| { | |
| "architectures": [ | |
| "SolarOpenForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 1, | |
| "dtype": "float32", | |
| "eos_token_id": 2, | |
| "first_k_dense_replace": 0, | |
| "head_dim": 16, | |
| "hidden_act": "silu", | |
| "hidden_size": 64, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 10240, | |
| "max_position_embeddings": 131072, | |
| "model_type": "solar_open", | |
| "moe_intermediate_size": 16, | |
| "n_group": 1, | |
| "n_routed_experts": 4, | |
| "n_shared_experts": 1, | |
| "norm_topk_prob": true, | |
| "num_attention_heads": 4, | |
| "num_experts_per_tok": 2, | |
| "num_hidden_layers": 2, | |
| "num_key_value_heads": 2, | |
| "pad_token_id": 2, | |
| "partial_rotary_factor": 1.0, | |
| "rms_norm_eps": 1e-05, | |
| "rope_parameters": { | |
| "factor": 2.0, | |
| "original_max_position_embeddings": 65536, | |
| "partial_rotary_factor": 1.0, | |
| "rope_theta": 1000000, | |
| "rope_type": "yarn", | |
| "type": "yarn" | |
| }, | |
| "routed_scaling_factor": 1.0, | |
| "tie_word_embeddings": false, | |
| "topk_group": 1, | |
| "transformers_version": "5.3.0.dev0", | |
| "use_cache": true, | |
| "vocab_size": 196608, | |
| "transformers.js_config": { | |
| "use_external_data_format": { | |
| "model.onnx": 1, | |
| "model_fp16.onnx": 1 | |
| }, | |
| "kv_cache_dtype": { | |
| "q4f16": "float16", | |
| "fp16": "float16" | |
| } | |
| } | |
| } |