Training in progress, step 100

adapter_config.json

@@ -1,9 +1,6 @@
 {
   "alpha_pattern": {},
-  "auto_mapping": {
-    "base_model_class": "LlamaForCausalLM",
-    "parent_library": "transformers.models.llama.modeling_llama"
-  },
+  "auto_mapping": null,
   "base_model_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
   "bias": "none",
   "fan_in_fan_out": false,
@@ -13,20 +10,22 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 8,
-  "lora_dropout": 0.0,
+  "lora_alpha": 16,
+  "lora_dropout": 0.5,
   "megatron_config": null,
   "megatron_core": "megatron.core",
   "modules_to_save": null,
   "peft_type": "LORA",
-  "r": 8,
+  "r": 32,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
+    "k_proj",
     "q_proj",
-    "v_proj"
+    "v_proj",
+    "output_proj"
   ],
-  "task_type": null,
+  "task_type": "CAUSAL_LM",
   "use_dora": false,
   "use_rslora": false
 }

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1e0588b866622755aec3743db5f6eb506bfe7e6be787e7ab5d1daac728d4ff8a
-size 13648432
+oid sha256:336827a62a0dc73cc7952bff05f1e05f85e45b5129eb09b2b85651c3c98b8b88
+size 75523312

special_tokens_map.json

@@ -13,5 +13,11 @@
     "rstrip": false,
     "single_word": false
   },
-  "pad_token": "<|eot_id|>"
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
 }

tokenizer.json

@@ -2306,6 +2306,15 @@
       "rstrip": false,
       "normalized": false,
       "special": true
+    },
+    {
+      "id": 128256,
+      "content": "[PAD]",
+      "single_word": false,
+      "lstrip": false,
+      "rstrip": false,
+      "normalized": false,
+      "special": true
     }
   ],
   "normalizer": null,

tokenizer_config.json

@@ -2047,6 +2047,14 @@
       "rstrip": false,
       "single_word": false,
       "special": true
+    },
+    "128256": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
     }
   },
   "bos_token": "<|begin_of_text|>",
@@ -2058,6 +2066,6 @@
     "attention_mask"
   ],
   "model_max_length": 1000000000000000019884624838656,
-  "pad_token": "<|eot_id|>",
+  "pad_token": "[PAD]",
   "tokenizer_class": "PreTrainedTokenizerFast"
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9ad6a32a727fdffacb412b493742e44eaa82b985e18717690d687441678b2c64
+oid sha256:1b308bbd5bfd8bfdbc7e0a30f801399a0940ce3ec129b3a0ce7f0b05d6174654
 size 5112