Model save

README.md

@@ -1,4 +1,5 @@
 ---
+base_model: Qwen/Qwen3-1.7B
 library_name: transformers
 model_name: PreThink_MemAgent
 tags:
@@ -10,7 +11,7 @@ licence: license
 
 # Model Card for PreThink_MemAgent
 
-This model is a fine-tuned version of [None](https://huggingface.co/None).
+This model is a fine-tuned version of [Qwen/Qwen3-1.7B](https://huggingface.co/Qwen/Qwen3-1.7B).
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start
@@ -26,7 +27,7 @@ print(output["generated_text"])
 
 ## Training procedure
 
-[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/mingzeli/PreThink_MemAgent/runs/z5tryx9w) 
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/mingzeli/PreThink_MemAgent/runs/9ek6f7el) 
 
 
 This model was trained with SFT.

all_results.json

@@ -1,8 +1,8 @@
 {
-    "total_flos": 33029095800832.0,
-    "train_loss": 1.1323178586634723,
-    "train_runtime": 824.9985,
-    "train_samples": 1386,
-    "train_samples_per_second": 3.36,
-    "train_steps_per_second": 0.107
+    "total_flos": 179772395159552.0,
+    "train_loss": 1.3804568826581873,
+    "train_runtime": 1345.8555,
+    "train_samples": 26745,
+    "train_samples_per_second": 19.872,
+    "train_steps_per_second": 0.311
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "total_flos": 33029095800832.0,
-    "train_loss": 1.1323178586634723,
-    "train_runtime": 824.9985,
-    "train_samples": 1386,
-    "train_samples_per_second": 3.36,
-    "train_steps_per_second": 0.107
+    "total_flos": 179772395159552.0,
+    "train_loss": 1.3804568826581873,
+    "train_runtime": 1345.8555,
+    "train_samples": 26745,
+    "train_samples_per_second": 19.872,
+    "train_steps_per_second": 0.311
 }