Merge branch 'new-flagembedding-v1' of https://github.com/hanhainebula/FlagEmbedding into new-flagembedding-v1

Author: 545999961

FlagEmbedding/abc/inference/AbsEmbedder.py

@@ -19,6 +19,25 @@ class AbsEmbedder(ABC):
     """
     Base class for embedder.
     Extend this class and implement `encode_queries`, `encode_passages`, `encode` for custom embedders.
+
+    Args:
+        model_name_or_path (str): If it's a path to a local model, it loads the model from the path. Otherwise tries to download and
+            load a model from HuggingFace Hub with the name.
+        normalize_embeddings (bool, optional): If True, normalize the embedding vector. Default: `True`.
+        use_fp16 (bool, optional): If true, use half-precision floating-point to speed up computation with a slight performance 
+            degradation. Default: `True`.
+        query_instruction_for_retrieval: (Optional[str], optional): Query instruction for retrieval tasks, which will be used with
+            with `query_instruction_format`. Default: `None`.
+        query_instruction_format: (str, optional): The template for `query_instruction_for_retrieval`. Default: `"{}{}"`.
+        devices (Optional[Union[str, int, List[str], List[int]]], optional): Devices to use for model inference. Default: `None`.
+        batch_size (int, optional): Batch size for inference. Default: `256`.
+        query_max_length (int, optional): Maximum length for query. Default: `512`.
+        passage_max_length (int, optional): Maximum length for passage. Default: `512`.
+        instruction (Optional[str], optional): Instruction for embedding. Default: `None`.
+        instruction_format (str, optional): Instruction format when using `instruction`. Default: `"{}{}"`.
+        convert_to_numpy (bool, optional): If True, the output embedding will be a Numpy array. Otherwise, it will be a Torch Tensor. 
+            Default: `True`.
+        kwargs (Dict[Any], optional): Additional parameters for HuggingFace Transformers config or children classes.
     """
     def __init__(
         self,

Tutorials/1_Embedding/1.1_Intro&Inference.ipynb

@@ -83,6 +83,18 @@
     "%pip install -U FlagEmbedding sentence_transformers openai cohere"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import os \n",
+    "os.environ['TRANSFORMERS_NO_ADVISORY_WARNINGS'] = 'true'\n",
+    "# single GPU is better for small tasks\n",
+    "os.environ['CUDA_VISIBLE_DEVICES'] = '0'"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -92,7 +104,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -138,19 +150,27 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "initial target device: 100%|██████████| 8/8 [00:31<00:00,  3.89s/it]\n",
+      "Chunks: 100%|██████████| 3/3 [00:04<00:00,  1.61s/it]\n"
+     ]
+    },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
       "Embeddings:\n",
       "(3, 768)\n",
       "Similarity scores:\n",
-      "[[1.         0.7900386  0.57525384]\n",
-      " [0.7900386  0.9999998  0.59190154]\n",
-      " [0.57525384 0.59190154 0.99999994]]\n"
+      "[[1.     0.79   0.575 ]\n",
+      " [0.79   0.9995 0.592 ]\n",
+      " [0.575  0.592  0.999 ]]\n"
      ]
     }
    ],
@@ -373,7 +393,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "base",
+   "display_name": "dev",
    "language": "python",
    "name": "python3"
   },
@@ -387,7 +407,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.13"
+   "version": "3.12.7"
   }
  },
  "nbformat": 4,