imfinethx's picture
new dns ckpts for extra experiments
e93b448
{
"_name_or_path": "checkpoints/microsoft/phi-1_5",
"anyprec": {
"arch_config": {
"layers_name": "layers",
"model_name": "model",
"module_names": [
"self_attn.q_proj",
"self_attn.k_proj",
"self_attn.v_proj",
"self_attn.dense",
"mlp.fc1",
"mlp.fc2"
]
},
"group_count": 1,
"parent_precision": 4,
"seed_precision": 2,
"sparse_numvals": {
"model.layers.0.mlp.fc1": 1239127,
"model.layers.0.mlp.fc2": 1195670,
"model.layers.0.self_attn.dense": 181030,
"model.layers.0.self_attn.k_proj": 308573,
"model.layers.0.self_attn.q_proj": 276310,
"model.layers.0.self_attn.v_proj": 203605,
"model.layers.1.mlp.fc1": 513800,
"model.layers.1.mlp.fc2": 734843,
"model.layers.1.self_attn.dense": 168023,
"model.layers.1.self_attn.k_proj": 216089,
"model.layers.1.self_attn.q_proj": 210873,
"model.layers.1.self_attn.v_proj": 192721,
"model.layers.10.mlp.fc1": 679814,
"model.layers.10.mlp.fc2": 721214,
"model.layers.10.self_attn.dense": 173300,
"model.layers.10.self_attn.k_proj": 212281,
"model.layers.10.self_attn.q_proj": 203800,
"model.layers.10.self_attn.v_proj": 192613,
"model.layers.11.mlp.fc1": 679739,
"model.layers.11.mlp.fc2": 710094,
"model.layers.11.self_attn.dense": 169678,
"model.layers.11.self_attn.k_proj": 206617,
"model.layers.11.self_attn.q_proj": 198897,
"model.layers.11.self_attn.v_proj": 186403,
"model.layers.12.mlp.fc1": 666839,
"model.layers.12.mlp.fc2": 720708,
"model.layers.12.self_attn.dense": 168552,
"model.layers.12.self_attn.k_proj": 211803,
"model.layers.12.self_attn.q_proj": 204967,
"model.layers.12.self_attn.v_proj": 190021,
"model.layers.13.mlp.fc1": 656663,
"model.layers.13.mlp.fc2": 706611,
"model.layers.13.self_attn.dense": 169732,
"model.layers.13.self_attn.k_proj": 212001,
"model.layers.13.self_attn.q_proj": 201431,
"model.layers.13.self_attn.v_proj": 186583,
"model.layers.14.mlp.fc1": 649519,
"model.layers.14.mlp.fc2": 750919,
"model.layers.14.self_attn.dense": 176740,
"model.layers.14.self_attn.k_proj": 208548,
"model.layers.14.self_attn.q_proj": 204873,
"model.layers.14.self_attn.v_proj": 190063,
"model.layers.15.mlp.fc1": 636168,
"model.layers.15.mlp.fc2": 731257,
"model.layers.15.self_attn.dense": 168984,
"model.layers.15.self_attn.k_proj": 209159,
"model.layers.15.self_attn.q_proj": 224143,
"model.layers.15.self_attn.v_proj": 183076,
"model.layers.16.mlp.fc1": 626682,
"model.layers.16.mlp.fc2": 767523,
"model.layers.16.self_attn.dense": 167080,
"model.layers.16.self_attn.k_proj": 208334,
"model.layers.16.self_attn.q_proj": 211502,
"model.layers.16.self_attn.v_proj": 177715,
"model.layers.17.mlp.fc1": 610333,
"model.layers.17.mlp.fc2": 744465,
"model.layers.17.self_attn.dense": 165331,
"model.layers.17.self_attn.k_proj": 198333,
"model.layers.17.self_attn.q_proj": 196969,
"model.layers.17.self_attn.v_proj": 176784,
"model.layers.18.mlp.fc1": 603584,
"model.layers.18.mlp.fc2": 737950,
"model.layers.18.self_attn.dense": 181416,
"model.layers.18.self_attn.k_proj": 207074,
"model.layers.18.self_attn.q_proj": 234332,
"model.layers.18.self_attn.v_proj": 191671,
"model.layers.19.mlp.fc1": 593992,
"model.layers.19.mlp.fc2": 706175,
"model.layers.19.self_attn.dense": 179214,
"model.layers.19.self_attn.k_proj": 208480,
"model.layers.19.self_attn.q_proj": 232108,
"model.layers.19.self_attn.v_proj": 188840,
"model.layers.2.mlp.fc1": 592377,
"model.layers.2.mlp.fc2": 724600,
"model.layers.2.self_attn.dense": 163369,
"model.layers.2.self_attn.k_proj": 216145,
"model.layers.2.self_attn.q_proj": 209454,
"model.layers.2.self_attn.v_proj": 191623,
"model.layers.20.mlp.fc1": 586654,
"model.layers.20.mlp.fc2": 696954,
"model.layers.20.self_attn.dense": 168154,
"model.layers.20.self_attn.k_proj": 205374,
"model.layers.20.self_attn.q_proj": 228084,
"model.layers.20.self_attn.v_proj": 175088,
"model.layers.21.mlp.fc1": 588068,
"model.layers.21.mlp.fc2": 706060,
"model.layers.21.self_attn.dense": 169767,
"model.layers.21.self_attn.k_proj": 199803,
"model.layers.21.self_attn.q_proj": 229813,
"model.layers.21.self_attn.v_proj": 179190,
"model.layers.22.mlp.fc1": 598655,
"model.layers.22.mlp.fc2": 789250,
"model.layers.22.self_attn.dense": 167244,
"model.layers.22.self_attn.k_proj": 193913,
"model.layers.22.self_attn.q_proj": 275941,
"model.layers.22.self_attn.v_proj": 173336,
"model.layers.23.mlp.fc1": 680534,
"model.layers.23.mlp.fc2": 1070400,
"model.layers.23.self_attn.dense": 210736,
"model.layers.23.self_attn.k_proj": 216070,
"model.layers.23.self_attn.q_proj": 388123,
"model.layers.23.self_attn.v_proj": 221935,
"model.layers.3.mlp.fc1": 635345,
"model.layers.3.mlp.fc2": 728654,
"model.layers.3.self_attn.dense": 177298,
"model.layers.3.self_attn.k_proj": 231769,
"model.layers.3.self_attn.q_proj": 223554,
"model.layers.3.self_attn.v_proj": 205803,
"model.layers.4.mlp.fc1": 684553,
"model.layers.4.mlp.fc2": 733946,
"model.layers.4.self_attn.dense": 172237,
"model.layers.4.self_attn.k_proj": 219073,
"model.layers.4.self_attn.q_proj": 211940,
"model.layers.4.self_attn.v_proj": 203401,
"model.layers.5.mlp.fc1": 675444,
"model.layers.5.mlp.fc2": 746508,
"model.layers.5.self_attn.dense": 170531,
"model.layers.5.self_attn.k_proj": 238923,
"model.layers.5.self_attn.q_proj": 234678,
"model.layers.5.self_attn.v_proj": 199692,
"model.layers.6.mlp.fc1": 680226,
"model.layers.6.mlp.fc2": 732111,
"model.layers.6.self_attn.dense": 182410,
"model.layers.6.self_attn.k_proj": 215481,
"model.layers.6.self_attn.q_proj": 211152,
"model.layers.6.self_attn.v_proj": 203868,
"model.layers.7.mlp.fc1": 685126,
"model.layers.7.mlp.fc2": 727309,
"model.layers.7.self_attn.dense": 167738,
"model.layers.7.self_attn.k_proj": 222298,
"model.layers.7.self_attn.q_proj": 216189,
"model.layers.7.self_attn.v_proj": 191654,
"model.layers.8.mlp.fc1": 692316,
"model.layers.8.mlp.fc2": 715828,
"model.layers.8.self_attn.dense": 176375,
"model.layers.8.self_attn.k_proj": 222524,
"model.layers.8.self_attn.q_proj": 215045,
"model.layers.8.self_attn.v_proj": 198133,
"model.layers.9.mlp.fc1": 688557,
"model.layers.9.mlp.fc2": 718275,
"model.layers.9.self_attn.dense": 166542,
"model.layers.9.self_attn.k_proj": 220268,
"model.layers.9.self_attn.q_proj": 213681,
"model.layers.9.self_attn.v_proj": 191284
}
},
"architectures": [
"PhiForCausalLM"
],
"attention_dropout": 0.0,
"auto_map": {
"AutoConfig": "configuration_phi.PhiConfig",
"AutoModelForCausalLM": "modeling_phi.PhiForCausalLM"
},
"bos_token_id": null,
"embd_pdrop": 0.0,
"eos_token_id": null,
"hidden_act": "gelu_new",
"hidden_size": 2048,
"initializer_range": 0.02,
"intermediate_size": 8192,
"layer_norm_eps": 1e-05,
"max_position_embeddings": 2048,
"model_type": "phi",
"num_attention_heads": 32,
"num_hidden_layers": 24,
"num_key_value_heads": 32,
"partial_rotary_factor": 0.5,
"qk_layernorm": false,
"resid_pdrop": 0.0,
"rope_scaling": null,
"rope_theta": 10000.0,
"tie_word_embeddings": false,
"torch_dtype": "float16",
"transformers_version": "4.39.3",
"use_cache": true,
"vocab_size": 51200
}