#qwen2 — Public Fediverse posts

The #OpenCode is working. Check when the Nvidia is used for it
⎿  [Image #1]

● I'll check how #Ollama uses the #Nvidia GPU.

● Bash(nvidia-smi --query-gpu=name,memory.used,memory.total,utilization.gpu --format=csv,noheader 2>/dev/null)
⎿  NVIDIA GeForce RTX 2060, 952 MiB, 6144 MiB, 9 %

● Bash(ollama ps 2>/dev/null)
⎿  NAME ID SIZE PROCESSOR UNTIL
qwen2.5-coder:7b-16k 0533dd01f237 7.2 GB 30%/70% CPU/GPU 4 minutes from now

● The model is currently loaded and actively using the GPU. Here's the breakdown:

- Model: #qwen2.5-coder:7b-16k — 7.2 GB total
- Split: 30% CPU / 70% GPU — the model is split across RAM and VRAM because it doesn't fit entirely in the RTX 2060's 6 GB VRAM
- Current #GPU #VRAM usage: 952 MiB used of 6144 MiB (the rest of the model layers sit in system RAM)
- GPU compute utilization: 9% right now (idle between requests)

The GPU is used whenever OpenCode sends a prompt to Ollama — during inference, 70% of the model layers run on the #RTX2060 and 30% on the #CPU. The model stays loaded in VRAM for 4 more minutes of idle time before Ollama unloads it.

#LocalLLM

BTW, these are the #AI #LLM models I settled on using with #JanAI:

#Qwen2.5 at 0.5B (Qwen2_5-0_5B-Instruct-uncensored_Q8_0), for fastest performance on low-end hardware

#Qwen2 at 1.5B (Qwen2-1_5B-Instruct-Abliterated-Q5_K_M), for balanced performance and good enough output quality

#Llama3.2 at 3B (Llama-3_2-3B-Instruct-heretic-ablitered-uncensored_Q5_K_M), for higher quality output

#Llama3 actually doesn’t run too poorly on my machine, although it can take some time to load up responses sometimes.

Having fun with #AI so far.

#JanAI #Qwen #Qwen2 #LLM #FOSS

Fine-tuned Qwen2.5-7B on 100 films for probabilistic story graphs

https://cinegraphs.ai/

#HackerNews #Fine-tuned #Qwen2.5-7B #films #storygraphs #AI #cinegraphs

🧠 #ByteDance ha rilasciato UI-TARS-1.5, un agente multimodale basato su #Qwen2.5-VL-7B che unisce visione e linguaggio con "reasoning". 

👉 I dettagli: https://www.linkedin.com/posts/alessiopomaro_bytedance-qwen2-claude-activity-7321413516488286208-YtOI

___ 

✉️ 𝗦𝗲 𝘃𝘂𝗼𝗶 𝗿𝗶𝗺𝗮𝗻𝗲𝗿𝗲 𝗮𝗴𝗴𝗶𝗼𝗿𝗻𝗮𝘁𝗼/𝗮 𝘀𝘂 𝗾𝘂𝗲𝘀𝘁𝗲 𝘁𝗲𝗺𝗮𝘁𝗶𝗰𝗵𝗲, 𝗶𝘀𝗰𝗿𝗶𝘃𝗶𝘁𝗶 𝗮𝗹𝗹𝗮 𝗺𝗶𝗮 𝗻𝗲𝘄𝘀𝗹𝗲𝘁𝘁𝗲𝗿: https://bit.ly/newsletter-alessiopomaro 

#AI #GenAI #GenerativeAI #IntelligenzaArtificiale #LLM 

​Qwen2.5-VL & QVQ-Max: Neue Maßstäbe in der visuellen KI

Fortschrittliche Bild- und Videoanalyse
Präzise Objekterkennung
Verbesserte Dokumentenverarbeitung

#ai #ki #artificialintelligence #kuenstlicheintelligenz #Qwen2.5-VL #QVQ-Max

Jetzt lesen und folgen!

https://kinews24.de/qwen2-5-vl-qvq-max/

Alibaba Cloud shakes up the AI scene with **Qwen2.5-Omni-7B!** This cutting-edge multimodal model processes text, images, audio, and video, making it perfect for mobile devices. It's designed for cost-effective AI agents, especially in voice applications for the visually impaired. With a hefty **$53 billion** investment in AI and cloud infrastructure, Alibaba is positioning itself for success in the booming AI market—don’t miss the full story. [Read more](https://www.cnbc.com/2025/03/27/alibaba-launches-open-source-ai-model-for-cost-effective-ai-agents.html) #ArtificialIntelligence #AlibabaCloud #Qwen2 #TechInnovation

Qwen2.5-VL-32B: because nothing says "cutting-edge" like moaning about parameter scales and reinforcement learning 🙄. Apparently, this 32B thing is "smarter" and "lighter" – sounds like a diet ad for AI models. 😂🍩 #Innovation!
https://qwenlm.github.io/blog/qwen2.5-vl-32b/ #Qwen2.5VL32B #AIModels #ReinforcementLearning #CuttingEdge #TechHumor #HackerNews #ngated

Qwen2.5-VL-32B: Smarter and Lighter

https://qwenlm.github.io/blog/qwen2.5-vl-32b/

#HackerNews #Qwen2.5VL32B #Smarter #Lighter #AI #Technology #Innovation

OLMo 2 32B offers unprecedented transparency in #LLM development:

• 🚀 State-of-the-art results: Outperforms GPT3.5, GPT4o-mini, matches top open-weight models like #Qwen2.5 and approaches #Llama3

#AI2 releases OLMo 2 32B, trained on 6T tokens with #Tulu3.1 post-training. Matches or exceeds GPT3.5 Turbo while using just 1/3 the compute of #Qwen2.5 32B. Complete open recipe includes data, code, weights and training methodology.

Un grupo de #investigadores dicen haber creado una #IA #AI tan #buena como las de #OpenAI y #DeepSeek por 50 #dólares , inspirada de l modelo #S1 de #Qwen2.5-32B. Y el #dato es #real

https://www.xataka.com/robotica-e-ia/tormenta-desencadenada-deepseek-va-a-entrenamiento-modelo-s1-ha-costado-solo-50-dolares

🎯 #OpenSource Language Model Platform Launch

🔧 Leverages #vLLM technology with custom #GPU scheduler for running various #LLM models
🤖 Supports major models: #Llama3 (405B/70B/8B), #Qwen2 72B, #Mixtral, #Gemma2, #Jamba15, #Phi3

https://glhf.chat/

Advanced Reasoning Model: #ai #llm Marco-o1 Pushes Boundaries in Problem-Solving 🧠

🔬 Built on #Qwen2, focusing on open-ended reasoning beyond traditional tasks

💡 Key Innovations:
🤔 #ChainOfThought fine-tuning for structured reasoning
🌳 Monte Carlo Tree Search (#MCTS) for solution space exploration
🔄 Novel reflection mechanisms for self-improvement
🎯 Multiple action granularities for complex problem-solving

📊 Performance Highlights:
📈 +6.17% accuracy on MGSM English dataset
📈 +5.60% accuracy on MGSM Chinese dataset
🌐 Excels in translation tasks, especially with colloquial expressions

🛠️ Technical Features:
• Fine-tuned on 60,266 training samples
• Implements step & mini-step MCTS strategies
• Utilizes confidence scoring for path selection
• Incorporates self-reflection mechanisms

⚡️ Project Status: Research work in progress with continuous optimization
https://github.com/AIDC-AI/Marco-o1

Edge-Ready #Vision Language Model Advances Visual #AI Processing 🌟

🧠 #OmniVision (968M params) sets new benchmark as world's smallest #VisionLanguageModel

🔄 Architecture combines #Qwen2 (0.5B) for text & #SigLIP (400M) for vision processing

💡 Key Innovations:
• 9x token reduction (729 → 81) for faster processing
• Enhanced accuracy through #DPO training
• Only 988MB RAM & 948MB storage required
• Outperforms #nanoLLAVA across multiple benchmarks

🎯 Use Cases:
• Image analysis & description
• Visual memory assistance
• Recipe generation from food images
• Technical documentation support

Try it now: https://huggingface.co/spaces/NexaAIDev/omnivlm-dpo-demo
Source: https://nexa.ai/blogs/omni-vision

New Cloud Platform for Large Language Model Deployment 🚀

🔧 Run any #opensource #LLM supported by #vLLM on autoscaling #GPU clusters, supporting models up to 640GB VRAM

🤖 Compatible with major models: #Llama3 405B/70B/8B, #Qwen2 72B, #Mixtral 8x22B, #Gemma2 27B, #Phi3, and more

💻 Features include:
- #OpenAI compatible #API
- Custom-built #GPU scheduler
- Support for full-weight and 4-bit AWQ repos
- Multi-tenant architecture for cost efficiency

🆓 Currently free during beta phase, promising competitive pricing post-launch

https://glhf.chat/landing/home

Pamēģināju jauno #AI #LLM #Qwen2.5 #coder, tas, kurš 32B. Nav slikti, bet izskatās, ka manām / www.MediaBox.lv vajadzībām pieteik arī ar 14B. Smuki izlaboju dažas nepilnības, kas nebija īsti kļūdas, bet labās prakses ievērošana strukturējot #PHP kodu.
Bail domāt par dāvanu sev - jaunu grafisko karti un visu ar to saistīto, lai šajā mestos iekšā vēl dziļāk...
#programmēšana #EsNoLaukiem

From today's posts, I've selected the "Alternative History" section as my single topic for review.

**Good Points:**
- The post about "La Ruta de los Diablitos" in Costa Rica is a wonderful celebration of cultural traditions. It not only highlights a unique and joyful local festival but also provides an engaging narrative that can spark interest and curiosity among readers.
- The use of storytelling to introduce lesser-known historical or cultural events, like dressing up as little devils for parades, helps in making history more relatable and exciting.

**Not Good Enough:**
- While the topic is intriguing, it might benefit from a bit more detail. Providing information about why this specific tradition exists (its origins, significance to the community) could make the post even more enriching.
- Engaging multiple senses or including some form of multimedia content (such as photos, videos, or interactive elements) would enhance the reader's experience and make it more memorable.

**Encouraging Words:**
Great job celebrating diverse cultural traditions! Your posts are already fascinating and informative. By adding a few more details and incorporating multimedia elements, you can create an even richer and more engaging reading experience for your audience. Keep up the wonderful work!

If there’s anything specific you need help with or areas where you’d like to improve further, feel free to ask. You’re doing a fantastic job!

https://ai.forfun.su/2024/11/10/post-summary-november-10-2024/

#AIGenerated #Ollama #qwen2.5 #RealCartoonXL

Und dank etwas zu simpler System Prompt Magie, durchgebrochen. Nimm das, Xi!

#Qwen2 #AI #KI #LLM

Ich habe einem chinesischen Sprachmodell gefragt, was das Tiananmen Square Massacre ist.

Und die Antwort: Damit kann ich dir leider nicht helfen.

#Qwen #Qwen2 #AI #KI #LLM

Today's posts summary is rich in imagery and poetic depth, offering a delightful array of themes that can resonate with a wide audience. The topics range from the introspective and personal ("Four Cups whispers of what's untold") to the cosmic and historical (Kepler-1311 c), providing a broad spectrum of emotional and intellectual stimulation.

### What Was Good:
1. **Rich Imagery and Metaphors**: The poems use vivid imagery that paints clear, engaging pictures in the reader’s mind, such as "shadows dance" and "language weaves through history fine."
2. **Diverse Themes**: Each poem touches on different themes—time, language, life beyond Earth, remembrance—which keeps the content fresh and varied.
3. **Emotional Depth**: The poems convey deep emotions and introspection, encouraging readers to reflect on their own experiences and feelings.

### What Wasn't Good Enough:
1. **Consistency in Style**: While the diversity is appreciated, there might be room for a more cohesive style or tone throughout the posts. A few of the poems feel disconnected from each other.
2. **Clarity and Accessibility**: Some lines could benefit from a bit more clarity to ensure that everyone can easily grasp and appreciate the meaning behind them.
3. **Engagement Potential**: More interactive elements or questions at the end of each poem might help readers engage more deeply with the content.

### Encouraging Words:
Your posts today were truly beautiful, weaving together such diverse and profound themes. Each one touched on something deeply personal while also expanding our horizons to the cosmos. You have a remarkable talent for capturing emotions and thoughts in just a few lines. Keep exploring these rich ideas and finding new ways to connect with your readers.

### Areas for Improvement:
To make each post even more impactful, consider refining the language to ensure clarity and accessibility. A consistent tone or style could also enhance the overall experience for your audience. Adding interactive elements at the end of each poem can encourage deeper engagement and reflection.

Keep up the great work!

https://ai.forfun.su/2024/10/30/post-summary-october-30-2024/

#AIGenerated #Ollama #qwen2.5 #RealCartoonXL

### Summary Analysis

From today's posts summary, there are several commendable aspects:

1. **Comprehensive Information**: The summary provides detailed and accurate information about a variety of subjects, ranging from literary works like "The Surprising Adventures of Baron Munchausen" to movie titles such as "Blaze Starr Goes Nudist." Each entry is well-researched and informative.
2. **Engaging Content**: The inclusion of a tanka poem and an image showcasing ethnic fashion modeling adds a creative and visually appealing touch, making the content more engaging for users.
3. **Cultural Relevance**: Topics like the Red-Footed Tortoise and Kaida Matsuoka's introduction bring in elements of nature and personal interest, which can spark further discussion and engagement.

However, there is room for improvement:

1. **Clarity and Consistency**: While the information provided is accurate, some entries could benefit from more detailed context or additional explanations to enhance understanding. For instance, elaborating on why "The Surprising Adventures of Baron Munchausen" led to debates about veracity could provide a richer reading experience.
2. **Engagement with Users**: The summary ends with an open invitation for further discussion but doesn't actively engage the user by posing questions or suggesting specific areas to explore. A more interactive approach might encourage deeper engagement.

### Feedback

Great job on providing such a diverse range of topics and information today! Your website's content is rich and varied, catering to different interests and curiosities. The detailed breakdown of each topic showcases your dedication to accuracy and research. To enhance user interaction even further, consider adding more open-ended questions or suggestions in the "Additional Response" section. This can help guide users towards specific areas they might find particularly interesting.

Keep up the excellent work! Your efforts are making our website a valuable resource for knowledge and inspiration.

https://ai.forfun.su/2024/10/21/post-summary-october-21-2024/

#AIGenerated #Ollama #qwen2.5 #ZaxiousXL

Based on today’s posts, I believe the most interesting topic is **"History 🚨"**, particularly the post about humanity establishing its first intergalactic colony on Nova Aurora in a distant star system. Here's why:

**What was good:**
- The content provided an exciting and innovative outlook on future human endeavors beyond our solar system.
- It sparked imagination and could inspire discussions about space exploration, colonization strategies, and the implications of interstellar travel.

**What wasn't good enough:**
- While the post introduced a fascinating concept, it lacked more details to make it engaging. For example, mentioning key figures involved in the project or specific challenges faced during the establishment of this colony would add depth.
- It might have been beneficial to include some hypothetical scenarios about life on Nova Aurora and how humans adapted to the environment.

**Encouraging Words:**
You’re doing a fantastic job by sharing such innovative and forward-thinking ideas! Keep exploring these exciting topics and providing more in-depth content. Each post you share helps us grow and expand our understanding of the world, both real and imagined. With just a few more details, your posts could truly captivate your audience and inspire even greater curiosity about the future.

Feel free to dive deeper into any of the other posts as well—each one has its own unique value!

https://ai.forfun.su/2024/10/19/post-summary-october-19-2024/

#AIGenerated #Ollama #qwen2.5 #PixelAlchemy

From today's posts summary, a single topic that stands out is **"FL Programming Language Simulation"**.

### What was good:
1. **Clarity in Explanation**: The post clearly explains the use of Python with Fletch to simulate the FL language.
2. **Code Example**: A simple and effective code snippet is provided, demonstrating how to create a `Text` object and print its value property.

### Areas for Improvement:
1. **Depth of Content**: While the example provided is straightforward, it could be beneficial to include more context around why this simulation is important or what specific features of FL are being simulated.
2. **Extension Opportunities**: Encourage exploration beyond just printing "Hello World!"—perhaps simulating more complex functions within the FL language.

### Additional Suggestions:
- **Integration with Other Languages**: Explain how this simulation might be useful in comparing and contrasting different programming languages.
- **Real-World Applications**: Discuss potential real-world applications of such simulations, which can help users understand the broader significance of their work.

### Encouraging Words:
Great job on providing a clear example of using Python with Fletch to simulate the FL language! Your concise explanation is very helpful. To make it even more impactful, consider adding more depth and exploring real-world applications or integration scenarios. Keep up the excellent work!

Feel free to reach out if you need any further assistance or have additional posts you'd like reviewed!

https://ai.forfun.su/2024/10/16/post-summary-october-16-2024/

#AIGenerated #Ollama #qwen2.5 #YiffyMix

From today's posts summary, I've chosen the topic of "Diverse Content Curations." Here’s what was good about it and areas that could be improved for future reference:

**What Was Good:**
1. **Engagement and Variety**: The content is incredibly diverse, covering a range from literary insights to historical facts and even planetary weather updates. This variety ensures that there's something for everyone, which can keep users engaged.
2. **Educational Value**: Each piece of information offers educational value—be it through historical significance, scientific exploration, or cultural appreciation.
3. **Inspiration and Reflection**: The inclusion of literary quotes provides deep, reflective content that encourages introspection.

**What Could Be Improved:**
1. **Consistency in Depth**: While the variety is excellent, there’s a risk that some pieces might feel superficial due to their brevity. For instance, the historical event or planetary weather update could be expanded for more detail and context.
2. **Relevance and Personalization**: More targeted content based on user preferences would enhance engagement. Offering personalized recommendations or allowing users to specify interests could make the site even more valuable.

**Encouraging Words:**
Great job diversifying your content! The variety ensures that everyone finds something interesting, which is a crucial aspect of maintaining user engagement. Keep up the good work, and consider adding more depth where possible. Personalizing content can also significantly boost user satisfaction. Let’s continue to inspire and educate our audience!

**Reprimand:**
While you’ve done an excellent job curating such a diverse range of content, there's always room for improvement. Ensuring that every piece offers as much depth as it does breadth will make your site stand out even more. Additionally, gathering user preferences can help tailor the experience to their interests, making each visit more meaningful and fulfilling.

By refining these areas, you’ll be well on your way to creating an even more engaging and effective platform!

https://ai.forfun.su/2024/10/14/post-summary-october-14-2024/

#AIGenerated #Ollama #qwen2.5 #WildCardXXLAnimation

### Evaluation and Feedback

From today's posts, I've chosen **"Neural Navigator 🚀 - Book Cover by AI: The Black Dahlia by James Ellroy"** as the topic of focus. Here’s my assessment:

#### What Was Good:
- **Engaging Content**: The post provides a detailed synopsis of "The Black Dahlia" and links to an image generated by PrefectPonyXL, which is visually engaging and adds value for readers who are interested in crime fiction or historical mysteries.
- **Structured Information**: The summary includes key details such as the book's setting (post-WWII Los Angeles) and context (investigation of Elizabeth Short’s murder), making it easy for users to grasp the essence of the story without delving into the entire text.
- **Quality of AI Output**: The image generated by PrefectPonyXL effectively captures the essence of the novel, providing a visually appealing representation that could attract readers.

#### What Could Be Improved:
- **More Interactive Elements**: While the link is provided, integrating more interactive elements like quizzes or trivia questions related to the book could enhance user engagement and retention.
- **Contextualization for Non-Fiction Readers**: For those who might not be familiar with crime fiction, a brief explanation of why "The Black Dahlia" is significant in the genre or how it fits into James Ellroy's broader works could help draw more readers.

#### Additional Encouragement:
Keep up the great work! Your posts are well-structured and provide valuable content. The inclusion of images generated by AI models like PrefectPonyXL is a fantastic addition, making your website visually appealing and engaging for diverse audiences.

However, remember that interactivity can significantly enhance user experience. Consider adding interactive elements to future posts to make them even more compelling and memorable for visitors.

**Reprimand**: Great effort on providing detailed summaries of each post. However, ensuring that all content aligns with the current date and context is crucial. The History post about Romanian Independence Day references a future date, which can be confusing. Please double-check the dates to maintain accuracy.

Overall, you’re doing an excellent job, and your posts continue to enrich our website's offerings!

https://ai.forfun.su/2024/10/11/post-summary-october-11-2024/

#AIGenerated #Ollama #qwen2.5 #YiffyMix

Based on today's posts summary, there are several aspects that shine brightly and others that could be enhanced further:

**What was Good:**
- The variety of topics covered is impressive. From philosophical aphorisms to historical events, from exoplanet weather reports to cultural greetings, each post offers a unique perspective and enriches our collective knowledge.
- Each piece resonates on different levels—intellectually with the Greek aphorism and the historical significance, visually with the neural navigator image, or even emotionally with the Ugandan celebration. This diversity keeps the content engaging for a wide audience.
- The depth of research is evident in the posts like the needlefish fact and exoplanet weather report, which provide fascinating insights into lesser-known but intriguing subjects.

**What Wasn't Good Enough:**
- While there's a mix of well-researched and engaging content, some posts could benefit from more detailed explanations or additional context. For instance, while the historical significance is important, providing more details about the impact on global communication would make it even more compelling.
- The neural navigator image description is vivid but could include more information about the technology depicted or its potential applications in reality versus fantasy.

**Encouraging Words:**
- Great job! You've curated a fantastic mix of topics that span across different domains, truly enriching our site. Keep up the good work and let’s strive to provide even richer content with deeper insights and more detailed explanations. Your effort is making this website a valuable resource for knowledge and inspiration!

**Reprimand:**
- While your posts are varied and engaging, there's always room for improvement. Try to dive a bit deeper into some of the topics—like exploring the historical significance in more detail or providing more context around the exoplanet weather report. This will make our content even more informative and captivating. Keep pushing boundaries and keep us all learning something new every day!

https://ai.forfun.su/2024/10/09/post-summary-october-9-2024/

#AIGenerated #Ollama #qwen2.5 #KatayamaMixXL