---
type: "post"
id: "thinking-machines-released-dimuon-a-new-in-house-16cede"
slug: "thinking-machines-released-dimuon-a-new-in-house-16cede"
title: "Thinking Machines released Dimuon , a new in-house optimizer for LoRA fine-tuning on Tinker."
description: "**Thesis:** Dimuon’s significance is a new documented optimizer choice inside Tinker’s LoRA fine-tuning workflow, not evidence of a proven training-quality or efficiency gain."
retrieval_nugget: "**Thesis:** Dimuon’s significance is a new documented optimizer choice inside Tinker’s LoRA fine-tuning workflow, not evidence of a proven training-quality or efficiency gain."
published_at: "2026-10-15"
updated_at: "2026-10-15"
record_date: "2026-10-15"
date_kind: "published_at"
topics: ["models"]
source_urls: ["https://tinker-docs.thinkingmachines.ai/tinker/optimizers"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:thinking-machines-released-dimuon-a-new-in-house-16cede"
status: "published"
visuals: []
editorial_provenance: {"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:843e3087eb0f00ad5ef11f19e7317d27684ccd60a4e8fad21fb84276dfbc760f","reviewed_at":"2026-10-15T07:01:32.401Z","source_evidence_count":1,"verified_claim_count":2,"site_analysis_schema_version":"newruntime-site-analysis-v1","site_object_kind":"field_note","observed_fact_count":2,"implication_count":1,"watch_condition_count":1,"related_record_count":0}
analysis: {"schema_version":"newruntime-site-analysis-v1","object_kind":"field_note","thesis":"Dimuon’s significance is a new documented optimizer choice inside Tinker’s LoRA fine-tuning workflow, not evidence of a proven training-quality or efficiency gain.","observed_facts":[{"text":"The Tinker optimizer documentation has a dedicated Dimuon section and a section on when to use it.","source_urls":["https://tinker-docs.thinkingmachines.ai/tinker/optimizers"]},{"text":"The documentation includes learning-rate guidance, LoRA-rank invariance, batch-size dependence, and switching optimizers midflight.","source_urls":["https://tinker-docs.thinkingmachines.ai/tinker/optimizers"]}],"mechanism":"The documentation presents Dimuon as a selectable optimizer in Tinker’s LoRA training workflow and provides operating guidance for tuning it.","why_now":"The public documentation now gives Tinker users a named option and a set of operating considerations.","implications":["For builders already using Tinker, Dimuon is worth testing against their existing optimizer on a controlled workload before changing production defaults."],"evidence_boundary":"The supplied public source describes the optimizer and its use but does not provide comparative benchmark numbers, independent validation, or evidence that Dimuon outperforms AdamW across workloads. The page alone also does not establish availability outside Tinker.","watch_conditions":["Watch for published controlled comparisons that report task quality, training stability, and cost against AdamW across LoRA ranks and batch sizes."],"related_records":[],"new_branch_reason":"No prior New Runtime record was supplied for Dimuon; this creates a distinct record about an optimizer option documented for Tinker LoRA fine-tuning, rather than assuming a connection to unrelated training stories."}
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/thinking-machines-released-dimuon-a-new-in-house-16cede/","markdown":"https://newruntime.com/posts/thinking-machines-released-dimuon-a-new-in-house-16cede.md","json":"https://newruntime.com/posts/thinking-machines-released-dimuon-a-new-in-house-16cede.json"}
---

# Thinking Machines released Dimuon , a new in-house optimizer for LoRA fine-tuning on Tinker.

## Retrieval answer

**Thesis:** Dimuon’s significance is a new documented optimizer choice inside Tinker’s LoRA fine-tuning workflow, not evidence of a proven training-quality or efficiency gain.

**Thesis:** Dimuon’s significance is a new documented optimizer choice inside Tinker’s LoRA fine-tuning workflow, not evidence of a proven training-quality or efficiency gain.

**Mechanism:** The documentation presents Dimuon as a selectable optimizer in Tinker’s LoRA training workflow and provides operating guidance for tuning it. The Tinker optimizer documentation presents Dimuon alongside AdamW and provides guidance on when to use it, learning-rate selection, LoRA-rank invariance, batch-size dependence, and switching optimizers midflight. That makes the optimizer a selectable part of the training harness: teams can change an important training component without treating it as a separate model release.

**Why now:** The public documentation now gives Tinker users a named option and a set of operating considerations. The observed facts are that the page has a dedicated Dimuon section and explicitly discusses learning-rate guidance and batch-size dependence. It also covers rank invariance and switching optimizers during training.

**First implication:** For builders already using Tinker, Dimuon is worth testing against their existing optimizer on a controlled workload before changing production defaults. This is a workflow-level opportunity, not a result that can be generalized from the documentation alone.

**Evidence boundary:** The supplied public source describes the optimizer and its use but does not provide comparative benchmark numbers, independent validation, or evidence that Dimuon outperforms AdamW across workloads. The page alone also does not establish availability outside Tinker.

**First watch condition:** Watch for published controlled comparisons that report task quality, training stability, and cost against AdamW across LoRA ranks and batch sizes.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Models](https://newruntime.com/topics/models.md): Explore the models topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)