PocketDoc commited on
Commit
2238c16
1 Parent(s): 6975d33

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +181 -0
README.md ADDED
@@ -0,0 +1,181 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: apache-2.0
4
+ language:
5
+ - en
6
+ base_model:
7
+ - mistralai/Mistral-Nemo-Base-2407
8
+ - Dans-DiscountModels/Mistral-Nemo-Base-2407-ChatML-Mod
9
+ tags:
10
+ - general-purpose
11
+ - roleplay
12
+ - storywriting
13
+ - chemistry
14
+ - biology
15
+ - code
16
+ - climate
17
+ - axolotl
18
+ - text-generation-inference
19
+ - finetune
20
+ datasets:
21
+ - PocketDoc/Dans-MemoryCore-CoreCurriculum-Small
22
+ - AquaV/Energetic-Materials-Sharegpt
23
+ - AquaV/Chemical-Biological-Safety-Applications-Sharegpt
24
+ - AquaV/US-Army-Survival-Sharegpt
25
+ - AquaV/Resistance-Sharegpt
26
+ - AquaV/Interrogation-Sharegpt
27
+ - AquaV/Multi-Environment-Operations-Sharegpt
28
+ - PocketDoc/Dans-Mathmaxx
29
+ - PocketDoc/Dans-Mathmaxx-Numina-CoT
30
+ - PJMixers/Math-Multiturn-1K-ShareGPT
31
+ - PocketDoc/Dans-Benchmaxx
32
+ - PocketDoc/Dans-Benchmaxx-COT
33
+ - PocketDoc/Dans-Codemaxx-LeetCode
34
+ - PocketDoc/Dans-Codemaxx-CodeFeedback-Conversations
35
+ - PocketDoc/Dans-Codemaxx-CodeFeedback-SingleTurn
36
+ - PocketDoc/Dans-Codemaxx-Bigcode-SelfInstruct
37
+ - PocketDoc/Dans-Taskmaxx
38
+ - PocketDoc/Dans-Taskmaxx-DataPrepper
39
+ - PocketDoc/Dans-Taskmaxx-ConcurrentQA-Reworked
40
+ - PocketDoc/Dans-Taskmaxx-TableGPT
41
+ - PocketDoc/Dans-Taskmaxx-SciRIFF
42
+ - PocketDoc/Dans-Taskmaxx-Edit
43
+ - PocketDoc/Dans-Systemmaxx
44
+ - PocketDoc/Dans-Toolmaxx-Agent
45
+ - PocketDoc/Dans-Toolmaxx-ShellCommands
46
+ - PocketDoc/Dans-Toolmaxx-Functions-Toolbench
47
+ - PocketDoc/Dans-Toolmaxx-Functions-ToolACE
48
+ - PocketDoc/Dans-Toolmaxx-Functions-apigen
49
+ - PocketDoc/Dans-ASCIIMaxx-Wordart
50
+ - PocketDoc/Dans-Prosemaxx-Gutenberg
51
+ - PocketDoc/Dans-Prosemaxx-Cowriter-M
52
+ - PocketDoc/Dans-Prosemaxx-Adventure
53
+ - PocketDoc/Dans-Prosemaxx-Gryphe-GPT4o-WritingPrompts
54
+ - PocketDoc/Dans-Assistantmaxx-Sharegpt
55
+ - PocketDoc/Dans-Assistantmaxx-OpenAssistant2
56
+ - PocketDoc/Dans-Assistantmaxx-Opus-Merge
57
+ - PocketDoc/Dans-Assistantmaxx-sonnetorca-subset
58
+ - PocketDoc/Dans-Assistantmaxx-sonnetorca-subset-2
59
+ - PocketDoc/Dans-Assistantmaxx-NoRobots
60
+ - PocketDoc/Dans-Assistantmaxx-Synthia
61
+ - PocketDoc/Dans-Assistantmaxx-ASL
62
+ - PocketDoc/Dans-Assistantmaxx-PersonaLLM-Opus
63
+ - PocketDoc/Dans-Assistantmaxx-UnnaturalInstructions-GPT4
64
+ - PocketDoc/Dans-Assistantmaxx-LongAlign
65
+ - PocketDoc/Dans-Assistantmaxx-EvolKit
66
+ - PocketDoc/Dans-Assistantmaxx-Camel-GPT4
67
+ - PocketDoc/Dans-Assistantmaxx-Tulu3-IF
68
+ - PocketDoc/Dans-Logicmaxx-Skunkworks
69
+ - PocketDoc/Dans-Logicmaxx-SAT-AP
70
+ - PocketDoc/Dans-Logicmaxx-Magpie-Ultra
71
+ - PJMixers/grimulkan_theory-of-mind-ShareGPT
72
+ - PJMixers/grimulkan_physical-reasoning-ShareGPT
73
+ - PocketDoc/Dans-Personamaxx
74
+ - PocketDoc/Dans-Personamaxx-Rainy
75
+ - PocketDoc/Dans-Personamaxx-Aesir
76
+ - PocketDoc/Dans-Kinomaxx-VanillaBackrooms
77
+ model-index:
78
+ - name: Mistral-12b-Test-V0.0.3
79
+ results: []
80
+ pipeline_tag: text-generation
81
+ ---
82
+ ## What is it?
83
+
84
+ This model series is intended to be multifarious in its capabilities and should be quite capable at both co-writing and roleplay as well as find itself quite at home performing sentiment analysis or summarization as part of a pipeline. It has been trained on a wide array of one shot instructions, multi turn instructions, role playing scenarios, text adventure games, co-writing, and much more. The full dataset is publicly available and can be found in the datasets section of the model page.
85
+
86
+ There has not been any form of harmfulness alignment done on this model, please take the appropriate precautions when using it in a production environment.
87
+
88
+ **Expected usable context length:** 32768 tokens
89
+
90
+
91
+ ## Prompting
92
+
93
+ The model has been trained on standard "ChatML" format prompting, an example of which is shown below:
94
+
95
+ ```
96
+ <|im_start|>system
97
+ system prompt<|im_end|>
98
+ <|im_start|>user
99
+ Hi there!<|im_end|>
100
+ <|im_start|>assistant
101
+ Nice to meet you!<|im_end|>
102
+ <|im_start|>user
103
+ Can I ask a question?<|im_end|>
104
+ <|im_start|>assistant
105
+ ```
106
+
107
+
108
+ ## SillyTavern templates
109
+
110
+ Below are Instruct and Context templates for use within SillyTavern.
111
+
112
+ <details><summary>context template</summary>
113
+
114
+ ```yaml
115
+ {
116
+ "story_string": "<|im_start|>system\n{{#if system}}{{system}}\n{{/if}}{{#if wiBefore}}{{wiBefore}}\n{{/if}}{{#if description}}{{description}}\n{{/if}}{{#if personality}}{{char}}'s personality: {{personality}}\n{{/if}}{{#if scenario}}Scenario: {{scenario}}\n{{/if}}{{#if wiAfter}}{{wiAfter}}\n{{/if}}{{#if persona}}{{persona}}\n{{/if}}{{trim}}<|im_end|>\n",
117
+ "example_separator": "",
118
+ "chat_start": "",
119
+ "use_stop_strings": false,
120
+ "allow_jailbreak": false,
121
+ "always_force_name2": false,
122
+ "trim_sentences": false,
123
+ "include_newline": false,
124
+ "single_line": false,
125
+ "name": "Dan-ChatML"
126
+ }
127
+ ```
128
+ </details><br>
129
+ <details><summary>instruct template</summary>
130
+
131
+ ```yaml
132
+ {
133
+ "system_prompt": "Write {{char}}'s actions and dialogue, user will write {{user}}'s.",
134
+ "input_sequence": "<|im_start|>user\n",
135
+ "output_sequence": "<|im_start|>assistant\n",
136
+ "first_output_sequence": "",
137
+ "last_output_sequence": "",
138
+ "system_sequence_prefix": "",
139
+ "system_sequence_suffix": "",
140
+ "stop_sequence": "<|im_end|>",
141
+ "wrap": false,
142
+ "macro": true,
143
+ "names": false,
144
+ "names_force_groups": false,
145
+ "activation_regex": "",
146
+ "skip_examples": false,
147
+ "output_suffix": "<|im_end|>\n",
148
+ "input_suffix": "<|im_end|>\n",
149
+ "system_sequence": "<|im_start|>system\n",
150
+ "system_suffix": "<|im_end|>\n",
151
+ "user_alignment_message": "",
152
+ "last_system_sequence": "",
153
+ "system_same_as_user": false,
154
+ "first_input_sequence": "",
155
+ "last_input_sequence": "",
156
+ "name": "Dan-ChatML"
157
+ }
158
+ ```
159
+ </details><br>
160
+
161
+
162
+ ## Example outputs
163
+
164
+ <details><summary>Prompt</summary>
165
+ > <|im_start|>system
166
+ > Take the users messages and use them as a writing prompt. Write 5 sentences.<|im_end|>
167
+ > <|im_start|>user
168
+ > Alpha Centauri has fallen, to the bears. This is a point of view tale about a soldier on the ground.<|im_end|>
169
+ > <|im_start|>assistant
170
+ </details><br>
171
+
172
+ > "Sir, they are gone sir." The young corporal was out of breath.
173
+ >
174
+ > The soldier looked up. His grey eyes had been locked on his holofac display for hours now. He was in his late thirties and wore the rank of colonel. He was tall, with a well-muscled but slightly over weight body.
175
+
176
+
177
+ ## Training
178
+
179
+ This model was full finetuned for 2 epochs on 1x H200 SXM equating to 88 hours of training.
180
+
181
+ [<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)