End of training

README.md

@@ -4,6 +4,7 @@ license: llama3.2
 base_model: secmlr/DS-Noisy-N_DS-Clean-N_DS-OSS-N_QWQ-OSS-N_QWQ-Clean-N_QWQ-Noisy-N_Llama-3.2-3B-Instruct_sft
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: Llama-3.2-3B-final_model
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Llama-3.2-3B-final_model
 
-This model is a fine-tuned version of [secmlr/DS-Noisy-N_DS-Clean-N_DS-OSS-N_QWQ-OSS-N_QWQ-Clean-N_QWQ-Noisy-N_Llama-3.2-3B-Instruct_sft](https://huggingface.co/secmlr/DS-Noisy-N_DS-Clean-N_DS-OSS-N_QWQ-OSS-N_QWQ-Clean-N_QWQ-Noisy-N_Llama-3.2-3B-Instruct_sft) on the None dataset.
+This model is a fine-tuned version of [secmlr/DS-Noisy-N_DS-Clean-N_DS-OSS-N_QWQ-OSS-N_QWQ-Clean-N_QWQ-Noisy-N_Llama-3.2-3B-Instruct_sft](https://huggingface.co/secmlr/DS-Noisy-N_DS-Clean-N_DS-OSS-N_QWQ-OSS-N_QWQ-Clean-N_QWQ-Noisy-N_Llama-3.2-3B-Instruct_sft) on the simplier_reasoning_ds_clean_32k, the simplier_reasoning_ds_noisy_32k, the simplier_reasoning_ds_ossfuzz_32k, the simplier_reasoning_qwq_clean_32k, the simplier_reasoning_qwq_noisy_small_32k and the simplier_reasoning_qwq_ossfuzz_32k datasets.
 
 ## Model description
 

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.998837479655894,
+    "total_flos": 88194404745216.0,
+    "train_loss": 0.6016569009234564,
+    "train_runtime": 11197.5198,
+    "train_samples_per_second": 4.609,
+    "train_steps_per_second": 0.192
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.998837479655894,
+    "total_flos": 88194404745216.0,
+    "train_loss": 0.6016569009234564,
+    "train_runtime": 11197.5198,
+    "train_samples_per_second": 4.609,
+    "train_steps_per_second": 0.192
+}