lm-evaluation-harness Model Context Protocol (MCP) Server

A framework for few-shot evaluation of language models.

14,301 stars

Python 104 files · ~126,916 tokens #evaluation-framework #language-model #transformer

lm-evaluation-harness MCP Server

MCP server for EleutherAI/lm-evaluation-harness

Ready-to-Use Client Configurations

{
  "mcpServers": {
    "lm-evaluation-harness": {
      "command": "npx",
      "args": [
        "-y",
        "@EleutherAI/lm-evaluation-harness"
      ]
    }
  }
}

{
  "mcpServers": {
    "lm-evaluation-harness": {
      "command": "npx",
      "args": [
        "-y",
        "@EleutherAI/lm-evaluation-harness"
      ]
    }
  }
}

{
  "mcpServers": {
    "lm-evaluation-harness": {
      "command": "npx",
      "args": [
        "-y",
        "@EleutherAI/lm-evaluation-harness"
      ]
    }
  }
}

{
  "mcpServers": {
    "lm-evaluation-harness": {
      "command": "npx",
      "args": [
        "-y",
        "@EleutherAI/lm-evaluation-harness"
      ],
      "disabled": false,
      "autoApprove": []
    }
  }
}

{
  "mcpServers": {
    "lm-evaluation-harness": {
      "command": "npx",
      "args": [
        "-y",
        "@EleutherAI/lm-evaluation-harness"
      ]
    }
  }
}

docker run -i --rm mcp/lm-evaluation-harness