Antonio0101 commited on
Commit
7d25f6e
·
verified ·
1 Parent(s): 9d5a11a

Thank you first commit

Browse files
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
chat_template.jinja ADDED
@@ -0,0 +1,112 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- macro render_content(msg) -%}
2
+ {%- set c = msg.get('content') -%}
3
+ {%- if c is string -%}
4
+ {{ c }}
5
+ {%- elif c is not none -%}
6
+ {% for content in c -%}
7
+ {% if content['type'] == 'image' or content['type'] == 'image_url' -%}
8
+ <|media_begin|>image<|media_content|><|media_pad|><|media_end|>
9
+ {% elif content['type'] == 'video' or content['type']== 'video_url'-%}
10
+ <|kimi_k25_video_placeholder|>
11
+ {% else -%}
12
+ {{ content['text'] }}
13
+ {%- endif -%}
14
+ {%- endfor -%}
15
+ {%- endif -%}
16
+ {%- endmacro -%}
17
+
18
+ {% macro set_roles(message) -%}
19
+ {%- set role_name = message.get('name') or message['role'] -%}
20
+ {%- if message['role'] == 'user' -%}
21
+ <|im_user|>{{role_name}}<|im_middle|>
22
+ {%- elif message['role'] == 'assistant' -%}
23
+ <|im_assistant|>{{role_name}}<|im_middle|>
24
+ {%- else -%}
25
+ <|im_system|>{{role_name}}<|im_middle|>
26
+ {%- endif -%}
27
+ {%- endmacro -%}
28
+
29
+
30
+ {%- macro render_toolcalls(message) -%}
31
+ <|tool_calls_section_begin|>
32
+ {%- for tool_call in message['tool_calls'] -%}
33
+ {%- set formatted_id = tool_call['id'] -%}
34
+ <|tool_call_begin|>{{ formatted_id }}<|tool_call_argument_begin|>{% if tool_call['function']['arguments'] is string %}{{ tool_call['function']['arguments'] }}{% else %}{{ tool_call['function']['arguments'] | tojson }}{% endif %}<|tool_call_end|>
35
+ {%- endfor -%}
36
+ <|tool_calls_section_end|>
37
+ {%- endmacro -%}
38
+
39
+
40
+ {%- set preserve_thinking = preserve_thinking | default(false) -%}
41
+ {# Find last non-tool-call assistant message. If preserve_thinking, keep -1 so hist is empty and all msgs use suffix (retain reasoning). #}
42
+ {%- set ns = namespace(last_non_tool_call_assistant_msg=-1) -%}
43
+ {%- if not preserve_thinking -%}
44
+ {%- for idx in range(messages|length-1, -1, -1) -%}
45
+ {%- if messages[idx]['role'] == 'assistant' and not messages[idx].get('tool_calls') -%}
46
+ {%- set ns.last_non_tool_call_assistant_msg = idx -%}
47
+ {%- break -%}
48
+ {%- endif -%}
49
+ {%- endfor -%}
50
+ {%- endif -%}
51
+
52
+ {# split all messages into history & suffix, reasoning_content in suffix should be reserved.#}
53
+ {%- set hist_msgs = messages[:ns.last_non_tool_call_assistant_msg+1] -%}
54
+ {%- set suffix_msgs = messages[ns.last_non_tool_call_assistant_msg+1:] -%}
55
+
56
+ {%- if tools -%}
57
+ {%- if tools_ts_str -%}
58
+ <|im_system|>tool_declare<|im_middle|>{{ tools_ts_str }}<|im_end|>
59
+ {%- else -%}
60
+ <|im_system|>tool_declare<|im_middle|>{{ tools | tojson(separators=(',', ':')) }}<|im_end|>
61
+ {%- endif -%}
62
+ {%- endif -%}
63
+
64
+
65
+ {%- for message in hist_msgs -%}
66
+ {{set_roles(message)}}
67
+ {%- if message['role'] == 'assistant' -%}
68
+ <think></think>{{render_content(message)}}
69
+ {%- if message.get('tool_calls') -%}
70
+ {{render_toolcalls(message)}}
71
+ {%- endif -%}
72
+ {%- elif message['role'] == 'tool' -%}
73
+ {%- set tool_call_id = message.tool_call_id -%}
74
+ ## Return of {{ tool_call_id }}
75
+ {{render_content(message)}}
76
+ {%- elif message['content'] is not none -%}
77
+ {{render_content(message)}}
78
+ {%- endif -%}
79
+ <|im_end|>
80
+ {%- endfor -%}
81
+
82
+ {%- for message in suffix_msgs -%}
83
+ {{set_roles(message)}}
84
+ {%- if message['role'] == 'assistant' -%}
85
+ {%- if thinking is defined and thinking is false and preserve_thinking is false -%}
86
+ <think></think>{{render_content(message)}}
87
+ {%- else -%}
88
+ {%- set rc = message.get('reasoning', message.get('reasoning_content', '')) -%}
89
+ <think>{{rc}}</think>{{render_content(message)}}
90
+ {%- endif -%}
91
+ {%- if message.get('tool_calls') -%}
92
+ {{render_toolcalls(message)}}
93
+ {%- endif -%}
94
+ {%- elif message['role'] == 'tool' -%}
95
+ {%- set tool_call_id = message.tool_call_id -%}
96
+ ## Return of {{ tool_call_id }}
97
+ {{render_content(message)}}
98
+ {%- elif message['content'] is not none -%}
99
+ {{render_content(message)}}
100
+ {%- endif -%}
101
+ <|im_end|>
102
+ {%- endfor -%}
103
+
104
+
105
+ {%- if add_generation_prompt -%}
106
+ <|im_assistant|>assistant<|im_middle|>
107
+ {%- if thinking is defined and thinking is false -%}
108
+ <think></think>
109
+ {%- else -%}
110
+ <think>
111
+ {%- endif -%}
112
+ {%- endif -%}
config.json ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DeepseekV3ForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "aux_loss_alpha": 0.001,
8
+ "bos_token_id": 163584,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 163585,
11
+ "ep_size": 1,
12
+ "first_k_dense_replace": 1,
13
+ "head_dim": 64,
14
+ "hidden_act": "silu",
15
+ "hidden_size": 2048,
16
+ "initializer_range": 0.02,
17
+ "intermediate_size": 11264,
18
+ "kv_lora_rank": 512,
19
+ "max_position_embeddings": 131072,
20
+ "model_type": "deepseek_v3",
21
+ "moe_intermediate_size": 1408,
22
+ "moe_layer_freq": 1,
23
+ "n_group": 1,
24
+ "n_routed_experts": 64,
25
+ "n_shared_experts": 2,
26
+ "norm_topk_prob": true,
27
+ "num_attention_heads": 16,
28
+ "num_experts_per_tok": 6,
29
+ "num_hidden_layers": 27,
30
+ "num_key_value_heads": 16,
31
+ "num_nextn_predict_layers": 1,
32
+ "pad_token_id": 163839,
33
+ "pretraining_tp": 1,
34
+ "q_lora_rank": null,
35
+ "qk_head_dim": 192,
36
+ "qk_nope_head_dim": 128,
37
+ "qk_rope_head_dim": 64,
38
+ "rms_norm_eps": 1e-05,
39
+ "rope_interleave": true,
40
+ "rope_parameters": {
41
+ "rope_theta": 800000.0,
42
+ "rope_type": "default"
43
+ },
44
+ "routed_scaling_factor": 2.446,
45
+ "scoring_func": "sigmoid",
46
+ "seq_aux": true,
47
+ "tie_word_embeddings": false,
48
+ "topk_group": 1,
49
+ "topk_method": "noaux_tc",
50
+ "transformers_version": "5.5.4",
51
+ "use_cache": false,
52
+ "v_head_dim": 128,
53
+ "vocab_size": 163840
54
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 163584,
4
+ "eos_token_id": 163585,
5
+ "pad_token_id": 163839,
6
+ "transformers_version": "5.5.4"
7
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d769d6941479cf7cc33975675e04be353e6a893000d6ffd4d2962207098b9432
3
+ size 31920893454
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1e66ac13dbffbd49e537ff08fb145286e6653cde5c94d56c618bcf314115b66e
3
+ size 19545447
tokenizer_config.json ADDED
@@ -0,0 +1,216 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "163584": {
4
+ "content": "[BOS]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "163585": {
12
+ "content": "[EOS]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "163586": {
20
+ "content": "<|im_end|>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "163587": {
28
+ "content": "<|im_user|>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "163588": {
36
+ "content": "<|im_assistant|>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "163590": {
44
+ "content": "<|start_header_id|>",
45
+ "lstrip": false,
46
+ "normalized": false,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": true
50
+ },
51
+ "163591": {
52
+ "content": "<|end_header_id|>",
53
+ "lstrip": false,
54
+ "normalized": false,
55
+ "rstrip": false,
56
+ "single_word": false,
57
+ "special": true
58
+ },
59
+ "163593": {
60
+ "content": "[EOT]",
61
+ "lstrip": false,
62
+ "normalized": false,
63
+ "rstrip": false,
64
+ "single_word": false,
65
+ "special": true
66
+ },
67
+ "163594": {
68
+ "content": "<|im_system|>",
69
+ "lstrip": false,
70
+ "normalized": false,
71
+ "rstrip": false,
72
+ "single_word": false,
73
+ "special": true
74
+ },
75
+ "163595": {
76
+ "content": "<|tool_calls_section_begin|>",
77
+ "lstrip": false,
78
+ "normalized": false,
79
+ "rstrip": false,
80
+ "single_word": false,
81
+ "special": false
82
+ },
83
+ "163596": {
84
+ "content": "<|tool_calls_section_end|>",
85
+ "lstrip": false,
86
+ "normalized": false,
87
+ "rstrip": false,
88
+ "single_word": false,
89
+ "special": false
90
+ },
91
+ "163597": {
92
+ "content": "<|tool_call_begin|>",
93
+ "lstrip": false,
94
+ "normalized": false,
95
+ "rstrip": false,
96
+ "single_word": false,
97
+ "special": false
98
+ },
99
+ "163598": {
100
+ "content": "<|tool_call_argument_begin|>",
101
+ "lstrip": false,
102
+ "normalized": false,
103
+ "rstrip": false,
104
+ "single_word": false,
105
+ "special": false
106
+ },
107
+ "163599": {
108
+ "content": "<|tool_call_end|>",
109
+ "lstrip": false,
110
+ "normalized": false,
111
+ "rstrip": false,
112
+ "single_word": false,
113
+ "special": false
114
+ },
115
+ "163601": {
116
+ "content": "<|im_middle|>",
117
+ "lstrip": false,
118
+ "normalized": false,
119
+ "rstrip": false,
120
+ "single_word": false,
121
+ "special": true
122
+ },
123
+ "163602": {
124
+ "content": "<|media_begin|>",
125
+ "lstrip": false,
126
+ "normalized": false,
127
+ "rstrip": false,
128
+ "single_word": false,
129
+ "special": true
130
+ },
131
+ "163603": {
132
+ "content": "<|media_content|>",
133
+ "lstrip": false,
134
+ "normalized": false,
135
+ "rstrip": false,
136
+ "single_word": false,
137
+ "special": true
138
+ },
139
+ "163604": {
140
+ "content": "<|media_end|>",
141
+ "lstrip": false,
142
+ "normalized": false,
143
+ "rstrip": false,
144
+ "single_word": false,
145
+ "special": true
146
+ },
147
+ "163605": {
148
+ "content": "<|media_pad|>",
149
+ "lstrip": false,
150
+ "normalized": false,
151
+ "rstrip": false,
152
+ "single_word": false,
153
+ "special": true
154
+ },
155
+ "163606": {
156
+ "content": "<think>",
157
+ "lstrip": false,
158
+ "normalized": false,
159
+ "rstrip": false,
160
+ "single_word": false,
161
+ "special": false
162
+ },
163
+ "163607": {
164
+ "content": "</think>",
165
+ "lstrip": false,
166
+ "normalized": false,
167
+ "rstrip": false,
168
+ "single_word": false,
169
+ "special": false
170
+ },
171
+ "163838": {
172
+ "content": "[UNK]",
173
+ "lstrip": false,
174
+ "normalized": false,
175
+ "rstrip": false,
176
+ "single_word": false,
177
+ "special": true
178
+ },
179
+ "163839": {
180
+ "content": "[PAD]",
181
+ "lstrip": false,
182
+ "normalized": false,
183
+ "rstrip": false,
184
+ "single_word": false,
185
+ "special": true
186
+ }
187
+ },
188
+ "additional_special_tokens": [
189
+ "<|im_end|>",
190
+ "<|im_user|>",
191
+ "<|im_assistant|>",
192
+ "<|start_header_id|>",
193
+ "<|end_header_id|>",
194
+ "[EOT]",
195
+ "<|im_system|>",
196
+ "<|im_middle|>",
197
+ "<|media_begin|>",
198
+ "<|media_content|>",
199
+ "<|media_end|>",
200
+ "<|media_pad|>"
201
+ ],
202
+ "bos_token": "[BOS]",
203
+ "clean_up_tokenization_spaces": false,
204
+ "eos_token": "[EOS]",
205
+ "extra_special_tokens": {},
206
+ "model_max_length": 1000000000000000019884624838656,
207
+ "pad_token": "[PAD]",
208
+ "tokenizer_class": "TikTokenTokenizer",
209
+ "unk_token": "[UNK]",
210
+ "auto_map": {
211
+ "AutoTokenizer": [
212
+ "tokenization_kimi.TikTokenTokenizer",
213
+ null
214
+ ]
215
+ }
216
+ }