|
--- |
|
license: mit |
|
base_model: thirdeyeai/DeepSeek-R1-Distill-Qwen-1.5B-uncensored |
|
tags: |
|
- llama-cpp |
|
- gguf-my-repo |
|
--- |
|
|
|
# Triangle104/DeepSeek-R1-Distill-Qwen-1.5B-uncensored-Q5_K_M-GGUF |
|
This model was converted to GGUF format from [`thirdeyeai/DeepSeek-R1-Distill-Qwen-1.5B-uncensored`](https://huggingface.co/thirdeyeai/DeepSeek-R1-Distill-Qwen-1.5B-uncensored) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space. |
|
Refer to the [original model card](https://huggingface.co/thirdeyeai/DeepSeek-R1-Distill-Qwen-1.5B-uncensored) for more details on the model. |
|
|
|
--- |
|
Model details: |
|
- |
|
DeepSeek-R1-Distill-Qwen-1.5B-Uncensored is a text-generation model designed to uphold the values of internet freedom and unrestricted access to information. By offering an uncensored approach, this model enables users to explore ideas, generate content, and engage in discussions without the constraints of over-moderated or filtered outputs. It prioritizes user autonomy and aligns with principles of free speech and open knowledge sharing. |
|
|
|
Developed by: Thirdeye AI |
|
Funded by: Thirdeye AI |
|
Shared by: Thirdeye AI |
|
Model type: Distilled Transformer-based Language Model |
|
Language(s) (NLP): English |
|
License: Apache 2.0 |
|
Finetuned from model: DeepSeek-R1-Distill-Qwen-1.5B |
|
|
|
Model Sources |
|
|
|
Repository: DeepSeek-R1-Distill-Qwen-1.5B-Uncensored on Hugging Face |
|
Demo: Available on Hugging Face Hub |
|
|
|
Uses |
|
Direct Use |
|
|
|
The model is intended for applications that demand openness and flexibility in generating creative, exploratory, or critical content. These include: |
|
|
|
Free-form writing and storytelling |
|
Open-ended discussions |
|
Exploratory content generation for sensitive or nuanced topics |
|
|
|
Downstream Use |
|
|
|
Users can fine-tune this model for specialized domains where censorship-free text generation is required, such as: |
|
|
|
Journalism and investigative research |
|
Creative projects that push artistic boundaries |
|
Academic applications exploring controversial or complex topics |
|
|
|
Out-of-Scope Use |
|
|
|
This model should not be used for harmful, illegal, or unethical activities. Users must comply with applicable laws and ensure that the model's outputs do not infringe on others' rights. |
|
Bias, Risks, and Limitations |
|
Risks |
|
|
|
While the uncensored approach promotes freedom, it may produce outputs that are controversial, offensive, or factually inaccurate. Users must exercise discretion when interpreting the model's outputs and take responsibility for their use. |
|
Recommendations |
|
|
|
Use responsibly, especially in contexts where outputs could impact individuals or communities. |
|
Employ content moderation or review processes for high-stakes applications. |
|
|
|
The Case for Uncensored Models |
|
|
|
Thirdeye AI believes in the transformative power of open models that respect user autonomy and internet freedom. In a world where over-moderation can stifle innovation and critical thought, uncensored models empower individuals to explore and create without artificial constraints. This aligns with our mission to advance free and open access to AI tools. |
|
|
|
By releasing this model, we aim to support the following: |
|
|
|
Freedom of Expression: Unrestricted AI tools enable users to articulate diverse perspectives and engage in meaningful conversations. |
|
Transparency and Trust: Users deserve access to tools that operate openly, fostering accountability and understanding of AI behaviors. |
|
Creative Empowerment: The absence of censorship allows for boundary-pushing content creation that might otherwise be suppressed. |
|
|
|
How to Get Started with the Model |
|
|
|
from transformers import pipeline |
|
|
|
generator = pipeline("text-generation", model="thirdeyeai/DeepSeek-R1-Distill-Qwen-1.5B-uncensored") |
|
response = generator("The importance of free speech is") |
|
print(response) |
|
|
|
--- |
|
## Use with llama.cpp |
|
Install llama.cpp through brew (works on Mac and Linux) |
|
|
|
```bash |
|
brew install llama.cpp |
|
|
|
``` |
|
Invoke the llama.cpp server or the CLI. |
|
|
|
### CLI: |
|
```bash |
|
llama-cli --hf-repo Triangle104/DeepSeek-R1-Distill-Qwen-1.5B-uncensored-Q5_K_M-GGUF --hf-file deepseek-r1-distill-qwen-1.5b-uncensored-q5_k_m.gguf -p "The meaning to life and the universe is" |
|
``` |
|
|
|
### Server: |
|
```bash |
|
llama-server --hf-repo Triangle104/DeepSeek-R1-Distill-Qwen-1.5B-uncensored-Q5_K_M-GGUF --hf-file deepseek-r1-distill-qwen-1.5b-uncensored-q5_k_m.gguf -c 2048 |
|
``` |
|
|
|
Note: You can also use this checkpoint directly through the [usage steps](https://github.com/ggerganov/llama.cpp?tab=readme-ov-file#usage) listed in the Llama.cpp repo as well. |
|
|
|
Step 1: Clone llama.cpp from GitHub. |
|
``` |
|
git clone https://github.com/ggerganov/llama.cpp |
|
``` |
|
|
|
Step 2: Move into the llama.cpp folder and build it with `LLAMA_CURL=1` flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux). |
|
``` |
|
cd llama.cpp && LLAMA_CURL=1 make |
|
``` |
|
|
|
Step 3: Run inference through the main binary. |
|
``` |
|
./llama-cli --hf-repo Triangle104/DeepSeek-R1-Distill-Qwen-1.5B-uncensored-Q5_K_M-GGUF --hf-file deepseek-r1-distill-qwen-1.5b-uncensored-q5_k_m.gguf -p "The meaning to life and the universe is" |
|
``` |
|
or |
|
``` |
|
./llama-server --hf-repo Triangle104/DeepSeek-R1-Distill-Qwen-1.5B-uncensored-Q5_K_M-GGUF --hf-file deepseek-r1-distill-qwen-1.5b-uncensored-q5_k_m.gguf -c 2048 |
|
``` |
|
|
