Upload config.json with huggingface_hub
Browse files- config.json +922 -0
config.json
ADDED
|
@@ -0,0 +1,922 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Glm4MoeForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": true,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"head_dim": 128,
|
| 8 |
+
"hidden_act": "silu",
|
| 9 |
+
"hidden_size": 5120,
|
| 10 |
+
"intermediate_size": 12288,
|
| 11 |
+
"max_position_embeddings": 202752,
|
| 12 |
+
"model_type": "glm4_moe",
|
| 13 |
+
"moe_intermediate_size": 1536,
|
| 14 |
+
"norm_topk_prob": true,
|
| 15 |
+
"scoring_func": "sigmoid",
|
| 16 |
+
"num_attention_heads": 96,
|
| 17 |
+
"n_group": 1,
|
| 18 |
+
"topk_group": 1,
|
| 19 |
+
"n_routed_experts": 160,
|
| 20 |
+
"n_shared_experts": 1,
|
| 21 |
+
"routed_scaling_factor": 2.5,
|
| 22 |
+
"topk_method": "noaux_tc",
|
| 23 |
+
"moe_router_dtype": "float32",
|
| 24 |
+
"num_experts_per_tok": 8,
|
| 25 |
+
"first_k_dense_replace": 3,
|
| 26 |
+
"use_qk_norm": true,
|
| 27 |
+
"num_hidden_layers": 92,
|
| 28 |
+
"num_key_value_heads": 8,
|
| 29 |
+
"num_nextn_predict_layers": 1,
|
| 30 |
+
"partial_rotary_factor": 0.5,
|
| 31 |
+
"rms_norm_eps": 1e-05,
|
| 32 |
+
"rope_scaling": null,
|
| 33 |
+
"rope_theta": 1000000,
|
| 34 |
+
"tie_word_embeddings": false,
|
| 35 |
+
"torch_dtype": "bfloat16",
|
| 36 |
+
"vocab_size": 151552,
|
| 37 |
+
"transformers_version": "4.52.3",
|
| 38 |
+
"eos_token_id": [
|
| 39 |
+
151329,
|
| 40 |
+
151336,
|
| 41 |
+
151338
|
| 42 |
+
],
|
| 43 |
+
"pad_token_id": 151329,
|
| 44 |
+
"quantization_config": {
|
| 45 |
+
"config_groups": {
|
| 46 |
+
"group_0": {
|
| 47 |
+
"input_activations": {
|
| 48 |
+
"actorder": null,
|
| 49 |
+
"block_structure": null,
|
| 50 |
+
"dynamic": true,
|
| 51 |
+
"group_size": null,
|
| 52 |
+
"num_bits": 8,
|
| 53 |
+
"observer": null,
|
| 54 |
+
"observer_kwargs": {},
|
| 55 |
+
"strategy": "token",
|
| 56 |
+
"symmetric": true,
|
| 57 |
+
"type": "float"
|
| 58 |
+
},
|
| 59 |
+
"output_activations": null,
|
| 60 |
+
"targets": [
|
| 61 |
+
"Linear"
|
| 62 |
+
],
|
| 63 |
+
"weights": {
|
| 64 |
+
"actorder": null,
|
| 65 |
+
"block_structure": null,
|
| 66 |
+
"dynamic": false,
|
| 67 |
+
"group_size": null,
|
| 68 |
+
"num_bits": 8,
|
| 69 |
+
"observer": "minmax",
|
| 70 |
+
"observer_kwargs": {},
|
| 71 |
+
"strategy": "channel",
|
| 72 |
+
"symmetric": true,
|
| 73 |
+
"type": "float"
|
| 74 |
+
}
|
| 75 |
+
}
|
| 76 |
+
},
|
| 77 |
+
"format": "float-quantized",
|
| 78 |
+
"ignore": [
|
| 79 |
+
"model.layers.46.post_attention_layernorm",
|
| 80 |
+
"model.layers.16.mlp.gate",
|
| 81 |
+
"model.layers.36.self_attn.k_proj.bias",
|
| 82 |
+
"model.layers.88.self_attn.k_norm",
|
| 83 |
+
"model.layers.17.self_attn.q_proj.bias",
|
| 84 |
+
"model.layers.26.self_attn.v_proj.bias",
|
| 85 |
+
"model.layers.5.mlp.gate",
|
| 86 |
+
"model.layers.59.input_layernorm",
|
| 87 |
+
"model.layers.24.self_attn.k_proj.bias",
|
| 88 |
+
"model.layers.40.self_attn.v_proj.bias",
|
| 89 |
+
"model.layers.55.mlp.gate.e_score_correction_bias",
|
| 90 |
+
"model.layers.72.self_attn.v_proj.bias",
|
| 91 |
+
"model.layers.62.self_attn.q_norm",
|
| 92 |
+
"model.layers.24.self_attn.q_norm",
|
| 93 |
+
"model.layers.7.self_attn.q_norm",
|
| 94 |
+
"model.layers.78.self_attn.k_proj.bias",
|
| 95 |
+
"model.layers.75.self_attn.q_norm",
|
| 96 |
+
"model.layers.45.post_attention_layernorm",
|
| 97 |
+
"model.layers.50.self_attn.k_norm",
|
| 98 |
+
"model.layers.22.self_attn.v_proj.bias",
|
| 99 |
+
"model.layers.15.post_attention_layernorm",
|
| 100 |
+
"model.layers.24.input_layernorm",
|
| 101 |
+
"model.layers.24.mlp.gate.e_score_correction_bias",
|
| 102 |
+
"model.layers.34.self_attn.q_norm",
|
| 103 |
+
"model.layers.21.post_attention_layernorm",
|
| 104 |
+
"model.layers.15.self_attn.k_norm",
|
| 105 |
+
"model.layers.82.self_attn.v_proj.bias",
|
| 106 |
+
"model.layers.60.self_attn.q_proj.bias",
|
| 107 |
+
"model.layers.86.self_attn.k_norm",
|
| 108 |
+
"model.layers.91.input_layernorm",
|
| 109 |
+
"model.layers.67.self_attn.q_proj.bias",
|
| 110 |
+
"model.layers.27.self_attn.k_norm",
|
| 111 |
+
"model.layers.76.self_attn.q_proj.bias",
|
| 112 |
+
"model.layers.53.self_attn.v_proj.bias",
|
| 113 |
+
"model.layers.8.self_attn.k_proj.bias",
|
| 114 |
+
"model.layers.81.self_attn.k_proj.bias",
|
| 115 |
+
"model.layers.20.self_attn.k_norm",
|
| 116 |
+
"model.layers.24.self_attn.v_proj.bias",
|
| 117 |
+
"model.layers.81.input_layernorm",
|
| 118 |
+
"model.layers.23.mlp.gate",
|
| 119 |
+
"model.layers.42.input_layernorm",
|
| 120 |
+
"model.layers.48.mlp.gate.e_score_correction_bias",
|
| 121 |
+
"model.layers.91.self_attn.q_proj.bias",
|
| 122 |
+
"model.layers.32.self_attn.q_norm",
|
| 123 |
+
"model.layers.23.input_layernorm",
|
| 124 |
+
"model.layers.49.mlp.gate",
|
| 125 |
+
"model.layers.54.self_attn.v_proj.bias",
|
| 126 |
+
"model.layers.62.self_attn.k_proj.bias",
|
| 127 |
+
"model.layers.90.self_attn.k_norm",
|
| 128 |
+
"model.layers.72.self_attn.k_proj.bias",
|
| 129 |
+
"model.layers.39.self_attn.q_norm",
|
| 130 |
+
"model.layers.28.self_attn.v_proj.bias",
|
| 131 |
+
"model.layers.63.self_attn.k_proj.bias",
|
| 132 |
+
"model.layers.17.self_attn.v_proj.bias",
|
| 133 |
+
"model.layers.72.post_attention_layernorm",
|
| 134 |
+
"model.layers.26.input_layernorm",
|
| 135 |
+
"model.layers.77.mlp.gate.e_score_correction_bias",
|
| 136 |
+
"model.layers.33.mlp.gate.e_score_correction_bias",
|
| 137 |
+
"model.layers.54.self_attn.k_proj.bias",
|
| 138 |
+
"model.layers.72.self_attn.k_norm",
|
| 139 |
+
"model.layers.87.self_attn.k_proj.bias",
|
| 140 |
+
"model.layers.22.input_layernorm",
|
| 141 |
+
"model.layers.90.mlp.gate",
|
| 142 |
+
"model.layers.38.self_attn.k_proj.bias",
|
| 143 |
+
"model.layers.20.input_layernorm",
|
| 144 |
+
"model.layers.4.self_attn.k_proj.bias",
|
| 145 |
+
"model.layers.43.mlp.gate",
|
| 146 |
+
"model.layers.37.post_attention_layernorm",
|
| 147 |
+
"model.layers.14.mlp.gate",
|
| 148 |
+
"model.layers.79.post_attention_layernorm",
|
| 149 |
+
"model.layers.84.self_attn.k_norm",
|
| 150 |
+
"model.layers.71.self_attn.q_norm",
|
| 151 |
+
"model.layers.56.mlp.gate.e_score_correction_bias",
|
| 152 |
+
"model.layers.6.self_attn.k_proj.bias",
|
| 153 |
+
"model.layers.70.self_attn.k_proj.bias",
|
| 154 |
+
"model.layers.54.post_attention_layernorm",
|
| 155 |
+
"model.layers.55.self_attn.q_norm",
|
| 156 |
+
"model.layers.4.mlp.gate",
|
| 157 |
+
"model.layers.81.mlp.gate",
|
| 158 |
+
"model.layers.7.mlp.gate",
|
| 159 |
+
"model.layers.8.self_attn.q_norm",
|
| 160 |
+
"model.layers.57.self_attn.v_proj.bias",
|
| 161 |
+
"model.layers.65.self_attn.q_norm",
|
| 162 |
+
"model.layers.88.post_attention_layernorm",
|
| 163 |
+
"model.layers.19.input_layernorm",
|
| 164 |
+
"model.layers.6.self_attn.v_proj.bias",
|
| 165 |
+
"model.layers.4.input_layernorm",
|
| 166 |
+
"model.layers.71.self_attn.k_norm",
|
| 167 |
+
"model.layers.32.self_attn.k_norm",
|
| 168 |
+
"model.layers.22.post_attention_layernorm",
|
| 169 |
+
"model.layers.6.self_attn.q_proj.bias",
|
| 170 |
+
"model.layers.61.self_attn.q_norm",
|
| 171 |
+
"model.layers.3.input_layernorm",
|
| 172 |
+
"model.layers.13.input_layernorm",
|
| 173 |
+
"model.layers.74.self_attn.k_norm",
|
| 174 |
+
"model.layers.59.self_attn.v_proj.bias",
|
| 175 |
+
"model.layers.60.self_attn.v_proj.bias",
|
| 176 |
+
"model.layers.35.self_attn.q_proj.bias",
|
| 177 |
+
"model.layers.18.post_attention_layernorm",
|
| 178 |
+
"model.layers.44.self_attn.k_proj.bias",
|
| 179 |
+
"model.layers.64.mlp.gate",
|
| 180 |
+
"model.layers.61.post_attention_layernorm",
|
| 181 |
+
"model.layers.46.self_attn.k_norm",
|
| 182 |
+
"model.layers.19.post_attention_layernorm",
|
| 183 |
+
"model.layers.66.input_layernorm",
|
| 184 |
+
"model.layers.40.mlp.gate.e_score_correction_bias",
|
| 185 |
+
"model.layers.18.self_attn.k_norm",
|
| 186 |
+
"model.layers.30.input_layernorm",
|
| 187 |
+
"model.layers.56.self_attn.v_proj.bias",
|
| 188 |
+
"model.layers.49.self_attn.v_proj.bias",
|
| 189 |
+
"model.layers.3.self_attn.v_proj.bias",
|
| 190 |
+
"model.layers.72.mlp.gate.e_score_correction_bias",
|
| 191 |
+
"model.layers.79.self_attn.k_norm",
|
| 192 |
+
"model.layers.92.self_attn.k_norm",
|
| 193 |
+
"model.layers.79.self_attn.v_proj.bias",
|
| 194 |
+
"model.layers.32.mlp.gate.e_score_correction_bias",
|
| 195 |
+
"model.layers.26.mlp.gate.e_score_correction_bias",
|
| 196 |
+
"model.embed_tokens",
|
| 197 |
+
"model.layers.18.mlp.gate",
|
| 198 |
+
"model.layers.62.self_attn.q_proj.bias",
|
| 199 |
+
"model.layers.44.post_attention_layernorm",
|
| 200 |
+
"model.layers.55.post_attention_layernorm",
|
| 201 |
+
"model.layers.18.input_layernorm",
|
| 202 |
+
"model.layers.46.mlp.gate.e_score_correction_bias",
|
| 203 |
+
"model.layers.75.mlp.gate",
|
| 204 |
+
"model.layers.63.self_attn.q_norm",
|
| 205 |
+
"model.layers.36.mlp.gate.e_score_correction_bias",
|
| 206 |
+
"model.layers.66.self_attn.q_norm",
|
| 207 |
+
"model.layers.45.mlp.gate",
|
| 208 |
+
"model.layers.24.post_attention_layernorm",
|
| 209 |
+
"model.layers.45.self_attn.k_norm",
|
| 210 |
+
"model.layers.0.self_attn.q_proj.bias",
|
| 211 |
+
"model.layers.42.mlp.gate.e_score_correction_bias",
|
| 212 |
+
"model.layers.64.self_attn.q_norm",
|
| 213 |
+
"model.layers.8.input_layernorm",
|
| 214 |
+
"model.layers.26.self_attn.q_proj.bias",
|
| 215 |
+
"model.layers.32.post_attention_layernorm",
|
| 216 |
+
"model.layers.40.self_attn.k_norm",
|
| 217 |
+
"model.layers.91.self_attn.q_norm",
|
| 218 |
+
"model.layers.69.post_attention_layernorm",
|
| 219 |
+
"model.layers.27.self_attn.v_proj.bias",
|
| 220 |
+
"model.layers.9.self_attn.k_norm",
|
| 221 |
+
"model.layers.62.self_attn.k_norm",
|
| 222 |
+
"model.layers.92.input_layernorm",
|
| 223 |
+
"model.layers.14.self_attn.q_proj.bias",
|
| 224 |
+
"model.layers.84.input_layernorm",
|
| 225 |
+
"model.layers.88.mlp.gate",
|
| 226 |
+
"model.layers.7.self_attn.v_proj.bias",
|
| 227 |
+
"model.layers.14.self_attn.q_norm",
|
| 228 |
+
"model.layers.17.self_attn.q_norm",
|
| 229 |
+
"model.layers.85.mlp.gate",
|
| 230 |
+
"model.layers.63.mlp.gate",
|
| 231 |
+
"model.layers.66.self_attn.q_proj.bias",
|
| 232 |
+
"model.layers.76.self_attn.k_proj.bias",
|
| 233 |
+
"model.layers.18.self_attn.q_norm",
|
| 234 |
+
"model.layers.77.post_attention_layernorm",
|
| 235 |
+
"model.layers.55.self_attn.k_norm",
|
| 236 |
+
"model.layers.37.self_attn.q_norm",
|
| 237 |
+
"model.layers.91.post_attention_layernorm",
|
| 238 |
+
"model.layers.53.self_attn.q_norm",
|
| 239 |
+
"model.layers.76.input_layernorm",
|
| 240 |
+
"model.layers.1.self_attn.v_proj.bias",
|
| 241 |
+
"model.layers.38.self_attn.v_proj.bias",
|
| 242 |
+
"model.layers.14.mlp.gate.e_score_correction_bias",
|
| 243 |
+
"model.layers.9.self_attn.v_proj.bias",
|
| 244 |
+
"model.layers.75.post_attention_layernorm",
|
| 245 |
+
"model.layers.9.self_attn.q_norm",
|
| 246 |
+
"model.layers.57.mlp.gate.e_score_correction_bias",
|
| 247 |
+
"model.layers.82.self_attn.k_norm",
|
| 248 |
+
"model.layers.7.post_attention_layernorm",
|
| 249 |
+
"model.layers.43.self_attn.q_norm",
|
| 250 |
+
"model.layers.91.self_attn.k_proj.bias",
|
| 251 |
+
"model.layers.29.self_attn.k_norm",
|
| 252 |
+
"model.layers.11.self_attn.k_proj.bias",
|
| 253 |
+
"model.layers.6.mlp.gate",
|
| 254 |
+
"model.layers.7.self_attn.q_proj.bias",
|
| 255 |
+
"model.layers.63.post_attention_layernorm",
|
| 256 |
+
"model.layers.11.self_attn.q_proj.bias",
|
| 257 |
+
"model.layers.25.input_layernorm",
|
| 258 |
+
"model.layers.21.input_layernorm",
|
| 259 |
+
"model.layers.61.input_layernorm",
|
| 260 |
+
"model.layers.37.mlp.gate.e_score_correction_bias",
|
| 261 |
+
"model.layers.65.mlp.gate",
|
| 262 |
+
"model.layers.20.self_attn.q_norm",
|
| 263 |
+
"model.layers.60.post_attention_layernorm",
|
| 264 |
+
"model.layers.35.self_attn.k_proj.bias",
|
| 265 |
+
"model.layers.70.mlp.gate.e_score_correction_bias",
|
| 266 |
+
"model.layers.85.self_attn.k_norm",
|
| 267 |
+
"model.layers.58.input_layernorm",
|
| 268 |
+
"model.layers.66.self_attn.v_proj.bias",
|
| 269 |
+
"model.layers.80.self_attn.q_norm",
|
| 270 |
+
"model.layers.59.self_attn.k_norm",
|
| 271 |
+
"model.layers.47.self_attn.k_proj.bias",
|
| 272 |
+
"model.layers.46.mlp.gate",
|
| 273 |
+
"model.layers.27.self_attn.k_proj.bias",
|
| 274 |
+
"model.layers.71.self_attn.k_proj.bias",
|
| 275 |
+
"model.layers.76.self_attn.q_norm",
|
| 276 |
+
"model.layers.16.input_layernorm",
|
| 277 |
+
"model.layers.56.self_attn.k_proj.bias",
|
| 278 |
+
"model.layers.30.self_attn.q_proj.bias",
|
| 279 |
+
"model.layers.89.self_attn.q_norm",
|
| 280 |
+
"model.layers.80.self_attn.k_norm",
|
| 281 |
+
"model.layers.40.mlp.gate",
|
| 282 |
+
"model.layers.10.mlp.gate.e_score_correction_bias",
|
| 283 |
+
"model.layers.89.self_attn.k_proj.bias",
|
| 284 |
+
"model.layers.41.mlp.gate.e_score_correction_bias",
|
| 285 |
+
"model.layers.79.self_attn.q_norm",
|
| 286 |
+
"model.layers.49.self_attn.k_proj.bias",
|
| 287 |
+
"model.layers.43.self_attn.k_norm",
|
| 288 |
+
"model.layers.21.self_attn.q_proj.bias",
|
| 289 |
+
"model.layers.69.self_attn.v_proj.bias",
|
| 290 |
+
"model.layers.70.self_attn.k_norm",
|
| 291 |
+
"model.layers.65.self_attn.k_norm",
|
| 292 |
+
"model.layers.59.mlp.gate.e_score_correction_bias",
|
| 293 |
+
"model.layers.92.post_attention_layernorm",
|
| 294 |
+
"model.layers.81.self_attn.v_proj.bias",
|
| 295 |
+
"model.layers.51.input_layernorm",
|
| 296 |
+
"model.layers.3.post_attention_layernorm",
|
| 297 |
+
"model.layers.28.post_attention_layernorm",
|
| 298 |
+
"model.layers.1.self_attn.q_proj.bias",
|
| 299 |
+
"model.layers.0.self_attn.v_proj.bias",
|
| 300 |
+
"model.layers.10.self_attn.q_norm",
|
| 301 |
+
"model.layers.34.mlp.gate.e_score_correction_bias",
|
| 302 |
+
"model.layers.52.self_attn.v_proj.bias",
|
| 303 |
+
"model.layers.1.self_attn.q_norm",
|
| 304 |
+
"model.layers.89.self_attn.k_norm",
|
| 305 |
+
"model.layers.27.mlp.gate.e_score_correction_bias",
|
| 306 |
+
"model.layers.53.self_attn.k_norm",
|
| 307 |
+
"model.layers.61.self_attn.v_proj.bias",
|
| 308 |
+
"model.layers.78.input_layernorm",
|
| 309 |
+
"model.layers.2.self_attn.v_proj.bias",
|
| 310 |
+
"model.layers.76.post_attention_layernorm",
|
| 311 |
+
"model.layers.36.self_attn.q_norm",
|
| 312 |
+
"model.layers.68.self_attn.q_proj.bias",
|
| 313 |
+
"model.layers.33.self_attn.k_proj.bias",
|
| 314 |
+
"model.layers.78.mlp.gate",
|
| 315 |
+
"model.layers.36.mlp.gate",
|
| 316 |
+
"model.layers.44.mlp.gate",
|
| 317 |
+
"model.layers.54.mlp.gate.e_score_correction_bias",
|
| 318 |
+
"model.layers.73.mlp.gate.e_score_correction_bias",
|
| 319 |
+
"model.layers.41.self_attn.k_proj.bias",
|
| 320 |
+
"model.layers.63.self_attn.k_norm",
|
| 321 |
+
"model.layers.29.self_attn.q_proj.bias",
|
| 322 |
+
"model.layers.22.self_attn.q_norm",
|
| 323 |
+
"model.layers.74.self_attn.q_proj.bias",
|
| 324 |
+
"model.layers.47.post_attention_layernorm",
|
| 325 |
+
"model.layers.64.input_layernorm",
|
| 326 |
+
"model.layers.77.self_attn.k_norm",
|
| 327 |
+
"model.layers.27.mlp.gate",
|
| 328 |
+
"model.layers.33.self_attn.q_proj.bias",
|
| 329 |
+
"model.layers.77.self_attn.v_proj.bias",
|
| 330 |
+
"model.layers.87.input_layernorm",
|
| 331 |
+
"model.layers.5.input_layernorm",
|
| 332 |
+
"model.layers.14.self_attn.k_proj.bias",
|
| 333 |
+
"model.layers.32.input_layernorm",
|
| 334 |
+
"model.layers.35.input_layernorm",
|
| 335 |
+
"model.layers.44.self_attn.q_norm",
|
| 336 |
+
"model.layers.51.self_attn.k_norm",
|
| 337 |
+
"model.layers.80.self_attn.v_proj.bias",
|
| 338 |
+
"model.layers.63.self_attn.v_proj.bias",
|
| 339 |
+
"model.layers.4.self_attn.k_norm",
|
| 340 |
+
"model.layers.89.self_attn.q_proj.bias",
|
| 341 |
+
"model.layers.61.self_attn.q_proj.bias",
|
| 342 |
+
"model.layers.9.post_attention_layernorm",
|
| 343 |
+
"model.layers.29.self_attn.v_proj.bias",
|
| 344 |
+
"model.layers.92.hnorm",
|
| 345 |
+
"model.layers.53.self_attn.q_proj.bias",
|
| 346 |
+
"model.layers.73.self_attn.k_norm",
|
| 347 |
+
"model.layers.91.mlp.gate.e_score_correction_bias",
|
| 348 |
+
"model.layers.75.self_attn.q_proj.bias",
|
| 349 |
+
"model.layers.47.self_attn.q_norm",
|
| 350 |
+
"model.layers.77.mlp.gate",
|
| 351 |
+
"model.layers.21.mlp.gate.e_score_correction_bias",
|
| 352 |
+
"model.layers.90.self_attn.v_proj.bias",
|
| 353 |
+
"model.layers.44.mlp.gate.e_score_correction_bias",
|
| 354 |
+
"model.layers.37.self_attn.k_proj.bias",
|
| 355 |
+
"model.layers.92.mlp.gate",
|
| 356 |
+
"model.layers.32.mlp.gate",
|
| 357 |
+
"model.layers.54.input_layernorm",
|
| 358 |
+
"model.layers.13.self_attn.q_proj.bias",
|
| 359 |
+
"model.layers.0.post_attention_layernorm",
|
| 360 |
+
"model.layers.84.mlp.gate.e_score_correction_bias",
|
| 361 |
+
"model.layers.31.mlp.gate.e_score_correction_bias",
|
| 362 |
+
"model.layers.57.self_attn.q_norm",
|
| 363 |
+
"model.layers.83.post_attention_layernorm",
|
| 364 |
+
"model.layers.61.mlp.gate.e_score_correction_bias",
|
| 365 |
+
"model.layers.49.input_layernorm",
|
| 366 |
+
"model.layers.50.mlp.gate.e_score_correction_bias",
|
| 367 |
+
"model.layers.48.self_attn.k_norm",
|
| 368 |
+
"model.layers.18.self_attn.v_proj.bias",
|
| 369 |
+
"model.layers.54.self_attn.q_proj.bias",
|
| 370 |
+
"model.layers.8.mlp.gate.e_score_correction_bias",
|
| 371 |
+
"model.layers.14.self_attn.v_proj.bias",
|
| 372 |
+
"model.layers.5.self_attn.q_norm",
|
| 373 |
+
"model.layers.90.mlp.gate.e_score_correction_bias",
|
| 374 |
+
"model.layers.83.mlp.gate.e_score_correction_bias",
|
| 375 |
+
"model.layers.69.input_layernorm",
|
| 376 |
+
"model.layers.82.self_attn.k_proj.bias",
|
| 377 |
+
"model.layers.86.self_attn.q_norm",
|
| 378 |
+
"model.layers.65.input_layernorm",
|
| 379 |
+
"model.layers.16.mlp.gate.e_score_correction_bias",
|
| 380 |
+
"model.layers.22.mlp.gate.e_score_correction_bias",
|
| 381 |
+
"model.layers.80.mlp.gate.e_score_correction_bias",
|
| 382 |
+
"model.layers.8.self_attn.q_proj.bias",
|
| 383 |
+
"model.layers.23.self_attn.q_norm",
|
| 384 |
+
"model.layers.11.self_attn.k_norm",
|
| 385 |
+
"model.layers.53.mlp.gate.e_score_correction_bias",
|
| 386 |
+
"model.layers.65.self_attn.v_proj.bias",
|
| 387 |
+
"model.layers.47.mlp.gate",
|
| 388 |
+
"model.layers.70.mlp.gate",
|
| 389 |
+
"model.layers.12.mlp.gate.e_score_correction_bias",
|
| 390 |
+
"model.layers.38.mlp.gate.e_score_correction_bias",
|
| 391 |
+
"model.layers.46.self_attn.k_proj.bias",
|
| 392 |
+
"model.layers.64.self_attn.k_norm",
|
| 393 |
+
"model.layers.89.mlp.gate",
|
| 394 |
+
"model.layers.44.self_attn.v_proj.bias",
|
| 395 |
+
"model.layers.33.self_attn.k_norm",
|
| 396 |
+
"model.layers.8.self_attn.k_norm",
|
| 397 |
+
"model.layers.1.self_attn.k_proj.bias",
|
| 398 |
+
"model.layers.51.self_attn.q_proj.bias",
|
| 399 |
+
"lm_head",
|
| 400 |
+
"model.layers.42.self_attn.k_norm",
|
| 401 |
+
"model.layers.60.input_layernorm",
|
| 402 |
+
"model.layers.59.self_attn.q_proj.bias",
|
| 403 |
+
"model.layers.52.self_attn.q_norm",
|
| 404 |
+
"model.layers.84.self_attn.q_proj.bias",
|
| 405 |
+
"model.layers.51.mlp.gate",
|
| 406 |
+
"model.layers.81.self_attn.k_norm",
|
| 407 |
+
"model.layers.1.input_layernorm",
|
| 408 |
+
"model.layers.33.self_attn.q_norm",
|
| 409 |
+
"model.layers.78.mlp.gate.e_score_correction_bias",
|
| 410 |
+
"model.layers.17.post_attention_layernorm",
|
| 411 |
+
"model.layers.50.self_attn.k_proj.bias",
|
| 412 |
+
"model.layers.73.input_layernorm",
|
| 413 |
+
"model.layers.16.self_attn.q_proj.bias",
|
| 414 |
+
"model.layers.70.self_attn.q_proj.bias",
|
| 415 |
+
"model.layers.26.post_attention_layernorm",
|
| 416 |
+
"model.layers.27.input_layernorm",
|
| 417 |
+
"model.layers.86.self_attn.q_proj.bias",
|
| 418 |
+
"model.layers.33.self_attn.v_proj.bias",
|
| 419 |
+
"model.layers.50.mlp.gate",
|
| 420 |
+
"model.layers.64.post_attention_layernorm",
|
| 421 |
+
"model.layers.73.self_attn.q_proj.bias",
|
| 422 |
+
"model.layers.72.input_layernorm",
|
| 423 |
+
"model.layers.66.self_attn.k_norm",
|
| 424 |
+
"model.layers.50.self_attn.v_proj.bias",
|
| 425 |
+
"model.layers.24.self_attn.k_norm",
|
| 426 |
+
"model.layers.69.self_attn.k_norm",
|
| 427 |
+
"model.layers.75.self_attn.k_norm",
|
| 428 |
+
"model.layers.36.input_layernorm",
|
| 429 |
+
"model.layers.40.self_attn.q_norm",
|
| 430 |
+
"model.layers.33.input_layernorm",
|
| 431 |
+
"model.layers.92.enorm",
|
| 432 |
+
"model.layers.23.self_attn.k_norm",
|
| 433 |
+
"model.layers.25.mlp.gate.e_score_correction_bias",
|
| 434 |
+
"model.layers.74.self_attn.q_norm",
|
| 435 |
+
"model.layers.13.self_attn.q_norm",
|
| 436 |
+
"model.layers.75.self_attn.v_proj.bias",
|
| 437 |
+
"model.layers.30.self_attn.v_proj.bias",
|
| 438 |
+
"model.layers.42.self_attn.v_proj.bias",
|
| 439 |
+
"model.layers.29.input_layernorm",
|
| 440 |
+
"model.layers.91.self_attn.k_norm",
|
| 441 |
+
"model.layers.92.self_attn.k_proj.bias",
|
| 442 |
+
"model.layers.12.self_attn.v_proj.bias",
|
| 443 |
+
"model.layers.19.self_attn.v_proj.bias",
|
| 444 |
+
"model.layers.70.self_attn.q_norm",
|
| 445 |
+
"model.layers.75.self_attn.k_proj.bias",
|
| 446 |
+
"model.layers.49.post_attention_layernorm",
|
| 447 |
+
"model.layers.50.self_attn.q_proj.bias",
|
| 448 |
+
"model.layers.84.post_attention_layernorm",
|
| 449 |
+
"model.layers.7.mlp.gate.e_score_correction_bias",
|
| 450 |
+
"model.layers.59.post_attention_layernorm",
|
| 451 |
+
"model.layers.86.input_layernorm",
|
| 452 |
+
"model.layers.22.self_attn.k_proj.bias",
|
| 453 |
+
"model.layers.31.self_attn.q_proj.bias",
|
| 454 |
+
"model.layers.30.post_attention_layernorm",
|
| 455 |
+
"model.layers.79.input_layernorm",
|
| 456 |
+
"model.layers.3.self_attn.k_proj.bias",
|
| 457 |
+
"model.layers.66.self_attn.k_proj.bias",
|
| 458 |
+
"model.layers.81.mlp.gate.e_score_correction_bias",
|
| 459 |
+
"model.layers.88.mlp.gate.e_score_correction_bias",
|
| 460 |
+
"model.layers.29.self_attn.k_proj.bias",
|
| 461 |
+
"model.layers.4.self_attn.q_proj.bias",
|
| 462 |
+
"model.layers.79.self_attn.q_proj.bias",
|
| 463 |
+
"model.layers.35.post_attention_layernorm",
|
| 464 |
+
"model.layers.63.input_layernorm",
|
| 465 |
+
"model.layers.12.self_attn.q_proj.bias",
|
| 466 |
+
"model.layers.0.self_attn.q_norm",
|
| 467 |
+
"model.layers.13.post_attention_layernorm",
|
| 468 |
+
"model.layers.58.self_attn.q_proj.bias",
|
| 469 |
+
"model.layers.90.self_attn.k_proj.bias",
|
| 470 |
+
"model.layers.45.self_attn.v_proj.bias",
|
| 471 |
+
"model.layers.86.self_attn.k_proj.bias",
|
| 472 |
+
"model.layers.38.input_layernorm",
|
| 473 |
+
"model.layers.70.post_attention_layernorm",
|
| 474 |
+
"model.layers.69.self_attn.k_proj.bias",
|
| 475 |
+
"model.layers.88.input_layernorm",
|
| 476 |
+
"model.layers.75.input_layernorm",
|
| 477 |
+
"model.layers.76.mlp.gate.e_score_correction_bias",
|
| 478 |
+
"model.layers.76.mlp.gate",
|
| 479 |
+
"model.layers.8.post_attention_layernorm",
|
| 480 |
+
"model.layers.30.mlp.gate",
|
| 481 |
+
"model.layers.38.self_attn.q_proj.bias",
|
| 482 |
+
"model.layers.19.self_attn.q_norm",
|
| 483 |
+
"model.layers.76.self_attn.k_norm",
|
| 484 |
+
"model.layers.79.mlp.gate",
|
| 485 |
+
"model.layers.23.self_attn.v_proj.bias",
|
| 486 |
+
"model.layers.27.self_attn.q_proj.bias",
|
| 487 |
+
"model.layers.57.post_attention_layernorm",
|
| 488 |
+
"model.layers.56.mlp.gate",
|
| 489 |
+
"model.layers.45.mlp.gate.e_score_correction_bias",
|
| 490 |
+
"model.layers.76.self_attn.v_proj.bias",
|
| 491 |
+
"model.layers.26.self_attn.q_norm",
|
| 492 |
+
"model.layers.17.self_attn.k_norm",
|
| 493 |
+
"model.layers.19.mlp.gate.e_score_correction_bias",
|
| 494 |
+
"model.layers.6.post_attention_layernorm",
|
| 495 |
+
"model.layers.12.post_attention_layernorm",
|
| 496 |
+
"model.layers.71.input_layernorm",
|
| 497 |
+
"model.layers.5.post_attention_layernorm",
|
| 498 |
+
"model.layers.25.self_attn.k_norm",
|
| 499 |
+
"model.layers.37.mlp.gate",
|
| 500 |
+
"model.layers.20.self_attn.k_proj.bias",
|
| 501 |
+
"model.layers.20.self_attn.v_proj.bias",
|
| 502 |
+
"model.layers.52.input_layernorm",
|
| 503 |
+
"model.layers.43.post_attention_layernorm",
|
| 504 |
+
"model.layers.51.post_attention_layernorm",
|
| 505 |
+
"model.layers.12.mlp.gate",
|
| 506 |
+
"model.layers.47.input_layernorm",
|
| 507 |
+
"model.layers.49.self_attn.q_norm",
|
| 508 |
+
"model.layers.78.self_attn.q_proj.bias",
|
| 509 |
+
"model.layers.53.input_layernorm",
|
| 510 |
+
"model.layers.11.mlp.gate",
|
| 511 |
+
"model.layers.88.self_attn.q_norm",
|
| 512 |
+
"model.layers.36.self_attn.q_proj.bias",
|
| 513 |
+
"model.layers.58.self_attn.q_norm",
|
| 514 |
+
"model.layers.77.input_layernorm",
|
| 515 |
+
"model.layers.41.post_attention_layernorm",
|
| 516 |
+
"model.layers.74.self_attn.v_proj.bias",
|
| 517 |
+
"model.layers.2.self_attn.q_norm",
|
| 518 |
+
"model.layers.5.self_attn.k_proj.bias",
|
| 519 |
+
"model.layers.92.self_attn.v_proj.bias",
|
| 520 |
+
"model.layers.58.self_attn.k_norm",
|
| 521 |
+
"model.layers.5.mlp.gate.e_score_correction_bias",
|
| 522 |
+
"model.layers.71.mlp.gate.e_score_correction_bias",
|
| 523 |
+
"model.layers.87.mlp.gate.e_score_correction_bias",
|
| 524 |
+
"model.layers.46.input_layernorm",
|
| 525 |
+
"model.layers.66.mlp.gate.e_score_correction_bias",
|
| 526 |
+
"model.layers.74.post_attention_layernorm",
|
| 527 |
+
"model.layers.80.self_attn.k_proj.bias",
|
| 528 |
+
"model.layers.25.self_attn.v_proj.bias",
|
| 529 |
+
"model.layers.14.self_attn.k_norm",
|
| 530 |
+
"model.layers.3.mlp.gate.e_score_correction_bias",
|
| 531 |
+
"model.layers.57.self_attn.q_proj.bias",
|
| 532 |
+
"model.layers.19.self_attn.q_proj.bias",
|
| 533 |
+
"model.layers.68.mlp.gate",
|
| 534 |
+
"model.layers.28.input_layernorm",
|
| 535 |
+
"model.layers.41.self_attn.q_norm",
|
| 536 |
+
"model.layers.4.mlp.gate.e_score_correction_bias",
|
| 537 |
+
"model.layers.37.input_layernorm",
|
| 538 |
+
"model.layers.91.mlp.gate",
|
| 539 |
+
"model.layers.35.mlp.gate.e_score_correction_bias",
|
| 540 |
+
"model.layers.39.mlp.gate",
|
| 541 |
+
"model.layers.30.self_attn.q_norm",
|
| 542 |
+
"model.layers.87.post_attention_layernorm",
|
| 543 |
+
"model.layers.52.self_attn.k_proj.bias",
|
| 544 |
+
"model.layers.11.self_attn.q_norm",
|
| 545 |
+
"model.layers.59.self_attn.k_proj.bias",
|
| 546 |
+
"model.layers.86.mlp.gate.e_score_correction_bias",
|
| 547 |
+
"model.layers.85.post_attention_layernorm",
|
| 548 |
+
"model.layers.83.mlp.gate",
|
| 549 |
+
"model.layers.70.input_layernorm",
|
| 550 |
+
"model.layers.69.self_attn.q_proj.bias",
|
| 551 |
+
"model.layers.92.eh_proj",
|
| 552 |
+
"model.layers.83.self_attn.k_norm",
|
| 553 |
+
"model.layers.85.self_attn.v_proj.bias",
|
| 554 |
+
"model.layers.34.input_layernorm",
|
| 555 |
+
"model.layers.70.self_attn.v_proj.bias",
|
| 556 |
+
"model.layers.34.self_attn.k_norm",
|
| 557 |
+
"model.layers.67.mlp.gate",
|
| 558 |
+
"model.layers.85.mlp.gate.e_score_correction_bias",
|
| 559 |
+
"model.layers.11.self_attn.v_proj.bias",
|
| 560 |
+
"model.layers.21.mlp.gate",
|
| 561 |
+
"model.layers.0.self_attn.k_proj.bias",
|
| 562 |
+
"model.layers.20.mlp.gate.e_score_correction_bias",
|
| 563 |
+
"model.layers.55.self_attn.k_proj.bias",
|
| 564 |
+
"model.layers.73.self_attn.k_proj.bias",
|
| 565 |
+
"model.layers.29.self_attn.q_norm",
|
| 566 |
+
"model.layers.41.self_attn.k_norm",
|
| 567 |
+
"model.layers.65.post_attention_layernorm",
|
| 568 |
+
"model.layers.46.self_attn.q_proj.bias",
|
| 569 |
+
"model.layers.83.self_attn.v_proj.bias",
|
| 570 |
+
"model.layers.73.self_attn.v_proj.bias",
|
| 571 |
+
"model.layers.82.post_attention_layernorm",
|
| 572 |
+
"model.layers.11.mlp.gate.e_score_correction_bias",
|
| 573 |
+
"model.layers.87.mlp.gate",
|
| 574 |
+
"model.layers.36.self_attn.v_proj.bias",
|
| 575 |
+
"model.layers.40.self_attn.k_proj.bias",
|
| 576 |
+
"model.layers.17.self_attn.k_proj.bias",
|
| 577 |
+
"model.layers.60.mlp.gate.e_score_correction_bias",
|
| 578 |
+
"model.layers.39.post_attention_layernorm",
|
| 579 |
+
"model.layers.61.self_attn.k_proj.bias",
|
| 580 |
+
"model.layers.91.self_attn.v_proj.bias",
|
| 581 |
+
"model.layers.49.mlp.gate.e_score_correction_bias",
|
| 582 |
+
"model.layers.19.self_attn.k_proj.bias",
|
| 583 |
+
"model.layers.32.self_attn.v_proj.bias",
|
| 584 |
+
"model.layers.77.self_attn.q_proj.bias",
|
| 585 |
+
"model.layers.33.post_attention_layernorm",
|
| 586 |
+
"model.layers.68.self_attn.k_norm",
|
| 587 |
+
"model.layers.73.self_attn.q_norm",
|
| 588 |
+
"model.layers.37.self_attn.v_proj.bias",
|
| 589 |
+
"model.layers.86.self_attn.v_proj.bias",
|
| 590 |
+
"model.layers.52.self_attn.k_norm",
|
| 591 |
+
"model.layers.6.self_attn.q_norm",
|
| 592 |
+
"model.layers.86.post_attention_layernorm",
|
| 593 |
+
"model.layers.47.self_attn.v_proj.bias",
|
| 594 |
+
"model.layers.15.self_attn.k_proj.bias",
|
| 595 |
+
"model.layers.21.self_attn.k_norm",
|
| 596 |
+
"model.layers.48.post_attention_layernorm",
|
| 597 |
+
"model.layers.82.mlp.gate",
|
| 598 |
+
"model.layers.88.self_attn.v_proj.bias",
|
| 599 |
+
"model.layers.85.self_attn.k_proj.bias",
|
| 600 |
+
"model.layers.74.mlp.gate.e_score_correction_bias",
|
| 601 |
+
"model.layers.43.mlp.gate.e_score_correction_bias",
|
| 602 |
+
"model.layers.74.mlp.gate",
|
| 603 |
+
"model.layers.84.self_attn.k_proj.bias",
|
| 604 |
+
"model.layers.29.post_attention_layernorm",
|
| 605 |
+
"model.layers.38.self_attn.k_norm",
|
| 606 |
+
"model.layers.7.self_attn.k_norm",
|
| 607 |
+
"model.layers.28.self_attn.k_proj.bias",
|
| 608 |
+
"model.layers.25.post_attention_layernorm",
|
| 609 |
+
"model.layers.35.self_attn.q_norm",
|
| 610 |
+
"model.layers.34.mlp.gate",
|
| 611 |
+
"model.layers.25.mlp.gate",
|
| 612 |
+
"model.layers.31.mlp.gate",
|
| 613 |
+
"model.layers.55.mlp.gate",
|
| 614 |
+
"model.layers.49.self_attn.q_proj.bias",
|
| 615 |
+
"model.layers.83.self_attn.q_proj.bias",
|
| 616 |
+
"model.layers.88.self_attn.k_proj.bias",
|
| 617 |
+
"model.layers.50.input_layernorm",
|
| 618 |
+
"model.layers.67.input_layernorm",
|
| 619 |
+
"model.layers.78.post_attention_layernorm",
|
| 620 |
+
"model.layers.13.mlp.gate.e_score_correction_bias",
|
| 621 |
+
"model.layers.60.self_attn.k_norm",
|
| 622 |
+
"model.layers.15.self_attn.q_norm",
|
| 623 |
+
"model.layers.72.mlp.gate",
|
| 624 |
+
"model.layers.1.self_attn.k_norm",
|
| 625 |
+
"model.layers.92.self_attn.q_norm",
|
| 626 |
+
"model.layers.20.post_attention_layernorm",
|
| 627 |
+
"model.layers.42.self_attn.k_proj.bias",
|
| 628 |
+
"model.layers.80.mlp.gate",
|
| 629 |
+
"model.layers.29.mlp.gate.e_score_correction_bias",
|
| 630 |
+
"model.layers.4.self_attn.q_norm",
|
| 631 |
+
"model.layers.60.self_attn.q_norm",
|
| 632 |
+
"model.layers.52.post_attention_layernorm",
|
| 633 |
+
"model.layers.48.self_attn.k_proj.bias",
|
| 634 |
+
"model.layers.37.self_attn.q_proj.bias",
|
| 635 |
+
"model.layers.65.self_attn.q_proj.bias",
|
| 636 |
+
"model.layers.31.post_attention_layernorm",
|
| 637 |
+
"model.layers.4.self_attn.v_proj.bias",
|
| 638 |
+
"model.layers.28.self_attn.k_norm",
|
| 639 |
+
"model.layers.38.self_attn.q_norm",
|
| 640 |
+
"model.layers.3.mlp.gate",
|
| 641 |
+
"model.layers.89.mlp.gate.e_score_correction_bias",
|
| 642 |
+
"model.layers.81.self_attn.q_proj.bias",
|
| 643 |
+
"model.layers.15.mlp.gate",
|
| 644 |
+
"model.layers.55.self_attn.v_proj.bias",
|
| 645 |
+
"model.layers.69.self_attn.q_norm",
|
| 646 |
+
"model.layers.77.self_attn.q_norm",
|
| 647 |
+
"model.layers.5.self_attn.k_norm",
|
| 648 |
+
"model.layers.9.input_layernorm",
|
| 649 |
+
"model.layers.62.mlp.gate.e_score_correction_bias",
|
| 650 |
+
"model.layers.23.post_attention_layernorm",
|
| 651 |
+
"model.layers.83.self_attn.k_proj.bias",
|
| 652 |
+
"model.layers.34.self_attn.q_proj.bias",
|
| 653 |
+
"model.layers.82.mlp.gate.e_score_correction_bias",
|
| 654 |
+
"model.layers.89.self_attn.v_proj.bias",
|
| 655 |
+
"model.layers.10.post_attention_layernorm",
|
| 656 |
+
"model.layers.42.self_attn.q_proj.bias",
|
| 657 |
+
"model.layers.2.input_layernorm",
|
| 658 |
+
"model.layers.66.mlp.gate",
|
| 659 |
+
"model.layers.87.self_attn.q_proj.bias",
|
| 660 |
+
"model.layers.46.self_attn.v_proj.bias",
|
| 661 |
+
"model.layers.17.mlp.gate.e_score_correction_bias",
|
| 662 |
+
"model.layers.16.self_attn.k_proj.bias",
|
| 663 |
+
"model.layers.2.post_attention_layernorm",
|
| 664 |
+
"model.layers.2.self_attn.k_norm",
|
| 665 |
+
"model.layers.90.self_attn.q_norm",
|
| 666 |
+
"model.layers.42.self_attn.q_norm",
|
| 667 |
+
"model.layers.37.self_attn.k_norm",
|
| 668 |
+
"model.layers.54.self_attn.k_norm",
|
| 669 |
+
"model.layers.68.input_layernorm",
|
| 670 |
+
"model.layers.3.self_attn.k_norm",
|
| 671 |
+
"model.layers.31.self_attn.k_proj.bias",
|
| 672 |
+
"model.layers.79.self_attn.k_proj.bias",
|
| 673 |
+
"model.layers.33.mlp.gate",
|
| 674 |
+
"model.layers.87.self_attn.q_norm",
|
| 675 |
+
"model.layers.68.post_attention_layernorm",
|
| 676 |
+
"model.layers.87.self_attn.k_norm",
|
| 677 |
+
"model.layers.0.self_attn.k_norm",
|
| 678 |
+
"model.layers.4.post_attention_layernorm",
|
| 679 |
+
"model.layers.26.self_attn.k_proj.bias",
|
| 680 |
+
"model.layers.39.self_attn.q_proj.bias",
|
| 681 |
+
"model.layers.71.mlp.gate",
|
| 682 |
+
"model.layers.43.self_attn.k_proj.bias",
|
| 683 |
+
"model.layers.50.post_attention_layernorm",
|
| 684 |
+
"model.layers.41.self_attn.v_proj.bias",
|
| 685 |
+
"model.layers.9.mlp.gate",
|
| 686 |
+
"model.layers.89.input_layernorm",
|
| 687 |
+
"model.layers.73.mlp.gate",
|
| 688 |
+
"model.layers.7.self_attn.k_proj.bias",
|
| 689 |
+
"model.layers.10.self_attn.v_proj.bias",
|
| 690 |
+
"model.layers.3.self_attn.q_norm",
|
| 691 |
+
"model.layers.12.self_attn.k_norm",
|
| 692 |
+
"model.layers.40.input_layernorm",
|
| 693 |
+
"model.layers.21.self_attn.v_proj.bias",
|
| 694 |
+
"model.layers.22.self_attn.k_norm",
|
| 695 |
+
"model.layers.16.post_attention_layernorm",
|
| 696 |
+
"model.layers.81.post_attention_layernorm",
|
| 697 |
+
"model.layers.56.self_attn.q_norm",
|
| 698 |
+
"model.layers.10.self_attn.k_norm",
|
| 699 |
+
"model.layers.35.mlp.gate",
|
| 700 |
+
"model.layers.65.self_attn.k_proj.bias",
|
| 701 |
+
"model.layers.41.input_layernorm",
|
| 702 |
+
"model.layers.80.input_layernorm",
|
| 703 |
+
"model.layers.31.input_layernorm",
|
| 704 |
+
"model.layers.30.self_attn.k_proj.bias",
|
| 705 |
+
"model.layers.49.self_attn.k_norm",
|
| 706 |
+
"model.layers.56.self_attn.k_norm",
|
| 707 |
+
"model.layers.92.embed_tokens",
|
| 708 |
+
"model.layers.42.mlp.gate",
|
| 709 |
+
"model.layers.84.self_attn.q_norm",
|
| 710 |
+
"model.layers.13.self_attn.v_proj.bias",
|
| 711 |
+
"model.layers.21.self_attn.k_proj.bias",
|
| 712 |
+
"model.layers.50.self_attn.q_norm",
|
| 713 |
+
"model.layers.44.self_attn.k_norm",
|
| 714 |
+
"model.layers.10.self_attn.k_proj.bias",
|
| 715 |
+
"model.layers.80.post_attention_layernorm",
|
| 716 |
+
"model.layers.12.input_layernorm",
|
| 717 |
+
"model.layers.57.self_attn.k_proj.bias",
|
| 718 |
+
"model.layers.27.post_attention_layernorm",
|
| 719 |
+
"model.layers.73.post_attention_layernorm",
|
| 720 |
+
"model.layers.90.input_layernorm",
|
| 721 |
+
"model.layers.68.self_attn.v_proj.bias",
|
| 722 |
+
"model.layers.19.mlp.gate",
|
| 723 |
+
"model.layers.9.self_attn.k_proj.bias",
|
| 724 |
+
"model.layers.54.mlp.gate",
|
| 725 |
+
"model.layers.78.self_attn.k_norm",
|
| 726 |
+
"model.layers.32.self_attn.q_proj.bias",
|
| 727 |
+
"model.layers.15.self_attn.q_proj.bias",
|
| 728 |
+
"model.layers.25.self_attn.q_norm",
|
| 729 |
+
"model.layers.23.self_attn.k_proj.bias",
|
| 730 |
+
"model.layers.57.self_attn.k_norm",
|
| 731 |
+
"model.layers.12.self_attn.q_norm",
|
| 732 |
+
"model.layers.56.self_attn.q_proj.bias",
|
| 733 |
+
"model.layers.48.input_layernorm",
|
| 734 |
+
"model.layers.44.self_attn.q_proj.bias",
|
| 735 |
+
"model.layers.45.self_attn.q_norm",
|
| 736 |
+
"model.layers.83.self_attn.q_norm",
|
| 737 |
+
"model.layers.39.self_attn.k_proj.bias",
|
| 738 |
+
"model.layers.40.self_attn.q_proj.bias",
|
| 739 |
+
"model.layers.84.self_attn.v_proj.bias",
|
| 740 |
+
"model.layers.48.mlp.gate",
|
| 741 |
+
"model.layers.15.mlp.gate.e_score_correction_bias",
|
| 742 |
+
"model.layers.67.self_attn.k_proj.bias",
|
| 743 |
+
"model.layers.62.mlp.gate",
|
| 744 |
+
"model.layers.28.mlp.gate.e_score_correction_bias",
|
| 745 |
+
"model.layers.18.self_attn.q_proj.bias",
|
| 746 |
+
"model.layers.9.self_attn.q_proj.bias",
|
| 747 |
+
"model.layers.26.self_attn.k_norm",
|
| 748 |
+
"model.layers.65.mlp.gate.e_score_correction_bias",
|
| 749 |
+
"model.layers.22.mlp.gate",
|
| 750 |
+
"model.layers.68.self_attn.k_proj.bias",
|
| 751 |
+
"model.layers.32.self_attn.k_proj.bias",
|
| 752 |
+
"model.layers.56.post_attention_layernorm",
|
| 753 |
+
"model.layers.54.self_attn.q_norm",
|
| 754 |
+
"model.layers.59.self_attn.q_norm",
|
| 755 |
+
"model.layers.71.post_attention_layernorm",
|
| 756 |
+
"model.layers.1.post_attention_layernorm",
|
| 757 |
+
"model.layers.64.self_attn.q_proj.bias",
|
| 758 |
+
"model.layers.68.mlp.gate.e_score_correction_bias",
|
| 759 |
+
"model.layers.41.mlp.gate",
|
| 760 |
+
"model.layers.22.self_attn.q_proj.bias",
|
| 761 |
+
"model.layers.31.self_attn.v_proj.bias",
|
| 762 |
+
"model.layers.74.input_layernorm",
|
| 763 |
+
"model.layers.27.self_attn.q_norm",
|
| 764 |
+
"model.layers.34.self_attn.v_proj.bias",
|
| 765 |
+
"model.layers.45.input_layernorm",
|
| 766 |
+
"model.layers.39.self_attn.v_proj.bias",
|
| 767 |
+
"model.layers.16.self_attn.k_norm",
|
| 768 |
+
"model.layers.62.input_layernorm",
|
| 769 |
+
"model.layers.24.self_attn.q_proj.bias",
|
| 770 |
+
"model.layers.17.input_layernorm",
|
| 771 |
+
"model.layers.17.mlp.gate",
|
| 772 |
+
"model.layers.29.mlp.gate",
|
| 773 |
+
"model.layers.51.self_attn.v_proj.bias",
|
| 774 |
+
"model.layers.14.post_attention_layernorm",
|
| 775 |
+
"model.layers.13.self_attn.k_norm",
|
| 776 |
+
"model.layers.43.self_attn.v_proj.bias",
|
| 777 |
+
"model.layers.56.input_layernorm",
|
| 778 |
+
"model.layers.77.self_attn.k_proj.bias",
|
| 779 |
+
"model.layers.79.mlp.gate.e_score_correction_bias",
|
| 780 |
+
"model.layers.82.input_layernorm",
|
| 781 |
+
"model.layers.26.mlp.gate",
|
| 782 |
+
"model.layers.78.self_attn.v_proj.bias",
|
| 783 |
+
"model.layers.5.self_attn.q_proj.bias",
|
| 784 |
+
"model.layers.39.mlp.gate.e_score_correction_bias",
|
| 785 |
+
"model.layers.67.self_attn.v_proj.bias",
|
| 786 |
+
"model.layers.51.self_attn.k_proj.bias",
|
| 787 |
+
"model.layers.35.self_attn.v_proj.bias",
|
| 788 |
+
"model.layers.67.self_attn.k_norm",
|
| 789 |
+
"model.layers.69.mlp.gate.e_score_correction_bias",
|
| 790 |
+
"model.layers.39.input_layernorm",
|
| 791 |
+
"model.layers.86.mlp.gate",
|
| 792 |
+
"model.layers.88.self_attn.q_proj.bias",
|
| 793 |
+
"model.layers.2.self_attn.k_proj.bias",
|
| 794 |
+
"model.layers.23.mlp.gate.e_score_correction_bias",
|
| 795 |
+
"model.layers.18.self_attn.k_proj.bias",
|
| 796 |
+
"model.layers.38.mlp.gate",
|
| 797 |
+
"model.layers.59.mlp.gate",
|
| 798 |
+
"model.layers.58.mlp.gate",
|
| 799 |
+
"model.layers.25.self_attn.q_proj.bias",
|
| 800 |
+
"model.layers.72.self_attn.q_proj.bias",
|
| 801 |
+
"model.layers.64.self_attn.k_proj.bias",
|
| 802 |
+
"model.layers.71.self_attn.v_proj.bias",
|
| 803 |
+
"model.layers.92.shared_head.norm",
|
| 804 |
+
"model.layers.43.input_layernorm",
|
| 805 |
+
"model.layers.30.self_attn.k_norm",
|
| 806 |
+
"model.layers.53.mlp.gate",
|
| 807 |
+
"model.layers.7.input_layernorm",
|
| 808 |
+
"model.layers.62.self_attn.v_proj.bias",
|
| 809 |
+
"model.layers.47.self_attn.k_norm",
|
| 810 |
+
"model.layers.53.post_attention_layernorm",
|
| 811 |
+
"model.layers.60.mlp.gate",
|
| 812 |
+
"model.layers.85.self_attn.q_proj.bias",
|
| 813 |
+
"model.layers.53.self_attn.k_proj.bias",
|
| 814 |
+
"model.layers.48.self_attn.q_proj.bias",
|
| 815 |
+
"model.layers.15.self_attn.v_proj.bias",
|
| 816 |
+
"model.layers.71.self_attn.q_proj.bias",
|
| 817 |
+
"model.layers.84.mlp.gate",
|
| 818 |
+
"model.layers.20.self_attn.q_proj.bias",
|
| 819 |
+
"model.layers.8.self_attn.v_proj.bias",
|
| 820 |
+
"model.layers.20.mlp.gate",
|
| 821 |
+
"model.layers.39.self_attn.k_norm",
|
| 822 |
+
"model.layers.57.mlp.gate",
|
| 823 |
+
"model.layers.40.post_attention_layernorm",
|
| 824 |
+
"model.layers.62.post_attention_layernorm",
|
| 825 |
+
"model.layers.68.self_attn.q_norm",
|
| 826 |
+
"model.layers.8.mlp.gate",
|
| 827 |
+
"model.layers.55.self_attn.q_proj.bias",
|
| 828 |
+
"model.layers.52.mlp.gate.e_score_correction_bias",
|
| 829 |
+
"model.layers.60.self_attn.k_proj.bias",
|
| 830 |
+
"model.layers.47.self_attn.q_proj.bias",
|
| 831 |
+
"model.layers.0.input_layernorm",
|
| 832 |
+
"model.layers.12.self_attn.k_proj.bias",
|
| 833 |
+
"model.layers.21.self_attn.q_norm",
|
| 834 |
+
"model.layers.51.self_attn.q_norm",
|
| 835 |
+
"model.layers.80.self_attn.q_proj.bias",
|
| 836 |
+
"model.layers.15.input_layernorm",
|
| 837 |
+
"model.layers.18.mlp.gate.e_score_correction_bias",
|
| 838 |
+
"model.layers.2.self_attn.q_proj.bias",
|
| 839 |
+
"model.layers.58.self_attn.k_proj.bias",
|
| 840 |
+
"model.layers.58.post_attention_layernorm",
|
| 841 |
+
"model.layers.5.self_attn.v_proj.bias",
|
| 842 |
+
"model.layers.3.self_attn.q_proj.bias",
|
| 843 |
+
"model.layers.75.mlp.gate.e_score_correction_bias",
|
| 844 |
+
"model.layers.45.self_attn.q_proj.bias",
|
| 845 |
+
"model.layers.43.self_attn.q_proj.bias",
|
| 846 |
+
"model.layers.28.self_attn.q_proj.bias",
|
| 847 |
+
"model.layers.10.mlp.gate",
|
| 848 |
+
"model.layers.36.post_attention_layernorm",
|
| 849 |
+
"model.layers.19.self_attn.k_norm",
|
| 850 |
+
"model.layers.66.post_attention_layernorm",
|
| 851 |
+
"model.layers.90.post_attention_layernorm",
|
| 852 |
+
"model.layers.23.self_attn.q_proj.bias",
|
| 853 |
+
"model.layers.87.self_attn.v_proj.bias",
|
| 854 |
+
"model.layers.36.self_attn.k_norm",
|
| 855 |
+
"model.layers.35.self_attn.k_norm",
|
| 856 |
+
"model.layers.44.input_layernorm",
|
| 857 |
+
"model.layers.24.mlp.gate",
|
| 858 |
+
"model.layers.34.post_attention_layernorm",
|
| 859 |
+
"model.layers.11.post_attention_layernorm",
|
| 860 |
+
"model.layers.13.mlp.gate",
|
| 861 |
+
"model.layers.85.input_layernorm",
|
| 862 |
+
"model.layers.52.mlp.gate",
|
| 863 |
+
"model.norm",
|
| 864 |
+
"model.layers.51.mlp.gate.e_score_correction_bias",
|
| 865 |
+
"model.layers.25.self_attn.k_proj.bias",
|
| 866 |
+
"model.layers.64.self_attn.v_proj.bias",
|
| 867 |
+
"model.layers.45.self_attn.k_proj.bias",
|
| 868 |
+
"model.layers.82.self_attn.q_proj.bias",
|
| 869 |
+
"model.layers.30.mlp.gate.e_score_correction_bias",
|
| 870 |
+
"model.layers.31.self_attn.k_norm",
|
| 871 |
+
"model.layers.67.mlp.gate.e_score_correction_bias",
|
| 872 |
+
"model.layers.52.self_attn.q_proj.bias",
|
| 873 |
+
"model.layers.57.input_layernorm",
|
| 874 |
+
"model.layers.69.mlp.gate",
|
| 875 |
+
"model.layers.85.self_attn.q_norm",
|
| 876 |
+
"model.layers.78.self_attn.q_norm",
|
| 877 |
+
"model.layers.10.self_attn.q_proj.bias",
|
| 878 |
+
"model.layers.48.self_attn.q_norm",
|
| 879 |
+
"model.layers.92.mlp.gate.e_score_correction_bias",
|
| 880 |
+
"model.layers.67.post_attention_layernorm",
|
| 881 |
+
"model.layers.28.self_attn.q_norm",
|
| 882 |
+
"model.layers.74.self_attn.k_proj.bias",
|
| 883 |
+
"model.layers.58.mlp.gate.e_score_correction_bias",
|
| 884 |
+
"model.layers.67.self_attn.q_norm",
|
| 885 |
+
"model.layers.46.self_attn.q_norm",
|
| 886 |
+
"model.layers.82.self_attn.q_norm",
|
| 887 |
+
"model.layers.9.mlp.gate.e_score_correction_bias",
|
| 888 |
+
"model.layers.16.self_attn.q_norm",
|
| 889 |
+
"model.layers.92.self_attn.q_proj.bias",
|
| 890 |
+
"model.layers.63.mlp.gate.e_score_correction_bias",
|
| 891 |
+
"model.layers.38.post_attention_layernorm",
|
| 892 |
+
"model.layers.72.self_attn.q_norm",
|
| 893 |
+
"model.layers.81.self_attn.q_norm",
|
| 894 |
+
"model.layers.63.self_attn.q_proj.bias",
|
| 895 |
+
"model.layers.48.self_attn.v_proj.bias",
|
| 896 |
+
"model.layers.6.input_layernorm",
|
| 897 |
+
"model.layers.61.mlp.gate",
|
| 898 |
+
"model.layers.14.input_layernorm",
|
| 899 |
+
"model.layers.31.self_attn.q_norm",
|
| 900 |
+
"model.layers.61.self_attn.k_norm",
|
| 901 |
+
"model.layers.83.input_layernorm",
|
| 902 |
+
"model.layers.89.post_attention_layernorm",
|
| 903 |
+
"model.layers.28.mlp.gate",
|
| 904 |
+
"model.layers.16.self_attn.v_proj.bias",
|
| 905 |
+
"model.layers.10.input_layernorm",
|
| 906 |
+
"model.layers.6.mlp.gate.e_score_correction_bias",
|
| 907 |
+
"model.layers.6.self_attn.k_norm",
|
| 908 |
+
"model.layers.11.input_layernorm",
|
| 909 |
+
"model.layers.34.self_attn.k_proj.bias",
|
| 910 |
+
"model.layers.41.self_attn.q_proj.bias",
|
| 911 |
+
"model.layers.58.self_attn.v_proj.bias",
|
| 912 |
+
"model.layers.90.self_attn.q_proj.bias",
|
| 913 |
+
"model.layers.55.input_layernorm",
|
| 914 |
+
"model.layers.13.self_attn.k_proj.bias",
|
| 915 |
+
"model.layers.64.mlp.gate.e_score_correction_bias",
|
| 916 |
+
"model.layers.47.mlp.gate.e_score_correction_bias",
|
| 917 |
+
"model.layers.42.post_attention_layernorm"
|
| 918 |
+
],
|
| 919 |
+
"quant_method": "compressed-tensors",
|
| 920 |
+
"quantization_status": "compressed"
|
| 921 |
+
}
|
| 922 |
+
}
|