jonathanjordan21 commited on
Commit
55ab28a
1 Parent(s): 6f4d127

Update constants/models.py

Browse files
Files changed (1) hide show
  1. constants/models.py +86 -24
constants/models.py CHANGED
@@ -9,10 +9,12 @@ MODEL_MAP = {
9
  "llama3-70b": "meta-llama/Meta-Llama-3-70B-Instruct",
10
  "zephyr-141b": "HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1",
11
  "default": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
12
- "Qwen2-72B": "Qwen/Qwen2-72B-Instruct",
13
- "Qwen2-7B-Instruct": "Qwen/Qwen2-7B-Instruct",
14
- "Qwen2-1.5B-Instruct" : "Qwen/Qwen2-1.5B-Instruct",
15
- "llama3-8b" : "meta-llama/Meta-Llama-3-8B-Instruct"
 
 
16
  }
17
 
18
  AVAILABLE_MODELS = list(MODEL_MAP.keys())
@@ -36,12 +38,14 @@ STOP_SEQUENCES_MAP = {
36
  # https://huggingface.co/CohereForAI/c4ai-command-r-plus-4bit/blob/main/tokenizer_config.json#L305
37
  "command-r-plus": "<|END_OF_TURN_TOKEN|>",
38
  # https://huggingface.co/Qwen/Qwen2-72B-Instruct/blob/main/tokenizer_config.json#L30
39
- "Qwen2-72B": "<|im_end|>",
40
- "Qwen2-7B-Instruct": "<|im_end|>",
41
- "Qwen2-1.5B-Instruct": "<|im_end|>",
42
 
43
  "llama3-8b" : "<|eot_id|>",
44
- "llama3-70b" : "<|eot_id|>"
 
 
45
  }
46
 
47
  TOKEN_LIMIT_MAP = {
@@ -55,23 +59,24 @@ TOKEN_LIMIT_MAP = {
55
  "llama3-70b": 8192,
56
  "zephyr-141b": 2048,
57
  "gpt-3.5-turbo": 8192,
58
- "Qwen2-72B": 32768,
59
- "Qwen2-7B-Instruct": 32768,
60
- "Qwen2-1.5B-Instruct": 32768,
61
  "llama3-8b": 8192,
 
 
62
  }
63
 
64
  TOKEN_RESERVED = 20
65
 
66
-
67
- # https://platform.openai.com/docs/api-reference/models/list
68
- AVAILABLE_MODELS_DICTS = [
69
  {
70
  "id": "mixtral-8x7b",
71
  "description": "[mistralai/Mixtral-8x7B-Instruct-v0.1]: https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1",
72
  "object": "model",
73
  "created": 1700000000,
74
  "owned_by": "mistralai",
 
75
  },
76
  {
77
  "id": "nous-mixtral-8x7b",
@@ -79,6 +84,7 @@ AVAILABLE_MODELS_DICTS = [
79
  "object": "model",
80
  "created": 1700000000,
81
  "owned_by": "NousResearch",
 
82
  },
83
  {
84
  "id": "mistral-7b",
@@ -86,6 +92,7 @@ AVAILABLE_MODELS_DICTS = [
86
  "object": "model",
87
  "created": 1700000000,
88
  "owned_by": "mistralai",
 
89
  },
90
  {
91
  "id": "yi-1.5-34b",
@@ -93,6 +100,7 @@ AVAILABLE_MODELS_DICTS = [
93
  "object": "model",
94
  "created": 1700000000,
95
  "owned_by": "01-ai",
 
96
  },
97
  {
98
  "id": "gemma-7b",
@@ -100,63 +108,92 @@ AVAILABLE_MODELS_DICTS = [
100
  "object": "model",
101
  "created": 1700000000,
102
  "owned_by": "Google",
 
103
  },
104
  {
105
  "id": "openchat-3.5",
106
  "description": "[openchat/openchat-3.5-0106]: https://huggingface.co/openchat/openchat-3.5-0106",
107
  "object": "model",
108
  "created": 1700000000,
109
- "owned_by": "openchat"
 
110
  },
111
  {
112
  "id": "command-r-plus",
113
  "description": "[CohereForAI/c4ai-command-r-plus]: https://huggingface.co/CohereForAI/c4ai-command-r-plus",
114
  "object": "model",
115
  "created": 1700000000,
116
- "owned_by": "CohereForAI"
 
117
  },
118
  {
119
  "id": "llama3-70b",
120
  "description": "[meta-llama/Meta-Llama-3-70B]: https://huggingface.co/meta-llama/Meta-Llama-3-70B",
121
  "object": "model",
122
  "created": 1700000000,
123
- "owned_by": "Meta"
 
124
  },
125
  {
126
  "id": "zephyr-141b",
127
  "description": "[HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1]: https://huggingface.co/HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1",
128
  "object": "model",
129
  "created": 1700000000,
130
- "owned_by": "Huggingface"
 
131
  },
132
  {
133
  "id": "gpt-3.5-turbo",
134
  "description": "[openai/gpt-3.5-turbo]: https://platform.openai.com/docs/models/gpt-3-5-turbo",
135
  "object": "model",
136
  "created": 1700000000,
137
- "owned_by": "OpenAI"
 
138
  },
139
  {
140
- "id": "Qwen2-72B",
141
  "description": "[Qwen/Qwen2-72B]: https://huggingface.co/Qwen/Qwen2-72B",
142
  "object": "model",
143
  "created": 1700000000,
144
- "owned_by": "Qwen"
 
145
  },
146
  {
147
  "id": "llama3-8b",
148
  "description": "[meta-llama/Meta-Llama-3-8B]: https://huggingface.co/meta-llama/Meta-Llama-3-8B",
149
  "object": "model",
150
  "created": 1700000000,
151
- "owned_by": "Meta"
 
152
  },
153
  {
154
- "id": "Qwen2-1.5B-Instruct",
155
  "description": "[Qwen/Qwen2-1.5B-Instruct]: https://huggingface.co/Qwen/Qwen2-1.5B-Instruct",
156
  "object": "model",
157
  "created": 1700000000,
158
- "owned_by": "Qwen"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
159
  },
 
 
 
 
160
  {
161
  "id": "nomic-text-embed",
162
  "description": "[nomic-ai/nomic-embed-text-v1.5]: https://huggingface.co/nomic-ai/nomic-embed-text-v1.5",
@@ -171,4 +208,29 @@ AVAILABLE_MODELS_DICTS = [
171
  "created": 1700000000,
172
  "owned_by": "mixedbread-ai"
173
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
174
  ]
 
 
 
 
 
9
  "llama3-70b": "meta-llama/Meta-Llama-3-70B-Instruct",
10
  "zephyr-141b": "HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1",
11
  "default": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
12
+ "qwen2-72b": "Qwen/Qwen2-72B-Instruct",
13
+ "qwen2-7b-instruct": "Qwen/Qwen2-7B-Instruct",
14
+ "qwen2-1.5b-instruct" : "Qwen/Qwen2-1.5B-Instruct",
15
+ "llama3-8b" : "meta-llama/Meta-Llama-3-8B-Instruct",
16
+ "mistral-nemo-instruct":"mistralai/Mistral-Nemo-Instruct-2407",
17
+ "zephyr-7b-beta":"HuggingFaceH4/zephyr-7b-beta"
18
  }
19
 
20
  AVAILABLE_MODELS = list(MODEL_MAP.keys())
 
38
  # https://huggingface.co/CohereForAI/c4ai-command-r-plus-4bit/blob/main/tokenizer_config.json#L305
39
  "command-r-plus": "<|END_OF_TURN_TOKEN|>",
40
  # https://huggingface.co/Qwen/Qwen2-72B-Instruct/blob/main/tokenizer_config.json#L30
41
+ "qwen2-72b": "<|im_end|>",
42
+ "qwen2-7b-instruct": "<|im_end|>",
43
+ "qwen2-1.5b-instruct": "<|im_end|>",
44
 
45
  "llama3-8b" : "<|eot_id|>",
46
+ "llama3-70b" : "<|eot_id|>",
47
+ "mistral-nemo-instruct": "</s>",
48
+ "zephyr-7b-beta":"</s>"
49
  }
50
 
51
  TOKEN_LIMIT_MAP = {
 
59
  "llama3-70b": 8192,
60
  "zephyr-141b": 2048,
61
  "gpt-3.5-turbo": 8192,
62
+ "qwen2-72b": 32768,
63
+ "qwen2-7b-instruct": 32768,
64
+ "qwen2-1.5b-instruct": 32768,
65
  "llama3-8b": 8192,
66
+ "mistral-nemo-instruct":128000,
67
+ "zephyr-7b-beta":32768,
68
  }
69
 
70
  TOKEN_RESERVED = 20
71
 
72
+ LLM_MODELS_DICTS = [
 
 
73
  {
74
  "id": "mixtral-8x7b",
75
  "description": "[mistralai/Mixtral-8x7B-Instruct-v0.1]: https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1",
76
  "object": "model",
77
  "created": 1700000000,
78
  "owned_by": "mistralai",
79
+ "type": "basic",
80
  },
81
  {
82
  "id": "nous-mixtral-8x7b",
 
84
  "object": "model",
85
  "created": 1700000000,
86
  "owned_by": "NousResearch",
87
+ "type": "basic",
88
  },
89
  {
90
  "id": "mistral-7b",
 
92
  "object": "model",
93
  "created": 1700000000,
94
  "owned_by": "mistralai",
95
+ "type": "basic",
96
  },
97
  {
98
  "id": "yi-1.5-34b",
 
100
  "object": "model",
101
  "created": 1700000000,
102
  "owned_by": "01-ai",
103
+ "type": "pro",
104
  },
105
  {
106
  "id": "gemma-7b",
 
108
  "object": "model",
109
  "created": 1700000000,
110
  "owned_by": "Google",
111
+ "type": "basic",
112
  },
113
  {
114
  "id": "openchat-3.5",
115
  "description": "[openchat/openchat-3.5-0106]: https://huggingface.co/openchat/openchat-3.5-0106",
116
  "object": "model",
117
  "created": 1700000000,
118
+ "owned_by": "openchat",
119
+ "type": "basic",
120
  },
121
  {
122
  "id": "command-r-plus",
123
  "description": "[CohereForAI/c4ai-command-r-plus]: https://huggingface.co/CohereForAI/c4ai-command-r-plus",
124
  "object": "model",
125
  "created": 1700000000,
126
+ "owned_by": "CohereForAI",
127
+ "type": "pro",
128
  },
129
  {
130
  "id": "llama3-70b",
131
  "description": "[meta-llama/Meta-Llama-3-70B]: https://huggingface.co/meta-llama/Meta-Llama-3-70B",
132
  "object": "model",
133
  "created": 1700000000,
134
+ "owned_by": "Meta",
135
+ "type": "pro",
136
  },
137
  {
138
  "id": "zephyr-141b",
139
  "description": "[HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1]: https://huggingface.co/HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1",
140
  "object": "model",
141
  "created": 1700000000,
142
+ "owned_by": "Huggingface",
143
+ "type": "pro",
144
  },
145
  {
146
  "id": "gpt-3.5-turbo",
147
  "description": "[openai/gpt-3.5-turbo]: https://platform.openai.com/docs/models/gpt-3-5-turbo",
148
  "object": "model",
149
  "created": 1700000000,
150
+ "owned_by": "OpenAI",
151
+ "type": "subscription",
152
  },
153
  {
154
+ "id": "qwen2-72b",
155
  "description": "[Qwen/Qwen2-72B]: https://huggingface.co/Qwen/Qwen2-72B",
156
  "object": "model",
157
  "created": 1700000000,
158
+ "owned_by": "Qwen",
159
+ "type": "pro",
160
  },
161
  {
162
  "id": "llama3-8b",
163
  "description": "[meta-llama/Meta-Llama-3-8B]: https://huggingface.co/meta-llama/Meta-Llama-3-8B",
164
  "object": "model",
165
  "created": 1700000000,
166
+ "owned_by": "Meta",
167
+ "type": "basic",
168
  },
169
  {
170
+ "id": "qwen2-1.5b-instruct",
171
  "description": "[Qwen/Qwen2-1.5B-Instruct]: https://huggingface.co/Qwen/Qwen2-1.5B-Instruct",
172
  "object": "model",
173
  "created": 1700000000,
174
+ "owned_by": "Qwen",
175
+ "type": "basic",
176
+ },
177
+ {
178
+ "id": "mistral-nemo-instruct",
179
+ "description": "[mistralai/Mistral-Nemo-Instruct-2407]: https://huggingface.co/mistralai/Mistral-Nemo-Instruct-2407",
180
+ "object": "model",
181
+ "created": 1700000000,
182
+ "owned_by": "mistralai",
183
+ "type": "basic",
184
+ },
185
+ {
186
+ "id": "zephyr-7b-beta",
187
+ "description": "[HuggingFaceH4/zephyr-7b-beta]: https://huggingface.co/HuggingFaceH4/zephyr-7b-beta",
188
+ "object": "model",
189
+ "created": 1700000000,
190
+ "owned_by": "HuggingFaceH4",
191
+ "type": "basic",
192
  },
193
+ ]
194
+
195
+
196
+ EMBEDDINGS_MODELS_DICTS = [
197
  {
198
  "id": "nomic-text-embed",
199
  "description": "[nomic-ai/nomic-embed-text-v1.5]: https://huggingface.co/nomic-ai/nomic-embed-text-v1.5",
 
208
  "created": 1700000000,
209
  "owned_by": "mixedbread-ai"
210
  },
211
+ {
212
+ "id": "mixbread-ai/mxbai-embed-large-v1",
213
+ "description": "[mixedbread-ai/mxbai-embed-large-v1]: https://huggingface.co/mixedbread-ai/mxbai-embed-large-v1",
214
+ "object": "embeddings",
215
+ "created": 1700000000,
216
+ "owned_by": "mixedbread-ai"
217
+ },
218
+ {
219
+ "id": "multilingual-e5-large-instruct",
220
+ "description": "[intfloat/multilingual-e5-large-instruct]: https://huggingface.co/intfloat/multilingual-e5-large-instruct",
221
+ "object": "embeddings",
222
+ "created": 1700000000,
223
+ "owned_by": "intfloat"
224
+ },
225
+ {
226
+ "id": "intfloat/multilingual-e5-large-instruct",
227
+ "description": "[intfloat/multilingual-e5-large-instruct]: https://huggingface.co/intfloat/multilingual-e5-large-instruct",
228
+ "object": "embeddings",
229
+ "created": 1700000000,
230
+ "owned_by": "intfloat"
231
+ },
232
  ]
233
+
234
+
235
+ # https://platform.openai.com/docs/api-reference/models/list
236
+ AVAILABLE_MODELS_DICTS = LLM_MODELS_DICTS + EMBEDDINGS_MODELS_DICTS