Configuration Parsing Warning:In config.json: "quantization_config.modules_to_not_convert" must be an array

IOL-AI 2026 — phi-4 AWQ

Structural variant of https://huggingface.co/arvindcr4/iolai-2026-qwen25-14b-awq

Identical script.py; the only change is the model. Six submissions built on Qwen2.5-14B-Instruct-AWQ showed the score was bounded by the model failing to solve the problems (exact match ~0.025 against the organizers' 0.0729 with the same weights), not by parsing or prompting — seven parser variants replayed over identical raw generations all scored exact match 0.0000.

microsoft/phi-4 is loadable under the sandbox's pinned transformers 4.44.1 (Phi3ForCausalLM, rope_scaling: None, standard AWQ gemm, no remote code), fits the T4 at 9.1 GB, and unlike Qwen2.5 ships no repetition_penalty in its generation config.

AWQ weights from stelterlab/phi-4-AWQ.

Downloads last month: 22

Safetensors

Model size

15B params

Tensor type

I32

F16

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for arvindcr4/iolai-2026-phi4-awq

Base model

microsoft/phi-4

Quantized

(162)

this model