Odysseyzju-vipa/Odyssey
Provides a training and fine-tuning solution for LLM agents in the open-world Minecraft environment,…
No ratings yet
Rate this skill
Stars409
Forks30
Watchers409
Open Issues3
InstallOne-line install
UnconfirmedWhether this repository follows the Agent Skills spec (that is, whether it contains a SKILL.md) has not been verified yet. Until that is settled, clone it and use what you need:
git clone https://github.com/zju-vipa/Odyssey.git
OverviewOverview
Key Features
Provides a training and fine-tuning solution for LLM agents in the open-world Minecraft environment, enabling models to complete long-horizon, multi-step embodied tasks. Key capabilities include: 1. Agent fine-tuning: fine-tunes large language models on Minecraft interaction trajectories so the model learns to decompose and execute long-horizon tasks in the game environment; 2. Embodied task execution: lets the agent perceive environment state in an open world and output executable action sequences to accomplish multi-step goals such as gathering, crafting and exploration; 3. Capability evaluation: provides Minecraft agent tasks and an evaluation pipeline for comparing how different models perform on long-horizon planning and execution.
Use Cases
training LLM agents that play Minecraft, validating an agent's long-horizon planning ability, comparing the effects of different fine-tuning strategies, building open-world embodied agent benchmarks.
Target Roles
AI / Algorithm Engineer, Backend Developer.
TopicsTopics
RatingUser ratings
Nobody has rated this skill yet. Yours would be the first on the site — your vote becomes the initial average.
RepositoryRepository
| Skill ID | 1999 |
|---|---|
| Data source | GitHub |
| Repository | zju-vipa/Odyssey |
| Category | AI & Agent Skills / Agent Training & Fine-Tuning |
| Primary language | Python |
| License | MIT License |
| Default branch | master |
| GitHub repository ID | 814366957 |
| Node ID | R_kgDOMIpA7Q |
| Rank | #1495 (position within the source ranking) |
| Site-wide popularity rank | #1713 (computed live from Stars, across all 5833 entries) |
GrowthStar / Fork growth
Still collecting
First record taken on 2026-09-26. Two days of data are needed before a trend line can be drawn.
TimelineTimeline
Repository created2024-06-12 21:50:582024-06-12
Data last updated2026-10-26 12:15:3423 days ago
Last commit2025-10-22 07:37:452025-10-22
CommitLatest commit
Committed by liushunyu. Update Category: Documentation Updates. Commit details: Update README.md. This update refines the documentation for easier onboarding.