End of training

README.md

@@ -2,11 +2,15 @@
 license: apache-2.0
 base_model: JackFram/llama-68m
 tags:
+- alignment-handbook
+- trl
+- sft
+- generated_from_trainer
 - trl
 - sft
 - generated_from_trainer
 datasets:
-- generator
+- HuggingFaceH4/ultrachat_200k
 model-index:
 - name: gpt2-sft-port
   results: []
@@ -17,7 +21,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # gpt2-sft-port
 
-This model is a fine-tuned version of [JackFram/llama-68m](https://huggingface.co/JackFram/llama-68m) on the generator dataset.
+This model is a fine-tuned version of [JackFram/llama-68m](https://huggingface.co/JackFram/llama-68m) on the HuggingFaceH4/ultrachat_200k dataset.
 It achieves the following results on the evaluation set:
 - Loss: 2.1059
 

all_results.json

@@ -1,5 +1,10 @@
 {
     "epoch": 2.0,
+    "eval_loss": 2.105895519256592,
+    "eval_runtime": 22.2286,
+    "eval_samples": 23110,
+    "eval_samples_per_second": 1438.643,
+    "eval_steps_per_second": 22.494,
     "train_loss": 2.1773926187554116,
     "train_runtime": 648.0178,
     "train_samples": 207865,

config.json

@@ -24,6 +24,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.38.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.0,
-    "eval_loss": 2.106666326522827,
-    "eval_runtime": 22.5513,
+    "eval_loss": 2.105895519256592,
+    "eval_runtime": 22.2286,
     "eval_samples": 23110,
-    "eval_samples_per_second": 1418.056,
-    "eval_steps_per_second": 22.172
+    "eval_samples_per_second": 1438.643,
+    "eval_steps_per_second": 22.494
 }

runs/Apr24_23-11-31_aga39/events.out.tfevents.1714019175.aga39.699328.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f59656302b1ad49448f5892c643bc118025e2586a737f3a44b46382b8be17101
+size 359