Cannot use the baml playground and a local client while connected to a machine via ssh

[this-josh]

From VS Code if I do Remote-SSH: Connect to Host...

Then navigate to my baml folder and set the client to

client<llm> Ollama {
  provider "openai-generic"
  options {
    base_url "http://localhost:11434/v1"
    model llama3.3:latest
  }
}

Now pressing run in the playground gives this error

vaibhav_resume
Unable to run
Unspecified error code: 2 reqwest::Error { kind: Request, source: "JsValue(TypeError: Failed to fetch\nTypeError: Failed to fetch\n    at __wbg_fetch_1e4e8ed1f64c7e28 (https://vscode-remote+ssh-002dremote-002b2a0c-003a5bc0-003a40-003a2e26-003ac27-003a598a-003ab050-003a306d.vscode-resource.vscode-cdn.net/home/jak121/.vscode-server/extensions/boundary.baml-extension-0.72.1/web-panel/dist/assets/baml_schema_build.js:2758:17)\n    at baml_schema_build.wasm.reqwest::wasm::client::js_fetch::h702657c4e8b70698 (wasm://wasm/baml_schema_build.wasm-02c1bb5a:wasm-function[8226]:0x785c64)\n    at baml_schema_build.wasm.reqwest::wasm::client::fetch::{{closure}}::h9dcb23421a40299d (wasm://wasm/baml_schema_build.wasm-02c1bb5a:wasm-function[402]:0x207ed0)\n    at baml_schema_build.wasm.baml_runtime::internal::llm_client::primitive::request::make_request::{{closure}}::h750be6e03bf00365 (wasm://wasm/baml_schema_build.wasm-02c1bb5a:wasm-function[424]:0x21c239)\n    at baml_schema_build.wasm.<baml_runtime::internal::llm_client::primitive::LLMPrimitiveProvider as baml_runtime::internal::llm_client::traits::WithStreamable>::stream::{{closure}}::h8e98f766dfb8589c (wasm://wasm/baml_schema_build.wasm-02c1bb5a:wasm-function[143]:0x341b0)\n    at baml_schema_build.wasm.baml_runtime::BamlRuntime::run_test::{{closure}}::{{closure}}::{{closure}}::h23629bb0be43580f (wasm://wasm/baml_schema_build.wasm-02c1bb5a:wasm-function[148]:0x53d71)\n    at baml_schema_build.wasm.wasm_bindgen_futures::future_to_promise::{{closure}}::{{closure}}::h0ef435efcf095f59 (wasm://wasm/baml_schema_build.wasm-02c1bb5a:wasm-function[264]:0x1690a7)\n    at baml_schema_build.wasm.wasm_bindgen_futures::queue::QueueState::run_all::h783ac55f5c0fa9d6 (wasm://wasm/baml_schema_build.wasm-02c1bb5a:wasm-function[4859]:0x689a5f)\n    at baml_schema_build.wasm.wasm_bindgen_futures::queue::Queue::new::{{closure}}::hc075082cb8ddbf92 (wasm://wasm/baml_schema_build.wasm-02c1bb5a:wasm-function[14598]:0x8237ff)\n    at baml_schema_build.wasm.<dyn core::ops::function::FnMut<(A,)>+Output = R as wasm_bindgen::closure::WasmClosure>::describe::invoke::h2aa238002d130c19 (wasm://wasm/baml_schema_build.wasm-02c1bb5a:wasm-function[14597]:0x8237f1))" }

If I take the generated curl request

curl -X POST 'http://localhost:11434/v1/chat/completions' -H "content-type: application/json" -d "{
  \"model\": \"llama3.3:latest\",
  \"messages\": [
    {
      \"role\": \"system\",
      \"content\": \"Extract from this content:\nVaibhav Gupta\[email protected]\n\nExperience:\n- Founder at BoundaryML\n- CV Engineer at Google\n- CV Engineer at Microsoft\n\nSkills:\n- Rust\n- C++\n\nAnswer in JSON using this schema:\n{\n  name: string,\n  email: string,\n  experience: string[],\n  skills: string[],\n}\"
    }
  ],
  \"stream\": true
}"

I can run this fine in my VS code (ssh) terminal, also if I open VS code up on the remote machine and try using the playground that works fine.

So this issue seems to be to do with BAML playground while in a VS SSH instance

[aaronvg]

Try changing vscode settings:

baml.enablePlaygroundProxy = false

[this-josh]

HI @aaronvg thanks for getting back to me, I've just tried this now and it hasn't worked, with what seems to be the same error message

Unspecified error code: 2 reqwest::Error { kind: Request, source: "JsValue(TypeError: Failed to fetch\nTypeError: Failed to fetch\n    at __wbg_fetch_c5d6726a1da3618f (https://vscode-remote+ssh-002dremote-002bcv-002diits-002dw01-002edept-002eic-002eac-002euk.vscode-resource.vscode-cdn.net/home/jak121/.vscode-server/extensions/boundary.baml-extension-0.73.4/web-panel/dist/assets/baml_schema_build.js:2758:17)\n    at baml_schema_build.wasm.reqwest::wasm::client::js_fetch::h0af2ae822c41a29d (wasm://wasm/baml_schema_build.wasm-02d549f2:wasm-function[8513]:0x7c37e4)\n    at baml_schema_build.wasm.reqwest::wasm::client::fetch::{{closure}}::h22fbbcbe205f38bf (wasm://wasm/baml_schema_build.wasm-02d549f2:wasm-function[403]:0x20c3da)\n    at baml_schema_build.wasm.baml_runtime::internal::llm_client::primitive::request::make_request::{{closure}}::hfbb3a3d328000cd5 (wasm://wasm/baml_schema_build.wasm-02d549f2:wasm-function[426]:0x221752)\n    at baml_schema_build.wasm.<baml_runtime::internal::llm_client::primitive::LLMPrimitiveProvider as baml_runtime::internal::llm_client::traits::WithStreamable>::stream::{{closure}}::h8c8151155b30a105 (wasm://wasm/baml_schema_build.wasm-02d549f2:wasm-function[143]:0x3461b)\n    at baml_schema_build.wasm.baml_runtime::BamlRuntime::run_test::{{closure}}::{{closure}}::{{closure}}::h4ab0d045b542953a (wasm://wasm/baml_schema_build.wasm-02d549f2:wasm-function[151]:0x5f9c7)\n    at baml_schema_build.wasm.wasm_bindgen_futures::future_to_promise::{{closure}}::{{closure}}::h5f746656074539c4 (wasm://wasm/baml_schema_build.wasm-02d549f2:wasm-function[268]:0x17084d)\n    at baml_schema_build.wasm.wasm_bindgen_futures::queue::QueueState::run_all::h783ac55f5c0fa9d6 (wasm://wasm/baml_schema_build.wasm-02d549f2:wasm-function[5031]:0x6be960)\n    at baml_schema_build.wasm.wasm_bindgen_futures::queue::Queue::new::{{closure}}::hc075082cb8ddbf92 (wasm://wasm/baml_schema_build.wasm-02d549f2:wasm-function[14967]:0x863cba)\n    at baml_schema_build.wasm.<dyn core::ops::function::FnMut<(A,)>+Output = R as wasm_bindgen::closure::WasmClosure>::describe::invoke::h2aa238002d130c19 (wasm://wasm/baml_schema_build.wasm-02d549f2:wasm-function[14966]:0x863cac))" }

Request options: {"model":"qwen2.5:14b-instruct-q6_K"}

[aaronvg]

Did you set the OLLAMA_ORIGINS=* env var?

Theres a var that allows ollama to bypass cors

[phiwi]

I am experience exactly the same situation. Starting Ollama with

singularity run --env  OLLAMA_ORIGINS=*  OLLAMA_KEEP_ALIVE=1h  --nv  ollama.sif 

didn't help.

[aaronvg]

If you join our discord I'm happy to debug with you in our office hours with screenshare https://discord.com/invite/yzaTpQ3tdT