Training in progress, step 30, checkpoint

checkpoint-30/adapter_config.json

@@ -23,11 +23,11 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "k_proj",
     "v_proj",
-    "o_proj",
-    "gate_proj",
+    "k_proj",
     "q_proj",
+    "gate_proj",
+    "o_proj",
     "down_proj",
     "up_proj"
   ],

checkpoint-30/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:555ae9dbdd5399f736ccf0e7d4364fc478ce7228e0831c5a78b55cfa91611f17
+oid sha256:7ab952efab01e33f03d0f19d1ee5cfc15024e4555aee826756ebcbe6c278b898
 size 145287696

checkpoint-30/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eb50010124ba399674d41831ed4fc601f93b9757ec615e236f0526b40d43b2d0
+oid sha256:29228ccd3a6abf030b6d73bf0611ee584514c664d35fe090c4459870489a1df3
 size 290833618

checkpoint-30/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:efe70a1ac954116cafd503663506e71c6dc98e58b8f6de74290b12eb58debe44
+oid sha256:2a0978cabf0cce48bd72f8b2a49841b57fc9d030790f9386a5448e5e10126d59
 size 1064

checkpoint-30/trainer_state.json

@@ -10,54 +10,54 @@
   "log_history": [
     {
       "epoch": 0.021739130434782608,
-      "grad_norm": 2.138256549835205,
-      "learning_rate": 4.5e-05,
-      "loss": 2.3657,
+      "grad_norm": 2.2049312591552734,
+      "learning_rate": 4.963768115942029e-05,
+      "loss": 2.3459,
       "step": 10
     },
     {
       "epoch": 0.021739130434782608,
-      "eval_loss": 1.5849242210388184,
-      "eval_runtime": 9.6002,
-      "eval_samples_per_second": 49.999,
-      "eval_steps_per_second": 1.562,
+      "eval_loss": 1.5330805778503418,
+      "eval_runtime": 9.6608,
+      "eval_samples_per_second": 49.685,
+      "eval_steps_per_second": 1.553,
       "step": 10
     },
     {
       "epoch": 0.043478260869565216,
-      "grad_norm": 1.1203280687332153,
-      "learning_rate": 4e-05,
-      "loss": 1.2125,
+      "grad_norm": 1.1927831172943115,
+      "learning_rate": 4.9275362318840584e-05,
+      "loss": 1.1669,
       "step": 20
     },
     {
       "epoch": 0.043478260869565216,
-      "eval_loss": 1.04584538936615,
-      "eval_runtime": 9.677,
-      "eval_samples_per_second": 49.602,
-      "eval_steps_per_second": 1.55,
+      "eval_loss": 1.0271039009094238,
+      "eval_runtime": 9.7215,
+      "eval_samples_per_second": 49.375,
+      "eval_steps_per_second": 1.543,
       "step": 20
     },
     {
       "epoch": 0.06521739130434782,
-      "grad_norm": 1.064527988433838,
-      "learning_rate": 3.5e-05,
-      "loss": 0.867,
+      "grad_norm": 1.0509072542190552,
+      "learning_rate": 4.891304347826087e-05,
+      "loss": 0.8327,
       "step": 30
     },
     {
       "epoch": 0.06521739130434782,
-      "eval_loss": 0.9389117360115051,
-      "eval_runtime": 9.7352,
-      "eval_samples_per_second": 49.305,
-      "eval_steps_per_second": 1.541,
+      "eval_loss": 0.9039635062217712,
+      "eval_runtime": 9.7818,
+      "eval_samples_per_second": 49.071,
+      "eval_steps_per_second": 1.533,
       "step": 30
     }
   ],
   "logging_steps": 10,
-  "max_steps": 100,
+  "max_steps": 1380,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 1,
+  "num_train_epochs": 3,
   "save_steps": 10,
   "stateful_callbacks": {
     "TrainerControl": {

checkpoint-30/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:17a1a7c75c85fbba54f1e97b2cee73c167834890dd0df62950989daaf8975814
+oid sha256:1cdf2cbd96a7d8012e1f3da0569783099ba26a3ecbec680ff36cde09793d0889
 size 5816