README: point to aws-neuron/vit-encoder-neuron-kernels (repo moved to aws-neuron org; docs only)
Browse files
README.md
CHANGED
|
@@ -23,7 +23,7 @@ tags:
|
|
| 23 |
implementation this was verified against, and rel_l2 4.2e-2 vs the fp32 torch encoder (the
|
| 24 |
model's own bf16-matmul sensitivity).
|
| 25 |
|
| 26 |
-
> **See also: [`
|
| 27 |
> a generalized version of these kernels. It adds batching (any batch size), LNC=1 (8-core) support,
|
| 28 |
> fused residual + LayerNorm epilogues, a SwiGLU FFN kernel, both RoPE pairings, and a DINOv3 layer
|
| 29 |
> alongside a V-JEPA2 one. This repo is unchanged and keeps its verified 16-frame / batch-1 path.
|
|
@@ -45,7 +45,7 @@ import torch
|
|
| 45 |
from transformers import AutoModel, KernelConfig
|
| 46 |
|
| 47 |
kernel_config = KernelConfig({
|
| 48 |
-
"VJEPA2Encoder": "
|
| 49 |
})
|
| 50 |
model = AutoModel.from_pretrained(
|
| 51 |
"facebook/vjepa2-vitl-fpc64-256",
|
|
|
|
| 23 |
implementation this was verified against, and rel_l2 4.2e-2 vs the fp32 torch encoder (the
|
| 24 |
model's own bf16-matmul sensitivity).
|
| 25 |
|
| 26 |
+
> **See also: [`aws-neuron/vit-encoder-neuron-kernels`](https://huggingface.co/kernels/aws-neuron/vit-encoder-neuron-kernels)**,
|
| 27 |
> a generalized version of these kernels. It adds batching (any batch size), LNC=1 (8-core) support,
|
| 28 |
> fused residual + LayerNorm epilogues, a SwiGLU FFN kernel, both RoPE pairings, and a DINOv3 layer
|
| 29 |
> alongside a V-JEPA2 one. This repo is unchanged and keeps its verified 16-frame / batch-1 path.
|
|
|
|
| 45 |
from transformers import AutoModel, KernelConfig
|
| 46 |
|
| 47 |
kernel_config = KernelConfig({
|
| 48 |
+
"VJEPA2Encoder": "aws-neuron/vjepa2-neuron-kernels:NeuronVJEPA2Encoder",
|
| 49 |
})
|
| 50 |
model = AutoModel.from_pretrained(
|
| 51 |
"facebook/vjepa2-vitl-fpc64-256",
|