grok-1Grok open release
grok-1基本信息
grok-1介绍
这个仓库包含了加载和运行 Grok-1 开源权重模型的 JAX 示例代码。
请确保下载检查点并将 ckpt-0 目录放在 checkpoints 中 - 参见下载权重
然后运行
pip install -r requirements.txt
python run.py
来测试代码。
该脚本加载检查点并对测试输入从模型中进行采样。
由于模型的大小(314B 参数),需要具有足够 GPU 内存的机器才能使用示例代码测试模型。此存储库中 MoE 层的实现效率不高。选择这种实现是为了避免需要自定义内核来验证模型的正确性。
模型规格
Grok-1 目前的设计规格如下:
- 参数: 314B
- 架构: 8 专家混合 (MoE)
- 专家利用率: 每个令牌使用 2 个专家
- 层数: 64
- 注意力头: 查询 48 个,键/值 8 个
- 嵌入大小: 6,144
- 分词: 131,072 个令牌的 SentencePiece 分词器
- 额外特性:
- 旋转嵌入 (RoPE)
- 支持激活分片和 8 位量化
- 最大序列长度(上下文): 8,192 个令牌
下载权重
你可以使用种子客户端和这个磁力链接下载权重:
magnet:?xt=urn:btih:5f96d43576e3d386c9ba65b883210a393b68210e&tr=https%3A%2F%2Facademictorrents.com%2Fannounce.php&tr=udp%3A%2F%2Ftracker.coppersurfer.tk%3A6969&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce
或直接从 HuggingFace 🤗 Hub 下载:
git clone https://github.com/xai-org/grok-1.git && cd grok-1
pip install huggingface_hub[hf_transfer]
huggingface-cli download xai-org/grok-1 --repo-type model --include ckpt-0/* --local-dir checkpoints --local-dir-use-symlinks False
grok-1相关项目
No-code LLM Platform to launch APIs and ETL Pipelines to structure unstructured documents
Dify is an open-source LLM app development platform. Dify's intuitive interface combines AI workflow, RAG pipeline, agent capabilities, model management, observability features and more, letting you quickly go from prototype to production.
A programming framework for agentic AI 🤖
fabric is an open-source framework for augmenting humans using AI. It provides a modular framework for solving specific problems using a crowdsourced set of AI prompts that can be used anywhere.
:robot: The free, Open Source alternative to OpenAI, Claude and others. Self-hosted and local-first. Drop-in replacement for OpenAI, running on consumer-grade hardware. No GPU required. Runs gguf, transformers, diffusers and many more models architectures. Features: Generate Text, Audio, Video, Images, Voice Cloning, Distributed inference
