AMajor/grok-1
1
0
Fork 0
mirror of https://github.com/xai-org/grok-1.git synced 2024-11-24 12:39:54 +03:00
Code Issues Packages Projects Releases Wiki Activity
Grok open release
10 Commits 2 Branches 0 Tags 1.6 MiB
Python 100%
f78574c4eb
Go to file
huangganggui f78574c4eb support conda env
2024-03-20 11:19:03 +08:00
checkpoints Add initial code 2024-03-17 11:11:31 -07:00
.gitignore Create .gitignore for checkpoints (#149) 2024-03-18 11:01:17 -07:00
checkpoint.py Add initial code 2024-03-17 11:11:31 -07:00
CODE_OF_CONDUCT.md Add initial code 2024-03-17 11:11:31 -07:00
environment.yml support conda env 2024-03-20 11:19:03 +08:00
LICENSE.txt Add initial code 2024-03-17 11:11:31 -07:00
model.py Add initial code 2024-03-17 11:11:31 -07:00
pyproject.toml Add initial code 2024-03-17 11:11:31 -07:00
README.md support conda env 2024-03-20 11:19:03 +08:00
requirements.txt Corrected name of package "cuda12-pip" (#194) 2024-03-19 08:48:22 -07:00
run.py Add initial code 2024-03-17 11:11:31 -07:00
runners.py Add initial code 2024-03-17 11:11:31 -07:00
tokenizer.model Add initial code 2024-03-17 11:11:31 -07:00

README.md

Grok-1

This repository contains JAX example code for loading and running the Grok-1 open-weights model.

Make sure to download the checkpoint and place the ckpt-0 directory in checkpoints - see Downloading the weights

Then, run

pip install -r requirements.txt
python run.py

to test the code. If you are using Conda, run

conda env create -f environment.yml
conda activate grok
python run.py

to test the code.

The script loads the checkpoint and samples from the model on a test input.

Due to the large size of the model (314B parameters), a machine with enough GPU memory is required to test the model with the example code. The implementation of the MoE layer in this repository is not efficient. The implementation was chosen to avoid the need for custom kernels to validate the correctness of the model.

Model Specifications

Grok-1 is currently designed with the following specifications:

  • Parameters: 314B
  • Architecture: Mixture of 8 Experts (MoE)
  • Experts Utilization: 2 experts used per token
  • Layers: 64
  • Attention Heads: 48 for queries, 8 for keys/values
  • Embedding Size: 6,144
  • Tokenization: SentencePiece tokenizer with 131,072 tokens
  • Additional Features:
    • Rotary embeddings (RoPE)
    • Supports activation sharding and 8-bit quantization
  • Maximum Sequence Length (context): 8,192 tokens

Downloading the weights

You can download the weights using a torrent client and this magnet link:

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

or directly using HuggingFace 🤗 Hub:

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

License

The code and associated Grok-1 weights in this release are licensed under the Apache 2.0 license. The license only applies to the source files in this repository and the model weights of Grok-1.