Model save

README.md

@@ -0,0 +1,70 @@
+---
+base_model: mistralai/Mistral-7B-Instruct-v0.1
+library_name: peft
+license: apache-2.0
+tags:
+- trl
+- sft
+- generated_from_trainer
+model-index:
+- name: mistral_gsm8k_gold_answer_lora
+  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_gsm8k_gold_answer_lora
+
+This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.5897
+
+## 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: 0.0002
+- train_batch_size: 4
+- eval_batch_size: 4
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 2
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 16
+- total_eval_batch_size: 8
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- training_steps: 20
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss |
+|:-------------:|:------:|:----:|:---------------:|
+| 2.6708        | 0.7692 | 5    | 1.9371          |
+| 1.627         | 1.5385 | 10   | 1.6907          |
+| 1.422         | 2.3077 | 15   | 1.6038          |
+| 1.1894        | 3.0769 | 20   | 1.5897          |
+
+
+### Framework versions
+
+- PEFT 0.12.0
+- Transformers 4.44.2
+- Pytorch 2.4.1+cu121
+- Datasets 3.0.0
+- Tokenizers 0.19.1

all_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 3.076923076923077,
+    "total_flos": 507088404480.0,
+    "train_loss": 1.75083886384964,
+    "train_runtime": 83.5061,
+    "train_samples": 100,
+    "train_samples_per_second": 3.832,
+    "train_steps_per_second": 0.24
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 3.076923076923077,
+    "total_flos": 507088404480.0,
+    "train_loss": 1.75083886384964,
+    "train_runtime": 83.5061,
+    "train_samples": 100,
+    "train_samples_per_second": 3.832,
+    "train_steps_per_second": 0.24
+}

trainer_state.json

@@ -0,0 +1,109 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.076923076923077,
+  "eval_steps": 5,
+  "global_step": 20,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 2.6500688910146555,
+      "learning_rate": 0.0001,
+      "loss": 3.1414,
+      "step": 1
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 4.60486835758179,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 2.6708,
+      "step": 5
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "eval_loss": 1.937098503112793,
+      "eval_runtime": 2.5272,
+      "eval_samples_per_second": 7.914,
+      "eval_steps_per_second": 1.187,
+      "step": 5
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 1.244018594872834,
+      "learning_rate": 0.00011736481776669306,
+      "loss": 1.627,
+      "step": 10
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "eval_loss": 1.690708875656128,
+      "eval_runtime": 1.2876,
+      "eval_samples_per_second": 15.533,
+      "eval_steps_per_second": 2.33,
+      "step": 10
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 1.2737122598991255,
+      "learning_rate": 3.5721239031346066e-05,
+      "loss": 1.422,
+      "step": 15
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "eval_loss": 1.603790283203125,
+      "eval_runtime": 1.2814,
+      "eval_samples_per_second": 15.608,
+      "eval_steps_per_second": 2.341,
+      "step": 15
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 1.3878615177459241,
+      "learning_rate": 0.0,
+      "loss": 1.1894,
+      "step": 20
+    },
+    {
+      "epoch": 3.076923076923077,
+      "eval_loss": 1.5896568298339844,
+      "eval_runtime": 1.2884,
+      "eval_samples_per_second": 15.523,
+      "eval_steps_per_second": 2.329,
+      "step": 20
+    },
+    {
+      "epoch": 3.076923076923077,
+      "step": 20,
+      "total_flos": 507088404480.0,
+      "train_loss": 1.75083886384964,
+      "train_runtime": 83.5061,
+      "train_samples_per_second": 3.832,
+      "train_steps_per_second": 0.24
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 20,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 507088404480.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}