arianraje commited on
Commit
36767e6
·
verified ·
1 Parent(s): 688cbdd

stage3-2.0B-OPD decayed final (H=32K, matched 1.8B recipe; run mimo-mathmix-h32k-runpod-2p0b-v1-20260905)

Browse files
.gitattributes CHANGED
@@ -33,3 +33,7 @@ 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
+ full_eval/reasoning_bench_aime24.generations.json filter=lfs diff=lfs merge=lfs -text
37
+ full_eval/reasoning_bench_aime25.generations.json filter=lfs diff=lfs merge=lfs -text
38
+ full_eval/reasoning_bench_think_math500.generations.json filter=lfs diff=lfs merge=lfs -text
39
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
added_tokens.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</tool_call>": 151658,
3
+ "<tool_call>": 151657,
4
+ "<|box_end|>": 151649,
5
+ "<|box_start|>": 151648,
6
+ "<|endoftext|>": 151643,
7
+ "<|file_sep|>": 151664,
8
+ "<|fim_middle|>": 151660,
9
+ "<|fim_pad|>": 151662,
10
+ "<|fim_prefix|>": 151659,
11
+ "<|fim_suffix|>": 151661,
12
+ "<|im_end|>": 151645,
13
+ "<|im_start|>": 151644,
14
+ "<|image_pad|>": 151655,
15
+ "<|object_ref_end|>": 151647,
16
+ "<|object_ref_start|>": 151646,
17
+ "<|quad_end|>": 151651,
18
+ "<|quad_start|>": 151650,
19
+ "<|repo_name|>": 151663,
20
+ "<|video_pad|>": 151656,
21
+ "<|vision_end|>": 151653,
22
+ "<|vision_pad|>": 151654,
23
+ "<|vision_start|>": 151652
24
+ }
chat_template.jinja ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0]['role'] == 'system' %}
4
+ {{- messages[0]['content'] }}
5
+ {%- else %}
6
+ {{- 'You are MiMo, an AI assistant developed by Xiaomi.' }}
7
+ {%- endif %}
8
+ {{- "\n\n# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
9
+ {%- for tool in tools %}
10
+ {{- "\n" }}
11
+ {{- tool | tojson }}
12
+ {%- endfor %}
13
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
14
+ {%- else %}
15
+ {%- if messages[0]['role'] == 'system' %}
16
+ {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
17
+ {%- else %}
18
+ {{- '<|im_start|>system\nYou are MiMo, an AI assistant developed by Xiaomi.<|im_end|>\n' }}
19
+ {%- endif %}
20
+ {%- endif %}
21
+ {%- for message in messages %}
22
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
23
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
24
+ {%- elif message.role == "assistant" %}
25
+ {{- '<|im_start|>' + message.role }}
26
+ {%- if message.content %}
27
+ {{- '\n' + message.content }}
28
+ {%- endif %}
29
+ {%- for tool_call in message.tool_calls %}
30
+ {%- if tool_call.function is defined %}
31
+ {%- set tool_call = tool_call.function %}
32
+ {%- endif %}
33
+ {{- '\n<tool_call>\n{"name": "' }}
34
+ {{- tool_call.name }}
35
+ {{- '", "arguments": ' }}
36
+ {{- tool_call.arguments | tojson }}
37
+ {{- '}\n</tool_call>' }}
38
+ {%- endfor %}
39
+ {{- '<|im_end|>\n' }}
40
+ {%- elif message.role == "tool" %}
41
+ {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
42
+ {{- '<|im_start|>user' }}
43
+ {%- endif %}
44
+ {{- '\n<tool_response>\n' }}
45
+ {{- message.content }}
46
+ {{- '\n</tool_response>' }}
47
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
48
+ {{- '<|im_end|>\n' }}
49
+ {%- endif %}
50
+ {%- endif %}
51
+ {%- endfor %}
52
+ {%- if add_generation_prompt %}
53
+ {{- '<|im_start|>assistant\n' }}
54
+ {%- if enable_thinking is defined and enable_thinking is false %}
55
+ {{- '<think>\n\n</think>\n' }}
56
+ {%- endif %}
57
+ {%- if enable_thinking is defined and enable_thinking is true %}
58
+ {{- '<think>\n' }}
59
+ {%- endif %}
60
+ {%- endif %}
config.json ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "MiMoGDNForCausalLM"
4
+ ],
5
+ "attention_dropout": 0.0,
6
+ "bos_token_id": 151643,
7
+ "dtype": "bfloat16",
8
+ "eos_token_id": 151645,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 4096,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 11008,
13
+ "layer_types": [
14
+ "linear_attention",
15
+ "linear_attention",
16
+ "linear_attention",
17
+ "full_attention",
18
+ "linear_attention",
19
+ "linear_attention",
20
+ "linear_attention",
21
+ "full_attention",
22
+ "linear_attention",
23
+ "linear_attention",
24
+ "linear_attention",
25
+ "full_attention",
26
+ "linear_attention",
27
+ "linear_attention",
28
+ "linear_attention",
29
+ "full_attention",
30
+ "linear_attention",
31
+ "linear_attention",
32
+ "linear_attention",
33
+ "full_attention",
34
+ "linear_attention",
35
+ "linear_attention",
36
+ "linear_attention",
37
+ "full_attention",
38
+ "linear_attention",
39
+ "linear_attention",
40
+ "linear_attention",
41
+ "full_attention",
42
+ "linear_attention",
43
+ "linear_attention",
44
+ "linear_attention",
45
+ "full_attention",
46
+ "linear_attention",
47
+ "linear_attention",
48
+ "linear_attention",
49
+ "full_attention"
50
+ ],
51
+ "linear_conv_kernel_dim": 4,
52
+ "linear_key_head_dim": 128,
53
+ "linear_num_key_heads": 16,
54
+ "linear_num_value_heads": 32,
55
+ "linear_value_head_dim": 128,
56
+ "max_position_embeddings": 65536,
57
+ "max_window_layers": 28,
58
+ "model_type": "mimo_gdn",
59
+ "num_attention_heads": 32,
60
+ "num_hidden_layers": 36,
61
+ "num_key_value_heads": 8,
62
+ "rms_norm_eps": 1e-05,
63
+ "rope_scaling": null,
64
+ "rope_theta": 640000,
65
+ "sliding_window": null,
66
+ "tie_word_embeddings": false,
67
+ "transformers_version": "4.57.3",
68
+ "use_cache": true,
69
+ "use_sliding_window": false,
70
+ "vocab_size": 151680
71
+ }
full_eval/gsm8k_1024.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model": "/workspace/mimo/checkpoints/wsd_runs/mimo-wsd-mathmix-h32k-2p0b-v1/snapshots/final",
3
+ "model_args": "pretrained=/workspace/mimo/checkpoints/wsd_runs/mimo-wsd-mathmix-h32k-2p0b-v1/snapshots/final,dtype=bfloat16,gpu_memory_utilization=0.85,max_model_len=8192",
4
+ "limit": null,
5
+ "lm_eval_version": "0.4.12",
6
+ "results": {
7
+ "gsm8k": {
8
+ "sample_len": 1319,
9
+ "exact_match,strict-match": 0.5807429871114481,
10
+ "exact_match_stderr,strict-match": 0.013591720959042115,
11
+ "exact_match,flexible-extract": 0.621683093252464,
12
+ "exact_match_stderr,flexible-extract": 0.013358407831777115
13
+ }
14
+ },
15
+ "task_status": {
16
+ "gsm8k": "ok (292s)"
17
+ },
18
+ "finished": "2026-09-06 20:49:16"
19
+ }
full_eval/likelihood_evals.json ADDED
@@ -0,0 +1,360 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model": "/workspace/mimo/checkpoints/wsd_runs/mimo-wsd-mathmix-h32k-2p0b-v1/snapshots/final",
3
+ "model_args": "pretrained=/workspace/mimo/checkpoints/wsd_runs/mimo-wsd-mathmix-h32k-2p0b-v1/snapshots/final,dtype=bfloat16,gpu_memory_utilization=0.85,max_model_len=8192",
4
+ "limit": null,
5
+ "lm_eval_version": "0.4.12",
6
+ "results": {
7
+ "piqa": {
8
+ "sample_len": 1838,
9
+ "acc,none": 0.7105549510337323,
10
+ "acc_stderr,none": 0.010581014740675618,
11
+ "acc_norm,none": 0.7247007616974973,
12
+ "acc_norm_stderr,none": 0.01042142927736953
13
+ },
14
+ "hellaswag": {
15
+ "sample_len": 10042,
16
+ "acc,none": 0.448814977096196,
17
+ "acc_stderr,none": 0.0049635670291290485,
18
+ "acc_norm,none": 0.5990838478390759,
19
+ "acc_norm_stderr,none": 0.004890824718530303
20
+ },
21
+ "arc_easy": {
22
+ "sample_len": 2376,
23
+ "acc,none": 0.6738215488215489,
24
+ "acc_stderr,none": 0.009619849417035165,
25
+ "acc_norm,none": 0.627104377104377,
26
+ "acc_norm_stderr,none": 0.009922743197129248
27
+ },
28
+ "arc_challenge": {
29
+ "sample_len": 1172,
30
+ "acc,none": 0.3575085324232082,
31
+ "acc_stderr,none": 0.014005494275916573,
32
+ "acc_norm,none": 0.39505119453924914,
33
+ "acc_norm_stderr,none": 0.014285898292938165
34
+ },
35
+ "winogrande": {
36
+ "sample_len": 1267,
37
+ "acc,none": 0.5880031570639306,
38
+ "acc_stderr,none": 0.013833112857645938
39
+ },
40
+ "mmlu_abstract_algebra": {
41
+ "sample_len": 100,
42
+ "acc,none": 0.42,
43
+ "acc_stderr,none": 0.049604496374885836
44
+ },
45
+ "mmlu_anatomy": {
46
+ "sample_len": 135,
47
+ "acc,none": 0.5037037037037037,
48
+ "acc_stderr,none": 0.04319223625811331
49
+ },
50
+ "mmlu_astronomy": {
51
+ "sample_len": 152,
52
+ "acc,none": 0.6381578947368421,
53
+ "acc_stderr,none": 0.03910525752849724
54
+ },
55
+ "mmlu_college_biology": {
56
+ "sample_len": 144,
57
+ "acc,none": 0.6944444444444444,
58
+ "acc_stderr,none": 0.03852084696008534
59
+ },
60
+ "mmlu_college_chemistry": {
61
+ "sample_len": 100,
62
+ "acc,none": 0.53,
63
+ "acc_stderr,none": 0.050161355804659205
64
+ },
65
+ "mmlu_college_computer_science": {
66
+ "sample_len": 100,
67
+ "acc,none": 0.51,
68
+ "acc_stderr,none": 0.05024183937956912
69
+ },
70
+ "mmlu_college_mathematics": {
71
+ "sample_len": 100,
72
+ "acc,none": 0.45,
73
+ "acc_stderr,none": 0.05
74
+ },
75
+ "mmlu_college_physics": {
76
+ "sample_len": 102,
77
+ "acc,none": 0.4411764705882353,
78
+ "acc_stderr,none": 0.04940635630605659
79
+ },
80
+ "mmlu_computer_security": {
81
+ "sample_len": 100,
82
+ "acc,none": 0.64,
83
+ "acc_stderr,none": 0.04824181513244218
84
+ },
85
+ "mmlu_conceptual_physics": {
86
+ "sample_len": 235,
87
+ "acc,none": 0.6425531914893617,
88
+ "acc_stderr,none": 0.031329417894764254
89
+ },
90
+ "mmlu_electrical_engineering": {
91
+ "sample_len": 145,
92
+ "acc,none": 0.593103448275862,
93
+ "acc_stderr,none": 0.04093793981266236
94
+ },
95
+ "mmlu_elementary_mathematics": {
96
+ "sample_len": 378,
97
+ "acc,none": 0.47883597883597884,
98
+ "acc_stderr,none": 0.025728230952130716
99
+ },
100
+ "mmlu_high_school_biology": {
101
+ "sample_len": 310,
102
+ "acc,none": 0.7258064516129032,
103
+ "acc_stderr,none": 0.025378139970885196
104
+ },
105
+ "mmlu_high_school_chemistry": {
106
+ "sample_len": 203,
107
+ "acc,none": 0.5862068965517241,
108
+ "acc_stderr,none": 0.03465304488406796
109
+ },
110
+ "mmlu_high_school_computer_science": {
111
+ "sample_len": 100,
112
+ "acc,none": 0.65,
113
+ "acc_stderr,none": 0.0479372485441102
114
+ },
115
+ "mmlu_high_school_mathematics": {
116
+ "sample_len": 270,
117
+ "acc,none": 0.4074074074074074,
118
+ "acc_stderr,none": 0.029958249250082114
119
+ },
120
+ "mmlu_high_school_physics": {
121
+ "sample_len": 151,
122
+ "acc,none": 0.4900662251655629,
123
+ "acc_stderr,none": 0.04081677107248437
124
+ },
125
+ "mmlu_high_school_statistics": {
126
+ "sample_len": 216,
127
+ "acc,none": 0.5601851851851852,
128
+ "acc_stderr,none": 0.033851779760448106
129
+ },
130
+ "mmlu_machine_learning": {
131
+ "sample_len": 112,
132
+ "acc,none": 0.32142857142857145,
133
+ "acc_stderr,none": 0.044328040552915185
134
+ },
135
+ "mmlu_business_ethics": {
136
+ "sample_len": 100,
137
+ "acc,none": 0.64,
138
+ "acc_stderr,none": 0.048241815132442176
139
+ },
140
+ "mmlu_clinical_knowledge": {
141
+ "sample_len": 265,
142
+ "acc,none": 0.6377358490566037,
143
+ "acc_stderr,none": 0.0295822451283843
144
+ },
145
+ "mmlu_college_medicine": {
146
+ "sample_len": 173,
147
+ "acc,none": 0.5664739884393064,
148
+ "acc_stderr,none": 0.03778621079092055
149
+ },
150
+ "mmlu_global_facts": {
151
+ "sample_len": 100,
152
+ "acc,none": 0.26,
153
+ "acc_stderr,none": 0.04408440022768078
154
+ },
155
+ "mmlu_human_aging": {
156
+ "sample_len": 223,
157
+ "acc,none": 0.5964125560538116,
158
+ "acc_stderr,none": 0.03292802819330314
159
+ },
160
+ "mmlu_management": {
161
+ "sample_len": 103,
162
+ "acc,none": 0.7184466019417476,
163
+ "acc_stderr,none": 0.044532548363264673
164
+ },
165
+ "mmlu_marketing": {
166
+ "sample_len": 234,
167
+ "acc,none": 0.7307692307692307,
168
+ "acc_stderr,none": 0.029058588303748842
169
+ },
170
+ "mmlu_medical_genetics": {
171
+ "sample_len": 100,
172
+ "acc,none": 0.58,
173
+ "acc_stderr,none": 0.049604496374885836
174
+ },
175
+ "mmlu_miscellaneous": {
176
+ "sample_len": 783,
177
+ "acc,none": 0.6679438058748404,
178
+ "acc_stderr,none": 0.016841174655295728
179
+ },
180
+ "mmlu_nutrition": {
181
+ "sample_len": 306,
182
+ "acc,none": 0.6470588235294118,
183
+ "acc_stderr,none": 0.027363593284684965
184
+ },
185
+ "mmlu_professional_accounting": {
186
+ "sample_len": 282,
187
+ "acc,none": 0.4148936170212766,
188
+ "acc_stderr,none": 0.0293922365846125
189
+ },
190
+ "mmlu_professional_medicine": {
191
+ "sample_len": 272,
192
+ "acc,none": 0.4375,
193
+ "acc_stderr,none": 0.030134614954403924
194
+ },
195
+ "mmlu_virology": {
196
+ "sample_len": 166,
197
+ "acc,none": 0.42168674698795183,
198
+ "acc_stderr,none": 0.038444531817709175
199
+ },
200
+ "mmlu_econometrics": {
201
+ "sample_len": 114,
202
+ "acc,none": 0.41228070175438597,
203
+ "acc_stderr,none": 0.04630653203366596
204
+ },
205
+ "mmlu_high_school_geography": {
206
+ "sample_len": 198,
207
+ "acc,none": 0.702020202020202,
208
+ "acc_stderr,none": 0.032586303838365555
209
+ },
210
+ "mmlu_high_school_government_and_politics": {
211
+ "sample_len": 193,
212
+ "acc,none": 0.7202072538860104,
213
+ "acc_stderr,none": 0.03239637046735703
214
+ },
215
+ "mmlu_high_school_macroeconomics": {
216
+ "sample_len": 390,
217
+ "acc,none": 0.5717948717948718,
218
+ "acc_stderr,none": 0.025088301454694834
219
+ },
220
+ "mmlu_high_school_microeconomics": {
221
+ "sample_len": 238,
222
+ "acc,none": 0.634453781512605,
223
+ "acc_stderr,none": 0.031282177063684614
224
+ },
225
+ "mmlu_high_school_psychology": {
226
+ "sample_len": 545,
227
+ "acc,none": 0.7651376146788991,
228
+ "acc_stderr,none": 0.018175110510343578
229
+ },
230
+ "mmlu_human_sexuality": {
231
+ "sample_len": 131,
232
+ "acc,none": 0.6106870229007634,
233
+ "acc_stderr,none": 0.04276486542814591
234
+ },
235
+ "mmlu_professional_psychology": {
236
+ "sample_len": 612,
237
+ "acc,none": 0.5,
238
+ "acc_stderr,none": 0.020227834851568375
239
+ },
240
+ "mmlu_public_relations": {
241
+ "sample_len": 110,
242
+ "acc,none": 0.5727272727272728,
243
+ "acc_stderr,none": 0.047381987035454834
244
+ },
245
+ "mmlu_security_studies": {
246
+ "sample_len": 245,
247
+ "acc,none": 0.6204081632653061,
248
+ "acc_stderr,none": 0.03106721126287249
249
+ },
250
+ "mmlu_sociology": {
251
+ "sample_len": 201,
252
+ "acc,none": 0.6915422885572139,
253
+ "acc_stderr,none": 0.03265819588512697
254
+ },
255
+ "mmlu_us_foreign_policy": {
256
+ "sample_len": 100,
257
+ "acc,none": 0.65,
258
+ "acc_stderr,none": 0.047937248544110196
259
+ },
260
+ "mmlu_formal_logic": {
261
+ "sample_len": 126,
262
+ "acc,none": 0.38095238095238093,
263
+ "acc_stderr,none": 0.04343525428949098
264
+ },
265
+ "mmlu_high_school_european_history": {
266
+ "sample_len": 165,
267
+ "acc,none": 0.6606060606060606,
268
+ "acc_stderr,none": 0.036974422050315946
269
+ },
270
+ "mmlu_high_school_us_history": {
271
+ "sample_len": 204,
272
+ "acc,none": 0.696078431372549,
273
+ "acc_stderr,none": 0.032282103870378935
274
+ },
275
+ "mmlu_high_school_world_history": {
276
+ "sample_len": 237,
277
+ "acc,none": 0.7172995780590717,
278
+ "acc_stderr,none": 0.029312814153955924
279
+ },
280
+ "mmlu_international_law": {
281
+ "sample_len": 121,
282
+ "acc,none": 0.6859504132231405,
283
+ "acc_stderr,none": 0.04236964753041018
284
+ },
285
+ "mmlu_jurisprudence": {
286
+ "sample_len": 108,
287
+ "acc,none": 0.5648148148148148,
288
+ "acc_stderr,none": 0.04792898170907061
289
+ },
290
+ "mmlu_logical_fallacies": {
291
+ "sample_len": 163,
292
+ "acc,none": 0.5214723926380368,
293
+ "acc_stderr,none": 0.03924746876751129
294
+ },
295
+ "mmlu_moral_disputes": {
296
+ "sample_len": 346,
297
+ "acc,none": 0.5260115606936416,
298
+ "acc_stderr,none": 0.026882643434022895
299
+ },
300
+ "mmlu_moral_scenarios": {
301
+ "sample_len": 895,
302
+ "acc,none": 0.26033519553072626,
303
+ "acc_stderr,none": 0.01467625200931947
304
+ },
305
+ "mmlu_philosophy": {
306
+ "sample_len": 311,
307
+ "acc,none": 0.572347266881029,
308
+ "acc_stderr,none": 0.02809924077580957
309
+ },
310
+ "mmlu_prehistory": {
311
+ "sample_len": 324,
312
+ "acc,none": 0.5524691358024691,
313
+ "acc_stderr,none": 0.02766713856942271
314
+ },
315
+ "mmlu_professional_law": {
316
+ "sample_len": 1534,
317
+ "acc,none": 0.38722294654498046,
318
+ "acc_stderr,none": 0.012441155326854922
319
+ },
320
+ "mmlu_world_religions": {
321
+ "sample_len": 171,
322
+ "acc,none": 0.7309941520467836,
323
+ "acc_stderr,none": 0.034010526201040885
324
+ },
325
+ "mmlu_stem": {
326
+ "sample_len": 3153,
327
+ "acc,none": 0.5496352679987314,
328
+ "acc_stderr,none": 0.008677976981655082
329
+ },
330
+ "mmlu_other": {
331
+ "sample_len": 3107,
332
+ "acc,none": 0.5857740585774058,
333
+ "acc_stderr,none": 0.008596016975905191
334
+ },
335
+ "mmlu_social_sciences": {
336
+ "sample_len": 3077,
337
+ "acc,none": 0.6243093922651933,
338
+ "acc_stderr,none": 0.00855938769188614
339
+ },
340
+ "mmlu_humanities": {
341
+ "sample_len": 4705,
342
+ "acc,none": 0.46524973432518596,
343
+ "acc_stderr,none": 0.006941343921612515
344
+ },
345
+ "mmlu": {
346
+ "sample_len": 14042,
347
+ "acc,none": 0.5457199829084176,
348
+ "acc_stderr,none": 0.0040424502772281245
349
+ }
350
+ },
351
+ "task_status": {
352
+ "piqa": "ok (327s)",
353
+ "hellaswag": "ok (375s)",
354
+ "arc_easy": "ok (238s)",
355
+ "arc_challenge": "ok (217s)",
356
+ "winogrande": "ok (210s)",
357
+ "mmlu": "ok (1887s)"
358
+ },
359
+ "finished": "2026-09-06 20:27:54"
360
+ }
full_eval/reasoning_bench_aime24.generations.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35e320babe6c2530aef78621a7afe7a4e418b25a03ac92e9e0585b19f8a4057e
3
+ size 15557367
full_eval/reasoning_bench_aime24.json ADDED
@@ -0,0 +1,1065 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model": "/workspace/mimo/checkpoints/wsd_runs/mimo-wsd-mathmix-h32k-2p0b-v1/snapshots/final",
3
+ "label": "stage3-2.0B-OPD-aime24-think",
4
+ "config": {
5
+ "tasks": "aime24",
6
+ "n": null,
7
+ "max_tokens": 32768,
8
+ "no_thinking": false,
9
+ "temperature": null,
10
+ "top_p": null,
11
+ "top_k": null,
12
+ "ruler_lengths": "4096,8192,16384,32768",
13
+ "ruler_tasks": "niah_single,niah_multikey,vt",
14
+ "ruler_item_tokenizer": null,
15
+ "ruler_items": 25,
16
+ "ruler_gen_tokens": 128,
17
+ "ruler_answer_prefill": false,
18
+ "limit": null,
19
+ "offset": 0,
20
+ "seed": 0,
21
+ "tp": 1,
22
+ "max_model_len": 40960,
23
+ "gpu_mem_util": 0.9,
24
+ "enforce_eager": false,
25
+ "max_num_seqs": 32
26
+ },
27
+ "thinking": true,
28
+ "sampling": {
29
+ "temperature": 0.6,
30
+ "top_p": 0.95,
31
+ "top_k": 20,
32
+ "max_tokens": 32768
33
+ },
34
+ "results": {
35
+ "aime24": {
36
+ "accuracy": 0.5958333333333333,
37
+ "n_problems": 30,
38
+ "n_samples": 8,
39
+ "truncated_frac": 0.2916666666666667,
40
+ "mean_gen_tokens": 21861.879166666666,
41
+ "per_problem": [
42
+ {
43
+ "idx": 0,
44
+ "gold": "204",
45
+ "correct": [
46
+ false,
47
+ true,
48
+ true,
49
+ true,
50
+ true,
51
+ true,
52
+ true,
53
+ true
54
+ ],
55
+ "truncated": [
56
+ false,
57
+ false,
58
+ false,
59
+ false,
60
+ false,
61
+ false,
62
+ false,
63
+ false
64
+ ],
65
+ "gen_tokens": [
66
+ 21810,
67
+ 5050,
68
+ 6887,
69
+ 5729,
70
+ 6982,
71
+ 5493,
72
+ 6791,
73
+ 5284
74
+ ]
75
+ },
76
+ {
77
+ "idx": 1,
78
+ "gold": "113",
79
+ "correct": [
80
+ false,
81
+ true,
82
+ true,
83
+ true,
84
+ false,
85
+ true,
86
+ false,
87
+ false
88
+ ],
89
+ "truncated": [
90
+ true,
91
+ false,
92
+ false,
93
+ false,
94
+ true,
95
+ false,
96
+ true,
97
+ false
98
+ ],
99
+ "gen_tokens": [
100
+ 32768,
101
+ 23431,
102
+ 21880,
103
+ 21275,
104
+ 32768,
105
+ 32686,
106
+ 32768,
107
+ 32378
108
+ ]
109
+ },
110
+ {
111
+ "idx": 2,
112
+ "gold": "371",
113
+ "correct": [
114
+ false,
115
+ false,
116
+ false,
117
+ false,
118
+ false,
119
+ false,
120
+ false,
121
+ false
122
+ ],
123
+ "truncated": [
124
+ true,
125
+ true,
126
+ false,
127
+ false,
128
+ false,
129
+ false,
130
+ true,
131
+ true
132
+ ],
133
+ "gen_tokens": [
134
+ 32768,
135
+ 32768,
136
+ 32152,
137
+ 32267,
138
+ 30978,
139
+ 29198,
140
+ 32768,
141
+ 32768
142
+ ]
143
+ },
144
+ {
145
+ "idx": 3,
146
+ "gold": "385",
147
+ "correct": [
148
+ false,
149
+ false,
150
+ false,
151
+ false,
152
+ false,
153
+ false,
154
+ false,
155
+ false
156
+ ],
157
+ "truncated": [
158
+ true,
159
+ true,
160
+ true,
161
+ true,
162
+ true,
163
+ true,
164
+ true,
165
+ true
166
+ ],
167
+ "gen_tokens": [
168
+ 32768,
169
+ 32768,
170
+ 32768,
171
+ 32768,
172
+ 32768,
173
+ 32768,
174
+ 32768,
175
+ 32768
176
+ ]
177
+ },
178
+ {
179
+ "idx": 4,
180
+ "gold": "110",
181
+ "correct": [
182
+ true,
183
+ true,
184
+ false,
185
+ true,
186
+ false,
187
+ true,
188
+ false,
189
+ true
190
+ ],
191
+ "truncated": [
192
+ false,
193
+ false,
194
+ true,
195
+ false,
196
+ true,
197
+ false,
198
+ false,
199
+ false
200
+ ],
201
+ "gen_tokens": [
202
+ 29067,
203
+ 20593,
204
+ 32768,
205
+ 25655,
206
+ 32768,
207
+ 19688,
208
+ 21025,
209
+ 16835
210
+ ]
211
+ },
212
+ {
213
+ "idx": 5,
214
+ "gold": "104",
215
+ "correct": [
216
+ false,
217
+ true,
218
+ false,
219
+ false,
220
+ true,
221
+ false,
222
+ true,
223
+ false
224
+ ],
225
+ "truncated": [
226
+ false,
227
+ false,
228
+ true,
229
+ true,
230
+ false,
231
+ true,
232
+ false,
233
+ true
234
+ ],
235
+ "gen_tokens": [
236
+ 20475,
237
+ 21519,
238
+ 32768,
239
+ 32768,
240
+ 32242,
241
+ 32768,
242
+ 12816,
243
+ 32768
244
+ ]
245
+ },
246
+ {
247
+ "idx": 6,
248
+ "gold": "721",
249
+ "correct": [
250
+ true,
251
+ true,
252
+ true,
253
+ true,
254
+ true,
255
+ true,
256
+ true,
257
+ true
258
+ ],
259
+ "truncated": [
260
+ false,
261
+ false,
262
+ false,
263
+ false,
264
+ false,
265
+ false,
266
+ false,
267
+ false
268
+ ],
269
+ "gen_tokens": [
270
+ 13135,
271
+ 10520,
272
+ 14191,
273
+ 20470,
274
+ 17455,
275
+ 19188,
276
+ 23239,
277
+ 31835
278
+ ]
279
+ },
280
+ {
281
+ "idx": 7,
282
+ "gold": "025",
283
+ "correct": [
284
+ true,
285
+ true,
286
+ true,
287
+ true,
288
+ true,
289
+ true,
290
+ true,
291
+ true
292
+ ],
293
+ "truncated": [
294
+ false,
295
+ false,
296
+ false,
297
+ false,
298
+ false,
299
+ false,
300
+ false,
301
+ false
302
+ ],
303
+ "gen_tokens": [
304
+ 15423,
305
+ 6682,
306
+ 7739,
307
+ 12095,
308
+ 9590,
309
+ 12370,
310
+ 9975,
311
+ 7581
312
+ ]
313
+ },
314
+ {
315
+ "idx": 8,
316
+ "gold": "809",
317
+ "correct": [
318
+ true,
319
+ true,
320
+ true,
321
+ true,
322
+ true,
323
+ true,
324
+ true,
325
+ true
326
+ ],
327
+ "truncated": [
328
+ false,
329
+ false,
330
+ false,
331
+ false,
332
+ false,
333
+ false,
334
+ false,
335
+ false
336
+ ],
337
+ "gen_tokens": [
338
+ 8974,
339
+ 13126,
340
+ 12952,
341
+ 16095,
342
+ 11811,
343
+ 14263,
344
+ 17397,
345
+ 13556
346
+ ]
347
+ },
348
+ {
349
+ "idx": 9,
350
+ "gold": "116",
351
+ "correct": [
352
+ true,
353
+ true,
354
+ true,
355
+ true,
356
+ true,
357
+ true,
358
+ true,
359
+ true
360
+ ],
361
+ "truncated": [
362
+ false,
363
+ false,
364
+ false,
365
+ false,
366
+ false,
367
+ false,
368
+ false,
369
+ false
370
+ ],
371
+ "gen_tokens": [
372
+ 8042,
373
+ 8629,
374
+ 8274,
375
+ 8696,
376
+ 9915,
377
+ 9335,
378
+ 7384,
379
+ 8647
380
+ ]
381
+ },
382
+ {
383
+ "idx": 10,
384
+ "gold": "104",
385
+ "correct": [
386
+ true,
387
+ true,
388
+ true,
389
+ true,
390
+ false,
391
+ true,
392
+ false,
393
+ false
394
+ ],
395
+ "truncated": [
396
+ false,
397
+ false,
398
+ false,
399
+ false,
400
+ true,
401
+ false,
402
+ true,
403
+ true
404
+ ],
405
+ "gen_tokens": [
406
+ 12855,
407
+ 11745,
408
+ 29288,
409
+ 23041,
410
+ 32768,
411
+ 10680,
412
+ 32768,
413
+ 32768
414
+ ]
415
+ },
416
+ {
417
+ "idx": 11,
418
+ "gold": "294",
419
+ "correct": [
420
+ true,
421
+ true,
422
+ true,
423
+ true,
424
+ true,
425
+ true,
426
+ true,
427
+ true
428
+ ],
429
+ "truncated": [
430
+ false,
431
+ false,
432
+ false,
433
+ false,
434
+ false,
435
+ false,
436
+ false,
437
+ false
438
+ ],
439
+ "gen_tokens": [
440
+ 12255,
441
+ 12460,
442
+ 10687,
443
+ 12768,
444
+ 11746,
445
+ 11805,
446
+ 11687,
447
+ 11356
448
+ ]
449
+ },
450
+ {
451
+ "idx": 12,
452
+ "gold": "540",
453
+ "correct": [
454
+ true,
455
+ true,
456
+ true,
457
+ true,
458
+ true,
459
+ true,
460
+ true,
461
+ true
462
+ ],
463
+ "truncated": [
464
+ false,
465
+ false,
466
+ false,
467
+ false,
468
+ false,
469
+ false,
470
+ false,
471
+ false
472
+ ],
473
+ "gen_tokens": [
474
+ 11068,
475
+ 11453,
476
+ 12629,
477
+ 7328,
478
+ 9126,
479
+ 16668,
480
+ 9325,
481
+ 10263
482
+ ]
483
+ },
484
+ {
485
+ "idx": 13,
486
+ "gold": "197",
487
+ "correct": [
488
+ false,
489
+ false,
490
+ false,
491
+ false,
492
+ false,
493
+ false,
494
+ false,
495
+ false
496
+ ],
497
+ "truncated": [
498
+ false,
499
+ true,
500
+ false,
501
+ false,
502
+ false,
503
+ false,
504
+ false,
505
+ false
506
+ ],
507
+ "gen_tokens": [
508
+ 20498,
509
+ 32768,
510
+ 30262,
511
+ 31206,
512
+ 27354,
513
+ 20487,
514
+ 29478,
515
+ 24112
516
+ ]
517
+ },
518
+ {
519
+ "idx": 14,
520
+ "gold": "480",
521
+ "correct": [
522
+ true,
523
+ true,
524
+ true,
525
+ true,
526
+ true,
527
+ true,
528
+ true,
529
+ false
530
+ ],
531
+ "truncated": [
532
+ false,
533
+ false,
534
+ false,
535
+ false,
536
+ false,
537
+ false,
538
+ false,
539
+ false
540
+ ],
541
+ "gen_tokens": [
542
+ 25691,
543
+ 21823,
544
+ 15204,
545
+ 16461,
546
+ 31263,
547
+ 27169,
548
+ 15274,
549
+ 19360
550
+ ]
551
+ },
552
+ {
553
+ "idx": 15,
554
+ "gold": "073",
555
+ "correct": [
556
+ true,
557
+ true,
558
+ true,
559
+ true,
560
+ false,
561
+ true,
562
+ false,
563
+ false
564
+ ],
565
+ "truncated": [
566
+ false,
567
+ false,
568
+ false,
569
+ false,
570
+ true,
571
+ false,
572
+ false,
573
+ true
574
+ ],
575
+ "gen_tokens": [
576
+ 9199,
577
+ 20464,
578
+ 11633,
579
+ 28321,
580
+ 32768,
581
+ 25488,
582
+ 30159,
583
+ 32768
584
+ ]
585
+ },
586
+ {
587
+ "idx": 16,
588
+ "gold": "468",
589
+ "correct": [
590
+ true,
591
+ true,
592
+ true,
593
+ false,
594
+ true,
595
+ true,
596
+ true,
597
+ true
598
+ ],
599
+ "truncated": [
600
+ false,
601
+ false,
602
+ false,
603
+ false,
604
+ false,
605
+ false,
606
+ false,
607
+ false
608
+ ],
609
+ "gen_tokens": [
610
+ 11540,
611
+ 30995,
612
+ 12387,
613
+ 29532,
614
+ 16504,
615
+ 15596,
616
+ 13938,
617
+ 12377
618
+ ]
619
+ },
620
+ {
621
+ "idx": 17,
622
+ "gold": "601",
623
+ "correct": [
624
+ true,
625
+ false,
626
+ false,
627
+ false,
628
+ false,
629
+ false,
630
+ true,
631
+ true
632
+ ],
633
+ "truncated": [
634
+ false,
635
+ true,
636
+ true,
637
+ true,
638
+ false,
639
+ true,
640
+ false,
641
+ false
642
+ ],
643
+ "gen_tokens": [
644
+ 22409,
645
+ 32768,
646
+ 32768,
647
+ 32768,
648
+ 30322,
649
+ 32768,
650
+ 24095,
651
+ 23676
652
+ ]
653
+ },
654
+ {
655
+ "idx": 18,
656
+ "gold": "023",
657
+ "correct": [
658
+ false,
659
+ true,
660
+ true,
661
+ false,
662
+ true,
663
+ false,
664
+ true,
665
+ true
666
+ ],
667
+ "truncated": [
668
+ true,
669
+ false,
670
+ false,
671
+ false,
672
+ false,
673
+ true,
674
+ false,
675
+ false
676
+ ],
677
+ "gen_tokens": [
678
+ 32768,
679
+ 16927,
680
+ 23729,
681
+ 25068,
682
+ 22820,
683
+ 32768,
684
+ 18415,
685
+ 25413
686
+ ]
687
+ },
688
+ {
689
+ "idx": 19,
690
+ "gold": "321",
691
+ "correct": [
692
+ true,
693
+ true,
694
+ true,
695
+ true,
696
+ true,
697
+ true,
698
+ true,
699
+ true
700
+ ],
701
+ "truncated": [
702
+ false,
703
+ false,
704
+ false,
705
+ false,
706
+ false,
707
+ false,
708
+ false,
709
+ false
710
+ ],
711
+ "gen_tokens": [
712
+ 13497,
713
+ 15707,
714
+ 11702,
715
+ 12854,
716
+ 13906,
717
+ 17959,
718
+ 13279,
719
+ 18084
720
+ ]
721
+ },
722
+ {
723
+ "idx": 20,
724
+ "gold": "211",
725
+ "correct": [
726
+ false,
727
+ false,
728
+ true,
729
+ false,
730
+ true,
731
+ false,
732
+ false,
733
+ false
734
+ ],
735
+ "truncated": [
736
+ true,
737
+ true,
738
+ false,
739
+ true,
740
+ true,
741
+ true,
742
+ true,
743
+ true
744
+ ],
745
+ "gen_tokens": [
746
+ 32768,
747
+ 32768,
748
+ 30340,
749
+ 32768,
750
+ 32768,
751
+ 32768,
752
+ 32768,
753
+ 32768
754
+ ]
755
+ },
756
+ {
757
+ "idx": 21,
758
+ "gold": "315",
759
+ "correct": [
760
+ false,
761
+ false,
762
+ false,
763
+ false,
764
+ false,
765
+ false,
766
+ false,
767
+ false
768
+ ],
769
+ "truncated": [
770
+ true,
771
+ false,
772
+ true,
773
+ true,
774
+ true,
775
+ true,
776
+ true,
777
+ true
778
+ ],
779
+ "gen_tokens": [
780
+ 32768,
781
+ 32605,
782
+ 32768,
783
+ 32768,
784
+ 32768,
785
+ 32768,
786
+ 32768,
787
+ 32768
788
+ ]
789
+ },
790
+ {
791
+ "idx": 22,
792
+ "gold": "236",
793
+ "correct": [
794
+ true,
795
+ false,
796
+ true,
797
+ true,
798
+ true,
799
+ true,
800
+ false,
801
+ true
802
+ ],
803
+ "truncated": [
804
+ false,
805
+ true,
806
+ false,
807
+ false,
808
+ false,
809
+ false,
810
+ true,
811
+ false
812
+ ],
813
+ "gen_tokens": [
814
+ 19150,
815
+ 32768,
816
+ 16801,
817
+ 17602,
818
+ 22551,
819
+ 23325,
820
+ 32768,
821
+ 21981
822
+ ]
823
+ },
824
+ {
825
+ "idx": 23,
826
+ "gold": "045",
827
+ "correct": [
828
+ true,
829
+ true,
830
+ true,
831
+ true,
832
+ true,
833
+ true,
834
+ false,
835
+ true
836
+ ],
837
+ "truncated": [
838
+ false,
839
+ false,
840
+ false,
841
+ false,
842
+ false,
843
+ false,
844
+ true,
845
+ false
846
+ ],
847
+ "gen_tokens": [
848
+ 16401,
849
+ 16570,
850
+ 21555,
851
+ 12283,
852
+ 15641,
853
+ 23611,
854
+ 32768,
855
+ 22799
856
+ ]
857
+ },
858
+ {
859
+ "idx": 24,
860
+ "gold": "033",
861
+ "correct": [
862
+ true,
863
+ true,
864
+ true,
865
+ true,
866
+ true,
867
+ true,
868
+ true,
869
+ true
870
+ ],
871
+ "truncated": [
872
+ false,
873
+ false,
874
+ false,
875
+ false,
876
+ false,
877
+ false,
878
+ false,
879
+ false
880
+ ],
881
+ "gen_tokens": [
882
+ 7228,
883
+ 6508,
884
+ 6269,
885
+ 12335,
886
+ 5647,
887
+ 7404,
888
+ 5959,
889
+ 9479
890
+ ]
891
+ },
892
+ {
893
+ "idx": 25,
894
+ "gold": "080",
895
+ "correct": [
896
+ true,
897
+ false,
898
+ false,
899
+ false,
900
+ false,
901
+ false,
902
+ false,
903
+ false
904
+ ],
905
+ "truncated": [
906
+ false,
907
+ true,
908
+ true,
909
+ true,
910
+ true,
911
+ false,
912
+ true,
913
+ false
914
+ ],
915
+ "gen_tokens": [
916
+ 24883,
917
+ 32768,
918
+ 32768,
919
+ 32768,
920
+ 32768,
921
+ 31707,
922
+ 32768,
923
+ 24763
924
+ ]
925
+ },
926
+ {
927
+ "idx": 26,
928
+ "gold": "055",
929
+ "correct": [
930
+ true,
931
+ true,
932
+ true,
933
+ true,
934
+ true,
935
+ true,
936
+ true,
937
+ true
938
+ ],
939
+ "truncated": [
940
+ false,
941
+ false,
942
+ false,
943
+ false,
944
+ false,
945
+ false,
946
+ false,
947
+ false
948
+ ],
949
+ "gen_tokens": [
950
+ 8012,
951
+ 9557,
952
+ 10938,
953
+ 8044,
954
+ 8832,
955
+ 10170,
956
+ 6247,
957
+ 10423
958
+ ]
959
+ },
960
+ {
961
+ "idx": 27,
962
+ "gold": "699",
963
+ "correct": [
964
+ true,
965
+ false,
966
+ true,
967
+ true,
968
+ false,
969
+ false,
970
+ true,
971
+ false
972
+ ],
973
+ "truncated": [
974
+ false,
975
+ true,
976
+ false,
977
+ false,
978
+ true,
979
+ true,
980
+ false,
981
+ true
982
+ ],
983
+ "gen_tokens": [
984
+ 26419,
985
+ 32768,
986
+ 18406,
987
+ 24700,
988
+ 32768,
989
+ 32768,
990
+ 21747,
991
+ 32768
992
+ ]
993
+ },
994
+ {
995
+ "idx": 28,
996
+ "gold": "127",
997
+ "correct": [
998
+ false,
999
+ false,
1000
+ false,
1001
+ false,
1002
+ false,
1003
+ false,
1004
+ false,
1005
+ false
1006
+ ],
1007
+ "truncated": [
1008
+ true,
1009
+ true,
1010
+ true,
1011
+ true,
1012
+ true,
1013
+ true,
1014
+ true,
1015
+ true
1016
+ ],
1017
+ "gen_tokens": [
1018
+ 32768,
1019
+ 32768,
1020
+ 32768,
1021
+ 32768,
1022
+ 32768,
1023
+ 32768,
1024
+ 32768,
1025
+ 32768
1026
+ ]
1027
+ },
1028
+ {
1029
+ "idx": 29,
1030
+ "gold": "902",
1031
+ "correct": [
1032
+ false,
1033
+ false,
1034
+ false,
1035
+ false,
1036
+ false,
1037
+ false,
1038
+ false,
1039
+ false
1040
+ ],
1041
+ "truncated": [
1042
+ true,
1043
+ false,
1044
+ false,
1045
+ true,
1046
+ false,
1047
+ false,
1048
+ true,
1049
+ false
1050
+ ],
1051
+ "gen_tokens": [
1052
+ 32768,
1053
+ 26642,
1054
+ 29341,
1055
+ 32768,
1056
+ 27938,
1057
+ 30864,
1058
+ 32768,
1059
+ 26774
1060
+ ]
1061
+ }
1062
+ ]
1063
+ }
1064
+ }
1065
+ }
full_eval/reasoning_bench_aime25.generations.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4892d26ec866cf37e1121192d08225cd6dc66b0092426904c6c1b7bbb7ccf4a7
3
+ size 17401784
full_eval/reasoning_bench_aime25.json ADDED
@@ -0,0 +1,1065 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model": "/workspace/mimo/checkpoints/wsd_runs/mimo-wsd-mathmix-h32k-2p0b-v1/snapshots/final",
3
+ "label": "stage3-2.0B-OPD-aime25-think",
4
+ "config": {
5
+ "tasks": "aime25",
6
+ "n": null,
7
+ "max_tokens": 32768,
8
+ "no_thinking": false,
9
+ "temperature": null,
10
+ "top_p": null,
11
+ "top_k": null,
12
+ "ruler_lengths": "4096,8192,16384,32768",
13
+ "ruler_tasks": "niah_single,niah_multikey,vt",
14
+ "ruler_item_tokenizer": null,
15
+ "ruler_items": 25,
16
+ "ruler_gen_tokens": 128,
17
+ "ruler_answer_prefill": false,
18
+ "limit": null,
19
+ "offset": 0,
20
+ "seed": 0,
21
+ "tp": 1,
22
+ "max_model_len": 40960,
23
+ "gpu_mem_util": 0.9,
24
+ "enforce_eager": false,
25
+ "max_num_seqs": 32
26
+ },
27
+ "thinking": true,
28
+ "sampling": {
29
+ "temperature": 0.6,
30
+ "top_p": 0.95,
31
+ "top_k": 20,
32
+ "max_tokens": 32768
33
+ },
34
+ "results": {
35
+ "aime25": {
36
+ "accuracy": 0.4708333333333333,
37
+ "n_problems": 30,
38
+ "n_samples": 8,
39
+ "truncated_frac": 0.38333333333333336,
40
+ "mean_gen_tokens": 24356.570833333335,
41
+ "per_problem": [
42
+ {
43
+ "idx": 0,
44
+ "gold": "70",
45
+ "correct": [
46
+ true,
47
+ true,
48
+ true,
49
+ true,
50
+ true,
51
+ true,
52
+ true,
53
+ true
54
+ ],
55
+ "truncated": [
56
+ false,
57
+ false,
58
+ false,
59
+ false,
60
+ false,
61
+ false,
62
+ false,
63
+ false
64
+ ],
65
+ "gen_tokens": [
66
+ 7478,
67
+ 7978,
68
+ 4478,
69
+ 9938,
70
+ 9092,
71
+ 5809,
72
+ 8761,
73
+ 5969
74
+ ]
75
+ },
76
+ {
77
+ "idx": 1,
78
+ "gold": "588",
79
+ "correct": [
80
+ false,
81
+ true,
82
+ true,
83
+ true,
84
+ true,
85
+ true,
86
+ false,
87
+ true
88
+ ],
89
+ "truncated": [
90
+ true,
91
+ false,
92
+ false,
93
+ false,
94
+ false,
95
+ false,
96
+ true,
97
+ false
98
+ ],
99
+ "gen_tokens": [
100
+ 32768,
101
+ 19849,
102
+ 16331,
103
+ 23864,
104
+ 20712,
105
+ 17420,
106
+ 32768,
107
+ 14248
108
+ ]
109
+ },
110
+ {
111
+ "idx": 2,
112
+ "gold": "16",
113
+ "correct": [
114
+ true,
115
+ true,
116
+ true,
117
+ true,
118
+ false,
119
+ true,
120
+ true,
121
+ true
122
+ ],
123
+ "truncated": [
124
+ false,
125
+ false,
126
+ false,
127
+ false,
128
+ false,
129
+ false,
130
+ false,
131
+ false
132
+ ],
133
+ "gen_tokens": [
134
+ 12792,
135
+ 19906,
136
+ 11292,
137
+ 12554,
138
+ 13328,
139
+ 27446,
140
+ 13223,
141
+ 12278
142
+ ]
143
+ },
144
+ {
145
+ "idx": 3,
146
+ "gold": "117",
147
+ "correct": [
148
+ true,
149
+ true,
150
+ true,
151
+ true,
152
+ true,
153
+ true,
154
+ true,
155
+ true
156
+ ],
157
+ "truncated": [
158
+ false,
159
+ false,
160
+ false,
161
+ false,
162
+ false,
163
+ false,
164
+ false,
165
+ false
166
+ ],
167
+ "gen_tokens": [
168
+ 14221,
169
+ 15613,
170
+ 13304,
171
+ 14656,
172
+ 16979,
173
+ 16526,
174
+ 15348,
175
+ 24066
176
+ ]
177
+ },
178
+ {
179
+ "idx": 4,
180
+ "gold": "279",
181
+ "correct": [
182
+ true,
183
+ true,
184
+ true,
185
+ true,
186
+ true,
187
+ true,
188
+ true,
189
+ false
190
+ ],
191
+ "truncated": [
192
+ false,
193
+ false,
194
+ false,
195
+ false,
196
+ false,
197
+ false,
198
+ false,
199
+ true
200
+ ],
201
+ "gen_tokens": [
202
+ 29079,
203
+ 29608,
204
+ 14642,
205
+ 21861,
206
+ 20669,
207
+ 29331,
208
+ 20097,
209
+ 32768
210
+ ]
211
+ },
212
+ {
213
+ "idx": 5,
214
+ "gold": "504",
215
+ "correct": [
216
+ true,
217
+ true,
218
+ true,
219
+ true,
220
+ true,
221
+ true,
222
+ true,
223
+ true
224
+ ],
225
+ "truncated": [
226
+ false,
227
+ false,
228
+ false,
229
+ false,
230
+ false,
231
+ false,
232
+ false,
233
+ false
234
+ ],
235
+ "gen_tokens": [
236
+ 6155,
237
+ 4249,
238
+ 4613,
239
+ 4116,
240
+ 5565,
241
+ 5476,
242
+ 8531,
243
+ 5774
244
+ ]
245
+ },
246
+ {
247
+ "idx": 6,
248
+ "gold": "821",
249
+ "correct": [
250
+ false,
251
+ false,
252
+ false,
253
+ false,
254
+ false,
255
+ false,
256
+ false,
257
+ false
258
+ ],
259
+ "truncated": [
260
+ false,
261
+ true,
262
+ true,
263
+ true,
264
+ true,
265
+ true,
266
+ false,
267
+ false
268
+ ],
269
+ "gen_tokens": [
270
+ 22220,
271
+ 32768,
272
+ 32768,
273
+ 32768,
274
+ 32768,
275
+ 32768,
276
+ 22241,
277
+ 29196
278
+ ]
279
+ },
280
+ {
281
+ "idx": 7,
282
+ "gold": "77",
283
+ "correct": [
284
+ false,
285
+ false,
286
+ false,
287
+ false,
288
+ false,
289
+ false,
290
+ false,
291
+ false
292
+ ],
293
+ "truncated": [
294
+ false,
295
+ false,
296
+ false,
297
+ false,
298
+ true,
299
+ false,
300
+ false,
301
+ false
302
+ ],
303
+ "gen_tokens": [
304
+ 29125,
305
+ 31294,
306
+ 25318,
307
+ 27828,
308
+ 32768,
309
+ 27951,
310
+ 27672,
311
+ 29340
312
+ ]
313
+ },
314
+ {
315
+ "idx": 8,
316
+ "gold": "62",
317
+ "correct": [
318
+ false,
319
+ false,
320
+ true,
321
+ false,
322
+ true,
323
+ false,
324
+ false,
325
+ false
326
+ ],
327
+ "truncated": [
328
+ true,
329
+ true,
330
+ false,
331
+ true,
332
+ false,
333
+ true,
334
+ true,
335
+ true
336
+ ],
337
+ "gen_tokens": [
338
+ 32768,
339
+ 32768,
340
+ 18565,
341
+ 32768,
342
+ 17592,
343
+ 32768,
344
+ 32768,
345
+ 32768
346
+ ]
347
+ },
348
+ {
349
+ "idx": 9,
350
+ "gold": "81",
351
+ "correct": [
352
+ false,
353
+ false,
354
+ false,
355
+ false,
356
+ false,
357
+ false,
358
+ false,
359
+ false
360
+ ],
361
+ "truncated": [
362
+ false,
363
+ false,
364
+ false,
365
+ true,
366
+ false,
367
+ false,
368
+ false,
369
+ false
370
+ ],
371
+ "gen_tokens": [
372
+ 26248,
373
+ 30994,
374
+ 29570,
375
+ 32768,
376
+ 28818,
377
+ 24970,
378
+ 29802,
379
+ 27175
380
+ ]
381
+ },
382
+ {
383
+ "idx": 10,
384
+ "gold": "259",
385
+ "correct": [
386
+ false,
387
+ false,
388
+ false,
389
+ true,
390
+ false,
391
+ false,
392
+ false,
393
+ false
394
+ ],
395
+ "truncated": [
396
+ true,
397
+ true,
398
+ true,
399
+ false,
400
+ true,
401
+ true,
402
+ true,
403
+ true
404
+ ],
405
+ "gen_tokens": [
406
+ 32768,
407
+ 32768,
408
+ 32768,
409
+ 27680,
410
+ 32768,
411
+ 32768,
412
+ 32768,
413
+ 32768
414
+ ]
415
+ },
416
+ {
417
+ "idx": 11,
418
+ "gold": "510",
419
+ "correct": [
420
+ true,
421
+ true,
422
+ true,
423
+ true,
424
+ false,
425
+ false,
426
+ false,
427
+ false
428
+ ],
429
+ "truncated": [
430
+ false,
431
+ false,
432
+ true,
433
+ false,
434
+ true,
435
+ true,
436
+ true,
437
+ true
438
+ ],
439
+ "gen_tokens": [
440
+ 24557,
441
+ 25043,
442
+ 32768,
443
+ 28568,
444
+ 32768,
445
+ 32768,
446
+ 32768,
447
+ 32768
448
+ ]
449
+ },
450
+ {
451
+ "idx": 12,
452
+ "gold": "204",
453
+ "correct": [
454
+ false,
455
+ false,
456
+ false,
457
+ false,
458
+ false,
459
+ false,
460
+ false,
461
+ false
462
+ ],
463
+ "truncated": [
464
+ true,
465
+ false,
466
+ true,
467
+ true,
468
+ true,
469
+ true,
470
+ true,
471
+ true
472
+ ],
473
+ "gen_tokens": [
474
+ 32768,
475
+ 28650,
476
+ 32768,
477
+ 32768,
478
+ 32768,
479
+ 32768,
480
+ 32768,
481
+ 32768
482
+ ]
483
+ },
484
+ {
485
+ "idx": 13,
486
+ "gold": "60",
487
+ "correct": [
488
+ false,
489
+ false,
490
+ false,
491
+ true,
492
+ false,
493
+ true,
494
+ false,
495
+ false
496
+ ],
497
+ "truncated": [
498
+ false,
499
+ true,
500
+ true,
501
+ true,
502
+ true,
503
+ true,
504
+ false,
505
+ true
506
+ ],
507
+ "gen_tokens": [
508
+ 32148,
509
+ 32768,
510
+ 32768,
511
+ 32768,
512
+ 32768,
513
+ 32768,
514
+ 29510,
515
+ 32768
516
+ ]
517
+ },
518
+ {
519
+ "idx": 14,
520
+ "gold": "735",
521
+ "correct": [
522
+ false,
523
+ false,
524
+ false,
525
+ false,
526
+ false,
527
+ false,
528
+ false,
529
+ false
530
+ ],
531
+ "truncated": [
532
+ true,
533
+ true,
534
+ false,
535
+ false,
536
+ true,
537
+ true,
538
+ true,
539
+ true
540
+ ],
541
+ "gen_tokens": [
542
+ 32768,
543
+ 32768,
544
+ 31815,
545
+ 30971,
546
+ 32768,
547
+ 32768,
548
+ 32768,
549
+ 32768
550
+ ]
551
+ },
552
+ {
553
+ "idx": 15,
554
+ "gold": "468",
555
+ "correct": [
556
+ true,
557
+ true,
558
+ true,
559
+ true,
560
+ true,
561
+ true,
562
+ true,
563
+ true
564
+ ],
565
+ "truncated": [
566
+ false,
567
+ false,
568
+ false,
569
+ false,
570
+ false,
571
+ false,
572
+ false,
573
+ false
574
+ ],
575
+ "gen_tokens": [
576
+ 7655,
577
+ 7730,
578
+ 8094,
579
+ 8197,
580
+ 12580,
581
+ 9355,
582
+ 6257,
583
+ 11207
584
+ ]
585
+ },
586
+ {
587
+ "idx": 16,
588
+ "gold": "49",
589
+ "correct": [
590
+ true,
591
+ true,
592
+ true,
593
+ true,
594
+ true,
595
+ true,
596
+ true,
597
+ true
598
+ ],
599
+ "truncated": [
600
+ false,
601
+ false,
602
+ false,
603
+ false,
604
+ false,
605
+ false,
606
+ false,
607
+ false
608
+ ],
609
+ "gen_tokens": [
610
+ 7843,
611
+ 6010,
612
+ 7368,
613
+ 9676,
614
+ 8938,
615
+ 8927,
616
+ 8027,
617
+ 8258
618
+ ]
619
+ },
620
+ {
621
+ "idx": 17,
622
+ "gold": "82",
623
+ "correct": [
624
+ false,
625
+ false,
626
+ false,
627
+ false,
628
+ true,
629
+ false,
630
+ false,
631
+ true
632
+ ],
633
+ "truncated": [
634
+ true,
635
+ true,
636
+ true,
637
+ false,
638
+ true,
639
+ true,
640
+ true,
641
+ true
642
+ ],
643
+ "gen_tokens": [
644
+ 32768,
645
+ 32768,
646
+ 32768,
647
+ 25332,
648
+ 32768,
649
+ 32768,
650
+ 32768,
651
+ 32768
652
+ ]
653
+ },
654
+ {
655
+ "idx": 18,
656
+ "gold": "106",
657
+ "correct": [
658
+ true,
659
+ true,
660
+ true,
661
+ true,
662
+ true,
663
+ true,
664
+ true,
665
+ true
666
+ ],
667
+ "truncated": [
668
+ false,
669
+ false,
670
+ false,
671
+ false,
672
+ false,
673
+ false,
674
+ false,
675
+ false
676
+ ],
677
+ "gen_tokens": [
678
+ 22122,
679
+ 10627,
680
+ 13617,
681
+ 11034,
682
+ 8075,
683
+ 7992,
684
+ 10870,
685
+ 11504
686
+ ]
687
+ },
688
+ {
689
+ "idx": 19,
690
+ "gold": "336",
691
+ "correct": [
692
+ false,
693
+ false,
694
+ false,
695
+ false,
696
+ false,
697
+ false,
698
+ false,
699
+ false
700
+ ],
701
+ "truncated": [
702
+ true,
703
+ false,
704
+ true,
705
+ false,
706
+ true,
707
+ false,
708
+ true,
709
+ true
710
+ ],
711
+ "gen_tokens": [
712
+ 32768,
713
+ 28701,
714
+ 32768,
715
+ 22840,
716
+ 32768,
717
+ 28152,
718
+ 32768,
719
+ 32768
720
+ ]
721
+ },
722
+ {
723
+ "idx": 20,
724
+ "gold": "293",
725
+ "correct": [
726
+ true,
727
+ true,
728
+ true,
729
+ true,
730
+ false,
731
+ true,
732
+ true,
733
+ true
734
+ ],
735
+ "truncated": [
736
+ false,
737
+ false,
738
+ false,
739
+ false,
740
+ true,
741
+ false,
742
+ false,
743
+ false
744
+ ],
745
+ "gen_tokens": [
746
+ 14824,
747
+ 15828,
748
+ 25268,
749
+ 30751,
750
+ 32768,
751
+ 20503,
752
+ 27700,
753
+ 12316
754
+ ]
755
+ },
756
+ {
757
+ "idx": 21,
758
+ "gold": "237",
759
+ "correct": [
760
+ false,
761
+ false,
762
+ false,
763
+ false,
764
+ false,
765
+ false,
766
+ false,
767
+ true
768
+ ],
769
+ "truncated": [
770
+ false,
771
+ false,
772
+ false,
773
+ false,
774
+ false,
775
+ false,
776
+ false,
777
+ false
778
+ ],
779
+ "gen_tokens": [
780
+ 22745,
781
+ 19232,
782
+ 22283,
783
+ 19125,
784
+ 19903,
785
+ 24782,
786
+ 17122,
787
+ 20915
788
+ ]
789
+ },
790
+ {
791
+ "idx": 22,
792
+ "gold": "610",
793
+ "correct": [
794
+ false,
795
+ true,
796
+ false,
797
+ false,
798
+ false,
799
+ false,
800
+ false,
801
+ true
802
+ ],
803
+ "truncated": [
804
+ true,
805
+ false,
806
+ true,
807
+ true,
808
+ false,
809
+ false,
810
+ false,
811
+ false
812
+ ],
813
+ "gen_tokens": [
814
+ 32768,
815
+ 30425,
816
+ 32768,
817
+ 32768,
818
+ 23801,
819
+ 31188,
820
+ 27706,
821
+ 32565
822
+ ]
823
+ },
824
+ {
825
+ "idx": 23,
826
+ "gold": "149",
827
+ "correct": [
828
+ true,
829
+ false,
830
+ true,
831
+ true,
832
+ true,
833
+ true,
834
+ true,
835
+ true
836
+ ],
837
+ "truncated": [
838
+ false,
839
+ false,
840
+ false,
841
+ false,
842
+ false,
843
+ false,
844
+ false,
845
+ false
846
+ ],
847
+ "gen_tokens": [
848
+ 15294,
849
+ 26139,
850
+ 17376,
851
+ 16672,
852
+ 16039,
853
+ 14525,
854
+ 15077,
855
+ 23237
856
+ ]
857
+ },
858
+ {
859
+ "idx": 24,
860
+ "gold": "907",
861
+ "correct": [
862
+ false,
863
+ false,
864
+ false,
865
+ false,
866
+ true,
867
+ true,
868
+ true,
869
+ false
870
+ ],
871
+ "truncated": [
872
+ true,
873
+ true,
874
+ true,
875
+ true,
876
+ false,
877
+ false,
878
+ false,
879
+ true
880
+ ],
881
+ "gen_tokens": [
882
+ 32768,
883
+ 32768,
884
+ 32768,
885
+ 32768,
886
+ 27744,
887
+ 24558,
888
+ 24444,
889
+ 32768
890
+ ]
891
+ },
892
+ {
893
+ "idx": 25,
894
+ "gold": "113",
895
+ "correct": [
896
+ true,
897
+ false,
898
+ false,
899
+ true,
900
+ true,
901
+ true,
902
+ false,
903
+ true
904
+ ],
905
+ "truncated": [
906
+ false,
907
+ false,
908
+ false,
909
+ true,
910
+ false,
911
+ false,
912
+ false,
913
+ false
914
+ ],
915
+ "gen_tokens": [
916
+ 22434,
917
+ 31945,
918
+ 29014,
919
+ 32768,
920
+ 32232,
921
+ 28927,
922
+ 22947,
923
+ 24442
924
+ ]
925
+ },
926
+ {
927
+ "idx": 26,
928
+ "gold": "19",
929
+ "correct": [
930
+ false,
931
+ true,
932
+ false,
933
+ true,
934
+ true,
935
+ true,
936
+ false,
937
+ true
938
+ ],
939
+ "truncated": [
940
+ true,
941
+ false,
942
+ true,
943
+ false,
944
+ false,
945
+ false,
946
+ true,
947
+ false
948
+ ],
949
+ "gen_tokens": [
950
+ 32768,
951
+ 26856,
952
+ 32768,
953
+ 21684,
954
+ 30355,
955
+ 20789,
956
+ 32768,
957
+ 24600
958
+ ]
959
+ },
960
+ {
961
+ "idx": 27,
962
+ "gold": "248",
963
+ "correct": [
964
+ false,
965
+ false,
966
+ false,
967
+ false,
968
+ false,
969
+ false,
970
+ false,
971
+ false
972
+ ],
973
+ "truncated": [
974
+ true,
975
+ true,
976
+ true,
977
+ true,
978
+ true,
979
+ true,
980
+ true,
981
+ true
982
+ ],
983
+ "gen_tokens": [
984
+ 32768,
985
+ 32768,
986
+ 32768,
987
+ 32768,
988
+ 32768,
989
+ 32768,
990
+ 32768,
991
+ 32768
992
+ ]
993
+ },
994
+ {
995
+ "idx": 28,
996
+ "gold": "104",
997
+ "correct": [
998
+ true,
999
+ false,
1000
+ false,
1001
+ true,
1002
+ false,
1003
+ false,
1004
+ true,
1005
+ true
1006
+ ],
1007
+ "truncated": [
1008
+ false,
1009
+ true,
1010
+ true,
1011
+ false,
1012
+ true,
1013
+ true,
1014
+ false,
1015
+ false
1016
+ ],
1017
+ "gen_tokens": [
1018
+ 21495,
1019
+ 32768,
1020
+ 32768,
1021
+ 31413,
1022
+ 32768,
1023
+ 32768,
1024
+ 23221,
1025
+ 21406
1026
+ ]
1027
+ },
1028
+ {
1029
+ "idx": 29,
1030
+ "gold": "240",
1031
+ "correct": [
1032
+ false,
1033
+ false,
1034
+ false,
1035
+ false,
1036
+ false,
1037
+ false,
1038
+ false,
1039
+ false
1040
+ ],
1041
+ "truncated": [
1042
+ true,
1043
+ true,
1044
+ true,
1045
+ true,
1046
+ true,
1047
+ true,
1048
+ true,
1049
+ true
1050
+ ],
1051
+ "gen_tokens": [
1052
+ 32768,
1053
+ 32768,
1054
+ 32768,
1055
+ 32768,
1056
+ 32768,
1057
+ 32768,
1058
+ 32768,
1059
+ 32768
1060
+ ]
1061
+ }
1062
+ ]
1063
+ }
1064
+ }
1065
+ }
full_eval/reasoning_bench_nothink_math500.generations.json ADDED
The diff for this file is too large to render. See raw diff
 
full_eval/reasoning_bench_nothink_math500.json ADDED
The diff for this file is too large to render. See raw diff
 
full_eval/reasoning_bench_think_math500.generations.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a47ed9ac131970cc704025c06cc3f6799bf18479f24d15151a2790f8c1f0635
3
+ size 11517271
full_eval/reasoning_bench_think_math500.json ADDED
The diff for this file is too large to render. See raw diff
 
full_eval/ruler_seed0.json ADDED
The diff for this file is too large to render. See raw diff
 
full_eval/ruler_seed1.json ADDED
The diff for this file is too large to render. See raw diff
 
full_eval/ruler_seed2.json ADDED
The diff for this file is too large to render. See raw diff
 
full_eval/ruler_seed3.json ADDED
The diff for this file is too large to render. See raw diff
 
full_eval/ruler_seed4.json ADDED
The diff for this file is too large to render. See raw diff
 
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 151643,
4
+ "eos_token_id": 151645,
5
+ "max_new_tokens": 2048,
6
+ "transformers_version": "4.57.3"
7
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f06594767a6412a3e4a22cec284f010c8f6c6e3ad7c7d3b9b6124707cb498590
3
+ size 4923777072
model-00002-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:043e9807a1fbb4f834d134e482d0b3d9e7013aad6b4c2a7f727876a930f253c8
3
+ size 4982186608
model-00003-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:512b16d1d2f9bbbab4225d28e24af8ea1659b4e130c13f7e1c19654d555bad27
3
+ size 4936491488
model-00004-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aaa018f6f002ec5e57b8ec53e9d95452561ecf5c3dc277e09c63e5b14c6905d8
3
+ size 1777391632
model.safetensors.index.json ADDED
@@ -0,0 +1,443 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_parameters": 8309898304,
4
+ "total_size": 16619796608
5
+ },
6
+ "weight_map": {
7
+ "lm_head.weight": "model-00004-of-00004.safetensors",
8
+ "model.embed_tokens.weight": "model-00001-of-00004.safetensors",
9
+ "model.layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors",
10
+ "model.layers.0.linear_attn.A_log": "model-00001-of-00004.safetensors",
11
+ "model.layers.0.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
12
+ "model.layers.0.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
13
+ "model.layers.0.linear_attn.in_proj_ba.weight": "model-00001-of-00004.safetensors",
14
+ "model.layers.0.linear_attn.in_proj_qkvz.weight": "model-00001-of-00004.safetensors",
15
+ "model.layers.0.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
16
+ "model.layers.0.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
17
+ "model.layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
18
+ "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
19
+ "model.layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
20
+ "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
21
+ "model.layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors",
22
+ "model.layers.1.linear_attn.A_log": "model-00001-of-00004.safetensors",
23
+ "model.layers.1.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
24
+ "model.layers.1.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
25
+ "model.layers.1.linear_attn.in_proj_ba.weight": "model-00001-of-00004.safetensors",
26
+ "model.layers.1.linear_attn.in_proj_qkvz.weight": "model-00001-of-00004.safetensors",
27
+ "model.layers.1.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
28
+ "model.layers.1.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
29
+ "model.layers.1.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
30
+ "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
31
+ "model.layers.1.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
32
+ "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
33
+ "model.layers.10.input_layernorm.weight": "model-00002-of-00004.safetensors",
34
+ "model.layers.10.linear_attn.A_log": "model-00002-of-00004.safetensors",
35
+ "model.layers.10.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
36
+ "model.layers.10.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
37
+ "model.layers.10.linear_attn.in_proj_ba.weight": "model-00002-of-00004.safetensors",
38
+ "model.layers.10.linear_attn.in_proj_qkvz.weight": "model-00002-of-00004.safetensors",
39
+ "model.layers.10.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
40
+ "model.layers.10.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
41
+ "model.layers.10.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
42
+ "model.layers.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
43
+ "model.layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
44
+ "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
45
+ "model.layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors",
46
+ "model.layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
47
+ "model.layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
48
+ "model.layers.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
49
+ "model.layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
50
+ "model.layers.11.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
51
+ "model.layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
52
+ "model.layers.11.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
53
+ "model.layers.11.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
54
+ "model.layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
55
+ "model.layers.11.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
56
+ "model.layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
57
+ "model.layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors",
58
+ "model.layers.12.linear_attn.A_log": "model-00002-of-00004.safetensors",
59
+ "model.layers.12.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
60
+ "model.layers.12.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
61
+ "model.layers.12.linear_attn.in_proj_ba.weight": "model-00002-of-00004.safetensors",
62
+ "model.layers.12.linear_attn.in_proj_qkvz.weight": "model-00002-of-00004.safetensors",
63
+ "model.layers.12.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
64
+ "model.layers.12.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
65
+ "model.layers.12.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
66
+ "model.layers.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
67
+ "model.layers.12.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
68
+ "model.layers.12.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
69
+ "model.layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors",
70
+ "model.layers.13.linear_attn.A_log": "model-00002-of-00004.safetensors",
71
+ "model.layers.13.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
72
+ "model.layers.13.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
73
+ "model.layers.13.linear_attn.in_proj_ba.weight": "model-00002-of-00004.safetensors",
74
+ "model.layers.13.linear_attn.in_proj_qkvz.weight": "model-00002-of-00004.safetensors",
75
+ "model.layers.13.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
76
+ "model.layers.13.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
77
+ "model.layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
78
+ "model.layers.13.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
79
+ "model.layers.13.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
80
+ "model.layers.13.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
81
+ "model.layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors",
82
+ "model.layers.14.linear_attn.A_log": "model-00002-of-00004.safetensors",
83
+ "model.layers.14.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
84
+ "model.layers.14.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
85
+ "model.layers.14.linear_attn.in_proj_ba.weight": "model-00002-of-00004.safetensors",
86
+ "model.layers.14.linear_attn.in_proj_qkvz.weight": "model-00002-of-00004.safetensors",
87
+ "model.layers.14.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
88
+ "model.layers.14.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
89
+ "model.layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
90
+ "model.layers.14.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
91
+ "model.layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
92
+ "model.layers.14.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
93
+ "model.layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors",
94
+ "model.layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
95
+ "model.layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
96
+ "model.layers.15.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
97
+ "model.layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
98
+ "model.layers.15.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
99
+ "model.layers.15.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
100
+ "model.layers.15.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
101
+ "model.layers.15.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
102
+ "model.layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
103
+ "model.layers.15.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
104
+ "model.layers.15.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
105
+ "model.layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors",
106
+ "model.layers.16.linear_attn.A_log": "model-00002-of-00004.safetensors",
107
+ "model.layers.16.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
108
+ "model.layers.16.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
109
+ "model.layers.16.linear_attn.in_proj_ba.weight": "model-00002-of-00004.safetensors",
110
+ "model.layers.16.linear_attn.in_proj_qkvz.weight": "model-00002-of-00004.safetensors",
111
+ "model.layers.16.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
112
+ "model.layers.16.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
113
+ "model.layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
114
+ "model.layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
115
+ "model.layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
116
+ "model.layers.16.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
117
+ "model.layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors",
118
+ "model.layers.17.linear_attn.A_log": "model-00002-of-00004.safetensors",
119
+ "model.layers.17.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
120
+ "model.layers.17.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
121
+ "model.layers.17.linear_attn.in_proj_ba.weight": "model-00002-of-00004.safetensors",
122
+ "model.layers.17.linear_attn.in_proj_qkvz.weight": "model-00002-of-00004.safetensors",
123
+ "model.layers.17.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
124
+ "model.layers.17.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
125
+ "model.layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
126
+ "model.layers.17.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
127
+ "model.layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
128
+ "model.layers.17.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
129
+ "model.layers.18.input_layernorm.weight": "model-00002-of-00004.safetensors",
130
+ "model.layers.18.linear_attn.A_log": "model-00002-of-00004.safetensors",
131
+ "model.layers.18.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
132
+ "model.layers.18.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
133
+ "model.layers.18.linear_attn.in_proj_ba.weight": "model-00002-of-00004.safetensors",
134
+ "model.layers.18.linear_attn.in_proj_qkvz.weight": "model-00002-of-00004.safetensors",
135
+ "model.layers.18.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
136
+ "model.layers.18.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
137
+ "model.layers.18.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
138
+ "model.layers.18.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
139
+ "model.layers.18.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
140
+ "model.layers.18.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
141
+ "model.layers.19.input_layernorm.weight": "model-00002-of-00004.safetensors",
142
+ "model.layers.19.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
143
+ "model.layers.19.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
144
+ "model.layers.19.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
145
+ "model.layers.19.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
146
+ "model.layers.19.self_attn.k_proj.bias": "model-00002-of-00004.safetensors",
147
+ "model.layers.19.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
148
+ "model.layers.19.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
149
+ "model.layers.19.self_attn.q_proj.bias": "model-00002-of-00004.safetensors",
150
+ "model.layers.19.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
151
+ "model.layers.19.self_attn.v_proj.bias": "model-00002-of-00004.safetensors",
152
+ "model.layers.19.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
153
+ "model.layers.2.input_layernorm.weight": "model-00001-of-00004.safetensors",
154
+ "model.layers.2.linear_attn.A_log": "model-00001-of-00004.safetensors",
155
+ "model.layers.2.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
156
+ "model.layers.2.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
157
+ "model.layers.2.linear_attn.in_proj_ba.weight": "model-00001-of-00004.safetensors",
158
+ "model.layers.2.linear_attn.in_proj_qkvz.weight": "model-00001-of-00004.safetensors",
159
+ "model.layers.2.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
160
+ "model.layers.2.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
161
+ "model.layers.2.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
162
+ "model.layers.2.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
163
+ "model.layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
164
+ "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
165
+ "model.layers.20.input_layernorm.weight": "model-00002-of-00004.safetensors",
166
+ "model.layers.20.linear_attn.A_log": "model-00002-of-00004.safetensors",
167
+ "model.layers.20.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
168
+ "model.layers.20.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
169
+ "model.layers.20.linear_attn.in_proj_ba.weight": "model-00002-of-00004.safetensors",
170
+ "model.layers.20.linear_attn.in_proj_qkvz.weight": "model-00002-of-00004.safetensors",
171
+ "model.layers.20.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
172
+ "model.layers.20.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
173
+ "model.layers.20.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
174
+ "model.layers.20.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
175
+ "model.layers.20.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
176
+ "model.layers.20.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
177
+ "model.layers.21.input_layernorm.weight": "model-00002-of-00004.safetensors",
178
+ "model.layers.21.linear_attn.A_log": "model-00002-of-00004.safetensors",
179
+ "model.layers.21.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
180
+ "model.layers.21.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
181
+ "model.layers.21.linear_attn.in_proj_ba.weight": "model-00002-of-00004.safetensors",
182
+ "model.layers.21.linear_attn.in_proj_qkvz.weight": "model-00002-of-00004.safetensors",
183
+ "model.layers.21.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
184
+ "model.layers.21.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
185
+ "model.layers.21.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
186
+ "model.layers.21.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
187
+ "model.layers.21.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
188
+ "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
189
+ "model.layers.22.input_layernorm.weight": "model-00003-of-00004.safetensors",
190
+ "model.layers.22.linear_attn.A_log": "model-00002-of-00004.safetensors",
191
+ "model.layers.22.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
192
+ "model.layers.22.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
193
+ "model.layers.22.linear_attn.in_proj_ba.weight": "model-00003-of-00004.safetensors",
194
+ "model.layers.22.linear_attn.in_proj_qkvz.weight": "model-00003-of-00004.safetensors",
195
+ "model.layers.22.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
196
+ "model.layers.22.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
197
+ "model.layers.22.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
198
+ "model.layers.22.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
199
+ "model.layers.22.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
200
+ "model.layers.22.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
201
+ "model.layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors",
202
+ "model.layers.23.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
203
+ "model.layers.23.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
204
+ "model.layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
205
+ "model.layers.23.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
206
+ "model.layers.23.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
207
+ "model.layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
208
+ "model.layers.23.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
209
+ "model.layers.23.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
210
+ "model.layers.23.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
211
+ "model.layers.23.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
212
+ "model.layers.23.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
213
+ "model.layers.24.input_layernorm.weight": "model-00003-of-00004.safetensors",
214
+ "model.layers.24.linear_attn.A_log": "model-00003-of-00004.safetensors",
215
+ "model.layers.24.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
216
+ "model.layers.24.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
217
+ "model.layers.24.linear_attn.in_proj_ba.weight": "model-00003-of-00004.safetensors",
218
+ "model.layers.24.linear_attn.in_proj_qkvz.weight": "model-00003-of-00004.safetensors",
219
+ "model.layers.24.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
220
+ "model.layers.24.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
221
+ "model.layers.24.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
222
+ "model.layers.24.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
223
+ "model.layers.24.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
224
+ "model.layers.24.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
225
+ "model.layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors",
226
+ "model.layers.25.linear_attn.A_log": "model-00003-of-00004.safetensors",
227
+ "model.layers.25.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
228
+ "model.layers.25.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
229
+ "model.layers.25.linear_attn.in_proj_ba.weight": "model-00003-of-00004.safetensors",
230
+ "model.layers.25.linear_attn.in_proj_qkvz.weight": "model-00003-of-00004.safetensors",
231
+ "model.layers.25.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
232
+ "model.layers.25.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
233
+ "model.layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
234
+ "model.layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
235
+ "model.layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
236
+ "model.layers.25.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
237
+ "model.layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors",
238
+ "model.layers.26.linear_attn.A_log": "model-00003-of-00004.safetensors",
239
+ "model.layers.26.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
240
+ "model.layers.26.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
241
+ "model.layers.26.linear_attn.in_proj_ba.weight": "model-00003-of-00004.safetensors",
242
+ "model.layers.26.linear_attn.in_proj_qkvz.weight": "model-00003-of-00004.safetensors",
243
+ "model.layers.26.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
244
+ "model.layers.26.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
245
+ "model.layers.26.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
246
+ "model.layers.26.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
247
+ "model.layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
248
+ "model.layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
249
+ "model.layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors",
250
+ "model.layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
251
+ "model.layers.27.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
252
+ "model.layers.27.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
253
+ "model.layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
254
+ "model.layers.27.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
255
+ "model.layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
256
+ "model.layers.27.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
257
+ "model.layers.27.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
258
+ "model.layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
259
+ "model.layers.27.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
260
+ "model.layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
261
+ "model.layers.28.input_layernorm.weight": "model-00003-of-00004.safetensors",
262
+ "model.layers.28.linear_attn.A_log": "model-00003-of-00004.safetensors",
263
+ "model.layers.28.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
264
+ "model.layers.28.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
265
+ "model.layers.28.linear_attn.in_proj_ba.weight": "model-00003-of-00004.safetensors",
266
+ "model.layers.28.linear_attn.in_proj_qkvz.weight": "model-00003-of-00004.safetensors",
267
+ "model.layers.28.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
268
+ "model.layers.28.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
269
+ "model.layers.28.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
270
+ "model.layers.28.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
271
+ "model.layers.28.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
272
+ "model.layers.28.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
273
+ "model.layers.29.input_layernorm.weight": "model-00003-of-00004.safetensors",
274
+ "model.layers.29.linear_attn.A_log": "model-00003-of-00004.safetensors",
275
+ "model.layers.29.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
276
+ "model.layers.29.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
277
+ "model.layers.29.linear_attn.in_proj_ba.weight": "model-00003-of-00004.safetensors",
278
+ "model.layers.29.linear_attn.in_proj_qkvz.weight": "model-00003-of-00004.safetensors",
279
+ "model.layers.29.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
280
+ "model.layers.29.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
281
+ "model.layers.29.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
282
+ "model.layers.29.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
283
+ "model.layers.29.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
284
+ "model.layers.29.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
285
+ "model.layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors",
286
+ "model.layers.3.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
287
+ "model.layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
288
+ "model.layers.3.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
289
+ "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
290
+ "model.layers.3.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
291
+ "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
292
+ "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
293
+ "model.layers.3.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
294
+ "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
295
+ "model.layers.3.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
296
+ "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
297
+ "model.layers.30.input_layernorm.weight": "model-00003-of-00004.safetensors",
298
+ "model.layers.30.linear_attn.A_log": "model-00003-of-00004.safetensors",
299
+ "model.layers.30.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
300
+ "model.layers.30.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
301
+ "model.layers.30.linear_attn.in_proj_ba.weight": "model-00003-of-00004.safetensors",
302
+ "model.layers.30.linear_attn.in_proj_qkvz.weight": "model-00003-of-00004.safetensors",
303
+ "model.layers.30.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
304
+ "model.layers.30.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
305
+ "model.layers.30.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
306
+ "model.layers.30.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
307
+ "model.layers.30.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
308
+ "model.layers.30.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
309
+ "model.layers.31.input_layernorm.weight": "model-00003-of-00004.safetensors",
310
+ "model.layers.31.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
311
+ "model.layers.31.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
312
+ "model.layers.31.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
313
+ "model.layers.31.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
314
+ "model.layers.31.self_attn.k_proj.bias": "model-00003-of-00004.safetensors",
315
+ "model.layers.31.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
316
+ "model.layers.31.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
317
+ "model.layers.31.self_attn.q_proj.bias": "model-00003-of-00004.safetensors",
318
+ "model.layers.31.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
319
+ "model.layers.31.self_attn.v_proj.bias": "model-00003-of-00004.safetensors",
320
+ "model.layers.31.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
321
+ "model.layers.32.input_layernorm.weight": "model-00003-of-00004.safetensors",
322
+ "model.layers.32.linear_attn.A_log": "model-00003-of-00004.safetensors",
323
+ "model.layers.32.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
324
+ "model.layers.32.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
325
+ "model.layers.32.linear_attn.in_proj_ba.weight": "model-00003-of-00004.safetensors",
326
+ "model.layers.32.linear_attn.in_proj_qkvz.weight": "model-00003-of-00004.safetensors",
327
+ "model.layers.32.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
328
+ "model.layers.32.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
329
+ "model.layers.32.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
330
+ "model.layers.32.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
331
+ "model.layers.32.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
332
+ "model.layers.32.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
333
+ "model.layers.33.input_layernorm.weight": "model-00003-of-00004.safetensors",
334
+ "model.layers.33.linear_attn.A_log": "model-00003-of-00004.safetensors",
335
+ "model.layers.33.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
336
+ "model.layers.33.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
337
+ "model.layers.33.linear_attn.in_proj_ba.weight": "model-00003-of-00004.safetensors",
338
+ "model.layers.33.linear_attn.in_proj_qkvz.weight": "model-00003-of-00004.safetensors",
339
+ "model.layers.33.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
340
+ "model.layers.33.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
341
+ "model.layers.33.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
342
+ "model.layers.33.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
343
+ "model.layers.33.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
344
+ "model.layers.33.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
345
+ "model.layers.34.input_layernorm.weight": "model-00004-of-00004.safetensors",
346
+ "model.layers.34.linear_attn.A_log": "model-00003-of-00004.safetensors",
347
+ "model.layers.34.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
348
+ "model.layers.34.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
349
+ "model.layers.34.linear_attn.in_proj_ba.weight": "model-00003-of-00004.safetensors",
350
+ "model.layers.34.linear_attn.in_proj_qkvz.weight": "model-00003-of-00004.safetensors",
351
+ "model.layers.34.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
352
+ "model.layers.34.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
353
+ "model.layers.34.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
354
+ "model.layers.34.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
355
+ "model.layers.34.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
356
+ "model.layers.34.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
357
+ "model.layers.35.input_layernorm.weight": "model-00004-of-00004.safetensors",
358
+ "model.layers.35.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
359
+ "model.layers.35.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
360
+ "model.layers.35.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
361
+ "model.layers.35.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
362
+ "model.layers.35.self_attn.k_proj.bias": "model-00004-of-00004.safetensors",
363
+ "model.layers.35.self_attn.k_proj.weight": "model-00004-of-00004.safetensors",
364
+ "model.layers.35.self_attn.o_proj.weight": "model-00004-of-00004.safetensors",
365
+ "model.layers.35.self_attn.q_proj.bias": "model-00004-of-00004.safetensors",
366
+ "model.layers.35.self_attn.q_proj.weight": "model-00004-of-00004.safetensors",
367
+ "model.layers.35.self_attn.v_proj.bias": "model-00004-of-00004.safetensors",
368
+ "model.layers.35.self_attn.v_proj.weight": "model-00004-of-00004.safetensors",
369
+ "model.layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors",
370
+ "model.layers.4.linear_attn.A_log": "model-00001-of-00004.safetensors",
371
+ "model.layers.4.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
372
+ "model.layers.4.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
373
+ "model.layers.4.linear_attn.in_proj_ba.weight": "model-00001-of-00004.safetensors",
374
+ "model.layers.4.linear_attn.in_proj_qkvz.weight": "model-00001-of-00004.safetensors",
375
+ "model.layers.4.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
376
+ "model.layers.4.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
377
+ "model.layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
378
+ "model.layers.4.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
379
+ "model.layers.4.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
380
+ "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
381
+ "model.layers.5.input_layernorm.weight": "model-00001-of-00004.safetensors",
382
+ "model.layers.5.linear_attn.A_log": "model-00001-of-00004.safetensors",
383
+ "model.layers.5.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
384
+ "model.layers.5.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
385
+ "model.layers.5.linear_attn.in_proj_ba.weight": "model-00001-of-00004.safetensors",
386
+ "model.layers.5.linear_attn.in_proj_qkvz.weight": "model-00001-of-00004.safetensors",
387
+ "model.layers.5.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
388
+ "model.layers.5.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
389
+ "model.layers.5.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
390
+ "model.layers.5.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
391
+ "model.layers.5.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
392
+ "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
393
+ "model.layers.6.input_layernorm.weight": "model-00001-of-00004.safetensors",
394
+ "model.layers.6.linear_attn.A_log": "model-00001-of-00004.safetensors",
395
+ "model.layers.6.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
396
+ "model.layers.6.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
397
+ "model.layers.6.linear_attn.in_proj_ba.weight": "model-00001-of-00004.safetensors",
398
+ "model.layers.6.linear_attn.in_proj_qkvz.weight": "model-00001-of-00004.safetensors",
399
+ "model.layers.6.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
400
+ "model.layers.6.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
401
+ "model.layers.6.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
402
+ "model.layers.6.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
403
+ "model.layers.6.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
404
+ "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
405
+ "model.layers.7.input_layernorm.weight": "model-00001-of-00004.safetensors",
406
+ "model.layers.7.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
407
+ "model.layers.7.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
408
+ "model.layers.7.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
409
+ "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
410
+ "model.layers.7.self_attn.k_proj.bias": "model-00001-of-00004.safetensors",
411
+ "model.layers.7.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
412
+ "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
413
+ "model.layers.7.self_attn.q_proj.bias": "model-00001-of-00004.safetensors",
414
+ "model.layers.7.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
415
+ "model.layers.7.self_attn.v_proj.bias": "model-00001-of-00004.safetensors",
416
+ "model.layers.7.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
417
+ "model.layers.8.input_layernorm.weight": "model-00001-of-00004.safetensors",
418
+ "model.layers.8.linear_attn.A_log": "model-00001-of-00004.safetensors",
419
+ "model.layers.8.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
420
+ "model.layers.8.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
421
+ "model.layers.8.linear_attn.in_proj_ba.weight": "model-00001-of-00004.safetensors",
422
+ "model.layers.8.linear_attn.in_proj_qkvz.weight": "model-00001-of-00004.safetensors",
423
+ "model.layers.8.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
424
+ "model.layers.8.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
425
+ "model.layers.8.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
426
+ "model.layers.8.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
427
+ "model.layers.8.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
428
+ "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
429
+ "model.layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors",
430
+ "model.layers.9.linear_attn.A_log": "model-00001-of-00004.safetensors",
431
+ "model.layers.9.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
432
+ "model.layers.9.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
433
+ "model.layers.9.linear_attn.in_proj_ba.weight": "model-00001-of-00004.safetensors",
434
+ "model.layers.9.linear_attn.in_proj_qkvz.weight": "model-00001-of-00004.safetensors",
435
+ "model.layers.9.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
436
+ "model.layers.9.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
437
+ "model.layers.9.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
438
+ "model.layers.9.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
439
+ "model.layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
440
+ "model.layers.9.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
441
+ "model.norm.weight": "model-00004-of-00004.safetensors"
442
+ }
443
+ }
snapshot_meta.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "final",
3
+ "k": 0,
4
+ "mark_tokens": 2000000000,
5
+ "consumed_tokens": 2000126188,
6
+ "step": 12887,
7
+ "lr_gdn": 1.1102230246251566e-20,
8
+ "lr_inherited": 1.1102230246251566e-20,
9
+ "horizon": 32768,
10
+ "wandb_id": "mimo-mathmix-h32k-runpod-2p0b-v1-20260905",
11
+ "full_state": false,
12
+ "saved": "2026-09-06 19:24:40"
13
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>",
5
+ "<|object_ref_start|>",
6
+ "<|object_ref_end|>",
7
+ "<|box_start|>",
8
+ "<|box_end|>",
9
+ "<|quad_start|>",
10
+ "<|quad_end|>",
11
+ "<|vision_start|>",
12
+ "<|vision_end|>",
13
+ "<|vision_pad|>",
14
+ "<|image_pad|>",
15
+ "<|video_pad|>"
16
+ ],
17
+ "eos_token": {
18
+ "content": "<|im_end|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ "pad_token": {
25
+ "content": "<|endoftext|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ }
31
+ }
surgery_report.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "total_params": 8309898304,
3
+ "inherited_params": 8301930496,
4
+ "inherited_fraction": 0.9990411666053525,
5
+ "dropped_teacher_bias_params": 165888,
6
+ "teacher": "/ocean/projects/cis250011p/araje1/cache/hf_cache/hub/models--XiaomiMiMo--MiMo-7B-RL-0530/snapshots/323400599af3903adc2a536d6340a23fee88d2e0",
7
+ "teacher_family": "mimo",
8
+ "teacher_unexpected_keys_ignored": 16,
9
+ "teacher_params": 7622619136,
10
+ "full_attention_layers": [
11
+ 3,
12
+ 7,
13
+ 11,
14
+ 15,
15
+ 19,
16
+ 23,
17
+ 27,
18
+ 31,
19
+ 35
20
+ ],
21
+ "surgery": {
22
+ "full_attention_layers": null,
23
+ "interval": 4,
24
+ "linear_num_key_heads": 16,
25
+ "linear_num_value_heads": 32,
26
+ "linear_key_head_dim": 128,
27
+ "linear_value_head_dim": 128,
28
+ "linear_conv_kernel_dim": 4,
29
+ "conv_init": "default"
30
+ }
31
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
3
+ size 11421896
tokenizer_config.json ADDED
@@ -0,0 +1,207 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_prefix_space": false,
4
+ "added_tokens_decoder": {
5
+ "151643": {
6
+ "content": "<|endoftext|>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "151644": {
14
+ "content": "<|im_start|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "151645": {
22
+ "content": "<|im_end|>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "151646": {
30
+ "content": "<|object_ref_start|>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "151647": {
38
+ "content": "<|object_ref_end|>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": true
44
+ },
45
+ "151648": {
46
+ "content": "<|box_start|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false,
51
+ "special": true
52
+ },
53
+ "151649": {
54
+ "content": "<|box_end|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false,
59
+ "special": true
60
+ },
61
+ "151650": {
62
+ "content": "<|quad_start|>",
63
+ "lstrip": false,
64
+ "normalized": false,
65
+ "rstrip": false,
66
+ "single_word": false,
67
+ "special": true
68
+ },
69
+ "151651": {
70
+ "content": "<|quad_end|>",
71
+ "lstrip": false,
72
+ "normalized": false,
73
+ "rstrip": false,
74
+ "single_word": false,
75
+ "special": true
76
+ },
77
+ "151652": {
78
+ "content": "<|vision_start|>",
79
+ "lstrip": false,
80
+ "normalized": false,
81
+ "rstrip": false,
82
+ "single_word": false,
83
+ "special": true
84
+ },
85
+ "151653": {
86
+ "content": "<|vision_end|>",
87
+ "lstrip": false,
88
+ "normalized": false,
89
+ "rstrip": false,
90
+ "single_word": false,
91
+ "special": true
92
+ },
93
+ "151654": {
94
+ "content": "<|vision_pad|>",
95
+ "lstrip": false,
96
+ "normalized": false,
97
+ "rstrip": false,
98
+ "single_word": false,
99
+ "special": true
100
+ },
101
+ "151655": {
102
+ "content": "<|image_pad|>",
103
+ "lstrip": false,
104
+ "normalized": false,
105
+ "rstrip": false,
106
+ "single_word": false,
107
+ "special": true
108
+ },
109
+ "151656": {
110
+ "content": "<|video_pad|>",
111
+ "lstrip": false,
112
+ "normalized": false,
113
+ "rstrip": false,
114
+ "single_word": false,
115
+ "special": true
116
+ },
117
+ "151657": {
118
+ "content": "<tool_call>",
119
+ "lstrip": false,
120
+ "normalized": false,
121
+ "rstrip": false,
122
+ "single_word": false,
123
+ "special": false
124
+ },
125
+ "151658": {
126
+ "content": "</tool_call>",
127
+ "lstrip": false,
128
+ "normalized": false,
129
+ "rstrip": false,
130
+ "single_word": false,
131
+ "special": false
132
+ },
133
+ "151659": {
134
+ "content": "<|fim_prefix|>",
135
+ "lstrip": false,
136
+ "normalized": false,
137
+ "rstrip": false,
138
+ "single_word": false,
139
+ "special": false
140
+ },
141
+ "151660": {
142
+ "content": "<|fim_middle|>",
143
+ "lstrip": false,
144
+ "normalized": false,
145
+ "rstrip": false,
146
+ "single_word": false,
147
+ "special": false
148
+ },
149
+ "151661": {
150
+ "content": "<|fim_suffix|>",
151
+ "lstrip": false,
152
+ "normalized": false,
153
+ "rstrip": false,
154
+ "single_word": false,
155
+ "special": false
156
+ },
157
+ "151662": {
158
+ "content": "<|fim_pad|>",
159
+ "lstrip": false,
160
+ "normalized": false,
161
+ "rstrip": false,
162
+ "single_word": false,
163
+ "special": false
164
+ },
165
+ "151663": {
166
+ "content": "<|repo_name|>",
167
+ "lstrip": false,
168
+ "normalized": false,
169
+ "rstrip": false,
170
+ "single_word": false,
171
+ "special": false
172
+ },
173
+ "151664": {
174
+ "content": "<|file_sep|>",
175
+ "lstrip": false,
176
+ "normalized": false,
177
+ "rstrip": false,
178
+ "single_word": false,
179
+ "special": false
180
+ }
181
+ },
182
+ "additional_special_tokens": [
183
+ "<|im_start|>",
184
+ "<|im_end|>",
185
+ "<|object_ref_start|>",
186
+ "<|object_ref_end|>",
187
+ "<|box_start|>",
188
+ "<|box_end|>",
189
+ "<|quad_start|>",
190
+ "<|quad_end|>",
191
+ "<|vision_start|>",
192
+ "<|vision_end|>",
193
+ "<|vision_pad|>",
194
+ "<|image_pad|>",
195
+ "<|video_pad|>"
196
+ ],
197
+ "bos_token": null,
198
+ "clean_up_tokenization_spaces": false,
199
+ "eos_token": "<|im_end|>",
200
+ "errors": "replace",
201
+ "extra_special_tokens": {},
202
+ "model_max_length": 131072,
203
+ "pad_token": "<|endoftext|>",
204
+ "split_special_tokens": false,
205
+ "tokenizer_class": "Qwen2Tokenizer",
206
+ "unk_token": null
207
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff