Model save

README.md

@@ -0,0 +1,67 @@
+---
+library_name: transformers
+license: llama3.1
+base_model: meta-llama/Llama-3.1-8B
+tags:
+- trl
+- sft
+- generated_from_trainer
+datasets:
+- generator
+model-index:
+- name: llama-3.1-8b-sft-full_vanilla_bon-sft
+  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. -->
+
+# llama-3.1-8b-sft-full_vanilla_bon-sft
+
+This model is a fine-tuned version of [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) on the generator dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.3956
+
+## 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: 2e-05
+- train_batch_size: 4
+- eval_batch_size: 4
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 8
+- total_train_batch_size: 32
+- total_eval_batch_size: 32
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 1
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss |
+|:-------------:|:-----:|:----:|:---------------:|
+| 1.3769        | 1.0   | 59   | 1.3956          |
+
+
+### Framework versions
+
+- Transformers 4.49.0
+- Pytorch 2.6.0+cu124
+- Datasets 3.3.2
+- Tokenizers 0.21.0

all_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 1.0,
+    "total_flos": 12353399685120.0,
+    "train_loss": 1.5075225143109339,
+    "train_runtime": 421.4624,
+    "train_samples": 18000,
+    "train_samples_per_second": 4.411,
+    "train_steps_per_second": 0.14
+}

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 128000,
+  "do_sample": true,
+  "eos_token_id": 128001,
+  "temperature": 0.6,
+  "top_p": 0.9,
+  "transformers_version": "4.49.0"
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 1.0,
+    "total_flos": 12353399685120.0,
+    "train_loss": 1.5075225143109339,
+    "train_runtime": 421.4624,
+    "train_samples": 18000,
+    "train_samples_per_second": 4.411,
+    "train_steps_per_second": 0.14
+}

trainer_state.json

@@ -0,0 +1,134 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 59,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01694915254237288,
+      "grad_norm": 60.88521024203674,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.8472,
+      "step": 1
+    },
+    {
+      "epoch": 0.0847457627118644,
+      "grad_norm": 8.632962220476115,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.8862,
+      "step": 5
+    },
+    {
+      "epoch": 0.1694915254237288,
+      "grad_norm": 2.728390119480778,
+      "learning_rate": 1.972022914080411e-05,
+      "loss": 1.701,
+      "step": 10
+    },
+    {
+      "epoch": 0.2542372881355932,
+      "grad_norm": 1.653033698027206,
+      "learning_rate": 1.8610436117673557e-05,
+      "loss": 1.5482,
+      "step": 15
+    },
+    {
+      "epoch": 0.3389830508474576,
+      "grad_norm": 1.5784684295339002,
+      "learning_rate": 1.6749830015182106e-05,
+      "loss": 1.5452,
+      "step": 20
+    },
+    {
+      "epoch": 0.423728813559322,
+      "grad_norm": 1.3995751820220748,
+      "learning_rate": 1.4300652022765207e-05,
+      "loss": 1.5062,
+      "step": 25
+    },
+    {
+      "epoch": 0.5084745762711864,
+      "grad_norm": 1.231955445588447,
+      "learning_rate": 1.1476465640024814e-05,
+      "loss": 1.4419,
+      "step": 30
+    },
+    {
+      "epoch": 0.5932203389830508,
+      "grad_norm": 1.1976848012990644,
+      "learning_rate": 8.52353435997519e-06,
+      "loss": 1.4424,
+      "step": 35
+    },
+    {
+      "epoch": 0.6779661016949152,
+      "grad_norm": 1.1650074053534814,
+      "learning_rate": 5.699347977234799e-06,
+      "loss": 1.4162,
+      "step": 40
+    },
+    {
+      "epoch": 0.7627118644067796,
+      "grad_norm": 1.086038145229086,
+      "learning_rate": 3.250169984817897e-06,
+      "loss": 1.3829,
+      "step": 45
+    },
+    {
+      "epoch": 0.847457627118644,
+      "grad_norm": 1.022039910463428,
+      "learning_rate": 1.3895638823264447e-06,
+      "loss": 1.3877,
+      "step": 50
+    },
+    {
+      "epoch": 0.9322033898305084,
+      "grad_norm": 1.0706264570716404,
+      "learning_rate": 2.7977085919589253e-07,
+      "loss": 1.3769,
+      "step": 55
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.3955787420272827,
+      "eval_runtime": 10.4762,
+      "eval_samples_per_second": 19.091,
+      "eval_steps_per_second": 0.668,
+      "step": 59
+    },
+    {
+      "epoch": 1.0,
+      "step": 59,
+      "total_flos": 12353399685120.0,
+      "train_loss": 1.5075225143109339,
+      "train_runtime": 421.4624,
+      "train_samples_per_second": 4.411,
+      "train_steps_per_second": 0.14
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 59,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 12353399685120.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}