config.json

{
  "architectures": [
    "MOLMoEForCausalLM"
  ],
  "auto_map": {
    "AutoConfig": "config_molmoe.MolmoeConfig",
    "AutoModelForCausalLM": "modeling_molmoe.MOLMoEForCausalLM"
  },
  "clip_qkv": null,
  "embedding_size": 152064,
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 1024,
  "layer_norm_eps": 1e-05,
  "max_position_embeddings": 4096,
  "model_type": "molmoe",
  "num_attention_heads": 16,
  "num_hidden_layers": 16,
  "num_key_value_heads": 16,
  "qkv_bias": false,
  "rope_theta": 10000.0,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.43.3",
  "use_cache": true,
  "use_position_ids": true,
  "vocab_size": 50304,
  "weight_tying": false
}