{"record":{"id":"cba447a574f990a3","repo":"sgl-project/sglang","slug":"generate-subcommand-is-not-yet-supported-for-model","errorCode":null,"errorMessage":"Generate subcommand is not yet supported for model: {model_path}","messagePattern":"Generate subcommand is not yet supported for model: (.+?)","errorType":"exception","errorClass":"Exception","httpStatus":null,"severity":"critical","filePath":"python/sglang/cli/generate.py","lineNumber":31,"sourceCode":"        parser = argparse.ArgumentParser(description=\"SGLang Multimodal Generation\")\n        add_multimodal_gen_generate_args(parser)\n        parser.parse_args(extra_argv)\n        return\n\n    model_path = get_model_path(extra_argv)\n    is_diffusion_model = get_is_diffusion_model(model_path)\n    if is_diffusion_model:\n        from sglang.multimodal_gen.runtime.entrypoints.cli.generate import (\n            add_multimodal_gen_generate_args,\n            generate_cmd,\n        )\n\n        parser = argparse.ArgumentParser(description=\"SGLang Multimodal Generation\")\n        add_multimodal_gen_generate_args(parser)\n        parsed_args, unknown_args = parser.parse_known_args(extra_argv)\n        generate_cmd(parsed_args, unknown_args)\n    else:\n        raise Exception(\n            f\"Generate subcommand is not yet supported for model: {model_path}\"\n        )\n","sourceCodeStart":13,"sourceCodeEnd":34,"githubUrl":"https://github.com/sgl-project/sglang/blob/0132848349585cfe6aae51c4941cbae872505f8a/python/sglang/cli/generate.py#L13-L34","documentation":"The float16 branch of deterministic_all_reduce only supports world sizes 2,4,6,8; other sizes hit the default case and throw.","triggerScenarios":"Deterministic allreduce on fp16 tensors with fa->world_size_ outside {2,4,6,8}.","commonSituations":"--enable-deterministic-inference with TP=3/5/7 or >8 GPUs on ROCm; misdetected world size from bad visibility masks.","solutions":["Use an even TP size within 2-8","Disable the deterministic custom allreduce for unsupported topologies","Verify GPU visibility/env masks"],"exampleFix":null,"handlingStrategy":"validation","validationCode":"assert fa.world_size_ in (2,4,6,8), 'deterministic fp16 allreduce needs TP in (2,4,6,8)'","typeGuard":null,"tryCatchPattern":null,"preventionTips":["Check TP size before enabling deterministic inference"],"tags":["rocm","allreduce","deterministic","tensor-parallel","float16"],"backgroundTag":"world-size-unsupported","analyzedSha":"0132848349585cfe6aae51c4941cbae872505f8a","analyzedAt":"2026-08-28T05:10:05.995Z","schemaVersion":2},"datasetVersion":"2026-08-28T06:17:29.519Z"}