End of training

README.md

@@ -3,11 +3,15 @@ library_name: transformers
 license: llama3
 base_model: meta-llama/Meta-Llama-3-8B-Instruct
 tags:
+- alignment-handbook
+- trl
+- sft
+- generated_from_trainer
 - trl
 - sft
 - generated_from_trainer
 datasets:
-- generator
+- train_data/story_generation
 model-index:
 - name: Meta-Llama-3-8B-IST_StoryGen
   results: []
@@ -18,9 +22,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Meta-Llama-3-8B-IST_StoryGen
 
-This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the generator dataset.
+This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the train_data/story_generation dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.4415
+- Loss: 1.4676
 
 ## Model description
 

all_results.json

@@ -1,5 +1,10 @@
 {
     "epoch": 1.0,
+    "eval_loss": 1.4675722122192383,
+    "eval_runtime": 19.4256,
+    "eval_samples": 614,
+    "eval_samples_per_second": 22.959,
+    "eval_steps_per_second": 0.721,
     "total_flos": 1.4371645474744566e+18,
     "train_loss": 0.45526161300328233,
     "train_runtime": 3331.5803,

config.json

@@ -24,6 +24,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.44.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 128256
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 1.0,
+    "eval_loss": 1.4675722122192383,
+    "eval_runtime": 19.4256,
+    "eval_samples": 614,
+    "eval_samples_per_second": 22.959,
+    "eval_steps_per_second": 0.721
+}