amd/Quark - Page 2

quark-torch-install

Install or verify the correct PyTorch build for a user's accelerator backend before Quark installation. Trigger for "install PyTorch", "pip install torch", "set up torch for ROCm", "set up torch for CUDA", "torch version mismatch", "CPU-only torch installed",…

Langue du texte source : anglais

quark-torch-llm-eval

End-to-end LLM accuracy evaluation on AMD ROCm (ROCm-only) — docker container setup OR host (no-docker) runtime, vLLM/SGLang/ATOM serving, lm-eval / lighteval / evalscope benchmarks. Use when the user wants to evaluate, benchmark, or compare an LLM's…

Langue du texte source : anglais

quark-torch-llm-ptq-eval

End-to-end Torch LLM PTQ recipe for AMD Quark — for PyTorch / HuggingFace transformers models (safetensors input): quantize, then validate, then evaluate accuracy, in one flow. Delegates the PTQ path (intake, planning, manifest, execution) to the…

Langue du texte source : anglais

quark-torch-model-intake

Inspect a HuggingFace transformers / safetensors model and prepare metadata for Quark Torch PTQ planning. Use for model path validation, architecture detection, quantization target discovery, layer counting, risk assessment, or transformers compatibility…

Langue du texte source : anglais

quark-torch-ptq

Torch LLM PTQ workflow for AMD Quark — for PyTorch / HuggingFace transformers models (safetensors input). Use when the user wants a complete PTQ pipeline: model inspection, quantization planning, script generation, and optional execution. Stops at the…

Langue du texte source : anglais

quark-torch-result-validator

Validate Quark Torch quantization output (HuggingFace safetensors + config.json) using four lightweight checks: auxiliary file copy alignment, excluded tensor MD5 byte-identity, config.json deep comparison after stripping quantization keys, and safetensors…

Langue du texte source : anglais