{ "architectures": [ "GPT2LMHeadModel" ], "model_type": "gpt2", "vocab_size": 4096, "n_positions": 256, "n_ctx": 256, "n_embd": 384, "n_layer": 6, "n_head": 6, "n_inner": 1536, "activation_function": "gelu_new", "layer_norm_epsilon": 1e-05, "resid_pdrop": 0.0, "embd_pdrop": 0.0, "attn_pdrop": 0.0, "initializer_range": 0.02, "tie_word_embeddings": true, "bos_token_id": 4095, "eos_token_id": 4095 }