Update README.md
Browse files
README.md
CHANGED
|
@@ -69,13 +69,81 @@ library_name: transformers
|
|
| 69 |
tags:
|
| 70 |
- mergekit
|
| 71 |
- merge
|
| 72 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 73 |
---
|
| 74 |
-
# merge
|
| 75 |
|
| 76 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 77 |
|
| 78 |
-
## Merge Details
|
| 79 |
### Merge Method
|
| 80 |
|
| 81 |
This model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using [xdrshjr/llama3.2_1b_uncensored_5000_8epoch_lora](https://huggingface.co/xdrshjr/llama3.2_1b_uncensored_5000_8epoch_lora) as a base.
|
|
|
|
| 69 |
tags:
|
| 70 |
- mergekit
|
| 71 |
- merge
|
| 72 |
+
- llama
|
| 73 |
+
- llama3.2
|
| 74 |
+
- rp
|
| 75 |
+
- roleplay
|
| 76 |
+
- nsfw
|
| 77 |
+
- 1b
|
| 78 |
+
- not-for-all-audiences
|
| 79 |
+
language:
|
| 80 |
+
- es
|
| 81 |
+
- en
|
| 82 |
+
datasets:
|
| 83 |
+
- HuggingFaceTB/smoltalk
|
| 84 |
+
- Guilherme34/uncensor
|
| 85 |
+
- teknium/OpenHermes-2.5
|
| 86 |
+
- passing2961/multifaceted-skill-of-mind
|
| 87 |
+
- PawanKrd/math-gpt-4o-200k
|
| 88 |
+
- V3N0M/Jenna-50K-Alpaca-Uncensored
|
| 89 |
+
- cognitivecomputations/dolphin-coder
|
| 90 |
+
- mlabonne/FineTome-100k
|
| 91 |
+
- microsoft/orca-math-word-problems-200k
|
| 92 |
+
- CarrotAI/ko-instruction-dataset
|
| 93 |
+
- Salesforce/xlam-function-calling-60k
|
| 94 |
+
- anthracite-org/kalo-opus-instruct-22k-no-refusal
|
| 95 |
+
- anthracite-org/stheno-filtered-v1.1
|
| 96 |
+
- anthracite-org/nopm_claude_writing_fixed
|
| 97 |
+
- AiAF/SCPWiki-Archive-02-March-2025-Datasets
|
| 98 |
+
- huihui-ai/QWQ-LONGCOT-500K
|
| 99 |
+
- huihui-ai/LONGCOT-Refine-500K
|
| 100 |
+
- Epiculous/Synthstruct-Gens-v1.1-Filtered-n-Cleaned
|
| 101 |
+
- Epiculous/SynthRP-Gens-v1.1-Filtered-n-Cleaned
|
| 102 |
+
- alexandreteles/AlpacaToxicQA_ShareGPT
|
| 103 |
+
- Nitral-AI/Active_RP-ShareGPT
|
| 104 |
+
- PJMixers/hieunguyenminh_roleplay-deduped-ShareGPT
|
| 105 |
+
- Nitral-AI/RP_Alignment-ShareGPT
|
| 106 |
+
- Chaser-cz/sonnet35-charcard-roleplay-sharegpt
|
| 107 |
+
- AiCloser/sharegpt_cot_dataset
|
| 108 |
+
- PJMixers/Gryphe_Opus-WritingPrompts-Story2Prompt-ShareGPT
|
| 109 |
+
- priveeai/pippa_sharegpt
|
| 110 |
+
- Locutusque/sharegpt_gpt4_uncensored_cleaned
|
| 111 |
+
- OpenCoder-LLM/opc-sft-stage1
|
| 112 |
+
- OpenCoder-LLM/opc-sft-stage2
|
| 113 |
+
- microsoft/orca-agentinstruct-1M-v1
|
| 114 |
+
- NousResearch/hermes-function-calling-v1
|
| 115 |
+
- AI-MO/NuminaMath-CoT
|
| 116 |
+
- AI-MO/NuminaMath-TIR
|
| 117 |
+
- allenai/tulu-3-sft-mixture
|
| 118 |
+
- cognitivecomputations/samantha-data
|
| 119 |
+
- m-a-p/CodeFeedback-Filtered-Instruction
|
| 120 |
+
- m-a-p/Code-Feedback
|
| 121 |
+
- FreedomIntelligence/medical-o1-reasoning-SFT
|
| 122 |
---
|
|
|
|
| 123 |
|
| 124 |
+
# PURGA
|
| 125 |
+
|
| 126 |
+
Este modelo fue purgado. La versión anterior contaba con errores debido a dos modelos que causaban asquerosos errores a la mezcla.
|
| 127 |
+
|
| 128 |
+
## Bahamuth 3.2 1B
|
| 129 |
+
<center>
|
| 130 |
+
<img src="https://i.ibb.co/HTpGdMQg/Behemoth.webp" alt="Behemoth" border="0">
|
| 131 |
+
</center>
|
| 132 |
+
|
| 133 |
+
|
| 134 |
+
*"Mira a Bahamut, criatura mía igual que tú"*
|
| 135 |
+
|
| 136 |
+
~ Job 40:10-19
|
| 137 |
+
|
| 138 |
+
|
| 139 |
+
Metafóricamente, su nombre ha llegado a ser usado para connotar algo extremadamente grande o poderoso.
|
| 140 |
+
|
| 141 |
+
**CARACTERISTICAS:**
|
| 142 |
+
- Más de 50 modelos mezclados.
|
| 143 |
+
- 39 datasets.
|
| 144 |
+
|
| 145 |
+
---
|
| 146 |
|
|
|
|
| 147 |
### Merge Method
|
| 148 |
|
| 149 |
This model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using [xdrshjr/llama3.2_1b_uncensored_5000_8epoch_lora](https://huggingface.co/xdrshjr/llama3.2_1b_uncensored_5000_8epoch_lora) as a base.
|