Pandas to_csv for multiple rows/timestamps of the same symbol

I am fetching a constant stream of data into a dataframe, which I now need to save into multiple CSVs.

For example,I am fetching OHLCV data from Binance and I get this dataframe:

                             sym              o              h              l  \2024-02-09 11:15:59.594  ETHUSDT  2419.56000000  2479.88000000  2419.16000000                                        c                v  barcomplete  2024-02-09 11:15:59.594  2471.79000000  170696.13700000        False                              sym           o           h           l  \2024-02-09 11:15:59.622  IDUSDT  0.54534000  0.64866000  0.53587000                                     c                  v  barcomplete  2024-02-09 11:15:59.622  0.60634000  93122120.00000000        False                               sym            o            h            l  \2024-02-09 11:15:59.658  ICPUSDT  12.18600000  12.81000000  12.16500000                                      c                 v  barcomplete  2024-02-09 11:15:59.658  12.62300000  1065607.61000000        False                               sym              o              h              l  \2024-02-09 11:15:59.594  ETHUSDT  2400.56000000  2422.88000000  2399.16000000                                        c                v  barcomplete  2024-02-08 11:15:59.594  2419.79000000  160696.13700000        False  

Index is the timestamp. I get multiple data rows for the same sym (ETHUSDT price today, yesterday, day before and so on). I want to save coins/ETHUSDT rows into their own separate CSV (ETHUSDT.csv, IDUSDT.csv, etc), with new data rows being appended to those CSVs as they get fetched.

I am using this but its slow:

    for coin in df.sym:        filename = r"{}.csv".format(coin)        print(filename)        df['sym'].to_csv(filename, mode='a', header=False)

But I just can't get this working properly. Please advise. (I just started learning pandas so please be kind) :)