-
Notifications
You must be signed in to change notification settings - Fork 29
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
feat(xla): support Gemma3n audio with dense PLE prefill
area:architectureArchitecture and code structure changesArchitecture and code structure changesarea:inferenceGeneration, sampling, decoding (incl. speculative, DRY)Generation, sampling, decoding (incl. speculative, DRY)area:modelsModel architectures, weights, loading, metadataModel architectures, weights, loading, metadatapriority:lowLow priorityLow prioritystatus:blockedBlocked by dependencies or other issuesBlocked by dependencies or other issuestype:enhancementNew features, capabilities, or significant additionsNew features, capabilities, or significant additionsStatus: Open.#878 In lablup/mlxcel;feat(xla): support Phi4MM audio and per-slot speech adapters
area:architectureArchitecture and code structure changesArchitecture and code structure changesarea:inferenceGeneration, sampling, decoding (incl. speculative, DRY)Generation, sampling, decoding (incl. speculative, DRY)area:modelsModel architectures, weights, loading, metadataModel architectures, weights, loading, metadatapriority:lowLow priorityLow prioritystatus:blockedBlocked by dependencies or other issuesBlocked by dependencies or other issuestype:enhancementNew features, capabilities, or significant additionsNew features, capabilities, or significant additionsStatus: Open.#877 In lablup/mlxcel;feat(xla): add the Gemma3n text backbone and dense PLE prefill
area:architectureArchitecture and code structure changesArchitecture and code structure changesarea:inferenceGeneration, sampling, decoding (incl. speculative, DRY)Generation, sampling, decoding (incl. speculative, DRY)area:modelsModel architectures, weights, loading, metadataModel architectures, weights, loading, metadatapriority:mediumMedium priorityMedium prioritystatus:blockedBlocked by dependencies or other issuesBlocked by dependencies or other issuestype:enhancementNew features, capabilities, or significant additionsNew features, capabilities, or significant additionsStatus: Open.#876 In lablup/mlxcel;feat(vlm): add a token-exact Gemma3n audio reference path
area:inferenceGeneration, sampling, decoding (incl. speculative, DRY)Generation, sampling, decoding (incl. speculative, DRY)area:modelsModel architectures, weights, loading, metadataModel architectures, weights, loading, metadatapriority:mediumMedium priorityMedium prioritystatus:readyReady to be worked onReady to be worked ontype:enhancementNew features, capabilities, or significant additionsNew features, capabilities, or significant additionsStatus: Open.#875 In lablup/mlxcel;feat(vlm): add a token-exact Phi4MM audio reference path
area:inferenceGeneration, sampling, decoding (incl. speculative, DRY)Generation, sampling, decoding (incl. speculative, DRY)area:modelsModel architectures, weights, loading, metadataModel architectures, weights, loading, metadatapriority:mediumMedium priorityMedium prioritystatus:readyReady to be worked onReady to be worked ontype:enhancementNew features, capabilities, or significant additionsNew features, capabilities, or significant additionsStatus: Open.#874 In lablup/mlxcel;feat(xla): add bounded audio preprocessing and serving plumbing
area:architectureArchitecture and code structure changesArchitecture and code structure changesarea:inferenceGeneration, sampling, decoding (incl. speculative, DRY)Generation, sampling, decoding (incl. speculative, DRY)priority:mediumMedium priorityMedium prioritystatus:blockedBlocked by dependencies or other issuesBlocked by dependencies or other issuestype:enhancementNew features, capabilities, or significant additionsNew features, capabilities, or significant additionsStatus: Open.#873 In lablup/mlxcel;feat(xla): support Youtu-VL windowed vision and MLA inference
area:inferenceGeneration, sampling, decoding (incl. speculative, DRY)Generation, sampling, decoding (incl. speculative, DRY)area:modelsModel architectures, weights, loading, metadataModel architectures, weights, loading, metadatapriority:lowLow priorityLow prioritystatus:blockedBlocked by dependencies or other issuesBlocked by dependencies or other issuestype:enhancementNew features, capabilities, or significant additionsNew features, capabilities, or significant additionsStatus: Open.#872 In lablup/mlxcel;feat(xla): support Molmo2 indexed attention pooling
area:inferenceGeneration, sampling, decoding (incl. speculative, DRY)Generation, sampling, decoding (incl. speculative, DRY)area:modelsModel architectures, weights, loading, metadataModel architectures, weights, loading, metadatapriority:lowLow priorityLow prioritystatus:blockedBlocked by dependencies or other issuesBlocked by dependencies or other issuestype:enhancementNew features, capabilities, or significant additionsNew features, capabilities, or significant additionsStatus: Open.#871 In lablup/mlxcel;feat(xla): support Molmo sparse additive visual embeddings
area:inferenceGeneration, sampling, decoding (incl. speculative, DRY)Generation, sampling, decoding (incl. speculative, DRY)area:modelsModel architectures, weights, loading, metadataModel architectures, weights, loading, metadatapriority:lowLow priorityLow prioritystatus:blockedBlocked by dependencies or other issuesBlocked by dependencies or other issuestype:enhancementNew features, capabilities, or significant additionsNew features, capabilities, or significant additionsStatus: Open.#870 In lablup/mlxcel;feat(xla): support Gemma3 VLM additive prefill masks
area:inferenceGeneration, sampling, decoding (incl. speculative, DRY)Generation, sampling, decoding (incl. speculative, DRY)area:modelsModel architectures, weights, loading, metadataModel architectures, weights, loading, metadatapriority:lowLow priorityLow prioritystatus:blockedBlocked by dependencies or other issuesBlocked by dependencies or other issuestype:enhancementNew features, capabilities, or significant additionsNew features, capabilities, or significant additionsStatus: Open.#869 In lablup/mlxcel;feat(xla): support Qwen3-VL and DeepStack vision features
area:inferenceGeneration, sampling, decoding (incl. speculative, DRY)Generation, sampling, decoding (incl. speculative, DRY)area:modelsModel architectures, weights, loading, metadataModel architectures, weights, loading, metadatapriority:lowLow priorityLow prioritystatus:blockedBlocked by dependencies or other issuesBlocked by dependencies or other issuestype:enhancementNew features, capabilities, or significant additionsNew features, capabilities, or significant additionsStatus: Open.#868 In lablup/mlxcel;feat(xla): add per-layer multimodal feature injection to prefill
area:architectureArchitecture and code structure changesArchitecture and code structure changesarea:coremlxcel-core: MLX FFI, primitives, KV cache, layersmlxcel-core: MLX FFI, primitives, KV cache, layersarea:inferenceGeneration, sampling, decoding (incl. speculative, DRY)Generation, sampling, decoding (incl. speculative, DRY)priority:mediumMedium priorityMedium prioritystatus:blockedBlocked by dependencies or other issuesBlocked by dependencies or other issuestype:enhancementNew features, capabilities, or significant additionsNew features, capabilities, or significant additionsStatus: Open.#867 In lablup/mlxcel;