Spaces:
Sleeping
Sleeping
jonathanjordan21
commited on
Commit
•
a8bfe22
1
Parent(s):
41f032f
Update constants/models.py
Browse files- constants/models.py +4 -1
constants/models.py
CHANGED
@@ -11,6 +11,7 @@ MODEL_MAP = {
|
|
11 |
"default": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
|
12 |
"Qwen2-72B": "Qwen/Qwen2-72B",
|
13 |
"Qwen2-7B-Instruct": "Qwen/Qwen2-7B-Instruct",
|
|
|
14 |
"llama3-8b" : "meta-llama/Meta-Llama-3-8B-Instruct"
|
15 |
}
|
16 |
|
@@ -34,7 +35,8 @@ STOP_SEQUENCES_MAP = {
|
|
34 |
"command-r-plus": "<|END_OF_TURN_TOKEN|>",
|
35 |
# https://huggingface.co/Qwen/Qwen2-72B-Instruct/blob/main/tokenizer_config.json#L30
|
36 |
"Qwen2-72B": "<|im_end|>",
|
37 |
-
"Qwen2-7B-Instruct": "
|
|
|
38 |
|
39 |
"llama3-8b" : "<|eot_id|>"
|
40 |
}
|
@@ -52,6 +54,7 @@ TOKEN_LIMIT_MAP = {
|
|
52 |
"gpt-3.5-turbo": 8192,
|
53 |
"Qwen2-72B": 32768,
|
54 |
"Qwen2-7B-Instruct": 32768,
|
|
|
55 |
"llama3-8b": 8192,
|
56 |
}
|
57 |
|
|
|
11 |
"default": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
|
12 |
"Qwen2-72B": "Qwen/Qwen2-72B",
|
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 |
|
|
|
35 |
"command-r-plus": "<|END_OF_TURN_TOKEN|>",
|
36 |
# https://huggingface.co/Qwen/Qwen2-72B-Instruct/blob/main/tokenizer_config.json#L30
|
37 |
"Qwen2-72B": "<|im_end|>",
|
38 |
+
"Qwen2-7B-Instruct": "<|im_end|>",
|
39 |
+
"Qwen/Qwen2-1.5B-Instruct": "<|im_end|>",
|
40 |
|
41 |
"llama3-8b" : "<|eot_id|>"
|
42 |
}
|
|
|
54 |
"gpt-3.5-turbo": 8192,
|
55 |
"Qwen2-72B": 32768,
|
56 |
"Qwen2-7B-Instruct": 32768,
|
57 |
+
"Qwen2-1.5B-Instruct": 32768,
|
58 |
"llama3-8b": 8192,
|
59 |
}
|
60 |
|