Add loadable model.safetensors + config for EMoE ViT-Base

README.md

@@ -1,9 +1,10 @@
 ---
-license: cc-by-4.0
-datasets:
-- ILSVRC/imagenet-1k
-- uoft-cs/cifar10
-- uoft-cs/cifar100
-metrics:
-- accuracy
----
+tags:
+- model_hub_mixin
+- pytorch_model_hub_mixin
+---
+
+This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration:
+- Code: [More Information Needed]
+- Paper: [More Information Needed]
+- Docs: [More Information Needed]

config.json

@@ -0,0 +1,18 @@
+{
+  "backbone_pretrained": false,
+  "moe_config": {
+    "alpha": 1.0,
+    "apply_to_patches_only": true,
+    "blocks": "last6",
+    "bottleneck": 192,
+    "experts": 8,
+    "freeze_backbone": true,
+    "ortho_lambda": 0.0005,
+    "r": 128,
+    "router_mode": "top1",
+    "tau": 1.0,
+    "unfreeze_layernorm": false
+  },
+  "num_classes": 1000,
+  "vit_model_name": "vit_base_patch16_224"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d3c017159c961a959e01cb09d594016198146b479049a36fd2ccb7c1f5bdf0a
+size 407673536