ibivibiv commited on
Commit
8c93fdc
1 Parent(s): 6e762a5

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +31 -29
README.md CHANGED
@@ -10,12 +10,42 @@ tags:
10
 
11
  ![img](./strix_rufipes.png)
12
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
13
  # Model Details
14
  * **Trained by**: [ibivibiv](https://huggingface.co/ibivibiv)
15
  * **Library**: [HuggingFace Transformers](https://github.com/huggingface/transformers)
16
  * **Model type:** **strix-rufipes-70b** is an auto-regressive language model fine tuned on the Llama 2 transformer architecture.
17
  * **Language(s)**: English
18
- * **Purpose**: Has specific training for logic enforcement, will do well in ARC or other logic testing as well as critical thinking tasks. This model is targeted towards planning exercises.
19
 
20
  # Benchmark Scores
21
 
@@ -87,35 +117,7 @@ tags:
87
 
88
 
89
 
90
- # Prompting
91
-
92
- ## Prompt Template for alpaca style
93
-
94
- ```
95
- ### Instruction:
96
 
97
- <prompt> (without the <>)
98
-
99
- ### Response:
100
- ```
101
-
102
- ## Sample Code
103
-
104
- ```python
105
- import torch
106
- from transformers import AutoModelForCausalLM, AutoTokenizer
107
-
108
- torch.set_default_device("cuda")
109
-
110
- model = AutoModelForCausalLM.from_pretrained("ibivibiv/strix-rufipes-70b", torch_dtype="auto", device_config='auto')
111
- tokenizer = AutoTokenizer.from_pretrained("ibivibiv/strix-rufipes-70b")
112
-
113
- inputs = tokenizer("### Instruction: Create a plan for developing the game of snake in python using pygame.\n### Response:\n", return_tensors="pt", return_attention_mask=False)
114
-
115
- outputs = model.generate(**inputs, max_length=200)
116
- text = tokenizer.batch_decode(outputs)[0]
117
- print(text)
118
- ```
119
 
120
  ## Citations
121
 
 
10
 
11
  ![img](./strix_rufipes.png)
12
 
13
+ # Prompting
14
+
15
+ ## Prompt Template for alpaca style
16
+
17
+ ```
18
+ ### Instruction:
19
+
20
+ <prompt> (without the <>)
21
+
22
+ ### Response:
23
+ ```
24
+
25
+ ## Sample Code
26
+
27
+ ```python
28
+ import torch
29
+ from transformers import AutoModelForCausalLM, AutoTokenizer
30
+
31
+ torch.set_default_device("cuda")
32
+
33
+ model = AutoModelForCausalLM.from_pretrained("ibivibiv/strix-rufipes-70b", torch_dtype="auto", device_config='auto')
34
+ tokenizer = AutoTokenizer.from_pretrained("ibivibiv/strix-rufipes-70b")
35
+
36
+ inputs = tokenizer("### Instruction: Create a plan for developing the game of snake in python using pygame.\n### Response:\n", return_tensors="pt", return_attention_mask=False)
37
+
38
+ outputs = model.generate(**inputs, max_length=200)
39
+ text = tokenizer.batch_decode(outputs)[0]
40
+ print(text)
41
+ ```
42
+
43
  # Model Details
44
  * **Trained by**: [ibivibiv](https://huggingface.co/ibivibiv)
45
  * **Library**: [HuggingFace Transformers](https://github.com/huggingface/transformers)
46
  * **Model type:** **strix-rufipes-70b** is an auto-regressive language model fine tuned on the Llama 2 transformer architecture.
47
  * **Language(s)**: English
48
+ * **Purpose**: Has specific training for logic enforcement. This model is targeted towards planning exercises.
49
 
50
  # Benchmark Scores
51
 
 
117
 
118
 
119
 
 
 
 
 
 
 
120
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
121
 
122
  ## Citations
123