test space

.github/workflows/hugging-face.yaml

@@ -15,6 +15,4 @@ jobs:
       - name: Push to hub
         env:
           HF_TOKEN: ${{ secrets.HF_TOKEN }}
-          HF_USERNAME: zaibutcooler
-          SPACE_NAME: yume
-        run: git push https://${{ secrets.HF_USERNAME }}:[email protected]/spaces/$HF_USERNAME/$SPACE_NAME main
+        run: git push https://zaibutcooler:[email protected]/spaces/zaibutcooler/burmese-gpt main

README.md

@@ -1 +1,14 @@
 # Burmese GPT
+
+---
+
+title: BurmeseGPT
+emoji: ✨
+colorFrom: yellow
+colorTo: blue
+sdk: streamlit
+sdk_version: 1.29.0
+app_file: space.py
+pinned: false
+license: openrail
+---

burmesegpt/__init__.py

@@ -1 +1 @@
-from .main import BurmeseGpt
+from .core import BurmeseGpt

burmesegpt/{main.py → core.py}

@@ -1,4 +1,6 @@
 import torch
+from huggingface_hub import login
+
 from .models import SelfAttention,MLP,GPT
 from .tokenizer import Tokenizer
 from .config import Config
@@ -16,18 +18,29 @@ class BurmeseGpt:
         self.train_data = Data().to(device)
         self.tokenizer = tk
 
-    def train(self):
+    def save_pretrained(self, name="burmese-gpt"):
+        print("Uploading model...")
+        self.model.save_pretrained(name)
+        print(f"Model saved locally as '{name}'")
+        self.model.push_to_hub(name)
+        print(f"Model '{name}' uploaded to the Hugging Face Model Hub")
+
+    def load_pretrained(self, model_id="zaibutcooler/burmese-gpt"):
+        print("Loading model...")
+        model = model.from_pretrained(model_id)
+        print(f"Model '{model_id}' loaded successfully")
+        return model
+    
+    def huggingface_login(self,token):
+        login(token)
+    
+    def pretrain(self):
+        pass
+
+    def fine_tune(self):
         pass
 
     def generate(self,entry=''):
         result = None
         
-        return result
-
-    def save_trained(self):
-        pass
-
-    def load_pretrained(self,name=""):
-        pass
-
-    
+        return result

burmesegpt/data_prep.py

@@ -2,12 +2,12 @@
 from datasets import load_dataset
 from torch.utils.data import Dataset
 
-dataset = load_dataset("")
 
 class Data(Dataset):
-    def __init__(self,trainset):
+    def __init__(self):
         super().__init__()
-        self.trainset = trainset
+        self.texts = None
+        self.data
 
     def __len__(self):
         return None

burmesegpt/utils.py

@@ -5,21 +5,3 @@ def get_stats():
 
 def merge():
     pass
-
-def upload_model(model, name="", token=""):
-    print("Uploading model...")
-    model.save_pretrained(name)
-    print(f"Model saved locally as '{name}'")
-    model.push_to_hub(name)
-    print(f"Model '{name}' uploaded to the Hugging Face Model Hub")
-
-def save_model(model, name="gpt"):
-    print("Saving model...")
-    model.save_pretrained(name)
-    print(f"Model saved locally as '{name}'")
-
-def load_model(model_name):
-    print("Loading model...")
-    model = model.from_pretrained(model_name)
-    print(f"Model '{model_name}' loaded successfully")
-    return model

requirements.txt

@@ -0,0 +1 @@
+streamlit