Config reference: `training`

Enterprise tuning surface

Defaults + constraints are rendered directly from Pydantic.
Env keys when available

Many fields have an env-style alias (from TriBridConfig.to_flat_dict()).
Tooltip-level guidance

If a matching glossary entry exists, you’ll see deeper tuning notes.

Config reference Config API & workflow Glossary

Total parameters: 36

Group index

(root)

`(root)`

JSON key	Env key(s)	Type	Default	Constraints	Summary
`training.learning_reranker_backend`	`LEARNING_RERANKER_BACKEND`	`Literal["auto", "mlx_qwen3"]`	`"auto"`	allowed="auto", "mlx_qwen3"	Learning reranker backend: auto (prefer MLX Qwen3 on Apple Silicon), mlx_qwen3 (force). Legacy values 'transformers'/'hf' normalize to 'auto'.
`training.learning_reranker_base_model`	`LEARNING_RERANKER_BASE_MODEL`	`str`	`"Qwen/Qwen3-Reranker-0.6B"`	—	Base model to fine-tune for MLX Qwen3 learning reranker
`training.learning_reranker_grad_accum_steps`	`LEARNING_RERANKER_GRAD_ACCUM_STEPS`	`int`	`8`	≥ 1, ≤ 128	Gradient accumulation steps per optimizer update for MLX Qwen3 learning reranker training
`training.learning_reranker_lora_alpha`	`LEARNING_RERANKER_LORA_ALPHA`	`float`	`32.0`	> 0.0, ≤ 512.0	LoRA alpha for MLX Qwen3 learning reranker
`training.learning_reranker_lora_dropout`	`LEARNING_RERANKER_LORA_DROPOUT`	`float`	`0.05`	≥ 0.0, ≤ 0.5	LoRA dropout for MLX Qwen3 learning reranker
`training.learning_reranker_lora_rank`	`LEARNING_RERANKER_LORA_RANK`	`int`	`16`	≥ 1, ≤ 128	LoRA rank for MLX Qwen3 learning reranker
`training.learning_reranker_lora_target_modules`	—	`list[str]`	`["q_proj", "k_proj", "v_proj", "o_proj"]`	min_length=1	Module name suffixes to apply LoRA to (MLX Qwen3)
`training.learning_reranker_negative_ratio`	`LEARNING_RERANKER_NEGATIVE_RATIO`	`int`	`5`	≥ 1, ≤ 20	Negative pairs per positive during learning reranker training
`training.learning_reranker_promote_epsilon`	`LEARNING_RERANKER_PROMOTE_EPSILON`	`float`	`0.0`	≥ 0.0, ≤ 1.0	Minimum improvement required to auto-promote (primary metric delta)
`training.learning_reranker_promote_if_improves`	`LEARNING_RERANKER_PROMOTE_IF_IMPROVES`	`int`	`1`	≥ 0, ≤ 1	Promote trained learning artifact to active path only if primary metric improves
`training.learning_reranker_telemetry_interval_steps`	`LEARNING_RERANKER_TELEMETRY_INTERVAL_STEPS`	`int`	`2`	≥ 1, ≤ 20	Emit trainer telemetry every N optimizer steps (plus first/final)
`training.learning_reranker_unload_after_sec`	`LEARNING_RERANKER_UNLOAD_AFTER_SEC`	`int`	`0`	≥ 0, ≤ 86400	Unload MLX learning reranker model after idle seconds (0 = never)
`training.ragweld_agent_backend`	`RAGWELD_AGENT_BACKEND`	`str`	`"mlx_qwen3"`	—	Ragweld agent backend (in-process chat model). Currently: mlx_qwen3
`training.ragweld_agent_base_model`	`RAGWELD_AGENT_BASE_MODEL`	`str`	`"mlx-community/Qwen3-1.7B-4bit"`	—	Shipped base model for the ragweld agent (MLX).
`training.ragweld_agent_grad_accum_steps`	`RAGWELD_AGENT_GRAD_ACCUM_STEPS`	`int`	`8`	≥ 1, ≤ 128	Gradient accumulation steps per optimizer update for ragweld agent training.
`training.ragweld_agent_lora_alpha`	`RAGWELD_AGENT_LORA_ALPHA`	`float`	`32.0`	> 0.0, ≤ 512.0	LoRA alpha for ragweld agent MLX fine-tuning.
`training.ragweld_agent_lora_dropout`	`RAGWELD_AGENT_LORA_DROPOUT`	`float`	`0.05`	≥ 0.0, ≤ 0.5	LoRA dropout for ragweld agent MLX fine-tuning.
`training.ragweld_agent_lora_rank`	`RAGWELD_AGENT_LORA_RANK`	`int`	`16`	≥ 1, ≤ 128	LoRA rank for ragweld agent MLX fine-tuning.
`training.ragweld_agent_lora_target_modules`	—	`list[str]`	`["q_proj", "k_proj", "v_proj", "o_proj"]`	min_length=1	Module name suffixes to apply LoRA to (ragweld agent; MLX Qwen3).
`training.ragweld_agent_model_path`	`RAGWELD_AGENT_MODEL_PATH`	`str`	`"models/learning-agent-epstein-files-1"`	—	Active ragweld agent adapter artifact path (directory containing adapter.npz + adapter_config.json).
`training.ragweld_agent_promote_epsilon`	`RAGWELD_AGENT_PROMOTE_EPSILON`	`float`	`0.0`	≥ 0.0, ≤ 10.0	Minimum eval_loss improvement required to auto-promote (baseline_loss - new_loss >= epsilon).
`training.ragweld_agent_promote_if_improves`	`RAGWELD_AGENT_PROMOTE_IF_IMPROVES`	`int`	`1`	≥ 0, ≤ 1	Auto-promote trained ragweld agent adapter only if eval_loss improves.
`training.ragweld_agent_reload_period_sec`	`RAGWELD_AGENT_RELOAD_PERIOD_SEC`	`int`	`60`	≥ 0, ≤ 600	Adapter reload check period (seconds). 0 = check every request.
`training.ragweld_agent_telemetry_interval_steps`	`RAGWELD_AGENT_TELEMETRY_INTERVAL_STEPS`	`int`	`2`	≥ 1, ≤ 20	Emit ragweld agent trainer telemetry every N optimizer steps (plus first/final).
`training.ragweld_agent_train_dataset_path`	`RAGWELD_AGENT_TRAIN_DATASET_PATH`	`str`	`""`	—	Training dataset path for the ragweld agent (empty = use evaluation.eval_dataset_path).
`training.ragweld_agent_unload_after_sec`	`RAGWELD_AGENT_UNLOAD_AFTER_SEC`	`int`	`0`	≥ 0, ≤ 86400	Unload ragweld agent model after idle seconds (0 = never).
`training.reranker_train_batch`	`RERANKER_TRAIN_BATCH`	`int`	`16`	≥ 1, ≤ 128	Training batch size
`training.reranker_train_epochs`	`RERANKER_TRAIN_EPOCHS`	`int`	`2`	≥ 1, ≤ 20	Training epochs for reranker
`training.reranker_train_lr`	`RERANKER_TRAIN_LR`	`float`	`2e-05`	≥ 1e-06, ≤ 0.001	Learning rate
`training.reranker_warmup_ratio`	`RERANKER_WARMUP_RATIO`	`float`	`0.1`	≥ 0.0, ≤ 0.5	Warmup steps ratio
`training.tribrid_reranker_mine_mode`	`TRIBRID_RERANKER_MINE_MODE`	`str`	`"replace"`	pattern=^(replace\|append)$	Triplet mining mode
`training.tribrid_reranker_mine_reset`	`TRIBRID_RERANKER_MINE_RESET`	`int`	`0`	≥ 0, ≤ 1	Reset triplets file before mining
`training.tribrid_reranker_model_path`	`TRIBRID_RERANKER_MODEL_PATH`	`str`	`"models/learning-reranker-epstein-files-1"`	—	Active learning reranker artifact path (MLX adapter directory).
`training.tribrid_triplets_path`	`TRIBRID_TRIPLETS_PATH`	`str`	`"data/training/triplets__epstein-files-1.jsonl"`	—	Training triplets file path
`training.triplets_min_count`	`TRIPLETS_MIN_COUNT`	`int`	`100`	≥ 10, ≤ 10000	Min triplets for training
`training.triplets_mine_mode`	`TRIPLETS_MINE_MODE`	`str`	`"replace"`	pattern=^(replace\|append)$	Triplet mining mode

Details (glossary)

training.learning_reranker_backend (LEARNING_RERANKER_BACKEND) — Learning Reranker Backend

Category: reranking

Backend used when RERANKER_MODE="learning". auto selects the MLX Qwen3 LoRA backend (Apple Silicon). mlx_qwen3 forces the same MLX backend. Legacy values (transformers/hf) are accepted for backward compatibility but normalize to auto.

Badges: - Backend selector

training.learning_reranker_base_model (LEARNING_RERANKER_BASE_MODEL) — Learning Reranker Base Model

Category: reranking

Base model identifier to fine-tune FROM when using the MLX Qwen3 learning backend (e.g. Qwen/Qwen3-Reranker-0.6B). Training produces a LoRA adapter artifact written under TRIBRID_RERANKER_MODEL_PATH and inference loads that adapter on top of this base. Changing the base model makes existing adapters incompatible.

Badges: - MLX only

training.learning_reranker_grad_accum_steps (LEARNING_RERANKER_GRAD_ACCUM_STEPS) — Learning Reranker Grad Accum Steps

Category: reranking

Number of micro-batches to accumulate gradients over before applying one optimizer update when training the MLX Qwen3 learning reranker. This increases effective batch size without increasing memory. Typical: 4–16; default 8.

Badges: - MLX only - Affects training speed

training.learning_reranker_lora_alpha (LEARNING_RERANKER_LORA_ALPHA) — Learning Reranker LoRA Alpha

Category: reranking

LoRA scaling (alpha) for MLX Qwen3 fine-tuning. Effective LoRA scale is alpha/rank. Typical: 16–64; start at 32.

Badges: - MLX only

training.learning_reranker_lora_dropout (LEARNING_RERANKER_LORA_DROPOUT) — Learning Reranker LoRA Dropout

Category: reranking

LoRA dropout probability for MLX Qwen3 fine-tuning. Small dropout (0.0–0.1) can reduce overfitting on small mined datasets. Start at 0.05.

Badges: - MLX only

training.learning_reranker_lora_rank (LEARNING_RERANKER_LORA_RANK) — Learning Reranker LoRA Rank

Category: reranking

LoRA rank (r) for MLX Qwen3 fine-tuning. Higher rank increases adapter capacity (and training/inference cost) and can improve quality with enough data. Typical: 8–32; start at 16.

Badges: - MLX only - Affects training cost

training.learning_reranker_negative_ratio (LEARNING_RERANKER_NEGATIVE_RATIO) — Learning Reranker Negative Ratio

Category: reranking

When converting mined triplets into labeled (query, document) pairs for pairwise training, use up to this many negatives per positive. Higher ratios can improve discrimination but increase training time. Typical: 3–5; default 5.

Badges: - Quality vs cost

training.learning_reranker_promote_epsilon (LEARNING_RERANKER_PROMOTE_EPSILON) — Learning Reranker Promotion Epsilon

Category: reranking

Minimum dev-metric improvement required to auto-promote a newly trained learning reranker artifact over the active baseline. Use a small epsilon (e.g., 0.002) to avoid promoting on noise.

Badges: - Prevents noise promotions

training.learning_reranker_promote_if_improves (LEARNING_RERANKER_PROMOTE_IF_IMPROVES) — Learning Reranker Promotion Gate

Category: reranking

If enabled (1), training only promotes the newly trained artifact to TRIBRID_RERANKER_MODEL_PATH if the primary dev metric improves over the active baseline by at least LEARNING_RERANKER_PROMOTE_EPSILON. If disabled (0), successful training always promotes.

Badges: - Safety

training.learning_reranker_telemetry_interval_steps (LEARNING_RERANKER_TELEMETRY_INTERVAL_STEPS) — Learning Reranker Telemetry Interval Steps

Category: reranking

Emit trainer telemetry every N optimizer steps (plus first and final events). Default: 2. Range: 1-20. Lower values give smoother live charts but increase event volume.

training.learning_reranker_unload_after_sec (LEARNING_RERANKER_UNLOAD_AFTER_SEC) — Learning Reranker Idle Unload

Category: reranking

If >0, unload the MLX Qwen3 reranker model from memory after this many seconds of inactivity. Set to 0 to keep the model resident (faster first rerank, higher memory use).

Badges: - MLX only - Affects latency

training.ragweld_agent_backend (RAGWELD_AGENT_BACKEND) — RAGWELD_AGENT_BACKEND

Category: general