1.5

adapter_config.json

@@ -20,13 +20,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "o_proj",
     "up_proj",
     "gate_proj",
-    "k_proj",
-    "q_proj",
     "v_proj",
-    "down_proj"
+    "o_proj",
+    "down_proj",
+    "k_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 sha256:5f4df28160aa2cacaeceebeac4b0ff96d4e2e2078aa9e6967cff0bfbe22f87ba
+oid sha256:2cf73554208b66e451d3bdaa3cf09c4c0809a22711e973fb5037e53716148ac2
 size 1556140392

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 0.6956521739130435,
-    "eval_loss": 0.04214267060160637,
-    "eval_runtime": 3.45,
-    "eval_samples_per_second": 2.899,
-    "eval_steps_per_second": 0.29,
+    "eval_loss": 0.04213493689894676,
+    "eval_runtime": 3.3693,
+    "eval_samples_per_second": 2.968,
+    "eval_steps_per_second": 0.297,
     "total_flos": 3880733380509696.0,
-    "train_loss": 0.029286871664226055,
-    "train_runtime": 89.5279,
-    "train_samples_per_second": 1.005,
-    "train_steps_per_second": 0.022
+    "train_loss": 0.02928687445819378,
+    "train_runtime": 86.3619,
+    "train_samples_per_second": 1.042,
+    "train_steps_per_second": 0.023
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 0.6956521739130435,
-    "eval_loss": 0.04214267060160637,
-    "eval_runtime": 3.45,
-    "eval_samples_per_second": 2.899,
-    "eval_steps_per_second": 0.29
+    "eval_loss": 0.04213493689894676,
+    "eval_runtime": 3.3693,
+    "eval_samples_per_second": 2.968,
+    "eval_steps_per_second": 0.297
 }

experiment.config

@@ -0,0 +1,25 @@
+vision_config {
+  vision_api: TEXT_DETECTION
+  feature_element: WORD
+  word_confidence_threshold: -0.1
+  return_raw_response: true
+}
+preprocess_config {
+  label_overlap_threshold: 0.4
+  num_processes: 4
+  rotation_fixer: VISION_BASED_FAST
+}
+train_config {
+  num_epochs: 1
+}
+version: "3.0.2"
+field_extraction {
+  sub_exp_config {
+    model_config {
+      architecture {
+        nova {
+        }
+      }
+    }
+  }
+}

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 0.6956521739130435,
     "total_flos": 3880733380509696.0,
-    "train_loss": 0.029286871664226055,
-    "train_runtime": 89.5279,
-    "train_samples_per_second": 1.005,
-    "train_steps_per_second": 0.022
+    "train_loss": 0.02928687445819378,
+    "train_runtime": 86.3619,
+    "train_samples_per_second": 1.042,
+    "train_steps_per_second": 0.023
 }

trainer_log.jsonl

@@ -1,3 +1,3 @@
-{"current_steps": 1, "total_steps": 2, "loss": 0.0288, "lr": 0.0001, "epoch": 0.34782608695652173, "percentage": 50.0, "elapsed_time": "0:00:52", "remaining_time": "0:00:52"}
-{"current_steps": 2, "total_steps": 2, "loss": 0.0298, "lr": 0.0, "epoch": 0.6956521739130435, "percentage": 100.0, "elapsed_time": "0:01:22", "remaining_time": "0:00:00"}
-{"current_steps": 2, "total_steps": 2, "epoch": 0.6956521739130435, "percentage": 100.0, "elapsed_time": "0:01:28", "remaining_time": "0:00:00"}
+{"current_steps": 1, "total_steps": 2, "loss": 0.0288, "lr": 0.0001, "epoch": 0.34782608695652173, "percentage": 50.0, "elapsed_time": "0:00:39", "remaining_time": "0:00:39"}
+{"current_steps": 2, "total_steps": 2, "loss": 0.0298, "lr": 0.0, "epoch": 0.6956521739130435, "percentage": 100.0, "elapsed_time": "0:01:07", "remaining_time": "0:00:00"}
+{"current_steps": 2, "total_steps": 2, "epoch": 0.6956521739130435, "percentage": 100.0, "elapsed_time": "0:01:24", "remaining_time": "0:00:00"}

trainer_state.json

@@ -10,14 +10,14 @@
   "log_history": [
     {
       "epoch": 0.34782608695652173,
-      "grad_norm": 0.07414738088846207,
+      "grad_norm": 0.0741637647151947,
       "learning_rate": 0.0001,
       "loss": 0.0288,
       "step": 1
     },
     {
       "epoch": 0.6956521739130435,
-      "grad_norm": 0.06888309866189957,
+      "grad_norm": 0.06889505684375763,
       "learning_rate": 0.0,
       "loss": 0.0298,
       "step": 2
@@ -26,10 +26,10 @@
       "epoch": 0.6956521739130435,
       "step": 2,
       "total_flos": 3880733380509696.0,
-      "train_loss": 0.029286871664226055,
-      "train_runtime": 89.5279,
-      "train_samples_per_second": 1.005,
-      "train_steps_per_second": 0.022
+      "train_loss": 0.02928687445819378,
+      "train_runtime": 86.3619,
+      "train_samples_per_second": 1.042,
+      "train_steps_per_second": 0.023
     }
   ],
   "logging_steps": 1,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:68b07d6117d888dc51d19e06cb48a53cb20d5edb3faef27e7da7fca291d4a00d
+oid sha256:91f9d3edd57c685ba0165f354d6d438df2d7a056465c7e502c3ff099b11d297e
 size 5496