Model card synced with fastforks 7204e33f

README.md

@@ -52,15 +52,21 @@ Tested up to (portable, fp16, RTX 5060 Ti 16 GB): `efficientnet_b0.tv_in1k` 224x
 
 | model | time (ms): original fp32 | original | portable | kernels | kernels speed-up: vs original | vs portable | peak memory (MiB, allocated / reserved): original | portable | kernels |
 | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
-| `efficientnet_b0.tv_in1k` | 5.64 | 7.91 | 4.98 | 0.87 | 6.5x | 5.7x | 64 / 84 | 48 / 96 | 78 / 162 |
-| `efficientnet_b4.tv_in1k` | 11.09 | 15.66 | 10.0 | 2.19 | 5.1x | 4.6x | 147 / 174 | 89 / 114 | 109 / 178 |
+| `efficientnet_b0.tv_in1k` | 5.64 | 7.91 | 4.98 | 0.769 | 7.3x | 6.5x | 64 / 84 | 48 / 96 | 78 / 118 |
+| `efficientnet_b4.tv_in1k` | 11.09 | 15.66 | 10.0 | 1.999 | 5.5x | 5.0x | 147 / 174 | 89 / 114 | 111 / 180 |
 
 ### Startup and first call
 
 | model | load (s): original | portable | kernels | first call (s): original | portable | kernels |
 | --- | --- | --- | --- | --- | --- | --- |
-| `efficientnet_b0.tv_in1k` | 1.25 | 0.17 | 0.24† | 2.86 | 0.38 | 0.01 |
-| `efficientnet_b4.tv_in1k` | 0.41 | 0.37 | 0.53† | 1.72 | 0.06 | 0.01 |
+| `efficientnet_b0.tv_in1k` | 1.25 | 0.17 | 1.53 | 2.86 | 0.38 | 0.06 |
+| `efficientnet_b4.tv_in1k` | 0.41 | 0.37 | 1.94 | 1.72 | 0.06 | 0.07 |
+
+### torch.compile and TensorRT
+
+| model | input | portable ms | compile ms | compile s (cold) | TensorRT fp16 ms | ONNX + build s | kernels ms | peak MiB (alloc / reserved): portable | compile | TensorRT (device) |
+| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
+| `efficientnet_b0.tv_in1k` | 224x224 | 4.98 | 2.9 | 9.1 | 0.5 | 26 | 0.769 | 48 / 96 | 80 / 96 | 38 |
 
 ### Fixed-size speed and peak memory, all models
 
@@ -79,14 +85,14 @@ Hardware, protocol and caveats: [benchmarks](https://github.com/kleinicke/fastfo
 
 ### Parity with the original code
 
-Every variant is checked against golden outputs of the original code at its pinned commit. CPU, float32: max |diff| <= 0.002 vs the original. GPU, float16 (the default), portable and kernels: mean |diff| <= 0.02 px and p99 <= 0.08 px.
+Every variant is checked against golden outputs of the original code at its pinned commit. CPU, float32: max |diff| <= 0.002 vs the original. GPU, float16 (the default), portable and kernels: mean |diff| <= 0.02 and p99 <= 0.08 (absolute error of the output `logits`).
 
 ## Usage
 
 ```python
 from fastforks import classification
 
-model = classification.load("efficientnet_b0.tv_in1k")  # portable PyTorch on the GPU if there is one, else the CPU
+model = classification.load("efficientnet_b0.tv_in1k")  # mode="auto": the preferred backend from recorded measurements
 result = model("photo.jpg")
 result.top(5)  # top-5 classes
 result.backend  # what ran, e.g. Backend(mode='portable', device='cuda:0', dtype='float16')
@@ -97,3 +103,5 @@ Install with `pip install fastforks` (see <https://github.com/kleinicke/fastfork
 ## Kernels backend
 
 `classification.load("efficientnet_b0.tv_in1k", mode="kernels")` runs our CUDA kernels (`cudnn_hacc`, `fvm_conv`, `fvm_extra`, `fvm_mb`, `fvm_r3`). It needs a prebuilt kernel for your GPU architecture (downloaded from <https://huggingface.co/fastforks/kernels> on first use) and a recent NVIDIA driver; if there is none it raises `BackendUnavailableError` with the reason and never falls back silently. `mode="auto"` picks `kernels` when it can run and logs the choice.
+
+<!-- staged from fastforks 7204e33f (2026-11-05) -->