Model save

README.md

@@ -1,20 +1,17 @@
 ---
 base_model: Qwen/Qwen3-1.7B-Base
-datasets: atomwalk12/linalg-debug-distilled
 library_name: transformers
 model_name: Qwen3-1.7B-Base-SFT
 tags:
 - generated_from_trainer
-- trl
-- tool-use
 - sft
-- linalg-zero
+- trl
 licence: license
 ---
 
 # Model Card for Qwen3-1.7B-Base-SFT
 
-This model is a fine-tuned version of [Qwen/Qwen3-1.7B-Base](https://huggingface.co/Qwen/Qwen3-1.7B-Base) on the [atomwalk12/linalg-debug-distilled](https://huggingface.co/datasets/atomwalk12/linalg-debug-distilled) dataset.
+This model is a fine-tuned version of [Qwen/Qwen3-1.7B-Base](https://huggingface.co/Qwen/Qwen3-1.7B-Base).
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start
@@ -30,7 +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/walks/huggingface/runs/gr54d3dr) 
+[<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/walks/huggingface/runs/l7yo1gpc) 
 
 
 This model was trained with SFT.

adapter_config.json

@@ -25,13 +25,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "o_proj",
-    "up_proj",
-    "down_proj",
-    "q_proj",
     "gate_proj",
+    "q_proj",
+    "down_proj",
+    "o_proj",
+    "k_proj",
     "v_proj",
-    "k_proj"
+    "up_proj"
   ],
   "task_type": "CAUSAL_LM",
   "trainable_token_indices": null,

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 1052153750937600.0,
-    "train_loss": 0.6244659423828125,
-    "train_runtime": 26.0114,
+    "train_loss": 0.0,
+    "train_runtime": 1.217,
     "train_samples": 5,
-    "train_samples_per_second": 0.961,
-    "train_steps_per_second": 0.192
+    "train_samples_per_second": 20.542,
+    "train_steps_per_second": 4.108
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 1052153750937600.0,
-    "train_loss": 0.6244659423828125,
-    "train_runtime": 26.0114,
+    "train_loss": 0.0,
+    "train_runtime": 1.217,
     "train_samples": 5,
-    "train_samples_per_second": 0.961,
-    "train_steps_per_second": 0.192
+    "train_samples_per_second": 20.542,
+    "train_steps_per_second": 4.108
 }

trainer_state.json

@@ -53,10 +53,10 @@
       "epoch": 5.0,
       "step": 5,
       "total_flos": 1052153750937600.0,
-      "train_loss": 0.6244659423828125,
-      "train_runtime": 26.0114,
-      "train_samples_per_second": 0.961,
-      "train_steps_per_second": 0.192
+      "train_loss": 0.0,
+      "train_runtime": 1.217,
+      "train_samples_per_second": 20.542,
+      "train_steps_per_second": 4.108
     }
   ],
   "logging_steps": 1,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c2f958acf6e02e9541571040742a17fb45f0fbd69b3baf647047a1c9a0c1ddfe
-size 6481
+oid sha256:8dc91a72cde04976580b8077db338a8c9562606254307974e93d689c396ab427
+size 6609