Upload externalmod.py

externalmod.py

@@ -9,7 +9,7 @@ import re
 import tempfile
 import warnings
 from pathlib import Path
-from typing import TYPE_CHECKING, Callable
+from typing import TYPE_CHECKING, Callable, Literal
 
 import httpx
 import huggingface_hub
@@ -33,11 +33,15 @@ if TYPE_CHECKING:
     from gradio.interface import Interface
 
 
+HF_TOKEN = os.environ.get("HF_TOKEN") if os.environ.get("HF_TOKEN") else None # If private or gated models aren't used, ENV setting is unnecessary.
+server_timeout = 600
+
+
 @document()
 def load(
     name: str,
     src: str | None = None,
-    hf_token: str | None = None,
+    hf_token: str | Literal[False] | None = None,
     alias: str | None = None,
     **kwargs,
 ) -> Blocks:
@@ -48,7 +52,7 @@ def load(
     Parameters:
         name: the name of the model (e.g. "gpt2" or "facebook/bart-base") or space (e.g. "flax-community/spanish-gpt2"), can include the `src` as prefix (e.g. "models/facebook/bart-base")
         src: the source of the model: `models` or `spaces` (or leave empty if source is provided as a prefix in `name`)
-        hf_token: optional access token for loading private Hugging Face Hub models or spaces. Find your token here: https://huggingface.co/settings/tokens.  Warning: only provide this if you are loading a trusted private Space as it can be read by the Space you are loading.
+        hf_token: optional access token for loading private Hugging Face Hub models or spaces. Will default to the locally saved token if not provided. Pass `token=False` if you don't want to send your token to the server. Find your token here: https://huggingface.co/settings/tokens.  Warning: only provide a token if you are loading a trusted private Space as it can be read by the Space you are loading.
         alias: optional string used as the name of the loaded model instead of the default name (only applies if loading a Space running Gradio 2.x)
     Returns:
         a Gradio Blocks object for the given model
@@ -65,7 +69,7 @@ def load(
 def load_blocks_from_repo(
     name: str,
     src: str | None = None,
-    hf_token: str | None = None,
+    hf_token: str | Literal[False] | None = None,
     alias: str | None = None,
     **kwargs,
 ) -> Blocks:
@@ -89,7 +93,7 @@ def load_blocks_from_repo(
     if src.lower() not in factory_methods:
         raise ValueError(f"parameter: src must be one of {factory_methods.keys()}")
 
-    if hf_token is not None:
+    if hf_token is not None and hf_token is not False:
         if Context.hf_token is not None and Context.hf_token != hf_token:
             warnings.warn(
                 """You are loading a model/Space with a different access token than the one you used to load a previous model/Space. This is not recommended, as it may cause unexpected behavior."""
@@ -100,12 +104,16 @@ def load_blocks_from_repo(
     return blocks
 
 
-def from_model(model_name: str, hf_token: str | None, alias: str | None, **kwargs):
+def from_model(
+    model_name: str, hf_token: str | Literal[False] | None, alias: str | None, **kwargs
+):
     model_url = f"https://huggingface.co/{model_name}"
     api_url = f"https://api-inference.huggingface.co/models/{model_name}"
     print(f"Fetching model from: {model_url}")
 
-    headers = {"Authorization": f"Bearer {hf_token}"} if hf_token is not None else {}
+    headers = (
+        {} if hf_token in [False, None] else {"Authorization": f"Bearer {hf_token}"}
+    )
     response = httpx.request("GET", api_url, headers=headers)
     if response.status_code != 200:
         raise ModelNotFoundError(
@@ -115,7 +123,7 @@ def from_model(model_name: str, hf_token: str | None, alias: str | None, **kwarg
 
     headers["X-Wait-For-Model"] = "true"
     client = huggingface_hub.InferenceClient(
-        model=model_name, headers=headers, token=hf_token
+        model=model_name, headers=headers, token=hf_token, timeout=server_timeout,
     )
 
     # For tasks that are not yet supported by the InferenceClient
@@ -365,10 +373,14 @@ def from_model(model_name: str, hf_token: str | None, alias: str | None, **kwarg
     else:
         raise ValueError(f"Unsupported pipeline type: {p}")
 
-    def query_huggingface_inference_endpoints(*data):
+    def query_huggingface_inference_endpoints(*data, **kwargs):
         if preprocess is not None:
             data = preprocess(*data)
-        data = fn(*data)  # type: ignore
+        try:
+            data = fn(*data, **kwargs)  # type: ignore
+        except huggingface_hub.utils.HfHubHTTPError as e:
+            if "429" in str(e):
+                raise TooManyRequestsError() from e
         if postprocess is not None:
             data = postprocess(data)  # type: ignore
         return data
@@ -380,7 +392,7 @@ def from_model(model_name: str, hf_token: str | None, alias: str | None, **kwarg
         "inputs": inputs,
         "outputs": outputs,
         "title": model_name,
-    #    "examples": examples,
+        #"examples": examples,
     }
 
     kwargs = dict(interface_info, **kwargs)
@@ -391,19 +403,12 @@ def from_model(model_name: str, hf_token: str | None, alias: str | None, **kwarg
 def from_spaces(
     space_name: str, hf_token: str | None, alias: str | None, **kwargs
 ) -> Blocks:
-    client = Client(
-        space_name,
-        hf_token=hf_token,
-        download_files=False,
-        _skip_components=False,
-    )
-
     space_url = f"https://huggingface.co/spaces/{space_name}"
 
     print(f"Fetching Space from: {space_url}")
 
     headers = {}
-    if hf_token is not None:
+    if hf_token not in [False, None]:
         headers["Authorization"] = f"Bearer {hf_token}"
 
     iframe_url = (
@@ -440,8 +445,7 @@ def from_spaces(
                 "Blocks or Interface locally. You may find this Guide helpful: "
                 "https://gradio.app/using_blocks_like_functions/"
             )
-        if client.app_version < version.Version("4.0.0b14"):
-            return from_spaces_blocks(space=space_name, hf_token=hf_token)
+        return from_spaces_blocks(space=space_name, hf_token=hf_token)
 
 
 def from_spaces_blocks(space: str, hf_token: str | None) -> Blocks:
@@ -486,7 +490,7 @@ def from_spaces_interface(
     config = external_utils.streamline_spaces_interface(config)
     api_url = f"{iframe_url}/api/predict/"
     headers = {"Content-Type": "application/json"}
-    if hf_token is not None:
+    if hf_token not in [False, None]:
         headers["Authorization"] = f"Bearer {hf_token}"
 
     # The function should call the API with preprocessed data
@@ -526,6 +530,56 @@ def gr_Interface_load(
     src: str | None = None,
     hf_token: str | None = None,
     alias: str | None = None,
-    **kwargs,
+    **kwargs, # ignore
 ) -> Blocks:
-    return load_blocks_from_repo(name, src, hf_token, alias)
+    try:
+        return load_blocks_from_repo(name, src, hf_token, alias)
+    except Exception as e:
+        print(e)
+        return gradio.Interface(lambda: None, ['text'], ['image'])
+
+
+def list_uniq(l):
+    return sorted(set(l), key=l.index)
+
+
+def get_status(model_name: str):
+    from huggingface_hub import AsyncInferenceClient
+    client = AsyncInferenceClient(token=HF_TOKEN, timeout=10)
+    return client.get_model_status(model_name)
+
+
+def is_loadable(model_name: str, force_gpu: bool = False):
+    try:
+        status = get_status(model_name)
+    except Exception as e:
+        print(e)
+        print(f"Couldn't load {model_name}.")
+        return False
+    gpu_state = isinstance(status.compute_type, dict) and "gpu" in status.compute_type.keys()
+    if status is None or status.state not in ["Loadable", "Loaded"] or (force_gpu and not gpu_state):
+        print(f"Couldn't load {model_name}. Model state:'{status.state}', GPU:{gpu_state}")
+    return status is not None and status.state in ["Loadable", "Loaded"] and (not force_gpu or gpu_state)
+
+
+def find_model_list(author: str="", tags: list[str]=[], not_tag="", sort: str="last_modified", limit: int=30, force_gpu=False, check_status=False):
+    from huggingface_hub import HfApi
+    api = HfApi(token=HF_TOKEN)
+    default_tags = ["diffusers"]
+    if not sort: sort = "last_modified"
+    limit = limit * 20 if check_status and force_gpu else limit * 5
+    models = []
+    try:
+        model_infos = api.list_models(author=author, #task="text-to-image",
+                                       tags=list_uniq(default_tags + tags), cardData=True, sort=sort, limit=limit)
+    except Exception as e:
+        print(f"Error: Failed to list models.")
+        print(e)
+        return models
+    for model in model_infos:
+        if not model.private and not model.gated or HF_TOKEN is not None:
+           loadable = is_loadable(model.id, force_gpu) if check_status else True
+           if not_tag and not_tag in model.tags or not loadable: continue
+           models.append(model.id)
+           if len(models) == limit: break
+    return models