| { | |
| "_name_or_path": "/glusterfs/dfs-gfs-dist/hallepat/rwkv-model", | |
| "architectures": [ | |
| "NGMEForCausalLM" | |
| ], | |
| "bos_token_id": 1, | |
| "eos_token_id": 0, | |
| "ffn_dim": 512, | |
| "hidden_act": "silu", | |
| "hidden_size": 768, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 11008, | |
| "max_position_embeddings": 2048, | |
| "model_type": "ngme", | |
| "num_attention_heads": 6, | |
| "num_hidden_layers": 8, | |
| "num_key_value_heads": 6, | |
| "pad_token_id": 0, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-06, | |
| "rope_scaling": null, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.31.0", | |
| "unk_idx": 1, | |
| "unk_token_id": 1, | |
| "use_cache": true, | |
| "use_flash_attn": false, | |
| "vocab_size": 36484 | |
| } | |