End of training

README.md

@@ -3,11 +3,16 @@ library_name: transformers
 license: llama3.1
 base_model: meta-llama/Llama-3.1-8B
 tags:
+- alignment-handbook
+- trl
+- sft
+- generated_from_trainer
 - trl
 - sft
 - generated_from_trainer
 datasets:
-- generator
+- oxford-llms/european_social_survey_2023_germany_sft
+- oxford-llms/ultrachat_filtered
 model-index:
 - name: llama3-1-ox-llms-8b-sft-only-germany-data-and-ultrafeedback
   results: []
@@ -18,7 +23,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # llama3-1-ox-llms-8b-sft-only-germany-data-and-ultrafeedback
 
-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.
+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 oxford-llms/european_social_survey_2023_germany_sft and the oxford-llms/ultrachat_filtered datasets.
 It achieves the following results on the evaluation set:
 - Loss: 1.0782
 

all_results.json

@@ -1,5 +1,10 @@
 {
     "epoch": 2.0,
+    "eval_loss": 1.07817542552948,
+    "eval_runtime": 43.2502,
+    "eval_samples": 9820,
+    "eval_samples_per_second": 75.907,
+    "eval_steps_per_second": 1.202,
     "total_flos": 169178761789440.0,
     "train_loss": 0.8271688128461933,
     "train_runtime": 2411.6836,

config.json

@@ -31,6 +31,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.45.1",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 128256
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.0,
+    "eval_loss": 1.07817542552948,
+    "eval_runtime": 43.2502,
+    "eval_samples": 9820,
+    "eval_samples_per_second": 75.907,
+    "eval_steps_per_second": 1.202
+}