support comma separated output_format

[mercury233]

No description provided.

[mercury233]

the code is wrotten by chatgpt

[NielsMayer]

Too bad ChatGPT wasn't able to find this: #228 -- and the consensus solution to support TSV format as one of the official options for whisper output formats.

[ryanheise]

ChatGPT can't know about Whisper since it was released in 2022 which is after ChatGPT's knowledge cutoff date of September 2021. If it did give you an answer, it would either be a hallucination or an apology.

I think you would be better off just reading the README. Or if you don't want to read the "README", I guess you could copy and paste the README into ChatGPT and ask it to read it for you (but honestly, reading it + CTRL+F would still be faster).

---

(Conversation with ChatGPT)

I want to know if whisper supports the TSV output format. I will copy and paste the README below. Can you tell me if it does?

-- COPY AND PASTE --

Based on the provided README, there is no specific mention of the Whisper model supporting the TSV (Tab-Separated Values) output format. The README primarily focuses on the setup, available models and languages, command-line usage, and Python usage of the Whisper model for speech recognition and related tasks.

To confirm whether the Whisper model supports TSV output, it would be necessary to refer to the official documentation, codebase, or any updates from the developers. Alternatively, you can explore the codebase or consult the OpenAI Whisper community or support channels for more information regarding the supported output formats.

Is there any way I could find out from the command line whether it supports it?

To determine whether the OpenAI Whisper model supports TSV (Tab-Separated Values) output format from the command line, you can utilize the whisper --help command. This command will display the available options and parameters for the Whisper command-line interface.

To execute the command, open your command-line interface (e.g., Terminal or Command Prompt) and run the following command:

whisper --help

The output will provide a list of available options and their descriptions. Look for any options related to output format or file type, such as specifying TSV or tab-separated output. If such an option exists, it indicates that the Whisper model supports TSV output. Additionally, the documentation or help information may provide further details on how to specify the output format.

Keep in mind that as of my knowledge cutoff in September 2021, there was no specific mention of TSV output format in the provided README. Therefore, the availability of TSV output format support may require further investigation by referring to the official documentation or by contacting the OpenAI Whisper community or support channels.

Here is the output of whisper --help. Can you tell me if it explains how to get TSV output?

-- COPY AND PASTE ---

Based on the output of whisper --help, there is no specific option mentioned for TSV (Tab-Separated Values) output format. The available output formats listed in the help message are txt, vtt, srt, tsv, json, and all. However, please note that the options and capabilities of the Whisper model might have been updated or changed after my knowledge cutoff in September 2021.

To obtain more accurate and up-to-date information regarding the TSV output format support in Whisper, I recommend referring to the official Whisper documentation, release notes, or contacting the OpenAI Whisper community or support channels. They will provide the most reliable information about the model's capabilities and supported output formats.

What about tsv?

Apologies for the confusion in my previous response. According to the output of whisper --help, the tsv format is indeed listed as one of the available output formats. You can specify the tsv format using the --output_format or -f option when running the whisper command.

Here is the relevant section from the whisper --help output:

--output_format {txt,vtt,srt,tsv,json,all}, -f {txt,vtt,srt,tsv,json,all}
                        format of the output file; if not specified, all available formats will be produced (default: all)

To obtain the output in the TSV format, you can add --output_format tsv to your command. For example:

whisper --output_format tsv audio.wav

This will generate the transcription output in the TSV format.

---

As you can see from the interaction above, it seems I'm a better reader than ChatGPT ;-)