sync 2026-04-24T07:32:24+00:00

sft/checkpoint-200/adapter_config.json

@@ -30,8 +30,8 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
     "k_proj",
+    "q_proj",
     "v_proj",
     "o_proj"
   ],

sft/checkpoint-200/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3969884eca31fa76f9aef7175b4cebfdc6d5b239094e713ec1d2b4d6ec6105f6
+oid sha256:23a9dd8fabae6de649147b472c2f3b3c0299587b7d29dc4690bd2203887ee683
 size 61380432

sft/checkpoint-200/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:88be034ce74fb909976c5f0af1eaca344b672bdf5766af980e753ff07ab28123
+oid sha256:fb695f7a88aa180a7dbc0f57a0f1e1e8bbedcef2aeb5c7875dc3038aa2f699af
 size 122930379

sft/checkpoint-200/trainer_state.json

@@ -11,142 +11,142 @@
   "log_history": [
     {
       "epoch": 0.04094165813715456,
-      "grad_norm": 6.233852386474609,
+      "grad_norm": 7.529000759124756,
       "learning_rate": 9.000000000000001e-07,
-      "loss": 6.206051254272461,
+      "loss": 6.639524841308594,
       "step": 10
     },
     {
       "epoch": 0.08188331627430911,
-      "grad_norm": 5.675178527832031,
+      "grad_norm": 7.1427388191223145,
       "learning_rate": 1.9000000000000002e-06,
-      "loss": 6.079831314086914,
+      "loss": 6.582770538330078,
       "step": 20
     },
     {
       "epoch": 0.12282497441146366,
-      "grad_norm": 5.132958889007568,
+      "grad_norm": 6.251121997833252,
       "learning_rate": 2.9e-06,
-      "loss": 5.749290466308594,
+      "loss": 6.114261245727539,
       "step": 30
     },
     {
       "epoch": 0.16376663254861823,
-      "grad_norm": 5.81948184967041,
+      "grad_norm": 7.014133453369141,
       "learning_rate": 3.900000000000001e-06,
-      "loss": 6.033080673217773,
+      "loss": 6.333962631225586,
       "step": 40
     },
     {
       "epoch": 0.2047082906857728,
-      "grad_norm": 6.442906379699707,
+      "grad_norm": 7.072969913482666,
       "learning_rate": 4.9000000000000005e-06,
-      "loss": 5.972381973266602,
+      "loss": 6.248543930053711,
       "step": 50
     },
     {
       "epoch": 0.24564994882292732,
-      "grad_norm": 5.304522514343262,
+      "grad_norm": 5.674424171447754,
       "learning_rate": 5.9e-06,
-      "loss": 5.472136306762695,
+      "loss": 5.6220745086669925,
       "step": 60
     },
     {
       "epoch": 0.2865916069600819,
-      "grad_norm": 5.261083126068115,
+      "grad_norm": 5.3739800453186035,
       "learning_rate": 6.9e-06,
-      "loss": 5.041688537597656,
+      "loss": 5.0251930236816404,
       "step": 70
     },
     {
       "epoch": 0.32753326509723646,
-      "grad_norm": 6.387202739715576,
+      "grad_norm": 6.372293472290039,
       "learning_rate": 7.9e-06,
-      "loss": 4.916830062866211,
+      "loss": 5.033017349243164,
       "step": 80
     },
     {
       "epoch": 0.368474923234391,
-      "grad_norm": 7.068334102630615,
+      "grad_norm": 7.22106409072876,
       "learning_rate": 8.900000000000001e-06,
-      "loss": 3.531498336791992,
+      "loss": 3.514456939697266,
       "step": 90
     },
     {
       "epoch": 0.4094165813715456,
-      "grad_norm": 4.787615776062012,
+      "grad_norm": 4.990958213806152,
       "learning_rate": 9.9e-06,
-      "loss": 2.2226877212524414,
+      "loss": 2.256962776184082,
       "step": 100
     },
     {
       "epoch": 0.4503582395087001,
-      "grad_norm": 3.4490268230438232,
+      "grad_norm": 3.2661993503570557,
       "learning_rate": 9.76923076923077e-06,
-      "loss": 1.955281639099121,
+      "loss": 2.0482555389404298,
       "step": 110
     },
     {
       "epoch": 0.49129989764585463,
-      "grad_norm": 2.435624361038208,
+      "grad_norm": 2.997323513031006,
       "learning_rate": 9.512820512820514e-06,
-      "loss": 1.5814408302307128,
+      "loss": 1.6077747344970703,
       "step": 120
     },
     {
       "epoch": 0.5322415557830092,
-      "grad_norm": 2.0870230197906494,
+      "grad_norm": 2.422741413116455,
       "learning_rate": 9.256410256410257e-06,
-      "loss": 1.4773900985717774,
+      "loss": 1.538028621673584,
       "step": 130
     },
     {
       "epoch": 0.5731832139201638,
-      "grad_norm": 2.1049163341522217,
+      "grad_norm": 2.4594290256500244,
       "learning_rate": 9e-06,
-      "loss": 1.508018970489502,
+      "loss": 1.5614049911499024,
       "step": 140
     },
     {
       "epoch": 0.6141248720573184,
-      "grad_norm": 2.7535016536712646,
+      "grad_norm": 2.7256577014923096,
       "learning_rate": 8.743589743589743e-06,
-      "loss": 1.6693695068359375,
+      "loss": 1.7180768966674804,
       "step": 150
     },
     {
       "epoch": 0.6550665301944729,
-      "grad_norm": 2.2710325717926025,
+      "grad_norm": 2.6614902019500732,
       "learning_rate": 8.487179487179488e-06,
-      "loss": 1.590962028503418,
+      "loss": 1.6412044525146485,
       "step": 160
     },
     {
       "epoch": 0.6960081883316275,
-      "grad_norm": 2.007903575897217,
+      "grad_norm": 2.1517934799194336,
       "learning_rate": 8.230769230769232e-06,
-      "loss": 1.6514751434326171,
+      "loss": 1.729467010498047,
       "step": 170
     },
     {
       "epoch": 0.736949846468782,
-      "grad_norm": 2.2933390140533447,
+      "grad_norm": 2.4588229656219482,
       "learning_rate": 7.974358974358975e-06,
-      "loss": 1.2684261322021484,
+      "loss": 1.305363941192627,
       "step": 180
     },
     {
       "epoch": 0.7778915046059366,
-      "grad_norm": 2.4877562522888184,
+      "grad_norm": 2.5282061100006104,
       "learning_rate": 7.717948717948718e-06,
-      "loss": 1.4245257377624512,
+      "loss": 1.4986873626708985,
       "step": 190
     },
     {
       "epoch": 0.8188331627430911,
-      "grad_norm": 2.4900834560394287,
+      "grad_norm": 3.699396848678589,
       "learning_rate": 7.461538461538462e-06,
-      "loss": 1.4417343139648438,
+      "loss": 1.5059691429138184,
       "step": 200
     }
   ],