Training in progress, step 28389, checkpoint

last-checkpoint/adapter_config.json

@@ -29,13 +29,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "q_proj",
-    "gate_proj",
     "up_proj",
-    "down_proj",
     "o_proj",
-    "k_proj"
+    "v_proj",
+    "down_proj",
+    "k_proj",
+    "q_proj",
+    "gate_proj"
   ],
   "target_parameters": null,
   "task_type": "CAUSAL_LM",

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:70f9b9fb741bedf906e82837997d6a0d8073aa26cd256f399f58548c10bfc88a
+oid sha256:10298979fc3e9c60b02cbc6faa90dad5b0e3f00fe3c62150bb8266a0abb8bf7d
 size 295488936

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d4c4b5d23d3b5265128e1809ad25d7a02bf6379a5d749ff61413f7a1b10fa010
+oid sha256:7533b9ed6ccfd5d3aae318e2b9e1f4b78f22e9cbbe0670de3ba41b66ccfbd8f5
 size 591209083

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ad72eb133cba517188ca2904582f9864218d10e150f5145ec013fe46f8e0e271
+oid sha256:e271799a9390b95549db5f989a0b1f5db31c050fe1eed3dcacd98c670ca2acea
 size 1383

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e6f4fc749262161a1078bb07093eccfbd5c45a8b5e8b40352b9175ec6bbac0c2
+oid sha256:0bcce9d801442e885424f2ae9fd67dda51ccf797bf351cdcc043eed717a0b682
 size 1465

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.9589812013686143,
+  "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 28000,
+  "global_step": 28389,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -20056,6 +20056,272 @@
       "eval_samples_per_second": 2.75,
       "eval_steps_per_second": 2.75,
       "step": 28000
+    },
+    {
+      "epoch": 2.9600380464219187,
+      "grad_norm": 1.5094038248062134,
+      "learning_rate": 2.2132211273220847e-08,
+      "loss": 0.8209236145019532,
+      "step": 28010
+    },
+    {
+      "epoch": 2.9610948914752235,
+      "grad_norm": 1.6858611106872559,
+      "learning_rate": 2.0982845894376713e-08,
+      "loss": 0.8495326995849609,
+      "step": 28020
+    },
+    {
+      "epoch": 2.9621517365285284,
+      "grad_norm": 1.41942298412323,
+      "learning_rate": 1.986411333650129e-08,
+      "loss": 0.8880720138549805,
+      "step": 28030
+    },
+    {
+      "epoch": 2.9632085815818328,
+      "grad_norm": 1.4312872886657715,
+      "learning_rate": 1.8776014971538803e-08,
+      "loss": 0.9381152153015136,
+      "step": 28040
+    },
+    {
+      "epoch": 2.9642654266351376,
+      "grad_norm": 1.4956817626953125,
+      "learning_rate": 1.771855213386908e-08,
+      "loss": 0.8022502899169922,
+      "step": 28050
+    },
+    {
+      "epoch": 2.965322271688442,
+      "grad_norm": 1.7753782272338867,
+      "learning_rate": 1.669172612030756e-08,
+      "loss": 0.7477090358734131,
+      "step": 28060
+    },
+    {
+      "epoch": 2.966379116741747,
+      "grad_norm": 1.0993744134902954,
+      "learning_rate": 1.5695538190091397e-08,
+      "loss": 0.8584689140319824,
+      "step": 28070
+    },
+    {
+      "epoch": 2.967435961795051,
+      "grad_norm": 2.141275644302368,
+      "learning_rate": 1.4729989564887803e-08,
+      "loss": 0.8031422615051269,
+      "step": 28080
+    },
+    {
+      "epoch": 2.968492806848356,
+      "grad_norm": 1.3441087007522583,
+      "learning_rate": 1.3795081428788493e-08,
+      "loss": 0.8645223617553711,
+      "step": 28090
+    },
+    {
+      "epoch": 2.9695496519016604,
+      "grad_norm": 1.0135029554367065,
+      "learning_rate": 1.2890814928315231e-08,
+      "loss": 0.8085169792175293,
+      "step": 28100
+    },
+    {
+      "epoch": 2.9706064969549653,
+      "grad_norm": 1.4154417514801025,
+      "learning_rate": 1.2017191172405962e-08,
+      "loss": 0.8781013488769531,
+      "step": 28110
+    },
+    {
+      "epoch": 2.9716633420082696,
+      "grad_norm": 1.3878732919692993,
+      "learning_rate": 1.1174211232417575e-08,
+      "loss": 0.8837569236755372,
+      "step": 28120
+    },
+    {
+      "epoch": 2.9727201870615745,
+      "grad_norm": 1.2050855159759521,
+      "learning_rate": 1.0361876142139792e-08,
+      "loss": 0.8309374809265136,
+      "step": 28130
+    },
+    {
+      "epoch": 2.973777032114879,
+      "grad_norm": 0.9934705495834351,
+      "learning_rate": 9.580186897767407e-09,
+      "loss": 0.8521280288696289,
+      "step": 28140
+    },
+    {
+      "epoch": 2.9748338771681837,
+      "grad_norm": 1.339559555053711,
+      "learning_rate": 8.829144457922489e-09,
+      "loss": 0.7942620754241944,
+      "step": 28150
+    },
+    {
+      "epoch": 2.9758907222214885,
+      "grad_norm": 1.5272369384765625,
+      "learning_rate": 8.108749743637734e-09,
+      "loss": 0.8327934265136718,
+      "step": 28160
+    },
+    {
+      "epoch": 2.976947567274793,
+      "grad_norm": 1.5825679302215576,
+      "learning_rate": 7.419003638364785e-09,
+      "loss": 0.8163703918457031,
+      "step": 28170
+    },
+    {
+      "epoch": 2.9780044123280973,
+      "grad_norm": 1.743831992149353,
+      "learning_rate": 6.759906987968689e-09,
+      "loss": 0.8478512763977051,
+      "step": 28180
+    },
+    {
+      "epoch": 2.979061257381402,
+      "grad_norm": 1.4517738819122314,
+      "learning_rate": 6.131460600725114e-09,
+      "loss": 0.8876804351806641,
+      "step": 28190
+    },
+    {
+      "epoch": 2.980118102434707,
+      "grad_norm": 1.115539789199829,
+      "learning_rate": 5.533665247325903e-09,
+      "loss": 0.9241631507873536,
+      "step": 28200
+    },
+    {
+      "epoch": 2.9811749474880114,
+      "grad_norm": 1.241997480392456,
+      "learning_rate": 4.9665216608735245e-09,
+      "loss": 0.8331340789794922,
+      "step": 28210
+    },
+    {
+      "epoch": 2.9822317925413158,
+      "grad_norm": 1.9293731451034546,
+      "learning_rate": 4.430030536878294e-09,
+      "loss": 0.7604406833648681,
+      "step": 28220
+    },
+    {
+      "epoch": 2.9832886375946206,
+      "grad_norm": 1.9037610292434692,
+      "learning_rate": 3.924192533263926e-09,
+      "loss": 0.755042839050293,
+      "step": 28230
+    },
+    {
+      "epoch": 2.9843454826479254,
+      "grad_norm": 1.1857556104660034,
+      "learning_rate": 3.449008270359211e-09,
+      "loss": 0.8992856979370117,
+      "step": 28240
+    },
+    {
+      "epoch": 2.98540232770123,
+      "grad_norm": 1.5789319276809692,
+      "learning_rate": 3.004478330900784e-09,
+      "loss": 0.9083795547485352,
+      "step": 28250
+    },
+    {
+      "epoch": 2.9864591727545347,
+      "grad_norm": 1.7061728239059448,
+      "learning_rate": 2.590603260038682e-09,
+      "loss": 0.9431068420410156,
+      "step": 28260
+    },
+    {
+      "epoch": 2.987516017807839,
+      "grad_norm": 1.4992685317993164,
+      "learning_rate": 2.2073835653252383e-09,
+      "loss": 0.8583915710449219,
+      "step": 28270
+    },
+    {
+      "epoch": 2.988572862861144,
+      "grad_norm": 1.2664291858673096,
+      "learning_rate": 1.8548197167150838e-09,
+      "loss": 0.8603200912475586,
+      "step": 28280
+    },
+    {
+      "epoch": 2.9896297079144483,
+      "grad_norm": 1.313591718673706,
+      "learning_rate": 1.532912146573473e-09,
+      "loss": 0.8772127151489257,
+      "step": 28290
+    },
+    {
+      "epoch": 2.990686552967753,
+      "grad_norm": 1.2386982440948486,
+      "learning_rate": 1.2416612496707335e-09,
+      "loss": 0.8571273803710937,
+      "step": 28300
+    },
+    {
+      "epoch": 2.9917433980210575,
+      "grad_norm": 1.1998481750488281,
+      "learning_rate": 9.810673831794903e-10,
+      "loss": 0.8896389961242676,
+      "step": 28310
+    },
+    {
+      "epoch": 2.9928002430743623,
+      "grad_norm": 1.4557340145111084,
+      "learning_rate": 7.511308666774408e-10,
+      "loss": 0.9232834815979004,
+      "step": 28320
+    },
+    {
+      "epoch": 2.993857088127667,
+      "grad_norm": 1.3959667682647705,
+      "learning_rate": 5.5185198214458e-10,
+      "loss": 0.9544175148010254,
+      "step": 28330
+    },
+    {
+      "epoch": 2.9949139331809715,
+      "grad_norm": 1.6930980682373047,
+      "learning_rate": 3.8323097396320006e-10,
+      "loss": 0.9191384315490723,
+      "step": 28340
+    },
+    {
+      "epoch": 2.995970778234276,
+      "grad_norm": 1.811262607574463,
+      "learning_rate": 2.4526804892344157e-10,
+      "loss": 0.8569545745849609,
+      "step": 28350
+    },
+    {
+      "epoch": 2.9970276232875808,
+      "grad_norm": 1.187766194343567,
+      "learning_rate": 1.379633762149668e-10,
+      "loss": 1.032188892364502,
+      "step": 28360
+    },
+    {
+      "epoch": 2.9980844683408856,
+      "grad_norm": 1.189721941947937,
+      "learning_rate": 6.131708742695974e-11,
+      "loss": 0.8739087104797363,
+      "step": 28370
+    },
+    {
+      "epoch": 2.99914131339419,
+      "grad_norm": 1.2776782512664795,
+      "learning_rate": 1.5329276556452776e-11,
+      "loss": 0.8936100006103516,
+      "step": 28380
     }
   ],
   "logging_steps": 10,
@@ -20070,12 +20336,12 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 1.5113475231050957e+18,
+  "total_flos": 1.5323224005185126e+18,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null