Hugging Face's logo

pankajmathur
/
model_007

Text Generation
Transformers
PyTorch
Safetensors
English
llama
Eval Results (legacy)
text-generation-inference
Model card Files
xet
 Community
5
pankajmathur commited on
Commit
e5334da
·
1 Parent(s): 213eef7

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +49 -24
README.md CHANGED
@@ -3,6 +3,12 @@ language:
3
  - en
4
  library_name: transformers
5
  license: llama2
 
 
 
 
 
 
6
  ---
7
 
8
 
@@ -30,7 +36,7 @@ A hybrid (explain + instruct) style Llama2-70b model, Pleae check examples below
30
 
31
 
32
  ### quantized versions
33
- Huge respect to man.. @TheBloke, here are the GGML/GPTQ/GGUF versions, go crazy :)
34
 
35
  https://huggingface.co/TheBloke/model_007-70B-GGML
36
 
@@ -52,19 +58,22 @@ We evaluated model_007 on a wide range of tasks using [Language Model Evaluation
52
 
53
  Here are the results on metrics used by [HuggingFaceH4 Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
54
 
55
- |||||
56
- |:------:|:--------:|:-------:|:--------:|
57
- |**Task**|**Metric**|**Value**|**Stderr**|
58
- |*arc_challenge*|acc_norm|0.7108|0.0141|
59
- |*hellaswag*|acc_norm|0.8765|0.0038|
60
- |*mmlu*|acc_norm|0.6904|0.0351|
61
- |*truthfulqa_mc*|mc2|0.6312|0.0157|
62
- |**Total Average**|-|**0.72729**||
 
 
 
63
 
64
 
65
  <br>
66
 
67
- ## Example Usage
68
 
69
  Here is the Orca prompt format
70
 
@@ -79,7 +88,35 @@ Tell me about Orcas.
79
 
80
  ```
81
 
82
- Below shows a code example on how to use this model
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
83
 
84
  ```python
85
  import torch
@@ -105,19 +142,7 @@ print(tokenizer.decode(output[0], skip_special_tokens=True))
105
 
106
  ```
107
 
108
-
109
- Here is the Alpaca prompt format
110
-
111
- ```
112
-
113
- ### User:
114
- Tell me about Alpacas.
115
-
116
- ### Assistant:
117
-
118
- ```
119
-
120
- Below shows a code example on how to use this model
121
 
122
  ```python
123
  import torch
 
3
  - en
4
  library_name: transformers
5
  license: llama2
6
+ datasets:
7
+ - pankajmathur/orca_mini_v1_dataset
8
+ - pankajmathur/dolly-v2_orca
9
+ - pankajmathur/WizardLM_Orca
10
+ - pankajmathur/alpaca_orca
11
+ - ehartford/dolphin
12
  ---
13
 
14
 
 
36
 
37
 
38
  ### quantized versions
39
+ Huge respect to @TheBloke, here are the GGML/GPTQ/GGUF versions, go crazy :)
40
 
41
  https://huggingface.co/TheBloke/model_007-70B-GGML
42
 
 
58
 
59
  Here are the results on metrics used by [HuggingFaceH4 Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
60
 
61
+ |||
62
+ |:------:|:--------:|
63
+ |**Task**|**Value**|
64
+ |*ARC*|0.7108|
65
+ |*HellaSwag*|0.8765|
66
+ |*MMLU*|0.6904|
67
+ |*TruthfulQA*|0.6312|
68
+ |*Winogrande*|0.8335|
69
+ |*GSM8K*|0.3715|
70
+ |*DROP*|0.3105|
71
+ |**Total Average**|**0.6320**|
72
 
73
 
74
  <br>
75
 
76
+ ## Prompt Format
77
 
78
  Here is the Orca prompt format
79
 
 
88
 
89
  ```
90
 
91
+ Here is the Alpaca prompt format
92
+
93
+ ```
94
+
95
+ ### User:
96
+ Tell me about Alpacas.
97
+
98
+ ### Assistant:
99
+
100
+ ```
101
+
102
+ #### OobaBooga Instructions:
103
+
104
+ This model required upto 45GB GPU VRAM in 4bit so it can be loaded directly on Single RTX 6000/L40/A40/A100/H100 GPU or Double RTX 4090/L4/A10/RTX 3090/RTX A5000
105
+ So, if you have access to Machine with 45GB GPU VRAM and have installed [OobaBooga Web UI](https://github.com/oobabooga/text-generation-webui) on it.
106
+ You can just download this model by using HF repo link directly on OobaBooga Web UI "Model" Tab/Page & Just use **load-in-4bit** option in it.
107
+
108
+ ![model_load_screenshot](https://huggingface.co/pankajmathur/model_101/resolve/main/oobabooga_model_load_screenshot.png)
109
+
110
+
111
+ After that go to Default Tab/Page on OobaBooga Web UI and **copy paste above prompt format into Input** and Enjoy!
112
+
113
+ ![default_input_screenshot](https://huggingface.co/pankajmathur/model_101/resolve/main/default_input_screenshot.png)
114
+
115
+ <br>
116
+
117
+ #### Code Instructions:
118
+
119
+ Below shows a code example on how to use this model via Orca prompt
120
 
121
  ```python
122
  import torch
 
142
 
143
  ```
144
 
145
+ Below shows a code example on how to use this model via Alpaca prompt
 
 
 
 
 
 
 
 
 
 
 
 
146
 
147
  ```python
148
  import torch