video/image: honor explicit Speed=off for companions + trim, probe explicit TE kernels, bench fidelity

Address the Codex review round on the video/quant work:

- Companion auto-quant now honors an explicit Speed=off. Both loaders already pin the DiT dense
  under an explicit off (bit-exact reference), but the unset text-encoder / VAE quant still promoted
  to auto and silently fp8/int8'd the companions, breaking the bit-exact request. An UNSET speed
  still auto-quantises; an explicit companion scheme still forces it.
- The HunyuanVideo joint-attention trim is a speed lever (it swaps to the fused SDPA kernel), so gate
  it on a non-off speed tier exactly like the adjacent attention-backend selection -- the off path
  keeps the stock dense-mask attention.
- Explicit torchao text-encoder modes (int8 / fp8_dynamic / nvfp4) now run the same kernel smoke
  test the auto ladder uses. They could clear the capability gate yet fail the real GEMM on a build
  where quantize_ wraps the encoder but the kernel is broken; the caster's try/except only covers the
  cast, not the first forward, so the load would report engaged then crash at generation. Now it
  falls back to dense. Layerwise fp8 has no torchao GEMM, so the probe is a no-op for it.
- The trim pre-hook's fallback restores the caller's original kwargs (it may have emptied the image
  stream / trimmed a text stream before failing), so the stock dense-mask path runs on exactly what
  it expects, matching the empty-prompt guard.
- video_speedmem_bench mirrors the loader: installs the Hunyuan trim before the backend set (gated on
  an active tier) and skips the auto int8 quant when it is the fp8-denied memory fallback and dense
  fits resident, so the shipped/auto rows measure what the loader actually runs.

Tests: TE explicit-mode kernel probe (+ layerwise-fp8 bypass), trim mid-trim restore, and loader-level
speed=off companion suppression + trim skip for both backends. 262 backend tests pass; ruff clean.

studio/backend/core/inference/diffusion_precision.py

@@ -253,6 +253,16 @@ def quantize_text_encoders(
         return None
     if not te_quant_supported(target, mode):
         return None
+    # Mirror select_te_quant_scheme's auto path: an EXPLICIT torchao TE mode (int8 / fp8_dynamic /
+    # nvfp4) can clear the capability gate above yet fail the real GEMM on a torchao/torch build
+    # where quantize_ wraps the encoder but the kernel is broken -- and the caster's try/except only
+    # catches the cast, not the first prompt-encoder forward. Run the same kernel smoke test the auto
+    # ladder uses so a failing kernel falls back to dense here instead of crashing at generation.
+    # Layerwise fp8 (no torchao GEMM) has no smoke scheme, so the probe is a no-op (True) for it.
+    device = str(getattr(target, "device", "cuda"))
+    if not _te_scheme_probe(mode, device):
+        _note(logger, f"text-encoder '{mode}' failed the kernel smoke test; staying dense")
+        return None
     if mode == TE_QUANT_INT8:
         first, last = skip  # type: ignore[misc]