Model save

README.md

@@ -0,0 +1,74 @@
+---
+library_name: peft
+license: apache-2.0
+base_model: HuggingFaceTB/SmolLM2-1.7B-Instruct
+tags:
+- trl
+- dpo
+- generated_from_trainer
+model-index:
+- name: smollm2-17b-dpo-cai-v1
+  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. -->
+
+# smollm2-17b-dpo-cai-v1
+
+This model is a fine-tuned version of [HuggingFaceTB/SmolLM2-1.7B-Instruct](https://huggingface.co/HuggingFaceTB/SmolLM2-1.7B-Instruct) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.6929
+- Rewards/chosen: -0.0007
+- Rewards/rejected: -0.0011
+- Rewards/accuracies: 0.4487
+- Rewards/margins: 0.0004
+- Logps/rejected: -228.6854
+- Logps/chosen: -245.1230
+- Logits/rejected: -1.2768
+- Logits/chosen: -1.2404
+
+## 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: 1e-06
+- train_batch_size: 4
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 16
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 1
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
+|:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
+| 0.6933        | 0.3883 | 100  | 0.6930          | -0.0013        | -0.0015          | 0.4358             | 0.0002          | -228.7255      | -245.1841    | -1.6940         | -1.6433       |
+| 0.6931        | 0.7767 | 200  | 0.6929          | -0.0007        | -0.0011          | 0.4487             | 0.0004          | -228.6854      | -245.1230    | -1.2768         | -1.2404       |
+
+
+### Framework versions
+
+- PEFT 0.14.0
+- Transformers 4.45.0
+- Pytorch 2.5.1+cu124
+- Datasets 3.2.0
+- Tokenizers 0.20.3

all_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.9980582524271845,
+    "total_flos": 0.0,
+    "train_loss": 0.6931419428220519,
+    "train_runtime": 4635.9681,
+    "train_samples": 4119,
+    "train_samples_per_second": 0.888,
+    "train_steps_per_second": 0.055
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.9980582524271845,
+    "total_flos": 0.0,
+    "train_loss": 0.6931419428220519,
+    "train_runtime": 4635.9681,
+    "train_samples": 4119,
+    "train_samples_per_second": 0.888,
+    "train_steps_per_second": 0.055
+}

trainer_state.json

@@ -0,0 +1,119 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9980582524271845,
+  "eval_steps": 100,
+  "global_step": 257,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003883495145631068,
+      "grad_norm": 0.06435743719339371,
+      "learning_rate": 3.846153846153846e-08,
+      "logits/chosen": -0.8751495480537415,
+      "logits/rejected": -0.7611837387084961,
+      "logps/chosen": -216.45455932617188,
+      "logps/rejected": -194.20330810546875,
+      "loss": 0.6931,
+      "rewards/accuracies": 0.0,
+      "rewards/chosen": 0.0,
+      "rewards/margins": 0.0,
+      "rewards/rejected": 0.0,
+      "step": 1
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 0.08497226983308792,
+      "learning_rate": 7.674539256930362e-07,
+      "logits/chosen": -1.1164360046386719,
+      "logits/rejected": -1.1212449073791504,
+      "logps/chosen": -261.63470458984375,
+      "logps/rejected": -252.7709197998047,
+      "loss": 0.6933,
+      "rewards/accuracies": 0.43686869740486145,
+      "rewards/chosen": -0.0006324161658994853,
+      "rewards/margins": -0.0002744389639701694,
+      "rewards/rejected": -0.0003579772310331464,
+      "step": 100
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "eval_logits/chosen": -1.6433279514312744,
+      "eval_logits/rejected": -1.6940118074417114,
+      "eval_logps/chosen": -245.18409729003906,
+      "eval_logps/rejected": -228.72549438476562,
+      "eval_loss": 0.6930427551269531,
+      "eval_rewards/accuracies": 0.4357594847679138,
+      "eval_rewards/chosen": -0.0013237559469416738,
+      "eval_rewards/margins": 0.00021979793382342905,
+      "eval_rewards/rejected": -0.001543553895317018,
+      "eval_runtime": 978.8839,
+      "eval_samples_per_second": 3.224,
+      "eval_steps_per_second": 0.404,
+      "step": 100
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.06872538477182388,
+      "learning_rate": 1.428588273890482e-07,
+      "logits/chosen": -1.0920249223709106,
+      "logits/rejected": -1.1872063875198364,
+      "logps/chosen": -267.94097900390625,
+      "logps/rejected": -240.1817169189453,
+      "loss": 0.6931,
+      "rewards/accuracies": 0.453125,
+      "rewards/chosen": -0.0006506562349386513,
+      "rewards/margins": 0.00018236432515550405,
+      "rewards/rejected": -0.0008330204873345792,
+      "step": 200
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "eval_logits/chosen": -1.2404372692108154,
+      "eval_logits/rejected": -1.2768234014511108,
+      "eval_logps/chosen": -245.1230010986328,
+      "eval_logps/rejected": -228.6853790283203,
+      "eval_loss": 0.6929394602775574,
+      "eval_rewards/accuracies": 0.4487341642379761,
+      "eval_rewards/chosen": -0.0007124166586436331,
+      "eval_rewards/margins": 0.0004301773733459413,
+      "eval_rewards/rejected": -0.0011425940319895744,
+      "eval_runtime": 981.2056,
+      "eval_samples_per_second": 3.216,
+      "eval_steps_per_second": 0.403,
+      "step": 200
+    },
+    {
+      "epoch": 0.9980582524271845,
+      "step": 257,
+      "total_flos": 0.0,
+      "train_loss": 0.6931419428220519,
+      "train_runtime": 4635.9681,
+      "train_samples_per_second": 0.888,
+      "train_steps_per_second": 0.055
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 257,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "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
+}