File size: 6,961 Bytes
e378b44
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
{
  "architectures": [
    "DeepseekV4ForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 0,
  "compress_ratios": [
    0,
    0,
    4,
    128,
    4
  ],
  "compress_rope_parameters": {
    "beta_fast": 32,
    "beta_slow": 1,
    "factor": 16,
    "original_max_position_embeddings": 65536,
    "partial_rotary_factor": 0.125,
    "rope_theta": 160000.0,
    "rope_type": "yarn",
    "type": "yarn"
  },
  "compress_rope_theta": 160000.0,
  "dtype": "bfloat16",
  "eos_token_id": 1,
  "first_k_dense_replace": null,
  "hc_eps": 1e-06,
  "hc_mult": 4,
  "hc_sinkhorn_iters": 20,
  "head_dim": 512,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "index_head_dim": 128,
  "index_n_heads": 64,
  "index_topk": 512,
  "initializer_range": 0.02,
  "intermediate_size": 18432,
  "kv_lora_rank": null,
  "max_position_embeddings": 1048576,
  "model_type": "deepseek_v4",
  "moe_intermediate_size": 2048,
  "n_group": null,
  "n_routed_experts": 256,
  "n_shared_experts": 1,
  "norm_topk_prob": true,
  "num_attention_heads": 64,
  "num_experts_per_tok": 6,
  "num_hash_layers": 3,
  "num_hidden_layers": 5,
  "num_key_value_heads": 1,
  "num_nextn_predict_layers": 0,
  "o_groups": 8,
  "o_lora_rank": 1024,
  "output_router_logits": false,
  "pad_token_id": null,
  "partial_rotary_factor": 0.125,
  "pretraining_tp": 1,
  "q_lora_rank": 1024,
  "qk_nope_head_dim": 448,
  "qk_rope_head_dim": 64,
  "quantization": {
    "bits": 4,
    "group_size": 32,
    "mode": "mxfp4",
    "model.layers.0.ffn.switch_mlp.down_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.0.ffn.switch_mlp.gate_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.0.ffn.switch_mlp.up_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.1.ffn.switch_mlp.down_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.1.ffn.switch_mlp.gate_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.1.ffn.switch_mlp.up_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.2.ffn.switch_mlp.down_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.2.ffn.switch_mlp.gate_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.2.ffn.switch_mlp.up_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.3.ffn.switch_mlp.down_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.3.ffn.switch_mlp.gate_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.3.ffn.switch_mlp.up_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.4.ffn.switch_mlp.down_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.4.ffn.switch_mlp.gate_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    },
    "model.layers.4.ffn.switch_mlp.up_proj": {
      "bits": 4,
      "group_size": 32,
      "mode": "mxfp4"
    }
  },
  "quantization_config": {
    "config_groups": {
      "group_0": {
        "format": "nvfp4-pack-quantized",
        "input_activations": {
          "actorder": null,
          "block_structure": null,
          "dynamic": "local",
          "group_size": 16,
          "num_bits": 4,
          "observer": "static_minmax",
          "observer_kwargs": {},
          "scale_dtype": "torch.float8_e4m3fn",
          "strategy": "tensor_group",
          "symmetric": true,
          "type": "float",
          "zp_dtype": null
        },
        "output_activations": null,
        "targets": [
          "re:model.*mlp.*(gate|up|down)_proj$"
        ],
        "weights": {
          "actorder": null,
          "block_structure": null,
          "dynamic": false,
          "group_size": 16,
          "num_bits": 4,
          "observer": "memoryless_minmax",
          "observer_kwargs": {},
          "scale_dtype": "torch.float8_e4m3fn",
          "strategy": "tensor_group",
          "symmetric": true,
          "type": "float",
          "zp_dtype": null
        }
      }
    },
    "format": "nvfp4-pack-quantized",
    "global_compression_ratio": null,
    "ignore": [
      "model.layers.0.self_attn.wq_a",
      "model.layers.0.self_attn.wq_b",
      "model.layers.0.self_attn.wkv",
      "model.layers.0.self_attn.wo_b",
      "model.layers.1.self_attn.wq_a",
      "model.layers.1.self_attn.wq_b",
      "model.layers.1.self_attn.wkv",
      "model.layers.1.self_attn.wo_b",
      "model.layers.2.self_attn.wq_a",
      "model.layers.2.self_attn.wq_b",
      "model.layers.2.self_attn.wkv",
      "model.layers.2.self_attn.wo_b",
      "model.layers.2.self_attn.compressor.wkv",
      "model.layers.2.self_attn.compressor.wgate",
      "model.layers.2.self_attn.compressor.indexer.wkv",
      "model.layers.2.self_attn.compressor.indexer.wgate",
      "model.layers.2.self_attn.compressor.indexer.wq_b",
      "model.layers.2.self_attn.compressor.indexer.weights_proj",
      "model.layers.3.self_attn.wq_a",
      "model.layers.3.self_attn.wq_b",
      "model.layers.3.self_attn.wkv",
      "model.layers.3.self_attn.wo_b",
      "model.layers.3.self_attn.compressor.wkv",
      "model.layers.3.self_attn.compressor.wgate",
      "model.layers.4.self_attn.wq_a",
      "model.layers.4.self_attn.wq_b",
      "model.layers.4.self_attn.wkv",
      "model.layers.4.self_attn.wo_b",
      "model.layers.4.self_attn.compressor.wkv",
      "model.layers.4.self_attn.compressor.wgate",
      "model.layers.4.self_attn.compressor.indexer.wkv",
      "model.layers.4.self_attn.compressor.indexer.wgate",
      "model.layers.4.self_attn.compressor.indexer.wq_b",
      "model.layers.4.self_attn.compressor.indexer.weights_proj",
      "lm_head"
    ],
    "kv_cache_scheme": null,
    "quant_method": "compressed-tensors",
    "quantization_status": "compressed",
    "sparsity_config": {},
    "transform_config": {},
    "version": "0.15.1.dev15+g11daf97.d20260427"
  },
  "rms_norm_eps": 1e-06,
  "rope_interleave": true,
  "rope_parameters": {
    "beta_fast": 32,
    "beta_slow": 1,
    "factor": 16,
    "original_max_position_embeddings": 65536,
    "partial_rotary_factor": 0.125,
    "rope_theta": 10000.0,
    "rope_type": "yarn",
    "type": "yarn"
  },
  "rope_theta": 10000.0,
  "routed_scaling_factor": 1.5,
  "router_aux_loss_coef": 0.001,
  "router_jitter_noise": 0.0,
  "scoring_func": "sqrtsoftplus",
  "sliding_window": 128,
  "swiglu_limit": 10.0,
  "tie_word_embeddings": false,
  "topk_group": null,
  "topk_method": "noaux_tc",
  "transformers_version": "5.7.0.dev0",
  "use_cache": true,
  "v_head_dim": null,
  "vocab_size": 129280
}