Training in progress, step 570

adapter_config.json

@@ -20,8 +20,8 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "v_proj"
+    "v_proj",
+    "q_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:010bf851e8d728cef779c2aa76071dc6d5dfba86fae2562e857b090b21a0401d
+oid sha256:acd6135ad9fc50bbaa654e28d65bf71c7b523772edbfbbacab4527d7944a4da9
 size 16794200

trainer_peft.log

@@ -343,3 +343,35 @@
 2024-05-30 20:16 - Setup PEFT
 2024-05-30 20:16 - Setup optimizer
 2024-05-30 20:16 - Resume from checkpoint - ./trainer_Wiki_lora/checkpoint-560
+2024-05-30 20:16 - Training complete!!!
+2024-05-30 20:30 - Cuda check
+2024-05-30 20:30 - True
+2024-05-30 20:30 - 1
+2024-05-30 20:30 - Configue Model and tokenizer
+2024-05-30 20:31 - Memory usage in   25.17 GB
+2024-05-30 20:31 - Dataset loaded successfully:
+ train-Jingmei/Pandemic_Books
+ test -Jingmei/Pandemic_WHO
+2024-05-30 20:33 - Tokenize data: DatasetDict({
+    train: Dataset({
+        features: ['input_ids', 'attention_mask'],
+        num_rows: 5966
+    })
+    test: Dataset({
+        features: ['input_ids', 'attention_mask'],
+        num_rows: 8264
+    })
+})
+2024-05-30 20:36 - Split data into chunks:DatasetDict({
+    train: Dataset({
+        features: ['input_ids', 'attention_mask'],
+        num_rows: 388208
+    })
+    test: Dataset({
+        features: ['input_ids', 'attention_mask'],
+        num_rows: 198964
+    })
+})
+2024-05-30 20:36 - Setup PEFT
+2024-05-30 20:36 - Setup optimizer
+2024-05-30 20:36 - Resume from checkpoint 

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:db34e7df3097f6e7ebe46bf2a510525d84a22126a709c1b9ebb02f3ecf43062a
+oid sha256:829885c7ba9fe77611a2e24a37ebfb3b536276478cf5304424ea54767009ac48
 size 5176