Upload config.json with huggingface_hub

config.json

@@ -0,0 +1,61 @@
+{
+  "architectures": [
+    "YiVLForConditionalGeneration"
+  ],
+  "ignore_index": -100,
+  "image_aspect_ratio": "resize",
+  "image_token_index": 9,
+  "mlp_post_norm": false,
+  "mlp_post_norm_element_affine": false,
+  "mm_projector": "resampler",
+  "mm_use_im_start_end": true,
+  "model_type": "yi_vl",
+  "pad_token_id": 2,
+  "projector_hidden_act": "gelu",
+  "text_config": {
+    "aux_loss_alpha": 0.001,
+    "first_k_dense_replace": 0,
+    "hidden_size": 2560,
+    "intermediate_size": 13568,
+    "max_position_embeddings": 65536,
+    "model_type": "deepseek",
+    "moe_grouped_gemm": null,
+    "moe_intermediate_size": 1664,
+    "moe_layer_freq": 1,
+    "n_routed_experts": 64,
+    "n_shared_experts": 2,
+    "norm_topk_prob": true,
+    "num_attention_heads": 20,
+    "num_experts_per_tok": 6,
+    "num_hidden_layers": 28,
+    "num_key_value_heads": 20,
+    "rope_theta": 5000000,
+    "scoring_func": "softmax",
+    "seq_aux": true,
+    "vocab_size": 100352
+  },
+  "tokens_per_image": [
+    128,
+    256
+  ],
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.42.4",
+  "vision_config": {
+    "_flash_attn_2_enabled": true,
+    "architectures": [
+      "CustomSiglipVisionModel"
+    ],
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 1152,
+    "image_size": 980,
+    "intermediate_size": 4304,
+    "layer_norm_eps": 1e-06,
+    "model_type": "custom_siglip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 27,
+    "patch_size": 14,
+    "torch_dtype": "float32"
+  },
+  "vision_feature_layer": -1,
+  "vision_feature_select_strategy": "full"
+}