mlx-bun library API
Documentation
Preparing search index...
@mlx-bun/inference
@mlx-bun/inference/artifacts/glm52
Module @mlx-bun/inference/artifacts/glm52
Classes
ColibriGlm52Container
ColibriGlm52ResidentWeights
ColibriGlm52Weights
Interfaces
ColibriQuantTensorInfo
ColibriTensorInfo
Glm52ContainerCapabilities
Glm52ExpertProjectionLayout
Glm52ExpertReadSegment
Glm52ExpertSlotLayout
Glm52LayoutValidation
Type Aliases
ColibriDtype
ColibriShardFamily
Glm52ExpertProjection
Variables
GLM52_EXPERT_MAX_READ_SEGMENTS
GLM52_EXPERT_SLOT_ALIGNMENT
Functions
buildGlm52ExpertSlotLayout
glm52CanonicalQ4MetalLayout
glm52CanonicalQ8MetalLayout
validateGlm52ContainerLayout
References
describeGlm52MemoryGeometry
→
describeGlm52MemoryGeometry
GLM52_G5_ASPIRATIONAL_DECODE_TPS
→
GLM52_G5_ASPIRATIONAL_DECODE_TPS
GLM52_G5_DEFAULT_ALLOCATOR_RESERVE_BYTES
→
GLM52_G5_DEFAULT_ALLOCATOR_RESERVE_BYTES
GLM52_G5_DEFAULT_BUN_NATIVE_RESERVE_BYTES
→
GLM52_G5_DEFAULT_BUN_NATIVE_RESERVE_BYTES
GLM52_G5_DEFAULT_CONTEXT_TOKENS
→
GLM52_G5_DEFAULT_CONTEXT_TOKENS
GLM52_G5_DEFAULT_MACHINE_BYTES
→
GLM52_G5_DEFAULT_MACHINE_BYTES
GLM52_G5_DEFAULT_MAX_GENERATION_TOKENS
→
GLM52_G5_DEFAULT_MAX_GENERATION_TOKENS
GLM52_G5_DEFAULT_PROCESS_LIMIT_BYTES
→
GLM52_G5_DEFAULT_PROCESS_LIMIT_BYTES
GLM52_G5_DEFAULT_SAFETY_MARGIN_BYTES
→
GLM52_G5_DEFAULT_SAFETY_MARGIN_BYTES
GLM52_G5_DIRECT_ORACLE_WARM_DECODE_TPS
→
GLM52_G5_DIRECT_ORACLE_WARM_DECODE_TPS
GLM52_G5_GIB
→
GLM52_G5_GIB
GLM52_G5_MEASURED_AT
→
GLM52_G5_MEASURED_AT
GLM52_G5_MEASURED_WARM_DECODE_TPS
→
GLM52_G5_MEASURED_WARM_DECODE_TPS
Glm52MemoryGeometry
→
Glm52MemoryGeometry
Glm52MemoryLineItems
→
Glm52MemoryLineItems
Glm52MemoryPlan
→
Glm52MemoryPlan
Glm52MemoryPlanOptions
→
Glm52MemoryPlanOptions
planGlm52Memory
→
planGlm52Memory
planGlm52MemoryForArtifact
→
planGlm52MemoryForArtifact
Settings
Member Visibility
Protected
Inherited
External
Theme
OS
Light
Dark
On This Page
Classes
Colibri
Glm52
Container
Colibri
Glm52
Resident
Weights
Colibri
Glm52
Weights
Interfaces
Colibri
Quant
Tensor
Info
Colibri
Tensor
Info
Glm52
Container
Capabilities
Glm52
Expert
Projection
Layout
Glm52
Expert
Read
Segment
Glm52
Expert
Slot
Layout
Glm52
Layout
Validation
Type Aliases
Colibri
Dtype
Colibri
Shard
Family
Glm52
Expert
Projection
Variables
GLM52_
EXPERT_
MAX_
READ_
SEGMENTS
GLM52_
EXPERT_
SLOT_
ALIGNMENT
Functions
build
Glm52
Expert
Slot
Layout
glm52
Canonical
Q4
Metal
Layout
glm52
Canonical
Q8
Metal
Layout
validate
Glm52
Container
Layout
References
describe
Glm52
Memory
Geometry
GLM52_
G5_
ASPIRATIONAL_
DECODE_
TPS
GLM52_
G5_
DEFAULT_
ALLOCATOR_
RESERVE_
BYTES
GLM52_
G5_
DEFAULT_
BUN_
NATIVE_
RESERVE_
BYTES
GLM52_
G5_
DEFAULT_
CONTEXT_
TOKENS
GLM52_
G5_
DEFAULT_
MACHINE_
BYTES
GLM52_
G5_
DEFAULT_
MAX_
GENERATION_
TOKENS
GLM52_
G5_
DEFAULT_
PROCESS_
LIMIT_
BYTES
GLM52_
G5_
DEFAULT_
SAFETY_
MARGIN_
BYTES
GLM52_
G5_
DIRECT_
ORACLE_
WARM_
DECODE_
TPS
GLM52_
G5_
GIB
GLM52_
G5_
MEASURED_
AT
GLM52_
G5_
MEASURED_
WARM_
DECODE_
TPS
Glm52
Memory
Geometry
Glm52
Memory
Line
Items
Glm52
Memory
Plan
Glm52
Memory
Plan
Options
plan
Glm52
Memory
plan
Glm52
Memory
For
Artifact
Documentation
mlx-bun library API
Loading...