Model save

.ipynb_checkpoints/lora_orpo-checkpoint.yaml

@@ -0,0 +1,43 @@
+### model
+model_name_or_path: mistralai/Mistral-7B-Instruct-v0.2
+
+### method
+stage: orpo
+do_train: true
+finetuning_type: lora
+lora_target: all
+
+### dataset
+dataset: dpo_mix_en
+dataset_dir: data
+template: mistral
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+### output
+output_dir: saves/Mistral-7B-Instruct-v0.2/lora/orpo
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+save_total_limit: 3
+load_best_model_at_end: true
+push_to_hub: true
+hub_model_id: chchen/Mistral-7B-Instruct-v0.2-ORPO
+
+### train
+per_device_train_batch_size: 2
+gradient_accumulation_steps: 8
+learning_rate: 0.000005
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+### eval
+val_size: 0.1
+per_device_eval_batch_size: 4
+evaluation_strategy: steps
+eval_steps: 500

README.md

@@ -0,0 +1,61 @@
+---
+license: apache-2.0
+library_name: peft
+tags:
+- trl
+- dpo
+- llama-factory
+- generated_from_trainer
+base_model: mistralai/Mistral-7B-Instruct-v0.2
+model-index:
+- name: Mistral-7B-Instruct-v0.2-ORPO
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# Mistral-7B-Instruct-v0.2-ORPO
+
+This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) on an unknown dataset.
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 5e-06
+- train_batch_size: 2
+- eval_batch_size: 4
+- seed: 42
+- gradient_accumulation_steps: 8
+- total_train_batch_size: 16
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_steps: 0.1
+- num_epochs: 3.0
+- mixed_precision_training: Native AMP
+
+### Training results
+
+
+
+### Framework versions
+
+- PEFT 0.10.0
+- Transformers 4.40.1
+- Pytorch 2.3.0
+- Datasets 2.19.0
+- Tokenizers 0.19.1

adapter_config.json

@@ -0,0 +1,34 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "mistralai/Mistral-7B-Instruct-v0.2",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "q_proj",
+    "k_proj",
+    "gate_proj",
+    "down_proj",
+    "v_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47477a7d72be6c59779358b5d59abe1532ce5697e206c4db569081d50b392cac
+size 83945296

lora_orpo.yaml

@@ -0,0 +1,43 @@
+### model
+model_name_or_path: mistralai/Mistral-7B-Instruct-v0.2
+
+### method
+stage: orpo
+do_train: true
+finetuning_type: lora
+lora_target: all
+
+### dataset
+dataset: dpo_mix_en
+dataset_dir: data
+template: mistral
+cutoff_len: 1024
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+### output
+output_dir: saves/Mistral-7B-Instruct-v0.2/lora/orpo
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+save_total_limit: 3
+load_best_model_at_end: true
+push_to_hub: true
+hub_model_id: chchen/Mistral-7B-Instruct-v0.2-ORPO
+
+### train
+per_device_train_batch_size: 2
+gradient_accumulation_steps: 8
+learning_rate: 0.000005
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_steps: 0.1
+fp16: true
+
+### eval
+val_size: 0.1
+per_device_eval_batch_size: 4
+evaluation_strategy: steps
+eval_steps: 500

special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "</s>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
+size 493443

tokenizer_config.json

@@ -0,0 +1,45 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [],
+  "bos_token": "<s>",
+  "chat_template": "{% if messages[0]['role'] == 'system' %}{% set system_message = messages[0]['content'] %}{% endif %}{{ '<s>' + system_message }}{% for message in messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ ' [INST] ' + content + ' [/INST]' }}{% elif message['role'] == 'assistant' %}{{ content + '</s>' }}{% endif %}{% endfor %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "legacy": true,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "</s>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "split_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

trainer_log.jsonl

@@ -0,0 +1,17 @@
+{"current_steps": 10, "total_steps": 168, "loss": 1.3004, "accuracy": 0.45625001192092896, "learning_rate": 4.957230266673969e-06, "epoch": 0.17777777777777778, "percentage": 5.95, "elapsed_time": "0:05:27", "remaining_time": "1:26:19"}
+{"current_steps": 20, "total_steps": 168, "loss": 1.1815, "accuracy": 0.48124998807907104, "learning_rate": 4.84529614391025e-06, "epoch": 0.35555555555555557, "percentage": 11.9, "elapsed_time": "0:10:47", "remaining_time": "1:19:52"}
+{"current_steps": 30, "total_steps": 168, "loss": 1.2768, "accuracy": 0.4375, "learning_rate": 4.667009949002349e-06, "epoch": 0.5333333333333333, "percentage": 17.86, "elapsed_time": "0:16:08", "remaining_time": "1:14:14"}
+{"current_steps": 40, "total_steps": 168, "loss": 1.2213, "accuracy": 0.48124998807907104, "learning_rate": 4.397288409237892e-06, "epoch": 0.7111111111111111, "percentage": 23.81, "elapsed_time": "0:21:22", "remaining_time": "1:08:22"}
+{"current_steps": 50, "total_steps": 168, "loss": 1.3385, "accuracy": 0.48124998807907104, "learning_rate": 4.097593272948218e-06, "epoch": 0.8888888888888888, "percentage": 29.76, "elapsed_time": "0:26:40", "remaining_time": "1:02:56"}
+{"current_steps": 60, "total_steps": 168, "loss": 1.1266, "accuracy": 0.550000011920929, "learning_rate": 3.712001321256744e-06, "epoch": 1.0666666666666667, "percentage": 35.71, "elapsed_time": "0:31:51", "remaining_time": "0:57:20"}
+{"current_steps": 70, "total_steps": 168, "loss": 1.1365, "accuracy": 0.512499988079071, "learning_rate": 3.284100248421622e-06, "epoch": 1.2444444444444445, "percentage": 41.67, "elapsed_time": "0:37:11", "remaining_time": "0:52:03"}
+{"current_steps": 80, "total_steps": 168, "loss": 1.1088, "accuracy": 0.4749999940395355, "learning_rate": 2.8288274295187397e-06, "epoch": 1.4222222222222223, "percentage": 47.62, "elapsed_time": "0:42:29", "remaining_time": "0:46:43"}
+{"current_steps": 90, "total_steps": 168, "loss": 1.1411, "accuracy": 0.512499988079071, "learning_rate": 2.3620757456122797e-06, "epoch": 1.6, "percentage": 53.57, "elapsed_time": "0:47:40", "remaining_time": "0:41:18"}
+{"current_steps": 100, "total_steps": 168, "loss": 1.1326, "accuracy": 0.5625, "learning_rate": 1.9001387874698865e-06, "epoch": 1.7777777777777777, "percentage": 59.52, "elapsed_time": "0:53:01", "remaining_time": "0:36:03"}
+{"current_steps": 110, "total_steps": 168, "loss": 1.1523, "accuracy": 0.4937500059604645, "learning_rate": 1.4591420711120425e-06, "epoch": 1.9555555555555557, "percentage": 65.48, "elapsed_time": "0:58:17", "remaining_time": "0:30:44"}
+{"current_steps": 120, "total_steps": 168, "loss": 1.1173, "accuracy": 0.48750001192092896, "learning_rate": 1.0544801205950327e-06, "epoch": 2.1333333333333333, "percentage": 71.43, "elapsed_time": "1:03:36", "remaining_time": "0:25:26"}
+{"current_steps": 130, "total_steps": 168, "loss": 1.0107, "accuracy": 0.48750001192092896, "learning_rate": 7.002790686106056e-07, "epoch": 2.311111111111111, "percentage": 77.38, "elapsed_time": "1:08:55", "remaining_time": "0:20:08"}
+{"current_steps": 140, "total_steps": 168, "loss": 1.1034, "accuracy": 0.6000000238418579, "learning_rate": 4.0890353469735717e-07, "epoch": 2.488888888888889, "percentage": 83.33, "elapsed_time": "1:14:09", "remaining_time": "0:14:49"}
+{"current_steps": 150, "total_steps": 168, "loss": 1.2074, "accuracy": 0.48124998807907104, "learning_rate": 1.905249951948071e-07, "epoch": 2.6666666666666665, "percentage": 89.29, "elapsed_time": "1:19:35", "remaining_time": "0:09:33"}
+{"current_steps": 160, "total_steps": 168, "loss": 1.0456, "accuracy": 0.581250011920929, "learning_rate": 5.2766712488021566e-08, "epoch": 2.8444444444444446, "percentage": 95.24, "elapsed_time": "1:24:53", "remaining_time": "0:04:14"}
+{"current_steps": 168, "total_steps": 168, "epoch": 2.986666666666667, "percentage": 100.0, "elapsed_time": "1:29:12", "remaining_time": "0:00:00"}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e875a0588d918bab72083c41753731da8efab74b78ea8d8dc9bfa65b5b65635
+size 5240