starlineventures/pilot-talk

adapter_config.json

@@ -23,12 +23,12 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "v_proj",
-    "k_proj",
-    "gate_proj",
     "down_proj",
+    "k_proj",
     "up_proj",
+    "v_proj",
+    "q_proj",
+    "gate_proj",
     "o_proj"
   ],
   "task_type": null,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9cbfaa87820050cb9b4a70ab1715a8fb6c4b0d72c5d031b6f8ee9ae548a0ae95
+oid sha256:9f3a1cea7d969cef21da106d3e9c7d41b5283af202e93bf96911eebc61146215
 size 94422368

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.75,
+    "epoch": 2.125,
     "total_flos": 0.0,
-    "train_loss": 0.2669122704437801,
-    "train_runtime": 143.3284,
-    "train_samples_per_second": 16.745,
-    "train_steps_per_second": 1.674
+    "train_loss": 0.2409752929911894,
+    "train_runtime": 174.6588,
+    "train_samples_per_second": 13.741,
+    "train_steps_per_second": 1.374
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.75,
+    "epoch": 2.125,
     "total_flos": 0.0,
-    "train_loss": 0.2669122704437801,
-    "train_runtime": 143.3284,
-    "train_samples_per_second": 16.745,
-    "train_steps_per_second": 1.674
+    "train_loss": 0.2409752929911894,
+    "train_runtime": 174.6588,
+    "train_samples_per_second": 13.741,
+    "train_steps_per_second": 1.374
 }

trainer_state.json

@@ -1,119 +1,140 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.75,
+  "epoch": 2.125,
   "eval_steps": 500,
-  "global_step": 140,
+  "global_step": 170,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.125,
-      "grad_norm": 26.014522552490234,
+      "grad_norm": 35.11359405517578,
       "learning_rate": 9.583333333333334e-05,
-      "loss": 1.2951,
+      "loss": 1.3709,
       "step": 10
     },
     {
       "epoch": 0.25,
-      "grad_norm": 38.19829177856445,
+      "grad_norm": 28.059900283813477,
       "learning_rate": 9.166666666666667e-05,
-      "loss": 0.4151,
+      "loss": 0.4762,
       "step": 20
     },
     {
       "epoch": 0.375,
-      "grad_norm": 35.403507232666016,
+      "grad_norm": 30.004230499267578,
       "learning_rate": 8.75e-05,
-      "loss": 0.2558,
+      "loss": 0.2966,
       "step": 30
     },
     {
       "epoch": 0.5,
-      "grad_norm": 37.772884368896484,
+      "grad_norm": 30.531251907348633,
       "learning_rate": 8.333333333333334e-05,
-      "loss": 0.2447,
+      "loss": 0.2166,
       "step": 40
     },
     {
       "epoch": 0.625,
-      "grad_norm": 36.102352142333984,
+      "grad_norm": 25.447418212890625,
       "learning_rate": 7.916666666666666e-05,
-      "loss": 0.2089,
+      "loss": 0.1816,
       "step": 50
     },
     {
       "epoch": 0.75,
-      "grad_norm": 37.57475662231445,
+      "grad_norm": 25.023717880249023,
       "learning_rate": 7.500000000000001e-05,
-      "loss": 0.1759,
+      "loss": 0.1653,
       "step": 60
     },
     {
       "epoch": 0.875,
-      "grad_norm": 47.88151550292969,
+      "grad_norm": 23.13313102722168,
       "learning_rate": 7.083333333333334e-05,
-      "loss": 0.1593,
+      "loss": 0.1506,
       "step": 70
     },
     {
       "epoch": 1.0,
-      "grad_norm": 51.923824310302734,
+      "grad_norm": 21.35039710998535,
       "learning_rate": 6.666666666666667e-05,
-      "loss": 0.1495,
+      "loss": 0.1369,
       "step": 80
     },
     {
       "epoch": 1.125,
-      "grad_norm": 52.507205963134766,
+      "grad_norm": 19.913026809692383,
       "learning_rate": 6.25e-05,
-      "loss": 0.1442,
+      "loss": 0.1315,
       "step": 90
     },
     {
       "epoch": 1.25,
-      "grad_norm": 51.443668365478516,
+      "grad_norm": 21.789073944091797,
       "learning_rate": 5.833333333333334e-05,
-      "loss": 0.1408,
+      "loss": 0.1296,
       "step": 100
     },
     {
       "epoch": 1.375,
-      "grad_norm": 46.47669982910156,
+      "grad_norm": 25.12034797668457,
       "learning_rate": 5.4166666666666664e-05,
-      "loss": 0.1379,
+      "loss": 0.1279,
       "step": 110
     },
     {
       "epoch": 1.5,
-      "grad_norm": 42.74077606201172,
+      "grad_norm": 25.35553550720215,
       "learning_rate": 5e-05,
-      "loss": 0.1379,
+      "loss": 0.1241,
       "step": 120
     },
     {
       "epoch": 1.625,
-      "grad_norm": 40.221805572509766,
+      "grad_norm": 23.74753761291504,
       "learning_rate": 4.5833333333333334e-05,
-      "loss": 0.1371,
+      "loss": 0.1208,
       "step": 130
     },
     {
       "epoch": 1.75,
-      "grad_norm": 34.954124450683594,
+      "grad_norm": 23.058244705200195,
       "learning_rate": 4.166666666666667e-05,
-      "loss": 0.1346,
+      "loss": 0.1188,
       "step": 140
     },
     {
-      "epoch": 1.75,
-      "step": 140,
+      "epoch": 1.875,
+      "grad_norm": 22.608642578125,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.1174,
+      "step": 150
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 22.311994552612305,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.1162,
+      "step": 160
+    },
+    {
+      "epoch": 2.125,
+      "grad_norm": 23.028535842895508,
+      "learning_rate": 2.916666666666667e-05,
+      "loss": 0.1158,
+      "step": 170
+    },
+    {
+      "epoch": 2.125,
+      "step": 170,
       "total_flos": 0.0,
-      "train_loss": 0.2669122704437801,
-      "train_runtime": 143.3284,
-      "train_samples_per_second": 16.745,
-      "train_steps_per_second": 1.674
+      "train_loss": 0.2409752929911894,
+      "train_runtime": 174.6588,
+      "train_samples_per_second": 13.741,
+      "train_steps_per_second": 1.374
     }
   ],
   "logging_steps": 10,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e3328c58653121484b735b79386a21a53daeb13b417b5ef0defe329a7fa35aec
+oid sha256:0bf77fa574b65faae23151f91dc99b2b3a380810ab2b3e9268e86e0152036049
 size 5432