Upload folder using huggingface_hub

config.json

@@ -0,0 +1,44 @@
+{
+  "architectures": [
+    "DeepseekV3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "aux_loss_alpha": 0.001,
+  "bos_token_id": 0,
+  "eos_token_id": 1,
+  "first_k_dense_replace": 2,
+  "head_dim": 16,
+  "hidden_act": "silu",
+  "hidden_size": 32,
+  "initializer_range": 0.02,
+  "intermediate_size": 37,
+  "kv_lora_rank": 16,
+  "max_position_embeddings": 512,
+  "model_type": "deepseek_v3",
+  "moe_intermediate_size": 12,
+  "n_group": 2,
+  "n_routed_experts": 8,
+  "n_shared_experts": 1,
+  "norm_topk_prob": true,
+  "num_attention_heads": 4,
+  "num_experts_per_tok": 2,
+  "num_hidden_layers": 5,
+  "num_key_value_heads": 4,
+  "pretraining_tp": 1,
+  "q_lora_rank": 32,
+  "qk_head_dim": 48,
+  "qk_nope_head_dim": 32,
+  "qk_rope_head_dim": 16,
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "routed_scaling_factor": 2.5,
+  "tie_word_embeddings": false,
+  "topk_group": 1,
+  "torch_dtype": "float32",
+  "transformers_version": "4.49.0.dev0",
+  "use_cache": true,
+  "v_head_dim": 32,
+  "vocab_size": 99
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 0,
+  "eos_token_id": 1,
+  "transformers_version": "4.49.0.dev0"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1005295dc5ef4163a98870643f1911abd96da33661c5f53a7b0c9a77671290d8
+size 526688