happyz
/
gemma.cpp
mirror of https://github.com/google/gemma.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Document Gemma 2 model names 

PiperOrigin-RevId: 659858832
This commit is contained in:
Jan Wassenberg 2024-08-06 01:43:49 -07:00 committed by Copybara-Service
parent 5e433e774a
commit 4154f5a910
1 changed files with 17 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
README.md
View File

@ -77,18 +77,21 @@ winget install --id Microsoft.VisualStudio.2022.BuildTools --force --override "-
### Step 1: Obtain model weights and tokenizer from Kaggle or Hugging Face Hub ### Step 1: Obtain model weights and tokenizer from Kaggle or Hugging Face Hub
Visit [the Gemma model page on Visit the
Kaggle](https://www.kaggle.com/models/google/gemma/frameworks/gemmaCpp) and select `Model Variations [Kaggle page for Gemma](https://www.kaggle.com/models/google/gemma/frameworks/gemmaCpp),
|> Gemma C++`. On this tab, the `Variation` dropdown includes the options below. or [Gemma-2](https://www.kaggle.com/models/google/gemma-2/gemmaCpp), and select
Note bfloat16 weights are higher fidelity, while 8-bit switched floating point `Model Variations |> Gemma C++`.
weights enable faster inference. In general, we recommend starting with the
`-sfp` checkpoints.
Alternatively, visit the [gemma.cpp](https://huggingface.co/models?other=gemma.cpp) On this tab, the `Variation` dropdown includes the options below. Note bfloat16
models on the Hugging Face Hub. First go the the model repository of the model of interest weights are higher fidelity, while 8-bit switched floating point weights enable
(see recommendations below). Then, click the `Files and versions` tab and download the faster inference. In general, we recommend starting with the `-sfp` checkpoints.
model and tokenizer files. For programmatic downloading, if you have `huggingface_hub`
installed, you can also download by running: Alternatively, visit the
[gemma.cpp](https://huggingface.co/models?other=gemma.cpp) models on the Hugging
Face Hub. First go the the model repository of the model of interest (see
recommendations below). Then, click the `Files and versions` tab and download
the model and tokenizer files. For programmatic downloading, if you have
`huggingface_hub` installed, you can also download by running:
``` ```
huggingface-cli login # Just the first time huggingface-cli login # Just the first time
@ -117,6 +120,9 @@ huggingface-cli download google/gemma-2b-sfp-cpp --local-dir build/
> **Important**: We strongly recommend starting off with the `2b-it-sfp` model to > **Important**: We strongly recommend starting off with the `2b-it-sfp` model to
> get up and running. > get up and running.
Gemma 2 models are named `gemma2-2b-it` for 2B and `9b-it` or `27b-it`. See the
`kModelFlags` definition in `common.cc`.
### Step 2: Extract Files ### Step 2: Extract Files
If you downloaded the models from Hugging Face, skip to step 3. If you downloaded the models from Hugging Face, skip to step 3.