|
{ |
|
"config_version": "1.0.0", |
|
"model_id": "Qwen/Qwen2.5-32B", |
|
"model_kinds": [ |
|
"ARTIFACT" |
|
], |
|
"model_class": { |
|
"module": "furiosa_models.architecture.models.qwen2", |
|
"name": "Qwen2ForCausalLM" |
|
}, |
|
"llm_config": { |
|
"optimization_config": { |
|
"attention_type": "PAGED_ATTENTION", |
|
"optimize_rope": true, |
|
"optimize_packed": true, |
|
"decompose_layernorm": false, |
|
"optimize_furiosa": false, |
|
"use_unsplit_packed": false, |
|
"compact_causal_mask": false, |
|
"use_rngd_gelu": false, |
|
"causal_mask_free_decoding": true, |
|
"kv_cache_sharing_across_beams": false, |
|
"inbound_beamsearch_softmax": false, |
|
"calculate_logit_only_for_last_token": false, |
|
"optimized_for_speculative_decoding": true, |
|
"use_2d_masks": false, |
|
"merged_kv_indices": false |
|
}, |
|
"quantization_config": { |
|
"weight": "bf16", |
|
"activation": "bf16", |
|
"kv_cache": "bf16", |
|
"use_mcp": false |
|
} |
|
}, |
|
"components_versions": { |
|
"furiosa_llm": { |
|
"version": "0.1.0-dev", |
|
"git_hash": "832055d", |
|
"build_time": null |
|
}, |
|
"furiosa_ir": { |
|
"version": "0.11.0-dev", |
|
"git_hash": "4467f6a699", |
|
"build_time": "2025-08-22T02:07:15Z" |
|
}, |
|
"furiosa_runtime": { |
|
"version": "2025.3.1", |
|
"git_hash": "832055d73", |
|
"build_time": "2025-08-22T02:07:20Z" |
|
}, |
|
"furiosa_model_compressor": { |
|
"version": "2025.3.0 (rev: 4cd9804)", |
|
"git_hash": null, |
|
"build_time": null |
|
} |
|
} |
|
} |