List of all items

Structs

• config::Config
• config::Defaults
• config::Experts
• config::Limits
• config::Overrides
• config::Server
• config::Source
• control::Listener
• control::state::ActiveRequest
• control::state::CacheStats
• control::state::Current
• control::state::MemoryStats
• control::state::SessionStats
• control::state::State
• control::state::Stats
• control::state::Versioned
• control::stats::Expert
• control::stats::Layer
• control::stats::Observation
• control::stats::Page
• control::stats::Rates
• control::stats::Snapshot
• control::stats::Summary
• download::Download
• download::Index
• download::progress::Reporter
• metal::MetalContext
• model::Checkpoint
• model::CheckpointFormat
• model::DataSpan
• model::MappedBytes
• model::MappedObjects
• model::ModelDescription
• model::NgramTable
• model::ObjectId
• model::ObjectInfo
• model::StoredTensor
• model::TableShard
• model::Tensor
• model::TensorId
• model::TensorType
• model::format::CheckpointFormat
• model::index::ArtifactDetails
• model::index::ArtifactLease
• model::index::DiskStore
• model::index::GcReport
• model::index::ModelDetails
• model::index::ModelEntry
• model::index::ModelIndex
• model::index::ModelSummary
• model::index::PackOptions
• model::index::Removal
• model::index::ResolveOptions
• model::index::catalog::LockedCatalog
• model::index::disk::DiskStore
• model::index::hub::Hub
• model::index::hub::RemoteObject
• model::index::hub::RemoteReader
• model::index::hub::Repo
• model::index::hub::Sibling
• model::index::hub::headers::Request
• model::index::packing::PackOptions
• model::index::records::Artifact
• model::index::records::ArtifactDetails
• model::index::records::Catalog
• model::index::records::GcReport
• model::index::records::ModelDetails
• model::index::records::ModelEntry
• model::index::records::ModelSummary
• model::index::records::Removal
• model::index::resolve::ResolveOptions
• model::index::store::ArtifactLease
• model::index::store::Pending
• model::inspect::Checkpoint
• model::inspect::packed::Records
• options::Options
• prefix_cache::Entry
• prefix_cache::Prefill
• prefix_cache::PrefillProgress
• prefix_cache::PrefixCache
• prompt::ChatTemplate
• prompt::Prompt
• quant::QLinear
• qwen4_exp::DenseEntry
• qwen4_exp::ExpertLayout
• qwen4_exp::Manifest
• qwen4_exp::NgramLayout
• qwen4_exp::Qwen4ExpConfig
• qwen4_exp::RopeParams
• qwen4_exp::config::Qwen4ExpConfig
• qwen4_exp::config::RopeParams
• qwen4_exp::cpu::AttnWeights
• qwen4_exp::cpu::CpuModel
• qwen4_exp::cpu::DeltaState
• qwen4_exp::cpu::DeltaWeights
• qwen4_exp::cpu::HcWeights
• qwen4_exp::cpu::KvCache
• qwen4_exp::cpu::Layer
• qwen4_exp::cpu::MoeWeights
• qwen4_exp::cpu::MtpWeights
• qwen4_exp::cpu::PleState
• qwen4_exp::cpu::PleWeights
• qwen4_exp::cpu::State
• qwen4_exp::gpu::Attn
• qwen4_exp::gpu::Delta
• qwen4_exp::gpu::ExpertActivity
• qwen4_exp::gpu::ExpertCounters
• qwen4_exp::gpu::GLayer
• qwen4_exp::gpu::Gpu
• qwen4_exp::gpu::HalfSet
• qwen4_exp::gpu::Hc
• qwen4_exp::gpu::HcBufs
• qwen4_exp::gpu::LaEntry
• qwen4_exp::gpu::LayerStats
• qwen4_exp::gpu::MemoryStats
• qwen4_exp::gpu::Moe
• qwen4_exp::gpu::Mtp
• qwen4_exp::gpu::PhaseStats
• qwen4_exp::gpu::Pipes
• qwen4_exp::gpu::Ple
• qwen4_exp::gpu::PredictionStats
• qwen4_exp::gpu::PrefixState
• qwen4_exp::gpu::Q
• qwen4_exp::gpu::QuantStats
• qwen4_exp::gpu::ReadStats
• qwen4_exp::gpu::Scratch
• qwen4_exp::gpu::T
• qwen4_exp::gpu::activity::ExpertActivity
• qwen4_exp::gpu::activity::ExpertCounters
• qwen4_exp::gpu::activity::layers::LayerStats
• qwen4_exp::gpu::activity::layers::PhaseStats
• qwen4_exp::gpu::activity::layers::PredictionStats
• qwen4_exp::gpu::activity::layers::QuantStats
• qwen4_exp::gpu::activity::reads::ReadSources
• qwen4_exp::gpu::activity::reads::ReadStats
• qwen4_exp::gpu::activity::reads::ReadTicket
• qwen4_exp::gpu::activity::reads::ReadTracker
• qwen4_exp::gpu::budget::PoolMemory
• qwen4_exp::gpu::memory::MemoryStats
• qwen4_exp::gpu::params::AttnPartParams
• qwen4_exp::gpu::params::ConvParams
• qwen4_exp::gpu::params::DeltaPrepParams
• qwen4_exp::gpu::params::GroupParams
• qwen4_exp::gpu::params::IndexParams
• qwen4_exp::gpu::params::KvQParams
• qwen4_exp::gpu::params::MoeBParams
• qwen4_exp::gpu::params::NormPrepParams
• qwen4_exp::gpu::params::PleConvParams
• qwen4_exp::gpu::params::QkRopeParams
• qwen4_exp::gpu::params::QmvParams
• qwen4_exp::gpu::phases::ClockSample
• qwen4_exp::gpu::phases::CpuWindow
• qwen4_exp::gpu::phases::MachTimebase
• qwen4_exp::gpu::phases::PhaseTimer
• qwen4_exp::gpu::prefill::ChunkStats
• qwen4_exp::gpu::prefill::MoeRef
• qwen4_exp::gpu::prefill::PrefillScratch
• qwen4_exp::gpu::prefill::PrefillStats
• qwen4_exp::gpu::prefill::attention::AttnGemmParams
• qwen4_exp::gpu::prefill::attention::GemmHParams
• qwen4_exp::gpu::prefill::attention::SelMaskParams
• qwen4_exp::gpu::prefill::experts::ExpertJob
• qwen4_exp::gpu::prefill::stats::PrefillStats
• qwen4_exp::gpu::residency::CopyPool
• qwen4_exp::gpu::residency::Landed
• qwen4_exp::gpu::residency::PlannedRead
• qwen4_exp::gpu::residency::ReadPlan
• qwen4_exp::gpu::residency::RecordRead
• qwen4_exp::gpu::residency::Residency
• qwen4_exp::gpu::streaming::BlockSignals
• qwen4_exp::gpu::streaming::PendingRead
• qwen4_exp::lowbit::Layout
• qwen4_exp::manifest::DenseEntry
• qwen4_exp::manifest::ExpertLayout
• qwen4_exp::manifest::Manifest
• qwen4_exp::manifest::NgramLayout
• qwen4_exp::pack::Progress
• qwen4_exp::pack::affine::Group
• qwen4_exp::pack::source::Quantized
• qwen4_exp::pack::source::Source
• qwen4_exp::pack::source::Tensor
• qwen4_exp::packed::ExpertRef
• qwen4_exp::packed::NgramMetadata
• qwen4_exp::packed::Packed
• runner::PrefillResume
• runner::decode::Decode
• sampling::Sampler
• sampling::Sampling
• server::Job
• server::UsageStats
• server::failure::Failure
• server::output::Output
• server::pacer::ChunkPacer
• server::registry::Registry
• server::registry::Ticket
• server::request::PreparedRequest
• server::request::Request
• server::request::SessionInput
• server::response::Response
• server::sessions::Commit
• server::sessions::Session
• server::sessions::SessionStats
• server::sessions::Store
• server::sessions::Turn
• server::stats::UsageStats
• server::tool_call::GeneratedToolCall
• server::tool_call::Parameter
• server::tool_call::ParsedToolCallOutput
• server::tool_call::RawParameter
• server::tool_call::RawToolCall
• server::tool_call::RegionParser
• server::tool_call::Terminal
• server::tool_call::Tool
• server::tool_call::ToolCallOutputContract
• server::tool_call::ToolCallOutputDecoder
• server::tool_call::ToolCallParseDiagnostics
• server::tool_call::WireToolCall
• server::worker::Active
• server::worker::GeneratedText
• server::worker::Pending
• server::worker::Worker
• storage::Paths
• tensors::ModelWeights
• tensors::TensorInfo
• tok::ChatTokenizer

Enums

• control::Command
• model::AffineOffset
• model::Architecture
• model::BitPacking
• model::CheckpointConventions
• model::CompatibilityIssue
• model::ContainerFormat
• model::Dtype
• model::GgmlEncoding
• model::NgramHash
• model::Preparation
• model::PreparationStep
• model::TensorEncoding
• model::TensorRole
• model::format::CheckpointConventions
• model::format::ContainerFormat
• model::index::ArtifactKind
• model::index::DiskLayout
• model::index::ModelEvent
• model::index::Source
• model::index::disk::DiskLayout
• model::index::events::ModelEvent
• model::index::records::ArtifactKind
• model::index::records::Source
• model::index::selector::Selector
• model::preparation::CompatibilityIssue
• model::preparation::Preparation
• model::preparation::PreparationStep
• options::PoolBudget
• qwen4_exp::cpu::Mixer
• qwen4_exp::gpu::GpuTiming
• qwen4_exp::gpu::Mix
• qwen4_exp::gpu::activity::reads::ReadSource
• qwen4_exp::gpu::phases::CpuPhase
• qwen4_exp::gpu::phases::GpuTiming
• qwen4_exp::gpu::prefill::experts::ExpertSource
• qwen4_exp::gpu::residency::Pool
• qwen4_exp::pack::Class
• qwen4_exp::pack::source::Conversion
• qwen4_exp::pack::source::Part
• server::ApiKind
• server::output::Frame
• server::tool_call::Disposition
• server::tool_call::FallbackReason
• server::tool_call::JsonValueKind
• server::worker::Phase
• tensors::Dtype

Traits

• model::ByteSource

Macros

• kernels::metal_library

Functions

• config::absolute
• config::default_socket
• control::activity::experts
• control::activity::layers
• control::activity::page
• control::activity::range
• control::activity::ratio
• control::activity::summary
• control::handle
• control::page_limit
• control::query
• control::read_frame
• control::same_user
• control::wait_readable
• control::write_frame
• download::missing_bytes
• download::publish
• download::run
• download::run_at
• download::shards
• model::index::anchor_selector
• model::index::catalog::valid_id
• model::index::catalog::validate
• model::index::disk::registered
• model::index::hub::ambient_token
• model::index::hub::endpoint
• model::index::hub::inspect
• model::index::id
• model::index::is_reference
• model::index::local::copy_store
• model::index::local::fingerprint
• model::index::packing::check_local
• model::index::packing::copy_runtime_metadata
• model::index::packing::remove_variants
• model::index::packing::sync_files
• model::index::prepared_precisions
• model::index::reference::display_source
• model::index::reference::find
• model::index::reference::is_revision
• model::index::reference::lookup
• model::index::reference::matches_selector
• model::index::reference::preferred
• model::index::reference::selects
• model::index::resolve_path
• model::index::resolve_runtime
• model::index::same_source
• model::index::selector::legacy
• model::index::selector::parse
• model::index::selector::revision
• model::index::selector::source
• model::index::selector::validate_revision
• model::index::store::bytes
• model::index::store::precisions
• model::index::store::rebuild_variants
• model::index::store::references
• model::index::validate_name
• model::index::validate_selector
• model::inspect::architecture
• model::inspect::declares_mlx
• model::inspect::describe_inventory
• model::inspect::describe_raw
• model::inspect::formats
• model::inspect::ngram_table
• model::inspect::packed::affine
• model::inspect::packed::dense
• model::inspect::packed::dense_part
• model::inspect::packed::experts
• model::inspect::packed::ngram
• model::inspect::packed::open
• model::inspect::shard_number
• model::inspect::unsigned
• model::preparation::compatibility
• model::qwen::tensor_role
• nn::l2_norm
• nn::rms_norm
• nn::sigmoid
• nn::silu
• nn::softmax
• nn::softplus
• options::cut
• options::positive
• prefix_cache::chunk_len
• prefix_cache::matches
• prefix_cache::pacing_sample
• prefix_cache::prefill_rows
• prompt::common_prefix
• prompt::embedded_template
• prompt::ensure_reasoning_content_absent
• prompt::function_call_name_and_arguments
• prompt::normalize_content
• prompt::text_messages
• prompt::tool_call_value
• prompt::valid_function_name
• prompt::validate_assistant_message
• prompt::validate_plain_message
• prompt::validate_tool_message
• qwen4_exp::config::d320
• qwen4_exp::config::eight
• qwen4_exp::config::four
• qwen4_exp::config::silu
• qwen4_exp::config::three
• qwen4_exp::config::yes
• qwen4_exp::cpu::bf16_dot
• qwen4_exp::cpu::bf16_matvec
• qwen4_exp::cpu::causal_conv
• qwen4_exp::cpu::grouped_rms_norm
• qwen4_exp::cpu::hc
• qwen4_exp::cpu::inject
• qwen4_exp::cpu::load_layer
• qwen4_exp::cpu::mlp
• qwen4_exp::cpu::rms_norm_shift
• qwen4_exp::cpu::rope_partial
• qwen4_exp::gpu::kv_q8_side
• qwen4_exp::gpu::layer_cap
• qwen4_exp::gpu::phases::accumulate_phases
• qwen4_exp::gpu::phases::clock_scale
• qwen4_exp::gpu::phases::ensure_indices
• qwen4_exp::gpu::phases::observation_delay
• qwen4_exp::gpu::phases::sample_clock
• qwen4_exp::gpu::phases::signed_delta
• qwen4_exp::gpu::phases::thread_cpu_seconds
• qwen4_exp::gpu::ple::prefetch_ngram_rows
• qwen4_exp::gpu::prefill::allocation::rows_fit
• qwen4_exp::gpu::prefill::allocation::scratch
• qwen4_exp::gpu::prefill::allocation::scratch_bytes
• qwen4_exp::gpu::residency::fetch_into_slots
• qwen4_exp::gpu::residency::read_at
• qwen4_exp::gpu::residency::read_record
• qwen4_exp::gpu::set_bytes
• qwen4_exp::gpu::union_of
• qwen4_exp::lowbit::announce_build
• qwen4_exp::lowbit::build
• qwen4_exp::lowbit::create_output
• qwen4_exp::lowbit::ensure
• qwen4_exp::lowbit::ensure_many
• qwen4_exp::lowbit::ensure_selected
• qwen4_exp::lowbit::ensure_with_policy
• qwen4_exp::lowbit::is_usable
• qwen4_exp::lowbit::pack_matrix
• qwen4_exp::lowbit::pack_record
• qwen4_exp::lowbit::pack_records
• qwen4_exp::lowbit::paths
• qwen4_exp::lowbit::q2_slots
• qwen4_exp::lowbit::q3_slots
• qwen4_exp::lowbit::spot_check
• qwen4_exp::lowbit::usable
• qwen4_exp::pack::affine::quantize
• qwen4_exp::pack::classify
• qwen4_exp::pack::copy_chat_metadata
• qwen4_exp::pack::dtype_name
• qwen4_exp::pack::missing_chat_metadata
• qwen4_exp::pack::pack
• qwen4_exp::pack::pack_dense
• qwen4_exp::pack::pack_experts
• qwen4_exp::pack::pack_ngram
• qwen4_exp::pack::pad_to
• qwen4_exp::pack::prepare
• qwen4_exp::pack::source::folded_norm
• qwen4_exp::pack::source::normalized_name
• qwen4_exp::pack::source::quantized_weight
• qwen4_exp::pack::source::write_norm
• qwen4_exp::packed::as_bf16
• qwen4_exp::packed::as_u32
• qwen4_exp::packed::map
• runner::check_budget
• runner::diagnostics::argmax
• runner::diagnostics::dump_decode
• runner::diagnostics::dump_prefill_chunk
• runner::diagnostics::dump_run
• runner::diagnostics::logits_diff
• runner::diagnostics::qwen4_exp_check_rows
• runner::prefill_engine
• runner::prefill_prompt
• runner::prefill_row_batches
• runner::qwen4_exp_gen_once
• runner::run
• sampling::nucleus
• sampling::rank
• server::failure::status
• server::http::error
• server::http::line
• server::http::read_http
• server::http::respond
• server::http::sse
• server::output::write_frames
• server::registry::id
• server::registry::tool_call_id
• server::request::context_tokens
• server::request::legacy_function_call
• server::request::parse_legacy_function_controls
• server::request::parse_parallel_tool_calls
• server::request::parse_request
• server::request::parse_tool_choice
• server::request::parse_tools
• server::request::render_prompt
• server::routes::connection
• server::routes::request_error
• server::routes::session_endpoint
• server::serve
• server::tool_call::admits_value
• server::tool_call::classify_json_value
• server::tool_call::compile_direct_types
• server::tool_call::compile_schema_types
• server::tool_call::encode_json_string
• server::tool_call::fallback
• server::tool_call::is_ws
• server::tool_call::json_number_is_integer
• server::tool_call::normalize_declared_parameter
• server::tool_call::normalize_parameter
• server::tool_call::normalize_raw_tool_call
• server::tool_call::parse_qwen_tool_call_output
• server::tool_call::same_tool
• server::tool_call::schema_bit
• server::tool_call::skip_ws
• server::tool_call::starts_at
• server::tool_call::strip_parameter_framing
• server::tool_call::trim_ws
• server::tool_call::valid_function_name
• storage::create_private_dir
• storage::default_model_reference
• storage::require_space
• storage::safe_component
• storage::validate_identity
• storage::xdg_dir

Type Aliases

• model::index::hub::headers::Shard
• qwen4_exp::gpu::Buf
• qwen4_exp::gpu::Enc
• qwen4_exp::gpu::Pso
• qwen4_exp::gpu::RouteStep
• qwen4_exp::gpu::residency::Buf

Constants

• control::MAX_FRAME
• control::TIMEOUT
• download::METADATA
• kernels::BATCH_MSL
• kernels::CLOCK_PROBE_MSL
• kernels::FORWARD_MSL
• metal::PAGE_SIZE
• model::index::hub::METADATA_LIMIT
• model::index::hub::headers::CONCURRENCY
• quant::GROUP_SIZE
• quant::NIBBLES_PER_WORD
• qwen4_exp::PAGE
• qwen4_exp::gpu::ARGMAX_TGS
• qwen4_exp::gpu::ATTN_MAX_WG
• qwen4_exp::gpu::ATTN_TB
• qwen4_exp::gpu::IDS_IN
• qwen4_exp::gpu::IDS_MTP_IN
• qwen4_exp::gpu::IDS_MTP_OUT
• qwen4_exp::gpu::IDS_OUT
• qwen4_exp::gpu::MAX_NB
• qwen4_exp::gpu::MAX_SNAP
• qwen4_exp::gpu::SLOT_STRIDE
• qwen4_exp::gpu::phases::SECONDS_PER_NANOSECOND
• qwen4_exp::gpu::prefill::GROUP
• qwen4_exp::gpu::prefill::MAX_PREFILL_ROWS
• qwen4_exp::gpu::prefill::QS
• qwen4_exp::gpu::prefill::RING
• qwen4_exp::lowbit::PAGE
• qwen4_exp::manifest::PAGE
• qwen4_exp::pack::CHAT_METADATA_FILES
• qwen4_exp::pack::DENSE_ALIGN
• server::MODEL
• server::registry::CANCELLED
• server::registry::COMPLETED
• server::registry::LIVE
• server::tool_call::FUNCTION_CLOSE
• server::tool_call::FUNCTION_OPEN
• server::tool_call::MAX_TOOL_NAME_LENGTH
• server::tool_call::PARAM_CLOSE
• server::tool_call::PARAM_OPEN
• server::tool_call::TOOL_CLOSE
• server::tool_call::TOOL_OPEN
• server::tool_call::T_ARRAY
• server::tool_call::T_BOOLEAN
• server::tool_call::T_INTEGER
• server::tool_call::T_NULL
• server::tool_call::T_NUMBER
• server::tool_call::T_OBJECT
• server::tool_call::T_STRING
• storage::DEFAULT_REPO
• storage::DEFAULT_REVISION
• units::BYTES_PER_GB
• units::BYTES_PER_KIB
• units::BYTES_PER_MIB