Model save

README.md

@@ -0,0 +1,58 @@
+---
+license: apache-2.0
+base_model: Qwen/Qwen2.5-7B
+tags:
+- generated_from_trainer
+model-index:
+- name: Qwen2.5-7B-dpo-2k
+  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. -->
+
+# Qwen2.5-7B-dpo-2k
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-7B](https://huggingface.co/Qwen/Qwen2.5-7B) 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-07
+- train_batch_size: 4
+- eval_batch_size: 4
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 32
+- total_eval_batch_size: 16
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 1
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.43.3
+- Pytorch 2.2.2+cu121
+- Datasets 2.20.0
+- Tokenizers 0.19.1

all_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.992,
+    "total_flos": 0.0,
+    "train_loss": 0.8402291834354401,
+    "train_runtime": 793.2299,
+    "train_samples": 1992,
+    "train_samples_per_second": 2.511,
+    "train_steps_per_second": 0.078
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "max_new_tokens": 2048,
+  "transformers_version": "4.43.3"
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.992,
+    "total_flos": 0.0,
+    "train_loss": 0.8402291834354401,
+    "train_runtime": 793.2299,
+    "train_samples": 1992,
+    "train_samples_per_second": 2.511,
+    "train_steps_per_second": 0.078
+}

trainer_state.json

@@ -0,0 +1,147 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.992,
+  "eval_steps": 200,
+  "global_step": 62,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.016,
+      "grad_norm": 54.68133828489169,
+      "learning_rate": 7.142857142857142e-08,
+      "logits/generated": -1.112015962600708,
+      "logits/real": -0.9725905656814575,
+      "logps/generated": -192.91046142578125,
+      "logps/real": -197.25250244140625,
+      "loss": 0.9186,
+      "rewards/accuracies": 0.0,
+      "rewards/generated": 0.0,
+      "rewards/margins": 0.0,
+      "rewards/real": 0.0,
+      "step": 1
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 78.08601698235012,
+      "learning_rate": 4.727272727272727e-07,
+      "logits/generated": -0.9562382102012634,
+      "logits/real": -0.9224010109901428,
+      "logps/generated": -202.88710021972656,
+      "logps/real": -216.11236572265625,
+      "loss": 0.8993,
+      "rewards/accuracies": 0.4722222089767456,
+      "rewards/generated": 0.16792699694633484,
+      "rewards/margins": 0.015514086931943893,
+      "rewards/real": 0.18344110250473022,
+      "step": 10
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 52.37055337168455,
+      "learning_rate": 3.818181818181818e-07,
+      "logits/generated": -0.9773967862129211,
+      "logits/real": -0.9421946406364441,
+      "logps/generated": -205.1346435546875,
+      "logps/real": -216.21044921875,
+      "loss": 0.8759,
+      "rewards/accuracies": 0.612500011920929,
+      "rewards/generated": 1.164185881614685,
+      "rewards/margins": 0.03111693263053894,
+      "rewards/real": 1.1953028440475464,
+      "step": 20
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 45.56925653007757,
+      "learning_rate": 2.909090909090909e-07,
+      "logits/generated": -1.0400158166885376,
+      "logits/real": -0.8696712255477905,
+      "logps/generated": -174.08009338378906,
+      "logps/real": -180.80616760253906,
+      "loss": 0.8296,
+      "rewards/accuracies": 0.6000000238418579,
+      "rewards/generated": 1.7348871231079102,
+      "rewards/margins": 0.13560794293880463,
+      "rewards/real": 1.870495080947876,
+      "step": 30
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 47.856709569705075,
+      "learning_rate": 2e-07,
+      "logits/generated": -1.0196596384048462,
+      "logits/real": -0.8638933897018433,
+      "logps/generated": -176.03512573242188,
+      "logps/real": -207.9772491455078,
+      "loss": 0.824,
+      "rewards/accuracies": 0.675000011920929,
+      "rewards/generated": 2.319814920425415,
+      "rewards/margins": 0.2503294050693512,
+      "rewards/real": 2.5701441764831543,
+      "step": 40
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 45.32128168371637,
+      "learning_rate": 1.0909090909090908e-07,
+      "logits/generated": -1.049173355102539,
+      "logits/real": -0.8328621983528137,
+      "logps/generated": -177.70408630371094,
+      "logps/real": -205.44827270507812,
+      "loss": 0.808,
+      "rewards/accuracies": 0.6499999761581421,
+      "rewards/generated": 2.6795811653137207,
+      "rewards/margins": 0.25618499517440796,
+      "rewards/real": 2.9357664585113525,
+      "step": 50
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 55.117145590325535,
+      "learning_rate": 1.818181818181818e-08,
+      "logits/generated": -1.1112569570541382,
+      "logits/real": -1.001379132270813,
+      "logps/generated": -170.5053253173828,
+      "logps/real": -174.80856323242188,
+      "loss": 0.8158,
+      "rewards/accuracies": 0.6625000238418579,
+      "rewards/generated": 2.7520089149475098,
+      "rewards/margins": 0.20990166068077087,
+      "rewards/real": 2.9619107246398926,
+      "step": 60
+    },
+    {
+      "epoch": 0.992,
+      "step": 62,
+      "total_flos": 0.0,
+      "train_loss": 0.8402291834354401,
+      "train_runtime": 793.2299,
+      "train_samples_per_second": 2.511,
+      "train_steps_per_second": 0.078
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 62,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}