SIERRA AI

New! Mistral’s awesome LLMs - Mistral 7B and Mixtral 8x7B - are coming to Bedrock. Let’s check it out… We are bringing Mistral AI to Amazon Bedrock as our 7th foundation model provider, joining other leading AI companies like AI21 Labs, Anthropic, Cohere, Meta, Stability AI, and Amazon. With these two Mistral AI models, you will have the flexibility to choose the optimal, high-performing LLM for your use case to build and scale generative AI applications using Amazon Bedrock. ??Mistral 7B is the first foundation model from Mistral AI, supporting English text generation tasks with natural coding capabilities. ?? It is optimized for low latency with a low memory requirement and high throughput for its size. This model is powerful and supports various use cases from text summarization and classification, to text completion and code completion. ?? Mixtral 8x7B is a popular, high-quality sparse Mixture-of-Experts (MoE) model that is ideal for text summarization, question and answering, text classification, text completion, and code generation. Mistral AI publicly available models are coming soon to Amazon Bedrock. Exciting! #genai #ai #amazon

New! Amazon introduces BASE TTS - a billion-parameter text-to-speech model trained on 100K hours of speech data (with fascinating emergent behavior). We just announced BASE TTS, the largest neural text-to-speech model to date with 1 billion parameters trained on 100K hours of public domain speech data. The model achieves state-of-the-art naturalness compared to other large-scale TTS systems. It's also fast. The decoder generates waveforms 70% faster than baseline diffusion-based decoding, allowing streaming while improving audio quality. As neural text-to-speech models scale up in parameters and training data, they begin to exhibit emergent abilities similar to large language models. That is, without being explicitly trained to handle certain phenomena, the models start acquiring new skills simply through exposure to more examples during training. ?? The small 1K hour, 150M parameter BASE TTS barely understands emotions and para-linguistics. ?? The 10K hour, 400M parameter version achieves competence across most categories, with excellence in English compound word pronunciation. ?? Further scaling to 100K hours and 1B parameters brings subtle continued improvements (some challenging semantic categories like emotions still need more work). BASE TTS sets a new benchmark for quality and efficiency at scale (across compound words, emotional cues, foreign words, unusual punctuations, questions, whispered speech, complex syntax, etc), and we hope the insights on model scaling could guide future work to achieve ever more robust, multi-lingual speech synthesis. #genai #ai #amazon

Use this year’s Emerging Tech Impact Radar to: ? Enhance your competitive edge in the smart world ? Prioritize prevalent and impactful GenAI use cases that already deliver real value to users ? Balance stimulating growth and mitigating risk ? Identify relevant emerging technologies that support your strategic product roadmap Explore all 30 technologies and trends: https://gtnr.it/3HVcDI4 #GartnerIT #EmergingTech #Technology #Tech #GenAI

If you watched the Grammy's, you saw #Llama2 in action. The The Recording Academy used IBM's WatsonX platform powered by Llama 2 + RAG+ few-shot learning to develop AI Stories that allowed the Academy to create content for the 1,000+ GRAMMY-nominated artists, across almost 100 award categories. It would have taken a small army to write, edit, and produce enough content to meet the demand. But using the WatsonX + Llama solution, the editorial the team quickly generated, reviewed and published content about a particular artist. https://lnkd.in/e_2_egwX

What better platforms than dynamic Dell Technologies Servers to test the latest NVIDIA L4 GPUs for low-power inferencing! ??... Love the PowerEdge XR7620 at the Edge ?? StorageReview.com strikes again! ?? They tested those beautiful GPUs not on one but on three different Dell Servers: ? PowerEdge XR7620 ? PowerEdge T560 ? PowerEdge R760 With all those, you won't be able to say that they haven't been tested in real-world conditions ?? Once again, thank you Brian Beeler for making the show! #Servers #GPU #Edge #IWork4Dell