Non-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over any text in a single forward pass, in 100+ languages, with a router that picks the right checkpoint per request.
Python calibrationclassificationdecision-modelhuggingfacejevmodernbertmultilingualnlppythonpytorchroutingtyped-decisionszero-shot
Update laya_finetune_typed_decisions_mps.py
Ang Zhange38c5bfa2f01
diff --git a/notebooks/laya_finetune_typed_decisions_mps.py b/notebooks/laya_finetune_typed_decisions_mps.pyindex 44033349..eb70be37 100644--- a/notebooks/laya_finetune_typed_decisions_mps.py+++ b/notebooks/laya_finetune_typed_decisions_mps.py@@ -1,4 +1,3 @@-#!/usr/bin/env python3"""Laya typed-decisions fine-tuning on Apple Silicon.This is a standalone replacement for the Kaggle 2xT4 notebook. It performs:@@ -8,26 +7,20 @@ This is a standalone replacement for the Kaggle 2xT4 notebook. It performs:4. temperature calibration5. checkpoint/model export-Run without arguments:-python notebooks/laya_finetune_typed_decisions_mps.py-Useful examples:python notebooks/laya_finetune_typed_decisions_mps.py --epochs 2 --micro-batch 1 --grad-accum 32python notebooks/laya_finetune_typed_decisions_mps.py --model-dir ./laya_base --items ./train_items.pt- python 2.py --device cpu+ python notebooks/laya_finetune_typed_decisions_mps.py --device cpu"""import argparseimport gcimport jsonimport math-import osimport random-import timefrom pathlib import Pathimport torch-from datasets import load_datasetfrom huggingface_hub import snapshot_downloadfrom safetensors.torch import load_file, save_filefrom transformers import AutoTokenizer@@ -109,13 +102,39 @@ def build_training_item(tokenizer, cfg, state, question, gold_question):def prepare_items(model_dir, items_path, force=False):items_path = Path(items_path)- if items_path.exists() and not force:- print(f"Using cached training items: {items_path}")- return-- with open(Path(model_dir) / "rl_agent_config.json") as f:+ model_path = Path(model_dir).resolve()+ with open(model_path / "rl_agent_config.json") as f:cfg = json.load(f)- tokenizer = AutoTokenizer.from_pretrained(Path(model_dir) / "tokenizer")+ max_len = cfg.get("max_len", 1024)+ head_max_len = cfg.get("head_max_len", 256)+ cache_meta_path = items_path.with_name(items_path.name + ".meta.json")+ cache_key = {+ "model_dir": str(model_path),+ "max_len": max_len,+ "head_max_len": head_max_len,+ "dataset": DATASET_ID,+ "split": "train",+ }+ if items_path.exists() and not force:+ if not cache_meta_path.exists():+ # Legacy item files predate the sidecar metadata. They remain+ # usable offline; newly written caches always receive a key.+ print(f"Using legacy cached training items without metadata: {items_path}")+ return+ try:+ with open(cache_meta_path) as f:+ cached_key = json.load(f)+ except (OSError, json.JSONDecodeError):+ cached_key = None+ if cached_key == cache_key:+ print(f"Using cached training items: {items_path}")+ return+ print("Training-item cache key changed; rebuilding the cache.")++ # Keep datasets optional when a compatible local cache is already available.+ from datasets import load_dataset++ tokenizer = AutoTokenizer.from_pretrained(model_path / "tokenizer")print(f"Downloading dataset {DATASET_ID} ...")dataset = load_dataset(DATASET_ID, "all", split="train")@@ -128,7 +147,13 @@ def prepare_items(model_dir, items_path, force=False):for qid, question in questions.items():if qid not in gold:continue- item = build_training_item(tokenizer, cfg, state, question, gold[qid])+ item = build_training_item(+ tokenizer,+ {**cfg, "max_len": max_len, "head_max_len": head_max_len},+ state,+ question,+ gold[qid],+ )if item is None:skipped += 1else:@@ -136,6 +161,8 @@ def prepare_items(model_dir, items_path, force=False):items_path.parent.mkdir(parents=True, exist_ok=True)torch.save(items, items_path)+ with open(cache_meta_path, "w") as f:+ json.dump(cache_key, f, indent=2)print(f"Saved {len(items)} training items to {items_path}; skipped={skipped}")@@ -191,7 +218,7 @@ def fit_temperature(samples):return float(torch.clamp(log_temperature.exp(), 0.1, 10.0).item())-def save_checkpoint(model, tokenizer, output_dir, epoch, final=False):+def save_checkpoint(model, tokenizer, cfg, output_dir, epoch, final=False):path = Path(output_dir) if final else Path(output_dir) / "checkpoint_latest"path.mkdir(parents=True, exist_ok=True)weights = {@@ -203,6 +230,8 @@ def save_checkpoint(model, tokenizer, output_dir, epoch, final=False):tokenizer.save_pretrained(path / "tokenizer")with open(path / "checkpoint_meta.json", "w") as f:json.dump({"epoch": epoch, "final": final}, f, indent=2)+ with open(path / "rl_agent_config.json", "w") as f:+ json.dump(cfg, f, indent=2)def train(args, model_dir, items_path, device):@@ -304,7 +333,7 @@ def train(args, model_dir, items_path, device):avg_loss = total_loss / max(1, n_batches)print(f"Epoch {epoch + 1}/{args.epochs} complete; avg_loss={avg_loss:.4f}")- save_checkpoint(model, tokenizer, args.output_dir, epoch + 1)+ save_checkpoint(model, tokenizer, cfg, args.output_dir, epoch + 1)print("Running temperature calibration ...")model.eval()@@ -323,7 +352,7 @@ def train(args, model_dir, items_path, device):)temperatures = [fit_temperature(group) if group else 1.2 for group in samples]- save_checkpoint(model, tokenizer, args.output_dir, args.epochs, final=True)+ save_checkpoint(model, tokenizer, cfg, args.output_dir, args.epochs, final=True)cfg.update({"fine_tuned": True,"model_name": "laya-typed-decisions",@@ -332,6 +361,8 @@ def train(args, model_dir, items_path, device):cfg.pop("temperature_by_options", None)with open(Path(args.output_dir) / "rl_agent_config.json", "w") as f:json.dump(cfg, f, indent=2)+ with open(Path(args.output_dir) / "checkpoint_latest" / "rl_agent_config.json", "w") as f:+ json.dump(cfg, f, indent=2)print(f"Model saved to {args.output_dir}")print(f"Temperatures: {temperatures}")