optimum-neuron-cache
/
neuronxcc-2.15.128.0+56dc5a86
/0_REGISTRY
/0.0.25.dev0
/inference
/distilbert
/hf-internal-testing
/tiny-random-DistilBertModel
/b0d504a6e98c193d8658.json
{"activation": "gelu", "architectures": ["DistilBertModel"], "attention_dropout": 0.1, "dim": 32, "dropout": 0.1, "hidden_act": "gelu", "hidden_dim": 37, "initializer_range": 0.02, "max_position_embeddings": 512, "model_type": "distilbert", "n_heads": 4, "n_layers": 5, "neuron": {"auto_cast": null, "auto_cast_type": null, "compiler_type": "neuronx-cc", "compiler_version": "2.15.128.0+56dc5a86", "disable_fallback": false, "disable_fast_relayout": false, "dynamic_batch_size": true, "inline_weights_to_neff": true, "optlevel": "2", "output_attentions": false, "output_hidden_states": false, "static_batch_size": 1, "static_sequence_length": 32}, "qa_dropout": 0.1, "seq_classif_dropout": 0.2, "sinusoidal_pos_embds": false, "task": "token-classification", "vocab_size": 1124} |