End of training

README.md

@@ -3,9 +3,15 @@ library_name: transformers
 license: llama3.1
 base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
 tags:
+- alignment-handbook
 - trl
 - sft
 - generated_from_trainer
+- trl
+- sft
+- generated_from_trainer
+datasets:
+- barc0/transduction_20k_gpt4o-mini_generated_problems_seed100.jsonl_messages_format_0.3
 model-index:
 - name: barc-llama3.1-8b-instruct-fft-transduction-gpt4omini20k_lr1e-5_epoch2
   results: []
@@ -16,7 +22,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # barc-llama3.1-8b-instruct-fft-transduction-gpt4omini20k_lr1e-5_epoch2
 
-This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct) on the barc0/transduction_20k_gpt4o-mini_generated_problems_seed100.jsonl_messages_format_0.3 dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.0623
 

all_results.json

@@ -1,5 +1,10 @@
 {
     "epoch": 1.993127147766323,
+    "eval_loss": 0.06226445734500885,
+    "eval_runtime": 107.317,
+    "eval_samples": 979,
+    "eval_samples_per_second": 9.123,
+    "eval_steps_per_second": 0.289,
     "total_flos": 8.781509506919629e+16,
     "train_loss": 0.08342721785964637,
     "train_runtime": 7557.4748,

config.json

@@ -35,6 +35,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.45.0.dev0",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 128256
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 1.993127147766323,
+    "eval_loss": 0.06226445734500885,
+    "eval_runtime": 107.317,
+    "eval_samples": 979,
+    "eval_samples_per_second": 9.123,
+    "eval_steps_per_second": 0.289
+}

runs/Sep23_06-23-32_ellis-compute-02.cs.cornell.edu/events.out.tfevents.1727094889.ellis-compute-02.cs.cornell.edu.2485545.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33bbd2ce248af6a12e0f7d1f8e844c99a43df34e19c27e5f840e1e9b348c4f72
+size 359