File size: 9,061 Bytes
57d76cc |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 |
#!/usr/bin/env python3
"""
Integration Guide: Add Authentication to Existing Training Code
This script shows how to integrate Hugging Face authentication into your
existing OpenLLM training code. Copy the relevant parts into your training script.
Usage:
Use this as a reference to update your existing training code.
"""
import os
import sys
import json
try:
from huggingface_hub import HfApi, login, whoami, create_repo
HF_AVAILABLE = True
except ImportError:
HF_AVAILABLE = False
print("❌ huggingface_hub not installed")
sys.exit(1)
def setup_hf_authentication():
"""
Set up Hugging Face authentication using GitHub secrets.
Add this function to your training script.
"""
print("🔐 Setting up Hugging Face Authentication")
print("-" * 40)
try:
# Get token from GitHub secrets
token = os.getenv("HF_TOKEN")
if not token:
raise ValueError("HF_TOKEN not found. Please set it in GitHub repository secrets.")
# Login
login(token=token)
# Get user info
api = HfApi()
user_info = whoami()
username = user_info["name"]
print(f"✅ Authentication successful!")
print(f" - Username: {username}")
print(f" - Source: GitHub secrets")
return api, username
except Exception as e:
print(f"❌ Authentication failed: {e}")
raise
def upload_model_after_training(api, username, model_dir, model_size="small", steps=8000):
"""
Upload the trained model to Hugging Face Hub.
Call this function after your training completes.
"""
try:
# Create repository name
repo_name = f"openllm-{model_size}-extended-{steps//1000}k"
repo_id = f"{username}/{repo_name}"
print(f"\n📤 Uploading model to {repo_id}")
# Create repository
create_repo(
repo_id=repo_id,
repo_type="model",
exist_ok=True,
private=False
)
# Create model configuration
config = {
"architectures": ["GPTModel"],
"model_type": "gpt",
"vocab_size": 32000,
"n_positions": 2048,
"n_embd": 768 if model_size == "small" else 1024 if model_size == "medium" else 1280,
"n_layer": 12 if model_size == "small" else 24 if model_size == "medium" else 32,
"n_head": 12 if model_size == "small" else 16 if model_size == "medium" else 20,
"bos_token_id": 1,
"eos_token_id": 2,
"pad_token_id": 0,
"unk_token_id": 3,
"transformers_version": "4.35.0",
"use_cache": True
}
config_path = os.path.join(model_dir, "config.json")
with open(config_path, "w") as f:
json.dump(config, f, indent=2)
# Create model card
model_card = f"""# OpenLLM {model_size.capitalize()} Model ({steps} steps)
This is a trained OpenLLM {model_size} model with extended training.
## Model Details
- **Model Type**: GPT-style decoder-only transformer
- **Architecture**: Custom OpenLLM implementation
- **Training Data**: SQUAD dataset (Wikipedia passages)
- **Vocabulary Size**: 32,000 tokens
- **Sequence Length**: 2,048 tokens
- **Model Size**: {model_size.capitalize()}
- **Training Steps**: {steps:,}
## Usage
This model can be used with the OpenLLM framework for text generation and language modeling tasks.
## License
This model is released under the GNU General Public License v3.0.
## Repository
This model is hosted on Hugging Face Hub: https://huggingface.co/{repo_id}
"""
readme_path = os.path.join(model_dir, "README.md")
with open(readme_path, "w") as f:
f.write(model_card)
# Upload all files
api.upload_folder(
folder_path=model_dir,
repo_id=repo_id,
repo_type="model",
commit_message=f"Add OpenLLM {model_size} model ({steps} steps)"
)
print(f"✅ Model uploaded successfully!")
print(f" - Repository: https://huggingface.co/{repo_id}")
return repo_id
except Exception as e:
print(f"❌ Upload failed: {e}")
raise
# ============================================================================
# INTEGRATION EXAMPLE: How to modify your existing training code
# ============================================================================
def example_integration():
"""
Example of how to integrate authentication into your existing training code.
"""
print("🚀 Example: Integrating Authentication into Training")
print("=" * 55)
# Step 1: Set up authentication at the start
print("\n1️⃣ Setting up authentication...")
api, username = setup_hf_authentication()
# Step 2: Your existing training code goes here
print("\n2️⃣ Running your existing training code...")
print(" - This is where your actual training happens")
print(" - Training saves model to: ./openllm-trained")
# Simulate training completion
model_dir = "./openllm-trained"
os.makedirs(model_dir, exist_ok=True)
# Create dummy model file
with open(os.path.join(model_dir, "best_model.pt"), "w") as f:
f.write("Dummy model file")
print(" ✅ Training completed!")
# Step 3: Upload model after training
print("\n3️⃣ Uploading model...")
repo_id = upload_model_after_training(
api=api,
username=username,
model_dir=model_dir,
model_size="small",
steps=8000
)
print(f"\n🎉 Success! Model available at: https://huggingface.co/{repo_id}")
# ============================================================================
# CODE SNIPPETS FOR YOUR EXISTING TRAINING SCRIPT
# ============================================================================
def get_code_snippets():
"""Show code snippets to add to your existing training script."""
snippets = """
# ============================================================================
# ADD THESE IMPORTS TO YOUR TRAINING SCRIPT
# ============================================================================
import os
from huggingface_hub import HfApi, login, whoami, create_repo
import json
# ============================================================================
# ADD THIS FUNCTION TO YOUR TRAINING SCRIPT
# ============================================================================
def setup_hf_authentication():
\"\"\"Set up Hugging Face authentication using GitHub secrets.\"\"\"
token = os.getenv("HF_TOKEN")
if not token:
raise ValueError("HF_TOKEN not found. Please set it in GitHub repository secrets.")
login(token=token)
api = HfApi()
user_info = whoami()
username = user_info["name"]
print(f"✅ Authentication successful: {username}")
return api, username
# ============================================================================
# ADD THIS FUNCTION TO YOUR TRAINING SCRIPT
# ============================================================================
def upload_model_after_training(api, username, model_dir, model_size="small", steps=8000):
\"\"\"Upload the trained model to Hugging Face Hub.\"\"\"
repo_name = f"openllm-{model_size}-extended-{steps//1000}k"
repo_id = f"{username}/{repo_name}"
# Create repository
create_repo(repo_id=repo_id, repo_type="model", exist_ok=True)
# Upload all files
api.upload_folder(
folder_path=model_dir,
repo_id=repo_id,
repo_type="model",
commit_message=f"Add OpenLLM {model_size} model ({steps} steps)"
)
print(f"✅ Model uploaded: https://huggingface.co/{repo_id}")
return repo_id
# ============================================================================
# MODIFY YOUR MAIN TRAINING FUNCTION
# ============================================================================
def main():
# Step 1: Set up authentication
api, username = setup_hf_authentication()
# Step 2: Your existing training code
# ... your training code here ...
# Step 3: Upload after training
model_dir = "./openllm-trained" # Your model directory
repo_id = upload_model_after_training(api, username, model_dir)
print(f"🎉 Training and upload completed!")
if __name__ == "__main__":
main()
"""
return snippets
def main():
"""Main function to demonstrate integration."""
print("🔧 Integration Guide: Add Authentication to Existing Training")
print("=" * 65)
# Show example integration
example_integration()
# Show code snippets
print("\n" + "="*65)
print("📝 CODE SNIPPETS FOR YOUR EXISTING TRAINING SCRIPT")
print("="*65)
print(get_code_snippets())
if __name__ == "__main__":
main()
|