Model save

README.md

@@ -0,0 +1,76 @@
+---
+library_name: transformers
+license: apache-2.0
+base_model: alignment-handbook/zephyr-7b-sft-full
+tags:
+- trl
+- sft
+- generated_from_trainer
+datasets:
+- generator
+model-index:
+- name: zephyr-7b-stem-case-3
+  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. -->
+
+# zephyr-7b-stem-case-3
+
+This model is a fine-tuned version of [alignment-handbook/zephyr-7b-sft-full](https://huggingface.co/alignment-handbook/zephyr-7b-sft-full) on the generator dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0406
+
+## 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-05
+- train_batch_size: 8
+- eval_batch_size: 8
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- total_train_batch_size: 32
+- total_eval_batch_size: 32
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.03
+- num_epochs: 10
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss |
+|:-------------:|:-----:|:----:|:---------------:|
+| 1.1002        | 1.0   | 2    | 1.2712          |
+| 1.1002        | 2.0   | 4    | 0.8801          |
+| 1.0859        | 3.0   | 6    | 0.6101          |
+| 1.0859        | 4.0   | 8    | 0.4419          |
+| 0.5465        | 5.0   | 10   | 0.2620          |
+| 0.5465        | 6.0   | 12   | 0.1413          |
+| 0.5465        | 7.0   | 14   | 0.0808          |
+| 0.1621        | 8.0   | 16   | 0.0535          |
+| 0.1621        | 9.0   | 18   | 0.0429          |
+| 0.0504        | 10.0  | 20   | 0.0406          |
+
+
+### Framework versions
+
+- 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": 10.0,
+    "total_flos": 1.1184101603357491e+17,
+    "train_loss": 0.46195218414068223,
+    "train_runtime": 339.5674,
+    "train_samples": 848,
+    "train_samples_per_second": 1.472,
+    "train_steps_per_second": 0.059
+}

generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "transformers_version": "4.44.2"
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 10.0,
+    "total_flos": 1.1184101603357491e+17,
+    "train_loss": 0.46195218414068223,
+    "train_runtime": 339.5674,
+    "train_samples": 848,
+    "train_samples_per_second": 1.472,
+    "train_steps_per_second": 0.059
+}

trainer_state.json

@@ -0,0 +1,157 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 20,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.5,
+      "grad_norm": 18.081775665283203,
+      "learning_rate": 1e-05,
+      "loss": 1.1002,
+      "step": 1
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.2711809873580933,
+      "eval_runtime": 5.1274,
+      "eval_samples_per_second": 9.752,
+      "eval_steps_per_second": 0.39,
+      "step": 2
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.8800549507141113,
+      "eval_runtime": 5.1265,
+      "eval_samples_per_second": 9.753,
+      "eval_steps_per_second": 0.39,
+      "step": 4
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 19.409046173095703,
+      "learning_rate": 8.94570254698197e-06,
+      "loss": 1.0859,
+      "step": 5
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.6101047992706299,
+      "eval_runtime": 5.1466,
+      "eval_samples_per_second": 9.715,
+      "eval_steps_per_second": 0.389,
+      "step": 6
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 0.44190600514411926,
+      "eval_runtime": 5.1566,
+      "eval_samples_per_second": 9.696,
+      "eval_steps_per_second": 0.388,
+      "step": 8
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 7.709940433502197,
+      "learning_rate": 5.412896727361663e-06,
+      "loss": 0.5465,
+      "step": 10
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 0.26203182339668274,
+      "eval_runtime": 5.1422,
+      "eval_samples_per_second": 9.724,
+      "eval_steps_per_second": 0.389,
+      "step": 10
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 0.14132770895957947,
+      "eval_runtime": 5.1462,
+      "eval_samples_per_second": 9.716,
+      "eval_steps_per_second": 0.389,
+      "step": 12
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 0.08078073710203171,
+      "eval_runtime": 5.1421,
+      "eval_samples_per_second": 9.724,
+      "eval_steps_per_second": 0.389,
+      "step": 14
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 1.5183640718460083,
+      "learning_rate": 1.6135921418712959e-06,
+      "loss": 0.1621,
+      "step": 15
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 0.05351748317480087,
+      "eval_runtime": 5.1503,
+      "eval_samples_per_second": 9.708,
+      "eval_steps_per_second": 0.388,
+      "step": 16
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 0.042922407388687134,
+      "eval_runtime": 5.1564,
+      "eval_samples_per_second": 9.697,
+      "eval_steps_per_second": 0.388,
+      "step": 18
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.851169228553772,
+      "learning_rate": 0.0,
+      "loss": 0.0504,
+      "step": 20
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 0.04061064124107361,
+      "eval_runtime": 5.054,
+      "eval_samples_per_second": 9.893,
+      "eval_steps_per_second": 0.396,
+      "step": 20
+    },
+    {
+      "epoch": 10.0,
+      "step": 20,
+      "total_flos": 1.1184101603357491e+17,
+      "train_loss": 0.46195218414068223,
+      "train_runtime": 339.5674,
+      "train_samples_per_second": 1.472,
+      "train_steps_per_second": 0.059
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 20,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "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": 1.1184101603357491e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}