformer/__init__.py
src/olmo_core/nn/transformer/block.py
src/olmo_core/nn/transformer/config.py
src/olmo_core/nn/transformer/init.py
src/olmo_core/nn/transformer/model.py
src/olmo_core/nn/vision/__init__.py
src/olmo_core/nn/vision/config.py
src/olmo_core/nn/vision/connector.py
src/olmo_core/nn/vision/image_vit.py
src/olmo_core/nn/vision/multimodal.py
src/olmo_core/ops/__init__.py
src/olmo_core/ops/emo_document_pool.py
src/olmo_core/ops/grad_accum.py
src/olmo_core/ops/moe.py
src/olmo_core/ops/moe_top16.py
src/olmo_core/ops/rounded_wgrad.py
src/olmo_core/ops/swiglu_pairwise.py
src/olmo_core/optim/__init__.py
src/olmo_core/optim/adam.py
src/olmo_core/optim/adamw.py
src/olmo_core/optim/config.py
src/olmo_core/optim/dion.py
src/olmo_core/optim/grad_debug.py
src/olmo_core/optim/lion.py
src/olmo_core/optim/moe_optimizer.py
src/olmo_core/optim/muon.py
src/olmo_core/optim/noop.py
src/olmo_core/optim/scheduler.py
src/olmo_core/optim/skip_step_optimizer.py
src/olmo_core/testing/__init__.py
src/olmo_core/testing/distributed.py
src/olmo_core/testing/utils.py
src/olmo_core/train/__init__.py
src/olmo_core/train/checkpoint.py
src/olmo_core/train/common.py
src/olmo_core/train/config.py
src/olmo_core/train/globals.py
src/olmo_core/train/trainer.py
src/olmo_core/train/utils.py
src/olmo_core/train/callbacks/__init__.py
src/olmo_core/train/callbacks/batch_size_scheduler.py
src/olmo_core/train/callbacks/beaker.py
src/olmo_core/train/callbacks/callback.py
src/olmo_core/train/callbacks/checkpoint_ready_notifier.py
src/olmo_core/train/callbacks/checkpointer.py
src/olmo_core/train/callbacks/comet.py
src/olmo_core/train/callbacks/config_saver.py
src/olmo_core/train/callbacks/console_logger.py
src/olmo_core/train/callbacks/evaluator_callback.py
src/olmo_core/train/callbacks/gap_monitor.py
src/olmo_core/train/callbacks/garbage_collector.py
src/olmo_core/train/callbacks/gpu_memory_monitor.py
src/olmo_core/train/callbacks/hf_converter.py
src/olmo_core/train/callbacks/list_checkpointer.py
src/olmo_core/train/callbacks/metric_saver.py
src/olmo_core/train/callbacks/model_merger.py
src/olmo_core/train/callbacks/monkey_patcher.py
src/olmo_core/train/callbacks/profiler.py
src/olmo_core/train/callbacks/sequence_length_scheduler.py
src/olmo_core/train/callbacks/slack_notifier.py
src/olmo_core/train/callbacks/speed_monitor.py
src/olmo_core/train/callbacks/stability_monitor.py
src/olmo_core/train/callbacks/wandb.py
src/olmo_core/train/train_module/__init__.py
src/olmo_core/train/train_module/config.py
src/olmo_core/train/train_module/train_module.py
src/olmo_core/train/train_module/transformer/__init__.py
src/olmo_core/train/train_module/transformer/common.py
src/olmo_core/train/train_module/transformer/config.py
src/olmo_core/train/train_module/transformer/ddp_train_module.py
src/olmo_core/train/train_module/transformer/pipeline_train_module.py
src/olmo_core/train/train_module/transformer/train_module.py
src/olmo_core/train/train_module/transformer/pipeline/__init__.py
src/olmo_core/train/train_module/transformer/pipeline/gpu_activation_offload.py
src/olmo_core/train/train_module/transformer/pipeline/helpers.py
src/olmo_core/train/train_module/transformer/pipeline/p2p_executor.py
src/olmo_core/train/train_module/transformer/pipeline/p2p_transport.py
src/olmo_core/train/train_module/transformer/pipeline/pipeline_schedule.py
src/olmo_core/train/train_module/transformer/pipeline/pipeline_stage.py