End of training

README.md

@@ -4,6 +4,7 @@ license: llama3.1
 base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: prm_gsm_2k_with_full_sol_mix_ref_hf
@@ -15,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # prm_gsm_2k_with_full_sol_mix_ref_hf
 
-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 prm_conversations_prm_gsm_2k_with_full_sol_full_gsm_2k_with_full_sol_mix_ref_hf dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0670
+- Loss: 0.0646
 
 ## Model description
 

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 0.9994008388256441,
+    "eval_loss": 0.06463462859392166,
+    "eval_runtime": 16.0963,
+    "eval_samples_per_second": 50.322,
+    "eval_steps_per_second": 6.337,
+    "total_flos": 161979572551680.0,
+    "train_loss": 0.08634093818035152,
+    "train_runtime": 5655.3143,
+    "train_samples_per_second": 14.165,
+    "train_steps_per_second": 0.221
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 0.9994008388256441,
+    "eval_loss": 0.06463462859392166,
+    "eval_runtime": 16.0963,
+    "eval_samples_per_second": 50.322,
+    "eval_steps_per_second": 6.337
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 0.9994008388256441,
+    "total_flos": 161979572551680.0,
+    "train_loss": 0.08634093818035152,
+    "train_runtime": 5655.3143,
+    "train_samples_per_second": 14.165,
+    "train_steps_per_second": 0.221
+}