TroyDoesAI
Llama-3.1-8B-Instruct
BlackSheep AFM 4.5B Q8 0 GGUF
MermaidMistral
BlackSheep-Llama3.2-3B
BlackSheep-24B
BlackSheep-24B-Q6_K
Llama-3.1-13B-Instruct
RAG-Qwen2.5-7B
BlackSheep-Llama3.2-3B-Context_Obedient-q4_k_m
gpt-oss-4B
BlackSheep-X-Dolphin
BlackSheep-16k-f16-RAM-14.5GB-gguf
Codestral-22B-RAG-Q8-gguf
BlackSheep-Llama3.2-5B-Q4_K_M
MermaidStable3B
BlackSheep-8k-f16-RAM-11GB-gguf
Qwen3-MoE-3B
BlackSheep-32k-f16-gguf
BlackSheep-4k-f16-RAM-9GB-gguf
Phi 3 Context Obedient RAG
Overview This model is meant to enhance adherence to provided context (e.g., for RAG applications) and reduce hallucinations, inspired by airoboros context-obedient question answer format. I know it's a bit verbose and annoying, but after much trial and error, using these explicit delimiters helps the model understand where to find the responses and how to associate specific sources with it. - `BEGININPUT` - denotes a new input block - `BEGINCONTEXT` - denotes the block of context (metadata key/value pairs) to associate with the current input block - `ENDCONTEXT` - denotes the end of the metadata block for the current input - [text] - Insert whatever text you want for the input block, as many paragraphs as can fit in the context. - `ENDINPUT` - denotes the end of the current input block - [repeat as many input blocks in this format as you want] - `BEGININSTRUCTION` - denotes the start of the list (or one) instruction(s) to respond to for all of the input blocks above. - [instruction(s)] - `ENDINSTRUCTION` - denotes the end of instruction set Here's a trivial, but important example to prove the point: As shown in the example, the dataset includes many examples of including source details in the response, when the question asks for source/citation/references. Why do this? Well, the R in RAG seems to be the weakest link in the chain. Retrieval accuracy, depending on many factors including the overall dataset size, can be quite low. This accuracy increases when retrieving more documents, but then you have the issue of actually using the retrieved documents in prompts. If you use one prompt per document (or document chunk), you know exactly which document the answer came from, so there's no issue. If, however, you include multiple chunks in a single prompt, it's useful to include the specific reference chunk(s) used to generate the response, rather than naively including references to all of the chunks included in the prompt. If the question being asked is `What color is the cat?`, I would only expect the 2nd document to be referenced in the response, as the other link is irrelevant.
UncensoredLM
Mermaid-Dolphin-MoE-2x7b_Story_Flow
Unrestricted-Knowledge-Will-Not-Refuse-15B
Agent-Flow-Phone_Demo_3GB_RAM
DigitalSoul-BlackSheep
BlackSheep-4k-Q6_K-RAM-5GB-gguf
Mini-Moo
TinyLlama-RAG
BlackSheep-8k-Q6_K-RAM-7GB-gguf
BlackSheep-Llama3.2-3B-Context_Obedient
Moo
BlackSheep-MermaidMistral-22B
BlackSheep-32k-Q6_K-gguf
BlackSheep-3.8B
Mermaid-Llama-6.7B-RAG
Mermaid-Llama-3-8B
CreativeWriter-Personality-12B
Mermaid-Llama-6.7B-RAG-Code-Instruct
MermaidMixtral-2x6.5b
Tiny-RAG-gguf
BlackSheep-16k-Q6_K-RAM-10.3GB-gguf
AgentFlow-3B
Mermaid-22B
BlackSheep-27.7B
BlackSheep-Qwen-14B
This Little DigitalSoul has all the gaurdrails removed, but no longer overly willing to push the limits unless you really ask for it. This new continuous training technique with the addition of ablation to reduce the toxicity post training has created BlackSheep's DigitalSoul without all the wild, untamed, or rude behavior that was once associated with its younger self. Use Alpaca Format and give me some feedback on it's responses