modelexpress
Model Express is a Rust-based component meant to be placed next to existing model inference systems to speed up their startup times and improve overall performance.
File Explorer
Download Latest Version (.zip)- settings.json
- rust.mdc
- devcontainer.json
- Dockerfile
- action.yml
- action.yml
- action.yml
- action.yml
- action.yml
- action.yml
- action.yml
- vram-sampling.sh
- action.yml
- action.yml
- action.yml
- build-wheels.yml
- ci.yml
- codeql.yml
- common.ps1
- copyright-check.ps1
- copyright-checks.yml
- label-pr.yml
- lint-pr-title.yml
- mirror_repo.sh
- modelexpress-ci-tests.yml
- trigger_ci.yml
- CODEOWNERS
- copilot-instructions.md
- copy-pr-bot.yaml
- dco.yml
- extensions.json
- settings.json
- Dockerfile
- Dockerfile.mooncake
- manifest-azure-mooncake.yaml
- manifest-azure-rolling-update.yaml
- manifest-azure-tp2.yaml
- manifest-azure.yaml
- Dockerfile
- manifest-azure-aggregated.yaml
- manifest-azure-disaggregated.yaml
- Dockerfile
- manifest-azure-dp2.yaml
- manifest-azure-ep2.yaml
- manifest-azure-fleet.yaml
- manifest-azure-multi-node-tp2.yaml
- manifest-azure-rolling-update.yaml
- manifest-azure-tp2.yaml
- manifest-azure.yaml
- manifest-gcp.yaml
- conftest.py
- kube_utils.py
- mx_sources.py
- test_dynamo_p2p.py
- test_fleet_scale.py
- test_p2p_artifacts.py
- test_p2p_k8s.py
- test_rolling_update.py
- test_stale_metadata.py
- crd-modelmetadata.yaml
- Dockerfile.server
- mx-server-redis.yaml
- mx-server.yaml
- rbac-modelmetadata.yaml
- TEST_PLAN.md
- docker-compose.yml
- Dockerfile
- Dockerfile.client-wheel
- benchmark-artifact-transfer.png
- benchmark-cold-start-loading.png
- benchmark-nixl-registration.png
- model-express-architecture.png
- model-express-key-visual.jpg
- model-express-runtime-paths.png
- ARCHITECTURE.md
- BENCHMARKS.md
- CLI.md
- DEPLOYMENT.md
- GCS_PROVIDER.md
- K8S_SERVICE_BACKEND.md
- metadata.md
- METRICS.md
- SGLANG.md
- agg.yaml
- README.md
- rbac-modelmetadata.yaml
- vllm-multi-node-aggregated.yaml
- vllm-single-node-disaggregated.yaml
- Dockerfile
- README.md
- README.md
- sources-tp2-single-pod.yaml
- sources-tp2.yaml
- target.yaml
- Dockerfile
- README.md
- sglang-single-node-streamer-azure.yaml
- sglang-single-node-streamer-local.yaml
- sglang-single-node-streamer-s3.yaml
- README.md
- vllm-single-node-streamer-azure.yaml
- vllm-single-node-streamer-local.yaml
- vllm-single-node-streamer-s3.yaml
- README.md
- Dockerfile
- sglang-single-node-p2p.yaml
- sglang-single-node-transfer-engine.yaml
- Dockerfile
- README.md
- trtllm-single-node-p2p.yaml
- Dockerfile
- README.md
- vllm-aws-efa.yaml
- Dockerfile
- vllm-multi-node.yaml
- vllm-single-node-metrics.yaml
- vllm-single-node.yaml
- README.md
- modelexpress-server-kubernetes.yaml
- rbac-modelmetadata.yaml
- modelexpress-server-redis.yaml
- redis-standalone.yaml
- README.md
- model-download.yaml
- README.md
- crds.yaml
- modelexpress-crds.yaml
- test-connection.yaml
- _helpers.tpl
- deployment.yaml
- ingress.yaml
- NOTES.txt
- pvc.yaml
- rbac.yaml
- service.yaml
- serviceaccount.yaml
- .helmignore
- Chart.yaml
- deploy.sh
- README.md
- test-values.yaml
- values-development.yaml
- values-local-storage.yaml
- values-production.yaml
- values.yaml
- api.pb.go
- api_grpc.pb.go
- health.pb.go
- health_grpc.pb.go
- model.pb.go
- model_grpc.pb.go
- p2p.pb.go
- p2p_grpc.pb.go
- generate_proto.sh
- go.mod
- go.sum
- README.md
- __init__.py
- base.py
- cuda.py
- xpu.py
- __init__.py
- adapter.py
- artifacts.py
- loader.py
- __init__.py
- adapter.py
- loader.py
- __init__.py
- _version.py
- patch_hf_snapshot_prefetch.py
- patch_humming_regex_ignore.py
- patch_object_storage_format_check.py
- __init__.py
- installer.py
- receiver.py
- __init__.py
- adapter.py
- artifacts.py
- loader.py
- mdl.py
- registration.py
- source_identity.py
- __init__.py
- __init__.py
- base.py
- context.py
- default_strategy.py
- gds_strategy.py
- instant_tensor_strategy.py
- model_streamer_strategy.py
- rdma_strategy.py
- server_cache_strategy.py
- __init__.py
- artifact_lifecycle.py
- artifact_manifest.py
- artifact_transfer.py
- client_factory.py
- k8s_service_client.py
- payload.py
- publish.py
- publisher.py
- source_id.py
- worker_server.py
- __init__.py
- receiver-driven-refit.svg
- reshard-while-moving.svg
- rl-refit-critical-path.svg
- __init__.py
- base.py
- nixl.py
- __init__.py
- cuda_pool.py
- geometry.py
- megatron_aliases.py
- megatron_publisher.py
- receiver.py
- rendezvous.py
- slice_plan.py
- transfer_plan.py
- types.py
- verify.py
- __init__.py
- README.md
- timing.py
- __init__.py
- _alloc_ext.cpp
- arena.py
- backend.py
- hook.py
- README.md
- runtime.py
- __init__.py
- _nixl.py
- adapter.py
- auth.py
- client.py
- envs.py
- gds_loader.py
- gds_transfer.py
- lifecycle.py
- metrics.py
- model_client.py
- model_pb2.py
- model_pb2_grpc.py
- model_prefetch.py
- model_snapshot.py
- nixl_transfer.py
- p2p_pb2.py
- p2p_pb2_grpc.py
- rank_utils.py
- refit_timing.py
- source_selection.py
- tensor_utils.py
- tracing.py
- transfer_safety.py
- types.py
- ucx_utils.py
- vllm_loader.py
- vllm_worker.py
- __init__.py
- adapter.py
- context.py
- installer.py
- __init__.py
- __init__.py
- layout.py
- receiver.py
- __init__.py
- __init__.py
- adapter.py
- client.py
- nixl_staged_transfer.py
- __init__.py
- adapter.py
- publisher.py
- __init__.py
- adapter.py
- aliases.py
- publisher.py
- selection.py
- __init__.py
- __init__.py
- adapter.py
- client.py
- manifest.py
- resources.py
- __init__.py
- control.py
- envs.py
- refit_pb2.py
- refit_pb2_grpc.py
- version.py
- conftest.py
- __init__.py
- conftest.py
- test_accelerator_backend.py
- test_adapter.py
- test_artifact_health_url.py
- test_artifact_lifecycle.py
- test_artifact_manifest.py
- test_artifact_transfer.py
- test_auth.py
- test_client.py
- test_envs.py
- test_gds_loader.py
- test_hf_snapshot_prefetch_patch.py
- test_instant_tensor_strategy.py
- test_k8s_service_client.py
- test_lifecycle.py
- test_mdl_fp8.py
- test_mdl_tp.py
- test_metrics.py
- test_metrics_deployment.py
- test_model_client.py
- test_model_prefetch.py
- test_model_snapshot.py
- test_model_streamer_strategy.py
- test_nixl_backend.py
- test_nixl_import.py
- test_nixl_peer_lifecycle.py
- test_patches.py
- test_payload.py
- test_pool_registration.py
- test_publisher.py
- test_publisher_peer_teardown.py
- test_rdma_peer_release.py
- test_rdma_transfer_timeout.py
- test_refit_control_client.py
- test_refit_envs.py
- test_refit_fsdp_adapter.py
- test_refit_fsdp_publisher.py
- test_refit_generator_client.py
- test_refit_megatron_adapter.py
- test_refit_megatron_selection.py
- test_refit_nixl_staged_transfer.py
- test_refit_timing.py
- test_refit_trainer_client.py
- test_refit_trainer_resources.py
- test_refit_vllm_adapter.py
- test_refit_vllm_installer.py
- test_reshard_megatron.py
- test_reshard_megatron_gqa.py
- test_reshard_refit_batch_install.py
- test_reshard_refit_ceiling.py
- test_reshard_refit_coverage.py
- test_reshard_refit_descriptor_cache.py
- test_reshard_refit_fused_wire.py
- test_reshard_refit_geometry.py
- test_reshard_refit_handshake.py
- test_reshard_refit_moe_experts.py
- test_reshard_refit_nixl_transport.py
- test_reshard_refit_rendezvous.py
- test_reshard_refit_replica_merge.py
- test_reshard_refit_slice_plan.py
- test_reshard_refit_stage_record.py
- test_reshard_refit_step_stamp.py
- test_reshard_refit_transfer.py
- test_reshard_refit_verify.py
- test_server_cache_strategy.py
- test_sglang_artifacts.py
- test_sglang_loader.py
- test_source_id.py
- test_source_selection.py
- test_tensor_utils.py
- test_tracing.py
- test_transfer_safety.py
- test_trtllm_adapter.py
- test_trtllm_loader.py
- test_trtllm_native_example.py
- test_types.py
- test_ucx_utils.py
- test_vllm_adapter.py
- test_vllm_artifacts.py
- test_vllm_loader.py
- test_vllm_source_identity.py
- test_vmm_arena.py
- test_vmm_backend.py
- test_vmm_hook.py
- .gitignore
- generate_proto.sh
- pyproject.toml
- README.md
- setup.py
- uv.lock
- args.rs
- handlers.rs
- output.rs
- payload.rs
- cli.rs
- fallback_test.rs
- test_client.rs
- auth.rs
- lib.rs
- Cargo.toml
- api.proto
- health.proto
- model.proto
- p2p.proto
- refit.proto
- cache_manifest.rs
- download_progress.rs
- download_task.rs
- downloader.rs
- model_dir.rs
- model_name.rs
- path_ext.rs
- provider_cache.rs
- gcs.rs
- huggingface.rs
- lock_file.rs
- ngc.rs
- s3.rs
- artifact_manifest.rs
- cache.rs
- client_config.rs
- config.rs
- download.rs
- envs.rs
- lib.rs
- models.rs
- providers.rs
- test_support.rs
- build.rs
- Cargo.toml
- layer.rs
- token.rs
- config_gen.rs
- exposition.rs
- kubernetes.rs
- memory.rs
- redis.rs
- backend.rs
- k8s_types.rs
- reaper.rs
- service.rs
- source_identity.rs
- state.rs
- create_weight_version.lua
- create_weight_version_shard.lua
- delete_version_lease.lua
- delete_weight_version.lua
- delete_weight_version_shard.lua
- register_version_lease.lua
- register_worker.lua
- redis.rs
- backend.rs
- service.rs
- kubernetes.rs
- memory.rs
- redis.rs
- backend.rs
- entry_key.rs
- k8s_types.rs
- state.rs
- auth.rs
- backend_config.rs
- cache.rs
- config.rs
- lib.rs
- main.rs
- metrics.rs
- p2p.rs
- refit.rs
- registry.rs
- server.rs
- services.rs
- auth_e2e.rs
- in_process_server.rs
- Cargo.toml
- performance.rs
- artifact_transfer_contract.rs
- helm_namespace_scope.rs
- integration_tests.rs
- k8s_multinode_tests.rs
- refit_service_redis.rs
- registry_backend_redis.rs
- Cargo.toml
- .coderabbit.yaml
- .dockerignore
- .gitignore
- .pre-commit-config.yaml
- Cargo.lock
- Cargo.toml
- CLAUDE.md
- CODE_OF_CONDUCT.md
- CONTRIBUTING.md
- deny.toml
- LICENSE
- modelexpress-cli-completion.bash
- NOTICES
- README.md
- run_integration_tests.sh
- rust-toolchain.toml
- rustfmt.toml
- SECURITY.md
- test_client.sh
- test_grpc_transfer_k8s.sh
- test_multinode_k8s.sh
// repository documentation
Was this content helpful?
(0 ratings)