tinyllama_mole_dpo_ep3 / config.json
hushell's picture
Training in progress, step 100
4f533ca verified
{
"_name_or_path": "ondevicellm/tinyllama_mole_sft_ultrachat_ep3",
"adapter_alpha": 0.01,
"adapter_dim": 16,
"architectures": [
"MixtralMoleForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"auto_map": {
"AutoConfig": "ondevicellm/tinyllama_mole_sft_ultrachat_ep3--configuration_mixtral_mole.MixtralMoleConfig",
"AutoModel": "ondevicellm/tinyllama_mole_sft_ultrachat_ep3--modeling_mixtral_mole.MixtralMoleModel",
"AutoModelForCausalLM": "ondevicellm/tinyllama_mole_sft_ultrachat_ep3--modeling_mixtral_mole.MixtralMoleForCausalLM"
},
"bos_token_id": 1,
"eos_token_id": 2,
"hidden_act": "silu",
"hidden_size": 2048,
"initializer_range": 0.02,
"intermediate_size": 5632,
"max_position_embeddings": 2048,
"model_type": "mixtralmole",
"num_attention_heads": 32,
"num_experts_per_tok": 2,
"num_hidden_layers": 22,
"num_key_value_heads": 4,
"num_local_experts": 8,
"output_router_logits": false,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 10000.0,
"router_aux_loss_coef": 0.001,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.37.0",
"trust_remote_code": true,
"use_cache": false,
"vocab_size": 32000
}