Can not using the ollama in docker container. ERROR: [LLM] http error

[meicale]

It is a great plugin and I love it. But I found an error here.

[LLM] http error: error sending request for url (http://localhost:11434/api/generate): connection closed before message completed

Following the config in the readme.

{
    "huggingface/llm.nvim",
    opts = {
      -- cf Setup
    },
    config = function()
      local llm = require("llm")
      llm.setup({
        api_token = nil, -- cf Install paragraph
        -- for ollama backend
        backend = "ollama", -- backend ID, "huggingface" | "" | "openai" | "tgi"
        model = "starcoder2:7b",
        url = "http://localhost:11434/api/generate",
        tokens_to_clear = { "<|endoftext|>" }, -- tokens to remove from the model's output
        -- parameters that are added to the request body, values are arbitrary, you can set any field:value pair here it will be passed as is to the backend
        request_body = {
          parameters = {
            max_new_tokens = 60,
            temperature = 0.2,
            top_p = 0.95,
          },
        },
        -- set this if the model supports fill in the middle
        fim = {
          enabled = true,
          prefix = "<fim_prefix>",
          middle = "<fim_middle>",
          suffix = "<fim_suffix>",
        },
        debounce_ms = 150,
        accept_keymap = "<C-y>",
        dismiss_keymap = "<C-n>",
        tls_skip_verify_insecure = false,
        -- llm-ls configuration, cf llm-ls section
        lsp = {
          bin_path = nil,
          host = nil,
          port = nil,
          version = "0.5.2",
        },
        tokenizer = {
          repository = "bigcode/starcoder2-7b",
         
        }, -- cf Tokenizer paragraph
        -- tokenizer = nil, -- cf Tokenizer paragraph
        context_window = 4096, -- max number of tokens for the context window
        enable_suggestions_on_startup = true,
        enable_suggestions_on_files = "*", -- pattern matching syntax to enable suggestions on specific files, either a string or a list of strings
      })
    end,
  } 

The MOST wirred thing is that I can curl the answer to the same model & api url, and my vscode continue plugin can communicate this is ollama which is running on a docker container but this plugin cannot!

Thank you for your time and reply!

[FlippingBinary]

This configuration worked for me as I try this plugin for the first time today. I only made one change, but it shouldn't have an impact on how the plugin behaves. According to lazy.nvim's documentation, "opts is the recommended way to configure plugins." So I replaced your empty opts dictionary with the one you passed to the setup function, then I eliminated the whole config property. My Ollama server is running in Docker on the same machine like you. I would need to make some tweaks to make it convenient to use in my workflow, but I didn't get the error you described. Maybe it's been fixed in the month since you posted?