config.json

{
  "model_type": "ArlowGPT",
  "clip_config": {
    "architectures": ["CLIPModel"],
    "hidden_size": 1024,
    "intermediate_size": 4096,
    "num_hidden_layers": 24,
    "num_attention_heads": 16,
    "initializer_range": 0.02,
    "layer_norm_eps": 1e-05,
    "vision_patch_size": 14,
    "vocab_size": 49408,
    "projection_dim": 768,
    "max_position_embeddings": 77,
    "image_size": 224
  },
  "clip_model_name": "yuchenxie/ArlowGPT-VLM-Untrained",
  "gpt2_config": {
    "architectures": ["GPT2LMHeadModel"],
    "vocab_size": 50257,
    "n_positions": 1024,
    "n_ctx": 1024,
    "n_embd": 768,
    "n_layer": 12,
    "n_head": 12,
    "attn_pdrop": 0.1,
    "resid_pdrop": 0.1,
    "embd_pdrop": 0.1,
    "layer_norm_epsilon": 1e-05,
    "initializer_range": 0.02
  },
  "gpt2_model_name": "yuchenxie/ArlowGPT-VLM-Untrained",
  "projection_dim": 768,
  "transformers_version": "4.47.0",
  "vocab_size": 50257
}