| { | |
| "_name_or_path": "/home/aiops/yangph/checkpoints/glide/checkpoint/experiments/vicuna-13b-16k.slim6b.glide.A100.tp1dp1.zero1.v1.0.sinkpi-slicing.longv2-32k-5e-6.lig-ce.no-mask.longsft-5e-6.s42/checkpoint-last", | |
| "architectures": [ | |
| "LlamaGlideDecoderLayer" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 5120, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 13824, | |
| "max_position_embeddings": 4096, | |
| "max_sequence_length": 16384, | |
| "mlp_bias": false, | |
| "model_type": "llama", | |
| "num_attention_heads": 40, | |
| "num_hidden_layers": 40, | |
| "num_key_value_heads": 40, | |
| "pad_token_id": 2, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": { | |
| "factor": 4.0, | |
| "rope_type": "linear", | |
| "type": "linear" | |
| }, | |
| "rope_theta": 10000.0, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.46.3", | |
| "use_cache": true, | |
| "vocab_size": 32000 | |
| } | |