Upload folder using huggingface_hub

README.md

@@ -0,0 +1,10 @@
+# SPEED: Specialized Position Experts for Efficient Speculative Decoding
+
+This repository provides the model checkpoint for **SPEED**, a speculative decoding method proposed in the anonymous paper:
+
+> **SPEED: Specialized Position Experts for Efficient Speculative Decoding**
+
+## 📦 Files
+
+- `pytorch_model.bin` — model weights  
+- `config.json` — model configuration

config.json

@@ -0,0 +1,31 @@
+{
+  "architectures": [
+    "LlamaForCausalLMPosSEagle3"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 128000,
+  "draft_vocab_size": 32000,
+  "dtype": "bfloat16",
+  "eos_token_id": 128001,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 2048,
+  "mlp_bias": false,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 3,
+  "num_key_value_heads": 8,
+  "pad_token_id": 0,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "tie_word_embeddings": false,
+  "transformers_version": "4.56.2",
+  "use_cache": true,
+  "vocab_size": 128256
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95d85eb3e230c698cc1103dcb921ee3947013878bdf717992625e4fde316e62e
+size 1822894912