Initial LoRA finetuned model and tokenizer for CoT

adapter_config.json

@@ -25,12 +25,12 @@
   "revision": null,
   "target_modules": [
     "down_proj",
-    "q_proj",
-    "up_proj",
-    "gate_proj",
     "k_proj",
+    "v_proj",
     "o_proj",
-    "v_proj"
+    "gate_proj",
+    "up_proj",
+    "q_proj"
   ],
   "task_type": "CAUSAL_LM",
   "trainable_token_indices": null,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fb833a72136ccf95c87319ee03832ebb09b86aa58ea30bc5d859ea59ccc73b6c
-size 1279323952
+oid sha256:b285968d22cd626569dfe97dea93f0cb395118921d3f63f125ede9ba83b046d9
+size 1986736152

added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "[PAD]": 43176
+}

checkpoint-7/adapter_config.json

@@ -25,12 +25,12 @@
   "revision": null,
   "target_modules": [
     "down_proj",
-    "q_proj",
-    "up_proj",
-    "gate_proj",
     "k_proj",
+    "v_proj",
     "o_proj",
-    "v_proj"
+    "gate_proj",
+    "up_proj",
+    "q_proj"
   ],
   "task_type": "CAUSAL_LM",
   "trainable_token_indices": null,

checkpoint-7/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fb833a72136ccf95c87319ee03832ebb09b86aa58ea30bc5d859ea59ccc73b6c
-size 1279323952
+oid sha256:b285968d22cd626569dfe97dea93f0cb395118921d3f63f125ede9ba83b046d9
+size 1986736152

checkpoint-7/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "[PAD]": 43176
+}

checkpoint-7/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7dbf62bc32d3c717ff792dabcee496dc79ce5933d1bfa09ab95bbd2806fbd42c
+oid sha256:a0161e9e9e182ee383d1226d219fb51eee7a0a7dee2d3cab2c3ef156e602996f
 size 650163476

checkpoint-7/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:08313fdbb29e3c0ee6b29e5901d0444354bcec03995b1fc6b3b45f2de0ebf332
+oid sha256:63c72f5771f7347224a4a09ed2354c66aedcc180caa4df8fc886e7d8acb9a307
 size 14244

checkpoint-7/special_tokens_map.json

@@ -13,7 +13,13 @@
     "rstrip": false,
     "single_word": false
   },
-  "pad_token": "</s>",
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
   "unk_token": {
     "content": "<unk>",
     "lstrip": false,

checkpoint-7/tokenizer.json

@@ -29,6 +29,15 @@
       "rstrip": false,
       "normalized": true,
       "special": true
+    },
+    {
+      "id": 43176,
+      "content": "[PAD]",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
     }
   ],
   "normalizer": null,

checkpoint-7/tokenizer_config.json

@@ -26,6 +26,14 @@
       "rstrip": false,
       "single_word": false,
       "special": true
+    },
+    "43176": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
     }
   },
   "bos_token": "<s>",
@@ -34,7 +42,7 @@
   "extra_special_tokens": {},
   "legacy": false,
   "model_max_length": 1000000000000000019884624838656,
-  "pad_token": "</s>",
+  "pad_token": "[PAD]",
   "padding_side": "right",
   "sp_model_kwargs": {},
   "spaces_between_special_tokens": false,

checkpoint-7/trainer_state.json

@@ -26,7 +26,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3970389867724800.0,
+  "total_flos": 4284174512947200.0,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null

checkpoint-7/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4a3f1b03b0f43731684eedb4e53eaa7608fa66b06438764d18682006c380ccb2
+oid sha256:35872fcc6d491acc7effa1c32a80c35245f0508a790d9d56ac5693a9faca89eb
 size 5368

special_tokens_map.json

@@ -13,7 +13,13 @@
     "rstrip": false,
     "single_word": false
   },
-  "pad_token": "</s>",
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
   "unk_token": {
     "content": "<unk>",
     "lstrip": false,

tokenizer.json

@@ -29,6 +29,15 @@
       "rstrip": false,
       "normalized": true,
       "special": true
+    },
+    {
+      "id": 43176,
+      "content": "[PAD]",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
     }
   ],
   "normalizer": null,

tokenizer_config.json

@@ -26,6 +26,14 @@
       "rstrip": false,
       "single_word": false,
       "special": true
+    },
+    "43176": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
     }
   },
   "bos_token": "<s>",
@@ -34,7 +42,7 @@
   "extra_special_tokens": {},
   "legacy": false,
   "model_max_length": 1000000000000000019884624838656,
-  "pad_token": "</s>",
+  "pad_token": "[PAD]",
   "padding_side": "right",
   "sp_model_kwargs": {},
   "spaces_between_special_tokens": false,