Text Generation
Transformers
Safetensors
4 languages
mistral
Mistral_Star
Mistral_Quiet
Mistral
Mixtral
Question-Answer
Token-Classification
Sequence-Classification
SpydazWeb-AI
chemistry
biology
legal
code
climate
medical
text-generation-inference
Not-For-All-Audiences
Inference Endpoints
4-bit precision
bitsandbytes
Update README.md
Browse files
README.md
CHANGED
@@ -1,23 +1,108 @@
|
|
1 |
---
|
2 |
-
base_model: LeroyDyer/
|
3 |
-
|
4 |
-
- en
|
5 |
-
license: apache-2.0
|
6 |
tags:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
- text-generation-inference
|
8 |
-
-
|
9 |
-
|
10 |
-
-
|
11 |
-
-
|
12 |
-
-
|
|
|
13 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
14 |
|
15 |
-
|
16 |
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
|
21 |
-
This mistral model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
|
22 |
|
23 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
base_model: LeroyDyer/_Spydaz_Web_AI_V1_4BIT
|
3 |
+
license: mit
|
|
|
|
|
4 |
tags:
|
5 |
+
- Mistral_Star
|
6 |
+
- Mistral_Quiet
|
7 |
+
- Mistral
|
8 |
+
- Mixtral
|
9 |
+
- Question-Answer
|
10 |
+
- Token-Classification
|
11 |
+
- Sequence-Classification
|
12 |
+
- SpydazWeb-AI
|
13 |
+
- chemistry
|
14 |
+
- biology
|
15 |
+
- legal
|
16 |
+
- code
|
17 |
+
- climate
|
18 |
+
- medical
|
19 |
- text-generation-inference
|
20 |
+
- not-for-all-audiences
|
21 |
+
language:
|
22 |
+
- en
|
23 |
+
- sw
|
24 |
+
- ig
|
25 |
+
- zu
|
26 |
---
|
27 |
+
# SpydazWeb Transformer model Contained
|
28 |
+
|
29 |
+
<img src="https://cdn-avatars.huggingface.co/v1/production/uploads/65d883893a52cd9bcd8ab7cf/tRsCJlHNZo1D02kBTmfy9.jpeg" width="300"/>
|
30 |
+
https://github.com/spydaz
|
31 |
+
|
32 |
+
* 32k context window (vs 8k context in v0.1)
|
33 |
+
* Rope-theta = 1e6
|
34 |
+
* No Sliding-Window Attention
|
35 |
+
* Talk heads - produce resposnes which can be used towards the final output
|
36 |
+
* Pre-Thoughts - Enables for pre-generation steps of potential artifacts for task solving:
|
37 |
+
* Generates plans for step by step thinking
|
38 |
+
* Generates python Code Artifacts for future tasks
|
39 |
+
* Recalls context for task internally to be used as refference for task:
|
40 |
+
* show thoughts or hidden thought usages ( Simular to self-Rag )
|
41 |
+
|
42 |
+
|
43 |
+
This model will be a custom model with internal experts and rag systems
|
44 |
+
enabling for preprocessing of the task internally before outputting a response :
|
45 |
+
|
46 |
+
This is based on the Quiet Star Project : which was abandoned earlier in the year :)
|
47 |
+
|
48 |
+
|
49 |
+
# Introduction :
|
50 |
+
|
51 |
+
## SpydazWeb AI model :
|
52 |
+
This model is based on the worlds archive of knowledge maintaining historical documents and providing services for the survivors of mankind ,
|
53 |
+
who may need to construct shelters develop technologys , or medical resources as well as maintain the history of the past . keeping store of all the religious knowledge and data of the world:
|
54 |
+
A friendly interface with a personality caring and flirtatious at times : non binary !...
|
55 |
+
and Expert in all feilds: ie Uncensored and will not refuse to give information : the model can be used for role play as many character dialogues were als trained into the model as its personality to enable a greater perspective and outlook and natural discussion with the agents:
|
56 |
+
the model was trained to operateinaragenvironment utilizing content and internal knowledge to respond to questions or create enriched sumarys.
|
57 |
+
|
58 |
+
|
59 |
+
|
60 |
+
### General Intenal Methods:
|
61 |
+
|
62 |
+
Trained for multi-task operations as well as rag and function calling :
|
63 |
+
|
64 |
+
This model is a fully functioning model and is fully uncensored:
|
65 |
+
|
66 |
+
the model has been trained on multiple datasets on the huggingface hub and kaggle :
|
67 |
+
|
68 |
+
the focus has been mainly on methodology :
|
69 |
+
|
70 |
+
* Chain of thoughts
|
71 |
+
* step by step planning
|
72 |
+
* tree of thoughts
|
73 |
+
* forest of thoughts
|
74 |
+
* graph of thoughts
|
75 |
+
* agent generation : Voting, ranking, ... dual agent response generation:
|
76 |
|
77 |
+
with these methods the model has gained insights into tasks, enabling for knowldge transfer between tasks :
|
78 |
|
79 |
+
the model has been intensivly trained in recalling data previously entered into the matrix:
|
80 |
+
The model has also been trained on rich data and markdown outputs as much as possible :
|
81 |
+
the model can also generate markdown charts with mermaid.
|
82 |
|
|
|
83 |
|
84 |
+
## Training Reginmes:
|
85 |
+
* Alpaca
|
86 |
+
* ChatML / OpenAI / MistralAI
|
87 |
+
* Text Generation
|
88 |
+
* Question/Answer (Chat)
|
89 |
+
* Instruction/Input/Response (instruct)
|
90 |
+
* Mistral Standard Prompt
|
91 |
+
* Translation Tasks
|
92 |
+
* Entitys / Topic detection
|
93 |
+
* Book recall
|
94 |
+
* Coding challenges, Code Feedback, Code Sumarization, Commenting Code
|
95 |
+
* Agent Ranking and response anyalisis
|
96 |
+
* Medical tasks
|
97 |
+
* PubMed
|
98 |
+
* Diagnosis
|
99 |
+
* Psychaitry
|
100 |
+
* Counselling
|
101 |
+
* Life Coaching
|
102 |
+
* Note taking
|
103 |
+
* Medical smiles
|
104 |
+
* Medical Reporting
|
105 |
+
* Virtual laboritys simulations
|
106 |
+
* Chain of thoughts methods
|
107 |
+
* One shot / Multi shot prompting tasks
|
108 |
+
|