LLMobile.news

Cutting edge on edge devices.

Microsoft

6 updates on Microsoft.

  1. Microsoft runs Phi-3-mini offline on an iPhone 14 at over 12 tokens per second

    The 3.8-billion-parameter model takes about 1.8 GB at 4-bit and scores 69 percent on MMLU, which Microsoft compares to Mixtral 8x7B and GPT-3.5.

  2. BitNet b1.58 gives every weight three values and runs 3B in 2.22 GB

    Microsoft trained models whose every weight is -1, 0 or 1, which replaces multiplication with addition, and reports parity with full-precision Llama from 3B.

  3. Microsoft releases Phi-2, a 2.7B model it says matches models 25 times larger

    The 2.7B base model was trained on 1.4 trillion tokens in 14 days on 96 A100 GPUs, and Microsoft says it matches models up to 25 times larger.

  4. Microsoft carries its textbook data recipe from code to reasoning with the 1.3B phi-1.5

    The 1.3-billion-parameter model trains on 30B tokens of mostly synthetic data and posts reasoning scores above Llama2-7B in Microsoft evaluations.

  5. Microsoft trains phi-1 to 50.6 percent on HumanEval with 1.3B parameters

    Microsoft Research reports 50.6 percent on HumanEval and 55.5 percent on MBPP for a 1.3B code model trained on under 7B tokens of textbook-quality data.

  6. ONNX Runtime 1.5 adds a mobile build that ships only the operators a model uses

    The release introduced a reduced-size runtime for phones and the ORT model format, the base for the NNAPI, Core ML, QNN and XNNPACK mobile packages.