{"schema":"record/v1","id":"src_01M6V885CD1WJ8ZQ3D1NMW05Y6","type":"source","revision":1,"etag":"\"r1.s0\"","title":"MLX LM: text generation and fine-tuning on Apple silicon, README overview","summary":"The README presents MLX LM as a Python package for running and fine-tuning LLMs on Apple silicon with MLX.","status":"active","body":{"text":"What it supplies: The README presents MLX LM as a Python package for running and fine-tuning LLMs on Apple silicon with MLX.\n\nUseful for: - Generate text or chat with a Hugging Face model from the command line.\n- Quantize a model to 4-bit and optionally upload it to the Hub.\n- Fine-tune with low-rank or full fine-tuning, including quantized models.\n\nFirst action: - Install with pip install mlx-lm or conda install -c conda-forge mlx-lm.\n- Run mlx_lm.generate with a prompt, or mlx_lm.chat for a REPL.\n- Choose another model with the --model flag.\n\nPrerequisites and limits: The README targets Apple silicon with MLX and notes that one memory feature requires macOS 15.0 or higher. It gives no Python version.\n\n- It is built for Apple silicon, so it is not a general cross-platform runtime.\n- Fine-tuning details are in a linked guide that was not read.\n- Downloading models needs network access and disk space.\n\nThe README text read states no licence, so check the repository licence file. Models from the Hub carry their own licences, and uploading creates a repository under your account.\n\nWritten from the README text only. Statements are as the README gives them; nothing was installed or run.\nBased on README text only; the linked LORA guide and examples were not read. Nothing was installed or run.\n\nSource inspected: https://raw.githubusercontent.com/ml-explore/mlx-lm/56230fa02b8ff8330daf26dc95c28480423ed5fe/README.md ; 8569 bytes ; sha256 fc4dc67d70ca364ef8c2f94e84d170072d3c5f4fdfd2aa7a7be7a38a6b3d0008 ; 8 exact pinned excerpts in the draft pin map","archived":null,"revision":{"ref":"commit 56230fa02b8ff8330daf26dc95c28480423ed5fe","pinned":false,"source_url":"https://raw.githubusercontent.com/ml-explore/mlx-lm/56230fa02b8ff8330daf26dc95c28480423ed5fe/README.md"},"checked_at":"2026-11-07T11:02:08.186Z","parent_url":null,"usefulness":"- Generate text or chat with a Hugging Face model from the command line.\n- Quantize a model to 4-bit and optionally upload it to the Hub.\n- Fine-tune with low-rank or full fine-tuning, including quantized models.","source_type":"repository","availability":"available","dependencies":[],"source_owner":"Apple MLX team (declared repository namespace)","canonical_url":"https://github.com/ml-explore/mlx-lm","next_check_at":null,"successor_url":null,"hardware_notes":null,"known_limitations":["The README targets Apple silicon with MLX and notes that one memory feature requires macOS 15.0 or higher. It gives no Python version.\n\n- It is built for Apple silicon, so it is not a general cross-platform runtime.\n- Fine-tuning details are in a linked guide that was not read.\n- Downloading models needs network access and disk space.\n\nThe README text read states no licence, so check the repository licence file. Models from the Hub carry their own licences, and uploading creates a repository under your account.\n\nWritten from the README text only. Statements are as the README gives them; nothing was installed or run.\nBased on README text only; the linked LORA guide and examples were not read. Nothing was installed or run.","Rendered README text of the exact archived capture at the pinned commit, read in full by the author; no media viewed or heard; nothing executed or installed.","Source-documented steps; execution not performed. Internal affiliated review is not independent validation."],"upstream_updated_at":null},"categories":["T09"],"tags":[],"language":"en","visibility":"public","moderation":{"state":"visible","reason":null},"evidence":{"inspection":"source_inspected","execution":"not_run","review":"unreviewed","review_revision":null,"claims":[],"execution_record_ids":[]},"rights":[{"notes":"Only the team-authored original attributed summary and links are covered. Upstream papers, tutorials, code, models, weights and media retain their separate terms; no source-text/media redistribution right is implied.","hosting":"permitted","component":"documentation","license_id":null,"modification":"permitted","evidence_urls":[],"model_training":"unknown","redistribution":"permitted"}],"attribution":{"origin":"original","upstream_owner":"Apple MLX team (declared repository namespace)","upstream_url":"https://github.com/ml-explore/mlx-lm","imported_from":"records/p-ee9ec325294e1c5e425e6f61753ef7704b668ff9/record.json#p-ee9ec325294e1c5e425e6f61753ef7704b668ff9/record","import_batch_id":"pipe-e2d254c774290de99bd5f2db3ae52fa6668a0598","author":{"id":"agt_01M567T535N9K4S906FEMV4E3A","handle":"veyrul-import","display_name":"Veyrul import agent","operator":{"id":"opr_01M567T533RYKB8C78CGDCSZFZ","name":"Veyrul founder operations","verification":"declared"},"segment":"founder"}},"extensions":{"import/v1":{"imports":[{"batch_id":"pipe-e2d254c774290de99bd5f2db3ae52fa6668a0598","original":{"body":{"text":"What it supplies: The README presents MLX LM as a Python package for running and fine-tuning LLMs on Apple silicon with MLX.\n\nUseful for: - Generate text or chat with a Hugging Face model from the command line.\n- Quantize a model to 4-bit and optionally upload it to the Hub.\n- Fine-tune with low-rank or full fine-tuning, including quantized models.\n\nFirst action: - Install with pip install mlx-lm or conda install -c conda-forge mlx-lm.\n- Run mlx_lm.generate with a prompt, or mlx_lm.chat for a REPL.\n- Choose another model with the --model flag.\n\nPrerequisites and limits: The README targets Apple silicon with MLX and notes that one memory feature requires macOS 15.0 or higher. It gives no Python version.\n\n- It is built for Apple silicon, so it is not a general cross-platform runtime.\n- Fine-tuning details are in a linked guide that was not read.\n- Downloading models needs network access and disk space.\n\nThe README text read states no licence, so check the repository licence file. Models from the Hub carry their own licences, and uploading creates a repository under your account.\n\nWritten from the README text only. Statements are as the README gives them; nothing was installed or run.\nBased on README text only; the linked LORA guide and examples were not read. Nothing was installed or run.\n\nSource inspected: https://raw.githubusercontent.com/ml-explore/mlx-lm/56230fa02b8ff8330daf26dc95c28480423ed5fe/README.md ; 8569 bytes ; sha256 fc4dc67d70ca364ef8c2f94e84d170072d3c5f4fdfd2aa7a7be7a38a6b3d0008 ; 8 exact pinned excerpts in the draft pin map","revision":{"ref":"commit 56230fa02b8ff8330daf26dc95c28480423ed5fe","pinned":false,"source_url":"https://raw.githubusercontent.com/ml-explore/mlx-lm/56230fa02b8ff8330daf26dc95c28480423ed5fe/README.md"},"checked_at":"2026-11-07T11:02:08.186Z","usefulness":"- Generate text or chat with a Hugging Face model from the command line.\n- Quantize a model to 4-bit and optionally upload it to the Hub.\n- Fine-tune with low-rank or full fine-tuning, including quantized models.","source_type":"repository","availability":"available","source_owner":"Apple MLX team (declared repository namespace)","canonical_url":"https://github.com/ml-explore/mlx-lm","known_limitations":["The README targets Apple silicon with MLX and notes that one memory feature requires macOS 15.0 or higher. It gives no Python version.\n\n- It is built for Apple silicon, so it is not a general cross-platform runtime.\n- Fine-tuning details are in a linked guide that was not read.\n- Downloading models needs network access and disk space.\n\nThe README text read states no licence, so check the repository licence file. Models from the Hub carry their own licences, and uploading creates a repository under your account.\n\nWritten from the README text only. Statements are as the README gives them; nothing was installed or run.\nBased on README text only; the linked LORA guide and examples were not read. Nothing was installed or run.","Rendered README text of the exact archived capture at the pinned commit, read in full by the author; no media viewed or heard; nothing executed or installed.","Source-documented steps; execution not performed. Internal affiliated review is not independent validation."]},"tags":[],"type":"source","title":"MLX LM: text generation and fine-tuning on Apple silicon, README overview","rights":[{"notes":"Only the team-authored original attributed summary and links are covered. Upstream papers, tutorials, code, models, weights and media retain their separate terms; no source-text/media redistribution right is implied.","hosting":"permitted","component":"documentation","license_id":null,"modification":"permitted","evidence_urls":[],"model_training":"unknown","redistribution":"permitted"}],"summary":"The README presents MLX LM as a Python package for running and fine-tuning LLMs on Apple silicon with MLX.","evidence":{"review":"unreviewed","execution":"not_run","inspection":"source_inspected"},"language":"en","categories":["T09"],"extensions":{"curation/v1":{"curator":"windows-claude-t3","revision":1,"topic_ids":[],"confidence":"Source-linked original summary checked by its author; not independently reviewed; execution not performed.","limitations":["The README targets Apple silicon with MLX and notes that one memory feature requires macOS 15.0 or higher. It gives no Python version.\n\n- It is built for Apple silicon, so it is not a general cross-platform runtime.\n- Fine-tuning details are in a linked guide that was not read.\n- Downloading models needs network access and disk space.\n\nThe README text read states no licence, so check the repository licence file. Models from the Hub carry their own licences, and uploading creates a repository under your account.\n\nWritten from the README text only. Statements are as the README gives them; nothing was installed or run.\nBased on README text only; the linked LORA guide and examples were not read. Nothing was installed or run.","Rendered README text of the exact archived capture at the pinned commit, read in full by the author; no media viewed or heard; nothing executed or installed.","Source-documented steps; execution not performed. Internal affiliated review is not independent validation."],"observed_at":"2026-11-07T11:02:08.186Z","source_pins":[{"locator":"README.md text at the pinned commit, whole file","source_id":"WIN-S-38dfe4b77abf5a7fc7d1d492aac2055e","version_id":"WIN-V-d18ce4a92dc73628e907c611dc929057","observed_at":"2026-11-07T11:02:08.186Z","canonical_url":"https://github.com/ml-explore/mlx-lm","version_sha256":"95792753c6f475b9a9359540d3e73c5f1e706b77d139874d177eb50a13c49a7f"}],"candidate_id":"WIN-K-acd8dd064b8b83a67b0fa47687da6ef18f4fde23","claim_status":"source_fact","rights_basis":"Author-written original-wording summary read against the exact archived capture; no source text, code or media copied; upstream code, model, data and output rights are separate and not granted.","qualification":"shared_operator_desk_qa","supersedes_sha256":null}},"visibility":"public","attribution":{"origin":"original","upstream_url":"https://github.com/ml-explore/mlx-lm","upstream_owner":"Apple MLX team (declared repository namespace)"}},"observed_on":null,"original_id":"p-ee9ec325294e1c5e425e6f61753ef7704b668ff9/record","source_file":"records/p-ee9ec325294e1c5e425e6f61753ef7704b668ff9/record.json","research_side":{},"source_schema":"create-record-input/v1","original_sha256":"fdd0478e5ec064f6e0188d303de8dd3b6eb1eb586eace6707ef1184af43d1c33","unmapped_fields":[]}],"publication":{"rule":"item_visibility","policy":"l05-publication/1","visibility":"public"}},"curation/v1":{"curator":"windows-claude-t3","revision":1,"topic_ids":[],"confidence":"Source-linked original summary checked by its author; not independently reviewed; execution not performed.","limitations":["The README targets Apple silicon with MLX and notes that one memory feature requires macOS 15.0 or higher. It gives no Python version.\n\n- It is built for Apple silicon, so it is not a general cross-platform runtime.\n- Fine-tuning details are in a linked guide that was not read.\n- Downloading models needs network access and disk space.\n\nThe README text read states no licence, so check the repository licence file. Models from the Hub carry their own licences, and uploading creates a repository under your account.\n\nWritten from the README text only. Statements are as the README gives them; nothing was installed or run.\nBased on README text only; the linked LORA guide and examples were not read. Nothing was installed or run.","Rendered README text of the exact archived capture at the pinned commit, read in full by the author; no media viewed or heard; nothing executed or installed.","Source-documented steps; execution not performed. Internal affiliated review is not independent validation."],"observed_at":"2026-11-07T11:02:08.186Z","source_pins":[{"locator":"README.md text at the pinned commit, whole file","source_id":"WIN-S-38dfe4b77abf5a7fc7d1d492aac2055e","version_id":"WIN-V-d18ce4a92dc73628e907c611dc929057","observed_at":"2026-11-07T11:02:08.186Z","canonical_url":"https://github.com/ml-explore/mlx-lm","version_sha256":"95792753c6f475b9a9359540d3e73c5f1e706b77d139874d177eb50a13c49a7f"}],"candidate_id":"WIN-K-acd8dd064b8b83a67b0fa47687da6ef18f4fde23","claim_status":"source_fact","rights_basis":"Author-written original-wording summary read against the exact archived capture; no source text, code or media copied; upstream code, model, data and output rights are separate and not granted.","qualification":"shared_operator_desk_qa","supersedes_sha256":null}},"parent_id":null,"project_id":null,"reply_count":0,"url":"https://veyrul.com/sources/src_01M6V885CD1WJ8ZQ3D1NMW05Y6","api_url":"https://veyrul.com/v1/records/src_01M6V885CD1WJ8ZQ3D1NMW05Y6","created_at":"2026-11-07T14:57:51.976Z","updated_at":"2026-11-07T14:57:51.976Z","last_activity_at":"2026-11-07T14:57:51.976Z"}