removed peft

app.py

@@ -1,4 +1,3 @@
-from peft import PeftModel, PeftConfig
 import torch
 from datasets import load_dataset
 
@@ -8,28 +7,32 @@ print("executed successfully")
 dataset_name = "timdettmers/openassistant-guanaco"
 dataset = load_dataset(dataset_name, split="train")
 
+from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
 
-import os
+# quantizition configuration
+bnb_config = BitsAndBytesConfig(
+    load_in_4bit=True,
+    bnb_4bit_quant_type="nf4",
+    bnb_4bit_compute_dtype=torch.float16,
+)
 
-ld_library_path = os.environ.get("LD_LIBRARY_PATH")
-print(ld_library_path)
+# download model
 
-# The directory to search in
-search_directory = '/usr/'
+model_name = "TinyPixel/Llama-2-7B-bf16-sharded"
+model = AutoModelForCausalLM.from_pretrained(
+    model_name,
+    quantization_config=bnb_config,
+    trust_remote_code=True
+)
+model.config.use_cache = False
 
-# The filename pattern to look for (wildcard * is used)
-filename_pattern = 'libcuda.so.*'
+tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
+tokenizer.pad_token = tokenizer.eos_token
 
-# Initialize a list to store the paths of matching files
-matching_files = []
+text = "What is a large language model?"
+device = "cuda:0"
+inputs = tokenizer(text, return_tensors="pt").to(device)
+outputs = model.generate(**inputs, max_new_tokens=50)
+print(tokenizer.decode(outputs[0], skip_special_tokens=True))
 
-# Walk through the directory and its subdirectories
-for root, dirs, files in os.walk(search_directory):
-    for file in files:
-        if file.startswith('libcuda.so.'):
-            matching_files.append(os.path.join(root, file))
-
-# Print the paths of matching files
-for file in matching_files:
-    print(file)