mirror of
https://github.com/nod-ai/AMD-SHARK-Studio.git
synced 2026-04-03 03:00:17 -04:00
Add Readme file for the bloom model
This commit is contained in:
36
tank/pytorch/bloom/README.md
Normal file
36
tank/pytorch/bloom/README.md
Normal file
@@ -0,0 +1,36 @@
|
||||
# Bloom model
|
||||
|
||||
## Installation
|
||||
|
||||
<details>
|
||||
<summary>Installation (Linux)</summary>
|
||||
|
||||
### Activate shark.venv Virtual Environment
|
||||
|
||||
```shell
|
||||
source shark.venv/bin/activate
|
||||
|
||||
# Some older pip installs may not be able to handle the recent PyTorch deps
|
||||
python -m pip install --upgrade pip
|
||||
```
|
||||
|
||||
### Install dependencies
|
||||
|
||||
```shell
|
||||
pip install transformers==4.21.2
|
||||
```
|
||||
Use this branch of Torch-MLIR for running the model: https://github.com/vivekkhandelwal1/torch-mlir/tree/bloom-ops
|
||||
|
||||
|
||||
### Run bloom model
|
||||
|
||||
```shell
|
||||
python bloom_model.py
|
||||
```
|
||||
|
||||
The runtime device, model config, and text prompt can be specified with `--device <device string>`, `--config <config string>`, `--prompt <prompt string>` respectively.
|
||||
|
||||
To run the complete 176B params bloom model, run the following command:
|
||||
```shell
|
||||
python bloom_model.py --config "bloom"
|
||||
```
|
||||
Reference in New Issue
Block a user