End of training

README.md

@@ -1,11 +1,15 @@
 ---
 base_model: kevin1010607/llama2-7b-mod
 tags:
+- alignment-handbook
+- trl
+- sft
+- generated_from_trainer
 - trl
 - sft
 - generated_from_trainer
 datasets:
-- generator
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: llama2-7b-mod-sft-full
   results: []
@@ -16,7 +20,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # llama2-7b-mod-sft-full
 
-This model is a fine-tuned version of [kevin1010607/llama2-7b-mod](https://huggingface.co/kevin1010607/llama2-7b-mod) on the generator dataset.
+This model is a fine-tuned version of [kevin1010607/llama2-7b-mod](https://huggingface.co/kevin1010607/llama2-7b-mod) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
 - Loss: 1.5823
 

all_results.json

@@ -1,5 +1,10 @@
 {
     "epoch": 0.9970559371933267,
+    "eval_loss": 1.5823049545288086,
+    "eval_runtime": 51.5775,
+    "eval_samples": 1000,
+    "eval_samples_per_second": 5.215,
+    "eval_steps_per_second": 0.659,
     "total_flos": 66244072439808.0,
     "train_loss": 2.2374156319250273,
     "train_runtime": 15313.1311,

config.json

@@ -24,6 +24,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "float32",
   "transformers_version": "4.41.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 0.9970559371933267,
+    "eval_loss": 1.5823049545288086,
+    "eval_runtime": 51.5775,
+    "eval_samples": 1000,
+    "eval_samples_per_second": 5.215,
+    "eval_steps_per_second": 0.659
+}

runs/Jun20_04-22-13_v1vh8qctr1718708466685-nzm8x/events.out.tfevents.1718844337.v1vh8qctr1718708466685-nzm8x.32192.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53035ea7875070d6a352d2bbb4c9595db017711785a2403982d355ea1d227cbe
+size 311