Upload folder using huggingface_hub

recipes/orpo-fixer-replanner-0.5b-iter2.yaml

@@ -0,0 +1,48 @@
+# ORPO iter-2 RE-PLANNER fixer on 0.5B base (per user 0.5B preference).
+# Teaches fixer to PRODUCE a correct alternative when given a failed planner attempt.
+# Starts from existing 0.5B ORPO fixer iter-1.
+model_name_or_path: ./output/qwen-coder0.5b-bird-fixer-orpo
+torch_dtype: bfloat16
+use_flash_attention_2: false
+
+dataset_mixer:
+  ../data/llm_alignment/scaleup_iter2_v3/hf_fixer_replanner: 1.0
+
+dataset_splits:
+- train_dpo
+- test_dpo
+preprocessing_num_workers: 12
+chat_template: "{{'<|im_start|>user\n' + messages['prompt'] + '<|im_end|>\n'}}{{'<|im_start|>assistant\n' + messages['completion'] + '<|im_end|>\n'}}"
+report_to: ["tensorboard"]
+
+bf16: true
+beta: 0.5
+do_eval: true
+eval_strategy: "steps"
+eval_steps: 100
+gradient_accumulation_steps: 8
+gradient_checkpointing: true
+gradient_checkpointing_kwargs:
+  use_reentrant: False
+learning_rate: 1.0e-6
+log_level: info
+logging_steps: 10
+lr_scheduler_type: inverse_sqrt
+max_length: 6144
+max_prompt_length: 5500
+num_train_epochs: -1
+max_steps: 400
+optim: adamw_torch
+output_dir: output/qwen-coder0.5b-scaleup-fixer-replanner-orpo-iter2
+overwrite_output_dir: true
+per_device_train_batch_size: 1
+per_device_eval_batch_size: 1
+push_to_hub: false
+remove_unused_columns: false
+save_strategy: "steps"
+save_steps: 400
+save_total_limit: 1
+seed: 42
+warmup_ratio: 0.05
+warmup_steps: 40
+max_grad_norm: 0.5

recipes/orpo-planner-collab-iter2.yaml

@@ -0,0 +1,46 @@
+# ORPO iter-2 on planner-COLLAB-iter1 — starts from prior ORPO checkpoint, uses new rollouts.
+model_name_or_path: ./output/qwen-coder3b-scaleup-planner-COLLAB-orpo
+torch_dtype: bfloat16
+use_flash_attention_2: false
+
+dataset_mixer:
+  ../data/llm_alignment/scaleup_iter2/hf_planner_collaborative: 1.0
+
+dataset_splits:
+- train_dpo
+- test_dpo
+preprocessing_num_workers: 12
+chat_template: "{{'<|im_start|>user\n' + messages['prompt'] + '<|im_end|>\n'}}{{'<|im_start|>assistant\n' + messages['completion'] + '<|im_end|>\n'}}"
+report_to: ["tensorboard"]
+
+bf16: true
+beta: 1.0
+do_eval: true
+eval_strategy: "steps"
+eval_steps: 100
+gradient_accumulation_steps: 16
+gradient_checkpointing: true
+gradient_checkpointing_kwargs:
+  use_reentrant: False
+learning_rate: 1.0e-6
+log_level: info
+logging_steps: 10
+lr_scheduler_type: inverse_sqrt
+max_length: 4000
+max_prompt_length: 3500
+num_train_epochs: -1
+max_steps: 200
+optim: adamw_torch
+output_dir: output/qwen-coder3b-scaleup-planner-COLLAB-orpo-iter2
+overwrite_output_dir: true
+per_device_train_batch_size: 1
+per_device_eval_batch_size: 1
+push_to_hub: false
+remove_unused_columns: false
+save_strategy: "steps"
+save_steps: 200
+save_total_limit: 1
+seed: 42
+warmup_ratio: 0.1
+warmup_steps: 50
+max_grad_norm: 0.5

recipes/validator-condition-fft-0.5b-v3.yaml

@@ -0,0 +1,41 @@
+# SFT Qwen2.5-Coder-0.5B Validator-Condition (v_c) — 0.5B base per user constraint.
+model_name_or_path: /home/datht/huggingface/Qwen/Qwen2.5-Coder-0.5B-Instruct
+model_revision: main
+torch_dtype: bfloat16
+use_flash_attention_2: false
+response_template: "<|im_start|>assistant"
+
+dataset_mixer:
+  ../data/multi-agents/fixed/sft-validator-condition-v3: 1.0
+dataset_splits:
+- train
+- test
+preprocessing_num_workers: 24
+chat_template: "{{'<|im_start|>user\n' + messages['prompt'] + '<|im_end|>\n'}}{{'<|im_start|>assistant\n' + messages['completion'] + '<|im_end|>\n'}}"
+
+bf16: true
+do_eval: false
+eval_strategy: "no"
+gradient_accumulation_steps: 4
+gradient_checkpointing: true
+learning_rate: 2.0e-05
+log_level: info
+logging_steps: 10
+logging_strategy: steps
+lr_scheduler_type: cosine
+max_seq_length: 5120
+truncation_side: left
+max_steps: -1
+num_train_epochs: 2
+optim: adamw_torch
+output_dir: output/qwen-coder0.5b-bird-validator-condition-sft-v3
+overwrite_output_dir: true
+per_device_train_batch_size: 8
+push_to_hub: false
+remove_unused_columns: true
+report_to:
+- tensorboard
+save_strategy: "epoch"
+save_total_limit: 1
+seed: 42
+tf32: true

recipes/validator-condition-fft-v3.yaml

@@ -0,0 +1,42 @@
+# SFT Qwen3-0.6B Validator-Condition (v_c) on section-specific data.
+# Per paper §Combined Validator, v_c critiques only the WHERE/HAVING/CASE conditions.
+model_name_or_path: /home/datht/huggingface/Qwen/Qwen3-0.6B
+model_revision: main
+torch_dtype: bfloat16
+use_flash_attention_2: false
+response_template: "<|im_start|>assistant"
+
+dataset_mixer:
+  ../data/multi-agents/fixed/sft-validator-condition-v3: 1.0
+dataset_splits:
+- train
+- test
+preprocessing_num_workers: 24
+chat_template: "{{'<|im_start|>user\n' + messages['prompt'] + '<|im_end|>\n'}}{{'<|im_start|>assistant\n' + messages['completion'] + '<|im_end|>\n'}}"
+
+bf16: true
+do_eval: false
+eval_strategy: "no"
+gradient_accumulation_steps: 8
+gradient_checkpointing: true
+learning_rate: 2.0e-05
+log_level: info
+logging_steps: 10
+logging_strategy: steps
+lr_scheduler_type: cosine
+max_seq_length: 5120
+truncation_side: left
+max_steps: -1
+num_train_epochs: 2
+optim: adamw_torch
+output_dir: output/qwen3-0.6b-bird-validator-condition-sft-v3
+overwrite_output_dir: true
+per_device_train_batch_size: 4
+push_to_hub: false
+remove_unused_columns: true
+report_to:
+- tensorboard
+save_strategy: "epoch"
+save_total_limit: 1
+seed: 42
+tf32: true

recipes/validator-selection-fft-0.5b-v3.yaml

@@ -0,0 +1,41 @@
+# SFT Qwen2.5-Coder-0.5B Validator-Selection (v_s) — 0.5B base per user constraint.
+model_name_or_path: /home/datht/huggingface/Qwen/Qwen2.5-Coder-0.5B-Instruct
+model_revision: main
+torch_dtype: bfloat16
+use_flash_attention_2: false
+response_template: "<|im_start|>assistant"
+
+dataset_mixer:
+  ../data/multi-agents/fixed/sft-validator-selection-v3: 1.0
+dataset_splits:
+- train
+- test
+preprocessing_num_workers: 24
+chat_template: "{{'<|im_start|>user\n' + messages['prompt'] + '<|im_end|>\n'}}{{'<|im_start|>assistant\n' + messages['completion'] + '<|im_end|>\n'}}"
+
+bf16: true
+do_eval: false
+eval_strategy: "no"
+gradient_accumulation_steps: 4
+gradient_checkpointing: true
+learning_rate: 2.0e-05
+log_level: info
+logging_steps: 10
+logging_strategy: steps
+lr_scheduler_type: cosine
+max_seq_length: 5120
+truncation_side: left
+max_steps: -1
+num_train_epochs: 2
+optim: adamw_torch
+output_dir: output/qwen-coder0.5b-bird-validator-selection-sft-v3
+overwrite_output_dir: true
+per_device_train_batch_size: 8
+push_to_hub: false
+remove_unused_columns: true
+report_to:
+- tensorboard
+save_strategy: "epoch"
+save_total_limit: 1
+seed: 42
+tf32: true