Csv

am getting this error and i dont see any way to fix it:

Traceback (most recent call last):
  File "/Users/ahmednadiir/Desktop/agency/app.py", line 11, in <module>
    from quran import quran_engine
  File "/Users/ahmednadiir/Desktop/agency/quran.py", line 19, in <module>
    quran_csv = CSVReader().load_data(file=csv_path);
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/ahmednadiir/Desktop/agency/localEvir/lib/python3.11/site-packages/llama_index/readers/file/tabular/base.py", line 48, in load_data
    return [Document(text="\n".join(text_list), metadata=extra_info)]
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/ahmednadiir/Desktop/agency/localEvir/lib/python3.11/site-packages/pydantic/v1/main.py", line 341, in __init__
    raise validation_error
pydantic.v1.error_wrappers.ValidationError: 1 validation error for Document
extra_info
  none is not an allowed value (type=type_error.none.not_allowed)

code:

import os
from llama_index.core import StorageContext, VectorStoreIndex, load_index_from_storage
from llama_index.readers.file import CSVReader

def get_index(data, index_name):
    index = None
    if not os.path.exists(index_name):
        print("building index", index_name)
        index = VectorStoreIndex.from_documents(data, show_progress=True)
        index.storage_context.persist(persist_dir=index_name)
    else:
        index = load_index_from_storage(
            StorageContext.from_defaults(persist_dir=index_name)
        )

    return index

csv_path = os.path.join("data", "quran-english-tafsir.csv");
quran_csv = CSVReader().load_data(file=csv_path);
quran_index = get_index(quran_csv, "quran")
quran_engine = quran_index.as_query_engine()
quran_engine.query()

9 comments

LLogan M

.load_data(..., extra_info={})

LLogan M

Should probably patch that reader to do that for you

CCOdXG

so like this quran_csv = CSVReader().load_data(file=csv_path, extra_info={});

LLogan M

I think so

CCOdXG

is there anyway i can put my gemini api on my local .env file so it can defaultly use it without definding it like it doing to openai

LLogan M

Nope. The easiest way is just setting it as a global default

Plain Text

from llama_index.core import Settings

Settings.llm = Gemini()

CCOdXG

i did this:

import os
from dotenv import load_dotenv
from llama_index.core import StorageContext, VectorStoreIndex, load_index_from_storage
from llama_index.readers.file import CSVReader
from llama_index.llms.gemini import Gemini
from llama_index.core import Settings

Settings.llm = Gemini()
load_dotenv()
GOOGLE_API_KEY = "api-ishere";
os.environ["GOOGLE_API_KEY"] = GOOGLE_API_KEY

def get_index(data, index_name):
    index = None
    if not os.path.exists(index_name):
        print("building index", index_name)
        index = VectorStoreIndex.from_documents(data, show_progress=True)
        index.storage_context.persist(persist_dir=index_name)
    else:
        index = load_index_from_storage(
            StorageContext.from_defaults(persist_dir=index_name)
        )

    return index

csv_path = os.path.join("data", "quran-english-tafsir.csv");
quran_csv = CSVReader().load_data(file=csv_path, extra_info={});
quran_index = get_index(quran_csv, "quran")
quran_engine = quran_index.as_query_engine()
quran_engine.query()

but i got :

google.api_core.exceptions.InvalidArgument: 400 API key not valid. Please pass a valid API key. [reason: "API_KEY_INVALID"
domain: "googleapis.com"
metadata {
  key: "service"
  value: "generativelanguage.googleapis.com"
}
]

i dont know the reason altho i pass the right api key in the file and also in the .env file?

LLogan M

Where is your api key from? Google is kind of confusing, Gemini and vertex are two separate platforms (not sure if you are using vertexai or something else)

ddoughboy

This is what works for me:

Plain Text

from dotenv import load_dotenv
load_dotenv()

Settings.llm = Gemini(
    api_key = os.getenv('GEMINI_API_KEY'),
    model_name = 'models/gemini-pro'
)

Add a reply

Find answers from the community

Csv