Training in progress, step 18, checkpoint

checkpoint-18/adapter_config.json

@@ -23,13 +23,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "k_proj",
-    "o_proj",
+    "down_proj",
     "v_proj",
-    "gate_proj",
+    "o_proj",
+    "q_proj",
     "up_proj",
-    "down_proj",
-    "q_proj"
+    "gate_proj",
+    "k_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

checkpoint-18/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:47b9997a9d2d5911259bada14c664df2faaf6843ee17130ebe444f0edf2bf39a
+oid sha256:1445afaff0e357be58d2ca70cf70c25ed19a85f7eb2045867bb74c70b2e0934a
 size 131146352

checkpoint-18/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c2c9bd780229cf332f00720e836e44718e64e26cf0cf9a82d12528dd6039e591
+oid sha256:712482c32a63dcb0ddb2a13ebeaf33fabf26de905d201125398564c738954430
 size 262615650

checkpoint-18/trainer_state.json

@@ -10,176 +10,176 @@
   "log_history": [
     {
       "epoch": 0.0006200589055960316,
-      "grad_norm": 0.39027148485183716,
+      "grad_norm": 0.5875879526138306,
       "learning_rate": 4.000000000000001e-06,
-      "loss": 1.1401,
+      "loss": 1.179,
       "step": 1
     },
     {
       "epoch": 0.0012401178111920632,
-      "grad_norm": 0.1772935390472412,
+      "grad_norm": 0.2559925317764282,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 0.9133,
+      "loss": 0.9296,
       "step": 2
     },
     {
       "epoch": 0.0018601767167880949,
-      "grad_norm": 0.192855104804039,
+      "grad_norm": 0.27975356578826904,
       "learning_rate": 1.2e-05,
-      "loss": 0.8036,
+      "loss": 0.8212,
       "step": 3
     },
     {
       "epoch": 0.0018601767167880949,
-      "eval_loss": 0.993510365486145,
-      "eval_runtime": 46.5106,
-      "eval_samples_per_second": 2.15,
-      "eval_steps_per_second": 2.15,
+      "eval_loss": 1.0227885246276855,
+      "eval_runtime": 47.9985,
+      "eval_samples_per_second": 2.083,
+      "eval_steps_per_second": 2.083,
       "step": 3
     },
     {
       "epoch": 0.0024802356223841263,
-      "grad_norm": 0.23705194890499115,
+      "grad_norm": 0.4077853858470917,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 0.9389,
+      "loss": 0.9553,
       "step": 4
     },
     {
       "epoch": 0.0031002945279801583,
-      "grad_norm": 0.22702816128730774,
+      "grad_norm": 0.3481099307537079,
       "learning_rate": 2e-05,
-      "loss": 0.7836,
+      "loss": 0.8071,
       "step": 5
     },
     {
       "epoch": 0.0037203534335761897,
-      "grad_norm": 0.22573529183864594,
+      "grad_norm": 0.33538541197776794,
       "learning_rate": 1.9200000000000003e-05,
-      "loss": 0.8874,
+      "loss": 0.9046,
       "step": 6
     },
     {
       "epoch": 0.0037203534335761897,
-      "eval_loss": 0.9887735843658447,
-      "eval_runtime": 47.2769,
-      "eval_samples_per_second": 2.115,
-      "eval_steps_per_second": 2.115,
+      "eval_loss": 1.0122885704040527,
+      "eval_runtime": 48.1202,
+      "eval_samples_per_second": 2.078,
+      "eval_steps_per_second": 2.078,
       "step": 6
     },
     {
       "epoch": 0.004340412339172222,
-      "grad_norm": 0.231398344039917,
+      "grad_norm": 0.2848486602306366,
       "learning_rate": 1.8400000000000003e-05,
-      "loss": 0.8755,
+      "loss": 0.8963,
       "step": 7
     },
     {
       "epoch": 0.004960471244768253,
-      "grad_norm": 0.28229379653930664,
+      "grad_norm": 0.29836800694465637,
       "learning_rate": 1.76e-05,
-      "loss": 0.971,
+      "loss": 0.9869,
       "step": 8
     },
     {
       "epoch": 0.005580530150364285,
-      "grad_norm": 0.21991951763629913,
+      "grad_norm": 0.22438423335552216,
       "learning_rate": 1.6800000000000002e-05,
-      "loss": 0.8546,
+      "loss": 0.8704,
       "step": 9
     },
     {
       "epoch": 0.005580530150364285,
-      "eval_loss": 0.9774181842803955,
-      "eval_runtime": 47.3982,
-      "eval_samples_per_second": 2.11,
-      "eval_steps_per_second": 2.11,
+      "eval_loss": 0.996995210647583,
+      "eval_runtime": 48.1511,
+      "eval_samples_per_second": 2.077,
+      "eval_steps_per_second": 2.077,
       "step": 9
     },
     {
       "epoch": 0.0062005890559603165,
-      "grad_norm": 0.26028916239738464,
+      "grad_norm": 0.26462429761886597,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 1.0344,
+      "loss": 1.0479,
       "step": 10
     },
     {
       "epoch": 0.0068206479615563476,
-      "grad_norm": 0.23867206275463104,
+      "grad_norm": 0.2650766968727112,
       "learning_rate": 1.5200000000000002e-05,
-      "loss": 0.8132,
+      "loss": 0.8352,
       "step": 11
     },
     {
       "epoch": 0.0074407068671523795,
-      "grad_norm": 0.30641764402389526,
+      "grad_norm": 0.3595620095729828,
       "learning_rate": 1.4400000000000001e-05,
-      "loss": 1.1168,
+      "loss": 1.1297,
       "step": 12
     },
     {
       "epoch": 0.0074407068671523795,
-      "eval_loss": 0.963714599609375,
-      "eval_runtime": 47.188,
-      "eval_samples_per_second": 2.119,
-      "eval_steps_per_second": 2.119,
+      "eval_loss": 0.9821679592132568,
+      "eval_runtime": 48.205,
+      "eval_samples_per_second": 2.074,
+      "eval_steps_per_second": 2.074,
       "step": 12
     },
     {
       "epoch": 0.008060765772748411,
-      "grad_norm": 0.25963348150253296,
+      "grad_norm": 0.2854643166065216,
       "learning_rate": 1.3600000000000002e-05,
-      "loss": 0.9735,
+      "loss": 0.9925,
       "step": 13
     },
     {
       "epoch": 0.008680824678344443,
-      "grad_norm": 0.2063651829957962,
+      "grad_norm": 0.21441830694675446,
       "learning_rate": 1.2800000000000001e-05,
-      "loss": 0.7957,
+      "loss": 0.8125,
       "step": 14
     },
     {
       "epoch": 0.009300883583940475,
-      "grad_norm": 0.2287115454673767,
+      "grad_norm": 0.21920695900917053,
       "learning_rate": 1.2e-05,
-      "loss": 0.8343,
+      "loss": 0.8518,
       "step": 15
     },
     {
       "epoch": 0.009300883583940475,
-      "eval_loss": 0.9510444402694702,
-      "eval_runtime": 47.3972,
-      "eval_samples_per_second": 2.11,
-      "eval_steps_per_second": 2.11,
+      "eval_loss": 0.9693244695663452,
+      "eval_runtime": 47.8177,
+      "eval_samples_per_second": 2.091,
+      "eval_steps_per_second": 2.091,
       "step": 15
     },
     {
       "epoch": 0.009920942489536505,
-      "grad_norm": 0.2136881947517395,
+      "grad_norm": 0.20094920694828033,
       "learning_rate": 1.1200000000000001e-05,
-      "loss": 0.7861,
+      "loss": 0.8003,
       "step": 16
     },
     {
       "epoch": 0.010541001395132537,
-      "grad_norm": 0.16134004294872284,
+      "grad_norm": 0.21595342457294464,
       "learning_rate": 1.04e-05,
-      "loss": 0.8375,
+      "loss": 0.8492,
       "step": 17
     },
     {
       "epoch": 0.01116106030072857,
-      "grad_norm": 0.20020240545272827,
+      "grad_norm": 0.20084670186042786,
       "learning_rate": 9.600000000000001e-06,
-      "loss": 0.8086,
+      "loss": 0.8238,
       "step": 18
     },
     {
       "epoch": 0.01116106030072857,
-      "eval_loss": 0.9415214657783508,
-      "eval_runtime": 47.2291,
-      "eval_samples_per_second": 2.117,
-      "eval_steps_per_second": 2.117,
+      "eval_loss": 0.9594716429710388,
+      "eval_runtime": 47.8775,
+      "eval_samples_per_second": 2.089,
+      "eval_steps_per_second": 2.089,
       "step": 18
     }
   ],
@@ -200,7 +200,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.854372503062528e+16,
+  "total_flos": 3.910655369324544e+16,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null

checkpoint-18/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:31008ae8f5fa3525c3b5855165b7a47a0cdb47e438b843022563680f47a27521
+oid sha256:b7726c9626d0a377a638b6674780a9e8d32aed5a1fc017243f1823b069fc15ac
 size 5688