config.json

{
  "_name_or_path": "./llava-calm2-siglip",
  "architectures": [
    "LlavaForConditionalGeneration"
  ],
  "eos_token_id": 0,
  "ignore_index": -100,
  "image_token_index": 65000,
  "model_type": "llava",
  "pad_token_id": 1,
  "projector_hidden_act": "gelu",
  "text_config": {
    "_name_or_path": "cyberagent/calm2-7b-chat",
    "architectures": [
      "LlamaForCausalLM"
    ],
    "bos_token_id": 0,
    "eos_token_id": 0,
    "max_position_embeddings": 32768,
    "model_type": "llama",
    "pad_token_id": 1,
    "rope_theta": 500000,
    "torch_dtype": "bfloat16",
    "vocab_size": 65024
  },
  "torch_dtype": "bfloat16",
  "transformers_version": "4.40.0",
  "vision_config": {
    "hidden_act": "gelu_pytorch_tanh",
    "hidden_size": 1152,
    "image_size": 384,
    "intermediate_size": 4304,
    "layer_norm_eps": 1e-06,
    "model_type": "siglip_vision_model",
    "num_attention_heads": 16,
    "num_hidden_layers": 27,
    "patch_size": 14
  },
  "vision_feature_layer": -2,
  "vision_feature_select_strategy": "default"
}