feat: add MiniMax as first-class LLM provider#144
Closed
octo-patch wants to merge 629 commits into
Closed
Conversation
Feat/profilev2 See merge request npc-work/aic/ai/evermemos-opensource!61
Feat/update demo See merge request npc-work/aic/ai/evermemos-opensource!62
…migrate_online_llm_clustering
…ne_llm_clustering Feat/migrate online llm clustering
…r lists Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…ind-AI#167) * docs: update README with project overview and structure changes - Uncomment the HuggingFace dataset badge link - Replace the detailed table of contents with a simplified project overview - Update project structure to include usecases directory and rename everos to EverCore - Refresh method descriptions with updated banner GIF links and documentation paths - Remove outdated benchmark results section to focus on current project state - Update use case table to reflect EverCore naming change * docs: update README table headers and remove outdated sections - Change "Use Case" column header to "Component" for clarity in the components table - Rename "everos Quick Start" section to "EverCore Quick Start" to match component name - Remove outdated Demo, Documentation, and GitHub Codespaces sections by commenting them out - Simplify contributing section by removing discussion mention and contribution guidelines - Move license section to bottom and adjust divider placement - Update benchmark image path to correct location
85a0b02 Update website code 8f2f6e2 Note results are partial, more coming soon git-subtree-dir: benchmarks/EvoAgentBench git-subtree-split: 85a0b02dde4ee89b22b60e873c2016a2bea644f0
* docs: update README with EverOS announcement * docs: update README to reflect project structure unification Replace outdated Memory Sparse Attention reference with current project name and structure. The EverCore, HyperMem, EverMemBench, EvoAgentBench, and usecases components have been unified into the EverOS repository.
Major restructure: - Reorganize into methods/ and benchmarks/ directory structure - Add EverMemBench and EvoAgentBench via git subtree - Add HyperMem source code under methods/HyperMem - Fix Citation with correct arXiv metadata and full author lists - Update README with new branding and project structure Old main preserved as branch 'EverMemOS'. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
6340655 Merge branch 'feat/everos-eval-fixes' into 'release/opensource-260410' df4d174 Add website link badge 494d47b Update results figure to res-0414 ce43079 Remove website (moved to separate repo) 9f18345 increase docker load and tmux setup timeouts for large swebench images a9919e2 fix swebench container name collision by using job_dir tag and increase docker client timeout 9cce83b adapt extract_skills for new EverOS: infinite wait, remove flush_clustering, add ghost filter and error handling git-subtree-dir: benchmarks/EvoAgentBench git-subtree-split: 6340655da556ab8f1d94a6d90af9ae965f516df9
…-AI#175) Update README.md and OVERVIEW.md to reference images via GitHub user attachment URLs instead of local paths. This ensures images display correctly when the documentation is viewed outside the local repository context, such as on the GitHub web interface.
- Add VS Code devcontainer setup with Docker Compose for local development - Include infrastructure service wait scripts and environment configuration - Add pre-commit hooks for code quality and conventional commit enforcement - Set up VS Code launch configurations for debugging - Remove old .gitlab/CODEOWNERS file as project structure changes
Update EverMemOS: optimize search perf, improve skill search
chore: add devcontainer configuration and development tooling
- Add MINIMAX_API_KEY and MINIMAX_BASE_URL to env.template with documentation and optional MINIMAX_WHITE_LIST for model restriction - Fix OpenAIProvider._build_request_data to exclude null fields (provider, response_format) from request payload, enabling compatibility with MiniMax and other strict OpenAI-compatible APIs - Add unit tests covering MiniMax instantiation, request building, generate calls, env config resolution, and whitelist enforcement
octo-patch
force-pushed
the
feature/add-minimax-provider
branch
from
020bac1 to
8b22348
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
12 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Add MiniMax as a dedicated LLM backend adapter alongside OpenAI, Anthropic, and Gemini. MiniMax provides an OpenAI-compatible API with powerful models:
Changes
src/core/component/llm/llm_adapter/minimax_adapter.py—MiniMaxAdapterextendingLLMBackendAdaptervia OpenAI SDK, with:<think>...</think>tag stripping for reasoning model responsesMINIMAX_API_KEYenvironment variable auto-detectionOpenAICompatibleClientto routeprovider: minimaxtoMiniMaxAdapterminimaxentry inllm_backends.yamlwith 4 modelsMINIMAX_API_KEYsection inenv.templateTest plan
minimaxbackend is properly configured inllm_backends.yamlOpenAICompatibleClientroutes toMiniMaxAdapterforprovider: minimaxMINIMAX_API_KEYand selectminimaxasdefault_backendinllm_backends.yaml, run the application7 files changed, 665 additions