Update README.md

README.md

@@ -23,16 +23,6 @@ from peft import PeftModel, PeftConfig
 from transformers import AutoModelForCausalLM, AutoTokenizer, GenerationConfig
 
 MODEL_NAME = "IlyaGusev/saiga_7b_lora"
-
-config = PeftConfig.from_pretrained(MODEL_NAME)
-model = AutoModelForCausalLM.from_pretrained(
-    config.base_model_name_or_path,
-    load_in_8bit=True,
-    device_map="auto"
-)
-model = PeftModel.from_pretrained(model, MODEL_NAME)
-tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
-
 DEFAULT_MESSAGE_TEMPLATE = "<s>{role}\n{content}</s>\n"
 DEFAULT_SYSTEM_PROMPT = "Ты — Сайга, русскоязычный автоматический ассистент. Ты разговариваешь с людьми и помогаешь им."
 
@@ -91,9 +81,14 @@ def generate(model, tokenizer, prompt, generation_config):
     output = tokenizer.decode(output_ids, skip_special_tokens=True)
     return output.strip()
 
-from transformers import GenerationConfig
-
-
+config = PeftConfig.from_pretrained(MODEL_NAME)
+model = AutoModelForCausalLM.from_pretrained(
+    config.base_model_name_or_path,
+    load_in_8bit=True,
+    device_map="auto"
+)
+model = PeftModel.from_pretrained(model, MODEL_NAME)
+tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
 generation_config = GenerationConfig.from_pretrained(MODEL_NAME)
 print(generation_config)