Not completing and not sure how to debug

[OneOfOne]

My current lazyvim config:

    {
        'milanglacier/minuet-ai.nvim',
        lazy = false,
        config = function()
            require('minuet').setup({
                provider_options = {
                    openai_compatible = {
                        model = 'llama3.2:3b',
                        end_point = 'http://localhost:11434/v1/chat/completions',
                        name = 'Ollama',
                        stream = false,
                    },
                },
            })
        end,
    },

....

-- nvim-cmp
            opts.sources = cmp.config.sources({
                { name = 'copilot' },
                { name = 'mineut' },

                { name = 'nvim_lsp' },
                { name = 'crates' },
            }, {
                { name = 'buffer' },
                { name = 'path' },
            })

It doesn't suggest anything and as far as I can tell it doesn't call the ollama server at all judging by journalctl log.

[milanglacier]

1. You need a fake API key (existed in your environement variables) even if ollama does not require the API key. This is mandatory.

2. I would suggest you take a look at the config suggested by README.

https://github.com/milanglacier/minuet-ai.nvim?tab=readme-ov-file#integration-with-lazyvim

Besides, I would recommend stream = true, because if the ollama cannot complete the chat completion request within the timeout, minuet can't receive anything.

[OneOfOne]

The provider specified as OpenAI compatible is not properly configured.

    {
        'milanglacier/minuet-ai.nvim',
        lazy = false,
        config = function()
            require('minuet').setup({
                provider = 'openai_compatible',
                provider_options = {
                    openai_compatible = {
                        model = 'llama3.2:3b',
                        end_point = 'http://localhost:11434/v1/chat/completions',
                        name = 'Ollama',
                        api_key = 'DEEPSEEK_API_KEY',
                        stream = true,
                        optional = {
                            stop = nil,
                            max_tokens = nil,
                        },
                    },
                },
            })
        end,
    },
    {
        'nvim-cmp',
        opts = function(_, opts)
            -- if you wish to use autocomplete
            table.insert(opts.sources, 1, {
                name = 'minuet',
                group_index = 1,
                priority = 100,
            })

            opts.performance = {
                -- It is recommended to increase the timeout duration due to
                -- the typically slower response speed of LLMs compared to
                -- other completion sources. This is not needed when you only
                -- need manual completion.
                fetching_timeout = 2000,
            }

            opts.mapping = vim.tbl_deep_extend('force', opts.mapping or {}, {
                -- if you wish to use manual complete
                ['<c-y>'] = require('minuet').make_cmp_map(),
                -- You don't need to worry about <CR> delay because lazyvim handles this situation for you.
                ['<CR>'] = nil,
            })
        end,
    },

[milanglacier]

Did you have your deep seek api key configured in your environment variables?

it need to set to an arbitrary value, it can be whatever, like aaa, bbb.

[OneOfOne]

I'm sorry, how do you set that? I thought you set it in the config as api_key?

[milanglacier]

I'm sorry, how do you set that? I thought you set it in the config as api_key?

In your shell configuration file, .profile, or .bash_profile, or .zprofile, depending on which shell you are using, write something like this:

export DEEPSEEK_API_KEY=xxx