Training in progress, step 20, checkpoint

checkpoint-20/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-20/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:660dcc2d67be5d2fe32409c7ba0d9cfed6b5b24881c298d60fdb89c42d0360bc
+oid sha256:7ce68484125ee1278f3f43bb94a8524b956115d080cda995d65d640647c62d0e
 size 145287696

checkpoint-20/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e8afddd0826ff2559097c030b60db735489c6dc5c4fb2ff00a9672846f5117d3
+oid sha256:251575bbb0fd1f71ca80a29fd408fa685078a87ac45c941b659d5bffbd689bbe
 size 290833618

checkpoint-20/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eba4b952d7bc6954e0a0bfd3e78364767ebd2dd8ca7e8db7026915eb33bdb17c
+oid sha256:373a75e9575452ae8f576522d53b5e4e93abe743c3f5bfba2ce3227409bcb5f9
 size 1064

checkpoint-20/trainer_state.json

@@ -10,39 +10,39 @@
   "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
     }
   ],
   "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-20/training_args.bin

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