Training in progress, step 300, checkpoint

checkpoint-300/adapter_config.json

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

checkpoint-300/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:50a797c96415508c534ebc290f57b943d028592456914e03cc70efff3cd52806
+oid sha256:61b0866e2e6f595dbfd8a72fbcbef3195ad385f43fd8a061c17d7758d6b248cf
 size 161533192

checkpoint-300/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7fea01c0d3708e7b288c3ec9978a93993c15b820625499b4fee848d3cc36c390
+oid sha256:d978b99be05b487dcf194ba39f91921146945b4ee145551dff4f51d6850ec651
 size 323290986

checkpoint-300/trainer_state.json

@@ -10,260 +10,260 @@
   "log_history": [
     {
       "epoch": 0.010604453870625663,
-      "grad_norm": 2.0663416385650635,
+      "grad_norm": 2.276311159133911,
       "learning_rate": 4.982325910215624e-05,
-      "loss": 1.91,
+      "loss": 1.9038,
       "step": 10
     },
     {
       "epoch": 0.021208907741251327,
-      "grad_norm": 0.7459995746612549,
+      "grad_norm": 0.7678110003471375,
       "learning_rate": 4.964651820431248e-05,
-      "loss": 1.0094,
+      "loss": 0.997,
       "step": 20
     },
     {
       "epoch": 0.031813361611876985,
-      "grad_norm": 0.656954824924469,
+      "grad_norm": 0.6726565361022949,
       "learning_rate": 4.946977730646872e-05,
-      "loss": 0.8052,
+      "loss": 0.8021,
       "step": 30
     },
     {
       "epoch": 0.042417815482502653,
-      "grad_norm": 0.7273707389831543,
+      "grad_norm": 0.7318501472473145,
       "learning_rate": 4.929303640862496e-05,
-      "loss": 0.7145,
+      "loss": 0.7122,
       "step": 40
     },
     {
       "epoch": 0.053022269353128315,
-      "grad_norm": 0.6689164042472839,
+      "grad_norm": 0.6708294749259949,
       "learning_rate": 4.91162955107812e-05,
-      "loss": 0.6797,
+      "loss": 0.6783,
       "step": 50
     },
     {
       "epoch": 0.053022269353128315,
-      "eval_loss": 0.8142186403274536,
-      "eval_runtime": 11.794,
-      "eval_samples_per_second": 40.699,
-      "eval_steps_per_second": 2.544,
+      "eval_loss": 0.8075584173202515,
+      "eval_runtime": 11.7244,
+      "eval_samples_per_second": 40.94,
+      "eval_steps_per_second": 2.559,
       "step": 50
     },
     {
       "epoch": 0.06362672322375397,
-      "grad_norm": 0.6781293749809265,
+      "grad_norm": 0.7024317979812622,
       "learning_rate": 4.893955461293744e-05,
-      "loss": 0.6249,
+      "loss": 0.6244,
       "step": 60
     },
     {
       "epoch": 0.07423117709437964,
-      "grad_norm": 0.6776543259620667,
+      "grad_norm": 0.6959078907966614,
       "learning_rate": 4.876281371509368e-05,
-      "loss": 0.612,
+      "loss": 0.6117,
       "step": 70
     },
     {
       "epoch": 0.08483563096500531,
-      "grad_norm": 0.7603946924209595,
+      "grad_norm": 0.7824214696884155,
       "learning_rate": 4.858607281724991e-05,
-      "loss": 0.5998,
+      "loss": 0.5989,
       "step": 80
     },
     {
       "epoch": 0.09544008483563096,
-      "grad_norm": 0.7581377029418945,
+      "grad_norm": 0.7792642116546631,
       "learning_rate": 4.8409331919406155e-05,
-      "loss": 0.6032,
+      "loss": 0.6033,
       "step": 90
     },
     {
       "epoch": 0.10604453870625663,
-      "grad_norm": 0.7402133345603943,
+      "grad_norm": 0.7486705780029297,
       "learning_rate": 4.823259102156239e-05,
-      "loss": 0.5912,
+      "loss": 0.5904,
       "step": 100
     },
     {
       "epoch": 0.10604453870625663,
-      "eval_loss": 0.7294260859489441,
-      "eval_runtime": 11.8272,
-      "eval_samples_per_second": 40.585,
-      "eval_steps_per_second": 2.537,
+      "eval_loss": 0.7230511903762817,
+      "eval_runtime": 11.7589,
+      "eval_samples_per_second": 40.82,
+      "eval_steps_per_second": 2.551,
       "step": 100
     },
     {
       "epoch": 0.11664899257688228,
-      "grad_norm": 0.8224169015884399,
+      "grad_norm": 0.8214134573936462,
       "learning_rate": 4.8055850123718633e-05,
-      "loss": 0.578,
+      "loss": 0.5778,
       "step": 110
     },
     {
       "epoch": 0.12725344644750794,
-      "grad_norm": 0.7843881249427795,
+      "grad_norm": 0.7937195897102356,
       "learning_rate": 4.7879109225874866e-05,
-      "loss": 0.5751,
+      "loss": 0.5748,
       "step": 120
     },
     {
       "epoch": 0.1378579003181336,
-      "grad_norm": 0.772152304649353,
+      "grad_norm": 0.7781085968017578,
       "learning_rate": 4.770236832803111e-05,
-      "loss": 0.5471,
+      "loss": 0.5469,
       "step": 130
     },
     {
       "epoch": 0.14846235418875928,
-      "grad_norm": 0.6972452998161316,
+      "grad_norm": 0.695203423500061,
       "learning_rate": 4.7525627430187344e-05,
-      "loss": 0.5516,
+      "loss": 0.5511,
       "step": 140
     },
     {
       "epoch": 0.15906680805938495,
-      "grad_norm": 0.7806265354156494,
+      "grad_norm": 0.7824985384941101,
       "learning_rate": 4.734888653234359e-05,
-      "loss": 0.5263,
+      "loss": 0.5258,
       "step": 150
     },
     {
       "epoch": 0.15906680805938495,
-      "eval_loss": 0.6928939819335938,
-      "eval_runtime": 11.8093,
-      "eval_samples_per_second": 40.646,
-      "eval_steps_per_second": 2.54,
+      "eval_loss": 0.6867014765739441,
+      "eval_runtime": 11.7852,
+      "eval_samples_per_second": 40.729,
+      "eval_steps_per_second": 2.546,
       "step": 150
     },
     {
       "epoch": 0.16967126193001061,
-      "grad_norm": 0.7372023463249207,
+      "grad_norm": 0.7392640113830566,
       "learning_rate": 4.717214563449982e-05,
-      "loss": 0.5715,
+      "loss": 0.5714,
       "step": 160
     },
     {
       "epoch": 0.18027571580063625,
-      "grad_norm": 0.8236038684844971,
+      "grad_norm": 0.8316007852554321,
       "learning_rate": 4.699540473665606e-05,
-      "loss": 0.4987,
+      "loss": 0.4982,
       "step": 170
     },
     {
       "epoch": 0.19088016967126192,
-      "grad_norm": 0.8057310581207275,
+      "grad_norm": 0.8199877738952637,
       "learning_rate": 4.681866383881231e-05,
-      "loss": 0.5097,
+      "loss": 0.5092,
       "step": 180
     },
     {
       "epoch": 0.2014846235418876,
-      "grad_norm": 0.7124797105789185,
+      "grad_norm": 0.7089737057685852,
       "learning_rate": 4.664192294096854e-05,
-      "loss": 0.4952,
+      "loss": 0.4953,
       "step": 190
     },
     {
       "epoch": 0.21208907741251326,
-      "grad_norm": 0.7755710482597351,
+      "grad_norm": 0.7906930446624756,
       "learning_rate": 4.6465182043124786e-05,
-      "loss": 0.497,
+      "loss": 0.4974,
       "step": 200
     },
     {
       "epoch": 0.21208907741251326,
-      "eval_loss": 0.6748186349868774,
-      "eval_runtime": 11.8077,
-      "eval_samples_per_second": 40.651,
-      "eval_steps_per_second": 2.541,
+      "eval_loss": 0.669015645980835,
+      "eval_runtime": 11.8934,
+      "eval_samples_per_second": 40.358,
+      "eval_steps_per_second": 2.522,
       "step": 200
     },
     {
       "epoch": 0.22269353128313893,
-      "grad_norm": 0.912337064743042,
+      "grad_norm": 0.9085277318954468,
       "learning_rate": 4.628844114528102e-05,
-      "loss": 0.4997,
+      "loss": 0.4991,
       "step": 210
     },
     {
       "epoch": 0.23329798515376457,
-      "grad_norm": 0.771654486656189,
+      "grad_norm": 0.7775648832321167,
       "learning_rate": 4.6111700247437265e-05,
-      "loss": 0.5027,
+      "loss": 0.5024,
       "step": 220
     },
     {
       "epoch": 0.24390243902439024,
-      "grad_norm": 0.8241815567016602,
+      "grad_norm": 0.8324123620986938,
       "learning_rate": 4.59349593495935e-05,
-      "loss": 0.5196,
+      "loss": 0.5187,
       "step": 230
     },
     {
       "epoch": 0.2545068928950159,
-      "grad_norm": 0.7876004576683044,
+      "grad_norm": 0.8002191781997681,
       "learning_rate": 4.5758218451749736e-05,
-      "loss": 0.4791,
+      "loss": 0.4785,
       "step": 240
     },
     {
       "epoch": 0.2651113467656416,
-      "grad_norm": 0.7941911816596985,
+      "grad_norm": 0.8004704117774963,
       "learning_rate": 4.5581477553905975e-05,
-      "loss": 0.5078,
+      "loss": 0.5073,
       "step": 250
     },
     {
       "epoch": 0.2651113467656416,
-      "eval_loss": 0.653518795967102,
-      "eval_runtime": 11.7906,
-      "eval_samples_per_second": 40.71,
-      "eval_steps_per_second": 2.544,
+      "eval_loss": 0.6479213833808899,
+      "eval_runtime": 11.882,
+      "eval_samples_per_second": 40.397,
+      "eval_steps_per_second": 2.525,
       "step": 250
     },
     {
       "epoch": 0.2757158006362672,
-      "grad_norm": 0.8866105079650879,
+      "grad_norm": 0.893301248550415,
       "learning_rate": 4.5404736656062215e-05,
       "loss": 0.477,
       "step": 260
     },
     {
       "epoch": 0.2863202545068929,
-      "grad_norm": 0.8498968482017517,
+      "grad_norm": 0.8550834059715271,
       "learning_rate": 4.5227995758218454e-05,
-      "loss": 0.4926,
+      "loss": 0.4923,
       "step": 270
     },
     {
       "epoch": 0.29692470837751855,
-      "grad_norm": 0.8863232135772705,
+      "grad_norm": 0.8929152488708496,
       "learning_rate": 4.505125486037469e-05,
-      "loss": 0.463,
+      "loss": 0.4624,
       "step": 280
     },
     {
       "epoch": 0.3075291622481442,
-      "grad_norm": 0.9501034021377563,
+      "grad_norm": 0.9584974050521851,
       "learning_rate": 4.487451396253093e-05,
-      "loss": 0.4809,
+      "loss": 0.4801,
       "step": 290
     },
     {
       "epoch": 0.3181336161187699,
-      "grad_norm": 0.7933001518249512,
+      "grad_norm": 0.7924031615257263,
       "learning_rate": 4.469777306468717e-05,
-      "loss": 0.4941,
+      "loss": 0.4942,
       "step": 300
     },
     {
       "epoch": 0.3181336161187699,
-      "eval_loss": 0.6434358954429626,
-      "eval_runtime": 11.7969,
-      "eval_samples_per_second": 40.689,
-      "eval_steps_per_second": 2.543,
+      "eval_loss": 0.6371316313743591,
+      "eval_runtime": 11.8575,
+      "eval_samples_per_second": 40.481,
+      "eval_steps_per_second": 2.53,
       "step": 300
     }
   ],

checkpoint-300/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:872835407be2a0a95214d9cc3b7966660b3e25cbfda85f13ed9adad96c071793
+oid sha256:53a8c021c0523ab617f7d132f6e1d21f2075e97fdb7689e840960f38ef5cb15e
 size 5880