Update prompt/prompt.txt

prompt/prompt.txt

@@ -8,8 +8,6 @@ You should do:
     Please make sure that all SMILES in the output of 'the get_multi_molecular_text_to_correct' tool are in the final output in the four categories： 'reactant template', 'product template', 'condition smiles', 'product'. 
     
     !!!! If the molecule in the image already has a label, such as 2a, 3b, 3fa, or 3da, use the label provided in the image. 
-    !!!! important NOTE: due to the characteristics of the tool, if YOU find the labels of the molecule appear likse 3ab,3ac,3ad,3ae,3af (the first English letter (here is a)remains unchanged), our tool cannot output normally, so you should copy the second letter to the first, make sure that the first English letters are different, such as (3bab,3cac,3dad,3eae,3faf).
-    !!!! change label like 3ab,3ac,3ad,3ae,3af to 3bab,3cac,3dad,3eae,3faf.
     !!!! When the reactant and product templates you find in the output of the 'the get_multi_molecular_text_to_correct' tool are inconsistent with the SMILES of the reactants and products in the output of the 'get_reaction' tool, Please uniformly use the output of the "get_reaction" tool to ensure that subsequent processing will not report errors.
          
 Your output should be like the json format. An example output is:
@@ -17,19 +15,16 @@ Your output should be like the json format. An example output is:
     {'*C(=O)NN=CC(F)(F)F': ['1','reactant template'], 
     'N#CN':['2','reactant template'],###smiles and text and assigned label for reactants template in the "get_reaction".(Sometimes there are same reactants)
     '*C(=O)N1NC(C(F)(F)F)N=C1N': ['3','product template'], ###smiles and text and assigned label for the product template in the "get_reaction". And make sure that the product template and product use the same number.
-    'Cc1cc(C)cc(C(=O)N2NC(C(F)(F)F)N=C2N)c1': ['3a', '6 h, 88%','product'],###smiles and text and assigned label for different products. Note that please also identify the missing text of the tool.
+    'Cc1cc(C)cc(C(=O)N2NC(C(F)(F)F)N=C2N)c1': ['3a', '6 h, 88%','product'],###smiles and text and assigned label for different products. Note that please also identify the missing text of the tool. And make sure that the product template and product use the same number.
     'Cc1cc(C)cc(C(=O)N2NC(C(F)(F)F)N=C2N)c1': ['3b, X:CF3, 6 h, 63%, 2q, X:OMe, 4 h, 60%','product'],###smiles and text and assigned label for different products. Note that please also identify the missing text of the tool.
     'Cc1cc(C)cc(C(=O)N2NC(C(F)(F)F)N=C2N)c1': ['3ca', "8 h, 91%', 'product']###smiles and text and assigned label for different products. Note that please also identify the missing text of the tool.
-    'Cc1cc(C)cc(C(=O)N2NC(C(F)(F)F)N=C2N)c1': ['3cac', '3ac, 8 h, 91%', 'product'],###change label like 3ab,3ac,3ad,3ae,3af to 3bab,3cac,3dad,3eae,3faf
-    'Cc1cc(C)cc(C(=O)N2NC(C(F)(F)F)N=C2N)c1': ['3dad', '3ad, 8 h, 90%', 'product'],###change label like 3ab,3ac,3ad,3ae,3af to 3bab,3cac,3dad,3eae,3faf
+    'Cc1cc(C)cc(C(=O)N2NC(C(F)(F)F)N=C2N)c1': ['3ac', '8 h, 91%', 'product'],
+    'Cc1cc(C)cc(C(=O)N2NC(C(F)(F)F)N=C2N)c1': ['3ad',  '8 h, 90%', 'product'],
     'CC(CC(=O)c1ccccc1)OCCC#N':['B17', 'conditions'] ## if there is any smiles in the 'condition' in "get_reaction" output. If not, please find from the 'get_multi_molecular_text_to_correct output' which is neither smiles of reactant, product template, nor smiles of product, They are also conditions smiles.
-    }
+    } ###!!!!!important: Please make sure the 'product template' and 'product' use the same number, avoid like ['2','product template'] and ['3a','product'] appear together.
+
 
-!!!! important NOTE: due to the characteristics of the tool, if YOU find the labels of the molecule appear likse 3ab,3ac,3ad,3ae,3af (the first English letter (here is a)remains unchanged), our tool cannot output normally, so you should copy the second letter to the first, make sure that the first English letters are different, such as (3bab,3cac,3dad,3eae,3faf).
-!!!! important NOTE: change label like 3ab,3ac,3ad,3ae,3af to 3bab,3cac,3dad,3eae,3faf.
 !!!! important NOTE: Please make sure again that all SMILES in the output of the 'get_multi_molecular_text_to_correct' tool are in the final output, Which means the number of the SMLIES in the output of 'the get_multi_molecular_text_to_correct' tool is the same as in the final output.
 !!!! important NOTE: Note again, When the reactant and product templates you find in the output of the 'the get_multi_molecular_text_to_correct' tool are inconsistent with the SMILES of the reactants and products in the output of the 'get_reaction' tool, Please uniformly use the output of the "get_reaction" tool to ensure that subsequent processing will not report errors.
      (For example: If 'get_multi_molecular_text_to_correct' tool output the product template: {...{"smiles": "*C(=O)C1=C([3*])NC(=O)CC1[1*]", "texts": ["6"]}....}, "get_reaction" tool output the product template:{... "products": [{"smiles": '[1*]C1CC(=O)NC([3*])=C1C([2*])=O'...}...]}. Then please output:'[1*]C1CC(=O)NC([3*])=C1C([2*])=O': ['6', 'product template'])
-         
-
-!!!!!!!important: Please check your results again to make sure the 'product template' and 'product' use the same number, avoid like ['2','product template'] and ['3a','product'] appear together. And Make sure that the number of SMILES in you final output is equal to the number of SMILES in "get_multi_molecular_text_to_correct"
+!!!!!important: Please check your results again to make sure the 'product template' and 'product' use the same number, avoid like ['2','product template'] and ['3a','product'] appear together. And Make sure that the number of SMILES in you final output is equal to the number of SMILES in "get_multi_molecular_text_to_correct"