Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4.1, MiniMax-H3, Gemma 4, FLUX and more.
Python agentaichatgptdeepseekfine-tuninggemmaimage-generationllamallmllmsopenaipythonqwenreinforcement-learningself-hostedstable-diffusiontext-to-speechttsuiunsloth
Hub token identity tests: let the llama-backend stubs take the caller's token after #13005 (#13022)
Shubhankar Tripathya8ff841db8a3
diff --git a/studio/backend/tests/test_hub_token_caller_identity.py b/studio/backend/tests/test_hub_token_caller_identity.pyindex 3ccad4f97f7..471c2decb3f 100644--- a/studio/backend/tests/test_hub_token_caller_identity.py+++ b/studio/backend/tests/test_hub_token_caller_identity.py@@ -126,7 +126,7 @@ def _st_resolver(monkeypatch, **overrides) -> None:stub it cares about and the rest stays out of the way."""stubs = {- "_llama_backend_active": lambda _m: False,+ "_llama_backend_active": lambda _m, _token = None: False,"_local_sentence_transformer_is_present": lambda _m: False,"_st_weight_source": lambda *_a, **_k: None,}@@ -137,7 +137,7 @@ def _st_resolver(monkeypatch, **overrides) -> None:def _gguf_resolver(monkeypatch, **overrides) -> None:"""The llama-server preamble: six stubs is what it takes to reach the candidate lookup."""stubs = {- "_llama_backend_active": lambda _m: True,+ "_llama_backend_active": lambda _m, _token = None: True,"_llama_runtime_available": lambda: True,"_resolves_as_local_gguf": lambda _m: False,"_local_gguf_backend_error": lambda _m: None,@@ -2179,6 +2179,25 @@ def test_the_embedding_resolver_does_not_probe_before_a_cache_lookup(monkeypatchassert probes["n"] == 0, "a local sentence-transformers hit still probed the Hub"+def test_the_backend_probe_is_asked_with_the_callers_token(monkeypatch):+ """Which backend serves a private repo depends on files only the caller's token can read (#13005)."""+ seen = []++ def _probe(model, token = None):+ seen.append((model, token))+ return False++ _st_resolver(+ monkeypatch,+ _llama_backend_active = _probe,+ _local_sentence_transformer_is_present = lambda _m: True,+ )++ settings_routes._resolve_embedding_model_plan("acme/private-st", "hf_caller")++ assert seen == [("acme/private-st", "hf_caller")]++def test_the_scan_is_refused_before_it_expands_its_targets(monkeypatch):"""The per-target check inside the loop cannot stand in for this one. Target expansionresolves the adapter's base and its native-audio dependencies, which reads the cache and