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-1
+  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-1
+
+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: 0.0500
+
+## 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 |
+|:-------------:|:-----:|:----:|:---------------:|
+| 1.1144        | 1.0   | 2    | 1.2928          |
+| 1.1144        | 2.0   | 4    | 0.8673          |
+| 1.0902        | 3.0   | 6    | 0.6944          |
+| 1.0902        | 4.0   | 8    | 0.4531          |
+| 0.5758        | 5.0   | 10   | 0.3128          |
+| 0.5758        | 6.0   | 12   | 0.1866          |
+| 0.5758        | 7.0   | 14   | 0.1056          |
+| 0.1998        | 8.0   | 16   | 0.0681          |
+| 0.1998        | 9.0   | 18   | 0.0533          |
+| 0.0645        | 10.0  | 20   | 0.0500          |
+
+
+### 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": 0.4837849587202072,
+    "train_runtime": 336.0147,
+    "train_samples": 851,
+    "train_samples_per_second": 1.429,
+    "train_steps_per_second": 0.06
+}

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": 0.4837849587202072,
+    "train_runtime": 336.0147,
+    "train_samples": 851,
+    "train_samples_per_second": 1.429,
+    "train_steps_per_second": 0.06
+}

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": 17.86797523498535,
+      "learning_rate": 1e-05,
+      "loss": 1.1144,
+      "step": 1
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.2928184270858765,
+      "eval_runtime": 5.1117,
+      "eval_samples_per_second": 9.39,
+      "eval_steps_per_second": 0.391,
+      "step": 2
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.8672829270362854,
+      "eval_runtime": 5.1217,
+      "eval_samples_per_second": 9.372,
+      "eval_steps_per_second": 0.39,
+      "step": 4
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 25.679489135742188,
+      "learning_rate": 8.94570254698197e-06,
+      "loss": 1.0902,
+      "step": 5
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.6943734288215637,
+      "eval_runtime": 5.1377,
+      "eval_samples_per_second": 9.343,
+      "eval_steps_per_second": 0.389,
+      "step": 6
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 0.4531029760837555,
+      "eval_runtime": 5.1394,
+      "eval_samples_per_second": 9.34,
+      "eval_steps_per_second": 0.389,
+      "step": 8
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 21.12933349609375,
+      "learning_rate": 5.412896727361663e-06,
+      "loss": 0.5758,
+      "step": 10
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 0.31279101967811584,
+      "eval_runtime": 5.1429,
+      "eval_samples_per_second": 9.333,
+      "eval_steps_per_second": 0.389,
+      "step": 10
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 0.18658435344696045,
+      "eval_runtime": 5.1385,
+      "eval_samples_per_second": 9.341,
+      "eval_steps_per_second": 0.389,
+      "step": 12
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 0.1056341603398323,
+      "eval_runtime": 5.1413,
+      "eval_samples_per_second": 9.336,
+      "eval_steps_per_second": 0.389,
+      "step": 14
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 2.3144567012786865,
+      "learning_rate": 1.6135921418712959e-06,
+      "loss": 0.1998,
+      "step": 15
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 0.0681278184056282,
+      "eval_runtime": 5.1435,
+      "eval_samples_per_second": 9.332,
+      "eval_steps_per_second": 0.389,
+      "step": 16
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 0.05328406020998955,
+      "eval_runtime": 5.1445,
+      "eval_samples_per_second": 9.33,
+      "eval_steps_per_second": 0.389,
+      "step": 18
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.7936142086982727,
+      "learning_rate": 0.0,
+      "loss": 0.0645,
+      "step": 20
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 0.05004188045859337,
+      "eval_runtime": 5.0849,
+      "eval_samples_per_second": 9.44,
+      "eval_steps_per_second": 0.393,
+      "step": 20
+    },
+    {
+      "epoch": 10.0,
+      "step": 20,
+      "total_flos": 1.1184101603357491e+17,
+      "train_loss": 0.4837849587202072,
+      "train_runtime": 336.0147,
+      "train_samples_per_second": 1.429,
+      "train_steps_per_second": 0.06
+    }
+  ],
+  "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
+}