Update README.md
Browse files
README.md
CHANGED
@@ -13,21 +13,42 @@ Mobius is a RWKV v6 arch model, a state based RNN+CNN+Transformer Mixed language
|
|
13 |
In comparison with the previous released Mobius, the improvements include:
|
14 |
|
15 |
* Only 24G Vram to run this model locally with fp16;
|
16 |
-
* Significant performance improvement;
|
17 |
-
* Multilingual support ;
|
18 |
* Stable support of 16K context length.
|
19 |
* function call support ;
|
20 |
|
21 |
|
22 |
## Usage
|
23 |
-
|
24 |
-
|
25 |
-
Recommend Temp and topp:
|
26 |
-
|
27 |
-
function call format:
|
28 |
-
|
29 |
-
|
30 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
31 |
|
32 |
Assistant: xxxx
|
33 |
|
@@ -35,6 +56,8 @@ Obersavtion: xxxx
|
|
35 |
|
36 |
Assistant: xxxx
|
37 |
|
|
|
|
|
38 |
## More details
|
39 |
Mobius 12B 16k based on RWKV v6 arch, which is leading state based RNN+CNN+Transformer Mixed large language model which focus opensouce community
|
40 |
* 10~100 trainning/inference cost reduce;
|
@@ -42,12 +65,12 @@ Mobius 12B 16k based on RWKV v6 arch, which is leading state based RNN+CNN+Trans
|
|
42 |
* community support.
|
43 |
|
44 |
## requirements
|
45 |
-
|
46 |
|
47 |
* [RWKV Runner](https://github.com/josStorer/RWKV-Runner)
|
48 |
* [Ai00 server](https://github.com/cgisky1980/ai00_rwkv_server)
|
49 |
|
50 |
## Benchmark
|
51 |
-
ceval
|
52 |
-
cmmlu
|
53 |
|
|
|
13 |
In comparison with the previous released Mobius, the improvements include:
|
14 |
|
15 |
* Only 24G Vram to run this model locally with fp16;
|
16 |
+
* Significant performance improvement in chinese;
|
|
|
17 |
* Stable support of 16K context length.
|
18 |
* function call support ;
|
19 |
|
20 |
|
21 |
## Usage
|
22 |
+
Chat format: User: xxxx\n\nAssistant: xxx\n\n
|
23 |
+
|
24 |
+
Recommend Temp and topp: 1 0.3
|
25 |
+
|
26 |
+
function call format example:
|
27 |
+
|
28 |
+
```
|
29 |
+
System: You are a helpful assistant with access to the following functions. Use them if required -{
|
30 |
+
"name": "get_exchange_rate",
|
31 |
+
"description": "Get the exchange rate between two currencies",
|
32 |
+
"parameters": {
|
33 |
+
"type": "object",
|
34 |
+
"properties": {
|
35 |
+
"base_currency": {
|
36 |
+
"type": "string",
|
37 |
+
"description": "The currency to convert from"
|
38 |
+
},
|
39 |
+
"target_currency": {
|
40 |
+
"type": "string",
|
41 |
+
"description": "The currency to convert to"
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"required": [
|
45 |
+
"base_currency",
|
46 |
+
"target_currency"
|
47 |
+
]
|
48 |
+
}
|
49 |
+
}
|
50 |
+
|
51 |
+
User: Hi, I need to know the exchange rate from USD to EUR
|
52 |
|
53 |
Assistant: xxxx
|
54 |
|
|
|
56 |
|
57 |
Assistant: xxxx
|
58 |
|
59 |
+
```
|
60 |
+
|
61 |
## More details
|
62 |
Mobius 12B 16k based on RWKV v6 arch, which is leading state based RNN+CNN+Transformer Mixed large language model which focus opensouce community
|
63 |
* 10~100 trainning/inference cost reduce;
|
|
|
65 |
* community support.
|
66 |
|
67 |
## requirements
|
68 |
+
21.9G vram to run fp16, 13.7G for int8, 7.2 for nf4 with Ai00 server.
|
69 |
|
70 |
* [RWKV Runner](https://github.com/josStorer/RWKV-Runner)
|
71 |
* [Ai00 server](https://github.com/cgisky1980/ai00_rwkv_server)
|
72 |
|
73 |
## Benchmark
|
74 |
+
ceval 63.53
|
75 |
+
cmmlu 76.07
|
76 |
|