changelog.json

{
    "1.1.0": {
        "upgraded": [
            "OpenAI package updated. The model gpt-3.5-turbo is available!"
        ],
        "updated": [
            "Context saving of conversations is working better by removing old messages. That allows to have an infinite conversation without the context resetting.",
            "Telegram response changed from simple text to text with Markdown V1 to highlight a code syntax."
        ],
        "fixed": [
            "Response max token decreased from 4000 to 1000.",
            "Temperature increased from 0.5 to 1 (by default)."
        ],
        "notes": [
            "Updating to 3.5 allows to decrease the cost from $0.020/1k tokens to $0.002/1k tokens. The language model is working in the same way."
        ]
    },
    "1.0.0": "The first version is implemented."
}