asiansoul commited on
Commit
c65696f
1 Parent(s): 12bdc21

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +35 -5
README.md CHANGED
@@ -25,12 +25,42 @@ tags:
25
 
26
  ✨ *Navigating the High Seas of Data: Crafting the Ultimate Yacht Insights with Merged LLMs* ✨
27
 
28
- ## Merge Details
29
- ### Merge Method
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
30
 
31
  This model was merged using the [DARE](https://arxiv.org/abs/2311.03099) [TIES](https://arxiv.org/abs/2306.01708) merge method using [NousResearch/Meta-Llama-3-8B](https://huggingface.co/NousResearch/Meta-Llama-3-8B) as a base.
32
 
33
- ### Models Merged
34
 
35
  The following models were included in the merge:
36
  * [NousResearch/Hermes-2-Pro-Llama-3-8B](https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B)
@@ -43,7 +73,7 @@ The following models were included in the merge:
43
  * [Danielbrdz/Barcenas-Llama3-8b-ORPO](https://huggingface.co/Danielbrdz/Barcenas-Llama3-8b-ORPO)
44
  * [aaditya/Llama3-OpenBioLLM-8B](https://huggingface.co/aaditya/Llama3-OpenBioLLM-8B)
45
 
46
- ### Configuration
47
 
48
  The following YAML configuration was used to produce this model:
49
 
@@ -104,4 +134,4 @@ parameters:
104
  dtype: bfloat16
105
 
106
 
107
- ```
 
25
 
26
  ✨ *Navigating the High Seas of Data: Crafting the Ultimate Yacht Insights with Merged LLMs* ✨
27
 
28
+ ## 🏟️ Merged Model Series Yacht Features
29
+
30
+ Welcome to the merged model series yacht! This provides an overview of the powerful features and functionalities that this series brings together, akin to a sleek, modern yacht sailing across the digital ocean.
31
+
32
+ ### 1. Function Calling & JSON Outputs
33
+ - Offers precise function calling and structured JSON outputs via specialized tokens like `<tools>`, `<tool_call>`, and `<tool_response>`. Streamlines system communication for developers.
34
+
35
+ ### 2. Conversational Interaction
36
+ - Avoids excessive "SYSTEM MESSAGE" chatter while delivering seamless, friendly dialogue.
37
+ - Specializes in answering questions with precision, handling arithmetic and tabular data effortlessly.
38
+
39
+ ### 3. Expanded Context Length
40
+ - Extends the context length to 256k tokens using PoSE, offering a broader field of data analysis.
41
+
42
+ ### 4. Multilingual Capabilities
43
+ - Transfers instruction-following from English to Korean for reliable interaction across languages.
44
+
45
+ ### 5. Optimized Dialogue & Safety
46
+ - Aligns with human preferences using fine-tuning (SFT) and reinforcement learning (RLHF), ensuring helpful and safe dialogues.
47
+
48
+ ### 6. Precision Merging
49
+ - Merges foundational and preview models for Korean language through task arithmetic, providing seamless integration.
50
+
51
+ ### 7. Specialized Biomedical Knowledge
52
+ - Specializes in biomedical tasks with accurate responses for healthcare professionals and researchers.
53
+
54
+ ### 8. Novel Training & Collaboration
55
+ - Combines ORPO methods and dolphin preference datasets for high-quality conversation and collaboration.
56
+
57
+ The merged model series yacht offers unparalleled functionality, drawing together a fleet of specialized models. Whether you need precise function calling, multilingual capabilities, or conversational AI, this yacht has every deck optimized to navigate the digital ocean with style and precision.
58
+
59
+ ## 👘 Merge Method
60
 
61
  This model was merged using the [DARE](https://arxiv.org/abs/2311.03099) [TIES](https://arxiv.org/abs/2306.01708) merge method using [NousResearch/Meta-Llama-3-8B](https://huggingface.co/NousResearch/Meta-Llama-3-8B) as a base.
62
 
63
+ ## 🩱 Models Merged
64
 
65
  The following models were included in the merge:
66
  * [NousResearch/Hermes-2-Pro-Llama-3-8B](https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B)
 
73
  * [Danielbrdz/Barcenas-Llama3-8b-ORPO](https://huggingface.co/Danielbrdz/Barcenas-Llama3-8b-ORPO)
74
  * [aaditya/Llama3-OpenBioLLM-8B](https://huggingface.co/aaditya/Llama3-OpenBioLLM-8B)
75
 
76
+ ## 🪭 Configuration
77
 
78
  The following YAML configuration was used to produce this model:
79
 
 
134
  dtype: bfloat16
135
 
136
 
137
+ ```