From fbcde1de382a6220f0c5ff9fa192351298cbfbf9 Mon Sep 17 00:00:00 2001
From: lxning <23464292+lxning@users.noreply.github.com>
Date: Wed, 10 May 2023 22:33:16 +0000
Subject: [PATCH] remove redundent Download_models.py (#2331)

---
 docs/large_model_inference.md                 |  2 +-
 examples/large_models/deepspeed/opt/Readme.md |  2 +-
 .../large_models/utils/Download_models.py     | 51 -------------------
 3 files changed, 2 insertions(+), 53 deletions(-)
 delete mode 100644 examples/large_models/utils/Download_models.py

diff --git a/docs/large_model_inference.md b/docs/large_model_inference.md
index 04f4dd4601..3d9b6dd815 100644
--- a/docs/large_model_inference.md
+++ b/docs/large_model_inference.md
@@ -181,7 +181,7 @@ torch-model-archiver --model-name bloom --version 1.0 --handler deepspeed_handle
 * Pre-install the model parallel library such as Deepspeed on the container or host.
 * Pre-download the model checkpoints. For example, if using HuggingFace pretrained model can be pre-downloaded via [Download_model.py](https://github.com/pytorch/serve/blob/75f66dc557b3b67a3ab56536a37d7aa21582cc04/examples/large_models/deepspeed/opt/Readme.md?plain=1#L7)
   * Set environment variable [HUGGINGFACE_HUB_CACHE](https://huggingface.co/docs/huggingface_hub/guides/manage-cache#understand-caching) and [TRANSFORMERS_CACHE](https://huggingface.co/transformers/v4.0.1/installation.html#caching-models)
-  * Download model to the HuggingFace cache dir via tool [Download_model.py](https://github.com/pytorch/serve/blob/75f66dc557b3b67a3ab56536a37d7aa21582cc04/examples/large_models/deepspeed/opt/Readme.md?plain=1#L7)
+  * Download model to the HuggingFace cache dir via tool [Download_model.py](https://github.com/pytorch/serve/blob/4fe5273cd6f98fb5abc570f802b402ac32ecd105/examples/large_models/Huggingface_pippy/Readme.md?plain=1#L20)
 
 #### Tune "[responseTimeout](https://github.com/pytorch/serve/blob/5ee02e4f050c9b349025d87405b246e970ee710b/docs/configuration.md?plain=1#L216)" (see [model config YAML file](https://github.com/pytorch/serve/blob/5ee02e4f050c9b349025d87405b246e970ee710b/model-archiver/README.md?plain=1#L164)) if high model loading or inference latency causes response timeout.
 
diff --git a/examples/large_models/deepspeed/opt/Readme.md b/examples/large_models/deepspeed/opt/Readme.md
index 0ee6b6ddd8..cb8c06a976 100644
--- a/examples/large_models/deepspeed/opt/Readme.md
+++ b/examples/large_models/deepspeed/opt/Readme.md
@@ -7,7 +7,7 @@ This document briefs on serving large HG models on multiple GPUs using deepspeed
 ### Step 1: Download model
 
 ```bash
-python ../../utils/Download_models.py --model_path model --model_name facebook/opt-350m --revision main
+python ../../utils/Download_model.py --model_path model --model_name facebook/opt-350m --revision main
 ```
 
 The script prints the path where the model is downloaded as below.
diff --git a/examples/large_models/utils/Download_models.py b/examples/large_models/utils/Download_models.py
deleted file mode 100644
index 5ad5db3c5f..0000000000
--- a/examples/large_models/utils/Download_models.py
+++ /dev/null
@@ -1,51 +0,0 @@
-import argparse
-import os
-
-from huggingface_hub import HfApi, snapshot_download
-
-
-def dir_path(path_str):
-    if os.path.isdir(path_str):
-        return path_str
-    elif input(f"{path_str} does not exist, create directory? [y/n]").lower() == "y":
-        os.makedirs(path_str)
-        return path_str
-    else:
-        raise NotADirectoryError(path_str)
-
-
-class HFModelNotFoundError(Exception):
-    def __init__(self, model_str):
-        super().__init__(f"HuggingFace model not found: '{model_str}'")
-
-
-def hf_model(model_str):
-    api = HfApi()
-    models = [m.modelId for m in api.list_models()]
-    if model_str in models:
-        return model_str
-    else:
-        raise HFModelNotFoundError(model_str)
-
-
-parser = argparse.ArgumentParser()
-parser.add_argument(
-    "--model_path",
-    "-o",
-    type=dir_path,
-    default="model",
-    help="Output directory for downloaded model files",
-)
-parser.add_argument(
-    "--model_name", "-m", type=hf_model, required=True, help="HuggingFace model name"
-)
-parser.add_argument("--revision", "-r", type=str, default="main", help="Revision")
-args = parser.parse_args()
-
-snapshot_path = snapshot_download(
-    repo_id=args.model_name,
-    revision=args.revision,
-    cache_dir=args.model_path,
-    use_auth_token=False,
-)
-print(f"Files for '{args.model_name}' is downloaded to '{snapshot_path}'")