hf-transformers-bot's picture
Update tiny models for GraniteMoeHybridForCausalLM
ee5241d verified
Raw
History Blame Contribute Delete
1.48 kB
{
"architectures": [
"GraniteMoeHybridForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"attention_multiplier": 1.0,
"attn_layer_indices": [
1,
3
],
"attn_rotary_emb": 8,
"bos_token_id": 0,
"dtype": "float32",
"embedding_multiplier": 1.0,
"eos_token_id": 0,
"hidden_act": "silu",
"hidden_size": 32,
"initializer_range": 0.02,
"intermediate_size": 64,
"is_decoder": true,
"layer_types": [
"linear_attention",
"full_attention",
"linear_attention",
"full_attention"
],
"logits_scaling": 1.0,
"mamba_chunk_size": 16,
"mamba_conv_bias": true,
"mamba_d_conv": 4,
"mamba_d_head": 4,
"mamba_d_state": 16,
"mamba_expand": 2,
"mamba_n_groups": 1,
"mamba_n_heads": 16,
"mamba_proj_bias": false,
"max_position_embeddings": 512,
"model_type": "granitemoehybrid",
"num_attention_heads": 4,
"num_experts_per_tok": 2,
"num_hidden_layers": 4,
"num_key_value_heads": 2,
"num_local_experts": 8,
"output_router_logits": false,
"pad_token_id": 0,
"position_embedding_type": null,
"residual_multiplier": 1.0,
"rms_norm_eps": 1e-06,
"rope_parameters": {
"rope_theta": 10000.0,
"rope_type": "default"
},
"router_aux_loss_coef": 0.001,
"shared_intermediate_size": 174,
"tie_word_embeddings": false,
"time_step_limit": [
0.0,
{
"__float__": "Infinity"
}
],
"transformers_version": "5.16.0.dev0",
"use_cache": true,
"vocab_size": 49156
}