File size: 1,147 Bytes
27693fd 0204d4a 27693fd c07da70 63e33f3 c07da70 63e33f3 c07da70 4236a98 c07da70 634e670 c07da70 63e33f3 c07da70 af4148c 63e33f3 27693fd c07da70 63e33f3 c07da70 17ffe56 c07da70 |
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 |
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
from config import HF_TOKEN, MODEL_ID
def load_model():
try:
print(f"🔄 Loading tokenizer and model: {MODEL_ID}")
tokenizer = AutoTokenizer.from_pretrained(
MODEL_ID,
token=HF_TOKEN or None,
trust_remote_code=True
)
model = AutoModelForCausalLM.from_pretrained(
MODEL_ID,
token=HF_TOKEN or None,
trust_remote_code=True,
device_map="auto" if torch.cuda.is_available() else "cpu",
torch_dtype=torch.float16 if torch.cuda.is_available() else torch.float32,
low_cpu_mem_usage=True
)
print("✅ Model loaded successfully.")
return pipeline(
"text-generation",
model=model,
tokenizer=tokenizer,
max_new_tokens=2048,
do_sample=True,
temperature=0.7,
top_p=0.9
)
except Exception as e:
print(f"❌ Failed to load model: {e}")
raise RuntimeError(f"Model loading failed: {e}")
|