| { | |
| "_name_or_path": "/home/rllaskowski_a100/mot-training/experiments/mot_medium_32_8/checkpoint-100000", | |
| "activation_function": "gelu_new", | |
| "architectures": [ | |
| "MoTLMHeadModel" | |
| ], | |
| "attn_pdrop": 0.1, | |
| "auto_map": { | |
| "AutoConfig": "configuration_mot.MoTConfig", | |
| "AutoModelForCausalLM": "modeling_mot.MoTLMHeadModel", | |
| "AutoModel": "modeling_mot.MoTModel" | |
| }, | |
| "bos_token_id": 50256, | |
| "embd_pdrop": 0.1, | |
| "emit_softmax_over_experts": false, | |
| "eos_token_id": 50256, | |
| "expert_size": 256, | |
| "group_size": 32, | |
| "init_scale": 1.0, | |
| "initializer_range": 0.02, | |
| "layer_norm_epsilon": 1e-05, | |
| "model_type": "mot", | |
| "n_embd": 512, | |
| "n_expert": 256, | |
| "n_head": 8, | |
| "n_inner": 65536, | |
| "n_layer": 8, | |
| "n_positions": 1024, | |
| "reorder_and_upcast_attn": false, | |
| "resid_pdrop": 0.1, | |
| "scale_attn_by_inverse_layer_idx": false, | |
| "scale_attn_weights": true, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.42.1", | |
| "use_cache": true, | |
| "use_discrete_routing": false, | |
| "vocab_size": 50257 | |
| } | |