| { |
| "transformers_version": "5.9.0", |
| "architectures": null, |
| "output_hidden_states": false, |
| "return_dict": true, |
| "dtype": null, |
| "chunk_size_feed_forward": 0, |
| "is_encoder_decoder": false, |
| "id2label": { |
| "0": "LABEL_0", |
| "1": "LABEL_1" |
| }, |
| "label2id": { |
| "LABEL_0": 0, |
| "LABEL_1": 1 |
| }, |
| "problem_type": null, |
| "num_hidden_layers": 6, |
| "num_local_experts": 8, |
| "vocab_size": 1024, |
| "hidden_size": 128, |
| "intermediate_size": 128, |
| "head_dim": 32, |
| "num_attention_heads": 4, |
| "num_key_value_heads": 1, |
| "sliding_window": 128, |
| "tie_word_embeddings": false, |
| "hidden_act": "silu", |
| "initializer_range": 0.02, |
| "max_position_embeddings": 4096, |
| "rms_norm_eps": 1e-05, |
| "rope_parameters": { |
| "rope_type": "yarn", |
| "factor": 4.0, |
| "original_max_position_embeddings": 1024, |
| "beta_fast": 32.0, |
| "beta_slow": 1.0, |
| "truncate": false |
| }, |
| "attention_dropout": 0.0, |
| "num_experts_per_tok": 4, |
| "router_aux_loss_coef": 0.0, |
| "output_router_logits": false, |
| "use_cache": true, |
| "layer_types": [ |
| "sliding_attention", |
| "full_attention", |
| "sliding_attention", |
| "full_attention", |
| "sliding_attention", |
| "full_attention" |
| ], |
| "pad_token_id": 1000, |
| "bos_token_id": 1000, |
| "eos_token_id": 1001, |
| "attention_bias": true, |
| "_name_or_path": "", |
| "initial_context_length": 1024, |
| "experts_per_token": 4, |
| "swiglu_limit": 7.0, |
| "rope_theta": 150000.0, |
| "model_type": "gpt_oss", |
| "output_attentions": false |
| } |
