Aakash098/ShinchanAI-small

1{2    "architectures": [3        "Gemma2ForCausalLM"4    ],5    "attention_bias": false,6    "attention_dropout": 0.0,7    "attn_logit_softcapping": 50.0,8    "bos_token_id": 2,9    "cache_implementation": "hybrid",10    "torch_dtype": "float16",11    "eos_token_id": 1,12    "final_logit_softcapping": 30.0,13    "head_dim": 256,14    "hidden_act": "gelu_pytorch_tanh",15    "hidden_activation": "gelu_pytorch_tanh",16    "hidden_size": 2304,17    "initializer_range": 0.02,18    "intermediate_size": 9216,19    "layer_types": [20        "sliding_attention",21        "full_attention",22        "sliding_attention",23        "full_attention",24        "sliding_attention",25        "full_attention",26        "sliding_attention",27        "full_attention",28        "sliding_attention",29        "full_attention",30        "sliding_attention",31        "full_attention",32        "sliding_attention",33        "full_attention",34        "sliding_attention",35        "full_attention",36        "sliding_attention",37        "full_attention",38        "sliding_attention",39        "full_attention",40        "sliding_attention",41        "full_attention",42        "sliding_attention",43        "full_attention",44        "sliding_attention",45        "full_attention"46    ],47    "max_position_embeddings": 8192,48    "model_type": "gemma2",49    "num_attention_heads": 8,50    "num_hidden_layers": 26,51    "num_key_value_heads": 4,52    "pad_token_id": 0,53    "query_pre_attn_scalar": 256,54    "rms_norm_eps": 1e-06,55    "rope_parameters": {56        "rope_theta": 10000.0,57        "rope_type": "default"58    },59    "sliding_window": 4096,60    "tie_word_embeddings": true,61    "unsloth_version": "2026.3.3",62    "use_bidirectional_attention": null,63    "use_cache": false,64    "vocab_size": 25600065}