From 3ddeacd83f534dfe48ec8b75e9d64dfe9234e752 Mon Sep 17 00:00:00 2001
From: marioggil <marioggil@gmail.com>
Date: Thu, 28 Sep 2023 14:48:59 -0500
Subject: [PATCH] Custom models of  huginface

---
 .gitignore          |  2 ++
 FindinDB.py         |  1 +
 loadModelinlocal.py |  9 +++++++++
 retrainEmbbeding.py | 13 +++++++++++++
 4 files changed, 25 insertions(+)
 create mode 100644 loadModelinlocal.py
 create mode 100644 retrainEmbbeding.py

diff --git a/.gitignore b/.gitignore
index e0276d8..5c17688 100644
--- a/.gitignore
+++ b/.gitignore
@@ -4,3 +4,5 @@ names/*
 nameshf/*
 photo_2023-09-24_00-25-17.jpg
 __pycache__/FindinDB.cpython-38.pyc
+embeddings/*
+tuned_models/*
\ No newline at end of file
diff --git a/FindinDB.py b/FindinDB.py
index e7deb46..8464d78 100644
--- a/FindinDB.py
+++ b/FindinDB.py
@@ -41,6 +41,7 @@ def read_main():
     return {"message": "This is your main app"}
 
 
+
 def loadModels():
     #model = GPT4All("orca-mini-3b.ggmlv3.q4_0.bin")
     callback_manager = BaseCallbackManager([StreamingStdOutCallbackHandler()])
diff --git a/loadModelinlocal.py b/loadModelinlocal.py
new file mode 100644
index 0000000..3572aae
--- /dev/null
+++ b/loadModelinlocal.py
@@ -0,0 +1,9 @@
+
+from langchain.embeddings import HuggingFaceEmbeddings
+df=HuggingFaceEmbeddings(
+    model_name="embeddings/all-MiniLM-L6-v2"
+)
+text = "This is a test document."
+
+query_result = df.embed_query(text)
+print(query_result)
\ No newline at end of file
diff --git a/retrainEmbbeding.py b/retrainEmbbeding.py
new file mode 100644
index 0000000..197b1fe
--- /dev/null
+++ b/retrainEmbbeding.py
@@ -0,0 +1,13 @@
+from sentence_transformers import SentenceTransformer
+# Preguntas y respuestas especializado en eso "multi-qa-mpnet-base-dot-v1"
+# uno de uso gereal el de mejor desempeño all-mpnet-base-v2
+# el mas rapido "paraphrase-MiniLM-L3-v2" y "all-MiniLM-L6-v2"
+# muy rappudo y muy acertado "all-MiniLM-L12-v2"
+models=["all-MiniLM-L12-v2","paraphrase-MiniLM-L3-v2" , "all-MiniLM-L6-v2","all-mpnet-base-v2","multi-qa-mpnet-base-dot-v1"]
+for model in models:
+    modelST = SentenceTransformer(model)
+    # Define the path where you want to save the model
+    save_path = './embeddings/%s/'%(model)
+    # Save the model
+    modelST.save(save_path)
+