Coverage for  / home / jenkins / .local / lib / python3.10 / site-packages / hyper_parallel / auto_parallel / config_adapter / _normalized_config.py: 100%

1# Copyright 2026 Huawei Technologies Co., Ltd 

2# 

3# Licensed under the Apache License, Version 2.0 (the "License"); 

4# you may not use this file except in compliance with the License. 

5# You may obtain a copy of the License at 

6# 

7# http://www.apache.org/licenses/LICENSE-2.0 

8# 

9# Unless required by applicable law or agreed to in writing, software 

10# distributed under the License is distributed on an "AS IS" BASIS, 

11# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 

12# See the License for the specific language governing permissions and 

13# limitations under the License. 

14# ============================================================================ 

15"""Shared type definitions for auto parallel strategy search configuration.""" 

16 

17from dataclasses import dataclass, field 

18from typing import Any, Dict, List, Optional, Literal 

19 

20 

21@dataclass 

22class NormalizedConfig: 

23 """Aggregate container for a parallel strategy search task. 

24 

25 Holds all configuration sections as plain dicts for maximum 

26 compatibility with PR631's Config class (sapp_nd.nd.common.config). 

27 The dict keys 

28 follow PR631's HyperParallel TOML naming conventions. 

29 

30 **Required model_spec fields**: ``n_layers``, ``dim``, ``n_heads``, 

31 ``vocab_size``. 

32 

33 **Optional model_spec fields**: ``inter_dim``, ``n_kv_heads``, 

34 ``seq_len``, ``local_batch_size``, ``params_dtype``, ``compute_dtype``, 

35 ``softmax_compute_dtype``, ``moe_enabled``, ``num_experts``, 

36 ``num_experts_per_tok``, ``num_shared_experts``, ``moe_inter_dim``, 

37 ``use_flash_attention``, ``use_clip_grad``, ``use_seq_parallel``, 

38 ``vocab_emb_dp``, ``enable_parallel_optimizer``, 

39 ``gradient_accumulation_shard``, ``optimizer_weight_shard_size``, 

40 ``enable_weight_tying``, ``multiple_of``, ``ffn_dim_multiplier``, 

41 ``mtp_depth``, ``n_dense_layers``, ``kv_lora_rank``, ``q_lora_rank``, 

42 ``qk_rope_head_dim``, ``v_head_dim``, ``qk_nope_head_dim``, 

43 ``capacity_factor``, ``first_k_dense_replace``, ``topk_group``, 

44 ``n_group``, ``routed_scaling_factor``. 

45 

46 Args: 

47 model_spec: Model architecture parameters. Must contain at least 

48 ``n_layers``, ``dim``, ``n_heads``, ``vocab_size``. 

49 cluster_spec: Hardware cluster description. 

50 search_space: Parallel dimension candidate values, e.g. 

51 ``{"dp": [1,2,4], "tp": [1,2,4,8], "pp": [1,2], "cp": [1], "ep": [1]}``. 

52 constraint: User-imposed constraints (global_batch_size, 

53 memory_limit_gb, fixed_*_degree). 

54 estimator: Estimation algorithm parameters. 

55 pp_config: Pipeline-parallel specific configuration. 

56 resolved_strategy: Final resolved strategy, populated after search. 

57 """ 

58 

59 model_spec: Dict[str, Any] = field(default_factory=dict) 

60 cluster_spec: Dict[str, Any] = field(default_factory=dict) 

61 search_space: Dict[str, List[int]] = field(default_factory=dict) 

62 constraint: Dict[str, Any] = field(default_factory=dict) 

63 estimator: Dict[str, Any] = field(default_factory=dict) 

64 pp_config: Dict[str, Any] = field(default_factory=dict) 

65 resolved_strategy: Optional[Dict[str, Any]] = None 

66 

67 def to_dict(self) -> Dict[str, Any]: 

68 """Serialize all config sections to a nested dictionary.""" 

69 result: Dict[str, Any] = { 

70 "model_spec": dict(self.model_spec), 

71 "cluster_spec": dict(self.cluster_spec), 

72 "search_space": dict(self.search_space), 

73 "constraint": dict(self.constraint), 

74 "estimator": dict(self.estimator), 

75 "pp_config": dict(self.pp_config), 

76 } 

77 if self.resolved_strategy is not None: 

78 result["resolved_strategy"] = dict(self.resolved_strategy) 

79 return result 

80 

81 

82@dataclass 

83class ValidationError: 

84 """A single validation error or warning discovered during config validation. 

85 

86 Args: 

87 field_path: Dot-separated path to the offending field 

88 (e.g. ``"model_spec.dim"``). 

89 message: Human-readable description of the problem. 

90 severity: Error severity level (``"error"`` or ``"warning"``). 

91 """ 

92 

93 field_path: str 

94 message: str 

95 severity: Literal["error", "warning"] = "error" 

96 

97 

98ValidationSeverity = Literal["error", "warning"]