Hugging Face's logo

PrunaAI
/
llama-moe-LLaMA-MoE-v1-3_5B-4_16-HQQ-2bit-smashed

Text Generation
Transformers
llama_moe
pruna-ai
custom_code
Model card Files Community
1
llama-moe-LLaMA-MoE-v1-3_5B-4_16-HQQ-2bit-smashed / config.json
sharpenb's picture
sharpenb
Upload folder using huggingface_hub (#1)
bc591ca verified
raw
history blame
7.48 kB
{
"_name_or_path": "llama-moe/LLaMA-MoE-v1-3_5B-4_16",
"add_weight_norm": false,
"architectures": [
"LlamaMoEForCausalLM"
],
"auto_map": {
"AutoConfig": "llama-moe/LLaMA-MoE-v1-3_5B-4_16--configuration_llama_moe.LlamaMoEConfig",
"AutoModel": "llama-moe/LLaMA-MoE-v1-3_5B-4_16--modeling_llama_moe_hf.LlamaMoEModel",
"AutoModelForCausalLM": "llama-moe/LLaMA-MoE-v1-3_5B-4_16--modeling_llama_moe_hf.LlamaMoEForCausalLM"
},
"bos_token_id": 1,
"calculator_type": "UniversalCalculator",
"capacity_factor": 1.25,
"drop_tokens": true,
"dropped_padding": "zero",
"eos_token_id": 2,
"gate_add_noise": true,
"gate_balance_loss_weight": 0.01,
"gate_network": "mlp",
"gate_noise_epsilon": 0.01,
"gate_type": "TopKBalancedNoisyGate",
"gate_use_balance": true,
"gate_use_softmax": true,
"gates": "mlp",
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 11008,
"max_position_embeddings": 4096,
"model_type": "llama_moe",
"multiply_gate_scores": true,
"num_attention_heads": 32,
"num_experts": 16,
"num_hidden_layers": 32,
"num_key_value_heads": 32,
"num_selects": 4,
"pad_token_id": 0,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"score_scale_factor": 4.0,
"size_experts": [
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
],
[
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688,
688
]
],
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.41.2",
"use_cache": true,
"vocab_size": 32000
}