Training in progress, step 27000

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:db604a690ace9df3e57323e7728a62d95ca6fe945a7a4f0390b60803c2b0d0f0
+oid sha256:738e63d241a0a1440d6120083eea9c47e57d65e6d8cb770ded25c0d8154e9153
 size 500897101

{checkpoint-23000 → checkpoint-26000/adapter_model}/adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b09fbcc84f52583b6b795b011e1c5f0c43a3db60e66b9f20591fcc7e4f4c5a5c
+oid sha256:db604a690ace9df3e57323e7728a62d95ca6fe945a7a4f0390b60803c2b0d0f0
 size 500897101

{checkpoint-23000/adapter_model → checkpoint-27000}/adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b09fbcc84f52583b6b795b011e1c5f0c43a3db60e66b9f20591fcc7e4f4c5a5c
+oid sha256:738e63d241a0a1440d6120083eea9c47e57d65e6d8cb770ded25c0d8154e9153
 size 500897101

{checkpoint-23000 → checkpoint-27000}/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cbfed6747bc76cf76b9beff591f4900ea3c29204b46ce6523ace5f8336472f0d
+oid sha256:3b9de76af61f1c178dbc3125a612d769b44995d2877a45c685215590ba3eeb50
 size 1001723453

{checkpoint-23000 → checkpoint-27000}/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ed2325c64a2e205df19d31c415cd90ef5d3f44b07c5241eed503aa4caf9c6bce
+oid sha256:1591f2d22b9863cfad258ea21db17a09b936886f6780542a64964d0cc000ccc5
 size 14575

{checkpoint-23000 → checkpoint-27000}/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:43a0198723866985e385225cfff961adcdcef3bedd49c15e274c7317774fe0f2
+oid sha256:490acdbaea2bee4323aee50f8dae99a50ef88531b049adbcba4f79582a4c910c
 size 627

{checkpoint-23000 → checkpoint-27000}/trainer_state.json

@@ -1,8 +1,8 @@
 {
-  "best_metric": 0.6180054545402527,
-  "best_model_checkpoint": "./qlora-out/checkpoint-23000",
-  "epoch": 0.8575370045859587,
-  "global_step": 23000,
+  "best_metric": 0.5985096096992493,
+  "best_model_checkpoint": "./qlora-out/checkpoint-27000",
+  "epoch": 1.0066738749487343,
+  "global_step": 27000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -1570,11 +1570,283 @@
       "eval_samples_per_second": 0.424,
       "eval_steps_per_second": 0.424,
       "step": 23000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001620327094880844,
+      "loss": 0.5795,
+      "step": 23100
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00016172596075378893,
+      "loss": 0.6025,
+      "step": 23200
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00016141827081809075,
+      "loss": 0.5669,
+      "step": 23300
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00016110964437266568,
+      "loss": 0.6172,
+      "step": 23400
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00016080008612346955,
+      "loss": 0.5899,
+      "step": 23500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00016048960079066636,
+      "loss": 0.5889,
+      "step": 23600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00016017819310855632,
+      "loss": 0.5893,
+      "step": 23700
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00015986586782550376,
+      "loss": 0.6363,
+      "step": 23800
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00015955262970386458,
+      "loss": 0.5876,
+      "step": 23900
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00015923848351991372,
+      "loss": 0.5964,
+      "step": 24000
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 0.6122664213180542,
+      "eval_runtime": 1255.2341,
+      "eval_samples_per_second": 0.432,
+      "eval_steps_per_second": 0.432,
+      "step": 24000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00015892343406377225,
+      "loss": 0.5943,
+      "step": 24100
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00015860748613933455,
+      "loss": 0.6008,
+      "step": 24200
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00015829064456419477,
+      "loss": 0.6123,
+      "step": 24300
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00015797291416957355,
+      "loss": 0.5819,
+      "step": 24400
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00015765429980024425,
+      "loss": 0.5731,
+      "step": 24500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00015733480631445926,
+      "loss": 0.593,
+      "step": 24600
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00015701443858387562,
+      "loss": 0.5764,
+      "step": 24700
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00015669320149348104,
+      "loss": 0.6037,
+      "step": 24800
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001563710999415193,
+      "loss": 0.5958,
+      "step": 24900
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00015604813883941535,
+      "loss": 0.6186,
+      "step": 25000
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 0.6086174249649048,
+      "eval_runtime": 1260.3923,
+      "eval_samples_per_second": 0.43,
+      "eval_steps_per_second": 0.43,
+      "step": 25000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00015572432311170096,
+      "loss": 0.597,
+      "step": 25100
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00015539965769593894,
+      "loss": 0.5657,
+      "step": 25200
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001550741475426484,
+      "loss": 0.6081,
+      "step": 25300
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00015474779761522894,
+      "loss": 0.5957,
+      "step": 25400
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00015442061288988525,
+      "loss": 0.6032,
+      "step": 25500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00015409259835555089,
+      "loss": 0.5662,
+      "step": 25600
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00015376375901381256,
+      "loss": 0.5607,
+      "step": 25700
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00015343409987883354,
+      "loss": 0.5727,
+      "step": 25800
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00015310362597727747,
+      "loss": 0.5762,
+      "step": 25900
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00015277234234823154,
+      "loss": 0.5841,
+      "step": 26000
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 0.6026987433433533,
+      "eval_runtime": 1292.1515,
+      "eval_samples_per_second": 0.419,
+      "eval_steps_per_second": 0.419,
+      "step": 26000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00015244025404312974,
+      "loss": 0.6015,
+      "step": 26100
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015210736612567588,
+      "loss": 0.5914,
+      "step": 26200
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015177368367176616,
+      "loss": 0.5799,
+      "step": 26300
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015143921176941205,
+      "loss": 0.6037,
+      "step": 26400
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00015110395551866255,
+      "loss": 0.5876,
+      "step": 26500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001507679200315264,
+      "loss": 0.5973,
+      "step": 26600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00015043111043189423,
+      "loss": 0.5957,
+      "step": 26700
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00015009353185546046,
+      "loss": 0.5696,
+      "step": 26800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00014975518944964478,
+      "loss": 0.5523,
+      "step": 26900
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0001494160883735139,
+      "loss": 0.5144,
+      "step": 27000
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 0.5985096096992493,
+      "eval_runtime": 1314.8131,
+      "eval_samples_per_second": 0.412,
+      "eval_steps_per_second": 0.412,
+      "step": 27000
     }
   ],
   "max_steps": 80463,
   "num_train_epochs": 3,
-  "total_flos": 6.446449236379976e+18,
+  "total_flos": 7.57131896695849e+18,
   "trial_name": null,
   "trial_params": null
 }