| { | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "auto_map": { | |
| "AutoConfig": "configuration_teleflm.TeleFLMConfig", | |
| "AutoModel": "modeling_teleflm.TeleFLMModel", | |
| "AutoModelForCausalLM": "modeling_teleflm.TeleFLMForCausalLM" | |
| }, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "hidden_act": "silu", | |
| "hidden_size": 20480, | |
| "initializer_range": 0.02, | |
| "input_mult": 1.0, | |
| "intermediate_size": 98304, | |
| "max_position_embeddings": 4096, | |
| "model_type": "TeleFLM", | |
| "mup_scale_factor": 80.0, | |
| "num_attention_heads": 160, | |
| "num_hidden_layers": 140, | |
| "num_key_value_heads": 160, | |
| "output_mult": 2.0, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": null, | |
| "rope_theta": 10000, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.40.2", | |
| "use_cache": true, | |
| "use_mup": true, | |
| "vocab_size": 81920 | |
| } | |