| { |
| "_name_or_path": "optimum-internal-testing/tiny-random-decilm", |
| "architectures": [ |
| "DeciLMForCausalLM" |
| ], |
| "attention_bias": false, |
| "auto_map": { |
| "AutoConfig": "configuration_decilm.DeciLMConfig", |
| "AutoModelForCausalLM": "modeling_decilm.DeciLMForCausalLM" |
| }, |
| "bos_token_id": 1, |
| "eos_token_id": 2, |
| "hidden_act": "silu", |
| "hidden_size": 256, |
| "initializer_range": 0.02, |
| "intermediate_size": 112, |
| "max_position_embeddings": 128, |
| "model_type": "deci", |
| "num_attention_heads": 4, |
| "num_hidden_layers": 4, |
| "num_key_value_heads": 4, |
| "num_key_value_heads_per_layer": [ |
| 1, |
| 2, |
| 2, |
| 1 |
| ], |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": { |
| "factor": 2.0, |
| "type": "dynamic" |
| }, |
| "rope_theta": 10000.0, |
| "tie_word_embeddings": false, |
| "tokenizer_class": "LlamaTokenizer", |
| "torch_dtype": "float32", |
| "transformers_version": "4.44.2", |
| "use_bfloat16": true, |
| "use_cache": true, |
| "vocab_size": 32000 |
| } |
|
|
