| { | |
| "_name_or_path": "VAIBHAV22334455/NOVA-3B", | |
| "architectures": [ | |
| "StableLmForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 50278, | |
| "eos_token_id": 50279, | |
| "head_dim": 256, | |
| "hidden_act": "silu", | |
| "hidden_activation": null, | |
| "hidden_dropout": 0.0, | |
| "hidden_size": 2560, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 6912, | |
| "layer_norm_eps": 1e-05, | |
| "max_position_embeddings": 4096, | |
| "model_type": "stablelm", | |
| "norm_eps": 1e-05, | |
| "num_attention_heads": 32, | |
| "num_experts_per_tok": 2, | |
| "num_hidden_layers": 32, | |
| "num_key_value_heads": 32, | |
| "num_local_experts": 8, | |
| "output_router_logits": false, | |
| "pad_token_id": 50279, | |
| "partial_rotary_factor": 0.25, | |
| "qk_layernorm": false, | |
| "rms_norm_eps": 1e-06, | |
| "rope_pct": 0.25, | |
| "rope_scaling": null, | |
| "rope_theta": 10000, | |
| "router_aux_loss_coef": 0.02, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "float16", | |
| "transformers_version": "4.42.4", | |
| "use_cache": false, | |
| "use_parallel_residual": false, | |
| "use_qkv_bias": false, | |
| "vocab_size": 50281 | |
| } | |