Commit
•
14ad760
1
Parent(s):
b30dc0e
Update README.md
Browse files
README.md
CHANGED
@@ -1,45 +1,83 @@
|
|
1 |
-
---
|
2 |
-
license: other
|
3 |
-
tags:
|
4 |
-
- generated_from_trainer
|
5 |
-
- axolotl
|
6 |
-
- llama-cpp
|
7 |
-
- gguf-my-repo
|
8 |
-
- Dllama
|
9 |
-
- Infero
|
10 |
-
base_model: meta-llama/Meta-Llama-3-8B
|
11 |
-
datasets:
|
12 |
-
- cognitivecomputations/Dolphin-2.9
|
13 |
-
- teknium/OpenHermes-2.5
|
14 |
-
- m-a-p/CodeFeedback-Filtered-Instruction
|
15 |
-
- cognitivecomputations/dolphin-coder
|
16 |
-
- cognitivecomputations/samantha-data
|
17 |
-
- microsoft/orca-math-word-problems-200k
|
18 |
-
- Locutusque/function-calling-chatml
|
19 |
-
- internlm/Agent-FLAN
|
20 |
-
model-index:
|
21 |
-
- name: out
|
22 |
-
results: []
|
23 |
-
---
|
24 |
|
25 |
# tinybiggames/dolphin-2.9.1-llama-3-8b-Q4_K_M-GGUF
|
26 |
This model was converted to GGUF format from [`cognitivecomputations/dolphin-2.9.1-llama-3-8b`](https://huggingface.co/cognitivecomputations/dolphin-2.9.1-llama-3-8b) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space.
|
27 |
Refer to the [original model card](https://huggingface.co/cognitivecomputations/dolphin-2.9.1-llama-3-8b) for more details on the model.
|
28 |
-
## Use with tinyBigGAMES's
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
45 |
```
|
|
|
1 |
+
---
|
2 |
+
license: other
|
3 |
+
tags:
|
4 |
+
- generated_from_trainer
|
5 |
+
- axolotl
|
6 |
+
- llama-cpp
|
7 |
+
- gguf-my-repo
|
8 |
+
- Dllama
|
9 |
+
- Infero
|
10 |
+
base_model: meta-llama/Meta-Llama-3-8B
|
11 |
+
datasets:
|
12 |
+
- cognitivecomputations/Dolphin-2.9
|
13 |
+
- teknium/OpenHermes-2.5
|
14 |
+
- m-a-p/CodeFeedback-Filtered-Instruction
|
15 |
+
- cognitivecomputations/dolphin-coder
|
16 |
+
- cognitivecomputations/samantha-data
|
17 |
+
- microsoft/orca-math-word-problems-200k
|
18 |
+
- Locutusque/function-calling-chatml
|
19 |
+
- internlm/Agent-FLAN
|
20 |
+
model-index:
|
21 |
+
- name: out
|
22 |
+
results: []
|
23 |
+
---
|
24 |
|
25 |
# tinybiggames/dolphin-2.9.1-llama-3-8b-Q4_K_M-GGUF
|
26 |
This model was converted to GGUF format from [`cognitivecomputations/dolphin-2.9.1-llama-3-8b`](https://huggingface.co/cognitivecomputations/dolphin-2.9.1-llama-3-8b) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space.
|
27 |
Refer to the [original model card](https://huggingface.co/cognitivecomputations/dolphin-2.9.1-llama-3-8b) for more details on the model.
|
28 |
+
## Use with tinyBigGAMES's LMEngine Inference Library
|
29 |
+
|
30 |
+
|
31 |
+
How to configure LMEngine:
|
32 |
+
|
33 |
+
```Delphi
|
34 |
+
Config_Init(
|
35 |
+
'C:/LLM/gguf', // path to model files
|
36 |
+
-1 // number of GPU layer, -1 to use all available layers
|
37 |
+
);
|
38 |
+
```
|
39 |
+
|
40 |
+
How to define model:
|
41 |
+
|
42 |
+
```Delphi
|
43 |
+
Model_Define('dolphin-2.9.1-llama-3-8b.Q4_K_M.gguf',
|
44 |
+
'dolphin-llama3:8B:Q4KM', 8000,
|
45 |
+
'<|im_start|>{role}\n{content}<|im_end|>\n',
|
46 |
+
'<|im_start|>assistant');
|
47 |
+
```
|
48 |
+
|
49 |
+
How to add a message:
|
50 |
+
|
51 |
+
```Delphi
|
52 |
+
Message_Add(
|
53 |
+
ROLE_USER, // role
|
54 |
+
'What is AI?' // content
|
55 |
+
);
|
56 |
+
```
|
57 |
+
|
58 |
+
`{role}` - will be substituted with the message "role"
|
59 |
+
`{content}` - will be substituted with the message "content"
|
60 |
+
|
61 |
+
How to do inference:
|
62 |
+
|
63 |
+
```Delphi
|
64 |
+
var
|
65 |
+
LTokenOutputSpeed: Single;
|
66 |
+
LInputTokens: Int32;
|
67 |
+
LOutputTokens: Int32;
|
68 |
+
LTotalTokens: Int32;
|
69 |
+
|
70 |
+
if Inference_Run('dolphin-llama3:8B:Q4KM', 1024) then
|
71 |
+
begin
|
72 |
+
Inference_GetUsage(nil, @LTokenOutputSpeed, @LInputTokens, @LOutputTokens,
|
73 |
+
@LTotalTokens);
|
74 |
+
Console_PrintLn('', FG_WHITE);
|
75 |
+
Console_PrintLn('Tokens :: Input: %d, Output: %d, Total: %d, Speed: %3.1f t/s',
|
76 |
+
FG_BRIGHTYELLOW, LInputTokens, LOutputTokens, LTotalTokens, LTokenOutputSpeed);
|
77 |
+
end
|
78 |
+
else
|
79 |
+
begin
|
80 |
+
Console_PrintLn('', FG_WHITE);
|
81 |
+
Console_PrintLn('Error: %s', FG_RED, Error_Get());
|
82 |
+
end;
|
83 |
```
|