All tags
Model: "deepseek-r1-distilled-qwen-1.5b"
small news items
gpt-4.5 gpt-5 deepseek-r1-distilled-qwen-1.5b o1-preview modernbert-0.3b qwen-0.5b o3 openai ollama mistral perplexity cerebras alibaba groq bytedance math benchmarking fine-tuning model-performance reinforcement-learning model-architecture partnerships funding jeremyphoward arankomatsuzaki sama nrehiew_ danhendrycks akhaliq
OpenAI announced plans for GPT-4.5 (Orion) and GPT-5, with GPT-5 integrating the o3 model and offering unlimited chat access in the free tier. DeepSeek R1 Distilled Qwen 1.5B outperforms OpenAI's o1-preview on math benchmarks, while ModernBERT 0.3b surpasses Qwen 0.5b at MMLU without fine-tuning. Mistral and Perplexity adopt Cerebras hardware for 10x performance gains. OpenAI's o3 model won a gold medal at the 2024 International Olympiad in Informatics. Partnerships include Qwen with Groq. Significant RLHF activity is noted in Nigeria and the global south, and Bytedance is expected to rise in AI prominence soon. "GPT5 is all you need."
not much happened today
zonos-v0.1 audiobox-aesthetics moshi sonar llama-3-70b gpt-4o-mini claude-3.5-haiku gpt-4o claude-3.5-sonnet deepseek-r1-distilled-qwen-1.5b reasonflux-32b o1-preview zyphra-ai meta-ai-fair kyutai-labs perplexity-ai cerebras uc-berkeley brilliant-labs google-deepmind text-to-speech speech-to-speech benchmarking model-performance reinforcement-learning math real-time-processing open-source cross-platform-integration multilinguality zero-shot-learning danhendrycks
Zyphra AI launched Zonos-v0.1, a leading open-weight text-to-speech model supporting multiple languages and zero-shot voice cloning. Meta FAIR released the open-source Audiobox Aesthetics model trained on 562 hours of audio data. Kyutai Labs introduced Moshi, a real-time speech-to-speech system with low latency. Perplexity AI announced the Sonar model based on Llama 3.3 70b, outperforming top models like GPT-4o and Claude 3.5 Sonnet with 1200 tokens/second speed, powered by Cerebras infrastructure. UC Berkeley open-sourced a 1.5B model trained with reinforcement learning that beats o1-preview on math tasks. ReasonFlux-32B achieved 91.2% on the MATH benchmark, outperforming OpenAI o1-preview. CrossPoster, an AI agent for cross-platform posting, was released using LlamaIndex workflows. Brilliant Labs integrated the Google DeepMind Gemini Live API into smart glasses for real-time translation and object identification.