All tags
Person: "tim_dettmers"
not much happened today
kling-2.5-turbo sora-2 gemini-2.5-flash granite-4.0 qwen-3 qwen-image-2509 qwen3-vl-235b openai google ibm alibaba kling_ai synthesia ollama huggingface arena artificialanalysis tinker scaling01 video-generation instruction-following physics-simulation image-generation model-architecture mixture-of-experts context-windows token-efficiency fine-tuning lora cpu-training model-benchmarking api workflow-automation artificialanlys kling_ai altryne teortaxestex fofrai tim_dettmers sundarpichai officiallogank andrew_n_carr googleaidevs clementdelangue wzhao_nlp alibaba_qwen scaling01 ollama
Kling 2.5 Turbo leads in text-to-video and image-to-video generation with competitive pricing. OpenAI Sora 2 shows strong instruction-following but has physics inconsistencies. Google Gemini 2.5 Flash "Nano Banana" image generation is now generally available with multi-image blending and flexible aspect ratios. IBM Granite 4.0 introduces a hybrid Mamba/Transformer architecture with large context windows and strong token efficiency, outperforming some peers on the Intelligence Index. Qwen models receive updates including fine-tuning API support and improved vision capabilities. Tinker offers a flexible fine-tuning API supporting LoRA sharing and CPU-only training loops. The ecosystem also sees updates like Synthesia 3.0 adding video agents.
not much happened today
gpt-5 kimi-k2-0905 glm-4.5 qwen3-asr opus-4.1 cognition founders-fund lux-capital 8vc neo vercel claude groq alibaba huggingface meta-ai-fair google theturingpost algoperf coding-agents agent-architecture open-source model-evaluation multilingual-models speech-recognition model-optimization kv-cache quantization algorithmic-benchmarking video-generation context-windows swyx tim_dettmers
Cognition raised $400M at a $10.2B valuation to advance AI coding agents, with swyx joining to support the "Decade of Agents" thesis. Vercel launched an OSS "vibe coding platform" using a tuned GPT-5 agent loop. Claude Code emphasizes minimalism in agent loops for reliability. Kimi K2-0905 achieved 94% on coding evals and improved agentic capabilities with doubled context length. Alibaba released Qwen3-ASR, a multilingual transcription model with <8% WER. Meta introduced Set Block Decoding for 3-5× faster decoding without architectural changes. Innovations in KV cache compression and quantization include AutoRound, QuTLASS v0.1.0, and AlgoPerf v0.6. Google's Veo 3 video generation API went GA with significant price cuts and vertical video support.
not much happened today
gpt-5 grok-code-fast-1 claude-sonnet glm-4.5 longcat-flash-chat fastvlm mobileclip2 internvl3.5 openai x-ai zhipu-ai meituan apple model-architecture moe adaptive-compute inference-speed model-training cost-efficiency coding developer-tools open-inference on-device-ai vision gdb martin_casado yanndubs elonmusk cline vikhyatk dzhng quixiai tim_dettmers casper_hansen_ reach_vb eliebakouch teortaxestex youjiacheng
OpenAI integrates GPT-5 into Xcode 26 with improved coding latency, though some UX trade-offs are noted. xAI's Grok Code Fast 1 gains momentum, surpassing Claude Sonnet in usage and praised for fast debugging. Zhipu's GLM-4.5 offers a cost-effective coding plan with strong performance against Claude Sonnet 4. Meituan releases the LongCat-Flash-Chat, a 560B parameter MoE model with adaptive compute and detailed technical insights. Apple debuts on-device vision-language models FastVLM and MobileCLIP2 alongside InternVL3.5.
FSDP+QLoRA: the Answer to 70b-scale AI for desktop class GPUs
qlora fsdp inflection-2.5 gpt-4 answer.ai hugging-face meta-ai-fair nvidia inflectionai model-training quantization memory-optimization gradient-checkpointing cpu-offloading fine-tuning model-sharding reinforcement-learning chain-of-thought benchmarking jeremy_howard tim_dettmers yann_lecun
Jeremy Howard and collaborators released a new tool combining FSDP, QLoRA, and HQQ to enable training 70b-parameter models on affordable consumer GPUs like RTX 4090s with only 24GB RAM, overcoming traditional memory constraints that required expensive data center GPUs costing over $150k. The approach shards quantized models across multiple GPUs and uses techniques like gradient checkpointing and CPU offloading to achieve efficient training on desktop-class hardware. The blogpost details challenges and solutions integrating these methods, highlighting a significant cost reduction from $150k to under $2.5k for training large language models. Additionally, Twitter recaps mention Inflection AI's Inflection-2.5 model rivaling GPT-4 in benchmarks with less compute, and Grok improving speed by 3x. Yann LeCun discusses multi-step reasoning training for LLMs.