End of training

all_results.json

@@ -1,5 +1,10 @@
 {
     "epoch": 0.991869918699187,
+    "eval_loss": 2.0331625938415527,
+    "eval_runtime": 0.7425,
+    "eval_samples": 25,
+    "eval_samples_per_second": 32.323,
+    "eval_steps_per_second": 1.347,
     "total_flos": 25439628165120.0,
     "train_loss": 0.9141011159928119,
     "train_runtime": 665.6488,

config.json

@@ -23,6 +23,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.46.3",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32770
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 0.991869918699187,
+    "eval_loss": 2.0331625938415527,
+    "eval_runtime": 0.7425,
+    "eval_samples": 25,
+    "eval_samples_per_second": 32.323,
+    "eval_steps_per_second": 1.347
+}

runs/Nov21_22-07-37_main-fft-mistral-summarization-0-0/events.out.tfevents.1732246520.main-fft-mistral-summarization-0-0.547.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e5984684c8443aade10cc1b386e7d6832a9b2ee091256f85a2a19ec4941a2a3
+size 354