Hugging Face's logo

cveavy
/
Test

Safetensors
quasar
Model card Files
xet
 Community
cveavy commited on
Commit
0f4946a
·
verified ·
1 Parent(s): 7e9cb91

Upload folder using huggingface_hub

Browse files
Files changed (12) hide show
  1. .gitattributes +1 -0
  2. config.json +138 -0
  3. generation_config.json +10 -0
  4. model-00001-of-00006.safetensors +3 -0
  5. model-00002-of-00006.safetensors +3 -0
  6. model-00003-of-00006.safetensors +3 -0
  7. model-00004-of-00006.safetensors +3 -0
  8. model-00005-of-00006.safetensors +3 -0
  9. model-00006-of-00006.safetensors +3 -0
  10. model.safetensors.index.json +1008 -0
  11. tokenizer.json +3 -0
  12. tokenizer_config.json +26 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
config.json ADDED
@@ -0,0 +1,138 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "allow_neg_eigval": false,
3
+ "architectures": [
4
+ "QuasarForCausalLM"
5
+ ],
6
+ "attn_mode": "chunk",
7
+ "bigmac_r": 0.25,
8
+ "bos_token_id": 2,
9
+ "conv_bias": false,
10
+ "conv_size": 4,
11
+ "d_ff": 11008,
12
+ "d_model": 4096,
13
+ "dense_input_layers": 4,
14
+ "dropout": 0.0,
15
+ "dtype": "bfloat16",
16
+ "eos_token_id": 1,
17
+ "expand_k": 0.5,
18
+ "expand_v": 1.0,
19
+ "fuse_cross_entropy": true,
20
+ "fuse_norm": true,
21
+ "fuse_swiglu": true,
22
+ "gated_layers": 2,
23
+ "gla_mode": "chunk",
24
+ "gradient_checkpointing": false,
25
+ "head_dim": 128,
26
+ "hidden_act": "silu",
27
+ "hidden_ratio": 4,
28
+ "hidden_size": 4096,
29
+ "hybrid_layer_types": [
30
+ "quasar",
31
+ "quasar",
32
+ "quasar",
33
+ "quasar",
34
+ "gla",
35
+ "gla",
36
+ "quasar",
37
+ "quasar",
38
+ "quasar",
39
+ "quasar",
40
+ "gla",
41
+ "gla",
42
+ "quasar",
43
+ "quasar",
44
+ "quasar",
45
+ "quasar",
46
+ "gla",
47
+ "gla",
48
+ "quasar",
49
+ "quasar",
50
+ "quasar",
51
+ "quasar",
52
+ "gla",
53
+ "gla",
54
+ "quasar",
55
+ "quasar",
56
+ "quasar",
57
+ "quasar",
58
+ "gla",
59
+ "gla",
60
+ "quasar",
61
+ "quasar"
62
+ ],
63
+ "initializer_range": 0.02,
64
+ "intermediate_size": 11008,
65
+ "layer_types": [
66
+ "linear_attention",
67
+ "linear_attention",
68
+ "linear_attention",
69
+ "linear_attention",
70
+ "linear_attention",
71
+ "linear_attention",
72
+ "linear_attention",
73
+ "linear_attention",
74
+ "linear_attention",
75
+ "linear_attention",
76
+ "linear_attention",
77
+ "linear_attention",
78
+ "linear_attention",
79
+ "linear_attention",
80
+ "linear_attention",
81
+ "linear_attention",
82
+ "linear_attention",
83
+ "linear_attention",
84
+ "linear_attention",
85
+ "linear_attention",
86
+ "linear_attention",
87
+ "linear_attention",
88
+ "linear_attention",
89
+ "linear_attention",
90
+ "linear_attention",
91
+ "linear_attention",
92
+ "linear_attention",
93
+ "linear_attention",
94
+ "linear_attention",
95
+ "linear_attention",
96
+ "linear_attention",
97
+ "linear_attention"
98
+ ],
99
+ "looped_injection_init": 0.1,
100
+ "max_position_embeddings": 16384,
101
+ "max_seq_len": 16384,
102
+ "memory_dim": 128,
103
+ "memory_slots": 128,
104
+ "model_type": "quasar",
105
+ "moe_aux_loss_coeff": 0.0001,
106
+ "moe_type": "bigmac",
107
+ "moe_z_loss_coeff": 0.0001,
108
+ "n_heads": 32,
109
+ "n_layers": 32,
110
+ "norm_eps": 1e-06,
111
+ "num_attention_heads": 32,
112
+ "num_heads": 32,
113
+ "num_hidden_layers": 32,
114
+ "num_key_value_heads": 32,
115
+ "num_loops": 1,
116
+ "num_routed_experts": 56,
117
+ "num_shared_experts": 1,
118
+ "num_v_heads": null,
119
+ "pad_token_id": 0,
120
+ "quasar_layers": 4,
121
+ "residual_scale": 0.1,
122
+ "rms_norm_eps": 1e-06,
123
+ "rope_theta": 1000000.0,
124
+ "routed_expert_size": 1024,
125
+ "shared_expert_size": 2048,
126
+ "smebu_beta": 0.5,
127
+ "smebu_kappa": 2.0,
128
+ "smebu_lambda": 0.002,
129
+ "tie_word_embeddings": true,
130
+ "top_k": 8,
131
+ "transformers_version": "5.3.0",
132
+ "use_cache": false,
133
+ "use_gla_first": false,
134
+ "use_l2warp": false,
135
+ "use_looped_injection": false,
136
+ "use_short_conv": true,
137
+ "vocab_size": 262144
138
+ }
generation_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 2,
4
+ "eos_token_id": 1,
5
+ "output_attentions": false,
6
+ "output_hidden_states": false,
7
+ "pad_token_id": 0,
8
+ "transformers_version": "5.7.0",
9
+ "use_cache": true
10
+ }
model-00001-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dc932cc219433954c2a615f1173ceca1a8ee47655ef91f4790f9b40433b4dcde
3
+ size 9994101706
model-00002-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7d911dc879ea7ec0a185b865fa4bc00909dc96a35266b4e82ac383735a5bd76
3
+ size 9996290910
model-00003-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf39c271742ff55d29bd84cc8492966fd68e40cecbee1fb58ec101bc5c14ca05
3
+ size 9626386346
model-00004-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:56dd03338d9f22d421ebd962dc9bcfa0713776669875d5d99ed896ce1c55294a
3
+ size 9327332134
model-00005-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af1e6db13066747b363cd6b1019a747971c853e86a01646a225306f1dcf0a066
3
+ size 9826041924
model-00006-of-00006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff95a6e5b672a3ffdb4261d30d4e548946591f45b4cdd9765905335a327d88ac
3
+ size 3124032584
model.safetensors.index.json ADDED
@@ -0,0 +1,1008 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 51894074604
4
+ },
5
+ "weight_map": {
6
+ "lm_head.weight": "model-00001-of-00006.safetensors",
7
+ "model.all_moe_bias": "model-00001-of-00006.safetensors",
8
+ "model.all_moe_max_vio": "model-00001-of-00006.safetensors",
9
+ "model.all_moe_momentum": "model-00001-of-00006.safetensors",
10
+ "model.embed_norm.weight": "model-00001-of-00006.safetensors",
11
+ "model.embed_tokens.weight": "model-00001-of-00006.safetensors",
12
+ "model.layers.0.attn.A_log": "model-00001-of-00006.safetensors",
13
+ "model.layers.0.attn.b_proj.weight": "model-00001-of-00006.safetensors",
14
+ "model.layers.0.attn.dt_bias": "model-00001-of-00006.safetensors",
15
+ "model.layers.0.attn.f_proj.weight": "model-00001-of-00006.safetensors",
16
+ "model.layers.0.attn.g_proj.0.weight": "model-00001-of-00006.safetensors",
17
+ "model.layers.0.attn.g_proj.1.bias": "model-00001-of-00006.safetensors",
18
+ "model.layers.0.attn.g_proj.1.weight": "model-00001-of-00006.safetensors",
19
+ "model.layers.0.attn.k_conv1d.weight": "model-00001-of-00006.safetensors",
20
+ "model.layers.0.attn.k_proj.weight": "model-00001-of-00006.safetensors",
21
+ "model.layers.0.attn.o_norm.weight": "model-00001-of-00006.safetensors",
22
+ "model.layers.0.attn.o_proj.weight": "model-00001-of-00006.safetensors",
23
+ "model.layers.0.attn.q_conv1d.weight": "model-00001-of-00006.safetensors",
24
+ "model.layers.0.attn.q_proj.weight": "model-00001-of-00006.safetensors",
25
+ "model.layers.0.attn.v_conv1d.weight": "model-00001-of-00006.safetensors",
26
+ "model.layers.0.attn.v_proj.weight": "model-00001-of-00006.safetensors",
27
+ "model.layers.0.ffn.down.weight": "model-00001-of-00006.safetensors",
28
+ "model.layers.0.ffn.gate.weight": "model-00001-of-00006.safetensors",
29
+ "model.layers.0.ffn.up.weight": "model-00001-of-00006.safetensors",
30
+ "model.layers.0.injection_gate": "model-00001-of-00006.safetensors",
31
+ "model.layers.0.ln1.weight": "model-00001-of-00006.safetensors",
32
+ "model.layers.0.ln1_out.weight": "model-00001-of-00006.safetensors",
33
+ "model.layers.0.ln2.weight": "model-00001-of-00006.safetensors",
34
+ "model.layers.0.ln2_out.weight": "model-00001-of-00006.safetensors",
35
+ "model.layers.1.attn.A_log": "model-00001-of-00006.safetensors",
36
+ "model.layers.1.attn.b_proj.weight": "model-00001-of-00006.safetensors",
37
+ "model.layers.1.attn.dt_bias": "model-00001-of-00006.safetensors",
38
+ "model.layers.1.attn.f_proj.weight": "model-00001-of-00006.safetensors",
39
+ "model.layers.1.attn.g_proj.0.weight": "model-00001-of-00006.safetensors",
40
+ "model.layers.1.attn.g_proj.1.bias": "model-00001-of-00006.safetensors",
41
+ "model.layers.1.attn.g_proj.1.weight": "model-00001-of-00006.safetensors",
42
+ "model.layers.1.attn.k_conv1d.weight": "model-00001-of-00006.safetensors",
43
+ "model.layers.1.attn.k_proj.weight": "model-00001-of-00006.safetensors",
44
+ "model.layers.1.attn.o_norm.weight": "model-00001-of-00006.safetensors",
45
+ "model.layers.1.attn.o_proj.weight": "model-00001-of-00006.safetensors",
46
+ "model.layers.1.attn.q_conv1d.weight": "model-00001-of-00006.safetensors",
47
+ "model.layers.1.attn.q_proj.weight": "model-00001-of-00006.safetensors",
48
+ "model.layers.1.attn.v_conv1d.weight": "model-00001-of-00006.safetensors",
49
+ "model.layers.1.attn.v_proj.weight": "model-00001-of-00006.safetensors",
50
+ "model.layers.1.ffn.down.weight": "model-00001-of-00006.safetensors",
51
+ "model.layers.1.ffn.gate.weight": "model-00001-of-00006.safetensors",
52
+ "model.layers.1.ffn.up.weight": "model-00001-of-00006.safetensors",
53
+ "model.layers.1.injection_gate": "model-00001-of-00006.safetensors",
54
+ "model.layers.1.ln1.weight": "model-00001-of-00006.safetensors",
55
+ "model.layers.1.ln1_out.weight": "model-00001-of-00006.safetensors",
56
+ "model.layers.1.ln2.weight": "model-00001-of-00006.safetensors",
57
+ "model.layers.1.ln2_out.weight": "model-00001-of-00006.safetensors",
58
+ "model.layers.10.C_to_hidden.weight": "model-00001-of-00006.safetensors",
59
+ "model.layers.10.W_alpha.bias": "model-00001-of-00006.safetensors",
60
+ "model.layers.10.W_alpha.weight": "model-00001-of-00006.safetensors",
61
+ "model.layers.10.attn.g_norm_swish_gate.weight": "model-00001-of-00006.safetensors",
62
+ "model.layers.10.attn.g_proj.weight": "model-00001-of-00006.safetensors",
63
+ "model.layers.10.attn.gk_proj.0.weight": "model-00001-of-00006.safetensors",
64
+ "model.layers.10.attn.gk_proj.1.bias": "model-00001-of-00006.safetensors",
65
+ "model.layers.10.attn.gk_proj.1.weight": "model-00001-of-00006.safetensors",
66
+ "model.layers.10.attn.k_proj.weight": "model-00001-of-00006.safetensors",
67
+ "model.layers.10.attn.o_proj.weight": "model-00001-of-00006.safetensors",
68
+ "model.layers.10.attn.q_proj.weight": "model-00001-of-00006.safetensors",
69
+ "model.layers.10.attn.v_proj.weight": "model-00001-of-00006.safetensors",
70
+ "model.layers.10.ffn.experts_w12": "model-00001-of-00006.safetensors",
71
+ "model.layers.10.ffn.experts_w3": "model-00001-of-00006.safetensors",
72
+ "model.layers.10.ffn.max_vio": "model-00001-of-00006.safetensors",
73
+ "model.layers.10.ffn.router.router_weights": "model-00001-of-00006.safetensors",
74
+ "model.layers.10.ffn.shared_experts.0.down.weight": "model-00001-of-00006.safetensors",
75
+ "model.layers.10.ffn.shared_experts.0.gate.weight": "model-00001-of-00006.safetensors",
76
+ "model.layers.10.ffn.shared_experts.0.up.weight": "model-00001-of-00006.safetensors",
77
+ "model.layers.10.ffn.w_down_proj.weight": "model-00001-of-00006.safetensors",
78
+ "model.layers.10.ffn.w_up_proj.weight": "model-00001-of-00006.safetensors",
79
+ "model.layers.10.injection_gate": "model-00001-of-00006.safetensors",
80
+ "model.layers.10.ln1.weight": "model-00001-of-00006.safetensors",
81
+ "model.layers.10.ln1_out.weight": "model-00001-of-00006.safetensors",
82
+ "model.layers.10.ln2.weight": "model-00001-of-00006.safetensors",
83
+ "model.layers.10.ln2_out.weight": "model-00001-of-00006.safetensors",
84
+ "model.layers.10.memory.W_eta.bias": "model-00001-of-00006.safetensors",
85
+ "model.layers.10.memory.W_eta.weight": "model-00001-of-00006.safetensors",
86
+ "model.layers.10.memory.W_qkv_mem.weight": "model-00001-of-00006.safetensors",
87
+ "model.layers.10.memory.compress_z.0.weight": "model-00001-of-00006.safetensors",
88
+ "model.layers.10.memory.compress_z.2.weight": "model-00001-of-00006.safetensors",
89
+ "model.layers.10.memory.eta_channels": "model-00001-of-00006.safetensors",
90
+ "model.layers.10.memory.input_norm.bias": "model-00001-of-00006.safetensors",
91
+ "model.layers.10.memory.input_norm.weight": "model-00001-of-00006.safetensors",
92
+ "model.layers.10.memory.summary_proj.bias": "model-00001-of-00006.safetensors",
93
+ "model.layers.10.memory.summary_proj.weight": "model-00001-of-00006.safetensors",
94
+ "model.layers.10.memory.summary_query": "model-00001-of-00006.safetensors",
95
+ "model.layers.10.memory.temperature": "model-00001-of-00006.safetensors",
96
+ "model.layers.11.C_to_hidden.weight": "model-00001-of-00006.safetensors",
97
+ "model.layers.11.W_alpha.bias": "model-00001-of-00006.safetensors",
98
+ "model.layers.11.W_alpha.weight": "model-00001-of-00006.safetensors",
99
+ "model.layers.11.attn.g_norm_swish_gate.weight": "model-00001-of-00006.safetensors",
100
+ "model.layers.11.attn.g_proj.weight": "model-00001-of-00006.safetensors",
101
+ "model.layers.11.attn.gk_proj.0.weight": "model-00001-of-00006.safetensors",
102
+ "model.layers.11.attn.gk_proj.1.bias": "model-00001-of-00006.safetensors",
103
+ "model.layers.11.attn.gk_proj.1.weight": "model-00001-of-00006.safetensors",
104
+ "model.layers.11.attn.k_proj.weight": "model-00001-of-00006.safetensors",
105
+ "model.layers.11.attn.o_proj.weight": "model-00001-of-00006.safetensors",
106
+ "model.layers.11.attn.q_proj.weight": "model-00001-of-00006.safetensors",
107
+ "model.layers.11.attn.v_proj.weight": "model-00001-of-00006.safetensors",
108
+ "model.layers.11.ffn.experts_w12": "model-00001-of-00006.safetensors",
109
+ "model.layers.11.ffn.experts_w3": "model-00001-of-00006.safetensors",
110
+ "model.layers.11.ffn.max_vio": "model-00001-of-00006.safetensors",
111
+ "model.layers.11.ffn.router.router_weights": "model-00001-of-00006.safetensors",
112
+ "model.layers.11.ffn.shared_experts.0.down.weight": "model-00001-of-00006.safetensors",
113
+ "model.layers.11.ffn.shared_experts.0.gate.weight": "model-00001-of-00006.safetensors",
114
+ "model.layers.11.ffn.shared_experts.0.up.weight": "model-00001-of-00006.safetensors",
115
+ "model.layers.11.ffn.w_down_proj.weight": "model-00001-of-00006.safetensors",
116
+ "model.layers.11.ffn.w_up_proj.weight": "model-00001-of-00006.safetensors",
117
+ "model.layers.11.injection_gate": "model-00001-of-00006.safetensors",
118
+ "model.layers.11.ln1.weight": "model-00001-of-00006.safetensors",
119
+ "model.layers.11.ln1_out.weight": "model-00001-of-00006.safetensors",
120
+ "model.layers.11.ln2.weight": "model-00001-of-00006.safetensors",
121
+ "model.layers.11.ln2_out.weight": "model-00001-of-00006.safetensors",
122
+ "model.layers.11.memory.W_eta.bias": "model-00001-of-00006.safetensors",
123
+ "model.layers.11.memory.W_eta.weight": "model-00001-of-00006.safetensors",
124
+ "model.layers.11.memory.W_qkv_mem.weight": "model-00001-of-00006.safetensors",
125
+ "model.layers.11.memory.compress_z.0.weight": "model-00001-of-00006.safetensors",
126
+ "model.layers.11.memory.compress_z.2.weight": "model-00001-of-00006.safetensors",
127
+ "model.layers.11.memory.eta_channels": "model-00001-of-00006.safetensors",
128
+ "model.layers.11.memory.input_norm.bias": "model-00001-of-00006.safetensors",
129
+ "model.layers.11.memory.input_norm.weight": "model-00001-of-00006.safetensors",
130
+ "model.layers.11.memory.summary_proj.bias": "model-00001-of-00006.safetensors",
131
+ "model.layers.11.memory.summary_proj.weight": "model-00001-of-00006.safetensors",
132
+ "model.layers.11.memory.summary_query": "model-00001-of-00006.safetensors",
133
+ "model.layers.11.memory.temperature": "model-00001-of-00006.safetensors",
134
+ "model.layers.12.attn.A_log": "model-00001-of-00006.safetensors",
135
+ "model.layers.12.attn.b_proj.weight": "model-00001-of-00006.safetensors",
136
+ "model.layers.12.attn.dt_bias": "model-00001-of-00006.safetensors",
137
+ "model.layers.12.attn.f_proj.weight": "model-00001-of-00006.safetensors",
138
+ "model.layers.12.attn.g_proj.0.weight": "model-00001-of-00006.safetensors",
139
+ "model.layers.12.attn.g_proj.1.bias": "model-00001-of-00006.safetensors",
140
+ "model.layers.12.attn.g_proj.1.weight": "model-00001-of-00006.safetensors",
141
+ "model.layers.12.attn.k_conv1d.weight": "model-00001-of-00006.safetensors",
142
+ "model.layers.12.attn.k_proj.weight": "model-00001-of-00006.safetensors",
143
+ "model.layers.12.attn.o_norm.weight": "model-00001-of-00006.safetensors",
144
+ "model.layers.12.attn.o_proj.weight": "model-00001-of-00006.safetensors",
145
+ "model.layers.12.attn.q_conv1d.weight": "model-00001-of-00006.safetensors",
146
+ "model.layers.12.attn.q_proj.weight": "model-00001-of-00006.safetensors",
147
+ "model.layers.12.attn.v_conv1d.weight": "model-00001-of-00006.safetensors",
148
+ "model.layers.12.attn.v_proj.weight": "model-00001-of-00006.safetensors",
149
+ "model.layers.12.ffn.experts_w12": "model-00001-of-00006.safetensors",
150
+ "model.layers.12.ffn.experts_w3": "model-00001-of-00006.safetensors",
151
+ "model.layers.12.ffn.max_vio": "model-00001-of-00006.safetensors",
152
+ "model.layers.12.ffn.router.router_weights": "model-00001-of-00006.safetensors",
153
+ "model.layers.12.ffn.shared_experts.0.down.weight": "model-00002-of-00006.safetensors",
154
+ "model.layers.12.ffn.shared_experts.0.gate.weight": "model-00002-of-00006.safetensors",
155
+ "model.layers.12.ffn.shared_experts.0.up.weight": "model-00002-of-00006.safetensors",
156
+ "model.layers.12.ffn.w_down_proj.weight": "model-00002-of-00006.safetensors",
157
+ "model.layers.12.ffn.w_up_proj.weight": "model-00002-of-00006.safetensors",
158
+ "model.layers.12.injection_gate": "model-00002-of-00006.safetensors",
159
+ "model.layers.12.ln1.weight": "model-00002-of-00006.safetensors",
160
+ "model.layers.12.ln1_out.weight": "model-00002-of-00006.safetensors",
161
+ "model.layers.12.ln2.weight": "model-00002-of-00006.safetensors",
162
+ "model.layers.12.ln2_out.weight": "model-00002-of-00006.safetensors",
163
+ "model.layers.13.attn.A_log": "model-00002-of-00006.safetensors",
164
+ "model.layers.13.attn.b_proj.weight": "model-00002-of-00006.safetensors",
165
+ "model.layers.13.attn.dt_bias": "model-00002-of-00006.safetensors",
166
+ "model.layers.13.attn.f_proj.weight": "model-00002-of-00006.safetensors",
167
+ "model.layers.13.attn.g_proj.0.weight": "model-00002-of-00006.safetensors",
168
+ "model.layers.13.attn.g_proj.1.bias": "model-00002-of-00006.safetensors",
169
+ "model.layers.13.attn.g_proj.1.weight": "model-00002-of-00006.safetensors",
170
+ "model.layers.13.attn.k_conv1d.weight": "model-00002-of-00006.safetensors",
171
+ "model.layers.13.attn.k_proj.weight": "model-00002-of-00006.safetensors",
172
+ "model.layers.13.attn.o_norm.weight": "model-00002-of-00006.safetensors",
173
+ "model.layers.13.attn.o_proj.weight": "model-00002-of-00006.safetensors",
174
+ "model.layers.13.attn.q_conv1d.weight": "model-00002-of-00006.safetensors",
175
+ "model.layers.13.attn.q_proj.weight": "model-00002-of-00006.safetensors",
176
+ "model.layers.13.attn.v_conv1d.weight": "model-00002-of-00006.safetensors",
177
+ "model.layers.13.attn.v_proj.weight": "model-00002-of-00006.safetensors",
178
+ "model.layers.13.ffn.experts_w12": "model-00002-of-00006.safetensors",
179
+ "model.layers.13.ffn.experts_w3": "model-00002-of-00006.safetensors",
180
+ "model.layers.13.ffn.max_vio": "model-00002-of-00006.safetensors",
181
+ "model.layers.13.ffn.router.router_weights": "model-00002-of-00006.safetensors",
182
+ "model.layers.13.ffn.shared_experts.0.down.weight": "model-00002-of-00006.safetensors",
183
+ "model.layers.13.ffn.shared_experts.0.gate.weight": "model-00002-of-00006.safetensors",
184
+ "model.layers.13.ffn.shared_experts.0.up.weight": "model-00002-of-00006.safetensors",
185
+ "model.layers.13.ffn.w_down_proj.weight": "model-00002-of-00006.safetensors",
186
+ "model.layers.13.ffn.w_up_proj.weight": "model-00002-of-00006.safetensors",
187
+ "model.layers.13.injection_gate": "model-00002-of-00006.safetensors",
188
+ "model.layers.13.ln1.weight": "model-00002-of-00006.safetensors",
189
+ "model.layers.13.ln1_out.weight": "model-00002-of-00006.safetensors",
190
+ "model.layers.13.ln2.weight": "model-00002-of-00006.safetensors",
191
+ "model.layers.13.ln2_out.weight": "model-00002-of-00006.safetensors",
192
+ "model.layers.14.attn.A_log": "model-00002-of-00006.safetensors",
193
+ "model.layers.14.attn.b_proj.weight": "model-00002-of-00006.safetensors",
194
+ "model.layers.14.attn.dt_bias": "model-00002-of-00006.safetensors",
195
+ "model.layers.14.attn.f_proj.weight": "model-00002-of-00006.safetensors",
196
+ "model.layers.14.attn.g_proj.0.weight": "model-00002-of-00006.safetensors",
197
+ "model.layers.14.attn.g_proj.1.bias": "model-00002-of-00006.safetensors",
198
+ "model.layers.14.attn.g_proj.1.weight": "model-00002-of-00006.safetensors",
199
+ "model.layers.14.attn.k_conv1d.weight": "model-00002-of-00006.safetensors",
200
+ "model.layers.14.attn.k_proj.weight": "model-00002-of-00006.safetensors",
201
+ "model.layers.14.attn.o_norm.weight": "model-00002-of-00006.safetensors",
202
+ "model.layers.14.attn.o_proj.weight": "model-00002-of-00006.safetensors",
203
+ "model.layers.14.attn.q_conv1d.weight": "model-00002-of-00006.safetensors",
204
+ "model.layers.14.attn.q_proj.weight": "model-00002-of-00006.safetensors",
205
+ "model.layers.14.attn.v_conv1d.weight": "model-00002-of-00006.safetensors",
206
+ "model.layers.14.attn.v_proj.weight": "model-00002-of-00006.safetensors",
207
+ "model.layers.14.ffn.experts_w12": "model-00002-of-00006.safetensors",
208
+ "model.layers.14.ffn.experts_w3": "model-00002-of-00006.safetensors",
209
+ "model.layers.14.ffn.max_vio": "model-00002-of-00006.safetensors",
210
+ "model.layers.14.ffn.router.router_weights": "model-00002-of-00006.safetensors",
211
+ "model.layers.14.ffn.shared_experts.0.down.weight": "model-00002-of-00006.safetensors",
212
+ "model.layers.14.ffn.shared_experts.0.gate.weight": "model-00002-of-00006.safetensors",
213
+ "model.layers.14.ffn.shared_experts.0.up.weight": "model-00002-of-00006.safetensors",
214
+ "model.layers.14.ffn.w_down_proj.weight": "model-00002-of-00006.safetensors",
215
+ "model.layers.14.ffn.w_up_proj.weight": "model-00002-of-00006.safetensors",
216
+ "model.layers.14.injection_gate": "model-00002-of-00006.safetensors",
217
+ "model.layers.14.ln1.weight": "model-00002-of-00006.safetensors",
218
+ "model.layers.14.ln1_out.weight": "model-00002-of-00006.safetensors",
219
+ "model.layers.14.ln2.weight": "model-00002-of-00006.safetensors",
220
+ "model.layers.14.ln2_out.weight": "model-00002-of-00006.safetensors",
221
+ "model.layers.15.attn.A_log": "model-00002-of-00006.safetensors",
222
+ "model.layers.15.attn.b_proj.weight": "model-00002-of-00006.safetensors",
223
+ "model.layers.15.attn.dt_bias": "model-00002-of-00006.safetensors",
224
+ "model.layers.15.attn.f_proj.weight": "model-00002-of-00006.safetensors",
225
+ "model.layers.15.attn.g_proj.0.weight": "model-00002-of-00006.safetensors",
226
+ "model.layers.15.attn.g_proj.1.bias": "model-00002-of-00006.safetensors",
227
+ "model.layers.15.attn.g_proj.1.weight": "model-00002-of-00006.safetensors",
228
+ "model.layers.15.attn.k_conv1d.weight": "model-00002-of-00006.safetensors",
229
+ "model.layers.15.attn.k_proj.weight": "model-00002-of-00006.safetensors",
230
+ "model.layers.15.attn.o_norm.weight": "model-00002-of-00006.safetensors",
231
+ "model.layers.15.attn.o_proj.weight": "model-00002-of-00006.safetensors",
232
+ "model.layers.15.attn.q_conv1d.weight": "model-00002-of-00006.safetensors",
233
+ "model.layers.15.attn.q_proj.weight": "model-00002-of-00006.safetensors",
234
+ "model.layers.15.attn.v_conv1d.weight": "model-00002-of-00006.safetensors",
235
+ "model.layers.15.attn.v_proj.weight": "model-00002-of-00006.safetensors",
236
+ "model.layers.15.ffn.experts_w12": "model-00002-of-00006.safetensors",
237
+ "model.layers.15.ffn.experts_w3": "model-00002-of-00006.safetensors",
238
+ "model.layers.15.ffn.max_vio": "model-00002-of-00006.safetensors",
239
+ "model.layers.15.ffn.router.router_weights": "model-00002-of-00006.safetensors",
240
+ "model.layers.15.ffn.shared_experts.0.down.weight": "model-00002-of-00006.safetensors",
241
+ "model.layers.15.ffn.shared_experts.0.gate.weight": "model-00002-of-00006.safetensors",
242
+ "model.layers.15.ffn.shared_experts.0.up.weight": "model-00002-of-00006.safetensors",
243
+ "model.layers.15.ffn.w_down_proj.weight": "model-00002-of-00006.safetensors",
244
+ "model.layers.15.ffn.w_up_proj.weight": "model-00002-of-00006.safetensors",
245
+ "model.layers.15.injection_gate": "model-00002-of-00006.safetensors",
246
+ "model.layers.15.ln1.weight": "model-00002-of-00006.safetensors",
247
+ "model.layers.15.ln1_out.weight": "model-00002-of-00006.safetensors",
248
+ "model.layers.15.ln2.weight": "model-00002-of-00006.safetensors",
249
+ "model.layers.15.ln2_out.weight": "model-00002-of-00006.safetensors",
250
+ "model.layers.16.C_to_hidden.weight": "model-00002-of-00006.safetensors",
251
+ "model.layers.16.W_alpha.bias": "model-00002-of-00006.safetensors",
252
+ "model.layers.16.W_alpha.weight": "model-00002-of-00006.safetensors",
253
+ "model.layers.16.attn.g_norm_swish_gate.weight": "model-00002-of-00006.safetensors",
254
+ "model.layers.16.attn.g_proj.weight": "model-00002-of-00006.safetensors",
255
+ "model.layers.16.attn.gk_proj.0.weight": "model-00002-of-00006.safetensors",
256
+ "model.layers.16.attn.gk_proj.1.bias": "model-00002-of-00006.safetensors",
257
+ "model.layers.16.attn.gk_proj.1.weight": "model-00002-of-00006.safetensors",
258
+ "model.layers.16.attn.k_proj.weight": "model-00002-of-00006.safetensors",
259
+ "model.layers.16.attn.o_proj.weight": "model-00002-of-00006.safetensors",
260
+ "model.layers.16.attn.q_proj.weight": "model-00002-of-00006.safetensors",
261
+ "model.layers.16.attn.v_proj.weight": "model-00002-of-00006.safetensors",
262
+ "model.layers.16.ffn.experts_w12": "model-00002-of-00006.safetensors",
263
+ "model.layers.16.ffn.experts_w3": "model-00002-of-00006.safetensors",
264
+ "model.layers.16.ffn.max_vio": "model-00002-of-00006.safetensors",
265
+ "model.layers.16.ffn.router.router_weights": "model-00002-of-00006.safetensors",
266
+ "model.layers.16.ffn.shared_experts.0.down.weight": "model-00002-of-00006.safetensors",
267
+ "model.layers.16.ffn.shared_experts.0.gate.weight": "model-00002-of-00006.safetensors",
268
+ "model.layers.16.ffn.shared_experts.0.up.weight": "model-00002-of-00006.safetensors",
269
+ "model.layers.16.ffn.w_down_proj.weight": "model-00002-of-00006.safetensors",
270
+ "model.layers.16.ffn.w_up_proj.weight": "model-00002-of-00006.safetensors",
271
+ "model.layers.16.injection_gate": "model-00002-of-00006.safetensors",
272
+ "model.layers.16.ln1.weight": "model-00002-of-00006.safetensors",
273
+ "model.layers.16.ln1_out.weight": "model-00002-of-00006.safetensors",
274
+ "model.layers.16.ln2.weight": "model-00002-of-00006.safetensors",
275
+ "model.layers.16.ln2_out.weight": "model-00002-of-00006.safetensors",
276
+ "model.layers.16.memory.W_eta.bias": "model-00002-of-00006.safetensors",
277
+ "model.layers.16.memory.W_eta.weight": "model-00002-of-00006.safetensors",
278
+ "model.layers.16.memory.W_qkv_mem.weight": "model-00002-of-00006.safetensors",
279
+ "model.layers.16.memory.compress_z.0.weight": "model-00002-of-00006.safetensors",
280
+ "model.layers.16.memory.compress_z.2.weight": "model-00002-of-00006.safetensors",
281
+ "model.layers.16.memory.eta_channels": "model-00002-of-00006.safetensors",
282
+ "model.layers.16.memory.input_norm.bias": "model-00002-of-00006.safetensors",
283
+ "model.layers.16.memory.input_norm.weight": "model-00002-of-00006.safetensors",
284
+ "model.layers.16.memory.summary_proj.bias": "model-00002-of-00006.safetensors",
285
+ "model.layers.16.memory.summary_proj.weight": "model-00002-of-00006.safetensors",
286
+ "model.layers.16.memory.summary_query": "model-00002-of-00006.safetensors",
287
+ "model.layers.16.memory.temperature": "model-00002-of-00006.safetensors",
288
+ "model.layers.17.C_to_hidden.weight": "model-00002-of-00006.safetensors",
289
+ "model.layers.17.W_alpha.bias": "model-00002-of-00006.safetensors",
290
+ "model.layers.17.W_alpha.weight": "model-00002-of-00006.safetensors",
291
+ "model.layers.17.attn.g_norm_swish_gate.weight": "model-00002-of-00006.safetensors",
292
+ "model.layers.17.attn.g_proj.weight": "model-00002-of-00006.safetensors",
293
+ "model.layers.17.attn.gk_proj.0.weight": "model-00002-of-00006.safetensors",
294
+ "model.layers.17.attn.gk_proj.1.bias": "model-00002-of-00006.safetensors",
295
+ "model.layers.17.attn.gk_proj.1.weight": "model-00002-of-00006.safetensors",
296
+ "model.layers.17.attn.k_proj.weight": "model-00002-of-00006.safetensors",
297
+ "model.layers.17.attn.o_proj.weight": "model-00002-of-00006.safetensors",
298
+ "model.layers.17.attn.q_proj.weight": "model-00002-of-00006.safetensors",
299
+ "model.layers.17.attn.v_proj.weight": "model-00002-of-00006.safetensors",
300
+ "model.layers.17.ffn.experts_w12": "model-00002-of-00006.safetensors",
301
+ "model.layers.17.ffn.experts_w3": "model-00002-of-00006.safetensors",
302
+ "model.layers.17.ffn.max_vio": "model-00002-of-00006.safetensors",
303
+ "model.layers.17.ffn.router.router_weights": "model-00002-of-00006.safetensors",
304
+ "model.layers.17.ffn.shared_experts.0.down.weight": "model-00002-of-00006.safetensors",
305
+ "model.layers.17.ffn.shared_experts.0.gate.weight": "model-00002-of-00006.safetensors",
306
+ "model.layers.17.ffn.shared_experts.0.up.weight": "model-00002-of-00006.safetensors",
307
+ "model.layers.17.ffn.w_down_proj.weight": "model-00002-of-00006.safetensors",
308
+ "model.layers.17.ffn.w_up_proj.weight": "model-00002-of-00006.safetensors",
309
+ "model.layers.17.injection_gate": "model-00002-of-00006.safetensors",
310
+ "model.layers.17.ln1.weight": "model-00002-of-00006.safetensors",
311
+ "model.layers.17.ln1_out.weight": "model-00002-of-00006.safetensors",
312
+ "model.layers.17.ln2.weight": "model-00002-of-00006.safetensors",
313
+ "model.layers.17.ln2_out.weight": "model-00002-of-00006.safetensors",
314
+ "model.layers.17.memory.W_eta.bias": "model-00002-of-00006.safetensors",
315
+ "model.layers.17.memory.W_eta.weight": "model-00002-of-00006.safetensors",
316
+ "model.layers.17.memory.W_qkv_mem.weight": "model-00002-of-00006.safetensors",
317
+ "model.layers.17.memory.compress_z.0.weight": "model-00002-of-00006.safetensors",
318
+ "model.layers.17.memory.compress_z.2.weight": "model-00002-of-00006.safetensors",
319
+ "model.layers.17.memory.eta_channels": "model-00002-of-00006.safetensors",
320
+ "model.layers.17.memory.input_norm.bias": "model-00002-of-00006.safetensors",
321
+ "model.layers.17.memory.input_norm.weight": "model-00002-of-00006.safetensors",
322
+ "model.layers.17.memory.summary_proj.bias": "model-00002-of-00006.safetensors",
323
+ "model.layers.17.memory.summary_proj.weight": "model-00002-of-00006.safetensors",
324
+ "model.layers.17.memory.summary_query": "model-00002-of-00006.safetensors",
325
+ "model.layers.17.memory.temperature": "model-00002-of-00006.safetensors",
326
+ "model.layers.18.attn.A_log": "model-00002-of-00006.safetensors",
327
+ "model.layers.18.attn.b_proj.weight": "model-00002-of-00006.safetensors",
328
+ "model.layers.18.attn.dt_bias": "model-00002-of-00006.safetensors",
329
+ "model.layers.18.attn.f_proj.weight": "model-00002-of-00006.safetensors",
330
+ "model.layers.18.attn.g_proj.0.weight": "model-00002-of-00006.safetensors",
331
+ "model.layers.18.attn.g_proj.1.bias": "model-00002-of-00006.safetensors",
332
+ "model.layers.18.attn.g_proj.1.weight": "model-00002-of-00006.safetensors",
333
+ "model.layers.18.attn.k_conv1d.weight": "model-00002-of-00006.safetensors",
334
+ "model.layers.18.attn.k_proj.weight": "model-00002-of-00006.safetensors",
335
+ "model.layers.18.attn.o_norm.weight": "model-00002-of-00006.safetensors",
336
+ "model.layers.18.attn.o_proj.weight": "model-00002-of-00006.safetensors",
337
+ "model.layers.18.attn.q_conv1d.weight": "model-00002-of-00006.safetensors",
338
+ "model.layers.18.attn.q_proj.weight": "model-00002-of-00006.safetensors",
339
+ "model.layers.18.attn.v_conv1d.weight": "model-00002-of-00006.safetensors",
340
+ "model.layers.18.attn.v_proj.weight": "model-00002-of-00006.safetensors",
341
+ "model.layers.18.ffn.experts_w12": "model-00002-of-00006.safetensors",
342
+ "model.layers.18.ffn.experts_w3": "model-00002-of-00006.safetensors",
343
+ "model.layers.18.ffn.max_vio": "model-00002-of-00006.safetensors",
344
+ "model.layers.18.ffn.router.router_weights": "model-00002-of-00006.safetensors",
345
+ "model.layers.18.ffn.shared_experts.0.down.weight": "model-00002-of-00006.safetensors",
346
+ "model.layers.18.ffn.shared_experts.0.gate.weight": "model-00002-of-00006.safetensors",
347
+ "model.layers.18.ffn.shared_experts.0.up.weight": "model-00002-of-00006.safetensors",
348
+ "model.layers.18.ffn.w_down_proj.weight": "model-00002-of-00006.safetensors",
349
+ "model.layers.18.ffn.w_up_proj.weight": "model-00002-of-00006.safetensors",
350
+ "model.layers.18.injection_gate": "model-00002-of-00006.safetensors",
351
+ "model.layers.18.ln1.weight": "model-00002-of-00006.safetensors",
352
+ "model.layers.18.ln1_out.weight": "model-00002-of-00006.safetensors",
353
+ "model.layers.18.ln2.weight": "model-00002-of-00006.safetensors",
354
+ "model.layers.18.ln2_out.weight": "model-00002-of-00006.safetensors",
355
+ "model.layers.19.attn.A_log": "model-00002-of-00006.safetensors",
356
+ "model.layers.19.attn.b_proj.weight": "model-00002-of-00006.safetensors",
357
+ "model.layers.19.attn.dt_bias": "model-00002-of-00006.safetensors",
358
+ "model.layers.19.attn.f_proj.weight": "model-00002-of-00006.safetensors",
359
+ "model.layers.19.attn.g_proj.0.weight": "model-00002-of-00006.safetensors",
360
+ "model.layers.19.attn.g_proj.1.bias": "model-00002-of-00006.safetensors",
361
+ "model.layers.19.attn.g_proj.1.weight": "model-00002-of-00006.safetensors",
362
+ "model.layers.19.attn.k_conv1d.weight": "model-00002-of-00006.safetensors",
363
+ "model.layers.19.attn.k_proj.weight": "model-00002-of-00006.safetensors",
364
+ "model.layers.19.attn.o_norm.weight": "model-00002-of-00006.safetensors",
365
+ "model.layers.19.attn.o_proj.weight": "model-00002-of-00006.safetensors",
366
+ "model.layers.19.attn.q_conv1d.weight": "model-00002-of-00006.safetensors",
367
+ "model.layers.19.attn.q_proj.weight": "model-00003-of-00006.safetensors",
368
+ "model.layers.19.attn.v_conv1d.weight": "model-00003-of-00006.safetensors",
369
+ "model.layers.19.attn.v_proj.weight": "model-00003-of-00006.safetensors",
370
+ "model.layers.19.ffn.experts_w12": "model-00003-of-00006.safetensors",
371
+ "model.layers.19.ffn.experts_w3": "model-00003-of-00006.safetensors",
372
+ "model.layers.19.ffn.max_vio": "model-00003-of-00006.safetensors",
373
+ "model.layers.19.ffn.router.router_weights": "model-00003-of-00006.safetensors",
374
+ "model.layers.19.ffn.shared_experts.0.down.weight": "model-00003-of-00006.safetensors",
375
+ "model.layers.19.ffn.shared_experts.0.gate.weight": "model-00003-of-00006.safetensors",
376
+ "model.layers.19.ffn.shared_experts.0.up.weight": "model-00003-of-00006.safetensors",
377
+ "model.layers.19.ffn.w_down_proj.weight": "model-00003-of-00006.safetensors",
378
+ "model.layers.19.ffn.w_up_proj.weight": "model-00003-of-00006.safetensors",
379
+ "model.layers.19.injection_gate": "model-00003-of-00006.safetensors",
380
+ "model.layers.19.ln1.weight": "model-00003-of-00006.safetensors",
381
+ "model.layers.19.ln1_out.weight": "model-00003-of-00006.safetensors",
382
+ "model.layers.19.ln2.weight": "model-00003-of-00006.safetensors",
383
+ "model.layers.19.ln2_out.weight": "model-00003-of-00006.safetensors",
384
+ "model.layers.2.attn.A_log": "model-00003-of-00006.safetensors",
385
+ "model.layers.2.attn.b_proj.weight": "model-00003-of-00006.safetensors",
386
+ "model.layers.2.attn.dt_bias": "model-00003-of-00006.safetensors",
387
+ "model.layers.2.attn.f_proj.weight": "model-00003-of-00006.safetensors",
388
+ "model.layers.2.attn.g_proj.0.weight": "model-00003-of-00006.safetensors",
389
+ "model.layers.2.attn.g_proj.1.bias": "model-00003-of-00006.safetensors",
390
+ "model.layers.2.attn.g_proj.1.weight": "model-00003-of-00006.safetensors",
391
+ "model.layers.2.attn.k_conv1d.weight": "model-00003-of-00006.safetensors",
392
+ "model.layers.2.attn.k_proj.weight": "model-00003-of-00006.safetensors",
393
+ "model.layers.2.attn.o_norm.weight": "model-00003-of-00006.safetensors",
394
+ "model.layers.2.attn.o_proj.weight": "model-00003-of-00006.safetensors",
395
+ "model.layers.2.attn.q_conv1d.weight": "model-00003-of-00006.safetensors",
396
+ "model.layers.2.attn.q_proj.weight": "model-00003-of-00006.safetensors",
397
+ "model.layers.2.attn.v_conv1d.weight": "model-00003-of-00006.safetensors",
398
+ "model.layers.2.attn.v_proj.weight": "model-00003-of-00006.safetensors",
399
+ "model.layers.2.ffn.down.weight": "model-00003-of-00006.safetensors",
400
+ "model.layers.2.ffn.gate.weight": "model-00003-of-00006.safetensors",
401
+ "model.layers.2.ffn.up.weight": "model-00003-of-00006.safetensors",
402
+ "model.layers.2.injection_gate": "model-00003-of-00006.safetensors",
403
+ "model.layers.2.ln1.weight": "model-00003-of-00006.safetensors",
404
+ "model.layers.2.ln1_out.weight": "model-00003-of-00006.safetensors",
405
+ "model.layers.2.ln2.weight": "model-00003-of-00006.safetensors",
406
+ "model.layers.2.ln2_out.weight": "model-00003-of-00006.safetensors",
407
+ "model.layers.20.attn.A_log": "model-00003-of-00006.safetensors",
408
+ "model.layers.20.attn.b_proj.weight": "model-00003-of-00006.safetensors",
409
+ "model.layers.20.attn.dt_bias": "model-00003-of-00006.safetensors",
410
+ "model.layers.20.attn.f_proj.weight": "model-00003-of-00006.safetensors",
411
+ "model.layers.20.attn.g_proj.0.weight": "model-00003-of-00006.safetensors",
412
+ "model.layers.20.attn.g_proj.1.bias": "model-00003-of-00006.safetensors",
413
+ "model.layers.20.attn.g_proj.1.weight": "model-00003-of-00006.safetensors",
414
+ "model.layers.20.attn.k_conv1d.weight": "model-00003-of-00006.safetensors",
415
+ "model.layers.20.attn.k_proj.weight": "model-00003-of-00006.safetensors",
416
+ "model.layers.20.attn.o_norm.weight": "model-00003-of-00006.safetensors",
417
+ "model.layers.20.attn.o_proj.weight": "model-00003-of-00006.safetensors",
418
+ "model.layers.20.attn.q_conv1d.weight": "model-00003-of-00006.safetensors",
419
+ "model.layers.20.attn.q_proj.weight": "model-00003-of-00006.safetensors",
420
+ "model.layers.20.attn.v_conv1d.weight": "model-00003-of-00006.safetensors",
421
+ "model.layers.20.attn.v_proj.weight": "model-00003-of-00006.safetensors",
422
+ "model.layers.20.ffn.experts_w12": "model-00003-of-00006.safetensors",
423
+ "model.layers.20.ffn.experts_w3": "model-00003-of-00006.safetensors",
424
+ "model.layers.20.ffn.max_vio": "model-00003-of-00006.safetensors",
425
+ "model.layers.20.ffn.router.router_weights": "model-00003-of-00006.safetensors",
426
+ "model.layers.20.ffn.shared_experts.0.down.weight": "model-00003-of-00006.safetensors",
427
+ "model.layers.20.ffn.shared_experts.0.gate.weight": "model-00003-of-00006.safetensors",
428
+ "model.layers.20.ffn.shared_experts.0.up.weight": "model-00003-of-00006.safetensors",
429
+ "model.layers.20.ffn.w_down_proj.weight": "model-00003-of-00006.safetensors",
430
+ "model.layers.20.ffn.w_up_proj.weight": "model-00003-of-00006.safetensors",
431
+ "model.layers.20.injection_gate": "model-00003-of-00006.safetensors",
432
+ "model.layers.20.ln1.weight": "model-00003-of-00006.safetensors",
433
+ "model.layers.20.ln1_out.weight": "model-00003-of-00006.safetensors",
434
+ "model.layers.20.ln2.weight": "model-00003-of-00006.safetensors",
435
+ "model.layers.20.ln2_out.weight": "model-00003-of-00006.safetensors",
436
+ "model.layers.21.attn.A_log": "model-00003-of-00006.safetensors",
437
+ "model.layers.21.attn.b_proj.weight": "model-00003-of-00006.safetensors",
438
+ "model.layers.21.attn.dt_bias": "model-00003-of-00006.safetensors",
439
+ "model.layers.21.attn.f_proj.weight": "model-00003-of-00006.safetensors",
440
+ "model.layers.21.attn.g_proj.0.weight": "model-00003-of-00006.safetensors",
441
+ "model.layers.21.attn.g_proj.1.bias": "model-00003-of-00006.safetensors",
442
+ "model.layers.21.attn.g_proj.1.weight": "model-00003-of-00006.safetensors",
443
+ "model.layers.21.attn.k_conv1d.weight": "model-00003-of-00006.safetensors",
444
+ "model.layers.21.attn.k_proj.weight": "model-00003-of-00006.safetensors",
445
+ "model.layers.21.attn.o_norm.weight": "model-00003-of-00006.safetensors",
446
+ "model.layers.21.attn.o_proj.weight": "model-00003-of-00006.safetensors",
447
+ "model.layers.21.attn.q_conv1d.weight": "model-00003-of-00006.safetensors",
448
+ "model.layers.21.attn.q_proj.weight": "model-00003-of-00006.safetensors",
449
+ "model.layers.21.attn.v_conv1d.weight": "model-00003-of-00006.safetensors",
450
+ "model.layers.21.attn.v_proj.weight": "model-00003-of-00006.safetensors",
451
+ "model.layers.21.ffn.experts_w12": "model-00003-of-00006.safetensors",
452
+ "model.layers.21.ffn.experts_w3": "model-00003-of-00006.safetensors",
453
+ "model.layers.21.ffn.max_vio": "model-00003-of-00006.safetensors",
454
+ "model.layers.21.ffn.router.router_weights": "model-00003-of-00006.safetensors",
455
+ "model.layers.21.ffn.shared_experts.0.down.weight": "model-00003-of-00006.safetensors",
456
+ "model.layers.21.ffn.shared_experts.0.gate.weight": "model-00003-of-00006.safetensors",
457
+ "model.layers.21.ffn.shared_experts.0.up.weight": "model-00003-of-00006.safetensors",
458
+ "model.layers.21.ffn.w_down_proj.weight": "model-00003-of-00006.safetensors",
459
+ "model.layers.21.ffn.w_up_proj.weight": "model-00003-of-00006.safetensors",
460
+ "model.layers.21.injection_gate": "model-00003-of-00006.safetensors",
461
+ "model.layers.21.ln1.weight": "model-00003-of-00006.safetensors",
462
+ "model.layers.21.ln1_out.weight": "model-00003-of-00006.safetensors",
463
+ "model.layers.21.ln2.weight": "model-00003-of-00006.safetensors",
464
+ "model.layers.21.ln2_out.weight": "model-00003-of-00006.safetensors",
465
+ "model.layers.22.C_to_hidden.weight": "model-00003-of-00006.safetensors",
466
+ "model.layers.22.W_alpha.bias": "model-00003-of-00006.safetensors",
467
+ "model.layers.22.W_alpha.weight": "model-00003-of-00006.safetensors",
468
+ "model.layers.22.attn.g_norm_swish_gate.weight": "model-00003-of-00006.safetensors",
469
+ "model.layers.22.attn.g_proj.weight": "model-00003-of-00006.safetensors",
470
+ "model.layers.22.attn.gk_proj.0.weight": "model-00003-of-00006.safetensors",
471
+ "model.layers.22.attn.gk_proj.1.bias": "model-00003-of-00006.safetensors",
472
+ "model.layers.22.attn.gk_proj.1.weight": "model-00003-of-00006.safetensors",
473
+ "model.layers.22.attn.k_proj.weight": "model-00003-of-00006.safetensors",
474
+ "model.layers.22.attn.o_proj.weight": "model-00003-of-00006.safetensors",
475
+ "model.layers.22.attn.q_proj.weight": "model-00003-of-00006.safetensors",
476
+ "model.layers.22.attn.v_proj.weight": "model-00003-of-00006.safetensors",
477
+ "model.layers.22.ffn.experts_w12": "model-00003-of-00006.safetensors",
478
+ "model.layers.22.ffn.experts_w3": "model-00003-of-00006.safetensors",
479
+ "model.layers.22.ffn.max_vio": "model-00003-of-00006.safetensors",
480
+ "model.layers.22.ffn.router.router_weights": "model-00003-of-00006.safetensors",
481
+ "model.layers.22.ffn.shared_experts.0.down.weight": "model-00003-of-00006.safetensors",
482
+ "model.layers.22.ffn.shared_experts.0.gate.weight": "model-00003-of-00006.safetensors",
483
+ "model.layers.22.ffn.shared_experts.0.up.weight": "model-00003-of-00006.safetensors",
484
+ "model.layers.22.ffn.w_down_proj.weight": "model-00003-of-00006.safetensors",
485
+ "model.layers.22.ffn.w_up_proj.weight": "model-00003-of-00006.safetensors",
486
+ "model.layers.22.injection_gate": "model-00003-of-00006.safetensors",
487
+ "model.layers.22.ln1.weight": "model-00003-of-00006.safetensors",
488
+ "model.layers.22.ln1_out.weight": "model-00003-of-00006.safetensors",
489
+ "model.layers.22.ln2.weight": "model-00003-of-00006.safetensors",
490
+ "model.layers.22.ln2_out.weight": "model-00003-of-00006.safetensors",
491
+ "model.layers.22.memory.W_eta.bias": "model-00003-of-00006.safetensors",
492
+ "model.layers.22.memory.W_eta.weight": "model-00003-of-00006.safetensors",
493
+ "model.layers.22.memory.W_qkv_mem.weight": "model-00003-of-00006.safetensors",
494
+ "model.layers.22.memory.compress_z.0.weight": "model-00003-of-00006.safetensors",
495
+ "model.layers.22.memory.compress_z.2.weight": "model-00003-of-00006.safetensors",
496
+ "model.layers.22.memory.eta_channels": "model-00003-of-00006.safetensors",
497
+ "model.layers.22.memory.input_norm.bias": "model-00003-of-00006.safetensors",
498
+ "model.layers.22.memory.input_norm.weight": "model-00003-of-00006.safetensors",
499
+ "model.layers.22.memory.summary_proj.bias": "model-00003-of-00006.safetensors",
500
+ "model.layers.22.memory.summary_proj.weight": "model-00003-of-00006.safetensors",
501
+ "model.layers.22.memory.summary_query": "model-00003-of-00006.safetensors",
502
+ "model.layers.22.memory.temperature": "model-00003-of-00006.safetensors",
503
+ "model.layers.23.C_to_hidden.weight": "model-00003-of-00006.safetensors",
504
+ "model.layers.23.W_alpha.bias": "model-00003-of-00006.safetensors",
505
+ "model.layers.23.W_alpha.weight": "model-00003-of-00006.safetensors",
506
+ "model.layers.23.attn.g_norm_swish_gate.weight": "model-00003-of-00006.safetensors",
507
+ "model.layers.23.attn.g_proj.weight": "model-00003-of-00006.safetensors",
508
+ "model.layers.23.attn.gk_proj.0.weight": "model-00003-of-00006.safetensors",
509
+ "model.layers.23.attn.gk_proj.1.bias": "model-00003-of-00006.safetensors",
510
+ "model.layers.23.attn.gk_proj.1.weight": "model-00003-of-00006.safetensors",
511
+ "model.layers.23.attn.k_proj.weight": "model-00003-of-00006.safetensors",
512
+ "model.layers.23.attn.o_proj.weight": "model-00003-of-00006.safetensors",
513
+ "model.layers.23.attn.q_proj.weight": "model-00003-of-00006.safetensors",
514
+ "model.layers.23.attn.v_proj.weight": "model-00003-of-00006.safetensors",
515
+ "model.layers.23.ffn.experts_w12": "model-00003-of-00006.safetensors",
516
+ "model.layers.23.ffn.experts_w3": "model-00003-of-00006.safetensors",
517
+ "model.layers.23.ffn.max_vio": "model-00003-of-00006.safetensors",
518
+ "model.layers.23.ffn.router.router_weights": "model-00003-of-00006.safetensors",
519
+ "model.layers.23.ffn.shared_experts.0.down.weight": "model-00003-of-00006.safetensors",
520
+ "model.layers.23.ffn.shared_experts.0.gate.weight": "model-00003-of-00006.safetensors",
521
+ "model.layers.23.ffn.shared_experts.0.up.weight": "model-00003-of-00006.safetensors",
522
+ "model.layers.23.ffn.w_down_proj.weight": "model-00003-of-00006.safetensors",
523
+ "model.layers.23.ffn.w_up_proj.weight": "model-00003-of-00006.safetensors",
524
+ "model.layers.23.injection_gate": "model-00003-of-00006.safetensors",
525
+ "model.layers.23.ln1.weight": "model-00003-of-00006.safetensors",
526
+ "model.layers.23.ln1_out.weight": "model-00003-of-00006.safetensors",
527
+ "model.layers.23.ln2.weight": "model-00003-of-00006.safetensors",
528
+ "model.layers.23.ln2_out.weight": "model-00003-of-00006.safetensors",
529
+ "model.layers.23.memory.W_eta.bias": "model-00003-of-00006.safetensors",
530
+ "model.layers.23.memory.W_eta.weight": "model-00003-of-00006.safetensors",
531
+ "model.layers.23.memory.W_qkv_mem.weight": "model-00003-of-00006.safetensors",
532
+ "model.layers.23.memory.compress_z.0.weight": "model-00003-of-00006.safetensors",
533
+ "model.layers.23.memory.compress_z.2.weight": "model-00003-of-00006.safetensors",
534
+ "model.layers.23.memory.eta_channels": "model-00003-of-00006.safetensors",
535
+ "model.layers.23.memory.input_norm.bias": "model-00003-of-00006.safetensors",
536
+ "model.layers.23.memory.input_norm.weight": "model-00003-of-00006.safetensors",
537
+ "model.layers.23.memory.summary_proj.bias": "model-00003-of-00006.safetensors",
538
+ "model.layers.23.memory.summary_proj.weight": "model-00003-of-00006.safetensors",
539
+ "model.layers.23.memory.summary_query": "model-00003-of-00006.safetensors",
540
+ "model.layers.23.memory.temperature": "model-00003-of-00006.safetensors",
541
+ "model.layers.24.attn.A_log": "model-00003-of-00006.safetensors",
542
+ "model.layers.24.attn.b_proj.weight": "model-00003-of-00006.safetensors",
543
+ "model.layers.24.attn.dt_bias": "model-00003-of-00006.safetensors",
544
+ "model.layers.24.attn.f_proj.weight": "model-00003-of-00006.safetensors",
545
+ "model.layers.24.attn.g_proj.0.weight": "model-00003-of-00006.safetensors",
546
+ "model.layers.24.attn.g_proj.1.bias": "model-00003-of-00006.safetensors",
547
+ "model.layers.24.attn.g_proj.1.weight": "model-00003-of-00006.safetensors",
548
+ "model.layers.24.attn.k_conv1d.weight": "model-00003-of-00006.safetensors",
549
+ "model.layers.24.attn.k_proj.weight": "model-00003-of-00006.safetensors",
550
+ "model.layers.24.attn.o_norm.weight": "model-00003-of-00006.safetensors",
551
+ "model.layers.24.attn.o_proj.weight": "model-00003-of-00006.safetensors",
552
+ "model.layers.24.attn.q_conv1d.weight": "model-00003-of-00006.safetensors",
553
+ "model.layers.24.attn.q_proj.weight": "model-00003-of-00006.safetensors",
554
+ "model.layers.24.attn.v_conv1d.weight": "model-00003-of-00006.safetensors",
555
+ "model.layers.24.attn.v_proj.weight": "model-00003-of-00006.safetensors",
556
+ "model.layers.24.ffn.experts_w12": "model-00003-of-00006.safetensors",
557
+ "model.layers.24.ffn.experts_w3": "model-00004-of-00006.safetensors",
558
+ "model.layers.24.ffn.max_vio": "model-00004-of-00006.safetensors",
559
+ "model.layers.24.ffn.router.router_weights": "model-00004-of-00006.safetensors",
560
+ "model.layers.24.ffn.shared_experts.0.down.weight": "model-00004-of-00006.safetensors",
561
+ "model.layers.24.ffn.shared_experts.0.gate.weight": "model-00004-of-00006.safetensors",
562
+ "model.layers.24.ffn.shared_experts.0.up.weight": "model-00004-of-00006.safetensors",
563
+ "model.layers.24.ffn.w_down_proj.weight": "model-00004-of-00006.safetensors",
564
+ "model.layers.24.ffn.w_up_proj.weight": "model-00004-of-00006.safetensors",
565
+ "model.layers.24.injection_gate": "model-00004-of-00006.safetensors",
566
+ "model.layers.24.ln1.weight": "model-00004-of-00006.safetensors",
567
+ "model.layers.24.ln1_out.weight": "model-00004-of-00006.safetensors",
568
+ "model.layers.24.ln2.weight": "model-00004-of-00006.safetensors",
569
+ "model.layers.24.ln2_out.weight": "model-00004-of-00006.safetensors",
570
+ "model.layers.25.attn.A_log": "model-00004-of-00006.safetensors",
571
+ "model.layers.25.attn.b_proj.weight": "model-00004-of-00006.safetensors",
572
+ "model.layers.25.attn.dt_bias": "model-00004-of-00006.safetensors",
573
+ "model.layers.25.attn.f_proj.weight": "model-00004-of-00006.safetensors",
574
+ "model.layers.25.attn.g_proj.0.weight": "model-00004-of-00006.safetensors",
575
+ "model.layers.25.attn.g_proj.1.bias": "model-00004-of-00006.safetensors",
576
+ "model.layers.25.attn.g_proj.1.weight": "model-00004-of-00006.safetensors",
577
+ "model.layers.25.attn.k_conv1d.weight": "model-00004-of-00006.safetensors",
578
+ "model.layers.25.attn.k_proj.weight": "model-00004-of-00006.safetensors",
579
+ "model.layers.25.attn.o_norm.weight": "model-00004-of-00006.safetensors",
580
+ "model.layers.25.attn.o_proj.weight": "model-00004-of-00006.safetensors",
581
+ "model.layers.25.attn.q_conv1d.weight": "model-00004-of-00006.safetensors",
582
+ "model.layers.25.attn.q_proj.weight": "model-00004-of-00006.safetensors",
583
+ "model.layers.25.attn.v_conv1d.weight": "model-00004-of-00006.safetensors",
584
+ "model.layers.25.attn.v_proj.weight": "model-00004-of-00006.safetensors",
585
+ "model.layers.25.ffn.experts_w12": "model-00004-of-00006.safetensors",
586
+ "model.layers.25.ffn.experts_w3": "model-00004-of-00006.safetensors",
587
+ "model.layers.25.ffn.max_vio": "model-00004-of-00006.safetensors",
588
+ "model.layers.25.ffn.router.router_weights": "model-00004-of-00006.safetensors",
589
+ "model.layers.25.ffn.shared_experts.0.down.weight": "model-00004-of-00006.safetensors",
590
+ "model.layers.25.ffn.shared_experts.0.gate.weight": "model-00004-of-00006.safetensors",
591
+ "model.layers.25.ffn.shared_experts.0.up.weight": "model-00004-of-00006.safetensors",
592
+ "model.layers.25.ffn.w_down_proj.weight": "model-00004-of-00006.safetensors",
593
+ "model.layers.25.ffn.w_up_proj.weight": "model-00004-of-00006.safetensors",
594
+ "model.layers.25.injection_gate": "model-00004-of-00006.safetensors",
595
+ "model.layers.25.ln1.weight": "model-00004-of-00006.safetensors",
596
+ "model.layers.25.ln1_out.weight": "model-00004-of-00006.safetensors",
597
+ "model.layers.25.ln2.weight": "model-00004-of-00006.safetensors",
598
+ "model.layers.25.ln2_out.weight": "model-00004-of-00006.safetensors",
599
+ "model.layers.26.attn.A_log": "model-00004-of-00006.safetensors",
600
+ "model.layers.26.attn.b_proj.weight": "model-00004-of-00006.safetensors",
601
+ "model.layers.26.attn.dt_bias": "model-00004-of-00006.safetensors",
602
+ "model.layers.26.attn.f_proj.weight": "model-00004-of-00006.safetensors",
603
+ "model.layers.26.attn.g_proj.0.weight": "model-00004-of-00006.safetensors",
604
+ "model.layers.26.attn.g_proj.1.bias": "model-00004-of-00006.safetensors",
605
+ "model.layers.26.attn.g_proj.1.weight": "model-00004-of-00006.safetensors",
606
+ "model.layers.26.attn.k_conv1d.weight": "model-00004-of-00006.safetensors",
607
+ "model.layers.26.attn.k_proj.weight": "model-00004-of-00006.safetensors",
608
+ "model.layers.26.attn.o_norm.weight": "model-00004-of-00006.safetensors",
609
+ "model.layers.26.attn.o_proj.weight": "model-00004-of-00006.safetensors",
610
+ "model.layers.26.attn.q_conv1d.weight": "model-00004-of-00006.safetensors",
611
+ "model.layers.26.attn.q_proj.weight": "model-00004-of-00006.safetensors",
612
+ "model.layers.26.attn.v_conv1d.weight": "model-00004-of-00006.safetensors",
613
+ "model.layers.26.attn.v_proj.weight": "model-00004-of-00006.safetensors",
614
+ "model.layers.26.ffn.experts_w12": "model-00004-of-00006.safetensors",
615
+ "model.layers.26.ffn.experts_w3": "model-00004-of-00006.safetensors",
616
+ "model.layers.26.ffn.max_vio": "model-00004-of-00006.safetensors",
617
+ "model.layers.26.ffn.router.router_weights": "model-00004-of-00006.safetensors",
618
+ "model.layers.26.ffn.shared_experts.0.down.weight": "model-00004-of-00006.safetensors",
619
+ "model.layers.26.ffn.shared_experts.0.gate.weight": "model-00004-of-00006.safetensors",
620
+ "model.layers.26.ffn.shared_experts.0.up.weight": "model-00004-of-00006.safetensors",
621
+ "model.layers.26.ffn.w_down_proj.weight": "model-00004-of-00006.safetensors",
622
+ "model.layers.26.ffn.w_up_proj.weight": "model-00004-of-00006.safetensors",
623
+ "model.layers.26.injection_gate": "model-00004-of-00006.safetensors",
624
+ "model.layers.26.ln1.weight": "model-00004-of-00006.safetensors",
625
+ "model.layers.26.ln1_out.weight": "model-00004-of-00006.safetensors",
626
+ "model.layers.26.ln2.weight": "model-00004-of-00006.safetensors",
627
+ "model.layers.26.ln2_out.weight": "model-00004-of-00006.safetensors",
628
+ "model.layers.27.attn.A_log": "model-00004-of-00006.safetensors",
629
+ "model.layers.27.attn.b_proj.weight": "model-00004-of-00006.safetensors",
630
+ "model.layers.27.attn.dt_bias": "model-00004-of-00006.safetensors",
631
+ "model.layers.27.attn.f_proj.weight": "model-00004-of-00006.safetensors",
632
+ "model.layers.27.attn.g_proj.0.weight": "model-00004-of-00006.safetensors",
633
+ "model.layers.27.attn.g_proj.1.bias": "model-00004-of-00006.safetensors",
634
+ "model.layers.27.attn.g_proj.1.weight": "model-00004-of-00006.safetensors",
635
+ "model.layers.27.attn.k_conv1d.weight": "model-00004-of-00006.safetensors",
636
+ "model.layers.27.attn.k_proj.weight": "model-00004-of-00006.safetensors",
637
+ "model.layers.27.attn.o_norm.weight": "model-00004-of-00006.safetensors",
638
+ "model.layers.27.attn.o_proj.weight": "model-00004-of-00006.safetensors",
639
+ "model.layers.27.attn.q_conv1d.weight": "model-00004-of-00006.safetensors",
640
+ "model.layers.27.attn.q_proj.weight": "model-00004-of-00006.safetensors",
641
+ "model.layers.27.attn.v_conv1d.weight": "model-00004-of-00006.safetensors",
642
+ "model.layers.27.attn.v_proj.weight": "model-00004-of-00006.safetensors",
643
+ "model.layers.27.ffn.experts_w12": "model-00004-of-00006.safetensors",
644
+ "model.layers.27.ffn.experts_w3": "model-00004-of-00006.safetensors",
645
+ "model.layers.27.ffn.max_vio": "model-00004-of-00006.safetensors",
646
+ "model.layers.27.ffn.router.router_weights": "model-00004-of-00006.safetensors",
647
+ "model.layers.27.ffn.shared_experts.0.down.weight": "model-00004-of-00006.safetensors",
648
+ "model.layers.27.ffn.shared_experts.0.gate.weight": "model-00004-of-00006.safetensors",
649
+ "model.layers.27.ffn.shared_experts.0.up.weight": "model-00004-of-00006.safetensors",
650
+ "model.layers.27.ffn.w_down_proj.weight": "model-00004-of-00006.safetensors",
651
+ "model.layers.27.ffn.w_up_proj.weight": "model-00004-of-00006.safetensors",
652
+ "model.layers.27.injection_gate": "model-00004-of-00006.safetensors",
653
+ "model.layers.27.ln1.weight": "model-00004-of-00006.safetensors",
654
+ "model.layers.27.ln1_out.weight": "model-00004-of-00006.safetensors",
655
+ "model.layers.27.ln2.weight": "model-00004-of-00006.safetensors",
656
+ "model.layers.27.ln2_out.weight": "model-00004-of-00006.safetensors",
657
+ "model.layers.28.C_to_hidden.weight": "model-00004-of-00006.safetensors",
658
+ "model.layers.28.W_alpha.bias": "model-00004-of-00006.safetensors",
659
+ "model.layers.28.W_alpha.weight": "model-00004-of-00006.safetensors",
660
+ "model.layers.28.attn.g_norm_swish_gate.weight": "model-00004-of-00006.safetensors",
661
+ "model.layers.28.attn.g_proj.weight": "model-00004-of-00006.safetensors",
662
+ "model.layers.28.attn.gk_proj.0.weight": "model-00004-of-00006.safetensors",
663
+ "model.layers.28.attn.gk_proj.1.bias": "model-00004-of-00006.safetensors",
664
+ "model.layers.28.attn.gk_proj.1.weight": "model-00004-of-00006.safetensors",
665
+ "model.layers.28.attn.k_proj.weight": "model-00004-of-00006.safetensors",
666
+ "model.layers.28.attn.o_proj.weight": "model-00004-of-00006.safetensors",
667
+ "model.layers.28.attn.q_proj.weight": "model-00004-of-00006.safetensors",
668
+ "model.layers.28.attn.v_proj.weight": "model-00004-of-00006.safetensors",
669
+ "model.layers.28.ffn.experts_w12": "model-00004-of-00006.safetensors",
670
+ "model.layers.28.ffn.experts_w3": "model-00004-of-00006.safetensors",
671
+ "model.layers.28.ffn.max_vio": "model-00004-of-00006.safetensors",
672
+ "model.layers.28.ffn.router.router_weights": "model-00004-of-00006.safetensors",
673
+ "model.layers.28.ffn.shared_experts.0.down.weight": "model-00004-of-00006.safetensors",
674
+ "model.layers.28.ffn.shared_experts.0.gate.weight": "model-00004-of-00006.safetensors",
675
+ "model.layers.28.ffn.shared_experts.0.up.weight": "model-00004-of-00006.safetensors",
676
+ "model.layers.28.ffn.w_down_proj.weight": "model-00004-of-00006.safetensors",
677
+ "model.layers.28.ffn.w_up_proj.weight": "model-00004-of-00006.safetensors",
678
+ "model.layers.28.injection_gate": "model-00004-of-00006.safetensors",
679
+ "model.layers.28.ln1.weight": "model-00004-of-00006.safetensors",
680
+ "model.layers.28.ln1_out.weight": "model-00004-of-00006.safetensors",
681
+ "model.layers.28.ln2.weight": "model-00004-of-00006.safetensors",
682
+ "model.layers.28.ln2_out.weight": "model-00004-of-00006.safetensors",
683
+ "model.layers.28.memory.W_eta.bias": "model-00004-of-00006.safetensors",
684
+ "model.layers.28.memory.W_eta.weight": "model-00004-of-00006.safetensors",
685
+ "model.layers.28.memory.W_qkv_mem.weight": "model-00004-of-00006.safetensors",
686
+ "model.layers.28.memory.compress_z.0.weight": "model-00004-of-00006.safetensors",
687
+ "model.layers.28.memory.compress_z.2.weight": "model-00004-of-00006.safetensors",
688
+ "model.layers.28.memory.eta_channels": "model-00004-of-00006.safetensors",
689
+ "model.layers.28.memory.input_norm.bias": "model-00004-of-00006.safetensors",
690
+ "model.layers.28.memory.input_norm.weight": "model-00004-of-00006.safetensors",
691
+ "model.layers.28.memory.summary_proj.bias": "model-00004-of-00006.safetensors",
692
+ "model.layers.28.memory.summary_proj.weight": "model-00004-of-00006.safetensors",
693
+ "model.layers.28.memory.summary_query": "model-00004-of-00006.safetensors",
694
+ "model.layers.28.memory.temperature": "model-00004-of-00006.safetensors",
695
+ "model.layers.29.C_to_hidden.weight": "model-00004-of-00006.safetensors",
696
+ "model.layers.29.W_alpha.bias": "model-00004-of-00006.safetensors",
697
+ "model.layers.29.W_alpha.weight": "model-00004-of-00006.safetensors",
698
+ "model.layers.29.attn.g_norm_swish_gate.weight": "model-00004-of-00006.safetensors",
699
+ "model.layers.29.attn.g_proj.weight": "model-00004-of-00006.safetensors",
700
+ "model.layers.29.attn.gk_proj.0.weight": "model-00004-of-00006.safetensors",
701
+ "model.layers.29.attn.gk_proj.1.bias": "model-00004-of-00006.safetensors",
702
+ "model.layers.29.attn.gk_proj.1.weight": "model-00004-of-00006.safetensors",
703
+ "model.layers.29.attn.k_proj.weight": "model-00004-of-00006.safetensors",
704
+ "model.layers.29.attn.o_proj.weight": "model-00004-of-00006.safetensors",
705
+ "model.layers.29.attn.q_proj.weight": "model-00004-of-00006.safetensors",
706
+ "model.layers.29.attn.v_proj.weight": "model-00004-of-00006.safetensors",
707
+ "model.layers.29.ffn.experts_w12": "model-00004-of-00006.safetensors",
708
+ "model.layers.29.ffn.experts_w3": "model-00004-of-00006.safetensors",
709
+ "model.layers.29.ffn.max_vio": "model-00004-of-00006.safetensors",
710
+ "model.layers.29.ffn.router.router_weights": "model-00004-of-00006.safetensors",
711
+ "model.layers.29.ffn.shared_experts.0.down.weight": "model-00004-of-00006.safetensors",
712
+ "model.layers.29.ffn.shared_experts.0.gate.weight": "model-00004-of-00006.safetensors",
713
+ "model.layers.29.ffn.shared_experts.0.up.weight": "model-00004-of-00006.safetensors",
714
+ "model.layers.29.ffn.w_down_proj.weight": "model-00004-of-00006.safetensors",
715
+ "model.layers.29.ffn.w_up_proj.weight": "model-00004-of-00006.safetensors",
716
+ "model.layers.29.injection_gate": "model-00004-of-00006.safetensors",
717
+ "model.layers.29.ln1.weight": "model-00004-of-00006.safetensors",
718
+ "model.layers.29.ln1_out.weight": "model-00004-of-00006.safetensors",
719
+ "model.layers.29.ln2.weight": "model-00004-of-00006.safetensors",
720
+ "model.layers.29.ln2_out.weight": "model-00004-of-00006.safetensors",
721
+ "model.layers.29.memory.W_eta.bias": "model-00004-of-00006.safetensors",
722
+ "model.layers.29.memory.W_eta.weight": "model-00004-of-00006.safetensors",
723
+ "model.layers.29.memory.W_qkv_mem.weight": "model-00004-of-00006.safetensors",
724
+ "model.layers.29.memory.compress_z.0.weight": "model-00004-of-00006.safetensors",
725
+ "model.layers.29.memory.compress_z.2.weight": "model-00004-of-00006.safetensors",
726
+ "model.layers.29.memory.eta_channels": "model-00004-of-00006.safetensors",
727
+ "model.layers.29.memory.input_norm.bias": "model-00004-of-00006.safetensors",
728
+ "model.layers.29.memory.input_norm.weight": "model-00004-of-00006.safetensors",
729
+ "model.layers.29.memory.summary_proj.bias": "model-00004-of-00006.safetensors",
730
+ "model.layers.29.memory.summary_proj.weight": "model-00004-of-00006.safetensors",
731
+ "model.layers.29.memory.summary_query": "model-00004-of-00006.safetensors",
732
+ "model.layers.29.memory.temperature": "model-00004-of-00006.safetensors",
733
+ "model.layers.3.attn.A_log": "model-00004-of-00006.safetensors",
734
+ "model.layers.3.attn.b_proj.weight": "model-00004-of-00006.safetensors",
735
+ "model.layers.3.attn.dt_bias": "model-00004-of-00006.safetensors",
736
+ "model.layers.3.attn.f_proj.weight": "model-00004-of-00006.safetensors",
737
+ "model.layers.3.attn.g_proj.0.weight": "model-00004-of-00006.safetensors",
738
+ "model.layers.3.attn.g_proj.1.bias": "model-00004-of-00006.safetensors",
739
+ "model.layers.3.attn.g_proj.1.weight": "model-00004-of-00006.safetensors",
740
+ "model.layers.3.attn.k_conv1d.weight": "model-00004-of-00006.safetensors",
741
+ "model.layers.3.attn.k_proj.weight": "model-00004-of-00006.safetensors",
742
+ "model.layers.3.attn.o_norm.weight": "model-00004-of-00006.safetensors",
743
+ "model.layers.3.attn.o_proj.weight": "model-00004-of-00006.safetensors",
744
+ "model.layers.3.attn.q_conv1d.weight": "model-00004-of-00006.safetensors",
745
+ "model.layers.3.attn.q_proj.weight": "model-00004-of-00006.safetensors",
746
+ "model.layers.3.attn.v_conv1d.weight": "model-00004-of-00006.safetensors",
747
+ "model.layers.3.attn.v_proj.weight": "model-00004-of-00006.safetensors",
748
+ "model.layers.3.ffn.down.weight": "model-00004-of-00006.safetensors",
749
+ "model.layers.3.ffn.gate.weight": "model-00004-of-00006.safetensors",
750
+ "model.layers.3.ffn.up.weight": "model-00004-of-00006.safetensors",
751
+ "model.layers.3.injection_gate": "model-00004-of-00006.safetensors",
752
+ "model.layers.3.ln1.weight": "model-00004-of-00006.safetensors",
753
+ "model.layers.3.ln1_out.weight": "model-00004-of-00006.safetensors",
754
+ "model.layers.3.ln2.weight": "model-00004-of-00006.safetensors",
755
+ "model.layers.3.ln2_out.weight": "model-00004-of-00006.safetensors",
756
+ "model.layers.30.attn.A_log": "model-00004-of-00006.safetensors",
757
+ "model.layers.30.attn.b_proj.weight": "model-00004-of-00006.safetensors",
758
+ "model.layers.30.attn.dt_bias": "model-00004-of-00006.safetensors",
759
+ "model.layers.30.attn.f_proj.weight": "model-00004-of-00006.safetensors",
760
+ "model.layers.30.attn.g_proj.0.weight": "model-00004-of-00006.safetensors",
761
+ "model.layers.30.attn.g_proj.1.bias": "model-00004-of-00006.safetensors",
762
+ "model.layers.30.attn.g_proj.1.weight": "model-00004-of-00006.safetensors",
763
+ "model.layers.30.attn.k_conv1d.weight": "model-00004-of-00006.safetensors",
764
+ "model.layers.30.attn.k_proj.weight": "model-00004-of-00006.safetensors",
765
+ "model.layers.30.attn.o_norm.weight": "model-00004-of-00006.safetensors",
766
+ "model.layers.30.attn.o_proj.weight": "model-00004-of-00006.safetensors",
767
+ "model.layers.30.attn.q_conv1d.weight": "model-00004-of-00006.safetensors",
768
+ "model.layers.30.attn.q_proj.weight": "model-00004-of-00006.safetensors",
769
+ "model.layers.30.attn.v_conv1d.weight": "model-00004-of-00006.safetensors",
770
+ "model.layers.30.attn.v_proj.weight": "model-00004-of-00006.safetensors",
771
+ "model.layers.30.ffn.experts_w12": "model-00005-of-00006.safetensors",
772
+ "model.layers.30.ffn.experts_w3": "model-00005-of-00006.safetensors",
773
+ "model.layers.30.ffn.max_vio": "model-00005-of-00006.safetensors",
774
+ "model.layers.30.ffn.router.router_weights": "model-00005-of-00006.safetensors",
775
+ "model.layers.30.ffn.shared_experts.0.down.weight": "model-00005-of-00006.safetensors",
776
+ "model.layers.30.ffn.shared_experts.0.gate.weight": "model-00005-of-00006.safetensors",
777
+ "model.layers.30.ffn.shared_experts.0.up.weight": "model-00005-of-00006.safetensors",
778
+ "model.layers.30.ffn.w_down_proj.weight": "model-00005-of-00006.safetensors",
779
+ "model.layers.30.ffn.w_up_proj.weight": "model-00005-of-00006.safetensors",
780
+ "model.layers.30.injection_gate": "model-00005-of-00006.safetensors",
781
+ "model.layers.30.ln1.weight": "model-00005-of-00006.safetensors",
782
+ "model.layers.30.ln1_out.weight": "model-00005-of-00006.safetensors",
783
+ "model.layers.30.ln2.weight": "model-00005-of-00006.safetensors",
784
+ "model.layers.30.ln2_out.weight": "model-00005-of-00006.safetensors",
785
+ "model.layers.31.attn.A_log": "model-00005-of-00006.safetensors",
786
+ "model.layers.31.attn.b_proj.weight": "model-00005-of-00006.safetensors",
787
+ "model.layers.31.attn.dt_bias": "model-00005-of-00006.safetensors",
788
+ "model.layers.31.attn.f_proj.weight": "model-00005-of-00006.safetensors",
789
+ "model.layers.31.attn.g_proj.0.weight": "model-00005-of-00006.safetensors",
790
+ "model.layers.31.attn.g_proj.1.bias": "model-00005-of-00006.safetensors",
791
+ "model.layers.31.attn.g_proj.1.weight": "model-00005-of-00006.safetensors",
792
+ "model.layers.31.attn.k_conv1d.weight": "model-00005-of-00006.safetensors",
793
+ "model.layers.31.attn.k_proj.weight": "model-00005-of-00006.safetensors",
794
+ "model.layers.31.attn.o_norm.weight": "model-00005-of-00006.safetensors",
795
+ "model.layers.31.attn.o_proj.weight": "model-00005-of-00006.safetensors",
796
+ "model.layers.31.attn.q_conv1d.weight": "model-00005-of-00006.safetensors",
797
+ "model.layers.31.attn.q_proj.weight": "model-00005-of-00006.safetensors",
798
+ "model.layers.31.attn.v_conv1d.weight": "model-00005-of-00006.safetensors",
799
+ "model.layers.31.attn.v_proj.weight": "model-00005-of-00006.safetensors",
800
+ "model.layers.31.ffn.experts_w12": "model-00005-of-00006.safetensors",
801
+ "model.layers.31.ffn.experts_w3": "model-00005-of-00006.safetensors",
802
+ "model.layers.31.ffn.max_vio": "model-00005-of-00006.safetensors",
803
+ "model.layers.31.ffn.router.router_weights": "model-00005-of-00006.safetensors",
804
+ "model.layers.31.ffn.shared_experts.0.down.weight": "model-00005-of-00006.safetensors",
805
+ "model.layers.31.ffn.shared_experts.0.gate.weight": "model-00005-of-00006.safetensors",
806
+ "model.layers.31.ffn.shared_experts.0.up.weight": "model-00005-of-00006.safetensors",
807
+ "model.layers.31.ffn.w_down_proj.weight": "model-00005-of-00006.safetensors",
808
+ "model.layers.31.ffn.w_up_proj.weight": "model-00005-of-00006.safetensors",
809
+ "model.layers.31.injection_gate": "model-00005-of-00006.safetensors",
810
+ "model.layers.31.ln1.weight": "model-00005-of-00006.safetensors",
811
+ "model.layers.31.ln1_out.weight": "model-00005-of-00006.safetensors",
812
+ "model.layers.31.ln2.weight": "model-00005-of-00006.safetensors",
813
+ "model.layers.31.ln2_out.weight": "model-00005-of-00006.safetensors",
814
+ "model.layers.4.C_to_hidden.weight": "model-00005-of-00006.safetensors",
815
+ "model.layers.4.W_alpha.bias": "model-00005-of-00006.safetensors",
816
+ "model.layers.4.W_alpha.weight": "model-00005-of-00006.safetensors",
817
+ "model.layers.4.attn.g_norm_swish_gate.weight": "model-00005-of-00006.safetensors",
818
+ "model.layers.4.attn.g_proj.weight": "model-00005-of-00006.safetensors",
819
+ "model.layers.4.attn.gk_proj.0.weight": "model-00005-of-00006.safetensors",
820
+ "model.layers.4.attn.gk_proj.1.bias": "model-00005-of-00006.safetensors",
821
+ "model.layers.4.attn.gk_proj.1.weight": "model-00005-of-00006.safetensors",
822
+ "model.layers.4.attn.k_proj.weight": "model-00005-of-00006.safetensors",
823
+ "model.layers.4.attn.o_proj.weight": "model-00005-of-00006.safetensors",
824
+ "model.layers.4.attn.q_proj.weight": "model-00005-of-00006.safetensors",
825
+ "model.layers.4.attn.v_proj.weight": "model-00005-of-00006.safetensors",
826
+ "model.layers.4.ffn.experts_w12": "model-00005-of-00006.safetensors",
827
+ "model.layers.4.ffn.experts_w3": "model-00005-of-00006.safetensors",
828
+ "model.layers.4.ffn.max_vio": "model-00005-of-00006.safetensors",
829
+ "model.layers.4.ffn.router.router_weights": "model-00005-of-00006.safetensors",
830
+ "model.layers.4.ffn.shared_experts.0.down.weight": "model-00005-of-00006.safetensors",
831
+ "model.layers.4.ffn.shared_experts.0.gate.weight": "model-00005-of-00006.safetensors",
832
+ "model.layers.4.ffn.shared_experts.0.up.weight": "model-00005-of-00006.safetensors",
833
+ "model.layers.4.ffn.w_down_proj.weight": "model-00005-of-00006.safetensors",
834
+ "model.layers.4.ffn.w_up_proj.weight": "model-00005-of-00006.safetensors",
835
+ "model.layers.4.injection_gate": "model-00005-of-00006.safetensors",
836
+ "model.layers.4.ln1.weight": "model-00005-of-00006.safetensors",
837
+ "model.layers.4.ln1_out.weight": "model-00005-of-00006.safetensors",
838
+ "model.layers.4.ln2.weight": "model-00005-of-00006.safetensors",
839
+ "model.layers.4.ln2_out.weight": "model-00005-of-00006.safetensors",
840
+ "model.layers.4.memory.W_eta.bias": "model-00005-of-00006.safetensors",
841
+ "model.layers.4.memory.W_eta.weight": "model-00005-of-00006.safetensors",
842
+ "model.layers.4.memory.W_qkv_mem.weight": "model-00005-of-00006.safetensors",
843
+ "model.layers.4.memory.compress_z.0.weight": "model-00005-of-00006.safetensors",
844
+ "model.layers.4.memory.compress_z.2.weight": "model-00005-of-00006.safetensors",
845
+ "model.layers.4.memory.eta_channels": "model-00005-of-00006.safetensors",
846
+ "model.layers.4.memory.input_norm.bias": "model-00005-of-00006.safetensors",
847
+ "model.layers.4.memory.input_norm.weight": "model-00005-of-00006.safetensors",
848
+ "model.layers.4.memory.summary_proj.bias": "model-00005-of-00006.safetensors",
849
+ "model.layers.4.memory.summary_proj.weight": "model-00005-of-00006.safetensors",
850
+ "model.layers.4.memory.summary_query": "model-00005-of-00006.safetensors",
851
+ "model.layers.4.memory.temperature": "model-00005-of-00006.safetensors",
852
+ "model.layers.5.C_to_hidden.weight": "model-00005-of-00006.safetensors",
853
+ "model.layers.5.W_alpha.bias": "model-00005-of-00006.safetensors",
854
+ "model.layers.5.W_alpha.weight": "model-00005-of-00006.safetensors",
855
+ "model.layers.5.attn.g_norm_swish_gate.weight": "model-00005-of-00006.safetensors",
856
+ "model.layers.5.attn.g_proj.weight": "model-00005-of-00006.safetensors",
857
+ "model.layers.5.attn.gk_proj.0.weight": "model-00005-of-00006.safetensors",
858
+ "model.layers.5.attn.gk_proj.1.bias": "model-00005-of-00006.safetensors",
859
+ "model.layers.5.attn.gk_proj.1.weight": "model-00005-of-00006.safetensors",
860
+ "model.layers.5.attn.k_proj.weight": "model-00005-of-00006.safetensors",
861
+ "model.layers.5.attn.o_proj.weight": "model-00005-of-00006.safetensors",
862
+ "model.layers.5.attn.q_proj.weight": "model-00005-of-00006.safetensors",
863
+ "model.layers.5.attn.v_proj.weight": "model-00005-of-00006.safetensors",
864
+ "model.layers.5.ffn.experts_w12": "model-00005-of-00006.safetensors",
865
+ "model.layers.5.ffn.experts_w3": "model-00005-of-00006.safetensors",
866
+ "model.layers.5.ffn.max_vio": "model-00005-of-00006.safetensors",
867
+ "model.layers.5.ffn.router.router_weights": "model-00005-of-00006.safetensors",
868
+ "model.layers.5.ffn.shared_experts.0.down.weight": "model-00005-of-00006.safetensors",
869
+ "model.layers.5.ffn.shared_experts.0.gate.weight": "model-00005-of-00006.safetensors",
870
+ "model.layers.5.ffn.shared_experts.0.up.weight": "model-00005-of-00006.safetensors",
871
+ "model.layers.5.ffn.w_down_proj.weight": "model-00005-of-00006.safetensors",
872
+ "model.layers.5.ffn.w_up_proj.weight": "model-00005-of-00006.safetensors",
873
+ "model.layers.5.injection_gate": "model-00005-of-00006.safetensors",
874
+ "model.layers.5.ln1.weight": "model-00005-of-00006.safetensors",
875
+ "model.layers.5.ln1_out.weight": "model-00005-of-00006.safetensors",
876
+ "model.layers.5.ln2.weight": "model-00005-of-00006.safetensors",
877
+ "model.layers.5.ln2_out.weight": "model-00005-of-00006.safetensors",
878
+ "model.layers.5.memory.W_eta.bias": "model-00005-of-00006.safetensors",
879
+ "model.layers.5.memory.W_eta.weight": "model-00005-of-00006.safetensors",
880
+ "model.layers.5.memory.W_qkv_mem.weight": "model-00005-of-00006.safetensors",
881
+ "model.layers.5.memory.compress_z.0.weight": "model-00005-of-00006.safetensors",
882
+ "model.layers.5.memory.compress_z.2.weight": "model-00005-of-00006.safetensors",
883
+ "model.layers.5.memory.eta_channels": "model-00005-of-00006.safetensors",
884
+ "model.layers.5.memory.input_norm.bias": "model-00005-of-00006.safetensors",
885
+ "model.layers.5.memory.input_norm.weight": "model-00005-of-00006.safetensors",
886
+ "model.layers.5.memory.summary_proj.bias": "model-00005-of-00006.safetensors",
887
+ "model.layers.5.memory.summary_proj.weight": "model-00005-of-00006.safetensors",
888
+ "model.layers.5.memory.summary_query": "model-00005-of-00006.safetensors",
889
+ "model.layers.5.memory.temperature": "model-00005-of-00006.safetensors",
890
+ "model.layers.6.attn.A_log": "model-00005-of-00006.safetensors",
891
+ "model.layers.6.attn.b_proj.weight": "model-00005-of-00006.safetensors",
892
+ "model.layers.6.attn.dt_bias": "model-00005-of-00006.safetensors",
893
+ "model.layers.6.attn.f_proj.weight": "model-00005-of-00006.safetensors",
894
+ "model.layers.6.attn.g_proj.0.weight": "model-00005-of-00006.safetensors",
895
+ "model.layers.6.attn.g_proj.1.bias": "model-00005-of-00006.safetensors",
896
+ "model.layers.6.attn.g_proj.1.weight": "model-00005-of-00006.safetensors",
897
+ "model.layers.6.attn.k_conv1d.weight": "model-00005-of-00006.safetensors",
898
+ "model.layers.6.attn.k_proj.weight": "model-00005-of-00006.safetensors",
899
+ "model.layers.6.attn.o_norm.weight": "model-00005-of-00006.safetensors",
900
+ "model.layers.6.attn.o_proj.weight": "model-00005-of-00006.safetensors",
901
+ "model.layers.6.attn.q_conv1d.weight": "model-00005-of-00006.safetensors",
902
+ "model.layers.6.attn.q_proj.weight": "model-00005-of-00006.safetensors",
903
+ "model.layers.6.attn.v_conv1d.weight": "model-00005-of-00006.safetensors",
904
+ "model.layers.6.attn.v_proj.weight": "model-00005-of-00006.safetensors",
905
+ "model.layers.6.ffn.experts_w12": "model-00005-of-00006.safetensors",
906
+ "model.layers.6.ffn.experts_w3": "model-00005-of-00006.safetensors",
907
+ "model.layers.6.ffn.max_vio": "model-00005-of-00006.safetensors",
908
+ "model.layers.6.ffn.router.router_weights": "model-00005-of-00006.safetensors",
909
+ "model.layers.6.ffn.shared_experts.0.down.weight": "model-00005-of-00006.safetensors",
910
+ "model.layers.6.ffn.shared_experts.0.gate.weight": "model-00005-of-00006.safetensors",
911
+ "model.layers.6.ffn.shared_experts.0.up.weight": "model-00005-of-00006.safetensors",
912
+ "model.layers.6.ffn.w_down_proj.weight": "model-00005-of-00006.safetensors",
913
+ "model.layers.6.ffn.w_up_proj.weight": "model-00005-of-00006.safetensors",
914
+ "model.layers.6.injection_gate": "model-00005-of-00006.safetensors",
915
+ "model.layers.6.ln1.weight": "model-00005-of-00006.safetensors",
916
+ "model.layers.6.ln1_out.weight": "model-00005-of-00006.safetensors",
917
+ "model.layers.6.ln2.weight": "model-00005-of-00006.safetensors",
918
+ "model.layers.6.ln2_out.weight": "model-00005-of-00006.safetensors",
919
+ "model.layers.7.attn.A_log": "model-00005-of-00006.safetensors",
920
+ "model.layers.7.attn.b_proj.weight": "model-00005-of-00006.safetensors",
921
+ "model.layers.7.attn.dt_bias": "model-00005-of-00006.safetensors",
922
+ "model.layers.7.attn.f_proj.weight": "model-00005-of-00006.safetensors",
923
+ "model.layers.7.attn.g_proj.0.weight": "model-00005-of-00006.safetensors",
924
+ "model.layers.7.attn.g_proj.1.bias": "model-00005-of-00006.safetensors",
925
+ "model.layers.7.attn.g_proj.1.weight": "model-00005-of-00006.safetensors",
926
+ "model.layers.7.attn.k_conv1d.weight": "model-00005-of-00006.safetensors",
927
+ "model.layers.7.attn.k_proj.weight": "model-00005-of-00006.safetensors",
928
+ "model.layers.7.attn.o_norm.weight": "model-00005-of-00006.safetensors",
929
+ "model.layers.7.attn.o_proj.weight": "model-00005-of-00006.safetensors",
930
+ "model.layers.7.attn.q_conv1d.weight": "model-00005-of-00006.safetensors",
931
+ "model.layers.7.attn.q_proj.weight": "model-00005-of-00006.safetensors",
932
+ "model.layers.7.attn.v_conv1d.weight": "model-00005-of-00006.safetensors",
933
+ "model.layers.7.attn.v_proj.weight": "model-00005-of-00006.safetensors",
934
+ "model.layers.7.ffn.experts_w12": "model-00005-of-00006.safetensors",
935
+ "model.layers.7.ffn.experts_w3": "model-00005-of-00006.safetensors",
936
+ "model.layers.7.ffn.max_vio": "model-00005-of-00006.safetensors",
937
+ "model.layers.7.ffn.router.router_weights": "model-00005-of-00006.safetensors",
938
+ "model.layers.7.ffn.shared_experts.0.down.weight": "model-00005-of-00006.safetensors",
939
+ "model.layers.7.ffn.shared_experts.0.gate.weight": "model-00005-of-00006.safetensors",
940
+ "model.layers.7.ffn.shared_experts.0.up.weight": "model-00005-of-00006.safetensors",
941
+ "model.layers.7.ffn.w_down_proj.weight": "model-00005-of-00006.safetensors",
942
+ "model.layers.7.ffn.w_up_proj.weight": "model-00005-of-00006.safetensors",
943
+ "model.layers.7.injection_gate": "model-00005-of-00006.safetensors",
944
+ "model.layers.7.ln1.weight": "model-00005-of-00006.safetensors",
945
+ "model.layers.7.ln1_out.weight": "model-00005-of-00006.safetensors",
946
+ "model.layers.7.ln2.weight": "model-00005-of-00006.safetensors",
947
+ "model.layers.7.ln2_out.weight": "model-00005-of-00006.safetensors",
948
+ "model.layers.8.attn.A_log": "model-00005-of-00006.safetensors",
949
+ "model.layers.8.attn.b_proj.weight": "model-00005-of-00006.safetensors",
950
+ "model.layers.8.attn.dt_bias": "model-00005-of-00006.safetensors",
951
+ "model.layers.8.attn.f_proj.weight": "model-00005-of-00006.safetensors",
952
+ "model.layers.8.attn.g_proj.0.weight": "model-00005-of-00006.safetensors",
953
+ "model.layers.8.attn.g_proj.1.bias": "model-00005-of-00006.safetensors",
954
+ "model.layers.8.attn.g_proj.1.weight": "model-00005-of-00006.safetensors",
955
+ "model.layers.8.attn.k_conv1d.weight": "model-00005-of-00006.safetensors",
956
+ "model.layers.8.attn.k_proj.weight": "model-00005-of-00006.safetensors",
957
+ "model.layers.8.attn.o_norm.weight": "model-00005-of-00006.safetensors",
958
+ "model.layers.8.attn.o_proj.weight": "model-00005-of-00006.safetensors",
959
+ "model.layers.8.attn.q_conv1d.weight": "model-00005-of-00006.safetensors",
960
+ "model.layers.8.attn.q_proj.weight": "model-00005-of-00006.safetensors",
961
+ "model.layers.8.attn.v_conv1d.weight": "model-00005-of-00006.safetensors",
962
+ "model.layers.8.attn.v_proj.weight": "model-00005-of-00006.safetensors",
963
+ "model.layers.8.ffn.experts_w12": "model-00006-of-00006.safetensors",
964
+ "model.layers.8.ffn.experts_w3": "model-00006-of-00006.safetensors",
965
+ "model.layers.8.ffn.max_vio": "model-00006-of-00006.safetensors",
966
+ "model.layers.8.ffn.router.router_weights": "model-00006-of-00006.safetensors",
967
+ "model.layers.8.ffn.shared_experts.0.down.weight": "model-00006-of-00006.safetensors",
968
+ "model.layers.8.ffn.shared_experts.0.gate.weight": "model-00006-of-00006.safetensors",
969
+ "model.layers.8.ffn.shared_experts.0.up.weight": "model-00006-of-00006.safetensors",
970
+ "model.layers.8.ffn.w_down_proj.weight": "model-00006-of-00006.safetensors",
971
+ "model.layers.8.ffn.w_up_proj.weight": "model-00006-of-00006.safetensors",
972
+ "model.layers.8.injection_gate": "model-00006-of-00006.safetensors",
973
+ "model.layers.8.ln1.weight": "model-00006-of-00006.safetensors",
974
+ "model.layers.8.ln1_out.weight": "model-00006-of-00006.safetensors",
975
+ "model.layers.8.ln2.weight": "model-00006-of-00006.safetensors",
976
+ "model.layers.8.ln2_out.weight": "model-00006-of-00006.safetensors",
977
+ "model.layers.9.attn.A_log": "model-00006-of-00006.safetensors",
978
+ "model.layers.9.attn.b_proj.weight": "model-00006-of-00006.safetensors",
979
+ "model.layers.9.attn.dt_bias": "model-00006-of-00006.safetensors",
980
+ "model.layers.9.attn.f_proj.weight": "model-00006-of-00006.safetensors",
981
+ "model.layers.9.attn.g_proj.0.weight": "model-00006-of-00006.safetensors",
982
+ "model.layers.9.attn.g_proj.1.bias": "model-00006-of-00006.safetensors",
983
+ "model.layers.9.attn.g_proj.1.weight": "model-00006-of-00006.safetensors",
984
+ "model.layers.9.attn.k_conv1d.weight": "model-00006-of-00006.safetensors",
985
+ "model.layers.9.attn.k_proj.weight": "model-00006-of-00006.safetensors",
986
+ "model.layers.9.attn.o_norm.weight": "model-00006-of-00006.safetensors",
987
+ "model.layers.9.attn.o_proj.weight": "model-00006-of-00006.safetensors",
988
+ "model.layers.9.attn.q_conv1d.weight": "model-00006-of-00006.safetensors",
989
+ "model.layers.9.attn.q_proj.weight": "model-00006-of-00006.safetensors",
990
+ "model.layers.9.attn.v_conv1d.weight": "model-00006-of-00006.safetensors",
991
+ "model.layers.9.attn.v_proj.weight": "model-00006-of-00006.safetensors",
992
+ "model.layers.9.ffn.experts_w12": "model-00006-of-00006.safetensors",
993
+ "model.layers.9.ffn.experts_w3": "model-00006-of-00006.safetensors",
994
+ "model.layers.9.ffn.max_vio": "model-00006-of-00006.safetensors",
995
+ "model.layers.9.ffn.router.router_weights": "model-00006-of-00006.safetensors",
996
+ "model.layers.9.ffn.shared_experts.0.down.weight": "model-00006-of-00006.safetensors",
997
+ "model.layers.9.ffn.shared_experts.0.gate.weight": "model-00006-of-00006.safetensors",
998
+ "model.layers.9.ffn.shared_experts.0.up.weight": "model-00006-of-00006.safetensors",
999
+ "model.layers.9.ffn.w_down_proj.weight": "model-00006-of-00006.safetensors",
1000
+ "model.layers.9.ffn.w_up_proj.weight": "model-00006-of-00006.safetensors",
1001
+ "model.layers.9.injection_gate": "model-00006-of-00006.safetensors",
1002
+ "model.layers.9.ln1.weight": "model-00006-of-00006.safetensors",
1003
+ "model.layers.9.ln1_out.weight": "model-00006-of-00006.safetensors",
1004
+ "model.layers.9.ln2.weight": "model-00006-of-00006.safetensors",
1005
+ "model.layers.9.ln2_out.weight": "model-00006-of-00006.safetensors",
1006
+ "model.norm.weight": "model-00006-of-00006.safetensors"
1007
+ }
1008
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:daab2354f8a74e70d70b4d1f804939b68a8c9624dd06cb7858e52dd8970e9726
3
+ size 33384567
tokenizer_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "boi_token": "<start_of_image>",
4
+ "bos_token": "<bos>",
5
+ "clean_up_tokenization_spaces": false,
6
+ "eoi_token": "<end_of_image>",
7
+ "eos_token": "<end_of_turn>",
8
+ "image_token": "<image_soft_token>",
9
+ "is_local": true,
10
+ "local_files_only": false,
11
+ "mask_token": "<mask>",
12
+ "model_max_length": 32768,
13
+ "model_specific_special_tokens": {
14
+ "boi_token": "<start_of_image>",
15
+ "eoi_token": "<end_of_image>",
16
+ "image_token": "<image_soft_token>"
17
+ },
18
+ "pad_token": "<pad>",
19
+ "padding_side": "left",
20
+ "processor_class": "Gemma3Processor",
21
+ "sp_model_kwargs": null,
22
+ "spaces_between_special_tokens": false,
23
+ "tokenizer_class": "GemmaTokenizer",
24
+ "unk_token": "<unk>",
25
+ "use_default_system_prompt": false
26
+ }