hf-transformers-bot's picture
Update tiny models for RecurrentGemmaModel
2ee1787 verified
Raw
History Blame Contribute Delete
941 Bytes
{
"architectures": [
"RecurrentGemmaModel"
],
"attention_bias": false,
"attention_dropout": 0.0,
"attention_window_size": 2048,
"block_types": [
"recurrent",
"attention"
],
"bos_token_id": 2,
"conv1d_width": 4,
"dtype": "float32",
"eos_token_id": 1,
"final_w_init_variance_scale": 1.0,
"head_dim": 16,
"hidden_activation": "gelu_pytorch_tanh",
"hidden_size": 32,
"intermediate_size": 32,
"logits_soft_cap": 30.0,
"lru_width": 32,
"model_type": "recurrent_gemma",
"num_attention_heads": 2,
"num_hidden_layers": 2,
"num_key_value_heads": 2,
"pad_token_id": 0,
"partial_rotary_factor": 0.5,
"rms_norm_eps": 1e-06,
"rope_parameters": {
"partial_rotary_factor": 0.5,
"rope_theta": 10000.0,
"rope_type": "default"
},
"tie_word_embeddings": false,
"transformers_version": "5.16.0.dev0",
"use_cache": true,
"vocab_size": 256000,
"w_init_variance_scale": 0.01
}