Update README.md

README.md

@@ -14,27 +14,45 @@ This is a fine-tuned version of LLAMA2 trained (7b) on spider, sql-create-contex
 
 To initialize the model:
 
-    
-    #from transformers import MBartForConditionalGeneration, MBart50TokenizerFast
-    #model = MBartForConditionalGeneration.from_pretrained("MRNH/mbart-english-grammar-corrector")
+bnb_config = BitsAndBytesConfig(
+    load_in_4bit=use_4bit,
+    bnb_4bit_quant_type=bnb_4bit_quant_type,
+    bnb_4bit_compute_dtype=compute_dtype,
+    bnb_4bit_use_double_quant=use_nested_quant,
+)
+
+model = AutoModelForCausalLM.from_pretrained(
+    model_name,
+    quantization_config=bnb_config,
+    device_map=device_map,
+    trust_remote_code=True
+)
     
     
 Use the tokenizer:
 
     
-    #tokenizer = MBart50TokenizerFast.from_pretrained("MRNH/mbart-english-grammar-corrector", src_lang="en_XX", tgt_lang="en_XX")
-    
-    #input = tokenizer("I was here yesterday to studying",
-    #                  text_target="I was here yesterday to study", return_tensors='pt')
+tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
+tokenizer.pad_token = tokenizer.eos_token
+tokenizer.padding_side = "right"
+
+To get the prompt:
+    dataset = dataset.map(
+    lambda example: {
+        "input": "### Instruction: \nYou are a powerful text-to-SQL model.   \
+                    Your job is to answer questions about a database. You are given \
+                    a question and context regarding one or more tables. \n\nYou must \
+                    output the SQL query that answers the question.   \
+                    \n\n \
+                    ### Dialect:\n\nsqlite\n\n \
+                    ### question:\n\n"+ example["question"]+" \
+                    \n\n### Context:\n\n"+example["context"],
+        "answer": example["answer"]
+    }
+    )
 
-To generate text using the model:
-
-    #output = model.generate(input["input_ids"],attention_mask=input["attention_mask"],
-    #                       forced_bos_token_id=tokenizer_it.lang_code_to_id["en_XX"])
 
+To generate text using the model:
 
-Training of the model is performed using the following loss computation based on the hidden state output h:
+    output = model.generate(input["input_ids"])
 
-    #h.logits, h.loss = model(input_ids=input["input_ids"],
-    #                                              attention_mask=input["attention_mask"],
-    #                                              labels=input["labels"])